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/W3101151231	https://hal.archives-ouvertes.fr/hal-02915648/document	English	null	Inner crust of a neutron star at the point of crystallization in a multicomponent approach	Astronomy & astrophysics	2,020	cc-by	10,653	Inner crust of a neutron star at the point of crystallization in a multicomponent approach T. Carreau, A. F. Fantina, F. Gulminelli To cite this version: T. Carreau, A. F. Fantina, F. Gulminelli. Inner crust of a neutron star at the point of crystallization in a multicomponent approach. Astronomy and Astrophysics - A&A, 2020, 640, pp.A77. 10.1051/0004- 6361/202038347. hal-02915648 To cite this version: T. Carreau, A. F. Fantina, F. Gulminelli. Inner crust of a neutron star at the point of crystallization in a multicomponent approach. Astronomy and Astrophysics - A&A, 2020, 640, pp.A77. 10.1051/0004- 6361/202038347. hal-02915648 1. Introduction e.g., Goriely et al. 2011). Therefore, a more realistic picture of the crust would be that of a multicomponent solid. It is generally assumed that the composition of an iso- lated neutron star (NS) is of “cold catalyzed matter”, mean- ing that determined in the ground state at zero temperature (see, e.g., Haensel et al. 2007; Chamel & Haensel 2008; Blaschke & Chamel 2018). In this hypothesis, the crust of an NS is supposed to be made of pure layers, each consisting of a one-component Coulomb crystal. However NSs, being born from core-collapse supernova explosions, are initially hot, with temperatures exceeding 1010 K. At such temperatures, the crust of a (proto-)NS is expected to be made of a Coulomb liquid com- posed of diﬀerent nuclear species in a charge-compensating elec- tron background: see Oertel & Hempel (2017) for a review. As the NS crust cools down, it is generally supposed that this multi- component plasma (MCP) remains in full thermodynamic equi- librium until the ground state is reached. However, it is unlikely for full equilibrium to be maintained, after crystallization occurs, until T = 0 K. Moreover, if the NS cools down rapidly enough, the composition of the crust could be frozen at ﬁnite tem- perature, Tf, above the crystallization temperature, Tm (see, p For the outer crust, the coexistence of diﬀerent nuclear species is not expected to signiﬁcantly impact the static prop- erties of the crust. Indeed, because of the relatively low crystal- lization temperatures, Tm ≲109 K, the most probable nucleus is very close to the ground-state one-component plasma (OCP) composition, and the contribution of other ions is typically very small. In the inner crust, the situation is a priori less obvi- ous, and deviations from the ground-state composition may be larger, due to the higher crystallization temperature, 109 ≲ Tm ≲1010 K (Haensel et al. 2007). Jones (1999, 2001) indeed suggested that thermal ﬂuctuation of the charge and neutron numbers may be quite signiﬁcant for mass densities of ρB ≳ 1013 g cm−3 near the crystallization temperature. The presence of amorphous and heterogeneous phases in the inner crust leads to a higher temperature-independent electrical resistivity and strong ohmic dissipation, and signiﬁcant consequences on the magnetic ﬁeld evolution were predicted by Jones (2004). More recently, Pons (2013) suggested that the increased resistivity due to the amorphous structure could reﬂect into observational timing properties of X-ray pulsars. ⋆The tables of the impurity parameter shown in Fig. 6 are only available at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsarc.u-strasbg.fr/viz-bin/ cat/J/A+A/640/A77 Open Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. p p 2 Grand Accélérateur National d’Ions Lourds (GANIL), CEA/DRF – CNRS/IN2P3, Boulevard Henri Becquerel, 14076 Caen, France p p 2 Grand Accélérateur National d’Ions Lourds (GANIL), CEA/DRF – CNRS/IN2P3, Boulevard Henri Becquerel Received 5 May 2020 / Accepted 3 June 2020 ABSTRACT This reﬂects on the behavior of the impurity parameter that monotonically increases with density reaching up to around 40 in the deeper regions of the inner crust. Conclusions. Our study shows that the contribution of impurities is non-negligible, thus potentially having an impact on the transport properties in the neutron-star crust. The obtained values of the impurity parameter represent a lower limit; larger values are expected in the presence of nonspherical geometries and/or fast cooling dynamics. Conclusions. Our study shows that the contribution of impurities is non-negligible, thus potentially having an impact on the transport properties in the neutron-star crust. The obtained values of the impurity parameter represent a lower limit; larger values are expected in the presence of nonspherical geometries and/or fast cooling dynamics. Key words. stars: neutron – dense matter – plasmas HAL Id: hal-02915648 https://hal.science/hal-02915648v1 Submitted on 14 Aug 2020 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. Astronomy & Astrophysics Astronomy & Astrophysics A&A 640, A77 (2020) https://doi.org/10.1051/0004-6361/202038347 c⃝T. Carreau et al. 2020 A&A 640, A77 (2020) https://doi.org/10.1051/0004-6361/202038347 c⃝T. Carreau et al. 2020 ABSTRACT Context. The possible presence of amorphous and heterogeneous phases in the inner crust of a neutron star is expected to reduce the electrical conductivity of the crust, potentially with signiﬁcant consequences on the magneto-thermal evolution of the star. In cooling simulations, the disorder is quantiﬁed by an impurity parameter, which is often taken as a free parameter. Context. The possible presence of amorphous and heterogeneous phases in the inner crust of a neutron star is expected to reduce the electrical conductivity of the crust, potentially with signiﬁcant consequences on the magneto-thermal evolution of the star. In cooling simulations, the disorder is quantiﬁed by an impurity parameter, which is often taken as a free parameter. Aims. We aim to give a quantitative prediction of the impurity parameter as a function of the density in the crust, performing micro- scopic calculations including up-to-date microphysics of the crust. , q y p y p , p Aims. We aim to give a quantitative prediction of the impurity parameter as a function of the density in the crust, performing micro- scopic calculations including up-to-date microphysics of the crust. Methods. A multicomponent approach was developed at a ﬁnite temperature using a compressible liquid-drop description of the ions with an improved energy functional based on recent microscopic nuclear models and optimized on extended Thomas-Fermi calcu- lations. Thermodynamic consistency was ensured by adding a rearrangement term, and deviations from the linear mixing rule were included in the liquid phase. Results. The impurity parameter is consistently calculated at the crystallization temperature as determined in the one-component plasma approximation for the diﬀerent functionals. Our calculations show that at the crystallization temperature, the composition of the inner crust is dominated by nuclei with charge number around Z ≈40, while the range of the Z distribution varies from about 20 near the neutron drip to about 40 closer to the crust-core transition. This reﬂects on the behavior of the impurity parameter that monotonically increases with density reaching up to around 40 in the deeper regions of the inner crust. the inner crust is dominated by nuclei with charge number around Z ≈40, while the range of the Z distribution varies from about 20 near the neutron drip to about 40 closer to the crust-core transition. Inner crust of a neutron star at the point of crystallization in a multicomponent approach⋆ T C 1 A F F ti 2 d F G l i lli1 T. Carreau1, A. F. Fantina2, and F. Gulminelli1 1 LPC (CNRS/ENSICAEN/Université de Caen Normandie), UMR6534, 14050 Caen Cédex, France e-mail: carreau@lpccaen.in2p3.fr p p ateur National d’Ions Lourds (GANIL), CEA/DRF – CNRS/IN2P3, Boulevard Henri Becquerel, 14076 Caen, France 1. Introduction More generally, the pres- ence of impurities in the crust has notable eﬀects on trans- port and magneto-rotational properties of the NS (see, e.g., Schmitt & Shternin 2018; Gourgouliatos & Esposito 2018 for A77, page 1 of 9 A&A 640, A77 (2020) volume V(j), such that p j is the frequency of occurrence or prob- ability of the component (j), with P j pj = 1. Thermodynamic quantities are deﬁned in terms of the ion densities of the diﬀer- ent species n( j) N , which are related to the probabilities p j through n( j) N = p j/⟨V⟩, where the bracket notation ⟨⟩indicates ensemble averages. recent reviews), which in turn aﬀect the NS thermal evolution. For these reasons, although cooling simulations are usually car- ried out using the ground-state composition, the presence of various nuclear species is taken into account via an “impurity factor”, often taken as a free parameter adjusted on observational cooling data: see, for instance, Viganò (2013). g g A microscopic calculation of the impurity parameter at the crystallization temperature for the outer crust of a non-accreting unmagnetized NS was recently performed in Fantina et al. (2020). In the latter work, the nuclear distributions of the multicomponent liquid plasma at the crystallization point was computed fully self-consistently, adapting a general formal- ism originally developed for the description of supernova cores (Gulminelli & Raduta 2015; Grams et al. 2018). The crystalliza- tion temperature was determined in the OCP approximation, using a microscopic nuclear mass model based on deformed Hartree–Fock–Bogoliubov calculations (HFB-24, Goriely et al. 2013). The study of Fantina et al. (2020), performed on the outer crust, has subsequently been extended in Carreau et al. (2020), who calculated the crystallization temperature and the associated composition in the inner crust using the compress- ible liquid-drop (CLD) model approach of Carreau et al. (2019), with parameters optimized on four diﬀerent microscopic mod- els: namely BSk22, BSk24, BSk25, and BSk26 (developed by Goriely et al. 2013). Shell eﬀects, as calculated in Pearson et al. (2019) for the same functionals, were added to the CLD model. The use of such an approach instead of a fully microscopic one not only reduces the computational time, but more importantly allows us to quantitatively estimate the model dependence of the results. The outcomes of Carreau et al. 1. Introduction (2020) suggest that, while shell eﬀects are important at the lowest densities close to the outer crust, the highest source of uncertainties comes from the smooth part of the nuclear functional, speciﬁcally the surface tension at extreme isospin values. The diﬀerent (A( j), Z( j)) conﬁgurations are associated with diﬀerent baryonic densities n(j) B , such that the total baryonic den- sity is nB = P j pjn( j) B (see Eq. (14)). Conversely, they share the same total pressure P imposed by the hydrostatic equilibrium and the same background densities of electrons, n(j) e = ne, and of free neutrons, n(j) g = ng. We also suppose that charge neutrality is realized in each cell, meaning that the proton density is the same in each cell (i.e. n(j) p = np) and equal to the electron density ne (i.e. ne = np = Z(j)/V( j)). p The free energy density of the multicomponent system is deﬁned as: F = X j n( j) N F(j) , (1) F = X j n( j) N F(j) , (1) where the free energy per ion of the component ( j) accounts for the contribution of the ion, the dripped neutrons, and the elec- trons: F(j) = F(j) i + F(j) n + F(j) e , (2) (2) including their mutual interactions1. For future convenience, the nuclear interactions between the ion and the neutron gas, and the Coulomb interactions between the ion and the electrons, are all included in the term F(j) i . Therefore, the free neutron and elec- tron components are simply given by: including their mutual interactions1. For future convenience, the nuclear interactions between the ion and the neutron gas, and the Coulomb interactions between the ion and the electrons, are all included in the term F(j) i . Therefore, the free neutron and elec- tron components are simply given by: F(j) n = V(j)Fg ; F(j) e = V(j)Fe , (3) (3) In the present work, we employed the same CLD model with parameters optimized on the same functionals as in Carreau et al. (2020), but we extended it by including a nuclear distribution in an MCP approach at equilibrium similar to that of Fantina et al. (2020). This also allows us to calculate the impurity parameter in the inner crust self-consistently, thus complementing the results obtained in Fantina et al. (2020) for the outer crust. 1. Introduction where Fg(e) is the free energy density of a uniform neutron (elec- tron) gas at density ng (ne). The explicit expression of these terms is discussed in Sect. 2.4. The ion contribution, F( j) i , can be writ- ten as: where Fg(e) is the free energy density of a uniform neutron (elec- tron) gas at density ng (ne). The explicit expression of these terms is discussed in Sect. 2.4. The ion contribution, F( j) i , can be writ- ten as: F(j) i = F( j),0 i + δF(j) . (4) (4) The formalism is described in Sect. 2. The numerical results are presented in Sect. 3; speciﬁcally, the composition of the inner crust is discussed in Sect. 3.1, and the impurity parameter in Sect. 3.2. Finally, we conclude in Sect. 4. The ﬁrst term in Eq. (4), F( j),0 i , noting mn (mp) the neutron (pro- ton) mass, is given by: The ﬁrst term in Eq. (4), F( j),0 i , noting mn (mp) the neutron (pro- ton) mass, is given by: F(j),0 i = (A( j) −Z(j))mnc2 + Z(j)mpc2 + F(j),nuc i +F(j),id i + F( j),int i , (5) (5) 2. Model of the inner crust where F(j),nuc i is the internal nuclear free energy and F(j),int i is the Coulomb interaction contribution. The explicit expressions of these terms, as well as of the last term in Eq. (4), δF( j), account- ing for the interaction between the ion and the surrounding (neu- tron) gas, depend on the adopted model and are discussed in Sects. 2.3 and 2.4. Finally, since in this work we are only inter- ested in temperatures higher or equal to the melting temperature, where the MCP is expected to be in the liquid phase, the “ideal” contribution, F(j),id i , accounts for the translational center-of-mass motion: where F(j),nuc i is the internal nuclear free energy and F(j),int i is the Coulomb interaction contribution. The explicit expressions of these terms, as well as of the last term in Eq. (4), δF( j), account- ing for the interaction between the ion and the surrounding (neu- tron) gas, depend on the adopted model and are discussed in Sects. 2.3 and 2.4. Finally, since in this work we are only inter- ested in temperatures higher or equal to the melting temperature, where the MCP is expected to be in the liquid phase, the “ideal” contribution, F(j),id i , accounts for the translational center-of-mass motion: 1 We denote with capital letters the (free) energy per ion, for example, F, while the notation F is used for the free energy density. 2.1. MCP in nuclear statistical equilibrium (8) (8) The probabilities p j and the densities n(j) N are calculated so as to maximize the thermodynamic potential in the canonical ensem- ble. Because of the chosen free energy decomposition, we can observe that the electron and free neutron part of the free energy density, Fe and Fg, do not depend on n(j) N , for instance, p j = N exp − ˜Ω(j) i kBT , (18) (18) with the normalization ) N = exp α kBT ! = X j exp − ˜Ω( j) i kBT . (19) F n n(j) N o = Fi n n(j) N o + Fe + Fg , (9 where F X ( j)F(j) (10 F n n(j) N o = Fi n n(j) N o + Fe + Fg , (9) (9) (19) where where The single-ion grand-canonical potential ˜Ω(j) N reads: (10) The single-ion grand-canonical potential ˜Ω(j) N reads: Fi = X j n( j) N F(j) i . (10 Fi = X j n( j) N F(j) i . (10) ˜Ω( j) i = Ω(j) i −µnN( j) −µpZ( j) , (20) (20) Therefore, the variation can be performed on the ion part only: Therefore, the variation can be performed on the ion part only: where µn and µp can be identiﬁed with the neutron and proton chemical potentials, respectively. In the deﬁnitions above, the ion free energy contains the rest-mass energy, thus the chemical potentials include the rest-mass energies as well. dFi = X j Ω( j) i + kBT ln n(j) N dn(j) N , (11) (11) where the single-ion canonical potential is given by: The calculation of the grand-canonical potential, ˜Ω(j) i , requires the evaluation of the chemical potentials µn, µp, as well as of the rearrangement term (last term in Eq. (12)): Ω(j) i = F( j) i −F(j),id i + kBT ln λ(j)3 g( j) s +n( j) N ∂ F( j) i −F(j),id i ∂n(j) N · (12) R( j) = n(j) N ∂ F(j) i −F( j),id i ∂n(j) N · (21) (21) (12) These terms are worked out in Sects. 2.2 and 2.5, respectively. Once the abundances of the diﬀerent ions are calculated via Eq. 2.1. MCP in nuclear statistical equilibrium To model a full statistical equilibrium of ions in the inner crust, we extended the formalism of Fantina et al. (2020), allowing for the presence of dripped neutrons, which are supposed to con- stitute a homogeneous gas. The possible contribution of a free proton gas is expected to be small at the temperatures we con- sidered, and thus it was neglected. This working hypothesis is a-posteriori conﬁrmed by the calculation of the proton fugacity, zp = exp[(µp −mpc2)/(kBT)], µp (mp) being the proton chemi- cal potential (mass), c the speed of light, and kB the Boltzmann constant, which never exceeds −20 MeV in the density and tem- perature domain studied in this paper. F(j),id i = kBT ln  n( j) N (λ(j))3 g(j) s −1 , (6) (6) The NS crust at a given depth in the star is supposed to contain diﬀerent ion species with mass and charge num- bers (A(j), Z(j)) associated with diﬀerent Wigner-Seitz cells of 1 We denote with capital letters the (free) energy per ion, for example, F, while the notation F is used for the free energy density. A77, page 2 of 9 T. Carreau et al.: Inner crust of a neutron star in a MCP T. Carreau et al.: Inner crust of a neutron star in a MCP where g(j) s is the spin degeneracy. For this, we take g(j) s = 1 for nuclei whose ground-state angular momentum is unknown, and the de Broglie wavelength of component ( j) is given by: The constraints Eqs. (13)–(15) are taken into account by introducing Lagrange multipliers (α, µn, µp) leading to the fol- lowing equations for the equilibrium densities n(j) N : λ(j) = s 2π(ℏc)2 M⋆( j)c2kBT , (7) X j Ω( j) i + kBT ln n(j) N −α dn(j) N −µn X j N(j)dn(j) N −µp X j Z( j)dn( j) N = 0 , (17) λ(j) = s 2π(ℏc)2 M⋆( j)c2kBT , (7) (17) ℏbeing the Planck-Dirac constant, and the eﬀective mass of the ion M∗( j) is deﬁned as ℏbeing the Planck-Dirac constant, and the eﬀective mass of the ion M∗( j) is deﬁned as with N(j) = A(j) 1 −ng/n(j) 0 −Z( j). Considering independent variations, the equilibrium distributions are given by M⋆( j)c2 = (A(j) −Z(j))mnc2 + Z(j)mpc2 + F(j),nuc i + δF(j) . 2.1. MCP in nuclear statistical equilibrium (18) at the crystallization temperature, it is also possible to calculate the impurity parameter of the solid crust, which repre- sents the variance of the ionic charge distributions and is deﬁned as (see, e.g., the discussion in Sect. 7 in Meisel et al. 2018 for a review) In Eq. (11), the variations dn(j) N are not independent, because of the normalization of probabilities and the baryonic number and charge conservation laws: 1 ⟨V⟩= X j n(j) N , (13) nB −ng = X j n(j) N A(j) 1 −ng n(j) 0 , (14) np = X j n(j) N Z(j) . (15) g as (see, e.g., the discussion in Sect. 7 in Meisel et al. 2018 for a review) Qimp = X j p(Z(j))(Z(j) −⟨Z⟩)2 , (22 where p(Z(j)) is the normalized probability distribution (inte grated over all N(j)) of the element Z(j). 1 ⟨V⟩= X j n(j) N , (13) nB −ng = X j n(j) N A(j) 1 −ng n(j) 0 , (14) np = X j n(j) N Z(j) . (15) as (se review Qimp where grated (13) Qimp = X j p(Z(j))(Z(j) −⟨Z⟩)2 , (22) −ng n(j) 0 , (14) Qimp = X j p(Z(j))(Z(j) −⟨Z⟩)2 , (22) (22) 0 Z(j) . (15) where p(Z(j)) is the normalized probability distribution (inte- grated over all N(j)) of the element Z(j). (15) where p(Z(j)) is the normalized probability distribution (inte- grated over all N(j)) of the element Z(j). (15) The correction factor on the right hand side of Eq. (14) accounts for the excluded volume, meaning the gas cannot occupy the nucleus volume. In the same equation, nB is the total baryonic density, and n(j) 0 is the average density of the ion (j). This latter can be calculated by imposing equilibrium with the nucleon gas via: 2.2. Evaluation of the chemical potentials In a given thermodynamic condition expressed by a temperature T and a pressure P, the proton and neutron chemical potentials can be determined using the thermodynamic relation F + P = µnnn + µpnp + µene, giving, together with the beta-equilibrium condition µn = µe+µp (µe being the electron chemical potential), n( j)2 0 A( j) ∂F(j),0 i ∂n( j) 0 = Pg , (16) n( j)2 0 A( j) ∂F(j),0 i ∂n( j) 0 = Pg , (16) µn = F + P nB ; µe = Fe + Pe np , (23) (23) where Pg = n2 gd(Fg/ng)/dng is the pressure of the neutron gas. This expression is explicitly demonstrated in Sect. 2.4 (see Eq. (38)). where Pg = n2 gd(Fg/ng)/dng is the pressure of the neutron gas. This expression is explicitly demonstrated in Sect. 2.4 (see Eq. (38)). where the baryon and proton densities, nB and np, are given by Eqs. (14) and (15), respectively, the free energy density A77, page 3 of 9 A&A 640, A77 (2020) we include the interactions of the nucleus with the neutrons and electrons in the term Fi: F is given by Eq. (1), and Fe (Pe = n2 ed(Fe/ne)/dne) is the free energy density (pressure) of the electron gas at den- sity np = ne. With this prescription, the equilibrium proba- bilities can only be determined by the solution of a complex nonlinear system of coupled equations which is a challenging numerical task. Such a complete nuclear statistical equilib- rium formalism has been adopted by diﬀerent authors (see Oertel & Hempel 2017; Burgio & Fantina 2018 for a review); however, simpliﬁed nuclear functionals were adopted, the den- sity (instead of the pressure) was imposed, and the rearrange- ment term was neglected. Fi = F0 i + δF , (28) Fi = F0 i + δF , (28) with F0 i = (A −Z)mnc2 + Zmpc2 + Fnuc i +Fid i + Fint i . (29) (29) In the OCP approximation, the translational motion is limited to the single Wigner-Seitz cell (nN = 1/V): In the outer crust regime, it was found by Fantina et al. (2020) that a perturbative implementation of the nuclear statis- tical equilibrium as proposed by Grams et al. (2018) leads to a very fast convergence, with reduced computational cost and increased numerical precision. We therefore adopt this same pre- scription in the inner crust. 2.2. Evaluation of the chemical potentials In the perturbative treatment, the equilibrium problem is solved in the OCP approximation, as detailed in Sect. 2.3 below. This gives a ﬁrst guess for the chem- ical potentials as: Fid i = kBT " ln λ3 Vgs ! −1 # , (30) (30) where the de Broglie wavelength λ is given by the same expres- sion as in Eq. (7), with M⋆(j) = M⋆. The interacting part of the ion free energy can be decomposed as: Fint i = Fii,liq + Fpol ie,liq . (31) (31) µOCP n = F OCP + P nOCP B ; µOCP p = µOCP n −Fe + Pe nOCP p , (24) Analytical formulae were derived by Potekhin & Chabrier (2000) for these two terms; see their Eqs. (16) and (19), respec- tively. For this study, only the ﬁrst term is included; indeed, the polarization correction is found to have no eﬀect in the density and temperature regime studied in the present paper and is there- fore neglected. In addition, the nuclear ﬁnite-size correction is also included. The latter is derived from the Gauss theorem and reads: Analytical formulae were derived by Potekhin & Chabrier (2000) for these two terms; see their Eqs. (16) and (19), respec- tively. For this study, only the ﬁrst term is included; indeed, the polarization correction is found to have no eﬀect in the density and temperature regime studied in the present paper and is there- fore neglected. In addition, the nuclear ﬁnite-size correction is also included. The latter is derived from the Gauss theorem and reads: (24) where F OCP is the equilibrium free energy density in the OCP approximation, and nOCP B , nOCP p are the baryon and proton densi- ties that, in the OCP approximation, lead to the pressure P (see Sect. 2.3). Similarly, the electron quantities Fe and Pe are cal- culated at ne = nOCP p . With this guess, the ion abundances are readily calculated via Eq. (18), and again using Eq. (23) we can get an improved estimation of the chemical potentials as Efs = 2np n0(1 −I) e2 r0 Z2 A1/3 , (32) (32) µn = P j n( j) N F(j) nB⟨V⟩ + P nB , (25) ypµe = P j n(j) N F(j) e nB⟨V⟩ + Pe nB , (26) with r0 = (4πn0/3)−1/3 and e as the elementary charge. 2.2. Evaluation of the chemical potentials (25) Finally, the interaction between the ion and the surrounding neutron gas is handled in the excluded volume approximation: (26) δF = −A no Fg . (33) (33) where yp = ⟨Z⟩/(nB⟨V⟩) is the average proton fraction of the mixture, with ⟨Z⟩= P j pjZ(j). The problem can thus be solved by iteration. It turns out that the diﬀerence between the initial guess, Eq. (24), and the result of the ﬁrst iteration, Eqs. (25)– (26), is so small for all pressures and temperatures considered in this work, that the simple OCP estimation, Eq. (24), can be kept. The equilibrium conﬁguration is obtained by minimizing Eq. (27) with respect to the variational variables using the baryon density constraint limited to a single cell, nB = ng + A V 1 −ng n0 ! . (34) (34) p q p The full MCP calculation becomes therefore computation- ally equivalent to the much simpler OCP one, with the additional advantage that the MCP results can be compared to the more standard OCP ones with no extra computational cost. This leads to the following system of coupled diﬀerential equa- tions2: ∂(F0 i /A) ∂A = 0, (35) 2 A  ∂F0 i ∂I − np 1 −I ∂F0 i ∂np = µe, (36) F0 i A + 1 −I A ∂F0 i ∂I = µB −Pg n0 , (37) n0 2 ∂(F0 i /A) ∂n0 = Pg, (38) (35) A77, page 4 of 9 2.5. Evaluation of the rearrangement term The computation of the equilibrium distributions, Eq. (18), associated with a thermodynamic condition characterized by a temperature T, and chemical potentials µn, µp, requires the eval- uation of the rearrangement term entering Eq. (12): 2.3. The OCP approximation 2014, 2018), meaning that the spread of the pre- dictions of those models can be taken as a reasonable estimation of the model dependence of our results. Concerning the nuclear models, we use the same functionals as in Carreau et al. (2020), namely the recent functionals of the BSk family BSk22, BSk24, BSk25, and BSk26 introduced by Goriely et al. (2013). These realistic microscopic models span a relatively large range in the symmetry energy parameters con- sistent with existing experimental constraints, thus covering the most important part of the present equation-of-state uncertainty (Pearson et al. 2014, 2018), meaning that the spread of the pre- dictions of those models can be taken as a reasonable estimation of the model dependence of our results. µB = 2npn0 n0(1 −I) −2np ∂(F0 i /A) ∂ng + dFg dng · (39) (39) In our parameterization (see Sect. 2.4), the in-medium modi- ﬁcation of the nuclear energy arising from the external gas is governed by a single parameter, p, which does not depend on the external neutron density, but only on the isospin asymme- try, I. Therefore, ∂F0 i /∂ng = 0, and the baryon chemical poten- tial can be identiﬁed with the chemical potential of the gas: µB = µg ≡dFg/dng. Finally, the free energy density Fe and pressure Pe of the electron gas are calculated within a relativistic Sommer- feld expansion. The complete expressions can be found in Haensel et al. (2007): see their Eqs. (2.65) and (2.67), respec- tively. Exchange and correlation contributions are found to be very small in the ranges of density and temperature explored in this work and can be safely neglected. At each value of the baryon density nOCP B and temperature T ≥Tm above the crystallization point, the system of coupled diﬀerential equations, Eqs. (35)–(38), is numerically solved as in Carreau et al. (2019). This procedure leads to the determina- tion of the favored liquid composition (A, I, n0, np, ng)\|OCP and to the evaluation of the total free energy and pressure, F OCP and P, as well as of the electron component, Fe(nOCP e ), Pe(nOCP e ). These quantities allow one to compute the chemical potentials of the MCP using Eq. (24). 2.4. The free energy functional The free energy functional for an isolated nucleus in the vacuum is modeled using the CLD model of Carreau et al. (2020), which we brieﬂy outline here. R( j) = n( j) N ∂ F( j) i −F(j),id i ∂n(j) N · (45) (45) The nuclear free energy Fnuc i at temperature T of a nucleus of mass number A, isospin asymmetry I, and average density n0, is decomposed into a bulk, surface, and Coulomb part as: As already discussed in Fantina et al. (2020), the rearrange- ment term arises from the self-consistency induced by the Coulomb part of the ion free energy. This stems from the fact that, due to the strong incompressibility of the electrons, we have imposed charge conservation at the level of each cell: (40) Fnuc i = A fb(n0, I, T) + Fsurf+curv + FCoul , (40) Fnuc i = A fb(n0, I, T) + Fsurf+curv + FCoul , where fb(nB, δ, T) represents the free energy per baryon of bulk nuclear matter, with nB = np +nn, δ = (nn −np)/nB, and np(nn) is the homogeneous proton (neutron) density. Assuming spherical nuclei, we write the Coulomb energy as: ne = np = X j n( j) N Z( j) = Z(j) V( j) · (46) (46) FCoul = 3 5 e2 r0 Z2 A1/3 , (41) (41) This is at variance with the baryonic density, which can ﬂuctuate from cell to cell (see Eq. (14)). As a consequence of that, any component of the free energy density that depends on the local cell proton density n(j) p = np leads to a dependence on the local density n(j) N through Eq. (46). Within the functional described in Sect. 2.4, this is only the case for the Coulomb interaction F( j),int i . The rearrangement term of component ( j) thus reduces to: This is at variance with the baryonic density, which can ﬂuctuate from cell to cell (see Eq. (14)). As a consequence of that, any component of the free energy density that depends on the local cell proton density n(j) p = np leads to a dependence on the local density n(j) N through Eq. (46). Within the functional described in Sect. 2.4, this is only the case for the Coulomb interaction F( j),int i . 2.3. The OCP approximation In the OCP approximation, the equilibrium conﬁguration of inhomogeneous dense matter in the inner crust in full thermo- dynamic equilibrium is obtained by minimizing the free energy density in a Wigner-Seitz cell of volume V with the constraint of a given baryon density, nB. (see Lattimer & Swesty 1991; Gulminelli & Raduta 2015; Carreau et al. 2020). (36) (37) ; ) Similarly to the general MCP case of Sect. 2.1, we write: (38) F (A, I, n0, np, ng) = Fi + Fn + Fe V , (27) (27) 2 These equations are equivalent to Eqs. (8)–(11) in Carreau et al. (2020). The notation Fi in Carreau et al. (2020) is indeed equivalent to the notation F0 i used in the present paper. We note, however, that there is a misprint in Eq. (9) in Carreau et al. (2020) (although the calcula- tions were done correctly); the term ∆mn,pc2 should not appear in their Eq. (9). where the variational variables are the mass number A and isospin ratio I = 1 −2Z/A of the ion, its internal density n0, the proton density in the cell np = ne, and the density of the homo- geneous gas of dripped neutrons ng. As in Sect. 2.1, see Eq. (4), T Carreau et al : Inner crust of a neutron star in a MCP T. Carreau et al.: Inner crust of a neutron star in a MCP where the gas pressure is given by Pg = n2 gd(Fg/ng)/dng = ngµB −Fg, and the baryon chemical potential µB results in: where the gas pressure is given by Pg = n2 gd(Fg/ng)/dng = ngµB −Fg, and the baryon chemical potential µB results in: away with the temperature, and we thus consider that they can be neglected in the temperature range we explore in this work. away with the temperature, and we thus consider that they can be neglected in the temperature range we explore in this work. Concerning the nuclear models, we use the same functionals as in Carreau et al. (2020), namely the recent functionals of the BSk family BSk22, BSk24, BSk25, and BSk26 introduced by Goriely et al. (2013). These realistic microscopic models span a relatively large range in the symmetry energy parameters con- sistent with existing experimental constraints, thus covering the most important part of the present equation-of-state uncertainty (Pearson et al. 2.4. The free energy functional The rearrangement term of component ( j) thus reduces to: with the surface and curvature free energies as in Newton et al. (2013), Lattimer & Swesty (1991): Fsurf+curv = 4πr2 0σsA2/3 +8πr0σs σ0,c σ0 α β −1 −I 2 ! A1/3, (42) (42) σ0 2 ! with α = 5.5, and an isospin-dependent surface tension given by: σs = σ0 2p+1 + bs (Z/A)−p + bs + (1 −Z/A)−p · (43) R( j) = n(j) N ∂F(j),int i ∂n(j) N {n(i) N }i,j = n(j) N Z(j) ∂F(j),int i ∂np , (47) 0 ! with α = 5.5, and an isospin-dependent surface tension given by: σs = σ0 2p+1 + bs (Z/A)−p + bs + (1 −Z/A)−p · (43) R( j) = n(j) N ∂F(j),int i ∂n(j) N {n(i) N }i,j = n(j) N Z(j) ∂F(j),int i ∂np , (47) with α = 5.5, and an isospin-dependent surface tension given by: R( j) = n(j) N ∂F(j),int i ∂n(j) N {n(i) N }i,j σs = σ0 2p+1 + bs (Z/A)−p + bs + (1 −Z/A)−p · (43) (43) (47) The surface and curvature parameters σ0, bs, p, σ0,c, and β are optimized on extended Thomas-Fermi (ETF) mass tables built with the same nuclear functional adopted for the bulk term, see Carreau et al. (2020) for details. The same functional is also used to compute the free energy density of the neutron gas, The surface and curvature parameters σ0, bs, p, σ0,c, and β are optimized on extended Thomas-Fermi (ETF) mass tables built with the same nuclear functional adopted for the bulk term, see Carreau et al. (2020) for details. The same functional is also used to compute the free energy density of the neutron gas, where we used Eq. (46), implying the relation ∂np/∂n(j) N = Z(j). Following Grams et al. (2018), to avoid the complication of a self-consistent resolution of Eq. (18), we looked for an approxi- mation of Eq. (47) using the requirement that the most probable ion in the MCP mixture should coincide with the OCP result, if nonlinear mixing terms in the MCP are omitted. This condition is a direct consequence of the principle of ensemble equivalence in the thermodynamic limit (see Gulminelli & Raduta 2015). Fg = ng fb(ng, 1, T) + ngmnc2 . (44) (44) In principle, a shell and pairing correction should be added to the nuclear free energy expression, Eq. (40). 3. Numerical results We computed the ﬁnite-temperature composition of the inner crust of non-accreting unmagnetized NSs within our MCP approach, thus including a distribution of nuclei in nuclear sta- tistical equilibrium. For the considered BSk functionals, the recent calculations of Pearson et al. (2020) show that nonspher- ical pasta structures are expected to be present at the highest densities above nB ≈0.05 fm−3, close to the crust-core transi- tion point. Since we only consider spherical nuclei in the present study, we limited our calculation to the density domain extend- ing from the neutron-drip point to nB = 0.04 fm−3. All the results presented in this section were obtained using the BSk CLD models, with the surface and curvature param- eters ﬁtted to the corresponding ETF calculations and crust- core transition densities (see Table 1 of Carreau et al. 2020 for the explicit parameter values). In Sect. 3.1, the results for the inner-crust composition at a ﬁnite temperature are shown for the BSk24 CLD model, as an illustrative example, while the impu- rity parameter is presented in Sect. 3.2 for all four of the con- sidered CLD models based on the BSk22, BSk24, BSk25, and BSk26 functionals. To evaluate the width of the distribution, in Fig. 2 we show the normalized probability distribution p(Z) for T = 1010 K and T = Tm and for two selected densities in the inner crust: nB = 5 × 10−4 fm−3 (panel a) and nB = 10−2 fm−3 (panel b). The peaks of the distributions, meaning the most probable Z, coincide with the charge numbers predicted in the OCP approximation (shown by the associated arrows), thus indicating that the linear mix- ing rule is a good approximation. To assess the importance of the rearrangement term, Eq. (47), we use vertical lines to mark the average values of the charge number, ⟨Z⟩, obtained when this term is not included in the calculations. We observe that the eﬀect of the rearrangement term is signiﬁcant, particularly at a higher density. Without taking this term into account, the distri- bution is systematically and considerably shifted toward a lower Z, proving that the rearrangement term is actually needed to sat- isfy the thermodynamic consistency. 2.4. The free energy functional However, it was shown in Carreau et al. (2020) that these corrections rapidly fade In principle, a shell and pairing correction should be added to the nuclear free energy expression, Eq. (40). However, it was shown in Carreau et al. (2020) that these corrections rapidly fade To look for the extremum of Eq. (18), one has to consider that in the MCP, both np and ng are imposed once the thermodynamic A77, page 5 of 9 A77, page 5 of 9 A77, page 5 of 9 A&A 640, A77 (2020) condition is speciﬁed. Therefore, these densities no longer act as constraints and should not be varied. The variation of Eq. (18) with respect to the ion variables A, I, n0 thus gives: 10−3 10−2 nB [fm−3] 0 200 400 600 800 1000 Acell A Z (a) MCP average MCP most probable OCP solution 10−3 10−2 nB [fm−3] Acell A Z (b) Fig. 1. Variation with baryon density nB of the average (solid lines) and most probable (dashed lines) values of the charge number Z (blue lines), cluster mass number A (orange lines), and total mass number Acell (red lines) in the inner crust at two selected temperatures: T = 1010 K (panel a), and T = Tm (panel b). Results obtained in the one-component plasma (OCP) approximation are also shown (dotted lines). 10−3 10−2 nB [fm−3] 0 200 400 600 800 1000 Acell A Z (a) MCP average MCP most probable OCP solution 10−3 10−2 nB [fm−3] Acell A Z (b) n02 A  ∂F0 i ∂n0 + ∂R ∂n0 = Pg, (48) 2 A  ∂F0 i ∂I + ∂R ∂I = µn −µp, (49) ∂F0 i ∂A + ∂R ∂A + 1 −I A  ∂F0 i ∂I + ∂R ∂I = µn −Pg n0 , (50) (48) (50) where the partial derivatives are calculated at the values corre- sponding to the equilibrium OCP solution, and F0 i is given by Eq. (29) using Eq. (30), meaning by the OCP functional, that is nonlinear mixing terms are excluded. By comparing Eqs. (48)–(50) to the OCP ones, Eqs. (35)– (38), and using Pg = ngµn −Fg, we can deduce that R(j) should not depend on n(j) 0 , that is R(j) = R(j)(A(j), I( j)), and that at the OCP solution, we should have Fig. 1. 2.4. The free energy functional Variation with baryon density nB of the average (solid lines) and most probable (dashed lines) values of the charge number Z (blue lines), cluster mass number A (orange lines), and total mass number Acell (red lines) in the inner crust at two selected temperatures: T = 1010 K (panel a), and T = Tm (panel b). Results obtained in the one-component plasma (OCP) approximation are also shown (dotted lines). 1 −I A ∂R ∂I = −∂R ∂A· (51) (51) This is satisﬁed if R(j) linearly depends on Z( j) = A(j)(1 −I(j))/2. Our ﬁnal expression for the rearrangement term is therefore: This is satisﬁed if R(j) linearly depends on Z( j) = A(j)(1 −I(j))/2. Our ﬁnal expression for the rearrangement term is therefore: Tf = 1010 K as an illustrative example. Indeed, a more realistic estimate of Tf would require dynamical simulations, which are beyond the scope of this paper. Tf = 1010 K as an illustrative example. Indeed, a more realistic estimate of Tf would require dynamical simulations, which are beyond the scope of this paper. R(j) ≃Z(j) *⟨n(j) N ⟩∂F(j),int i ∂np + j = Z(j)  1 V ∂Fint i ∂np  OCP , (52) (52) 3.1. Equilibrium composition of the MCP where the quantity in the parentheses is calculated at the OCP solution. The average and most probable mass and charge number in the MCP are displayed in Fig. 1 as a function of the baryon density in the inner crust for T = 1010 K (panel a) and T = Tm (panel b). For comparison, the results obtained in the OCP approxima- tion are also shown (dotted lines). We can see that the average and most probable values in the MCP approach follow the OCP ones very closely. This means that the deviations from the lin- ear mixing rule in the liquid phase are small, as already noted in Fantina et al. (2020) for the outer crust. While the mass num- bers increase with density, the charge number is almost constant, Z ≈40. The latter value is very close to that obtained at zero temperature (see also, the dotted curve in Fig. 6, panel (b), in Carreau et al. 2020 and Fig. 12 in Pearson et al. 2018), suggest- ing that the presence of Z ≈40 ions in the inner crust is a robust result. 3. Numerical results Normalized probability distribution p(Z) for nB = 5 × 10−4 fm−3 (panel a) and nB = 10−2 fm−3 (panel b) at two selected temperatures: 10 Fig. 2. Normalized probability distribution p(Z) for nB = 5 × 10−4 fm−3 (panel a) and nB = 10−2 fm−3 (panel b) at two selected temperatures: T = Tm (orange squares), and T = 1010 K (blue circles). Arrows indi- cate the OCP solutions. Vertical dashed lines correspond to the value of ⟨Z⟩obtained without considering the rearrangement term (see text for details). 30 Fig. 3. Normalized probability distribution p(Z) with increasing baryon density nB in the inner crust at the crystallization temperature Tm. less reliable. The ﬂattening of the distribution is more clearly vis- ible in Fig. 3, where the normalized probability distribution p(Z) at the crystallization temperature is plotted for diﬀerent increas- ing baryon densities in the inner crust. While the average value of Z is centered around 40 throughout the inner crust, the range of Z of the distribution varies from ≈20 closer to the neutron drip up to ≈40 near the crust-core transition. 75 100 125 30 40 50 Z (a) 50 100 30 40 50 (b) 100 150 200 N 30 40 50 Z (c) 100 200 N 30 40 50 (d) 0.00 0.01 0.02 0.0000 0.0025 0.0050 0.0075 0.000 0.002 0.004 0.006 0.000 0.002 0.004 Fig. 4. Normalized probability distribution of nuclei p(N, Z) for four chosen thermodynamic conditions. Panel a: nB = 5×10−4 fm−3, T = Tm; panel b: nB = 5 × 10−4 fm−3, T = 1010 K; panel c: nB = 10−2 fm−3, T = Tm; panel d: nB = 10−2 fm−3, T = 1010 K. In each panel, the OCP solution coincides with the intersection of the black lines. 50 100 30 40 50 (b) 100 200 N 30 40 50 (d) 00 01 02 0.0000 0.0025 0.0050 0.0075 000 002 004 006 0.000 0.002 0.004 75 100 125 30 40 50 Z (a) 100 150 200 N 30 40 50 Z (c) 50 100 30 40 50 (b) 0 1 2 To better assess the evolution of the nuclear distribution, both in charge and mass number with density and tempera- ture, in Fig. 3. Numerical results 4 we show the normalized probability distribution p(Z, N) for two selected densities in the inner crust: nB = 5 × 10−4 fm−3 (panels (a) and (b)) and nB = 10−2 fm−3 (panels (c) and (d)), both at T = 1010 K (panels b and d) and T = Tm (panels (a) and (c)). As expected, going from lower to higher densities (upper to lower panels), we observe that the ion species become more neutron rich and that the distribution, both in Z and N, broadens when going from lower to higher temperatures (left to right panels). 75 100 125 30 100 150 200 N 30 40 50 Z (c) Fig. 4. Normalized probability distribution of nuclei p(N, Z) for four chosen thermodynamic conditions. Panel a: nB = 5×10−4 fm−3, T = Tm; panel b: nB = 5 × 10−4 fm−3, T = 1010 K; panel c: nB = 10−2 fm−3, T = Tm; panel d: nB = 10−2 fm−3, T = 1010 K. In each panel, the OCP solution coincides with the intersection of the black lines. 3. Numerical results We can also notice that, as expected, the distributions become broader with increasing temperature and density, thus making the OCP approximation Our calculations of the liquid MCP were performed at the crystallization temperature Tm and, for comparison, at 1010 K = T > Tm. The reason of this choice stems from the fact that, depending on the NS cooling timescales, the com- position may be already frozen at some temperature Tf > Tm (see e.g., Goriely et al. 2011). In Carreau et al. (2020), the crys- tallization temperature of the inner crust was estimated to lie between ≈2.5 × 109 K and ≈8 × 109 K for the considered CLD models (see their Figs. 5 and 7, panel (a)). Therefore, we chose A77, page 6 of 9 T. Carreau et al.: Inner crust of a neutron star in a MCP 30 40 50 Z 0.00 0.05 0.10 0.15 0.20 0.25 p(Z) (a) 30 40 50 Z (b) Fig. 2. Normalized probability distribution p(Z) for nB = 5 × 10−4 fm−3 (panel a) and nB = 10−2 fm−3 (panel b) at two selected temperatures: T = Tm (orange squares), and T = 1010 K (blue circles). Arrows indi- cate the OCP solutions. Vertical dashed lines correspond to the value of ⟨Z⟩obtained without considering the rearrangement term (see text for details). 3e−04 1e−03 3e−03 1e−02 3e−02 20 30 40 50 60 Z Baryon density [fm−3] Fig. 3. Normalized probability distribution p(Z) with increasing baryon density nB in the inner crust at the crystallization temperature Tm. 3e−04 1e−03 3e−03 1e−02 3e−02 20 30 40 50 60 Z Baryon density [fm−3] Fig. 3. Normalized probability distribution p(Z) with increasing baryon density nB in the inner crust at the crystallization temperature Tm. 30 40 50 Z 0.00 0.05 0.10 0.15 0.20 0.25 p(Z) (a) 30 40 50 Z (b) Fig. 2. Normalized probability distribution p(Z) for nB = 5 × 10−4 fm−3 (panel a) and nB = 10−2 fm−3 (panel b) at two selected temperatures: T = Tm (orange squares), and T = 1010 K (blue circles). Arrows indi- cate the OCP solutions. Vertical dashed lines correspond to the value of ⟨Z⟩obtained without considering the rearrangement term (see text for details). 3e−04 1e−03 3e−03 1e−02 3e−02 20 30 40 50 60 Z Baryon density [fm−3] 30 40 50 Z 0.00 0.05 0.10 0.15 0.20 0.25 p(Z) (a) 30 40 50 Z (b) Fig. 2. 3.2. Impurity parameter Variation with baryon density nB of the impurity parameter Qimp in the inner-crust regime at two selected temperatures: T = 1010 K, and T = Tm, based on BSk22 (red lines), BSk24 (black lines), BSk25 (blue lines), and BSk26 (green lines) CLD calculations. Fig. 5. Variation with baryon density nB of the impurity parameter Qimp in the crust at the crystallization temperature T = Tm. In the outer crust regime, the solid (dashed) line represents the BSk24 CLD (HFB-24) prediction. In the inner crust regime, the impurity parameter is calculated in the CLD approximation. Points indicate the neutron-drip transition. tical equilibrium based on a CLD model for the nuclear part of the ion energetics. To achieve thermodynamical consistency, a rearrangement term is explicitly worked out. This term has an important eﬀect on the distributions and it is necessary to recover the correct limit at zero temperature. Since NSs are born hot, the equilibrium composition of a mature NS can be determined assuming a liquid phase for the MCP, at the lowest temperature at which strong and weak equilibrium are attained. In the absence of a dynamical estimation of the associated reaction rates, we consider the lowest temperature limit as given by the OCP crys- tallization temperature Tm. We show that at that temperature, the OCP approximation gives a very good estimation of the aver- age composition of the inner crust, with nonlinear mixing terms playing a very small role in the liquid phase. However, an impor- tant contribution of impurities is obtained, favoring the picture of a temperature-independent high resistivity in the inner crust for all T < Tm. We expect that a fully microscopic calculation would still present oscillations in Qimp beyond the drip point, that these oscillations should be progressively damped going deeper in the star, and that our calculation can be taken as a smooth interpolation of that oscillating behavior. g To have a quantitative prediction of the impurity factor, the problem of model dependence has to be addressed. Apart from the modeling of ﬁnite temperature shell eﬀects discussed above, the main source of uncertainty of the calculation comes from the choice of the nuclear functional. We show, in Fig. 6, the impu- rity parameter (Eq. (22)) as a function of the baryon density in the inner crust, at the crystallization temperature Tm (solid line), for the four considered BSk CLD models. 3.2. Impurity parameter The impurity parameter at the crystallization temperature is shown in the whole crust in Fig. 5, as calculated with the BSk24 CLD model (solid line). The black dot marks the neutron-drip point. We can see that the impurity parameter in the inner crust is higher than in the outer crust, meaning that the distribution is less peaked, thus the OCP approximation is less reliable than in the outer crust. Indeed, larger values of Qimp indicate more appre- ciable deviations from the OCP predictions. For comparison, we also plot the impurity parameter, taken from Fantina et al. (2020), calculated in the outer crust with the HFB-24 model (dashed line). The latter calculations show more prominent vari- ations of Qimp with respect to the CLD model calculations. This is due to the natural inclusion of shell eﬀects in the fully micro- scopic calculations, which exhibit bimodal distributions around values of pressure corresponding to the simultaneous presence of the two characteristic elements of adjacent layers (see Fig. 6 in Fantina et al. 2020). These strong ﬂuctuations are naturally smoothed out in the CLD model, because the nuclear functional varies continuously with A and Z, and so does the probability. However, we can observe that the CLD calculation nicely inter- polates the microscopic results, with an average impurity factor steadily increasing with the density and lying in the Qimp ≈ 0, 1−2 interval. In the inner crust, neutrons drip out of the ﬁnite ion volume, and the associated shell eﬀects naturally disappear (Chamel 2006). The inclusion of proton shell eﬀects in the inner crust would require a formidable numerical investment, which is far beyond the scope of the paper. Moreover, it was suggested in Carreau et al. (2020) that these eﬀects are small at the higher melting temperature of the inner crust, and that their eﬀect on the observables is smaller than the uncertainty brought by our imper- fect knowledge of the smooth part of the energy functional. For this reason, shell eﬀects were completely neglected in our study. A77, page 7 of 9 A77, page 7 of 9 A77, page 7 of 9 A&A 640, A77 (2020) 10−3 10−2 nB [fm−3] 10 20 30 40 Qimp T = 1010 K T = Tm BSk22 BSk24 BSk25 BSk26 Fig. 6. Acknowledgements. This work has been partially supported by the IN2P3 Master Project MAC and the CNRS PICS07889. Discussions with N. Chamel are gratefully acknowledged. 3.2. Impurity parameter Variation with baryon density nB of the impurity parameter Qimp in the inner-crust regime at two selected temperatures: T = 1010 K, and T = Tm, based on BSk22 (red lines), BSk24 (black lines), BSk25 (blue lines), and BSk26 (green lines) CLD calculations. 10−3 10−2 nB [fm−3] 10 20 30 40 Qimp T = 1010 K T = Tm BSk22 BSk24 BSk25 BSk26 10−8 10−7 10−6 10−5 10−4 10−3 10−2 nB [fm−3] 10−3 10−2 10−1 100 101 102 Qimp T = Tm Fig. 5. Variation with baryon density nB of the impurity parameter Qimp in the crust at the crystallization temperature T = Tm. In the outer crust regime, the solid (dashed) line represents the BSk24 CLD (HFB-24) prediction. In the inner crust regime, the impurity parameter is calculated in the CLD approximation. Points indicate the neutron-drip transition. 10−3 10−2 nB [fm−3] 10 20 30 40 Qimp T = 1010 K T = Tm BSk22 BSk24 BSk25 BSk26 Fig. 6. Variation with baryon density nB of the impurity parameter Qimp in the inner-crust regime at two selected temperatures: T = 1010 K, and T = Tm, based on BSk22 (red lines), BSk24 (black lines), BSk25 (blue lines), and BSk26 (green lines) CLD calculations. tical equilibrium based on a CLD model for the nuclear part of 10−8 10−7 10−6 10−5 10−4 10−3 10−2 nB [fm−3] 10−3 10−2 10−1 100 101 102 Qimp T = Tm Fig. 5. Variation with baryon density nB of the impurity parameter Qimp in the crust at the crystallization temperature T = Tm. In the outer crust regime, the solid (dashed) line represents the BSk24 CLD (HFB-24) prediction. In the inner crust regime, the impurity parameter is calculated in the CLD approximation. Points indicate the neutron-drip transition. 10−8 10−7 10−6 10−5 10−4 10−3 10−2 nB [fm−3] 10−3 10−2 10−1 100 101 102 Qimp T = Tm 10−8 10−7 10−6 10−5 10−4 10−3 10−2 nB [fm−3] 10−3 10−2 10−1 100 101 102 Qimp T = Tm Fig. 5. Variation with baryon density nB of the impurity parameter Qimp in the crust at the crystallization temperature T = Tm. In the outer crust regime, the solid (dashed) line represents the BSk24 CLD (HFB-24) prediction. In the inner crust regime, the impurity parameter is calculated in the CLD approximation. Points indicate the neutron-drip transition. Fig. 6. 3.2. Impurity parameter These data are avail- able in tabular format at the CDS. Considering that the chosen models are believed to cover the main uncertainty on the nuclear equation of state at subsaturation density (Pearson et al. 2018), we can take the spread of Qimp values obtained by the four cal- culations as a reasonable estimation of the uncertainty on the impurity parameter. Since this latter represents the variance of the charge distribution, low values of Qimp indicate that the dis- tribution is quite peaked, and thus the OCP approach is a good approximation, which can also be seen from Figs. 1 and 2, panel (a). The monotonic increase of the impurity parameter with den- sity is also in accordance with Fig. 3, which clearly shows the growth of the width of the charge distribution with increasing density. While at lower densities all the models predict similar values of the impurity parameter (≲5), at higher densities the spread among the models becomes larger, and the model associ- ated with the lower (larger) symmetry energy coeﬃcient at sat- uration has the larger (lower) Qimp. The same trend is observed at T = 1010 K (dashed lines in Fig. 6), although the hierarchy of the models is not preserved. In order to reach quantitative predictions for the associated impurity parameter, we considered four diﬀerent realistic micro- scopic nuclear functionals of the BSk family, which cover the present uncertainty in the nuclear modeling below the satura- tion density of nuclear matter. The impurity parameter is seen to increase with the density, and values in the Qimp ≈20−40 inter- val are reached at the highest densities considered in this study, namely nB = 0.04 fm−3. Higher values of the impurity parameter might be expected in the deepest region of the inner crust, close to the core-crust transition, due to the presence of nonspheri- cal pasta phases, which have not been considered in the present study. ( p g ) p y p Burgio, G. F., & Fantina, A. F. 2018, The Physics and Astrophysics of Neutron Stars, eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & I. Vidaña (Springer), Astrophys. Space Sci. Libr., 457, 255 4. Conclusions Blaschke, D., & Chamel, N. 2018, The Physics and Astrophysics of Neutron Stars, eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & I. Vidaña (Springer), Astrophys. Space Sci. Libr., 457, 337 In this paper, we present a multicomponent approach to the mod- eling of the crust of isolated, unmagnetized NSs. Completing the work of Fantina et al. (2020) on the outer crust, the composition of the inner crust is evaluated with an extended nuclear statis- ( p g ) p y p Burgio, G. F., & Fantina, A. F. 2018, The Physics and Astrophysics of Neutron Stars, eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & I. Vidaña (Springer), Astrophys. Space Sci. Libr., 457, 255 A77, page 8 of 9 T. Carreau et al.: Inner crust of a neutron star in a MCP Carreau, T., Gulminelli, F., & Margueron, J. 2019, Eur. Phys. J. A, 55, 188 Jones, P. B. 2001, MNRAS, 321, 167 Jones, P. B. 2004, Phys. Rev. Lett., 93, 221101 Carreau, T., Gulminelli, F., Chamel, N., Fantina, A. F., & Pearson, J. M. 2020, A&A, 635, A84 Jones, P. B. 2004, Phys. Rev. Lett., 93, 221101 Lattimer, J. M., & Swesty, F. D. 1991, Nucl. Phys. A, 535, 331 , , Chamel, N. 2006, Nucl. Phys. A, 773, 263 Meisel, Z., Deibel, A., Keek, L., Shternin, P., & Elfritz, J. 2018, J. Phys. G, 45, 093001 y Chamel, N., & Haensel, P. 2008, Liv. Rev. Relativ., 11, 10 Newton, W. G., Gearheart, M., & Li, B. A. 2013, ApJS, 204, 9 Fantina, A. F., De Ridder, S., Chamel, N., & Gulminelli, F. 2020, A&A, 633, A149 Oertel, M., Hempel, M., & Klähn, T., & Typel, S., 2017, Rev. Mod. Phys., 89, 015007 Goriely, S., Chamel, N., Janka, H.-T., & Pearson, J. M. 2011, A&A, 531, A78 Pearson, J. M., Chamel, N., Fantina, A. F., & Goriely, S. 2014, Eur. Phys. J A, 50, 43 riely, S., Chamel, N., & Pearson, J. M. 2013, Phys. Rev. C, 88, 024 Pearson, J. M., Chamel, N., Potekhin, A. Y., et al. 2018, MNRAS, 481, 2994 Gourgouliatos, K. N., & Esposito, P. 2018, The Physics and Astrophysics of Gourgouliatos, K. N., & Esposito, P. 2018, The Physics and Astrophysics of Neutron Stars, eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & I. Vidaña (Springer), Astrophys. Space Sci. Libr., 457, 57 Grams, G., Giraud, S., Fantina, A. F., & Gulminelli, F. 2018, Phys. Rev. C, 97, 035807 g , , p , , y p y Neutron Stars, eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & I. Vidaña (Springer), Astrophys. Space Sci. Libr., 457, 57 g p y p y Neutron Stars, eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & I Vid ñ (S i ) A t h S S i Lib 457 57 Pearson, J. M., Chamel, N., Potekhin, A. Y., et al. 2019, MNRAS, 486, 768 I. Vidaña (Springer), Astrophys. Space Sci. Libr., 457, 57 earson, J. M., Chamel, N., & Potekhin, A. Y. 2020, Phys. Rev. C, Pons, J. A., & Viganò D. & Rea, N., 2013, Nature Phys., 9, 431 Grams, G., Giraud, S., Fantina, A. F., & Gulminelli, F. 2018, Phys. T. Carreau et al.: Inner crust of a neutron star in a MCP Rev. C, 97, 035807 Grams, G., Giraud, S., Fantina, A. F., & Gulminelli, F. 2018, g y Potekhin, A. Y., & Chabrier, G. 2000, Phys. Rev. E, 62, 8554 Gulminelli, F., & Raduta, Ad R 2015, Phys. Rev. C, 92, 055803 y Schmitt, A., & Shternin, P. 2018, The Physics and Astrophysics of Neutron Stars, y Schmitt, A., & Shternin, P. 2018, The Physics and Astrophysic , , , , y p y eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & y Haensel, P., Potekhin, A. Y., & Yakovlev, D. G. 2007, Neutron Stars 1. Equation of state and structure (New York: Springer) Jones, P. B. 1999, Phys. Rev. Lett., 83, 3589 y p y eds. L. Rezzolla, P. Pizzochero, D. I. Jones, N. Rea, & I. Vidaña (Springer), Haensel, P., Potekhin, A. Y., & Yakovlev, D. G. 2007, Neutron Stars 1. Equation of state and structure (New York: Springer) Astrophys. Space Sci. Libr., 457, 455 Astrophys. Space Sci. Libr., 457, 455 p g Jones, P. B. 1999, Phys. Rev. Lett., 83, 3589 p y p Viganò, D., Rea, N., Pons, J. A., et al. 2013, MNRAS, 434, 123 A77, page 9 of 9
https://openalex.org/W2800679141	http://repository.essex.ac.uk/22062/1/fpsyg-09-00623.pdf	English	null	The Association Between Believing in Free Will and Subjective Well-Being Is Confounded by a Sense of Personal Control	Frontiers in psychology	2,018	cc-by	5,696	ORIGINAL RESEARCH published: 07 May 2018 doi: 10.3389/fpsyg.2018.00623 Edited by: Edited by: Baruch Eitam, University of Haifa, Israel Reviewed by: Jill Ann Jacobson, Queen’s University, Canada Guillermo B. Willis, Universidad de Granada, Spain Correspondence: Peter L. T. Gooding pltgoo@essex.ac.uk Reviewed by: Jill Ann Jacobson, Queen’s University, Canada Guillermo B. Willis, Universidad de Granada, Spain Keywords: free will, choice, control, satisfaction with life, subjective well-being, perceived stress, depression The Association Between Believing in Free Will and Subjective Well-Being Is Confounded by a Sense of Personal Control Peter L. T. Gooding, Mitchell J. Callan and Gethin Hughes Department of Psychology, University of Essex, Essex, United Kingdom The extent to which an individual believes in free will is associated with a number of positive life outcomes, including their own subjective well-being. However, it is not known whether the belief that one has free will per se is uniquely associated with subjective well-being over and above potential confounding variables. We examined a sense of personal control as one such confound—speciﬁcally, whether the association between free will belief (FWB) and subjective well-being is based, in part, on the degree to which an individual feels a sense of personal control over their life. In Study, 1 trait-level belief in personal control signiﬁcantly uniquely predicted satisfaction with life and stress, over and above the contribution of FWB. In Study 2, within-person daily ﬂuctuations in stress and depression were not signiﬁcantly predicted by daily changes in FWB over and above the contribution of personal control/choice. The ﬁndings provide new insight into the relationship between FWB and subjective well-being. Keywords: free will, choice, control, satisfaction with life, subjective well-being, perceived stress, depression INTRODUCTION *Correspondence: Peter L. T. Gooding pltgoo@essex.ac.uk A growing body of evidence has shown that believing in free will is associated with a variety of positive life outcomes, including feeling grateful for past events (MacKenzie et al., 2014), better job performance (Stillman et al., 2010), higher academic achievement (Feldman et al., 2016), passionate love (Boudesseul et al., 2016), satisfaction with life (Li et al., 2017), and lower levels of perceived stress (Crescioni et al., 2015). Specialty section: This article was submitted to Personality and Social Psychology, a section of the journal Frontiers in Psychology Received: 22 June 2017 Accepted: 12 April 2018 Published: 07 May 2018 Specialty section: This article was submitted to Personality and Social Psychology, a section of the journal Frontiers in Psychology Nonetheless, the extent to which belief in free will per se is associated with positive life outcomes or whether some third variable is driving these associations remains to be explored. One possibility is that the relationship between free will beliefs (FWBs) and positive life outcomes, such as satisfaction with one’s life, might be confounded by a sense of personal control. Indeed, it is well-established that a sense of personal control is positively associated with many of the same positive life outcomes that relate to FWB, including subjective well-being (for reviews, see Myers and Diener, 1995; Peterson, 1999; Ross and Mirowsky, 2013). Thus, it is unclear whether FWB are uniquely associated with indicators of subjective well-being over and above a sense of personal control. Received: 22 June 2017 Accepted: 12 April 2018 Published: 07 May 2018 Citation: Gooding PLT, Callan MJ and Hughes G (2018) The Association Between Believing in Free Will and Subjective Well-Being Is Confounded by a Sense of Personal Control. Front. Psychol. 9:623. doi: 10.3389/fpsyg.2018.00623 In their work exploring lay understandings of free will, Monroe and Malle (2010, 2014) found that people’s deﬁnitions of what it means to have free will diﬀered from philosophical understandings that typically view free will as the ability for our conscious minds (or a soul) to make decisions, regardless of brain states or prior causal events (Harris, 2012). Rather, May 2018 \| Volume 9 \| Article 623 1 Frontiers in Psychology \| www.frontiersin.org Control Not Free Will Predicts Well-Being Gooding et al. people deﬁned free will as their freedom to make choices and the ability to act without constraints—that is, their sense of personal control (see also Baumeister and Monroe, 2014). Thus, insofar that our participants’ lay concepts of FWB are speciﬁcally tied to having a sense of personal control, then individual diﬀerences in a sense of personal control might better predict subjective well-being than individual diﬀerences in FWB. Consistent with this idea, Monroe et al. (2017) found that people’s beliefs that an agent who committed an immoral act had the capacity to choose their actions better predicted judgments of their blameworthiness than did their beliefs that the agent had free will. We reasoned that the known association between FWB and subjective well-being might be confounded by a sense of personal control. people deﬁned free will as their freedom to make choices and the ability to act without constraints—that is, their sense of personal control (see also Baumeister and Monroe, 2014). Thus, insofar that our participants’ lay concepts of FWB are speciﬁcally tied to having a sense of personal control, then individual diﬀerences in a sense of personal control might better predict subjective well-being than individual diﬀerences in FWB. Consistent with this idea, Monroe et al. (2017) found that people’s beliefs that an agent who committed an immoral act had the capacity to choose their actions better predicted judgments of their blameworthiness than did their beliefs that the agent had free will. We reasoned that the known association between FWB and subjective well-being might be confounded by a sense of personal control. Results and Discussion Table 1 presents descriptive statistics, alpha reliabilities, and correlations among the measures. All of the measures correlated signiﬁcantly in the expected directions. FWB positively correlated with sense of personal control, and both correlated positively with SWL and negatively with perceived stress. Method Participants Participants from the United States were recruited through Amazon’s Mechanical Turk (N = 284). Demographic information was not collected (but see Levay et al., 2016, for information on the typical demographic composition of Mechanical Turk workers). Nineteen additional participants were excluded because of duplicate IP addresses (n = 6) or failing a basic attention check item (n = 13). A power analysis showed that our sample size had 80% power to detect “small-to-medium” eﬀect sizes (f 2 = 0.028; α = 0.05, two-tailed) in our multiple regression analysis. A multiple regression analysis showed that sense of personal control, b = 0.85, β = 0.49, SE = 0.098, t(281) = 8.65, p < 0.001, sr2 = 0.20, but not FWB, b = 0.003, β = 0.05, SE = 0.004, t(281) = 0.81, p = 0.42, sr2 = 0.002, uniquely predicted scores on the SWLS. Likewise, personal control, b = −0.67, β = −0.443, SE = 0.090, t(281) = −7.42, p < 0.001, sr2 = 0.20, but not FWB, b = 0.003, β = 0.04, SE = 0.004, t(281) = 0.73, p = 0.46, sr2 = 0.002, uniquely predicted perceived stress. Because confounding relationships are a special case of indirect relationships (MacKinnon et al., 2000), we tested whether there was a signiﬁcant decrease in the regression weight for FWB when modeled with a sense of personal control to predict each of SWL and perceived stress compared to Citation: Participants’ sense of personal control was gauged using a ﬁve-item measure (e.g., “Other people determine most of what I can and cannot do”; “There is little I can do to change many of the important things in my life”; “I can do just about anything I really set my mind to”; Chou et al., 2016, adapted from Lachman and Weaver, 1998). Participants indicated their level of agreement on a ﬁve-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). Higher scores indicate a greater sense of personal control. Participants’ perceived stress was measured using two items: “In the past year, how would you rate the amount of stress in your life (at home and at work)?” (1 = no stress to 6 = extreme stress; Littman et al., 2006) and “Stress means a situation in which a person feels tense, restless, nervous, or anxious or is unable to sleep at night because his/her mind is troubled all the time. Do you feel this kind of stress these days?” (1 = not at all to 6 = very much; Elo et al., 2003). Responses to the two items were highly correlated (r = 0.73, p < 0.001) and therefore averaged to form a composite measure of perceived stress. Across two studies, we compared the relative predictive utility of perceived control/choice and FWB across various indicators of subjective well-being. Study 1 investigated the degree to which personal control and FWB uniquely predicted satisfaction with life and perceived stress. Study 2 assessed how daily changes in perceived choice/control and FWB predicted life stress and depression across a 2-week period. Given the foregoing analysis, we predicted that the known associations between FWB and subjective well-being could be explained, in part, by people’s perceived ability to have choice and to control their lives. In Study 2 we also assessed participants’ qualitative deﬁnitions of free will, to investigate whether they ﬁt with previously reported lay conceptions of FWB (cf. Monroe and Malle, 2010). Participants’ life satisfaction was measured using Diener et al. (1985) widely used Satisfaction With Life Scale (SWLS), which is comprised of ﬁve items (e.g., “In most ways my life is close to my ideal”; 1 = strongly disagree to 7 = strongly agree). Alpha reliabilities for all measures with more than one item are shown in Table 1. Procedure and Materials Participants were instructed that they would complete a survey about their beliefs and opinions. We measured participants’ belief in free will using a single-item, graphical slider scale (“Using the slider provided, please indicate the extent to which you believe in free will”). The scale ranged from 0 (no belief in free will) to 100 (absolute belief in free will), and the starting position of the slider was set to the mid-point of the scale. This measure has been shown to have good convergent (Schooler et al., 2014) and predictive (e.g., Feldman et al., 2016) validity, and single-item free will measures have been shown to be sensitive to experimental manipulations of FWBs (MacKenzie et al., 2014; Nahmias et al., 2014; Monroe et al., 2017). TABLE 1 \| Descriptive statistics and correlations among measures in Study 1. Measures Mean (SD) 1 2 3 4 (1) FWB 82.52 (19.57) – (2) Control 3.82 (0.83) 0.426∗∗ (0.83) (3) SWLS 4.20 (1.44) 0.254∗∗ 0.510∗∗ (0.97) (4) Stress 3.61 (1.25) −0.145∗ −0.424∗∗ −0.409∗∗ (0.83) SWLS, the Satisfaction With Life Scale. When applicable, alpha reliabilities are presented in parentheses along the diagonal. ∗p < 0.05; ∗∗p < 0.01. SWLS, the Satisfaction With Life Scale. When applicable, alpha reliabilities are presented in parentheses along the diagonal. ∗p < 0.05; ∗∗p < 0.01. May 2018 \| Volume 9 \| Article 623 Frontiers in Psychology \| www.frontiersin.org 2 Control Not Free Will Predicts Well-Being Gooding et al. when FWB was modeled alone. Using Preacher and Hayes’s (2008) bootstrapping procedure for testing indirect eﬀects (10,000 resamples for each analysis), the relationships between FWB and SWL (indirect relationship = 0.015, β = 0.21, 95% bias-corrected and accelerated conﬁdence interval: 0.011, 0.022) and FWB and perceived stress (indirect relationship = −0.012, β = −0.19, 95% bias-corrected and accelerated conﬁdence interval: −0.017, −0.008) were signiﬁcantly reduced from their zero-order validities when statistically controlling for a sense of personal control. Of relevance here, during this initial session participants were asked to respond to an open-ended question about their FWBs: “Please explain what you think it means to have free will” (Monroe and Malle, 2010). Responses to this question were coded by two raters using Monroe and Malle’s (2010) original coding scheme. We included this question to replicate Monroe and Malle’s (2010) ﬁndings surrounding what “free will” means to people. Method Participants We coded participants’ open-ended responses using Monroe and Malle’s (2010) coding scheme. Speciﬁcally, we coded the responses the question “Please explain what you think it means to have free will” in terms of whether participants noted: (a) making decision or choices, (b) doing what they want, and (c) acting without internal or external constraints. Shown in Table 2, consistent with Monroe and Malle (2010, 2014), the majority of participants’ deﬁnitions of free will referred to the ability to decide/choose, doing what one wants, and/or being free of constraints. During the coding and analysis it was also clear that none of our participants deﬁned free will as reliant upon notions of indeterminism, magical causation or other qualities needed for the type of free will debated by philosophers (see Monroe and Malle, 2010; Baumeister and Monroe, 2014, for discussions). The ﬁnal sample of participants were 88 staﬀor students from the University of Essex (Mage = 24.18, SDage = 6.50; 77% female) who participated in exchange for a monetary reward (£1 for an initial session and £1 for every daily diary completed) and the chance to win gift cards. Two additional participants completed measures during an initial session but did not complete any of our focal daily measures. The ﬁnal sample size was determined by how many participants we could recruit within our monetary budget and time constraints. Procedure and Materials We measured participants’ daily FWB using a single-item, graphical slider scale (“Using the slider provided, please indicate the extent to which you believed you had free will today”). The scale ranged from 0 (no belief in free will today) to 100 (absolute belief in free will today). We measured participants’ sense that they controlled their actions and were free to choose that day using single-item, graphical slider scales (“Using the slider provided, please indicate the extent to which you believed you were in control of your actions today”; “. . .you were free to choose whatever you wanted to do today”). Scores could range from 0 (no control/no choice at all today) to 100 (complete control/complete choice today). Scores on these two daily measures were averaged to form a composite control/choice variable (within-person reliability = 0.54; see Nezlek, 2017). As our focal criterion variables, we measured participants’ daily stress (“Today, I felt stressed”) and daily depression (“Today, I felt depressed”) using four-point scales (1 = not at all, 4 = very much). Depression is an element of the unpleasant aﬀect component of subjective well-being (Diener et al., 1999). Procedure and Materials At the end of the initial session, participants were informed that they would receive daily emails including a link to a 10- min survey. The daily surveys were emailed to participants every day for 14 days at 5:00 PM; they had until 3:00 AM to complete the daily surveys. Participants who failed to complete more than ﬁve daily surveys were removed from the study (i.e., no longer sent the email links), but all data from participants who completed at least one daily survey were retained for analysis. Along with several questions unrelated to the current research interests, participants completed the following daily measures: These results suggest that the associations between FWB and SWL and FWB and perceived stress are largely due to co-variation between FWB and a sense of personal control. One limitation, however, is that our multiple regression approach fails to take measurement unreliability into account, which Westfall and Yarkoni (2016) showed can lead to spurious conclusions when testing for incremental validity. To address this issue, we replicated our results using structural equation modeling. Speciﬁcally, using the lavaan package in R (Rosseel, 2012), we speciﬁed measurement models for each predictor (sense of personal control and FWB) and each outcome to predict the latent outcome variables (separately for SWL and perceived stress) from the latent predictor variables. Because FWB was measured by a single indicator and therefore reliability could not be estimated empirically, we had to constrain the reliability of the FWB slider scores in our models. Following Westfall and Yarkoni’s (2016) recommendation, we tested a range of assumed reliabilities for our FWB measure (from 0.2 to 1 in increments of 0.2). Assuming the reliability for the FWB scores was as low as 0.4 (models failed to converge when the reliability of the FWB scores was assumed to be 0.2), analyses showed that a sense of personal control signiﬁcantly predicted both SWL, Z = 3.99, p < 0.001, and perceived stress, Z = −3.88, p < 0.001, over and above FWB. In no case did FWB signiﬁcantly predict the outcome variables over and above the contributions of sense of personal control (all ps > 0.13; the partial relationship between FWB and stress tended to become more positive when we assumed lower reliability for FWB scores). Daily Stress and Depression by participants; including correlations among the random eﬀects led to problems with convergence; Barr et al., 2013). All predictors were person-centered to control for between-person variance in the predictors. We did not model time (days) because we had no theoretical reason to expect time to inﬂuence daily changes in stress or depression across the 14-days. Given the nested structure of the data (daily responses nested within participants), analyses were performed using multilevel modeling (Nezlek, 2012). Analyses were performed using the lme4 package (Bates et al., 2015) in R, with maximal but uncorrelated random eﬀects (i.e., random slopes and intercepts On average participants completed 10.74 (SD = 3.75) of the 14 daily surveys (range = 13; total daily surveys completed = 944). TABLE 2 \| Content coding of the folk deﬁnitions of free will. Coding category Percentage coder agreement Kappa of agreement Percentage of participants mentioning the category Ability to make a decision/choice 91 0.81 64 Doing what you want 84 0.69 50 Acting without constraints 87 0.72 69 Deﬁnitions of coding categories were taken from Monroe and Malle (2010). TABLE 2 \| Content coding of the folk deﬁnitions of free will. TABLE 3 \| Means, standard deviations, and proportion of variance in the predictors and outcome variables at the within- and between-person levels. TABLE 4 \| Linear mixed effects models predicting daily stress and daily depression from daily FWB and daily choice/control (alone and with both simultaneously). Daily FWB Daily choice/control b SE Wald 95% CI b SE Wald 95% CI Daily stress FWB alone −0.007∗ 0.002 [−0.012, −0.002] – – – Choice alone – – – −0.010∗ 0.003 [−0.015, −0.004] FWB and choice −0.002 0.002 [−0.006, 0.002] −0.009∗ 0.003 [−0.014, −0.003] Daily depression FWB alone −0.008∗ 0.003 [−0.013, −0.003] – – – Choice alone – – – −0.011∗ 0.003 [−0.017, −0.007] FWB and choice −0.002 −0.002 [−0.007, 0.002] −0.01∗ 0.003 [−0.016, −0.005] FWB, free will belief. ∗p < 0.05 (based on the Wald 95% conﬁdence interval not containing zero). FIGURE 1 \| Mean levels of the two main predictor variables (combined choice/control and free will beliefs) and the two criterion variables (stress and depression) across days. Stress and depression have been rescaled (from 1–4 to 0–100) for illustration. Procedure and Measures Participants attended an initial laboratory session where they completed a variety of measures unrelated to the current project. May 2018 \| Volume 9 \| Article 623 Frontiers in Psychology \| www.frontiersin.org 3 Control Not Free Will Predicts Well-Being Gooding et al. Daily Stress and Depression Frontiers in Psychology \| www frontiersin org 4 May 2018 \| Volume 9 \| Article 623 TABLE 4 \| Linear mixed effects models predicting daily stress and daily depression from daily FWB and daily choice/control (alone and with both simultaneously). Daily FWB Daily choice/control b SE Wald 95% CI b SE Wald 95% CI Daily stress FWB alone −0.007∗ 0.002 [−0.012, −0.002] – – – Choice alone – – – −0.010∗ 0.003 [−0.015, −0.004] FWB and choice −0.002 0.002 [−0.006, 0.002] −0.009∗ 0.003 [−0.014, −0.003] Daily depression FWB alone −0.008∗ 0.003 [−0.013, −0.003] – – – Choice alone – – – −0.011∗ 0.003 [−0.017, −0.007] FWB and choice −0.002 −0.002 [−0.007, 0.002] −0.01∗ 0.003 [−0.016, −0.005] FWB, free will belief. ∗p < 0.05 (based on the Wald 95% conﬁdence interval not containing zero). FIGURE 1 \| Mean levels of the two main predictor variables (combined choice/control and free will beliefs) and the two criterion variables (stress and depression) across days. Stress and depression have been rescaled (from 1–4 to 0–100) for illustration. Frontiers in Psychology \| www frontiersin org 4 May 2018 \| Volume 9 \| Article 623 odels predicting daily stress and daily depression from daily FWB and daily choice/control (alone and with both simultaneously). FIGURE 1 \| Mean levels of the two main predictor variables (combined choice/control and free will beliefs) and the two criterion variables (stress and depression) across days. Stress and depression have been rescaled (from 1–4 to 0–100) for illustration. May 2018 \| Volume 9 \| Article 623 Frontiers in Psychology \| www.frontiersin.org 4 Control Not Free Will Predicts Well-Being Gooding et al. Table 3 shows descriptive statistics and the proportion of variance at the within- and between-person levels for each of the measures we employed. control/choice. Crescioni et al. (2015) showed that although both self-eﬃcacy and locus of control were correlated with FWB, they did not explain the association between FWB and life outcomes (meaning in life and SWL). We chose to focus on measures of control/choice that more closely reﬂected the nature of layperson conceptions of free will (Monroe and Malle, 2010). Unlike Monroe et al. (2017), who manipulated/measured choice using vignettes, we used a self-report measure of the degree to which participants believed in the ability to control their behavior or have the capacity for choice. Daily Stress and Depression These measures eﬀectively captured the key elements of the lay concepts of free will to the extent that they reduced the predictive utility of FWB on perceived stress and depression. p y As expected, daily ﬂuctuations in choice/control were signiﬁcantly associated with daily ﬂuctuations in participants’ FWB, b = 0.51, SE = 0.07 (95% Wald conﬁdence interval [CI]: 0.38, 0.65; here, FWB was the outcome variable in the analysis). Shown in Table 4, both daily FWBs and daily choice/control beliefs signiﬁcantly predicted daily ﬂuctuations in stress and depression when modeled alone. However, when daily FWBs and daily choice/control were modeled together to predict daily stress and depression, only daily choice/control emerged as a signiﬁcant predictor. Put diﬀerently, at the within-person level, daily changes in FWBs did not account for signiﬁcant variability in daily stress and depression over and above the contributions of daily changes in choice/control. Figure 1 shows the means of FWB, choice/control, stress, and depression across the 14 days. Much recent research has investigated the role of FWBs in a number of life outcomes, as well as psychological well-being. Here, we provide evidence for the role of personal control/choice in explaining why FWB predicts stress and depression. For laypeople, belief in free will fundamentally means having the capacity to make choices and control one’s life (Monroe and Malle, 2010), and our Study 2 ﬁndings of participants’ deﬁnitions of free will conﬁrm this. This perception of personal control appears to be protective of perceived stress and depression such that individuals with strong belief in the degree to which they control their lives may be less likely to negatively react to stressful life events. We also provide further evidence for the role of perceived control in stress and depression. This goes beyond previous research, by utilizing measures of control/choice that are closely aligned to high level beliefs in free will. Future research should investigate the relative power of these diﬀerent aspects of choice in predicting stress and depression. Like in Study 1, we tested whether there was a signiﬁcant decrease in the regression weight for daily FWB when modeled with daily choice/control to predict each of daily stress and daily depression compared to when FWB was modeled alone. We used Rockwood and Hayes’s (2017; see Zhang et al., 2009) MLmed SPSS macro for multilevel mediation to perform these analyses (with 10,000 Monte Carlo samples). Daily Stress and Depression Analyses showed that the relations between daily FWB and daily stress (within-subject indirect relationship = −0.004, Z = −2.77, p = 0.006, 95% Monte Carlo CI: −0.007, −0.001) and daily FWB and daily depression (within-subject indirect relationship = −0.005, Z = −3.38, p = 0.001, 95% Monte Carlo CI: −0.008, −0.002) were signiﬁcantly reduced from their zero-order relationships when statistically controlling for a sense of personal control. Although we show that the predictive utility of FWB on personal life outcomes is abolished when controlling for personal choice, it remains possible that FWB does have unique predictive utility in other contexts. Indeed, the modest correlation between FWB and personal control suggests that FWB and personal control are not precisely the same thing. Nonetheless, recent work (Monroe et al., 2017) shows that the relationship between FBW and morality is similarly explained by notions of personal control. As such future research should seek to determine which behaviors or outcomes might be predicted by FWB over and above personal control. These ﬁndings are consistent with our trait-level ﬁndings reported in Study 1: associations between participants’ subjective well-being (in this case, daily stress and depression) and FWBs appear to be due to the co-variation between FWB and beliefs about having control and being able to choose. GENERAL DISCUSSION Across two studies we investigated the role of personal control and choice in the relationship between FWB and subjective well-being. Previous research has provided evidence for the predictive value of FWB on such outcomes (e.g., Crescioni et al., 2015). Here, we show that this association can be explained by perceived control/choice. Study 1 showed that trait-level belief in personal control signiﬁcantly uniquely predicted SWL and stress, whereas FWB did not. Study 2 conﬁrmed that within-person daily ﬂuctuations in stress and depression are not signiﬁcantly predicted by daily changes in FWB over and above the contribution of personal control/choice. Further research should also attempt to identify the direction of these relationships. For instance, much research on FWB assumes that belief or disbelief in free will drives life outcomes and personal well-being. However, while control beliefs inﬂuence how someone copes with a stressful event, this coping also feeds back into the individual’s sense of personal control (Anderson, 1977). As such, while belief in free will/choice may be protective of subjective well-being, stressful life events may also lead to a reduction in a sense of personal control. Previous research has shown that the association between FWB and judgments of others’ morality/blame is due to perceived capacity for choice (Monroe et al., 2017). The current studies extend this by showing that like judgments of others’ behavior, the relationship between FWB and personal life outcomes, relevant to subjective well-being, is also due to co-variation with Frontiers in Psychology \| www.frontiersin.org REFERENCES Methods 40, 879–891. doi: 10.3758/BRM.40.3.879 Diener, E., Suh, E. M., Lucas, R. E., and Smith, H. L. (1999). Subjective well-being: three decades of progress. Psychol. Bull. 125, 276–302. doi: 10.1037/0033-2909. 125.2.276 Rockwood, N. J., and Hayes, A. F. (2017). “MLmed: an SPSS macro for multilevel mediation and conditional process analysis,” in Paper Presented at the Annual Meeting of the Association of Psychological Science (APS), Boston, MA. Diener, E. D., Emmons, R. A., Larsen, R. J., and Griﬃn, S. (1985). The satisfaction with life scale. J. Pers. Assess. 49, 71–75. doi: 10.1207/s15327752jpa4901_13 Ross, C. E., and Mirowsky, J. (2013). “The sense of personal control: social structural causes and emotional consequences,” in Handbook of the Sociology of Mental Health. Handbooks of Sociology and Social Research, eds C. S. Aneshensel, J. C. Phelan, and A. Bierman (Dordrecht: Springer). Elo, A. L., Leppänen, A., and Jahkola, A. (2003). Validity of a single-item measure of stress symptoms. Scand. J. Work Environ. Health 29, 444–451. doi: 10.5271/ sjweh.752 Rosseel, Y. (2012). lavaan: an R Package for Structural Equation Modeling. J. Stat. Softw. 48, 1–36. doi: 10.3389/fpsyg.2014.01521 Feldman, G., Chandrashekar, S. P., and Wong, K. F. E. (2016). The freedom to excel: belief in free will predicts better academic performance. Pers. Individ. Dif. 90, 377–383. Schooler, J., Nadelhoﬀer, T., Nahmias, E., and Vohs, K. D. (2014). “Measuring and manipulating beliefs and behaviors associated with free will. the good, the bad, and the ugly,” in Surrounding Free Will: Philosophy, Psychology, Neuroscience, ed. A. R. Mele (Oxford: Oxford University Press). Harris, S. (2012). Free Will. New York, NY: Simon and Schuster. doi: 10.1016/j.paid. 2015.11.043 Stillman, T. F., Baumeister, R. F., Vohs, K. D., Lambert, N. M., Fincham, F. D., and Brewer, L. E. (2010). Personal philosophy and personnel achievement: belief in free will predicts better job performance. Soc. Psychol. Pers. Sci. 1, 43–50. doi: 10.1177/1948550609351600 Lachman, M. E., and Weaver, S. L. (1998). The sense of control as a moderator of social class diﬀerences in health and well-being. J. Pers. Soc. Psychol. 74, 763–773. doi: 10.1037/0022-3514.74.3.763 Levay, K. E., Freese, J., and Druckman, J. N. (2016). The demographic and political composition of Mechanical Turk samples. SAGE Open 6, 1–16. doi: 10.1177/ 2158244016636433 Westfall, J., and Yarkoni, T. (2016). Statistically controlling for confounding constructs is harder than you think. PLoS One 11:e0152719. doi: 10.1371/ journal.pone.0152719 Li, C., Wang, S., Zhao, Y., Kong, F., and Li, J. (2017). REFERENCES Monroe, A. E., and Malle, B. F. (2010). From uncaused will to conscious choice: the need to study, not speculate about people’s folk concept of free will. Rev. Philos. Psychol. 1, 211–224. doi: 10.1007/s13164-009-0010-7 Anderson, C. R. (1977). Locus of control, coping behaviors, and performance in a stress setting: a longitudinal study. J. Appl. Psychol. 62, 446–451. doi: 10.1037/ 0021-9010.62.4.446 Monroe, A. E., and Malle, B. F. (2014). “Free will without metaphysics,” in Surrounding Free Will, ed. A. R. Mele (New York, NY: Oxford University Press), 25–48. doi: 10.1093/acprof:oso/9780199333950.003.0003 Barr, D. J., Levy, R., Scheepers, C., and Tily, H. J. (2013). Random eﬀects structure for conﬁrmatory hypothesis testing: keep it maximal. J. Mem. Lang. 68, 255–278. doi: 10.1016/j.jml.2012.11.001 Myers, D. G., and Diener, E. (1995). Who is happy? Psychol. Sci. 6, 10–19. doi: 10.1111/j.1467-9280.1995.tb00298.x Bates, D., Maechler, M., Bolker, B., and Walker, S. (2015). Fitting linear mixed- eﬀects models using lme4. J. Stat. Softw. 67, 1–48. doi: 10.18637/jss.v067.i01 Nahmias, E., Shepard, J., and Reuter, S. (2014). It’s OK if ‘my brain made me do it’: people’s intuitions about free will and neuroscientiﬁc prediction. Cognition 133, 502–516. doi: 10.1016/j.cognition.2014.07.009 502–516. doi: 10.1016/j.cognition.2014.07.009 Baumeister, R. F., and Monroe, A. E. (2014). Recent research on free will: conceptualizations, beliefs, and processes. Adv. Exp. Soc. Psychol. 50, 1–52. doi: 10.1016/B978-0-12-800284-1.00001-1 Nezlek, J. B. (2012). “Multilevel modeling of diary-style data,” in Handbook of Research Methods for Studying Daily Life, eds M. R. Mehl and T. S. Conner (New York, NY: Guilford Press), 357–383. Boudesseul, J., Lantian, A., Cova, F., and Bègue, L. (2016). Free love? On the relation between belief in free will, determinism, and passionate love. Conscious. Cogn. 46, 47–59. doi: 10.1016/j.concog.2016.09.003 Nezlek, J. B. (2017). A practical guide to understanding reliability of within-person variability. J. Res. Pers. 69, 149–155. doi: 10.1016/j.jrp.2016.06.020 Chou, E. Y., Parmar, B. L., and Galinsky, A. D. (2016). Economic insecurity increases physical pain. Psychol. Sci. 27, 443–454. doi: 10.1177/095679 7615625640 Peterson, C. (1999). “Personal control and well-being,” in Well-Being, eds D. Kahnemann, E. Diener, and N. Schwarz (New York, NY: Russell Sage Foundation), 288–301. Crescioni, A. W., Baumeister, R. F., Ainsworth, S. E., Ent, M., and Lambert, N. M. (2015). Subjective correlates and consequences of belief in free will. Philos. Psychol. 29, 1–23. Preacher, K. J., and Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect eﬀects in multiple mediator models. Behav. Res. ETHICS STATEMENT Ethical approval was granted by the University of Essex Faculty of Science and Health Ethics Sub-Committee (Approval Nos. PG1602 and PG1603). Prior to testing, all participants were May 2018 \| Volume 9 \| Article 623 5 Control Not Free Will Predicts Well-Being Gooding et al. contributed to writing the manuscript, and approved the ﬁnal version. contributed to writing the manuscript, and approved the ﬁnal version. informed of their right to withdraw from the research at any time. All participants signaled their consent to participate. informed of their right to withdraw from the research at any time. All participants signaled their consent to participate. FUNDING This work was supported by studentship ES/J500045/1 from the Economic and Social Research Council. This work was supported by studentship ES/J500045/1 from the Economic and Social Research Council. The research was conducted online and analyzed by PG and MC. All three authors conceived of and designed the studies, REFERENCES The freedom to pursue happiness: belief in free will predicts life satisfaction and positive aﬀect among Chinese adolescents. Front. Psychol. 7:2027. doi: 10.3389/fpsyg.2016.02027 Zhang, Z., Zyphur, M. J., and Preacher, K. J. (2009). Testing multilevel mediation using hierarchical linear models problems and solutions. Organ. Res. Methods 12, 695–719. doi: 10.1177/1094428108327450 Littman, A. J., White, E., Satia, J. A., Bowen, D. J., and Kristal, A. R. (2006). Reliability and validity of 2 single-item measures of psychosocial stress. Epidemiology 17, 398–403. doi: 10.1097/01.ede.0000219721.89552.51 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. MacKenzie, M. J., Vohs, K. D., and Baumeister, R. F. (2014). You didn’t have to do that belief in free will promotes gratitude. Pers. Soc. Psychol. Bull. 40, 1423–1434. doi: 10.1177/0146167214549322 Copyright © 2018 Gooding, Callan and Hughes. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). Copyright © 2018 Gooding, Callan and Hughes. 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 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. MacKinnon, D. P., Krull, J. L., and Lockwood, C. M. (2000). Equivalence of the mediation, confounding and suppression eﬀect. Prev. Sci. 1, 173–181. doi: 10.1023/A:1026595011371 Monroe, A. E., Brady, G., and Malle, B. F. (2017). This isn’t the free will worth looking for general free will beliefs do not inﬂuence moral judgments, agent- speciﬁc choice ascriptions do. Soc. Psychol. Pers. Sci. 8, 191–199. doi: 10.1177/ 1948550616667616 May 2018 \| Volume 9 \| Article 623 Frontiers in Psychology \| www.frontiersin.org 6
W2767206052.txt	https://dash.harvard.edu/bitstream/1/37160395/1/5763512.pdf	en		Association between within-visit systolic blood pressure variability and development of pre-diabetes and diabetes among overweight/obese individuals	Journal of human hypertension	2,017	cc-by	6,361	ASSOCIATION BETWEEN WITHIN-VISIT SYSTOLIC BLOOD PRESSURE VARIABILITY AND DEVELOPMENT OF PRE-DIABETES AND DIABETES AMONG OVERWEIGHT/OBESE INDIVIDUALS Citation Joshipura, Kaumudi J., Francisco J. Muñoz-Torres, Maribel Campos, Alba D. Rivera-Díaz, and Juan C. Zevallos. 2017. “ASSOCIATION BETWEEN WITHIN-VISIT SYSTOLIC BLOOD PRESSURE VARIABILITY AND DEVELOPMENT OF PRE-DIABETES AND DIABETES AMONG OVERWEIGHT/OBESE INDIVIDUALS.” Journal of human hypertension 32 (1): 26-33. doi:10.1038/ s41371-017-0009-y. http://dx.doi.org/10.1038/s41371-017-0009-y. Published Version doi:10.1038/s41371-017-0009-y Permanent link http://nrs.harvard.edu/urn-3:HUL.InstRepos:37160395 Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA Share Your Story The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . Accessibility HHS Public Access Author manuscript Author Manuscript J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Published in final edited form as: J Hum Hypertens. 2017 December ; 32(1): 26–33. doi:10.1038/s41371-017-0009-y. ASSOCIATION BETWEEN WITHIN-VISIT SYSTOLIC BLOOD PRESSURE VARIABILITY AND DEVELOPMENT OF PREDIABETES AND DIABETES AMONG OVERWEIGHT/OBESE INDIVIDUALS Author Manuscript Kaumudi J. Joshipura, BDS, MS, ScDa,b, Francisco J. Muñoz-Torres, MPHa, Maribel Campos, MD, MSc, MBAa, Alba D. Rivera-Díaz, MD, LNDa, and Juan C. Zevallos, MDc aCenter for Clinical Research and Health Promotion, University of Puerto Rico Medical Sciences Campus, School of Dental Medicine, San Juan, Puerto Rico bDepartment of Epidemiology, Harvard T. H. Chan School of Public Health, Boston, MA, United States cDepartment of Medical and Population Sciences Research, Herbert Wertheim College of Medicine, Florida International University, Miami, FL, United States Abstract Author Manuscript Author Manuscript Short-term blood pressure variability is associated with pre-diabetes/diabetes cross-sectionally, but there are no longitudinal studies evaluating this association. The objective of this study is to evaluate the association between within-visit systolic and diastolic blood pressure variability and development of pre-diabetes/diabetes longitudinally. The study was conducted among eligible participants from the San Juan Overweight Adults Longitudinal Study (SOALS), who completed the three-year follow-up exam. Participants were Hispanics, 40–65 years of age, and free of diabetes at baseline. Within-visit systolic and diastolic blood pressure variability was defined as the maximum difference between three measures, taken a few minutes apart, of systolic and diastolic blood pressure respectively. Diabetes progression was defined as development of prediabetes/diabetes over the follow-up period. We computed multivariate incidence rate ratios adjusting for baseline age, gender, smoking, physical activity, waist circumference and hypertension status. Participants with systolic blood pressure variability ≥10 mm Hg compared to those with <10 mm Hg, showed higher progression to pre-diabetes/diabetes (RR=1.77, 95% CI: 1.30–2.42). The association persisted among never smokers. Diastolic blood pressure variability ≥ 10 mm Hg (compared to < 10 mm Hg) did not show an association with diabetes status progression (RR=1.20, 95% CI: 0.71–2.01). Additional adjustment of baseline glycemia, C- Users may view, print, copy, and download text and data-mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use: http://www.nature.com/authors/editorial_policies/license.html#terms Corresponding author: Kaumudi J. Joshipura, BDS, MS, ScD, Center for Clinical Research and Health Promotion, University of Puerto Rico Medical Sciences Campus, School of Dental Medicine; Department of Epidemiology, Harvard T. H. Chan School of Public Health. Mailing address: Center for Clinical Research and Health promotion, Office A107, University of Puerto Rico Medical Sciences Campus, School of Dental Medicine, PO Box 365067 San Juan, PR, 00936-5067, Tel: 787-237-0009; 787-758-2525 ext. 2585 Fax: 787-763-4868, kaumudi.joshipura@upr.edu. Conflict of Interest The authors have nothing to disclose. Joshipura et al. Page 2 Author Manuscript reactive protein, and lipids (reported dyslipidemia or baseline HDL or triglycerides) did not change the estimates. Systolic blood pressure variability may be a novel independent risk factor and an early predictor for diabetes, which can be easily incorporated into a single routine outpatient visit at none to minimal additional cost. Keywords Within-visit blood pressure variability; pre-diabetes; diabetes; hypertension; overweight; obesity Introduction Author Manuscript The published literature has shown that blood pressure (BP) is not constant, and it undergoes natural oscillations (modulation) over the long-term (visit-to-visit) and short-term (withinvisit or within 24 hours) (1–8). Several studies have shown that high visit-to-visit blood pressure variability (BPV) and high BP are strongly associated with increased cardiometabolic disorders including carotid artery atherosclerosis and stiffness, stroke, organ damage, and all-cause mortality (2, 3, 9, 10). One recent study among Japanese adults, where long-term visit-to-visit BPV was computed across baseline and three annual visits, showed a small elevated risk of diabetes of around 10% for an increment of 1 standard deviation (SD) or of 6 mm mercury (Hg) for systolic blood pressure variability (SBPV), and an increment of 1 SD or 4 mm Hg for diastolic (DBPV) (11). A study comparing 24 hour ambulatory blood pressure between 18 diabetics and 18 non-diabetics showed significantly higher crude systolic (SBP), diastolic (DBP) and mean daytime arterial blood pressure among diabetics compared to non-diabetics (5). Another small study also showed increased BP variability among diabetics during the day (12). Author Manuscript A bi-directional relationship between high BP and diabetes is suggested (13). A recent report showed that 20 mm Hg higher SBP and 10 mm Hg higher DBP were associated with a 58% and a 52% higher risk of diabetes respectively in a large cohort (14). The same report also showed a 77% higher pooled relative risk of diabetes for a 20 mm Hg higher than usual SBP in a meta-analysis among 30 studies with 17,388 incident diabetes events (14). Withinvisit BPV assessed from repeated measures over a few minutes during an outpatient visit reflects a physiological transient fluctuation of autonomic stimulation, leading to a humoral response (1). Results from two cross-sectional studies suggest that higher within-visit BPV is associated with higher fasting plasma glucose, and with pre-diabetes/diabetes (8, 15). Author Manuscript Although there is evidence relating high blood pressure and long-term BPV with increased risk of diabetes, there are no longitudinal studies published to date evaluating within-visit BPV as a cause or consequence of pre-diabetes/diabetes. Although the causality may be bidirectional, in our study we hypothesized that the autonomic dysfunction, seen as shortterm BPV, may impact the diagnosis of diabetes. Short-term BPV can be assessed easily during a routine clinical visit, with minimal if any additional cost compared to a single BP measure. Hence, within-visit BPV, could be potentially used in identifying subjects at increased risk of developing diabetes, and may have large potential for impacting clinical practice. Accordingly, we evaluate within-visit SBPV and DBPV as potential predictors of J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 3 Author Manuscript development of pre-diabetes/diabetes within the San Juan Overweight Adults Longitudinal Study (SOALS). Materials and Methods Study overview Author Manuscript Author Manuscript These analyses include SOALS participants who completed key components of the baseline and follow-up examination. The study was approved by the University of Puerto Rico Human Research Subjects Protection Office Institutional Review Board, and participants signed a written informed consent. SOALS is a longitudinal study conducted among civilian, non-institutionalized Hispanic adults recruited primarily from San Juan metropolitan area using flyers and various other means. The sample size was pre-determined for the primary aims relating periodontitis and pre-diabetes. Recruitment and baseline data collection started in 2011 and the follow up exam (planned for around 3 year follow up period) was completed in 2016. Our study population consists of multiracial individuals of Hispanics ethnicity, who reported their race as White (25%), Black (14%), and Mixed race (61%). A total of 1451 came for the baseline exam. Eligibility criteria for this cohort study include: 1) age between 40 and 65 years, 2) overweight or obese (body mass index of at least 25.0 kg/m2), 3) free of clinically diagnosed diabetes prior to the baseline exam. The baseline exclusion criteria are as follows: 1) physician-diagnosed type 1 or type 2 diabetes or taking either insulin or oral anti-hyperglycemic agents; 2) pregnancy; 3) physician-diagnosed hypoglycemia, congenital heart murmurs, heart valve disease, congenital heart disease, endocarditis, rheumatic fever, and hemophilia or bleeding disorders; 4) active dialysis treatment; 5) having undergone procedures related to cardiovascular disease; 6) severe health conditions or psychological or physical disabilities that would interfere with participation in the study; and 7) plans on moving away in the next three-year period. Participants with a provisional diagnosis of type 2 diabetes based on fasting plasma glucose ≥ 7 mmol/l (126 mg/dl), two-hour oral glucose tolerance ≥ 11.1 mmol/l (200 mg/dl), or HbA1c > 6.5% (48 mmol/mol), detected from the baseline blood tests, were further excluded. We completed baseline blood samples, interviews, and anthropometric measurements among 1206 participants who provided extensive contact information including additional contacts. Retention efforts included phone calls, letters and tokens. The follow-up exam had similar data collection procedures. From the 1206 participants, 950 (79%) who came to the follow-up visit were included; of the remaining, 6 were deceased, 68 refused to participate, 87 were not reached by phone, letters or e-mail, 41 moved out of Puerto Rico, 53 were reached but were unable to come (see figure 1), and 1 excluded from the analyses for missing physical activity data. Author Manuscript Blood Pressure Variability Assessment Study nurses experienced in clinical research were individually trained utilizing audiovisual techniques to minimize observer bias. The double stethoscope was also utilized for training purposes and to reduce within and between observer variability (16). Calibration was conducted between the trainees and an experienced trainer using the double stethoscope simultaneously. Retraining was conducted as necessary during the three-year study followup. The criteria for passing the training (17), which includes 4 readings of a videotape and 3 simultaneous reading using a double stethoscope with an experienced trainer, were as J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 4 Author Manuscript Author Manuscript follows. a) The overall systolic and diastolic mean must lie within 1.96 SD from the standard mean. This criterion detects consistent bias in blood pressure readings. b) No more than one systolic or diastolic pair difference may lie beyond ± 1.96 SD from the expected zero value. This criterion tests the repeatability of blood pressure readings, which corresponds to an allowable range of ± 2 mm Hg. The gold standard Korotkoff auscultatory method with a mercury sphygmomanometer was utilized to assess the blood pressure of each participant after 5 minutes of rest in a quiet and relaxing setting (18). The palpatory method was used to obtain an approximation of the SBP and to ensure an adequate level of inflation (19). At the baseline exam, three serial measurements (with intervals of 1 minute between measures) were conducted in the upper arm of all participants in a sitting position, after the appropriate sphygmomanometer cuff width was selected based on the mid-arm circumference. The three readings were recorded and averaged for SBP and DBP. Several different measures of SBPV have been used in the literature such as standard deviation or coefficient of variation; we chose the maximum difference between the three measures taken over few minutes apart as the simplest and most clinically translatable measure. A cross-sectional report used cutoffs as follows: low-BPV (≤ 10 mmHg), moderate-BPV (11–20 mmHg), and high-BPV (> 20 mmHg) for evaluating the association between BPV and progression to pre-diabetes/diabetes (15). In our study, over 90% had SBPV and DBPV below 10 mm Hg; hence, individuals were categorized as having high (≥ 10 mm Hg) or low (< 10 mm Hg) SBPV and DBPV. Glycemia assessment Author Manuscript Glucose and insulin levels were evaluated at baseline and follow-up at fasting, and after administration of a 75-g glucose load at 30, 60 and 120 minutes. Glucose was measured using an enzymatic colorimetric assay. Plasma insulin concentrations were analyzed using an immunochemiluminometric assay. The lab staff conducting these assays were unaware of the participants’ BPV status, hence assessments were unbiased. Insulin resistance was estimated using HOMA-IR (Fasting glucose × Fasting insulin)/405. HbA1c was measured with an assay based on a latex immunoagglutination inhibition method (DCA 2000+ Analyzer, Siemens Healthcare Diagnostics, NY, US). Author Manuscript We defined diabetes at each visit based on American Diabetes Association thresholds as having fasting glucose ≥ 7 mmol/l (126 mg/dl), 2-hour post load glucose ≥ 11.1 mmol/l (200 mg/dl) or HbA1c ≥ 6.5% (48 mmol/mol); pre-diabetes as having fasting glucose ≥ 5.6 mmol/l (100 mg/dl) but < 7 mmol/l (126 mg/dl), 2-hour post load glucose ≥ 7.8 mmol/l (140 mg/dl) but < 11.1 mmol/l (200 mg/dl) or HbA1c ≥ 5.7% (39 mmol/mol) but < 6.5% (48 mmol/mol); and normal glycemia if the three measurements were below the mentioned thresholds. We defined diabetes progression as a change in the diabetes classification from normal glycemia to pre-diabetes or from normal or pre-diabetes to diabetes during the follow-up period. In addition, participants who reported physician diagnosed diabetes during the follow-up period were classified as having diabetes progression at the follow-up exam. Covariate assessment Anthropometric measurements were taken in duplicate according to the NHANES III procedures, a third measure was taken when the first two measures differed by 5.0 mm, and the measures recorded were averaged for each individual. Body weight was measured using J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 5 Author Manuscript a Tanita scale (Tanita Body Composition Analyzer-TBF-310A), and height was measured using a portable stadiometer (Seca Corporation, Hanover, MD). Body Mass Index (BMI) was calculated as weight in kg/height in m2. Waist circumference was measured with a Gulick tape. A questionnaire was administered by trained interviewers and included sociodemographic characteristics, lifestyle factors, medical and family history, and time and frequency of physical activity during a typical week. Each activity was assigned a metabolic equivalent (MET) score based on the intensity. Activities from 3.0 to 6.0 metabolic equivalents (METs) were classified as moderate and activities with more than 6 METs as vigorous. Physical activity was categorized as meeting or not meeting WHO recommendations of at least 150 minutes of moderate-intensity aerobic physical activity per week, or at least 75 minutes of vigorous-intensity aerobic physical activity per week, or an equivalent combination of moderate and vigorous activity (20). Participants were classified as hypertensive if they had physician diagnosis of hypertension, if they reported taking high blood pressure medication, and/or had high blood pressure at the baseline exam with systolic blood pressure (SBP) ≥ 140 or diastolic blood pressure (DBP) ≥ 90 (mmHg). Participants were classified as pre-hypertensive if they were not classified as hypertensive nor reported high blood pressure medication, and had SBP between 120 and 140 or DBP between 80 and 90 (mmHg). Author Manuscript Data analyses methods The proportional hazards assumption for Cox was violated with crossing hazards, but the assumptions for the Poisson models were met. The results were similar for Cox and Poisson, and we present only the Poisson results here, which inherently factors variations in followup time (21, 22). Since we were analyzing a binary outcome, we obtained robust standard errors for the Poisson parameter estimates (23). Author Manuscript Major potential confounders were selected and included based on the literature (age, gender, smoking, physical activity, waist circumference, and hypertension status). Several additional potential confounders including fasting glucose, 2-hour post load glucose, HbA1c, insulin resistance measured as HOMA, C-reactive protein, or lipids (reported dyslipidemia or baseline HDL or triglycerides), family history of diabetes, dietary factors (fiber rich foods including fruits, vegetable, whole grain bread, and beans), and alcohol intake, were considered using the change in estimate procedure. We also explored the associations between SBPV and diabetes progression among sub-groups defined by age (≥55 vs. <55 years), gender, smoking (ever vs. never), physical activity (compliance with the WHO recommendations), BMI (overweight vs. obese), SBP (< 120 vs. ≥ 120 mmHg) and DBP (< 80 vs. ≥ 80 mmHg). Author Manuscript Results For this cohort study, the mean follow-up was 2.97 years (SD = 0.24, range: 2.26–4.48, median = 2.96 years). From the 951 participants who completed follow-up, 1 person was excluded for missing physical activity data; the remaining 950 with complete data on the key variables were included in the analyses. The SBPV ranged from 0–36 mm Hg (mean=4.3, SD=3.4) and DBPV ranged from 0–28 mm Hg (mean=3.5 and SD=2.8). Table 1 presents the J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 6 Author Manuscript description of participant characteristics at baseline, by high and low SBPV and DBPV. The group with SBPV ≥ 10 mm Hg was older, more physically active, less obese, more likely to have prior hypertension diagnosis, and smoked more compared to persons with SBPV < 10 mm Hg. The group with DBPV ≥ 10 mm Hg was more physically active, less obese, higher alcohol consumption, and sleep disordered breathing compared with persons with DBPV < 10 mm Hg. The groups with SBPV ≥ 10 mm Hg and with DBPV ≥ 10 mm Hg had higher diabetes status progression compared to low SBPV and low DBPV groups. The group with high SBPV also showed higher DBPV and vice versa. Individual SBP and DBP measures reduced with each subsequent measure in all groups. SBP was higher in groups with higher SBPV or DBPV, but there was no consistent difference for DBP. Author Manuscript Author Manuscript A total of 213 participants developed pre-diabetes/diabetes. Of the normoglycemic individuals at baseline, 35% developed pre-diabetes and 2% developed diabetes; of the persons with pre-diabetes at baseline, 12% developed diabetes in the 3 years of follow-up. Table 2 shows results from multivariate Poisson models adjusted for important confounders. Participants with SBPV ≥ 10 mm Hg showed higher progression to diabetes compared to those with SBPV < 10 mm Hg (RR=1.77, 95% CI: 1.30–2.42) after adjustment for major risk factors for diabetes including age, gender, smoking, physical activity, waist circumference, and hypertension status. Participants with DBPV ≥ 10 mm Hg (compared to those with DBPV < 10 mm Hg) did not show a significant association with diabetes status progression (RR=1.20, 95% CI: 0.71–2.01) after similar adjustment. Adjustment of baseline fasting glucose, 2-hour post load glucose, HbA1c, insulin resistance measured as HOMA, Creactive protein, or lipids (reported dyslipidemia or baseline HDL or triglycerides), family history of diabetes, dietary factors (fiber rich foods including fruits, vegetable, whole grain bread, and beans) and alcohol intake, showed that none had a substantial impact on the associations. Additional adjustment for mean SBP or DBP, or for hypertension medications associated with BPV in the literature (24) (calcium channel blocker and diuretics) also did not change the estimates. The data is sparse to draw conclusions about differences in effects estimates across subgroups (not shown in tables), but importantly, the SBPV and diabetes progression association persisted among never smokers (RR=1.85, 95% CI: 1.21–2.84), and among people who did not take calcium channel blocker and diuretics (RR=1.83, 95% CI: 1.33– 2.54). All sub-groups showed elevated diabetes status progression among high SBPV compared to low SBPV. Although the association was not significant among several subgroups due to small numbers, consistency of the association within different subgroups suggest robustness of the associations. Author Manuscript Discussion Our study shows a 77% higher progression to pre-diabetes/diabetes over a three-year followup period among participants with SBPV ≥ 10 mm Hg compared to participants with SBPV < 10 mm Hg. These results are independent of major risk factors and stronger than findings from a previously published cross-sectional study showing significant associations between high vs. low SBPV (> 10 mm Hg vs. ≤ 10 mm Hg) and pre-diabetes (16%) and diabetes J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 7 Author Manuscript (40%) as detected by HbA1c (15). We did not find a significant association relating DBPV with diabetes status progression. SBPV is associated with progression to pre-diabetes/diabetes independently of major risk factors for diabetes including age, low physical activity, adiposity, and baseline SBP, baseline glycemia, insulin resistance and dyslipidemia. The individual SBP and DBP measures reduced over time as may be expected with subsequent measurements in the same visit. Importantly, the SBPV association was not driven by SBP or DBP, which were controlled in the analyses. Author Manuscript Author Manuscript The SBPV association remained similar after controlling for C-reactive protein, suggesting that the association was not likely to be explained by inflammation, although we do not have data on IL-6 or TNF-α or oxidative stress markers, which may be more important in this context. Known risk factors for BPV include extrinsic environmental and behavioral risk factors such as, physical activity, sleep disturbances and emotional stimuli such as stress(25– 28); and intrinsic cardiovascular regulatory mechanisms such as autonomic dysfunction, disturbed baroreceptor sensitivity, humoral response and rheological factors have also been proposed (29). Another potential risk factor for BPV may be neuroendocrine system stimulation via increase of sympathetic tone. Emotional stress elicits autonomic response, which increase sympathetic activity, resulting in elevated blood pressure. Prolonged stimulation and eventual dysregulation of the sympathetic stimuli may lead to BPV (28, 30, 31). Based on our crude descriptive data presented in Table 1, SBPV seems associated with smoking and physical activity in our study, whereas DBPV is positively associated with male gender, physical activity, alcohol consumption and sleep disturbances, and inversely related with hypertension diagnosis and medications. The exact mechanism of physiological and environmental factors involved in the variability of blood pressure has not been well described (32). The primary mechanism of blood pressure modulation, which impacts BPV, is postulated to occur via autonomic nervous system outflow and humoral response (33). Diabetes free participants who presented an increase of 10 mmHg in SBPV at baseline, had significantly higher risk of developing pre-diabetes/diabetes approximately 3 years later. The potential association between high BPV and increased risk of pre-diabetes/diabetes may be mediated by persistent autonomic dysregulation, which in turn may lead to a humoral response by the pancreas (34, 35). Increased sympathetic tone has been proposed as a mechanism of increased reactivity of vasculature of the neuroendocrine system (36). BPV is a promoter of endothelial dysfunction (37), and autonomic dysregulation resulting from SBPV may lead to end organ damage and increased glycemia. Author Manuscript Since some of the factors impacting BPV may also independently increase the risk for hyperglycemia, it is possible that such unmeasured common risk factors may explain part of the association between SBPV and pre-diabetes/diabetes. One important such factor is arterial stiffness or lack of arterial distensibility (38). Although the assessment of clinical and neuroendocrine markers of arterial stiffness such as pulse pressure or hormone modulation was beyond the scope of this study and we did not collect resting heart rate measures at baseline, we did control for several well recognized contributing risk factors for arterial stiffness, such as age, obesity, physical activity and blood pressure. The association between short term systolic BPV and pre-diabetes/diabetes remained significant independent J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 8 Author Manuscript of such factors. Measures of arterial stiffness are not easy to assess at the primary care level, whereas, the capacity to determine the blood pressure variability is within reach of the most basic clinic settings, and can be easily assessed in day to day clinical practice. Importantly, this is the first longitudinal study suggesting that SBPV may be an early predictor of prediabetes/diabetes, independent of major traditional risk factors for diabetes and may be a surrogate for arterial stiffness. The association between SBP and pre-diabetes/diabetes is seen within different subgroups defined by age, gender, smoking, physical activity, overweight/obese, SBP and DBP (suggesting robustness, although power to detect significant associations is limited in many subgroups). Importantly, the association is significant among never smokers and among people who did not take pertinent hypertension medications, suggesting that it is independent of confounding by smoking and medications that may affect BPV. Author Manuscript Author Manuscript In this population of overweight/obese adults, high systolic blood pressure variability showed a significant association with subsequent increased progression to pre-diabetes/ diabetes. The associations between BPV and pre-diabetes/diabetes need to be replicated in additional longitudinal studies, in different populations, and mechanistic pathways need to be evaluated. Regardless of whether the association is causal, High SBPV could alert the health care provider to consider the patient as high risk for diabetes. SBPV could thus be an important marker for progression of diabetes in a population and could be used for identifying high-risk groups for preventive interventions. Among people identified with high SBPV, promotion of preventive lifestyle measures could help delay or prevent diabetes; furthermore, timely screening for diabetes as appropriate for this high risk group, could help delay or prevent diabetes related comorbidities. The potential return of investment through the prevention or delay in onset of diabetes and associated co-morbid conditions justifies targeted additional health promotion, screening and follow up efforts in such high risk individuals. In conclusion, the findings suggest that high BPV could potentially be a novel modifiable risk factor amenable to lifestyle and pharmacological intervention, which could impact both cardiovascular as well as metabolic health through the prevention or delay of development of pre-diabetes/diabetes. BPV could be easily assessed clinically to identify individuals for primary and secondary prevention interventions. Acknowledgments Sources of Funding Author Manuscript Research reported in this publication was supported by the National Institute of Dental and Craniofacial Research Grant R01DE020111 and the National Institute on Minority Health and Health Disparities Grant 2U54MD007587 of the National Institutes of Health. The authors acknowledge the SOALS team (Dr. Oelisoa M. Andriankaja, PhD, DDS; Ms. Tania Ginebra, RN; Ms. Carla León, MPH; Ms. Yashira Maldonado, BS; Dr. Sasha Martinez, PhD; Ms. Xiomara O’Farrill; Ms. Samantha Ordaz, BS; Dr. Cynthia Perez, PhD; Ms. Elaine Rodriguez, MT; Ms. Rosalyn Roman, MT; Mr. Rafael Ruiz, BS; Ms. Yadiris Santaella, RN; Ms. Grace Velez, BS; Mr. Jose Vergara, BS, Ms. Lay Wah, MPH; Mr. Jeanpaul Fernández) and PRCTRC laboratory personnel (Ms. Aracelis Arroyo, MS, HIM, MT and Ms. Nilda Gonzalez, MS, MT) who contributed to the conduct/oversight/planning of data collection, and Ms. Katya Giovannetti, MPH, MS for conducting a program review, for their help with the study. The authors have no conflicts of interest. J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 9 Author Manuscript References Author Manuscript Author Manuscript Author Manuscript 1. Conway J, Boon N, Davies C, Jones JV, Sleight P. Neural and humoral mechanisms involved in blood pressure variability. Journal of hypertension. 1984; 2(2):203–8. [PubMed: 6398335] 2. Faramawi MF, Delongchamp R, Said Q, Jadhav S, Abouelenien S. Metabolic syndrome is associated with visit-to-visit systolic blood pressure variability in the US adults. Hypertension research : official journal of the Japanese Society of Hypertension. 2014; 37(9):875–9. [PubMed: 24804612] 3. Li Y, Liu J, Wang W, Zhao D. The association between within-visit blood pressure variability and carotid artery atherosclerosis in general population. PLoS One. 2014; 9(5):e97760. [PubMed: 24835667] 4. Mancia G, Grassi G. Mechanisms and clinical implications of blood pressure variability. Journal of cardiovascular pharmacology. 2000; 35(7 Suppl 4):S15–9. [PubMed: 11346215] 5. Mokhtar RH, Ayob A, Mohd Noor N. Blood pressure variability in patients with diabetes mellitus. Asian cardiovascular & thoracic annals. 2010; 18(4):344–8. [PubMed: 20719784] 6. Ozawa M, Tamura K, Iwatsubo K, Matsushita K, Sakai M, Tsurumi-Ikeya Y, et al. Ambulatory blood pressure variability is increased in diabetic hypertensives. Clin Exp Hypertens. 2008; 30(3): 213–24. [PubMed: 18425701] 7. Schillaci G, Bilo G, Pucci G, Laurent S, Macquin-Mavier I, Boutouyrie P, et al. Relationship between short-term blood pressure variability and large-artery stiffness in human hypertension: findings from 2 large databases. Hypertension. 2012; 60(2):369–77. [PubMed: 22753222] 8. Shin JH, Shin J, Kim BK, Lim YH, Park HC, Choi SI, et al. Within-visit blood pressure variability: relevant factors in the general population. Journal of human hypertension. 2013; 27(5):328–34. [PubMed: 22971753] 9. Rossignol P, Kessler M, Zannad F. Visit-to-visit blood pressure variability and risk for progression of cardiovascular and renal diseases. Current opinion in nephrology and hypertension. 2013; 22(1): 59–64. [PubMed: 23165110] 10. Nagai M, Hoshide S, Ishikawa J, Shimada K, Kario K. Visit-to-visit blood pressure variations: new independent determinants for carotid artery measures in the elderly at high risk of cardiovascular disease. J Am Soc Hypertens. 2011; 5(3):184–92. [PubMed: 21531344] 11. Yano Y, Fujimoto S, Kramer H, Sato Y, Konta T, Iseki K, et al. Long-Term Blood Pressure Variability, New-Onset Diabetes Mellitus, and New-Onset Chronic Kidney Disease in the Japanese General Population. Hypertension. 2015; 66(1):30–6. [PubMed: 25987664] 12. McKinlay S, Foster C, Clark A, Clark S, Kemp F, Denver E, et al. Increased blood pressure variability during 24h blood pressure monitoring as an early sign of autonomic dysfunction in noninsulin-dependent diabetics. Journal of human hypertension. 1994; 8(12):887–90. [PubMed: 7884786] 13. Oktay AA, Akturk HK, Jahangir E. Diabetes mellitus and hypertension: a dual threat. Current opinion in cardiology. 2016 14. Emdin CAAS, Woodward M, Rahimi K. Usual Blood Pressure and Risk of New-Onset Diabetes Evidence From 4.1 Million Adults and a Meta-Analysis of Prospective Studies. Journal of The American College of Cardiology. 2015; 66(14):1552–62. [PubMed: 26429079] 15. Okada R, Yasuda Y, Tsushita K, Wakai K, Hamajima N, Matsuo S. Within-visit blood pressure variability is associated with prediabetes and diabetes. Scientific reports. 2015; 5:7964. [PubMed: 25589061] 16. Perloff D, Grim C, Flack J, Frohlich ED, Hill M, McDonald M, et al. Human blood pressure determination by sphygmomanometry. Circulation. 1993; 88(5 Pt 1):2460–70. [PubMed: 8222141] 17. Curb JD, Labarthe DR, Cooper SP, Cutter GR, Hawkins CM. Training and certification of blood pressure observers. Hypertension. 1983; 5(4):610–4. [PubMed: 6862583] 18. Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, et al. Recommendations for blood pressure measurement in humans and experimental animals: part 1: blood pressure measurement in humans: a statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research. Circulation. 2005; 111(5):697–716. [PubMed: 15699287] J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 10 Author Manuscript Author Manuscript Author Manuscript Author Manuscript 19. King GE. Influence of rate of cuff inflation and deflation on observed blood pressure by sphygmomanometry. American heart journal. 1963; 65:303–6. [PubMed: 14032831] 20. WHO Guidelines Approved by the Guidelines Review Committee. Geneva: 2010. Global Recommendations on Physical Activity for Health. 21. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute. 1959; 22(4):719–48. [PubMed: 13655060] 22. Rothman, KJ., Boice, John, D. Epidemiologic analysis with a programmable calculator. Vol. vii. Bethesda, Md. Washington: U.S. Dept. of Health, Education, and Welfare, Public Health Service for sale by the Supt. of Docs; 1979. p. 142 23. Zou G. A modified poisson regression approach to prospective studies with binary data. American journal of epidemiology. 2004; 159(7):702–6. [PubMed: 15033648] 24. Levi-Marpillat N, Macquin-Mavier I, Tropeano AI, Parati G, Maison P. Antihypertensive drug classes have different effects on short-term blood pressure variability in essential hypertension. Hypertens Res. 2014; 37(6):585–90. [PubMed: 24671016] 25. James GD, Toledano T, Datz G, Pickering TG. Factors influencing the awake-sleep difference in ambulatory blood pressure: main effects and sex differences. Journal of human hypertension. 1995; 9(10):821–6. [PubMed: 8576898] 26. Kario K, Schwartz JE, Pickering TG. Ambulatory physical activity as a determinant of diurnal blood pressure variation. Hypertension. 1999; 34(4 Pt 1):685–91. [PubMed: 10523347] 27. Loredo JS, Ancoli-Israel S, Dimsdale JE. Sleep quality and blood pressure dipping in obstructive sleep apnea. American journal of hypertension. 2001; 14(9 Pt 1):887–92. [PubMed: 11587154] 28. Kario K, Schwartz JE, Gerin W, Robayo N, Maceo E, Pickering TG. Psychological and physical stress-induced cardiovascular reactivity and diurnal blood pressure variation in women with different work shifts. Hypertension research : official journal of the Japanese Society of Hypertension. 2002; 25(4):543–51. [PubMed: 12358139] 29. Chenniappan M. Blood Pressure Variability: Assessment, Prognostic Significance and Management. Journal of The Association of Physicians of India. 2015; 63:47–53. 30. de Vente W, van Amsterdam JG, Olff M, Kamphuis JH, Emmelkamp PM. Burnout Is Associated with Reduced Parasympathetic Activity and Reduced HPA Axis Responsiveness, Predominantly in Males. BioMed research international. 2015; 2015:431725. [PubMed: 26557670] 31. Schulte W, Neus H, Thones M, von Eiff AW. Basal blood pressure variability and reactivity of blood pressure to emotional stress in essential hypertension. Basic research in cardiology. 1984; 79(1):9–16. 32. Filipovsky J, Ducimetiere P, Eschwege E, Richard JL, Rosselin G, Claude JR. The relationship of blood pressure with glucose, insulin, heart rate, free fatty acids and plasma cortisol levels according to degree of obesity in middle-aged men. Journal of hypertension. 1996; 14(2):229–35. [PubMed: 8728301] 33. Parati G, Ochoa JE, Salvi P, Lombardi C, Bilo G. Prognostic value of blood pressure variability and average blood pressure levels in patients with hypertension and diabetes. Diabetes care. 2013; 36(Suppl 2):S312–24. [PubMed: 23882065] 34. Kiba T. Relationships between the autonomic nervous system and the pancreas including regulation of regeneration and apoptosis: recent developments. Pancreas. 2004; 29(2):e51–8. [PubMed: 15257115] 35. Ortiz-Alonso FJ, Herman WH, Zobel DL, Perry TJ, Smith MJ, Halter JB. Effect of epinephrine on pancreatic beta-cell and alpha-cell function in patients with NIDDM. Diabetes. 1991; 40(9):1194– 202. [PubMed: 1936624] 36. McEwen BS. Stress, adaptation, and disease. Allostasis and allostatic load. Annals of the New York Academy of Sciences. 1998; 840:33–44. [PubMed: 9629234] 37. Diaz KM, Veerabhadrappa P, Kashem MA, Thakkar SR, Feairheller DL, Sturgeon KM, et al. Visitto-visit and 24-h blood pressure variability: association with endothelial and smooth muscle function in African Americans. Journal of human hypertension. 2013; 27(11):671–7. [PubMed: 23615389] J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 11 Author Manuscript 38. Liu YP, Gu YM, Thijs L, Asayama K, Jin Y, Jacobs L, et al. Do level and variability of systolic blood pressure predict arterial properties or vice versa? Journal of human hypertension. 2014; 28(5):316–22. [PubMed: 24152823] Author Manuscript Author Manuscript Author Manuscript J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 12 Author Manuscript Summary Table What is known about this topic? • Several studies have shown that high visit-to-visit blood pressure variability (BPV) and high blood pressure (BP) are strongly associated with increased cardiometabolic disorders, stroke, organ damage, and all-cause mortality. • Long-term visit-to-visit BPV and variation in BP assessed by ambulatory blood pressure monitoring has been associated with elevated cardiometabolic risk. • Only two published cross-sectional studies evaluated within-visit BPV, and found higher within visit BPV associated with higher fasting plasma glucose, and with pre-diabetes/diabetes. Author Manuscript What this study adds? • This is the first longitudinal study evaluating whether within-visit BPV is associated with development of pre-diabetes/diabetes. • The results show that within visit systolic blood pressure variability was associated with 77% higher pre-diabetes/diabetes, independent of major diabetes risk factors. • This study suggests a novel early predictor and potential novel risk factor for pre-diabetes/diabetes that can be easily evaluated in routine clinical visits with none to minimal cost. Author Manuscript Author Manuscript J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 13 Author Manuscript Author Manuscript Figure 1. Flow Chart of the San Juan Adults Longitudinal Study (SOALS) Participants’ Tracking Author Manuscript Author Manuscript J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 14 Table 1 Author Manuscript Baseline Characteristics by Blood Pressure Variability SBPV < 10, mm Hg (N = 867) SBPV ≥10, mm Hg (N = 83) DBPV < 10, mm Hg (N = 912) DBPV ≥ 10, mm Hg (N = 38) Age, mean (SD), years 50.4 (6.8) 52.6 (6.8) 50.6 (6.8) 49.7 (6.4) Female, No. (%) 638 (73.6) 66 (79.5) 680 (74.6) 24 (63.2) Current smoker, No. (%) 154 (17.8) 19 (22.9) 167 (18.3) 6 (15.8) Meeting WHO physical activity goals, No. (%) 472 (54.4) 51 (61.5) 499 (54.7) 24 (63.2) Obese, No. (%) 558 (64.4) 48 (57.8) 586 (64.3) 20 (52.6) Waist circumference, mean (SD), mm Author Manuscript Author Manuscript 1062.1 (144.4) 1048.5 (130.3) 1061.2 (142.6) 1052.3 (158.2) Alcohol consumption, mean (SD), grams per day 2.3 (5.8) 2.0 (4.2) 2.2 (5.6) 3.5 (6.5) Sleep breathing disorder, No. (%) 38 (4.4) 3 (3.6) 37 (4.1) 4 (10.5) 271 (31.3) 404 (46.6) 21 (25.3) 44 (53.0) 282 (30.9) 430 (47.2) 10 (26.3) 18 (47.4) SBPV, mean (SD), mm Hg 3.6 (2.1) 12.3 (4.2) 4.2 (3.3) 7.2 (5.0) DBPV, mean (SD), mm Hg 3.3 (2.5) 5.3 (4.6) 3.1 (2.0) 12.3 (4.8) Fasting glucose, mean (SD), mmol/L 5.1 (0.5) 5.2 (0.5) 5.1 (0.5) 5.2 (0.5) 2-hour post load glucose, mean (SD), mmol/L 6.4 (1.6) 6.5 (1.7) 6.4 (1.7) 6.1 (1.6) HbA1c % (SD) mmol/mol 5.7 (0.3) 5.7 (0.3) 5.7 (0.3) 5.7 (0.3) HOMA-IR, mean (SD) 2.5 (1.7) 2.3 (1.3) 2.5 (1.7) 2.1 (0.9) C reactive protein, mean (SD), mmol/L 55.8 (60.7) 42.4 (43.8) 55.7 (60.2) 28.2 (26.3) Diabetes status progression, No. (%) 182 (21.0) 31 (37.4) 202 (22.2) 11 (29.0) measurement 128.6(16.6) 136.8(22.9) 129.2(17.4) 132.3(17.9) measurement 127.9(16.4) 133.4(22.4) 128.2(17.0) 130.8(18.9) SBP 3rd measurement 127.5(16.1) 130.0(21.8) 127.6(16.5) 130.5(21.1) DBP 1st measurement 81.0(9.6) 82.0(12.5) 81.0(9.7) 85.4(13.4) DBP 2nd measurement 80.8(9.8) 80.6(12.6) 80.8(9.7) 80.6(16.8) DBP 3rd measurement 80.7(9.6) 79.6(12.0) 80.7(9.6) 79.6(14.6) Hypertension status, No. (%) Pre-hypertension Hypertension SBP 1st SBP 2nd SD= Standard Deviation. SBPV= Systolic Blood Pressure Variability. DBPV= Diastolic Blood Pressure Variability. HbA1c= Hemoglobin A1C. HOMA=Homeostatic Model Assessment-Insulin Resistance Index. Author Manuscript J Hum Hypertens. Author manuscript; available in PMC 2018 May 07. Joshipura et al. Page 15 Table 2 Author Manuscript Incidence Rate Ratios (IRR) relating within-visit blood pressure variability and diabetes status progression Diabetes status progression SBPV ≥ 10 mm Hg (Ref = SBPV < 10 mm Hg) IRR 95% CI Adjusted for age, gender and smoking 1.81 1.33–2.47 Multivariate* 1.77 1.30–2.42 DBPV ≥ 10 mm Hg (Ref = DBPV < 10 mm Hg) IRR 95% CI Adjusted for age, gender and smoking 1.24 0.75–2.07 Multivariate* 1.20 0.71–2.01 SBPV= Systolic Blood Pressure Variability DBPV= Diastolic Blood Pressure Variability Author Manuscript * Adjusted for age (years), gender, smoking (never, former, current), physical activity (meeting WHO guidelines), waist circumference (mm), and hypertension status: hypertension if reported physician diagnosis of hypertension or high blood pressure medication, and/or high blood pressure at the baseline exam with systolic blood pressure ≥ 140 or diastolic blood pressure ≥ 90 (mmHg); and pre-hypertension if no hypertension and either SBP between 80 and 140 or DBP between 80 and 90 (mmHg). Author Manuscript Author Manuscript J Hum Hypertens. Author manuscript; available in PMC 2018 May 07.
https://openalex.org/W2157135994	https://actavetscand.biomedcentral.com/counter/pdf/10.1186/1751-0147-43-231	English	null	Aspects on feed related prophylactic measures aiming to prevent post weaning diarrhoea in pigs.	Acta veterinaria Scandinavica	2,002	cc-by	9,394	Acta vet. scand. 2002, 43, 231-245. Acta vet. scand. 2002, 43, 231-245. E. coli; lactose; zinc oxide; probiotic; prevention. E. coli; lactose; zinc oxide; probiotic; prevention. Aspects on Feed Related Prophylactic Measures Aiming to Prevent Post Weaning Diarrhoea in Pigs By L. Melin1,2 and P. Wallgren1,2 By L. Melin1,2 and P. Wallgren1,2 1Department of Ruminant and Porcine Diseases, National Veterinary Institute, Uppsala, Sweden, 2Department of Large Animal Clinical sciences, Swedish University of Agricultural Sciences, Uppsala, Sweden. Melin L, Wallgren P: Aspects on feed related prophylactic measures aiming to pre- vent post weaning diarrhoea in pigs. Acta vet. scand. 2002, 43, 231-245. – The abil- ity of feed related measures to prevent or reduce post weaning diarrhoea (PWD) was ex- amined in a split litter study including 30 pigs from 6 litters allotted into 5 groups. Four groups were exposed to 3 pathogenic strains of E. coli via the environment at weaning. Three of them were given zinc oxide, lactose+ﬁbres or non-pathogenic strains of E. coli as probiotics. The challenged and the unchallenged control groups were given a stan- dard creep feed. Diarrhoea was observed in all challenged groups but not among unin- fected animals, and the incidence of diarrhoea was lower in the group given non- pathogenic E. coli compared to all other challenged groups. The severity of PWD also differed between litters. When corrected for mortality due to PWD, a decreased inci- dence of diarrhoea was also seen in the groups given zinc oxide or lactose+ﬁbres. The dominating serotype of E. coli within faecal samples varied from day to day, also among diarrhoeic pigs, indicating that diarrhoea was not induced by one single serotype alone. The diversity of the faecal coliform populations decreased in all piglets during the ﬁrst week post weaning, coinciding with an increased similarity between these populations among pigs in the challenged groups. This indicated an inﬂuence of the challenge strains, which ceased during the second week. The group given lactose+ﬁbres was least affected with respect to these parameters. In conclusion feed related measures may al- leviate symptoms of PWD. E. coli; lactose; zinc oxide; probiotic; prevention. Introduction are mixed at weaning, which will amplify the stress by ﬁghts that will last until a social rank is established (Spencer et al. 1989). Taken to- gether, these stress factors may affect the im- mune functions negatively post weaning (Ble- cha et al. 1983, Bailey et al. 1992, Puppe et al. 1997, Wattrang et al. 1998). This will coincide in time with alterations of the intestinal popula- tion, in terms of a less diversiﬁed faecal col- iform ﬂora, which is induced by the weaning (Kühn et al. 1993, Melin et al. 1997&2000a, Katouli et al. 1999). Weaning is one of the most dangerous situa- tions in the life of a pig and introduces a num- ber of stress factors. Some of these may be of infectious origin, such as E. coli (Hampson 1994), rotavirus (Saif et al. 1994), Clostridium perfringens (Estrada Correa & Taylor 1988). Other stressors are of non-infectious origin such as an abrupt separation from the sow and a sudden change of feed from sow milk to a ce- real based creep feed. The latter also include withdrawal of the protective IgA that is secreted in milk (Klobasa et al. 1981) and act locally in the intestine of the piglets. Sometimes piglets The drawbacks of weaning described above Acta vet. scand. vol. 43 no. 4, 2002 L. Melin & P. Wallgren 232 may contribute to outbreaks of post weaning di- arrhoea (PWD), and such outbreaks are often related to infections with E. coli. However, many authors have suggested PWD to mirror a syndrome rather than a speciﬁc infection, be- cause single infections/provocations have failed to induce PWD (Smith & Jones 1963, Hampson et al. 1985, Wathes et al. 1989, Nabu- urs et al. 1993, Madec et al. 1998 & 2000, Melin et al. 2000a). Consequently efforts aim- ing to reduce the negative impact of the wean- ing have been practised. The introduction of age segregated rearing systems that provides a good environment and a low pathogen load have been proven efﬁcient in preventing PWD (Madec et al. 1998), and the importance of us- ing relevant feeding systems to weaned piglets have been discussed (Rantzer 1997). ria to increase in number on behalf of other clones (Katouli et al. 1999). However, it should be noted that such administrations never should exceed 14 days due to the toxicity of zinc (Jensen-Waern et al. 1998). Acta vet. scand. vol. 43 no. 4, 2002 Introduction Also antibiotics may prevent PWD, but antibiotics as feed in- gredients have been prohibited in Sweden since 1986. The European Communities (EC) has followed this example regarding 8 out of 12 permitted substances 1999 (Council directive 70/524/EEC on Feed additives) and the future of the remaining substances are to be discussed. Yet another strategy to prevent PWD has been to introduce non-pathogenic microorganisms, aiming to obstruct colonisation of pathogenic microorganism of indigenous or exotic origin by competition for nutrients and receptor sites (Kyriakis 1999, Underdahl 1983). ( ) High protein concentration enhances growth (Gracia et al. 1999) and preheating of the feed facilitates feed utilization (Graham et al. 1989). However, such a feed may also contribute to PWD in several ways. The protective inﬂuence of chewing and saliva is reduced. The gastric passage rate is increased and the feed has a high acid binding capacity, resulting in a decreased effect of hydrochloric acid and proteolytic en- zymes (Bolduan 1992, Spencer et al. 1994). Also individual ingredients, such as soya (Jager 1986, Nabuurs 1986), have occasionally been proven provocative. Consequently, an interest has been paid to feed composition. The aim of the present study was to scrutinize the efﬁcacy of some strategies aiming to pre- vent development of PWD in pigs exposed to pathogenic strains of E. coli in a way that previ- ously had been proven to induce PWD. These strategies included feed composition, admix- ture of zinc oxide and administration of non- pathogenic bacteria. Materials and methods Animals, initial health status and experimental design The animals originated from a conventional herd free from diseases according to the A-list of International ofﬁce of epizootics, Aujeszky´s disease, Atrophic rhinitis, Transmissible gas- tro-enteritis, Porcine epidemic diarrhoea, Por- cine reproductive and respiratory syndrome, Brachyspira hyodysenteriae and Salmonellosis. To further reduce the pathogen load, sows were given antiparasitic treatment prior to farrowing (Ivomec® vet, MSD, Rahway, N J, USA). They were also vaccinated to prevent erysipelas and parvovirus (Nordpremum® Plus vet, Pharmacia & Upjohn Animal Health, Helsingborg, Swe- By adding pure lactose to the feed the abrupt switch of general energy source at weaning may be moderated and by using non-heated meal feed with extra dietary ﬁbres the intestinal pas- sage time may be prolonged (Johansen & Bach Knudsen 1994). Other efforts to prevent PWD have included admixture of ingredients that sta- bilise the intestinal ﬂora around weaning. For instance high amounts of feed administered zinc oxide preserve the intestinal ﬂora post weaning by preventing certain clones of bacte- Feed related to post weaning diarrhoea 233 Table 1. The experimental design of a study aiming to scrutinise the effect of different feed related prophylac- tic measures in pigs exposed to three pathogenic serotypes of E. coli via the environment. The pigs were weaned on living day 35. Table 1. The experimental design of a study aiming to scrutinise the effect of different feed related prophylac- tic measures in pigs exposed to three pathogenic serotypes of E. coli via the environment. The pigs were weaned on living day 35. Exposed to Feed Group Pathogenic Non-pathogenic Structure Heat processed Protein ZnO Lactose E. coli E. coli (75°C for 20 sek) (%) (2500 ppm) A - - Pelletedc Yes 15.5 - - B Yesa - Pelletedc Yes 15.5 Yes - C Yesa - Meald - 14.5 - Yes D Yesa - Pelletedc Yes 15.5 - - E Yes Yesb Pelletedc Yes 15.5 - - a) E. coli O147; K89, STb at the day of weaning; E. coli O141; K85, STb, VT2 and E. coli O149; K91, K88, STa, STb, LT three days post weaning b) A mixture of 106 CFU of each of 60 deﬁned non pathogenic strains of E. coli given per os. Materials and methods c) Startgris Fiber, Lantmännen, Svalöv, Sweden d) Meal feed with lactose, dietary ﬁbres and char cole (Nibble, Tillberga, Sweden) each pig was given an oral dose with 106 colony forming units (CFU) of each of 60 deﬁned non- pathogenic strains of E. coli 15 min prior to the challenge with pathogenic strains of E. coli (Table 1). den), as well as neonatal infections with E.coli in the offspring (Piliguard vet, Scanvet, Fre- densborg, Denmark). No haemolytic strains of E. coli were found in the faeces from any of the 30 piglets one week before weaning. When initiating the trial, the groups were housed in separated rooms at the National Vet- erinary Institute (NVI) with separated urine and manure handling. The rooms were free from draught, illuminated for 14 h per day and kept at 20°C. To prevent spread of E. coli (including probiotic strains) to previously not exposed pigs, the groups were always visited in alpha- betical order. Boots and tools were designated to and kept within each room. The present study included 30 piglets, repre- senting 6 litters (1 to 6) designated to 5 experi- mental groups (A to E) with 6 pigs at weaning. Each group included one pig from each litter of origin. Each animal was given a group, litter identiﬁcation, i.e. pigs with the same letter were group mates, and pigs with the same number were littermates. All groups were fed ad libitum through feeding automates (Piggomat, Skälby Maskin, Enkö- ping, Sweden). Group A was left as an unin- fected control group and offered a preheated standard feed (Startgris Fiber, Lantmännen, Svalöv, Sweden). The other groups were ex- posed to 3 pathogenic strains of E. coli as de- scribed below. Group D was left as an infected control group, while group B was offered a feed with 2,500 ppm ZnO and group C was offered a non-heated meal feed with lactose and ﬁbres (produced by Nibble, Tillberga, Sweden). Group E was also offered the standard feed, but Acta vet. scand. vol. 43 no. 4, 2002 Inducement of post weaning diarrhoea nducement of post weaning dia hoea PWD was induced as brieﬂy described below with a model earlier used (Melin et al. 2000a&b). At the day of weaning (living day 35) the animals were transported for 1 h in a joint closed horse trailer to the NVI. All but the control pigs were exposed to pathogenic strains of E. coli via the environment. One h before the arrival of the animals a broth with a pathogenic strain of E. coli (O147; K89, STb) was spread Acta vet. scand. vol. 43 no. 4, 2002 234 L. Melin & P. Wallgren Table 2. Results obtained from 6 control animals and 24 piglets exposed to three pathogenic strains of E. coli at weaning on living day 35. The different prophylactic regimes used in the in the study are described in Table 1. Demonstration of rotavirus and/or the challenge strains of E. coli in faeces are shown on daily bases. Days with diarrhoea are shaded. Inducement of post weaning diarrhoea F4 Sampling day (Day 0 = day of weaning) Piglet Receptor 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 A:1 Neg A:2 Neg A:3 Pos R A:4 Pos A:5 Pos A:6 Pos B:1 Pos 12 7 9 9 9 9,1 1,9,7 1 1 1,7 7,1,9 1 1 B:2 Neg 7 7 7 1,9 1,7 1,9,7 1 1 1,7 1 1 1 B:3 Pos R 7, R 7, R 7, R 9,1,7 7,9 1 9,1,7 1,9 1 1,7 1 1 B:4 Pos R 7, R 7, R 7, R 7,1 7,1 9,7 7,1,9 7,1 1,7 1,7 1 1 1 B:5 Pos 9 1 1 1 1 1,7,9 1 1 1,7,9 7 B:6 Pos 7 7 9 1,7,9 7,1 1,7 1,7,9 1,7 1 1 1 1 R C:1 Neg 7 7 1,7,9 1,7 1,7 7 7,1 1,7 7 7,1 7 7 7 C:2 Neg 1,7,9 1,7,9 1 1 1 1 1,7 7,1 7 7,1 7 C:3 Pos R R R 7 9 9 9 9 9 1,9 7,1 1 7,9 7 R C:4 Pos 7 7 7 1,7,9 1,7 1,7 1,7 9,1 1,7 1 1,7 1,7 C:5 Pos 7 7 1,9 1,9 9,1 1,7 1,7,9 7,1,9 1,7 7 7,9 9,7 C:6 Pos 7 7 7 1,7 1,7 1 1 1 1,7 7,1 7 7 7 Dead D:1 Pos 7 9 9 9,1 9,7,1 7 7,1,9 7 7 7 D:2 Neg 7 7 9 1 1 1 1 1,7 1 7,1 7 7 D:3 Pos 7 9,1 1,9 1,7,9 1,7 1,9 1,7 1,7,9 7,1 7 1,7 D:4 Pos R R 7, R 9,1 7 1 1 1,7 7,1 D:5 Neg 7 7 7 1 1 1 1 1 1 1 9 9 9 D:6 Pos R 7 7 7 1,9,7 7,9 -------------Piglet (D6) dead from day 6 post weaning ------------ E:1 Neg 7 7 1,9 9,1 9,1 1,9 1 1 7,1 9 9 E:2 Neg 1 1,9 1 1 1 1 7,1 E:3 Pos 7 1 1,9 1,9 1 1 1 9 E:4 Pos R R 9,1 9 9 9 1 1 1 9 1 1 E:5 Neg 1,9 1,9 1 1 1 1 1,7,9 9 9 9 9 E:6 Pos 1 1 1 1,9 1 1 1 1 9,7 Diarrhoea: light grey = “diarrhoea”; dark grey = “watery diarrhoea”. 1, 7 and 9 = presence of E. coli O141, O147 and O149 respectively. Inducement of post weaning diarrhoea The most frequent serotype is given ﬁrst. If bolded the challenge strains comprised more than 25 % of the total coliform ﬂora. R = presence of rotavirus. Diarrhoea: light grey = “diarrhoea”; dark grey = “watery diarrhoea”. 1, 7 and 9 = presence of E. coli O141, O147 and O149 respectively. The most frequent serotype is given ﬁrst. If bolded the challenge strains comprised more than 25 % of the total coliform ﬂora. R = presence of rotavirus. to a density of 2 x 106 CFU per square meter on the ﬂoor of empty and previously disinfected pens. In the pen for the control group a sterile BHI-broth was used. One h after the arrival of the piglets, the pens were bedded with sawdust and the animals were given access to feed and water. Three days post weaning the animals were exposed a second time in the same way with a broth comprising both E. coli O141 (K85, STb, VT2) and E. coli O149 (K91, K88, STa, STb, LT). The trial was terminated 14 days post weaning by sacriﬁcing the animals. At that time the in- testinal epithelium of all piglets was tested for presence of receptors to the adhesion factor F4/K88 post mortem (Edfors-Lilja et al. 1995). Feed related to post weaning diarrhoea 235 All 3 E. coli strains used were previously tested positive for toxins using PCR-technique (Melin et al. 2000a&b), and by loop tests (Smith & Halls 1967) they were proven pathogenic (Melin et al. 2000a&b). from the pen ﬂoor (3×10 g per pen). All micro- bial analyses were initiated within 2 h after sampling with the exception of detection of ro- tavirus. These samples were stored at -20°C un- til analysed all at one single occasion. After the termination of the study the entire Ileum from all animals were stored in -80°C for analysis of Lawsonia intracellularis by PCR. Health status The health status of the animals was inspected at least 3 times per day, with special attention to faecal consistency. If the consistency allowed a collected sample to adapt to the shape of any container it was characterised as "diarrhoea". A watery consistence was denoted "watery diar- rhoea". These terms are separated in table 2, but the common term "diarrhoea" is used for both types of loose stool in the text. The results are expressed as number of pigs with diarrhoea (at any time) per group, and as number of pig days with diarrhoea per group. The latter corre- sponds to the sum of all days with diarrhoea per pig within group, and is also compared to total number of pig days at risk. Detection of the challenge strains Faecal samples were spread on blood agar plates (blood agar base No. 2; LabM, Salford, England + 5% horse blood) and incubated for 18 h at 37°C. No haemolytic strains of E. coli were determined prior to weaning why occur- rence of ß-haemolytic E. coli were denoted as potential isolates of the challenge strains. They were estimated as percentage of the total num- ber of coliforms, and tested for presence of cap- sule antigen (n = 5 per pig and day) by aggluti- nation with rabbit serum (Söderlind 1971). If positive (K85 = O141; K89 = O147; K91 = O149), they were considered as a reisolated challenge strain. Sampling procedures Sampling procedures Rectal samples for microbial analyses were col- lected daily at 9 a.m. with cotton swabs and transported to the laboratory in Aimes transport medium (Copan Italia, Brescia, Italy). The presence of Brachyspira spp was investigated in all animals at weaning and on day 7 post wean- ing. The occurrence of Isospora suis was anal- ysed on day 5 post weaning in faeces collected Acta vet. scand. vol. 43 no. 4, 2002 Daily weight gain, feed intake and feed conver- sion Detection of other pathogenic microorganisms The presence of Brachyspira spp was investi- gated by culturing on Fastidious Anaerobe agar, (LabM LAB 90, Salford, England) for 6 days under anaerobic conditions at 37°C (Fellström et al. 1995). Rotavirus was detected by an ELISA demonstrating group A rotavirus anti- gen in faecal samples (de Verdier Klingenberg & Esfandiari 1996). Isospora suis was analysed using a modiﬁed version of a ﬂotation/McMas- ter technique (Thienpont et al. 1979) used in routine diagnostics at NVI. The presence of Lawsonia intracellularis was analysed by PCR (Jacobson et al. 2000). From 7 days before weaning the piglets were weighed once a week on an electronic scale (Epescale 1045, Alfa-Laval, Södertälje, Swe- den) and the daily weight gains (DWG) were calculated as gram gained per day. The weight of given, as well as consumed, creep feed was noted. Feed conversion ratios were calculated as kg feed consumed per kg weight gained. Biochemical ﬁngerprinting Faecal samples were spread on MacConkey agar (Oxoid, Basingstoke, Hampshire, Eng- Acta vet. scand. vol. 43 no. 4, 2002 L. Melin & P. Wallgren 236 land) and incubated for 18 h at 37°C. From each sample 24 colonies of coliforms were picked randomly and inoculated on PhP-RE plates (Pheneplates®, PhPlate AB, Stockholm, Sweden). Each isolate was spread to 11 differ- ent substrates on a microtiter plate and the ab- sorption values (A650) were measured with a photometer (Titertek Multiscan MCC/340, Labsystems OY, Helsinki, Finland) after 16, 40 and 64 h of incubation at 37°C. The ability to utilise the various substrates was compared and isolates showing similarity coefﬁcients higher than 97.5 were regarded as identical (Kühn 1985) and assigned to the same biochemical phenotype (BPT). or litters, respectively, was calculated with the Mann-Whitney U test. The signiﬁcance of dif- ferences within groups or litters over time was determined by the Wilcoxon signed-rank test. The signiﬁcance of differences regarding clini- cal signs between groups or litters, respectively, was calculated by χ2-tests. Reisolation of challenge strains f g None of the pathogenic strains of E. coli used for challenge was found in any faecal sample collected prior to the study, or in any faecal sample collected from the control pigs. In con- trast, pathogenic E. coli challenge strains were frequently isolated from all exposed pigs (Table 2). The distribution in faeces of these 3 chal- lenge strains differed between experimental groups (Table 2). In group E, given a mixture of non-pathogenic E. coli strains prior to chal- lenge, the proportion of E. coli O147 was lower (p<0.001) and the extent of E. coli O141 higher (p<0.001-0.05), compared to all other groups. Also the litter of origin inﬂuenced this distribu- tion. In litter 2, where no animal expressed the F4 receptor in their jejunal epithelium (Tables 2 & 3a), the shedding of O149 was low compared to litters 3, 4 and 5 (p<0.001-0.05). On the con- trary, the proportion of O141 in litter 2 was high compared to litters 3 and 4 (p<0.01-0.05). The dominating serotype within each faecal sample varied from day to day, also among diarrhoeic pigs (Table 2). The phenotypic diversity of the coliform popu- lations was measured as Simpson's index of di- versity (Hunter & Gaston 1988). Diversity is high (maximum value of 1) for a population constituting many different and evenly dis- tributed BPTs and low (minimum value of 0) if one BPT is dominant. The mean diversities of the faecal coliform populations of each sam- pling occasion post weaning are presented as a percentage of each group mean value on the day of weaning. The ﬂoras of different piglets were compared using "Population Similarity" as described by Kühn et al. (1991). In this model the similarity, expressed as a SP- value, is high (maximum value of 1) if the 2 compared populations are identical and low (minimum value of 0) if they are totally different. Within each experimental group and sampling occasion all piglet ﬂoras were compared to each other giving a matrix of SP-values. From this matrix a mean SP-value for each group and sampling occasion was cal- culated. Further, within each group all isolates from each sampling occasion, i.e. days 3, 7, 10 and 14, were compared to the ﬂora at weaning. Acta vet. scand. vol. 43 no. 4, 2002 Reisolation of challenge strains Rotavirus, Brachyspira spp, Lawsonia intracel- lularis and Isospora suis As shown in Table 2, rotavirus was demon- strated on 20 sampling occasions in samples collected from 7 piglets representing all exper- imental groups. All except 2 of these samples were colleted during the ﬁrst 4 days post wean- ing. These rotavirus positive piglets did all orig- inate from litters 3, 4 and 5. Two additional Feed related to post weaning diarrhoea Feed related to post weaning diarrhoea Table 3a. Incidence of diarrhoea in one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning (I). One infected group was left as an infected control group, while the other three groups were given feed related prophylactics (for details see Table 1). Comparisons with the infected control group are hatched. Table 3a. Incidence of diarrhoea in one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning (I). One infected group was left as an infected control group, while the other three groups were given feed related prophylactics (for details see Table 1). Comparisons with the infected control group are hatched. g p The table also shows (II) the incidence of diarrhoea in exposed pigs (n = 24) with respect to litter of origin (1- 6). For both categories the presence of the F4-receptor in the intestine is given. Ratio Pigs Days F4 At risk With diarrhoea At risk With diarrhoea Signiﬁcance of difference Group / Litter Pos/Neg (n) (n) (%) (n) (n) (%) I) Group B C D E A: Uninfected control 4/2 6 0 0% 84 0 0% * * * * B: ZnO 5/1 6 5 83% 84 38 45% ** C: Meal feed 4/2 6 5 83% 83 30 36% * D: Infected control 4/2 6 6 100% 75 37 49% ** E: Probiotic 3/3 6 6 100% 84 20 24% II) Litter (exposed pigs) 2 3 4 5 6 1 2/2 4 4 100% 56 24 43% * 2 0/4 4 4 100% 56 13 23% * ** 3 4/0 4 4 100% 56 25 45% 4 4/0 4 4 100% 56 18 32% * 5 2/2 4 3 75% 56 20 36% 6 4/0 4 4 100% 46 25 51% Signiﬁcant differences: * = p < 0.05; = p< 0.01 and * = p < 0.001 (only shown at the row with the lowest group or litter number) piglets in litter 2 excreted rotavirus 7 days be- fore weaning, but not after weaning. Neither Brachyspira spp, Lawsonia Intracellularis nor Isospora suis were detected in any sample col- lected. challenge strains (Table 3a). Also the onset of clinical signs was slower and milder in this group, with 5 pig days of diarrhoea during the ﬁrst week post weaning (Table 2). Feed related to post weaning diarrhoea The corre- sponding ﬁgures were: 22 days in Group B (p<0.001) and 18 days in Groups C and D (p<0.01). Rotavirus, Brachyspira spp, Lawsonia intracel- lularis and Isospora suis Rotavirus, Brachyspira spp, Lawsonia intracel- lularis and Isospora suis Statistical analyses The signiﬁcance of differences between groups 237 Health status All control pigs remained healthy throughout the study. Diarrhoea was recorded in 5 to 6 piglets in each group exposed to pathogenic E. coli (Tables 2 and 3). As shown in Table 2, diarrhoea was seen in cor- relation to shed of rotavirus in 5 out of 7 ro- tavirus positive pigs (B3, B4, C3, D4 and E4). However, 4 of these 5 diarrhoeic piglets also shed E. coli O147. One animal in the control group (A3) excreted rotavirus on one occasion (at weaning), without any correlation to diar- rhoea. When presented as pig days with diarrhoea, all groups exposed to pathogenic E. coli showed a higher incidence of diarrhoea (p< 0.001) than the uninfected control group (Tables 3a & b). The incidence of diarrhoea in group E (given non-pathogenic strains of E. coli at weaning) was signiﬁcantly (p<0.05-0.01) lower when compared to all other groups exposed to the Another piglet (D6) was rotavirus positive at weaning, but did not show any clinical signs of diarrhoea at that time. E. coli O147 was demon- strated in faeces of that pig during day 1 to 3. Acta vet. scand. vol. 43 no. 4, 2002 L. Melin & P. Wallgren 238 Table 3b. Group wise incidence of diarrhoea when the litter with mortality (litter 6) were excluded from the animals presented in table 3a. Comparisons with the infected control group are hatched. Table 3b. Group wise incidence of diarrhoea when the litter with mortality (litter 6) were excluded from the animals presented in table 3a. Comparisons with the infected control group are hatched. Table 3b. Group wise incidence of diarrhoea when the litter with mortality (litter 6) were excluded from the animals presented in table 3a. Comparisons with the infected control group are hatched. Incidence of Days with Diarrhoea (%) Group and Signiﬁcance of differences between groups Day 1-7 Day 8-14 Whole period, Day 1-14 (%) B C D E (%) B C D E (%) B C D E A: Uninfected control (n=5) 0% * * *** * 0% * * * * 0% * * * * B: ZnO (n=5) 43% *** 26% * 34% p=0.06 C: Meal feed (n=5) 43% ** 23% * 33% * D: Infected control (n=5) 49% * 51% 50% E: Probiotic (n=5) 11% 43% 27% Signiﬁcant differences: * = p < 0.05; = p< 0.01 and * = p < 0.001. Acta vet. scand. vol. 43 no. 4, 2002 Health status (only shown at the row with the lowest group or litter number) = p< 0.01 and * = p < 0.001. (only shown at the row with the lowest group or litter number) From day 4 all 3 challenge strains were demon- strated and the pig developed diarrhoea. On day 6 post weaning it died in PWD. Another pig from the same litter (C6, offered the meal feed) also died due to PWD. This pig died on day 13 post weaning after having had PWD for 9 days (Table 2). From day 4 all 3 challenge strains were demon- strated and the pig developed diarrhoea. On day 6 post weaning it died in PWD. Another pig from the same litter (C6, offered the meal feed) also died due to PWD. This pig died on day 13 post weaning after having had PWD for 9 days (Table 2). was lower in litter 2 (lacking the F4-receptor) than in litters 1, 3 (p<0.05) and 6 (p<0.01; Table 3a). The pigs that died due to PWD originated from litter 6 and would presumably have con- tributed to a higher number of days with diar- rhoea in their groups if they had survived. In spite of this, litter 6 had the highest incidence of days with diarrhoea, 51% (Table 3a). This dif- ference between litter 6 and all other groups was most evident during the second week post weaning (p<0.05). When the results from litter All but one of the challenged pigs expressed di- arrhoea during the observation period). How- ever, when the results were stratiﬁed according to litter the number of pig days with diarrhoea Figure 1. The diversity of the faecal coliform ﬂora in one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning. The results are presented as mean diversity val- ues within group in relation to the mean diversity of that group at weaning. Figure 1. The diversity of the faecal coliform ﬂora in one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning. The results are presented as mean diversity val- ues within group in relation to the mean diversity of that group at weaning. Feed related to post weaning diarrhoea 239 Figure 2. Similarity (SP) within group between the individual faecal coliform populations at each sampling occasion. Biochemical ﬁngerprinting Biochemical ﬁngerprinting The mean diversities of the faecal coliform populations of each sampling occasion post weaning are presented as a percentage of the mean values obtained within group at the day of weaning (Fig. 1). The diversity of the faecal coliform population decreased in all piglets during the ﬁrst week following weaning. How- ever, the ﬂora was less affected in group C (given a meal feed with lactose and ﬁbres). During the second week post weaning the di- versity in Group C continued to decrease slightly, thereby reaching a similar level as groups A, B and D which in turn had regained an increased diversity of the coliform ﬂora dur- ing the second week post weaning. Group E (given non-pathogenic E. coli strains at wean- ing) developed a less diversiﬁed intestinal col- iform ﬂora during the ﬁrst week post weaning, and remained at that level during the second week. (Fig. 1). Health status The study comprises one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning. One infected group was left as an infected control group, while the other three groups were given feed related prophylactics (for details see Table 1). Figure 2. Similarity (SP) within group between the individual faecal coliform populations at each sampling occasion. The study comprises one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning. One infected group was left as an infected control group, while the other three groups were given feed related prophylactics (for details see Table 1). Figure 2. Similarity (SP) within group between the individual faecal coliform populations at each sampling occasion. The study comprises one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning. One infected group was left as an infected control group, while the other three groups were given feed related prophylactics (for details see Table 1). day) in the infected control group (D). No sig- niﬁcant differences in DWG between experi- mental groups or between litters were recorded. 6 were excluded (Table 3b) a lower incidence of diarrhoea (p<0,05) was revealed in groups B (ZnO) and C (meal feed) when compared to group D (infected control) during the second week post weaning. Acta vet. scand. vol. 43 no. 4, 2002 DWG and Feed consumption The highest DWG, both during the ﬁrst (154 ± 73 gram per day) and the second (354 ± 39 gram per day) week post weaning was recorded in the uninfected Group A. In the slowest grow- ing group (C) the corresponding ﬁgures were 29 ± 113 gram per day and 226 ± 213 gram per day respectively. The highest DWG among in- fected groups was seen in Group E, 136 ± 98 gram per day during the ﬁrst week post weaning and 314 ± 104 gram per day during the second week post weaning. The DWG reﬂected the feed consumption. Dur- ing the ﬁrst week post weaning the mean feed consumption ranged from 343 gram per pig and day in the group given meal feed (Group C) to 388 gram per pig in Groups A and E. During the second week post weaning the highest feed consumption was recorded in Group A (617 gram per day) and the lowest (524 gram per A comparison of the coliform ﬂoras between Acta vet. scand. vol. 43 no. 4, 2002 240 L. Melin & P. Wallgren Figure 3. Similarity (SP) between the total faecal coliform population of each group and sampling occasion compared to the total coliform population of that group at weaning. The study comprises one uninfected control group and four groups exposed to pathogenic serotypes of E. coli in connection with weaning. One infected group was left as an infected control group, while the other three groups were given feed related prophylactics (for details see Table 1). the members within each experimental group and sampling occasion revealed a mean SP- value of around 0.1 (range 0.08 to 0.11) at weaning and mixing (Fig. 2). In all groups ex- posed to pathogenic strains of E. coli the mean SP value increased to a maximum level (0.24 to 0.46) on day 7 post weaning, indicating a more homologous ﬂora within the groups at that time. Thereafter it decreased again to a level similar to that at weaning. This was obtained on day 14 post weaning (Fig. 2). In the uninfected group (Group A) the similarity within group was relatively constant over time. Acta vet. scand. vol. 43 no. 4, 2002 Discussion The use of non-pathogenic strains of a disease provoking bacteria, both as a prophylaxis and as treatment of infections, has earlier been proven useful (Winberg et al. 1993, Methner 1999, Barrow & Page 2000). Enteropathogenic strains of E. coli are dependent on receptor speciﬁc ad- hesions. Therefore, a competitive inhibition of these receptor sites by non-pathogenic strains of E. coli could decrease adhesion of pathogenic strains of E. coli and thereby pre- vent or reduce PWD. Group E was given a de- ﬁned mixture of 60 strains of non-pathogenic E. coli at weaning. Interestingly, the incidence of diarrhoea was signiﬁcantly lower in this group compared to all other challenged groups (Ta- bles 2 & 3b). This was most obvious during the ﬁrst week post weaning (Tables 2, 3a & 3b). The proportion of E. coli O147 was also lower than in the other infected groups. The diversity in the faecal coliform ﬂora de- creased in all groups, including the unexposed control group, during the ﬁrst week post wean- ing. This effect of weaning has previously been observed in healthy as well as in diarrhoeic pigs (Kühn et al. 1993, Katouli et al. 1995 & 1999, Melin et al. 1996, 2000a & 2001). In the pigs exposed to pathogenic strains of E. coli this dis- turbance may last long (Melin et al. 2000a), probably due to a continuous inﬂuence of those strains. In the present study the diversity was somewhat restored in all but one group at the end of the second week post weaning. Group E (that was given non-pathogenic strains of E. coli at weaning) conserved a low diversity among the coliform intestinal ﬂora throughout the study. This phenomenon may be associated to an initially good colonisation of at least some of the strains of E. coli given as probiotic at weaning as indicated by the increasing similar- ity (SP-values) between the faecal coliform populations of the pigs in this group during the ﬁrst week post weaning (Fig. 2). The decrease in similarity between the faecal populations of the pigs during the subsequent week may indi- cate that these probiotic strains fade away. Discussion Diarrhoea was detected in all groups exposed to pathogenic strains of E. coli but not in the con- trol group. This conﬁrmed earlier observations that the present combination of 3 pathogenic strains of E. coli can induce PWD (Melin et al. 2000b). Interestingly, and as found before (Nabuurs et al. 1993, Katouli et al. 1995, Melin et al. 2001), the dominating serotype within each animal varied from day to day (Table 2), indicating that a diarrhoeic pig does not neces- sarily excrete one single or a even dominant serotype of E. coli throughout a session of diar- rhoea. Also concurring earlier observations (Melin et al. 2001), a genetic predisposition to develop PWD was indicated as the number of days with diarrhoea varied signiﬁcantly be- tween litters, and as both pigs that died em- anated from the most affected litter (litter 6). In this context it was also notable that pigs from litter 2 (that lacked the F4-recepteor) were not fully protected against PWD. Still, this litter ex- pressed fewest pig days with diarrhoea, and E. The total coliform population for each group sampling (i.e. all isolates from all piglets within the group taken together) was compared to the total coliform population of that group at wean- ing. Overall a decreasing similarity to the ﬂora at weaning was seen with time (Fig. 3). This in- dicates development of an altered intestinal co- liform ﬂora following weaning. Feed related to post weaning diarrhoea 241 coli O149 (F4+) was only demonstrated occa- sionally during the ﬁrst 3 days post exposure to that strain (performed on day 3 post weaning; Tables 1 & 2). CFU collected per pig and sampling occasion for biochemical ﬁngerprinting; 144 CFU were analysed per group and day). Still, that might be of less importance as also the similarity be- tween the intestinal coliform populations be- tween pigs in the infected control group ceased, indicating a decreased inﬂuence of the chal- lenge strains with time. The negative impact of PWD was indicated by a higher DWG in the healthy control group. However, due to the low number of pigs and to large variations in initial weight, differences in weight gain and feed intake were not signiﬁ- cant. In spite of the fact that PWD was induced in all challenged groups, many of the parame- ters measured varied between different feeding regimes, as discussed below. Acta vet. scand. vol. 43 no. 4, 2002 Discussion The different feeding regimes and the use of good colonisers of non-pathogenic strains of E. coli appeared able to inﬂuence the intestinal co- liform microﬂora of piglets following weaning and thereby inﬂuence the severity of infections gained at that time. Heat-processing of the feed increases the pas- sage rate through the digestive channel, which in turn will reduce the effect of protective ele- ments such as digestive enzymes and hy- drochloric acid. This may result in a large num- ber of ingested microorganisms entering the intestine in a viable form and thereby increas- ing the risk for infections with pathogenic mi- croorganisms. A non-heated meal feed rich in dietary ﬁbres would instead prolong the gastric passage (Johansen & Bach Knudsen 1994). Group C was given such a feed (enriched with lactose) and the decline in diversity of the fae- cal coliform ﬂora was less pronounced in this group with no accentuated dip on day 7, which indicates that the ﬂora was less dominated by the challenge strains used (Fig. 1). Also the comparably low similarity between the col- iform ﬂoras of different piglets in this group at each sampling occasion indicates that the pigs managed to reduce the external inﬂuence on their intestinal ﬂora (Fig. 2). The incidence of diarrhoea was lower (p<0.05-0.01) than in the infected control group when the results were corrected for mortality (i.e. when litter 6 was excluded; Table 3b). Regarding the comparably low DWG in Group C (especially during the ﬁrst week post weaning), respect should be paid to the fact that the pigs were offered pellets dur- ing suckling and that they consumed fairly little of the non-heated meal feed. This feed also con- It is possible that the results obtained would have been different if for instance the probiotic strains of E. coli had been administered repeat- edly or well before weaning, as well as if the pigs offered the non-heated meal feed with lac- tose had been adapted to that food before wean- ing. However, the design of the study priori- tized genotype (litter of origin) and phenotype (conditions during the suckling period), factors of importance to the health status. Further it must be noted that we managed to standardise some important factors that could contribute to development of PWD, such as temperature, draught, and also had a low general pathogen load in this experiment. Discussion However, it may also actually indicate a good and evenly distributed colonisation of the pro- biotic strains, which ought to be further evalu- ated (the 60 probiotic strains outnumber the 24 Feed supplemented with 2500 ppm ZnO has previously been shown to prevent or decrease the severity of PWD (Holm 1990 & 1993, Poulsen 1995). Katouli et al. (1999) reported a less affected diversity of the coliform ﬂora post weaning and an improved DWG in healthy piglets given ZnO supplemented feed for 2 weeks post weaning. A high diversity is sug- gested to indicate a stabile ecosystem (Pielou 1975) and a stabile microﬂora is considered to have a higher colonisation resistance (Atlas 1984). In this study, however, the ZnO treated group did not differ from the infected control group with respect to diversity of the coliform ﬂora post weaning, nor with respect to pig days with diarrhoea. However, it ought to be remem- Acta vet. scand. vol. 43 no. 4, 2002 L. Melin & P. Wallgren 242 bered that the pig from the most affected litter of the infected control group already died on day 6 post weaning due to PWD. That death probably decreased the overall percentage of pig days with diarrhoea in the infected control group, since that pig probably would have ex- pressed diarrhoea for more days if it had sur- vived (Table 2). If the results from litter 6 were excluded (Table 3b) the incidence of diarrhoea was lower in Group B compared to the infected control (Group D) during the second week post weaning. tained a lower amount of protein than the stan- dard diet given to the other groups. This study conﬁrmed that the post weaning pe- riod is dangerous, and our results further sug- gest that the intestinal ﬂoras that develop post weaning with time will differ completely from those present before weaning (Fig. 3), as earlier proven by Katouli et al. (1995). However, many strains of E. coli could be regarded as intestinal transients (Kühn et al. 1993, Nabuurs et al. 1993, Katouli et al. 1995), which further points to the complex interaction between different microorganisms in the intestine. Efforts must therefore always be undertaken to facilitate the life for piglets at weaning. These measures in- clude a good environment, a low pathogen load, and the results obtained in this study indicate the importance of well-designed creep feeds. Acta vet. scand. vol. 43 no. 4, 2002 References References Atlas RM: Diversity of microbial communities. Adv. Microbial Ecol. 1984, 7, 1-47. Holm A: E. coli associated diarrhoea in weaner pigs: zinc oxide added to the feed as a preventive mea- sure. Proc. Int. Pig Vet. Soc., Lausanne, Switzer- land. 1990, 11, 154. Bailey M, Clarke CJ, Wilson AD, Williams NA, Stokes CR: Depressed potential for interleukin-2 pro- duction following early weaning of piglets. Vet. Immunol. Immunopathol. 1992, 34, 197-207. Holm A: Zinkkoncentration i vaev hos slagtesvin. Tillsaetning av zinkoxid till foder. (Zinc concen- tration in pigs at slaughter. Effect of creep feed- ing with high levels of zinc oxide). Dan. Vet. Tidsskr. 1993, 76, 10-11. (In Danish). Barrow PA, Page K: Inhibition of colonisation of the alimentary tract in young chickens with Campy- lobacter jejuni by pre-colonisation with strains of C. jejuni. FEMS Microbiol. Letters 2000, 182, 87-91. Hunter PR, Gaston MA: Numerical index of the dis- criminatory ability of typing systems: an applica- tion of Simpson's index of diversity. J. Clin. Mi- crobiol. 1988, 26, 2465-2466. Blecha F, Pollmann DS, Nichols DA: Weaning pigs at an early age decreases cellular immunity. J. Ani- mal Sci. 1983, 56, 396-400. Bolduan G: Acid-binding capacity of pig feeds. Muhle + Mischfuttertechnik 1992, 129, 87-89. Jacobson M, Fellström C, Heldtander M, Gunnars- son A: The prevalence of Lawsonia intracellularis in Swedish pigs submitted for autopsy. Proc. Int. Pig Vet. Soc. Congr. Melbourne, Australia. 2000, 16, 74. de Verdier Klingenberg K, Esfandiari J: Evaluation of a one-step test for rapid, in practice detection of rotavirus in farm animals. Vet. Rec. 1996, 138, 393-395. Jager LP, Zijlstra FJ, Hoogendoorn A, Nabuurs MJ: Enteropooling in piglets induced by soya-pep- tone mediated via an increased biosynthesis of prostanoids. Vet. Res.Commun. 1986, 10, 407- 412. Edfors-Lilja I, Gustafsson U, Duval-Iﬂah Y, Eller- gren H, Johansson M, Juneja RK, Marklund L, Andersson L: The porcine intestinal receptor for Escherichia coli K88ab, K88ac: regional local- ization on chromosome 13 and inﬂuence of IgG response to the K88 antigen. Anim. Genet. 1995, 26, 237-242. Jensen-Waern M, Melin L, Lindberg R, Johannisson A, Petersson L, Wallgren P: Dietary zinc oxide in weaned pigs-effects on performance, tissue con- centrations, morphology, neutrophil functions and faecal microﬂora. Res. Vet. Sci. 1998, 64, 225-231. Estrada Correa A, Taylor DJ: Enterotoxigenic Clo- stridium perfringens type A as a cause of diar- rhoea in weaned pigs. Proc. Int. Pig Vet. Soc. Discussion The efﬁcacy of the measures studied in practise must therefore be further evaluated, and the results obtained may well differ between different herds due to un- speciﬁc management errors. Feed related to post weaning diarrhoea 243 tein concentrate on productivity of early weaned pigs. VIII jornadas sobre prod. animal, 1999, 20, 448-450. Acknowledgement We thank Sigbrit Mattsson for the skilful work con- cerning all analyses regarding the coliform popula- tions, Helena Reineck-Bosaeus for the detection of rotavirus, Ulla Zimmerman for the detection of Brachyspira, Brittlouise Ljungström for the analyses regarding Isospora suis and Dr Inger Edfors-Lilja for the investigation concerning presence of F4 receptors on the jejunal epithelium. This study was ﬁnanced by grants from the Swedish Council of Forestry and Agricultural Research and from The Swedish Meat producing Farmers R&D Program. Graham H, Fadel JG, Newman CW, Newman RK: Ef- fect of pelleting and beta-glucanase supplemen- tation on the ileal and fecal digestibility of a bar- ley-based diet in the pig. J. Anim. Sci. 1989, 67, 1293-1298. Hampson DJ: Postweaning Escherichia coli Diar- rhoea in Pigs. in Escherichia coli in Domestic Animals and Humans, Ed: Gyles C. L., CABI Int., Walingford, 1994, 171-191. Hampson DJ, Hinton M, Kidder DE: Coliform num- bers in the stomach and small intestine of healthy pigs following weaning at three weeks of age. J. Comp. Pathol. 1985, 95, 353-362. References Rio de Janeiro, Brazil, 1988, 10, 138. Johansen HN, Bach Knudsen KE: Effects of reducing the starch content in oat-based diets with cellu- lose on jejunal ﬂow and absorption of glucose over an isolated loop of jejunum in pigs. B. J. Nutr. 1994, 72, 717-729. Fellström C, Petterson B, Uhlén M, Gunnarsson A, Johansson K-E: Phylogeny of Serpulina based on sequence analyses of the 16S rRNA gene and comparison with a scheme involving biochemical classiﬁcation. Res Vet Sci 1995, 59, 5-9. Katouli M, Lund A, Wallgren P, Kühn I, Söderlind O, Möllby R: Phenotypic Characterization of Intesti- Gracia MI, Medel P, Castellanos I, Mateos GG: Ef- fect of replacement of ﬁsh meal LT by a soya pro- Acta vet. scand. vol. 43 no. 4, 2002 L. Melin & P. Wallgren 244 nal Escherichia coli of Pigs during Suckling Post- weaning and Fattening Periods. Appl. Environ. Microbiol. 1995, 61, 778-783. Melin L, Katouli M, Lindberg A, Fossum C, Wallgren P: Weaning of piglets. Effects of an exposure to a pathogenic strain of Escherichia coli. J. Vet. Med. B 2000a, 47, 663-675. Katouli M, Melin L, Jensen-Waern M, Wallgren P, Möllby R: The effect of zinc oxide supplementa- tion on the stability of the intestinal ﬂora with special reference to composition of coliforms in weaned pigs. J. Appl. Microbiol. 1999, 87, 564- 573. Melin L., Mattsson S. Wallgren P.: Challenge with pathogenic strains of E. coli at weaning (I. Clini- cal signs and reisolation of challenge strains). Proc. Int. Pig Vet. Soc. Congr. Melbourne, Aus- tralia. 2000b, 16, 22. Klobasa F, Werhahn E, Butler JE: Regulation of hu- moral immunity in the piglet by immunoglobulin of maternal origin. Res. Vet. Sci. 1981, 31, 195- 206. Melin L.: Weaning of Pigs with Special Focus on the Intestinal Health. Thesis, Department of Large Animal Clinical Sciences, Swedish University of Agricultural Sciences, Uppsala, 2001. Kühn I: Biochemical ﬁngerprinting of Escherichia coli. a simple method for epidemiological inves- tigations. J. Microbiol. Methods 1985, 3, 159- 170. Methner U, Barrow PA, Berndt A, Steinbach G: Com- bination of vaccination and competitive exclu- sion to prevent Salmonella colonization in chick- ens: experimental studies. Int. J. Food Microbiol. 1999, 49, 1-2. Kühn I, Allestam G, Stenström TA, Möllby R: Bio- chemical ﬁngerprinting of water coliform bacte- ria, a new method for measuring phenotypic di- versity and for comparing different bacterial populations. Appl. Environ. Microbiol. 1991, 57, 3171-3177. References Nabuurs MJA: Thermostable factor(s) in soya pro- ducing a net excess of secretion in the ligated gut test in pigs. Vet. Res. Commun. 1986, 10, 399- 405. Nabuurs MJ, van Zijderveld FG, de Leeuw PW: Clin- ical and microbiological ﬁeld studies in The Netherlands of diarrhoea in pigs at weaning. Res. Vet. Sci. 1993, 55, 70-77. Kühn I, Katouli M, Lund A, Wallgren P, Möllby R: Phenotypic diversity and stability of the intestinal coliform ﬂora in piglets during the ﬁrst 3 months of age. Microbial Ecol. Health Disease 1993, 6, 101-107. Pielou EC: Ecological Diversity. Wiley Interscience, New York, 1975, 129-135. Poulsen HD: Zinc oxide for weanling piglets. Acta Agri. Scand. Section A, Animal Science 1995, 45, 159-167. Kyriakis SC, Tsiloyiannis VK, Vlemmas J, Sarris K, Tsinas AC, Alexopoulos C, Jansegers L: The ef- fect of probiotic LSP 122 on the control of post- weaning diarrhoea syndrome of piglets. Res. Vet. Sci. 1999, 67, 223-228. Puppe B, Tuchscherer M, Tuchscherer A, Vega Lopez MA, Bailey M, Telemo E, Stokes CR, Bosi P: The effect of housing conditions and social environ- ment immediately after weaning on the agonistic behaviour, neutrophil/lymphocyte ratio, and plasma glucose level in pigs Effect of early wean- ing on the development of immune cells in the pig small intestine Modulation of immune re- sponse and barrier function in the piglet gut by dietary means. Livestock Prod. Sci. 1997, 48, 157-164. Madec F, Bridoux N, Bounaix S, Cariolet R, Duval Iﬂah Y, Hampson DJ, Jestin A: Experimental models of porcine post-weaning colibacillosis and their relationship to post-weaning diarrhoea and digestive disorders as encountered in the ﬁeld. Vet. Microbiol. 2000, 72, 295-310. Madec F, Bridoux N, Bounaix S, Jestin A: Measure- ment of digestive disorders in the piglet at wean- ing and related risk factors. Prev. Vet. Med. 1998, 35, 53-72. Rantzer D: Weaning of Pigs. Exocrine pancreas se- cretion and the inﬂuence of preweaning housing and postweaning strategic feeding. Thesis, De- partment of Agricultural Biosystems and 0Tech- nology, Swedish University of Agricultural Sci- ences, Alnarp, 1997 Melin L, Jensen-Waern M, Johannisson A, Ederoth M, Katouli M, Wallgren P: Development of Se- lected Faecal Microﬂoras and of Phagocytic and Killing Capacity of Neutrophils in Young Pigs. Vet. Microbiol. 1997, 54, 287-300. Saif LJ, Rosen BI, Parwani AV: Animal Rotaviruses in Viral Infections of the Gastrointestinal Tract, Marcel Dekker, Inc., 1994, 335-367. Sammanfattning Avvänjningsdiarré hos gris - effekter av foderrelate- rade åtgärder. Avvänjningsdiarré hos gris - effekter av foderrelate- rade åtgärder. Smith HW, Jones JET: Observations on the alimen- tary tract and its bacterial ﬂora in healthy and dis- eased pigs. J. Pathol. Bacteriol. 1963, 86, 387- 412. Möjligheten att motverka avvänjningsdiarré genom olika foderrelaterade profylaktiska åtgärder under- söktes i en "split litter studie" som varade i 2 veckor och omfattande 5 grupper om vardera sex grisar från sex olika kullar. I 4 av grupperna exponerades gri- sarna i samband med avvänjningen för tre patogena stammar av E. coli. I 3 av dessa grupper undersöktes den profylaktiska effekten av: 2500 ppm ZnO i ett pelleterat foder, ett laktosberikat ﬁberrikt mjölfoder respektive en enstaka peroral giva av 60 stammar av apatogena E. coli som probiotika. Två kontrollgrup- per, en infekterad och en oinfekterad, fodrades med pelleterat standardfoder. Söderlind O: Studies on Escherichia coli in pigs. II. Serological investigations. Zentralbl. Veterin- armed [B] 1971, 18, 569-590. Spencer BT, Howell PG, Hillman K, Murdoch TA, Spencer RJ, Stewart CS: Some husbandry factors inﬂuencing weaning stresses in piglets. J. S. Afr. Vet. Assoc. 1989, 60, 62-64. Spencer BT, Tilley TJ, Poomvises P, Ingkaninun P: Low acid binding feed for weaner piglets. Proc. Int. Pig Vet. Soc. Congr. Bangkok, Thailand, 1994, 13, 279. Diarré påvisades i samtliga grupper som infekterats med patogena E. coli däremot inte i den oinfekterade kontrollgruppen. Den inom enskilda individer domi- nerande serotypen av faekala E. coli varierade från dag till dag även hos grisar med diarré. Detta indike- rar att diarrén inte inducerades av en enskild serotyp. Den infekterade kontrollgruppen företedde de gra- vaste symptomen på diarré emedan gruppen som gavs apatogena E. coli som probiotika uppvisade sig- niﬁkant lägre förekomst av diarré jämfört med samt- liga andra grupper. Diarréfrekvensen varierade även med avseende på kulltillhörighet. Thienpont D, Rochette F, van Parijs OFJ: Diagnosing Helmenthiasis through corpological examina- tion. Jansen Research Foundation, Beerse, Bel- gium, 1979, 40-41. Underdahl NR: The effect of feeding Streptococcus faecium upon Escherichia coli induced diarrhea in gnotobiotic pigs. Prog. Food Nutr. Sci. 1983, 7, 5-12. Wathes CM, Miller BG, Bourne FJ: Cold stress and post-weaning diarrhea in piglets inoculated orally or by aerosol. Anim. Prod. 1989, 49, 483- 496. Under första veckan efter avvänjning sjönk diversite- ten i den fekala koliforma populationen i alla grupper samtidigt som likheten mellan olika grisars ﬂoror inom de infekterade grupperna ökade. Reprints may be obtained from: Lennart Melin, Department of Ruminant and Porcine Diseases, National Ve- terinary Institute, S-751 89 Uppsala, Sweden. E-mail: Lennart.Melin@sva.se, tel: +46 18 674000, fax: +46 18 309162. References Melin L, Katouli M, Jensen-Waern M, Wallgren P: The inﬂuence of zincoxide of the intestinal mi- croﬂora of piglets at weaning. Proc. Int. Pig Vet. Soc. Congr. Bologna, Italy. 1996, 14, 465. Smith HW, Halls S: Observations by the ligated in- Acta vet. scand. vol. 43 no. 4, 2002 Feed related to post weaning diarrhoea 245 testinal segment and oral inoculation methods on Escherichia coli infections in pigs, calves, lambs and rabbits. J. Pathol. Bacteriol. 1967, 93, 499- 529. (Received September 19, 2001; accepted August 12, 2002). Acta vet. scand. vol. 43 no. 4, 2002 Sammanfattning Detta indike- rade en påverkan av infektionsstammarna. Denna på- verkan minskade under försökets andra vecka. Grup- pen som fodrades med ett laktosberikat ﬁberrikt mjölfoder uppvisade de minsta förändringarna inom den faekala koliforma ﬂoran. Sammanfattningsvis verkar de foderrelaterade profylaktiska åtgärder som testats i denna studie ha potential att minska före- komsten av, eller mildra de kliniska symptom vid, av- vänjningsdiarré. Wattrang E, Wallgren P, Lindberg Å, Fossum C: Signs of infections and reduced immune functions at weaning of conventionally reared and speciﬁc pathogen free pigs. Zentralbl Veterinarmed [B] 1998, 45, 7-17. Winberg J, Herthelius-Elman M, Möllby R, Nord CE: Pathogenesis of urinary tract infection-experi- mental studies of vaginal resistance to coloniza- tion. Pediatr. Nephrol. 1993, 7, 509-514. (Received September 19, 2001; accepted August 12, 2002). (Received September 19, 2001; accepted August 12, 2002). Reprints may be obtained from: Lennart Melin, Department of Ruminant and Porcine Diseases, National Ve- terinary Institute, S-751 89 Uppsala, Sweden. E-mail: Lennart.Melin@sva.se, tel: +46 18 674000, fax: +46 18 309162. Acta vet. scand. vol. 43 no. 4, 2002
https://openalex.org/W2808176413	https://europepmc.org/articles/pmc5997338?pdf=render	English	null	Probabilistic motor sequence learning in a virtual reality serial reaction time task	PloS one	2,018	cc-by	4,962	RESEARCH ARTICLE Florian Sense1,2, Hedderik van Rijn1,2 Florian Sense1,2, Hedderik van Rijn1,2 1 Department of Experimental Psychology, University of Groningen, Groningen, The Netherlands, 2 Behavioral and Cognitive Neuroscience, University of Groningen, Groningen, The Netherlands * f.sense@rug.nl a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Abstract The serial reaction time task is widely used to study learning and memory. The task is tradi- tionally administered by showing target positions on a computer screen and collecting responses using a button box or keyboard. By comparing response times to random or sequenced items or by using different transition probabilities, various forms of learning can be studied. However, this traditional laboratory setting limits the number of possible experi- mental manipulations. Here, we present a virtual reality version of the serial reaction time task and show that learning effects emerge as expected despite the novel way in which responses are collected. We also show that response times are distributed as expected. The current experiment was conducted in a blank virtual reality room to verify these basic principles. For future applications, the technology can be used to modify the virtual reality environment in any conceivable way, permitting a wide range of previously impossible experimental manipulations. Editor: Jane Elizabeth Aspell, Anglia Ruskin University, UNITED KINGDOM Editor: Jane Elizabeth Aspell, Anglia Ruskin University, UNITED KINGDOM Received: November 18, 2017 Accepted: May 24, 2018 Published: June 12, 2018 Copyright: © 2018 Sense, van Rijn. 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. Editor: Jane Elizabeth Aspell, Anglia Ruskin University, UNITED KINGDOM Received: November 18, 2017 Accepted: May 24, 2018 Published: June 12, 2018 Copyright: © 2018 Sense, van Rijn. 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. OPEN ACCESS Citation: Sense F, van Rijn H (2018) Probabilistic motor sequence learning in a virtual reality serial reaction time task. PLoS ONE 13(6): e0198759. https //doi org/10 1371/journal pone 0198759 Virtual reality serial reaction time task performance in sequenced blocks, the target position on any given trial has a probabilistic dependency on the target position of the previous trial. Specifically, there are two sequences, one with a high and one with a low probability. The two sequences used in Schvaneveldt and Gomez [4] are A = (1, 2, 4, 3) and B = (1, 3, 4, 2) and are recycled sequentially to generate stim- ulus sequences for the experiment. If A is the high-probability sequence for a given participant, for example, a target in position 1 (on screen) will be followed by either a target in position 2 (selected from sequence A) or 3 (from sequence B), with a high and low probability, respec- tively. Which sequence a target is drawn from on any given trial is determined by a weighted coin flip. In each case, every other target position is independent of whether A or B is the high- probability sequence (i.e., 1 and 4 are at the same location in sequence A and B). Since both A and B contain all target positions equally often, targets will, on average, appear equally often at each position. Thus, it is not that participants learn that one position is generally more fre- quent but that certain positions are more likely to be the target than another given the previous target. Hence, participants learn to anticipate certain probabilistic transitions rather than a fixed sequence. and analysis, decision to publish, or preparation of the manuscript. and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. In this context, learning of the probabilistic transitions can be assessed by contrasting responses on trials corresponding to the probable and improbable sequences. Manipulating the conditional probabilities of transitions has several advantages: Error rates become informa- tive, especially on improbable transitions, because most incorrect responses correspond to the probable target, which indicates learning and anticipation of the probable sequence [4]. Fur- thermore, the learning of the probabilistic sequence can be more easily distinguished from overall on-task speed-ups in RTs. One would expect participants to get faster as a function of the number of completed trials even if there is no embedded sequence. Introduction Nissen and Bullemer [1] introduced the serial reaction time task (SRTT) to study differences between introspective and performance measures of learning. Since their introduction of the task, it has been used widely as a way to “explore the processes underlying a broad range of behaviors, including the cognitive and biological principles of learning and memory” ([2], p. 10073). In the SRTT, learning is operationalized as a speed-up in response times (RTs) to a sequence of stimuli. Specifically, four target positions are shown horizontally on screen and are mapped onto four buttons on the keyboard. The participant simply presses the corre- sponding key as quickly as possible when a target lights up. A baseline for RTs can be obtained by starting the task with a block in which the targets light up randomly (excluding repetitions of the same position). Then, a repeating sequence of target positions is introduced. By using relatively long sequences (e.g., 10 items as in [1]), participants will remain unaware of a repeat- ing pattern even though their RTs decrease (for an explicit learning variation, see, e.g., [3]). Learning measures are then derived by comparing RTs on random and sequenced blocks [2]. Roughly a decade after its introduction, Schvaneveldt and Gomez [4] proposed a probabi- Data Availability Statement: The raw data, all scripts to produce statistics and analyses in the manuscript as well as the figure, and supplemental material are available at https://osf.io/exnvd/. Funding: Sense was supported by SNN VIA grant "VIRTu CPR VI16SN006" financed by the European Regional Development Fund; Van Rijn was partially supported by the research program "Interval Timing in the Real World: A functional, Funding: Sense was supported by SNN VIA grant "VIRTu CPR VI16SN006" financed by the European Regional Development Fund; Van Rijn was partially supported by the research program "Interval Timing in the Real World: A functional, computational and neuroscience approach", project number 453-16-005, financed by the Netherlands Organisation for Scientific Research (NWO). The funders had no role in study design, data collection Roughly a decade after its introduction, Schvaneveldt and Gomez [4] proposed a probabi- listic version of the SRTT. Instead of comparing performance in un-sequenced blocks with PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 1 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 If each trial is a proba- ble or improbable transition, one can scrutinize the differences in RTs (or error rates) between each transition type as an interaction with trial number. Due to its simplicity, the SRTT is a popular lab task used to study a wide range of phenom- ena related to learning and memory [2]. With the advance of new technology comes the opportunity to expand the set of possible experiments to run and manipulations to implement. The advent of wearable virtual reality (VR) headsets as well as the increased accessibility and usability of software to create VR environments is particularly exciting in this regard [5,6]. Rizzo and Koenig [7] surveyed the development of VR technology for clinical applications and concluded that VR is ready for primetime and has similar potential for many areas of psychol- ogy. VR also has great potential for the neuroscientific study of social processes because it can provide stimulus material that more closely resembles real world activities and interactions, and Parsons, Gaggioli, and Riva [8] argue that VR allows experimenters to maintain control while elevating ecological validity. Neguţ, Matu, Sava, and David [9] present a meta-analytic review of task difficulty when neuropsychological tests are administered either in VR or as classical pen-and-paper or computerized versions. They conclude that cognitive performance is poorer in VR, likely due to higher task complexity, which might consume additional cogni- tive resources. In another meta-analysis of VR measures of neuropsychological assessment, however, they show that VR measures have the necessary sensitivity to detect cognitive impairment, and suggest that VR measures have potential for many neuropsychological assess- ment applications [10]. If a task or test is to be implemented in VR, however, certain aspects will have to be adapted to the new environment. For the SRTT, for example, the response options are traditionally pre- sented horizontally on the computer screen and associated with keys on the keyboard or a response box. A keyboard is usually not available in a VR environment and using the VR con- troller(s) is more natural than having participants hold a response box. However, VR control- lers require different muscle movements, especially as the selection of alternatives is typically PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 2 / 9 Virtual reality serial reaction time task performed by hand- or arm-movements to move a pointer in 3D-space. Methods A total of 29 participants completed the experiment. Of those, 19 were female and the average age across all participants was 20.5 (range: [18; 27], SD: 2.1). Participants were recruited from the participant pool of the University of Groningen and participated for course credit. All par- ticipants gave written informed consent and the study was approved by the Ethics Committee Psychology (ID: 16300-S-NE). All students in the participant pool were eligible for participa- tion and none were excluded based on their on-task performance. An experimental session started with the experimenter describing in detail what the participant was asked to do and a detailed explanation of the information on the informed consent forms, explaining in particu- lar that participants can stop at any moment (without any penalty) and that they should inform the experimenter in case they are uncomfortable at any point. None of the participants indi- cated any discomfort or withdrew from the study. The HTC Vive accommodates most pre- scription glasses and we did not specify that wearing glasses would exclude participants from participating. A handful of participants did wear their glasses during the experiment but did not report any discomfort and none of the glasses were too large to cause any issues. Although this is a very naturalistic response, which is immediately understood and easily performed by participants, the increased complexity of these movements, and increased noise levels associated with these movements, could potentially result in reduced signal-to-noise ratios. Here, we present data from an implementation of the probabilistic SRTT in VR, using a VR controller to provide responses, in an attempt to verify that expected learning effects can be reproduced in VR. Spe- cifically, we expected to replicate response patterns from traditional, two-dimensional imple- mentations of the SRTT: A general reduction of RTs over time that interacts with transition probability such that high-probability transitions become increasingly faster over time com- pared to low-probability transitions. PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 Procedure Shown is the arrangement of the four target positions while one target is lit up, indicating the participant should reach out and touch that target as quickly as possible. https://doi.org/10.1371/journal.pone.0198759.g001 trials each. Every block was followed by a self-paced break to minimize fatigue. Completing the task took between 4 and 8 minutes from the moment the practice part was completed. In the context of this project a number of distractors and their possible impact on RTs were tested. The distractors were different sounds and slight changes in the VR environment (e.g., flickering lights) and were found not to influence RTs at all. Therefore, the general learning effects reported in the Results section will be presented independently of the presence/absence of distractors. We refer the interested reader to the online supplement at https://osf.io/exnvd/, which includes a detailed description of the individual distractors, a number of annotated anal- yses of the possible effects of these distractors, information about the randomization procedure used for distractors, as well as the audio files used as distractors (see sub-section “Virtual Real- ity Serial Reaction Time Task (VR-SRTT)” in https://osf.io/vxyka/ as well as the sound files in task/stressors/ at https://osf.io/exnvd/). Procedure Participants received general instructions for the virtual reality serial reaction time task (VR-SRTT) and then sat in a chair, wearing a HTC Vive VR headset and a single hand-held controller (in their dominant hand). The four possible target positions were presented as gray spheres in the VR environment such that they constituted the four corners of an invisible square. Participants were instructed to position themselves (and the chair they sat in) such that they could reach all four targets easily, with minimal arm movement, from a position at the center of the invisible square. A screenshot of how the environment looked like to the partici- pant is shown in Fig 1. The two sequences from Schvaneveldt and Gomez [4] were used to generate probabilistic target sequences. Which of the two served as the high probability sequence for each participant was determined randomly, and high-transition probabilities occurred in 65% of trials (com- pared to 80% in the original study). On each trial, the target sphere’s color changed to blue until the participant used the controller to reach out to one of the spheres. As soon as the con- troller intersected with any sphere, a response was recorded. Each response had two compo- nents: accuracy (correct if target sphere was touched, incorrect otherwise) and RT (the time, in milliseconds, between the lighting up of the target sphere and the moment any sphere was touched). After a response was detected, visual (corrective) feedback was provided, and the 500 ms inter-stimulus interval commenced. First, participants completed 25 practice trials in which the order of the targets was entirely random (but target positions could not repeat). Next, they completed four blocks with 150 3 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 Virtual reality serial reaction time task Fig 1. Screenshot of the VR environment. Shown is the arrangement of the four target positions while one target is lit up, indicating the participant should reach out and touch that target as quickly as possible. https://doi.org/10.1371/journal.pone.0198759.g001 Fig 1. Screenshot of the VR environment. Shown is the arrangement of the four target positions while one target is lit up, indicating the participant should reach out and touch that target as quickly as possible. https://doi.org/10.1371/journal.pone.0198759.g001 Fig 1. Screenshot of the VR environment. PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 Results Performance during the 25 practice trials was near-perfect for all participants: Overall 98.8% of responses were correct and no-one made more than two errors. This indicates that the task was understood intuitively and immediately. The data from the practice trials were discarded. Performance during the task itself was also near-perfect: Incorrect responses only accounted for 0.6% of all trials. Responses were also rather fast: The median response time (RT) across all correct responses was 425 ms and 90% of the responses were between 302 and 603 ms, with only 0.5% of all correct trials resulting in RTs longer than one second. For all subsequent anal- yses, we removed incorrect trials and those with RTs longer than one second. This leaves data from 17,214 trials across 29 participants. We expected participants to respond faster on high-probability transitions than on low probability transitions and also expected a general speed-up as the task progresses. The aggre- gate data are presented in Fig 2 and follow the expected pattern. Bins with 30 trials each were created and means and within-subject standard errors [11] were computed for each bin. The RTs on the high-probability transitions are consistently faster than those on the low-probabil- ity transitions. Furthermore, RTs are lower for later blocks and the difference between the two PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 4 / 9 Virtual reality serial reaction time task Fig 2. Average response time across trials. Trials have been binned to emphasize the overall pattern. Error bars are within-subject standard errors [11]. Note that there are fewer trials in the low-probability transition condition, resulting in slightly wider error bars. https://doi.org/10.1371/journal.pone.0198759.g002 Fig 2. Average response time across trials. Trials have been binned to emphasize the overall pattern. Error bars are within-subject standard errors [11]. Note that there are fewer trials in the low-probability transition condition, resulting in slightly wider error bars. https://doi.org/10.1371/journal.pone.0198759.g002 https://doi.org/10.1371/journal.pone.0198759.g002 transition probability conditions increases for later blocks. The plot also highlights that differ- ences in RTs emerge early on. For the statistical analysis, a series of Bayesian linear mixed-effects regression models were fit to the RTs of each trial, using transition probability, trial number, and their interaction as predictors. To account for between-subject variance in RTs, random intercepts for participants were added. The models are shown in Table 1 along with the Bayes factors. PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 https://doi.org/10.1371/journal.pone.0198759.t001 Discussion The goal of the current study was to test whether probabilistic motor sequence learning effects observed in the traditional serial reaction time task (SRTT) can be reproduced in a virtual reality (VR) environment. The graphical overview in Fig 2 along with the statistical analysis confirm that we see the expected interaction effects: Participants respond faster to high-proba- bility transitions and their general speed-up over trials is more extreme for high-probability transitions. These findings are reassuring and open a new avenue for controlled experimental studies using a VR SRTT. In the current experiment, the VR environment was a featureless gray room. One of the advantages of VR, however, is that the environment can be manipulated in any conceivable way [6]. A series of studies have demonstrated the viability and usefulness of such an approach for the Stroop task: Parsons, Courtney, and Dawson [14] have shown that traditional Stroop effects can be reproduced in VR while varying the threat-level of the VR environment (also see [15]). Parsons and Barnett [16] showed that Stroop effects still emerge when the task is presented in a VR apartment environment and are comparable to pen-and- paper and computerized Stroop tasks. Similarly, Parsons and Carlew [17] had participants complete a Stroop task in a virtual classroom and showed that individuals with autism spec- trum disorder performed worse in the presence of distractors. This line of research sets prom- ising precedents and analogous manipulations–that are much more immersive than the featureless gray room used in the present study–could be implemented for the SRTT to test their potential effects on motor sequence learning. The current work could also be extended by recording additional measures while the task is performed. For example, Kachergis, Berends, de Kleijn, and Hommel [18] have presented analyses of mouse trajectories recorded during the performance of an SRTT. Their work could be extended into the third dimension by tracking the trajectories of the VR controller, reveal- ing anticipation and prediction of the learned probabilistic structure of the task. Furthermore, several commercial options are available to record eye-movements in a VR headset, providing yet another way to measure anticipation, prediction, and the deployment of attention during the task. An additional question is to which extent the responses collected in the VR environment are comparable to those collected in traditional lab settings. Results To ease interpreta- tion, the Bayes factors are shown relative to the worst model, revealing that, for example, the model including only transition probability as a main effect (model 2) is approximately one quadrillion times more likely to have generated the observed data than the model including only trial number as a predictor (model 1). The best-fitting model included all listed predictors (model 4) and is approximately 5 times more likely than the model including both main effects but no interaction (model 3; 1.161×1034 / 2.258×1033 = 5.14). See the supplement for the model fitting and selection procedure using Bayes factors (using the BayesFactor package, Table 1. Summary of the Bayesian linear mixed-effects regression model comparison results. Linear Mixed-Effects Models Bayes factors relative to 1. 1. Trial Number 1 2. Transition Probability 1.057×1015 3. Trial + Probability 2.258×1033 4. Trial + probability + Trial:Probability 1.161×1034 All four models include random effects for participants and Bayes factors are expressed relative to the worst-fitting model to ease interpretation. https://doi.org/10.1371/journal.pone.0198759.t001 Table 1. Summary of the Bayesian linear mixed-effects regression model comparison results. Linear Mixed-Effects Models Bayes factors relative to 1. 1. Trial Number 1 2. Transition Probability 1.057×1015 3. Trial + Probability 2.258×1033 4. Trial + probability + Trial:Probability 1.161×1034 All four models include random effects for participants and Bayes factors are expressed relative to the worst-fitting model to ease interpretation. https://doi.org/10.1371/journal.pone.0198759.t001 Table 1. Summary of the Bayesian linear mixed-effects regression model comparison results. 5 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 Virtual reality serial reaction time task [12]). Also included in the supplement are estimates of the model’s coefficients as well as a tra- ditional linear mixed-effects regression (using the lme4 package, [13]) for comparison. See sub-section “Traditional analysis using lme4” in the file “VR-SRTT_analyses.html” at https:// osf.io/exnvd/. The best-fitting model revealed by the Bayes model comparison confirms the pattern appar- ent in Fig 2: RTs on probable transitions are estimated to be faster than those on improbable transitions and each additional trial reduces the expected RT further. However, this reduction as a function of trial number interacts significantly with transition probability such that RTs decrease more for probable transitions than for improbable transitions as trials progress. PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 Discussion Table 2 presents RTs and error rates from the original SRTT studies by Nissen and Bullemer [1] and Schvaneveldt and Gomez [4] along with a number of more recent studies. For comparison, the overall mean RT across all 17,214 trials included in our analyses is 439ms, while it is 433ms and 450ms for the probable and improbable transition conditions, respectively, and these numbers are also listed in the table. Mean RTs from the present study seem to be in line with most of the means in Table 2. The RTs reported in the table come from a range of different populations and experimental setups so it is not surprising that there is some variance. Given that learning in the SRTT is 6 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 Virtual reality serial reaction time task Table 2. Comparison of reaction times across studies. Study Reaction Times (ms) Error Rates (%) Present study Overall mean: 439 Probable: 433 Improbable: 450 Overall: 0.603 Probable: 0.450 Improbable: 0.875 Nissen & Bullemer [1] Sequence: 216 Random: 346 Sequence: 3.250 Random: 4.625 Schvaneveldt & Gomez [4] Probable: 368 Improbable: 447 Probable: 3.954 Improbable: 10.190 Franklin, Smallwood, Zedelius, Broadway, & Schooler [19] Sequence: 420 Random: 436 Overall: 7.5 Du, Prashad, Schoenbrun, & Clark [20] Overall mean: 448 N/A Kraeutner, Gaughan, Eppler, & Boe [21] Implicit: 610 Random: 641 Implicit: 1.55 Random: 2.75 Guzma´n Muños [22] Overall mean: 391 Overall: 4.775 All reaction times are in milliseconds (ms) and all error rates are in percent (%). https://doi.org/10.1371/journal.pone.0198759.t002 All reaction times are in milliseconds (ms) and all error rates are in percent (%). All reaction times are in milliseconds (ms) and all error rates are in percent (%). All reaction times are in milliseconds (ms) and all error rates are in percent (%). https://doi.org/10.1371/journal.pone.0198759.t002 https://doi.org/10.1371/journal.pone.0198759.t002 conceptualized as a relative speed-up in RTs rather than their absolute values, however, makes it more important that our analyses confirm the presence of the expected pattern. It is reassur- ing, however, that the RTs collected in our experiment are not out of line with other studies, although they required a more complex motor response (moving of hand/wrist/arm rather than a button press). Our error rates, on the other hand, are lower than those reported in any other listed study. Discussion One would expect participants to make more errors on improbably transition trials (in the probabilistic paradigm) or random blocks (in the deterministic paradigm) and this pattern is observed both in our data as well as all other studies reported in Table 2. The error rates col- umn suggests that error rates vary greatly between studies but that none of them are as low as the ones reported here. The current study does not allow us to pinpoint the source of this dif- ference. We believe that the type of motor response that is required to give a response (pressing a button vs. reaching out to a sphere) would be a prime candidate for further investigation into this difference and that recording the response trajectories through 3D space would be a prom- ising way to illuminate this issue. To summarize, the experiment reported here is comparable to traditional versions of the (probabilistic) serial reaction time task despite having been implemented in a novel virtual reality environment. Learning effects emerge as expected and can be traced using the change in response time distributions as in the traditional setting. These results are reassuring and open the door to innovative experimental manipulations for an established experimental para- digm using state-of-the-art technology. Writing – review & editing: Florian Sense, Hedderik van Rijn. Writing – review & editing: Florian Sense, Hedderik van Rijn. Acknowledgments We would like to thank Charlotte Schlu¨ter for collecting the data and Stark Learning in Gro- ningen, The Netherlands for developing the virtual reality environment for this task. Sense was supported by SNN VIA grant "VIRTu CPR VI16SN006" financed by the European Regional Development Fund, Van Rijn was partially supported by the research program "Interval Timing in the Real World: A functional, computational and neuroscience approach", project number 453-16-005, financed by the Netherlands Organisation for Scientific Research (NWO). The funders had no role in study design, data collection and analysis, decision to pub- lish, or preparation of the manuscript. The authors have declared that no competing interests PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 7 / 9 Virtual reality serial reaction time task exist. The online supplement including the raw data, the analyses reported here, and additional analyses and visualizations can be found at https://osf.io/exnvd/. We would like to thank George Kachergis and two anonymous reviewers for their helpful feedback. exist. The online supplement including the raw data, the analyses reported here, and additional analyses and visualizations can be found at https://osf.io/exnvd/. We would like to thank George Kachergis and two anonymous reviewers for their helpful feedback. Author Contributions Conceptualization: Florian Sense, Hedderik van Rijn. Data curation: Florian Sense. Visualization: Florian Sense. Visualization: Florian Sense. Writing – original draft: Florian Sense. References 1. Nissen MJ, Bullemer P. Attentional requirements of learning: Evidence from performance measures. Cogn Psychol. 1987; 19: 1–32. https://doi.org/10.1016/0010-0285(87)90002-8 2. Robertson EM. The Serial Reaction Time Task: Implicit Motor Skill Learning? J Neurosci. 2007; 27: 10073–10075. https://doi.org/10.1523/JNEUROSCI.2747-07.2007 PMID: 17881512 3. Willingham DB, Salidis J, Gabrieli JDE. Direct comparison of neural systems mediating conscious and unconscious skill learning. J Neurophysiol. 2002; 88: 1451–60. https://doi.org/10.1152/jn.2002.88.3. 1451 PMID: 12205165 4. Schvaneveldt RW, Gomez RL. Attention and probabilistic sequence learning. Psychol Res. 1998; 61: 175–190. https://doi.org/10.1007/s004260050023 5. Bohil CJ, Alicea B, Biocca FA. Virtual reality in neuroscience research and therapy. Nat Rev Neurosci. Nature Publishing Group; 2011; 12. https://doi.org/10.1038/nrn3122 PMID: 22048061 6. Loomis JM, Blascovich JJ, Beall AC. Immersive virtual environment technology as a basic research tool in psychology. Behav Res Methods, Instruments, Comput. 1999; 31: 557–564. https://doi.org/10.3758/ BF03200735 7. Rizzo AS, Koenig ST. Is clinical virtual reality ready for primetime? Neuropsychology. 2017; 31: 877– 899. https://doi.org/10.1037/neu0000405 PMID: 29376669 8. Parsons TD, Gaggioli A, Riva G. Virtual reality for research in social neuroscience. Brain Sci. 2017; 7: 1–21. https://doi.org/10.3390/brainsci7040042 PMID: 28420150 9. Neguţ A, Matu SA, Sava FA, David D. Task difficulty of virtual reality-based assessment tools compared to classical paper-and-pencil or computerized measures: A meta-analytic approach. Comput Human Behav. 2016; 54: 414–424. https://doi.org/10.1016/j.chb.2015.08.029 10. Neguţ A, Matu SA, Sava FA, David D. Virtual reality measures in neuropsychological assessment: A meta-analytic review. Clin Neuropsychol. Routledge; 2016; 30: 165–184. https://doi.org/10.1080/ 13854046.2016.1144793 PMID: 26923937 11. Morey RD. Confidence Intervals from Normalized Data: A correction to Cousineau (2005). Tutor Quant Methods Psychol. 2008; 4: 61–64. 12. Morey RD, Rouder JN. BayesFactor 0.9.12–2 CRAN. 2015. 12. Morey RD, Rouder JN. BayesFactor 0.9.12–2 CRAN. 2015. 13. Bates D, Ma¨chler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Using lme4. J Stat Softw. 2015; 67: 1–48. https://doi.org/10.18637/jss.v067.i01 14. Parsons TD, Courtney CG, Dawson ME. Virtual reality Stroop task for assessment of supervisory atten- tional processing. J Clin Exp Neuropsychol. Routledge; 2013; 35: 812–826. https://doi.org/10.1080/ 13803395.2013.824556 PMID: 23961959 8 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 Virtual reality serial reaction time task 15. Parsons TD, Courtney C. Interactions between Threat and Executive Control in a Virtual Reality Stroop Task. IEEE Trans Affect Comput. 2016; 16. Parsons TD, Barnett M. Virtual Apartment-Based Stroop for assessing distractor inhibition in healthy aging. Appl Neuropsychol Adult. Taylor & Francis; 2017; https://doi.org/10.1080/23279095.2017. 1373281 PMID: 28976213 17. Parsons TD, Carlew AR. PLOS ONE \| https://doi.org/10.1371/journal.pone.0198759 June 12, 2018 References Bimodal Virtual Reality Stroop for Assessing Distractor Inhibition in Autism Spectrum Disorders. J Autism Dev Disord. Springer US; 2016; 46: 1255–1267. https://doi.org/10.1007/ s10803-015-2663-7 PMID: 26614084 18. Kachergis G, Berends F, Kleijn R De, Hommel B. Trajectory Effects in a Novel Serial Reaction Time Task. IEEE Conference on Development and Learning / EpiRob 2014. 2014. 19. Franklin MS, Smallwood J, Zedelius CM, Broadway JM, Schooler JW. Unaware yet reliant on attention: Experience sampling reveals that mind-wandering impedes implicit learning. Psychon Bull Rev. 2016; 23: 223–229. https://doi.org/10.3758/s13423-015-0885-5 PMID: 26122895 20. Du Y, Prashad S, Schoenbrun I, Clark JE. Probabilistic Motor Sequence Yields Greater Offline and Less Online Learning than Fixed Sequence. Front Hum Neurosci. 2016; 10: 1–11. 21. Kraeutner SN, Gaughan TC, Eppler SN, Boe SG. Motor imagery-based implicit sequence learning depends on the formation of stimulus-response associations. Acta Psychol (Amst). 2017; 178: 48–55. https://doi.org/10.1016/j.actpsy.2017.05.009 PMID: 28577488 22. Guzma´n Muñoz FJ. The influence of personality and working memory capacity on implicit learning. Q J Exp Psychol. 2018; https://doi.org/10.1177/1747021817749582 PMID: 29313740 9 / 9
https://openalex.org/W4283381827	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0270598&type=printable	English	null	Effects of supplemental oxygen on systemic and cerebral hemodynamics in experimental hypovolemia: Protocol for a randomized, double blinded crossover study	PloS one	2,022	cc-by	4,749	PLOS ONE PLOS ONE STUDY PROTOCOL Effects of supplemental oxygen on systemic and cerebral hemodynamics in experimental hypovolemia: Protocol for a randomized, double blinded crossover study Sole Lindvåg LieID1,2,3, Jonny Hisdal2,3, Marius RehnID1,4,5, Lars Øivind HøisethID1,6 1 Department of Research and Development, Norwegian Air Ambulance Foundation, Oslo, Norway, 2 Faculty of Medicine, University of Oslo, Oslo, Norway, 3 Section of Vascular Investigations, Oslo University Hospital, Oslo, Norway, 4 Division of Prehospital Services, Air Ambulance Department, Oslo University Hospital, Oslo, Norway, 5 Faculty of Health Sciences, University of Stavanger, Stavanger, Norway, 6 Division of Emergencies and Critical Care, Department of Anesthesiology, Oslo University Hospital, Oslo, Norway a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 sole.lindvaag.lie@norskluftambulanse.no * sole.lindvaag.lie@norskluftambulanse.no Editor: Quan Jiang, Henry Ford Health System, UNITED STATES Editor: Quan Jiang, Henry Ford Health System, UNITED STATES Received: January 14, 2022 Accepted: May 4, 2022 Published: June 24, 2022 Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0270598 OPEN ACCESS Supplemental oxygen is widely administered in trauma patients, often leading to hyperoxia. However, the clinical evidence for providing supplemental oxygen in all trauma patients is scarce, and hyperoxia has been found to increase mortality in some patient populations. Hypovolemia is a common finding in trauma patients, which affects many hemodynamic parameters, but little is known about how supplemental oxygen affects systemic and cere- bral hemodynamics during hypovolemia. We therefore plan to conduct an experimental, ran- domized, double blinded crossover study to investigate the effect of 100% oxygen compared to room air delivered by a face mask with reservoir on systemic and cerebral hemodynamics during simulated hypovolemia in the lower body negative pressure model in 15 healthy volunteers. We will measure cardiac output, stroke volume, blood pressure, mid- dle cerebral artery velocity and tolerance to hypovolemia continuously in all subjects at two visits to investigate whether oxygen affects the cardiovascular response to simulated hypo- volemia. The effect of oxygen on the outcome variables will be analyzed with mixed linear regression. The study is registered in the European Union Drug Regulating Authorities Clini- cal Trials Database (EudraCT, registration number 2021-003238-35). Citation: Lindvåg Lie S, Hisdal J, Rehn M, Høiseth LØ (2022) Effects of supplemental oxygen on systemic and cerebral hemodynamics in experimental hypovolemia: Protocol for a randomized, double blinded crossover study. PLoS ONE 17(6): e0270598. https://doi.org/10.1371/ journal.pone.0270598 * sole.lindvaag.lie@norskluftambulanse.no Introduction Copyright: © 2022 Lindvåg Lie 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. Supplemental oxygen is frequently administered in acutely and critically ill patients to avoid arterial hypoxemia and tissue hypoxia [1]. For trauma patients, this is stated in the ATLS (Advanced Trauma Life Support) guidelines: “Supplemental oxygen must be administered to all severely injured trauma patients” [2]. Accordingly, supplemental oxygen is often given to trauma patients, frequently resulting in hyperoxia [3]. However, the clinical evidence for pro- viding supplemental oxygen in all trauma patients is scarce [4] and the liberal use has been Data Availability Statement: No datasets were generated or analysed during the current study. All 1 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia largely founded on a presumption that supplemental oxygen is harmless. There is an increas- ing focus on possible deleterious effects of hyperoxia [1], and a recent retrospective cohort study on trauma patients receiving supplemental oxygen found higher mortality rates in patients with a higher SpO2 [5]. relevant data from this study will be made available upon study completion. Funding: This study is funded by the Norwegian Air Ambulance Foundation and Oslo University Hospital. The funders had and will not have a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. In the initial treatment of trauma patients, detection and treatment of hypovolemia is of paramount importance. The overriding goal for the resuscitation of these patients is to ensure adequate oxygen delivery to the vital organs, which is given by the product of cardiac output and arterial oxygen content. Hypovolemia leads to reduced cardiac filling, stroke volume and cardiac output [6]. Under normal circumstances in unanesthetized humans, this is compen- sated by an increase in systemic vascular resistance and heart rate to maintain a normal or near-normal mean arterial pressure (MAP). Normobaric hyperoxia induces vasoconstriction and reduced blood flow to several organs in normovolemic healthy volunteers, including the brain, heart and skeletal muscle [7,8]. Accordingly, hyperoxia may lead to an increased toler- ance to hypovolemia mediated by vasoconstriction and thereby maintained MAP as well as a potential increase in arterial oxygen content. Introduction However, hyperoxia may lead to reduced toler- ance to hypovolemia due to reduced cerebral blood flow. Competing interests: The authors have declared that no competing interests exist. Competing interests: The authors have declared that no competing interests exist. There is a lack of studies investigating the effect of supplemental oxygen on systemic hemo- dynamics during hypovolemia in a controlled, experimental setting. We therefore plan to con- duct an experimental, randomized, double blinded crossover study where healthy subjects will inhale 100% oxygen or room air administered on a face mask with reservoir during simulated hypovolemia in the lower body negative pressure (LBNP) model. LBNP is an experimental model of central hypovolemia where blood is redistributed from the upper to the lower body [9]. While the separated effects of hyperoxia and LBNP on healthy volunteers are described previously [7,9], the potential effects of hyperoxia on the hemodynamic response to LBNP need elucidation. The aim of the present study is therefore to investigate the effect of supple- mental oxygen on systemic and cerebral hemodynamics during LBNP. The primary hypothesis of this study is that supplemental oxygen will induce a different response in cardiac output compared to room air during LBNP. Secondary hypotheses are that supplemental oxygen induces different responses in stroke volume, middle cerebral artery velocity (MCAV) or time to decompensation during LBNP. Organization and conduct The study protocol is written according to the Norwegian Clinical Research Infrastructure Network (NorCRIN) guidelines and registered in the European Union Drug Regulating Authorities Clinical Trials Database (EudraCT, registration number 2021-003238-35). The Norwegian Medical Agency (21/15284-9) and the Regional Ethics Committee (REK South East D, ref. 285164) have assessed and approved the protocol. The original protocol is found in the supporting information file “S1 Protocol” and the spirit checklist in “S1 Checklist”. The sponsor of this trial is Oslo University Hospital, Norway. Experiments will be con- ducted at The Section of Vascular Investigations, Oslo University Hospital, Oslo, Norway. We will obtain written informed consent from all subjects before the start of the study. PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 Design In this single-center, experimental, randomized, double blinded, crossover trial we will study the effects of supplemental oxygen on systemic and cerebral hemodynamics during simulated hypovolemia in 15 healthy subjects. The schedule of enrolment, interventions, and assessments is shown in Fig 1, and study design is illustrated in Fig 2. All subjects will participate on two 2 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia Fig 1. SPIRIT schedule of enrolment, interventions, and assessments. https://doi.org/10.1371/journal.pone.0270598.g001 Fig 1. SPIRIT schedule of enrolment, interventions, and assessments. Fig 1. SPIRIT schedule of enrolment, interventions, and assessments. https://doi.org/10.1371/journal.pone.0270598.g001 https://doi.org/10.1371/journal.pone.0270598.g001 different visits, with at least one day between each visit. On both visits the subjects will be exposed to LBNP and inhale either 100% oxygen or room air, in a block-randomized order. Except from the inhalation gas, the experiments on Visit 1 and 2 are identical. different visits, with at least one day between each visit. On both visits the subjects will be exposed to LBNP and inhale either 100% oxygen or room air, in a block-randomized order. Except from the inhalation gas, the experiments on Visit 1 and 2 are identical. Prior to the start of the experiment on either visit, the subject will be familiarized to the set- up and rest for 20–30 minutes in the supine position to stabilize hemodynamic parameters before data sampling begins. After a baseline period, a 5 min run-in time for the inhalation gas will follow. The subjects will be exposed to stepwise LBNP starting at 0 mmHg with 10 mmHg increments every 3 minutes until reaching LBNP 80 mmHg or aborting the experiment (see Prior to the start of the experiment on either visit, the subject will be familiarized to the set- up and rest for 20–30 minutes in the supine position to stabilize hemodynamic parameters before data sampling begins. After a baseline period, a 5 min run-in time for the inhalation gas will follow. The subjects will be exposed to stepwise LBNP starting at 0 mmHg with 10 mmHg increments every 3 minutes until reaching LBNP 80 mmHg or aborting the experiment (see 3 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia Fig 2. Schematic illustration of the study design per visit. Eligibility criteria Subjects will be recruited according to the inclusion and exclusion criteria given in Table 2. In addition, subjects must abstain from caffeine containing products for 6 hours before each visit, nicotine containing products for 12 hours before each visit and strenuous exercise for 3 hours before each visit. Subjects are allowed to have a light meal on the day of the experiment before the experiment begins. Due to potential effects of circadian rhythm on the hemodynamic response to LBNP, we will to the extent possible conduct both visits at a similar time of the day for each subject. However, the evidence supporting the effect of circadian rhythm on the hemodynamic response to LBNP in the literature seems weak [13]. Randomization At enrolment, subjects will be randomly assigned (block randomization) in a 1:1 ratio to receive oxygen or room air on Visit 1, and the other on Visit 2. To get at balanced design, the subjects will be randomized with permuted blocks of size 4 or 6, using the “blockrand” package [10] in R [11] /Rstudio [12]. The randomization list will be automatically generated by the principal investigator as a.pdf-document and handed to a 3rd party who will prepare hosing for oxygen or room air administration and envelopes for emergency unblinding. Randomiza- tion lists will not be available to the investigators collecting data until after end of the study. Each subject will be dispensed blinded study intervention. Design The subject receives either 100% oxygen or room air as inhalation gas on Visit 1, and the other on Visit 2, throughout the entire experiment. Lower body negative pressure (LBNP) is increased stepwise with increments of 10 mmHg every 3 min from 0 mmHg until reaching 80 mmHg or aborting. Inh. gas = inhalation gas, MAP = mean arterial pressure, HR = heart rate. Fig 2. Schematic illustration of the study design per visit. The subject receives either 100% oxygen or room air as inhalation gas on Visit 1, and the other on Visit 2, throughout the entire experiment. Lower body negative pressure (LBNP) is increased stepwise with increments of 10 mmHg every 3 min from 0 mmHg until reaching 80 mmHg or aborting. Inh. gas = inhalation gas, MAP = mean arterial pressure, HR = heart rate. https://doi.org/10.1371/journal.pone.0270598.g002 Table 1). As all subjects receive both oxygen and room air, they will act as their own controls due to the crossover design. Both the subjects and the investigators will be blinded to the inha- lation gas. Table 1). As all subjects receive both oxygen and room air, they will act as their own controls due to the crossover design. Both the subjects and the investigators will be blinded to the inha- lation gas. Stop-criteria blood is displaced from the central compartment of the upper body to the lower extremities and pelvis. The subject is placed in the supine position in the LBNP chamber which is sealed just above the iliac crest. The model has been used for more than half a century and is consid- ered a safe and useful model for studying hypovolemia in conscious volunteers. Inhalation gas: Oxygen and room air. At each visit a subject will inhale either 100% oxy- gen or room air during the entire experiment as shown in Fig 2. The inhalation gas will be administered on a face mask with reservoir from a gas cylinder connected to a flow meter to ensure an output flow of 15 L/min. Administration of normobaric oxygen at 100% is not recommended for >6 h due to forma- tion of reactive oxygen species (ROS) [14] and their possible side-effects, primarily affecting the lungs. During the study, administration of 100% oxygen will in most subjects be limited to approximately 30 min, and never exceed 60 min. In essence, we are not aware of significant medical risks with the short-term use of oxygen in healthy adults. There are no absolute con- traindications to normobaric oxygen supplementation [14]. Table 2. Inclusion and exclusion criteria. Inclusion criteria Age 18 and < 50 years at the time of signing the consent Overtly healthy as determined by medical evaluation including medical history, heart and lung auscultation, focused cardiac ultrasound and measurement of cardiac conduction times Woman of childbearing potential (WOCBP) must 1) use adequate birth control or 2) have a negative pregnancy test less than 14 days before visit Capable of giving a signed informed consent Exclusion criteria Any medical condition limiting physical exertional capacity or requiring regular medication (allergy and contraceptives excepted) Pregnancy Breastfeeding History of syncope (syncope of presumed vasovagal nature with known precipitating factor excepted) Any known cardiac arrhythmia Any drug (contraceptives excepted) used on a regular basis for a chronic condition (allergy excepted) See “S1 Protocol” for specific requirements to adequate birth control. https://doi.org/10.1371/journal.pone.0270598.t002 Table 2. Inclusion and exclusion criteria. Stop-criteria Inclusion criteria Age 18 and < 50 years at the time of signing the consent Overtly healthy as determined by medical evaluation including medical history, heart and lung auscultation, focused cardiac ultrasound and measurement of cardiac conduction times Woman of childbearing potential (WOCBP) must 1) use adequate birth control or 2) have a negative pregnancy test less than 14 days before visit Capable of giving a signed informed consent Exclusion criteria Any medical condition limiting physical exertional capacity or requiring regular medication (allergy and contraceptives excepted) Pregnancy Breastfeeding History of syncope (syncope of presumed vasovagal nature with known precipitating factor excepted) Any known cardiac arrhythmia Any drug (contraceptives excepted) used on a regular basis for a chronic condition (allergy excepted) See “S1 Protocol” for specific requirements to adequate birth control. https://doi.org/10.1371/journal.pone.0270598.t002 Table 2. Inclusion and exclusion criteria. PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 Interventions Lower body negative pressure. LBNP is a method to simulate central hypovolemia where negative pressure is applied to the body from the waist-down [9] as shown on Fig 3. Thereby, 4 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia Table 1. Stop-criteria. Stop-criteria Symptoms or signs of impending circulatory collapse • Symptoms of pre-syncope 1. Light-headedness 2. Nausea 3. Sweating • Occurrence of hemodynamic thresholds preceding circulatory collapse (determined from measurements at baseline) 1. MAP-reduction to less than 75% of baseline values (measured at normovolemia) for >3 s 2. HR-reduction to less than 75% baseline values (measured at normovolemia) for >3 s Subject request for reasons other than above MAP = mean arterial pressure, HR = heart rate. https://doi.org/10.1371/journal.pone.0270598.t001 Table 1. Stop-criteria. Stop-criteria Symptoms or signs of impending circulatory collapse • Symptoms of pre-syncope 1. Light-headedness 2. Nausea 3. Sweating • Occurrence of hemodynamic thresholds preceding circulatory collapse (determined from measurements at baseline) 1. MAP-reduction to less than 75% of baseline values (measured at normovolemia) for >3 s 2. HR-reduction to less than 75% baseline values (measured at normovolemia) for >3 s Subject request for reasons other than above MAP = mean arterial pressure, HR = heart rate. https://doi.org/10.1371/journal.pone.0270598.t001 Table 1. Stop-criteria. Outcome measures During each visit we will measure heart rate with a three-lead ECG (Powerlab; ADInstruments, Dunedin, New Zealand). MAP and cardiac stroke volume will be measured with the volume- clamp method on the third finger of the left hand (Nexfin; Edwards Lifesciences corp., CA, USA) and by suprasternal Doppler ultrasound (SD-50 (SD-50; Vingmed Ultrasound, Horten, Norway). Cardiac output is calculated as the product of stroke volume from the Doppler ultrasound and heart rate from the ECG. Middle cerebral artery velocity (MCAV) will be measured using triplex ultrasound (GE E95; General Electric/ Vingmed, Horten, Norway) as a surrogate for cerebral blood flow. Arterial pulse oximetry will be obtained (Masimo Radical 7; Maximo corp., CA, USA) in addition to cerebral oxygen saturation by near infrared spectroscopy (Invos 5100C cerebral/ somatic oximeter; Somanetics, Troy, MI, USA). We will use laser Doppler flowmetry to measure acral skin blood flow (PeriFlux 4001 Master; Perimed AB, Ja¨rfa¨lla, Sweden), and volumetric cap- nography to measure respiratory frequency and end-tidal CO2 (Medlab CAP 10; Medlab GmbH, Stutensee, Germany). Tolerance to hypovolemia will be estimated as time from the start of LBNP 0 to hemodynamic decompensation, where decompensation is defined by Table 1. Stop-criteria. All data will be sampled continuously and stored on the hospital’s secured server. Our primary and secondary objectives with corresponding endpoints are shown in Table 3. Inclusion criteria 5 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia Fig 3. Illustration showing the test subject inside 1) the lower body negative pressure (LBNP) chamber. The chamber is 2) sealed just above the iliac crest and connected to 3) a vacuum pump controlled by 4) a pressure control unit. The applied negative pressure is displayed on 5) a pressure monitor. Measurements such as 6) ECG for heart rate (HR), 7) mean arterial pressure (MAP) and 8) stroke volume (SV) are connected to 9) a data acquisition device and 10) sampled on a laptop continuously. The inhalation gas is administered on 11) a face mask connected to 12) a gas cylinder. https://doi.org/10.1371/journal.pone.0270598.g003 Fig 3. Illustration showing the test subject inside 1) the lower body negative pressure (LBNP) chamber. The chamber is 2) sealed just above the iliac crest and connected to 3) a vacuum pump controlled by 4) a pressure control unit. The applied negative pressure is displayed on 5) a pressure monitor. Measurements such as 6) ECG for heart rate (HR), 7) mean arterial pressure (MAP) and 8) stroke volume (SV) are connected to 9) a data acquisition device and 10) sampled on a laptop continuously. The inhalation gas is administered on 11) a face mask connected to 12) a gas cylinder. https://doi.org/10.1371/journal.pone.0270598.g003 https://doi.org/10.1371/journal.pone.0270598.g003 https://doi.org/10.1371/journal.pone.0270598.g003 given inhalation gas will we present at all visits for the purpose of emergency unblinding due to medical considerations. given inhalation gas will we present at all visits for the purpose of emergency unblinding due to medical considerations. Statistical methods The effect of oxygen on cardiac output will be analyzed in a mixed linear regression model to account for repeated measurements within subjects. The effect of oxygen on MCAV and car- diac stroke volume will be analyzed in a similar fashion. LBNP-tolerance (time to decompensa- tion) will be analyzed in a mixed proportional hazards model. No interim analysis will be performed. Sample size The estimated effect of LBNP on cardiac output with its standard deviation was estimated from the raw data from a previous study [15]. A change of 15% in cardiac output is often used as a threshold when evaluating interventions to increase cardiac output [16]. We assume a mean cardiac output of 4.85 ± 1.08 L/min at baseline, and a change of -0.489 L/min for each LBNP- level (Δ-20 mmHg/level). Error within subjects is assumed independent between LBNP-levels with SD 0.385 L/min. Assuming that a 15% reduction in cardiac output during oxygen inhala- tion compared to air is significant, this would give an increased reduction (interaction effect) of 0.18 L/min per LBNP level. If assuming a SD of 0.18 L/min for this interaction effect, including 15 subjects would give a 1−β = 0.87 to detect this effect with α = 0.05, based on simulations. Trial oversight This study will be monitored by the Clinical Trials Unit (CTU) at Oslo University Hospital to ensure all procedures follow Good clinical practice (GCP) guidelines. Adverse events (AEs) and serious adverse advents (SAEs) will be collected from the start of the experiment on Visit 1 and until the end of Visit 2. All SAEs will be recorded and reported to the sponsor or designee imme- diately. The investigator will submit any updated SAE data to the sponsor within 24 hours. Fatal or life threatening suspected unexpected adverse reactions (SUSARs) will be reported to The Nor- wegian Medicines Agency within 7 days, and other SUSARs within 15 days. Discontinuation of study LBNP is released after 3 min at LBNP 80 mmHg or sooner by occurrence of any of the stop- criteria in given in Table 1. For safety reasons, an envelope containing a paper stating the 6 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia Table 3. Primary and secondary objectives and endpoints. Objectives Endpoints Primary Study the effect of supplemental oxygen on cardiac output during LBNP Difference in the change in cardiac output between oxygen and room air during LBNP Secondary Study the effect of supplemental oxygen on cardiac stroke volume during LBNP Difference in the change in cardiac stroke volume between oxygen and room air during LBNP Study the effect of supplemental oxygen on MCAV during LBNP Difference in the change in MCAV between oxygen and room air during LBNP Study the effect of supplemental oxygen on time to hemodynamic decompensation during LBNP Difference in time to decompensation between oxygen and room air during LBNP LBNP = lower body negative pressure, MCAV = middle cerebral artery velocity. h //d i /10 1371/j l 0270598 003 Discussion There are few experimental studies investigating the effect of supplemental oxygen on systemic hemodynamics during simulated hypovolemia. This is unfortunate since trauma patients 7 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia often receive supplemental oxygen and may suffer from hypovolemia. To our knowledge, only one study has previously exposed healthy volunteers to LBNP and 100% oxygen while measur- ing systemic hemodynamics [17]. They found no difference in hemodynamic response to LBNP between 100% oxygen and room air. A limitation to this study was that the authors only applied one level of LBNP, which was also low to moderate (-40 mmHg). In our planned study we will use graded LBNP from 0 to -80 mmHg to induce a greater span of hypovolemia and also estimate cerebral blood flow. We hope that our results can contribute to the understand- ing of the effect of oxygen on systemic and cerebral hemodynamics during hypovolemia. When designing this study, we had to weigh the duration of each LBNP-level against the desire to reach a sufficiently high (negative) level of LBNP. By increasing the duration of each LBNP-level, we could potentially increase the number of decompensations at the cost of fewer observations at high LBNP-levels. Based on the decompensation rate in prior work [15,18], we believe that the present LBNP protocol will be able to reveal an effect of oxygen on time to decompensation, i.e. LBNP tolerance. We also believe that the MAP stop-criterion of 25% below baseline values is suitable to detect hemodynamic decompensation, as the change rela- tive to the individual subject’s habitual blood pressure is considered. Also, a MAP reduction to less than 75% of baseline values largely coincides with a substantial reduction in systolic blood pressure using an absolute threshold of e.g. 80 mmHg. There are a few considerations regarding the validity of the suprasternal Doppler ultra- sound which is used to measure our main outcome variable. The velocity profile in the ascend- ing aorta is rectangular and preserved for the first 3 cm distal to the aortic orifice, even if the aortic diameter changes [19]. Consequently, slight changes in sample volume location in either lateral or caudal direction will have minor influence on the obtained velocity. Supporting information Supporting information S1 Checklist. (DOC) S1 Protocol. (PDF) S1 Checklist. (DOC) S1 Protocol. (PDF) Discussion In addition, since the suprasternal ultrasound probe is pointed in a craniocaudal direction, a theoretical caudal displacement of the heart with LBNP should have negligible influence on the angle of insonation and hence the obtained velocity. Trial status The study is planned to enroll test subjects from December 2021 to June 2022. PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 References 1. Horncastle E, Lumb AB. Hyperoxia in anaesthesia and intensive care. BJA Education. 2019; 19: 176– 182. https://doi.org/10.1016/j.bjae.2019.02.005 PMID: 33456888 2. American College of Surgeons, Committee on Trauma. Advanced trauma life support: student course manual. 2018. 3. Eskesen TG, Baekgaard JS, Christensen RE, Lee JM, Velmahos GC, Steinmetz J, et al. Supplemental oxygen and hyperoxemia in trauma patients: A prospective, observational study. Acta Anaesthesiol Scand. 2019; 63: 531–536. https://doi.org/10.1111/aas.13301 PMID: 30520014 4. Eskesen TG, Baekgaard JS, Steinmetz J, Rasmussen LS. Initial use of supplementary oxygen for trauma patients: a systematic review. BMJ Open. 2018; 8: e020880. https://doi.org/10.1136/bmjopen- 2017-020880 PMID: 29982208 5. Christensen MA, Steinmetz J, Velmahos G, Rasmussen LS. Supplemental oxygen therapy in trauma patients: An exploratory registry-based study. Acta Anaesthesiol Scand. 2021; 65: 967–978. https://doi. org/10.1111/aas.13829 PMID: 33840093 6. Vincent J-L, De Backer D. Circulatory Shock. New England Journal of Medicine. 2013; 369: 1726– 1734. https://doi.org/10.1056/NEJMra1208943 7. Brugniaux JV, Coombs GB, Barak OF, Dujic Z, Sekhon MS, Ainslie PN. Highs and lows of hyperoxia: physiological, performance, and clinical aspects. Am J Physiol Regul Integr Comp Physiol. 2018; 315: R1–R27. https://doi.org/10.1152/ajpregu.00165.2017 PMID: 29488785 8. Smit B, Smulders YM, van der Wouden JC, Oudemans-van Straaten HM, Spoelstra-de Man AME. Hemodynamic effects of acute hyperoxia: systematic review and meta-analysis. Crit Care. 2018; 22: 45. https://doi.org/10.1186/s13054-018-1968-2 PMID: 29477145 9. Goswami N, Blaber AP, Hinghofer-Szalkay H, Convertino VA. Lower Body Negative Pressure: Physio- logical Effects, Applications, and Implementation. Physiological reviews. 2018/12/13 ed. 2019; 99: 807– 851. https://doi.org/10.1152/physrev.00006.2018 PMID: 30540225 10. Snow G. blockrand: Randomization for block random clinical trials (R package version 1.3, 2013). 2020. Available: https://CRAN.R-project.org/package=blockrand. 11. R Core Team. R: A language and environment for statistical computing. Vienna, Austria; 2021. Avail- able: https://www.R-project.org/. 12. Team RStudio. RStudio: Integrated development environment for r. Boston, MA; 2021. Available: http:// www.rstudio.com/. 13. Gillen CM, Nishiyasu T, Langhans G, Weseman C, Mack GW, Nadel ER. Cardiovascular and renal function during exercise-induced blood volume expansion in men. Journal of Applied Physiology. 1994; 76: 2602–2610. https://doi.org/10.1152/jappl.1994.76.6.2602 PMID: 7928889 14. 06-4046.pdf. Available: https://www.legemiddelsok.no/_layouts/15/Preparatomtaler/Spc/06-4046.Pdf. 15. Hisdal J, Landsverk SA, Hoff IE, Hagen OA, Kirkeboen KA, Hoiseth LO. Associations between changes in precerebral blood flow and cerebral oximetry in the lower body negative pressure model of hypovole- mia in healthy volunteers. PLoS One. 2019/06/30 ed. 2019; 14: e0219154. https://doi.org/10.1371/ journal.pone.0219154 PMID: 31251778 16. Cecconi M, Parsons AK, Rhodes A. What is a fluid challenge? Curr Opin Crit Care. 2011; 17: 290–295. Author Contributions Conceptualization: Sole Lindvåg Lie, Jonny Hisdal, Marius Rehn, Lars Øivind Høiseth. Funding acquisition: Jonny Hisdal, Marius Rehn, Lars Øivind Høiseth. Methodology: Sole Lindvåg Lie, Jonny Hisdal, Marius Rehn, Lars Øivind Høiseth. Project administration: Sole Lindvåg Lie, Lars Øivind Høiseth. Supervision: Jonny Hisdal, Marius Rehn, Lars Øivind Høiseth. Writing – original draft: Sole Lindvåg Lie. Writing – review & editing: Jonny Hisdal, Marius Rehn, Lars Øivind Høiseth. 8 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 PLOS ONE Supplemental oxygen in hypovolemia PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022 References https://doi.org/10.1097/MCC.0b013e32834699cd PMID: 21508838 17. Kim YK, Jun IG, Kim SR, Hwang JH, Cho SK, Han SM, et al. Using 100% oxygen does not alter the car- diovascular autonomic regulation during non-invasively simulated haemorrhage in healthy volunteers. J Int Med Res. 2008; 36: 227–236. https://doi.org/10.1177/147323000803600203 PMID: 18380931 18. Lie SL, Hisdal J, Høiseth LØ. Cerebral blood flow velocity during simultaneous changes in mean arterial pressure and cardiac output in healthy volunteers. Eur J Appl Physiol. 2021; 121: 2207–2217. https:// doi.org/10.1007/s00421-021-04693-6 PMID: 33890157 19. Eriksen M, Walløe L. Improved method for cardiac output determination in man using ultrasound Dopp- ler technique. Med Biol Eng Comput. 1990; 28: 555–560. https://doi.org/10.1007/BF02442607 PMID: 2287179 9 / 9 PLOS ONE \| https://doi.org/10.1371/journal.pone.0270598 June 24, 2022
https://openalex.org/W4376609119	https://www.scielo.br/j/bjps/a/XW8rq83BvGnkKygnKQwTRYM/?lang=en&format=pdf	English	null	Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats	Brazilian Journal of Pharmaceutical Sciences	2,023	cc-by	8,501	Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats Pratik Prakash Maske1,2, Popat Sonappa Kumbhar3, Ashok Gurulingappa Wali4, John Intru Disouza3, Maya Sharma1,5 1Pacific Academy of Higher Education and Research, Udaipur, Rajasthan, India, 2Department of Pharmaceutical Chemistry, Genesis Institute of Pharmacy, Sonyachi Shiroli, Tal: Radhanagari, Dist: Kolhapur Maharashtra 416212, India, 3Tatyasaheb Kore College of Pharmacy, Warananagar, Tal: Panhala, Dist: Kolhapur Maharashtra, 416113, India, 4Hon. Shri Annasaheb Dange Ayurved Medical College, Ashta Tal: Walawa, Dist: Sangli Maharashtra, 416301, India, 5Pacific College of Pharmacy, Pacific University, Udaipur Rajsthan, India Diabetes is a life-threatening disease, and currently available synthetic medicines for treating diabetes are associated with various side effects. Therefore, there is an unmet need to develop herbal remedies against diabetes as an alternative to synthetic medicines. Although local healers use the roots of Spermadicyton suaveolens (SS) to manage diabetes, there is negligible research to validate its antidiabetic properties. The present investigation aims to the assess the antioxidant, antidiabetic, and antihyperlipidemic potential of the ethanolic extract of S. Suaveolen’s roots (EESS) on streptozotocin (STZ) induced diabetic rats. The extract was screened for in vitro antioxidant and antidiabetic activity. The in vivo antidiabetic potential of EESS (at 200 and 400 mg/kg) was studied on STZ-induced diabetic rats for 20 days. The EESS displayed significant (p<0.05) antidiabetic and antioxidant properties. The administration of 200 mg/kg and 400 mg/kg EESS in STZ-induced diabetic rats significantly reduced hyperglycemia, and restored antioxidant enzymes and lipid profile–a high density lipoprotein (HDL) increased by the administration of a single dose of streptozotocin. Thus, EESS could be a promising herbal medicine in the treatment of diabetes and hyperlipidemia. Keywords: Spermadicyton suaveolens extract. Antioxidant. Antidiabetic. Hyperlipidemic. Histopathology. LIST OF ABBREVIATIONS SS–Spermadicyton Suaveolens STZ–Streptozotocin EESS–Ethanolic extract of Spermadicyton Suaveolens CESS–Chloroform extract of Spermadicyton Suaveolens PESS–Petroleum extract of Spermadicyton Suaveolens AESS–Aqueous extract of Spermadicyton Suaveolens NO–Nitric oxide SOD–Superoxide radical scavenging HDL–High-density lipoprotein CAM–Complementary and alternative medicine NBT–Nitro blue tetrazolium DMSO–Dimethyl sulfoxide P-NPG–Para-nitrophenyl–α–D-glucopyranoside Na2CO3–Sodium carbonate OGTT–Oral glucose tolerance test MDA–Malondialdehyde GSH–Glutathione CAT–Catalase HDL–High-density lipoprotein LDL–Low-density lipoprotein VLDL–Very low-density lipoprotein Correspondence: J. I. Disouza. Tatyasaheb Kore College of Pharmacy. Warananagar, Tal: Panhala, Dist: Kolhapur Maharashtra. 416113, India. E-mail: johnsir4u@gmail.com. ORCID: https://orcid.org/0000-0002-9807- 7932 Correspondence: J. I. Disouza. Tatyasaheb Kore College of Pharmacy. Warananagar, Tal: Panhala, Dist: Kolhapur Maharashtra. 416113, India. E-mail: johnsir4u@gmail.com. ORCID: https://orcid.org/0000-0002-9807- 7932 Article Brazilian Journal of Pharmaceutical Sciences INTRODUCTION metabolism, leading to a cytotoxic action on pancreatic β-cells. This phenomenon brings on hyperglycemia in rodents, just like diabetic people. The oxidative stress disrupts endoplasmic reticulum function leading to cell necrosis. Hyperglycemia engenders variance in the metabolism of proteins and lipids (Prabhakar 2016; Kunwar, Priyadarsini, 2011; Sajid et al., 2020). Management and treatment of chronic illnesses are one of the most significant difficulties faced by healthcare societies. Despite significant advances in the sector, clinicians have not been able to eradicate severe chronic illnesses such as diabetes mellitus (Asif et al., 2019). Diabetes is the third-most reason for mortality. As per a 2021 report, approximately 537 million adults are living with diabetes. This number is projected to grow to 643 million by 2030 and 783 million by 2045. Globally, more than 90% of people have type 2 diabetes (Cho et al., 2018; IDF Atlas 2021). The Spermadictyon suaveolens (SS) commonly called as Forest Champa, Van-Champa, Gidesa, Jitsaya, etc. has diversified therapeutic applications. It is found in Maharashtra, the Himalayan region, Kashmir, and the Northern Areas of Pakistan. Local healers in Maharashtra use the herb’s stem and root for various ailments like bone and muscle wounds, herpes, diabetes, etc. Ayurvedic practitioners or vaidyas use this plant’s stem powder to treat viral ailments like herpes and diabetes (Musmade, Rakshe, Mokat, 2016). Recognising this severity, investigators worldwide emphasise the need to discover complementary and alternative medicine for diabetes. Over 80 percent of the population in the economically growing and emerging nations use herbal remedies to treat their diseases. Natural antidiabetic products are a game-changer in the treatment of diabetes due to the higher costs and adverse effects associated with allopathic therapy. (Birdee, Yeh, 2018; Falguni et al., 2017; Uddin et al., 2018) p ( ) The methanolic, chloroform, and petroleum ether extract of bark and leaves exhibit potent antioxidant and antimicrobial action. The roots were evaluated for their wound healing properties in Wistar rats (Ajaib, Khalid, Hanif, 2016). The stem and leaf contain phytoconstituents including Azulene, Tetratetracontane, n-hexadecanoic acid, Ergost–5-en–3–ol,22,23-dimethyl-,acetate,(3β), Phenol,2-methoxy-4-(1-propenyl)-,(E)(9), etc., and they have reported antioxidant, anti-inflammatory, antibacterial, and antiulcerogenic activity (Kulkarni, Sathe, 2013). The flower and leaf contain 3,7,11,15– tetramethyl–2–hexadecen–1–ol, Phytol, 3,4– dianhydro–2–deoxy–. beta–d–lyxo–hexo–pyranose, 3,7,11,15–tetramethyl–2–hexadecen–1–ol, etc., and has shown analgesic, antimicrobial, anticancer, anti- diuretic, anti-inflammatory, and antipyretic properties. (Papitha, Ravi, Selvaraj, 2017). LIST OF ABBREVIATIONS HDL–High-density lipoprotein CAM–Complementary and alternative medicine NBT–Nitro blue tetrazolium DMSO–Dimethyl sulfoxide P-NPG–Para-nitrophenyl–α–D-glucopyranoside Na2CO3–Sodium carbonate OGTT–Oral glucose tolerance test MDA–Malondialdehyde GSH–Glutathione CAT–Catalase HDL–High-density lipoprotein LDL–Low-density lipoprotein VLDL–Very low-density lipoprotein LIST OF ABBREVIATIONS SS–Spermadicyton Suaveolens STZ–Streptozotocin EESS–Ethanolic extract of Spermadicyton Suaveolens CESS–Chloroform extract of Spermadicyton Suaveolens PESS–Petroleum extract of Spermadicyton Suaveolens AESS–Aqueous extract of Spermadicyton Suaveolens NO–Nitric oxide SOD–Superoxide radical scavenging HDL–High-density lipoprotein CAM–Complementary and alternative medicine NBT–Nitro blue tetrazolium DMSO–Dimethyl sulfoxide P-NPG–Para-nitrophenyl–α–D-glucopyranoside Na2CO3–Sodium carbonate OGTT–Oral glucose tolerance test MDA–Malondialdehyde GSH–Glutathione CAT–Catalase HDL–High-density lipoprotein LDL–Low-density lipoprotein VLDL–Very low-density lipoprotein Correspondence: J. I. Disouza. Tatyasaheb Kore College of Pharmacy. Warananagar, Tal: Panhala, Dist: Kolhapur Maharashtra. 416113, India. E-mail: johnsir4u@gmail.com. ORCID: https://orcid.org/0000-0002-9807- 7932 SS–Spermadicyton Suaveolens STZ–Streptozotocin EESS–Ethanolic extract of Spermadicyton Suaveolens CESS–Chloroform extract of Spermadicyton Suaveolens PESS–Petroleum extract of Spermadicyton Suaveolens AESS–Aqueous extract of Spermadicyton Suaveolens NO–Nitric oxide SOD–Superoxide radical scavenging SS–Spermadicyton Suaveolens Page 1/17 Braz. J. Pharm. Sci. 2023;59: e21820 Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma Plant Material The root of SS was collected in the month of November from the hilly vicinity of Panhala Fort in Kolhapur district, Maharashtra. The plant was identified and authenticated by taxonomist Dr. Kavale from the Department of Botany, Shivraj College, Gadhinglaj, Maharashtra. A voucher specimen was deposited (SCG/ BOT/HERB/07-2019) in Shivrai College. After collecting the roots, they were chopped into small pieces and dried before being used as a raw material for the treatment. The dried roots were pulverized on a Rising Automatic DP Pulverizer and later powedered. The powder was then surpassed through a 40# numbered sieve and stored in an airtight container. INTRODUCTION Diabetes mellitus is a fuel-metabolism disorder caused by increased blood glucose due to starving insulin secretion or inability to meet the demand of target tissues. Hyperglycemia is the first clinical sign, followed by metabolic interruption of biomolecules. Prolonged hyperglycemia causes chronic microvascular and macrovascular complications, such as neuropathy, nephropathy, retinopathy, and arteriosclerosis (Satyanarayana, Chakrapani, 2012; Prabhakar, 2016). Oxidation is the fuel for the development of diabetes by lipid autoxidation, DNA, and protein damage. The development of molecular oxygen is an integral part of routine work, and incomplete reduction is responsible for the generation of oxygen free radicals. Antioxidants neutralise the free radicals and reactive oxygen species. The imbalance between generation and neutralisation leads to oxidative stress which increases risk of complications like cardiovascular diseases, diabetes, and neurogenerative diseases, etc. (Gudise, Chowdhury, Manjappa 2019; Kunwar, Priyadarsini, 2011). As far as we know, this is the first work to provide insight into the antioxidant and hypoglycemic effects of EESS. The current investigation was based upon the evaluation of in vitro and in vivo hypoglycemic action on STZ-NA induced rats. Besides, serum investigation of biomarkers along with in vivo antioxidant behavior was evaluated. Thus, we have investigated the efficiency of crude extract in diabetes-related complications in vivo. Streptozotocin is a diabetes inducer. Its administration causes a burst of free radicals due to Page 2/17 Page 2/17 Page 2/17 Braz. J. Pharm. Sci. 2023;59: e21820 Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats MATERIAL AND METHODS extracts of SS–the aqueous extract of SS (AESS), ethanol extract of SS (EESS), chloroform extract of SS (CESS), and petroleum ether extract of SS (PESS)–were prepared using four solvents with different polarity. Thus, the solvents’ systematic extraction of plant material had increasing polarity. Successive extraction was done using 50 g powdered root material via the soxhlet apparatus (Figure 1). The PESS using petroleum ether (250 mL, 45ºC 15 cycles) was obtained, and after successive extraction, CESS using chloroform (250 mL, 45ºC 15 cycles) and EESS employing ethanol (250 mL, 60ºC 15-17 cycles) were obtained. After completing the extracting and filtering, the concentration of the residue weighing 50 ml in the water bath. The extract was then transferred to a previously weighed evaporating dish. The total weight of the evaporating dish containing the extract was measured and placed in the water bath for evaporation until it became viscous. The difference in weight was calculated every 10 minutes until a constant weight was obtained. Superoxide radical scavenging assay (SOD) The dioxide ion engulfed propensity of the extract was calculated using the alkaline DMSO process, which was slightly modified from Elizabeth and Rao’s (1990) method. The decreased nitro blue tetrazolium (NBT) was evaluated by this method. Both extract (30 μL) and standard were dissolved in DMSO, 100 μl of alkaline DMSO, and 10 μl of NBT were added to fulfil the volume 140 μL. Finally, the reading was taken at 560 nm by a microplate reader. Ascorbic acid served as a positive control (Harput et al., 2011). α- glucosidase activity The Griess Illosvory reaction was utilised to assess free radical scavenging. The Griess Illosvory reagent commercially contains naphthyl ethylene diamine dihydrochloride (0.1% w/v), 1 mL of 10 mm sodium nitroprusside in 0.5 mL phosphate buffer saline (pH 7.4) was mixed with 1 ml each of extract of different concentrations (200, 400, 600, 800, 1000 µg/mL). The mixture was incubated at 25ºC for 150 min. After incubation, the reaction mixture was mixed with 1.0 ml of pre-prepared Griess reagent (1.0 mL sulfanilic acid reagent 0.33% in 20% glacial acetic acid at room temperature for 5 min with 1 mL of naphthyl ethylenediamine dichloride 0.1% w/v). The mixture was then incubated at room temperature for 30 minutes, and the nitrite concentration was estimated at 546 nm using nitrite solutions as the control. Buffer solution served as blank, whereas the reference solution was ascorbic acid (Patel et al., 2010). The decreasing absorbance indicates a high nitric oxide scavenging activity. The percentage inhibition was determined using the equation, The α-glucosidase inhibitory activity of SS extracts was analysed using a standard method with little modifications. Briefly, the concoction of SS extracts (50 µL), glutathione (50 µL), 10 μL α-glucosidase in phosphate buffer was mixed in a 96-well plate and incubated for 15 minutes at 37°C. After incubation, 20 μL P-NPG (5 mM) was added and then incubated for 15 min at 37°C. The addition of 50 μL Na2CO3 (0.1 M) terminated the reaction. The blank was prepared using a similar process but without adding enzyme (α - glucosidase) solution. Without a test sample, each procedure was done thrice to act as a control. The sample and blank absorbances were read at 400 nm. The p-nitrophenol produced from p-NPG determined the α-glucosidase activity. Acarbose was positive control (Telagari, Hullatti, 2015). Animals Male Wistar rats weighing between 190- 220 gm were chosen for the experiment. They were placed in rat cages made of stainless steel. Except where fasting was needed, food and water were given ad libitum. The rats were kept at a constant temperature of 20-250C and were subjected to a regular 12-hour light and 12-hour dark periods. They were given a two-week acclimatisation period before the experiment. Soxhlet extraction The solvents employed for the extraction process were selected based on their polarity. The four different FIGURE 1 - Schematic representation of Soxhlet extraction. FIGURE 1 - Schematic representation of Soxhlet extraction. Page 3/17 Braz. J. Pharm. Sci. 2023;59: e21820 Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma II. Estimation of total cholesterol (TC) and HDL cholesterol (HDLC) The TC and HDLC estimation were performed as per the previously reported protocol using a CHOD-PAP (Cholesterol oxidase phenol4-aminoantipyrine) method kit. After a brief gap, 1 mL of the enzyme reagent was added to 10 μL of EESS. The mixture was shaken well and incubated at 47°C for 5 minutes and optical I. Estimation of triglycerides (TG) The estimation was performed using a GPO-TOPS AGAPPE kit. Briefly, the enzyme reagent (1 mL) was added to 10 μL of the sample. The mixture was then shaken well and incubated at 37°C for 10 min. Finally, the optical density (OD) was recorded at 546 nm using a spectrophotometer. Biochemical parameters measurement For the induction T2DM, firstly intraperitoneal (i.p.) injection of nicotinamide (NA); 120 mg/kg in saline was administered in animals. After completion of 15 min of nicotinamide administration, freshly prepared streptozotocin in 0.1M citrate buffer (pH 4.5), was administered in single dose 60 mg/kg to the animals (Furman et al., 2021) To avoid death from STZ-NA induced hypoglycemia, all rats were given a 10% glucose solution for 12 hr. After three days rats with fasting plasma glucose > 170 mg/dL were diabetic and were integrated in this study. The STZ-NA-induced diabetic rats were split into different subgroups, each with six animals. (Furman et al., 2021). Acute toxicity study The lethal dose of plant extract was investigated using OECD guidelines (test 423: Acute oral toxicity- acute toxic class method, 2002). (OECD Library, 2002). Before the trial, the animals were split into four classes (n=6) and were kept on fasting overnight. The EESS was administered orally, starting from 5, 50, and 300 to increasing to 2000 mg/kg body weight. At 30 minutes and 2, 4, 8, and 24 hours after the dosage, all groups were closely monitored for the occurrence of any clinical or toxicological effects and there after every 24 hours for 14 days. The animals were monitored for Page 4/17 Braz. J. Pharm. Sci. 2023;59: e21820 tioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats Group- I Normal control: Animals receiving normal diet Group-II Positive control (Diabetic control): Diabetic animals receiving normal diet and free from any treatment Group-III Standard (glibenclamide) treated: Diabetic animals treated with glibenclamide at a dose of 5 mg/ kg body weight Group-IV Test (EESS treated at a dose of 200 mg): Diabetic animals treated with ethanolic extract of SS at a dose of 200 mg/kg body weight Group-V Test (EESS treated at a dose of 400 mg): Diabetic animals treated with ethanolic extract of SS at a dose of 400 mg/kg body weight Group- I Normal control: Animals receiving normal diet Group-II Positive control (Diabetic control): Diabetic animals receiving normal diet and free from any treatment Group-III Standard (glibenclamide) treated: Diabetic animals treated with glibenclamide at a dose of 5 mg/ kg body weight 14 days for the long-term possible lethal outcome. The body weights of the animals were measured on days 1, 7, and 14. Oral glucose tolerance test (OGTT) Healthy rats were chosen for the OGTT. The rats were feed in an interval of 12 hours and were divided into 4 groups. Group-IV Test (EESS treated at a dose of 200 mg): Diabetic animals treated with ethanolic extract of SS at a dose of 200 mg/kg body weight The OGTT was done after starving normal rats for 2 hours. Distilled water, EESS extract (200 mg/ kg and 400mg/kg), and glibenclamide (2 mg/kg) were administered to all the four groups of rats. Glucose (2 g/kg) was fed 30 minutes after the pretreatment using distilled water, EESS, and glibenclamide. To measure the blood sugar level, blood was taken from the tail vein at 0 minutes (as a baseline before the glucose solution loading), and then at 30, 60, and 120 minutes after glucose loading, to evaluate the outcome of the extract on glucose level. The serum sugar was estimated by blood glucose test strips and glucometer (Changsha sinocare Inc; China). Both standard and test samples were given to the animals administered to the rats by gastric incubation using oral gavage. Briefly, glibenclamide was administered to the animals of group III at a dose of 5 mg/kg. Similarly, EESS was administered to the animals of group IV and V at dose of 200 and 400 mg/ kg respectively. The blood samples were taken from tail vein before the experiment and on the 1st, 10th and 20th day of the treatment and fasting blood glucose level was estimated. Also, body weights of animals were recorded. Effect of long-term treatment with EESS on glycemic control The animals were grouped into 5 groups and 6 rats were placed in each group. Braz. J. Pharm. Sci. 2023;59: e21820 Page 5/17 Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma Histopathological analysis The animal pancreas was removed at the end of the experiment and fixed in buffered formalin (10%) at room temperature. Then it was stained with Hematoxylin and Eosin (H and E) stains, and observed using a 40X magnifications under a microscope on albumenized glass slides (Bancroft, Gamble, 2008). Preparation of Tissue Homogenate The liver tissue collected from the animals were homogenated using a phosphate buffer (200 mM; pH 6.6). This obtained liver tissue homogenate was centrifuged and used for analysis. III. Estimation of VLDL and LDL cholesterol (Friedewald et al.,1972) The level of VLDL and LDL cholesterol were calculated based on the parameters calculated in the two procedures using a Friedewald’s formula. Statistical Analysis The liver tissue homogenate 0.4 mL (10%) was compounded with sodium dodecyl sulphate (1.5 mL; 8.1%). The above mixture, acetate buffer pH 3.5 (1.5 mL; 20%), and TBA solution (1.5 mL; 0.8%) were mixed and the mixture was completely vortexed. It was left to stand after cooling until organic and aqueous layers were split, 5 mL of n-butanol-pyridine was added in the process (15:1). Finally, the absorbance of organic layer was recorded at 532 nm using a UV-visible spectrophotometer (Sani, Kouhsari, Moradabadi, 2012; Merghem, Dahamna, Khennouf, 2019). The outcomes of samples were displayed as mean ± standard error means. The statistical analysis was performed by using GraphPad Prism software version 5 (GraphPad Software, Inc., La Jolla, CA, USA). The results were analyzed by a one-way analysis of variance. Values of p < 0.01 and p < 0.05 were considered to be statistically significant. C) Determination of reduced glutathione (GSH) density (OD) reading was recorded at 505 nm using a spectrophotometer. (Palaniappan et al., 2020). density (OD) reading was recorded at 505 nm using a spectrophotometer. (Palaniappan et al., 2020). For this assay, 1 mL liver tissue homogenate was mixed with 3mL of DTNB solution in phosphate buffer (0.1M pH 8) and 7.5 mL 0.3 mM NAPDH (pH 7.3). The resultant solution was then diluted with 4 mL 0.1 M phosphate buffer (pH 8). Finally, the absorbance was taken at 412 nm against blank solution prepared (Sani, Kouhsari, Moradabadi, 2012; Merghem, Dahamna, Khennouf, 2019). B) Determination of catalase enzyme activity The four extract of S. Suaveolens were prepared by hot continuous (soxhlet extraction) process. These extract were further characterized for in vitro and in vivo activities. The catalase enzyme activity was determined by mixing 200 mL of diluted liver tissue homogenate with 1.0 mL of phosphate buffer, 0.4 mL of distilled water, and 0.5 mL H2O2. To terminate the reaction, 2 mL potassium dichromate acetic acid was added after 1 minute of incubation at 37°C. Finally, the samples were maintained in a boiling water bath for 15 minutes (Sani, Kouhsari, Moradabadi, 2012; Merghem, Dahamna, Khennouf, 2019). Nitric oxide scavenging assay Nitric oxide is a highly reactive compound which in the presence of oxygen generates the stable compounds Page 6/17 Braz. J. Pharm. Sci. 2023;59: e21820 Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic R potential with an augment in the concentration. The EESS showed more potency (354.69 ± 0.92 μg/mL) for elimination of free nitrite radical as compared to other. However the EESS (354.69 ± 0.92 μg/mL) scavenges less free radical than the standard (73.06 ± 0.42 μg/mL). (nitrates and nitrite). The Griess reagent was used for its measurement. The concentration of nitrous acid is decreased due to the ability of scavenging action of test compounds. The antioxidant potential of extracts is depicted in (Figure 2A). The obtained results revealed increased antioxidant FIGURE 2A - Scavenging potential of different extracts of Spermadicyton suaveolens against Nitric oxide. FIGURE 2A - Scavenging potential of different extracts of Spermadicyton suaveolens against Nitric oxide. Nitric oxide (NO) is synthesized by vascular endothelial cells, phagocytes, and neurons which cause harmful effects from amino acid L-arginine. When radical superoxide responds, the lethality of NO is enhanced due to the formation of another reactive compound peroxynitrite anion (ONOO-). Naturally sourced antioxidants showed promise in the reduction of oxidative stress and therefore can be used as an choice to the synthetic antioxidants in the treatment of diabetes (disease associated with oxidative stress) (Habu, Ibeh, 2015). In the current investigation, EESS and CESS demonstrated good nitrite free radical scavenging action in a dose dependent manner which is comparable with standard (ascorbic acid). Among these two extracts, EESS displayed potent nitrite free radical scavenging action than CESS that may be due to the presence of more polyphenolic contents extracted by ethanol from SS through the better penetration into cell membrane of plants (Boora, Chirisa, Mukanganyama, 2014). Braz. J. Pharm. Sci. 2023;59: e21820 Superoxide radical scavenging This assay measured the dioxide scavenging potential of extracts. The nonenzymatic phenazine methosulfate- nicotinamide adenine dinucleotide (PMS/NADH) system produces dioxide ions, which lessen NBT to a purple formazan. The generated superoxide was reduced to nitro blue tetrazolium, which forms purple formazan at 560 nm. Among the four extracts, EESS displayed significant (p<0.01) superoxide radical scavenging activity (low IC50 value: 0.278 ± 0.028 μg/mL) when compared to AESS, CESS and PESS. However, the free radical scavenging potential of EESS was observed to be less than standard (IC50: 0.092 ± 0.011) depicted in (Figure 2B). Braz. J. Pharm. Sci. 2023;59: e21820 Page 7/17 Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma FIGURE 2B - Scavenging potential of different extracts of Spermadicyton suaveolens against Superoxide. FIGURE 2B - Scavenging potential of different extracts of Spermadicyton suaveolens against Superoxide. The extracts were tested for the NBT assay to determine its ability to remove superoxide anions. Superoxide anion is the prevalent free radical in living system and since it is starting material for other reactive oxygen species such as hydroxyl radical, hydrogen peroxide etc. causes tissue damage with potential reaction with biomolecules. The concentration of superoxide anion is the prime for oxidative stress (Hazra, Biswas, Mandal, 2008). In addition, at the cellular level, superoxide offers much more toxic effects. The extracts and standard was demonstrated increase in superoxide scavenging action with increase in concentration. The superoxide scavenging action of standard (ascorbic acid) was found to be higher than that of EESS and PESS. Several novel bioactive phyto-constituents (secondary metabolites) derived from plants with hypoglycemic and anti-hyperglycemic activities showed remarkable anti-diabetic efficacy that is comparable to, and often even more effective than currently available marketed oral hypoglycemic medications (Rai et al., 2013; Watal et al., 2014). α-glucosidase activity The half-maximal inhibitory activity of SS extracts on α-glucosidase was represented (Figure 3). There was a significant difference (P < 0.05) in the inhibitory potential of the standard (acarbose) and all extracts tested on α-glucosidase activities. The EESS displayed good inhibitory potential against α-glucosidase (IC50 727 ± 0.94 μg/mL) than other extracts but not promising while compared to standard acarbose (IC50 94.18 ± 0.692 μg/mL). In vitro antidiabetic activity Diabetes mellitus is associated with malfunctioning of the metabolic system which becomes chronic that leads to several complications. Due to the constraints of publicly known pharmacological mediators for the management of diabetes, there is an unmet need to discover novel antidiabetic medicines with assorted modes of action (Adedayo et al., 2014). Braz. J. Pharm. Sci. 2023;59: e21820 Page 8/17 Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats FIGURE 3 - Inhibitory effects of different extracts of Spermadicyton suaveolens on the α-glucosidase. GURE 3 - Inhibitory effects of different extracts of Spermadicyton suaveolens on the α-glucosidase. Carbohydrates are long chain compound that undergoes breakdown into the monosaccharide like glucose in the digestive tract through enzyme namely α-glucosidase. Moreover, this enzyme is responsible for the reabsorption of glucose in the intestine. Therefore, this enzyme can serve as a chief target in the treatment of diabetes. Thus, enzyme suppression is one of the effective approaches used to reduce postprandial hyperglycemia in diabetics by slugging the absorption of intestinal glucose. So, the enzyme inhibitors are vital in management and treatment of diabetes (Khan et al., 2016). In this study, EESS showed strongest α-glucosidase inhibitory action when compared to other extracts however; it was less than standard (Alimi, Ashafa, 2017). Thus, better antidiabetic activity of the EESS extract against T2DM could be attributed to the delaying of carbohydrate digestion due to the significant inhibition of α-glucosidase activity. Acute toxicity study Oral administration of the crude EESS was found to be safe at a dose of 2 g/kg body weight, which showed no any toxicity. Thus, after 14 days, no animal mortality was seen with EESS. The lethal dose of EESS was considered to be more than 2 g/kg. The 200 mg/kg and 400 mg/kg doses were considered for study. The body weights of animals of the control and EESS treated groups were increased progressively throughout the study period (Table I). Moreover, behavioral observations of the all animals treated at different doses of EESS were found to be normal as control group animal behaviour (Table II). Thus, these obtained results revealed safety and tolerability (absence of toxicity) in the animals treated with various doses of EESS. Braz. J. Pharm. Sci. 2023;59: e21820 Page 9/17 TABLE I - Effect of extract on body weight of rats in Acute toxicity study Group Body Weight (gm) 1st Day 7th Day 14th Day Control 180.33 ±0.39 187.5 ±0.49 200.16 ±0.69 EESS 300 mg/kg 184.66 ±0.42 187.5 ±0.37 198.5 ±0.45 Braz. J. Pharm. Sci. 2023;59: e21820 Page 9/17 TABLE I - Effect of extract on body weight of rats in Acute toxicity study Group Body Weight (gm) 1st Day 7th Day 14th Day Control 180.33 ±0.39 187.5 ±0.49 200.16 ±0.69 EESS 300 mg/kg 184.66 ±0.42 187.5 ±0.37 198.5 ±0.45 TABLE I - Effect of extract on body weight of rats in Acute toxicity study Braz. J. Pharm. Sci. Acute toxicity study 2023;59: e21820 Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma TABLE I - Effect of extract on body weight of rats in Acute toxicity study Group Body Weight (gm) 1st Day 7th Day 14th Day EESS 1000 mg/kg 181.33 ±0.33 186.33 ±0.44 200.33 ±0.38 EESS 2000 mg/kg 184.16 ±0.41 185.66 ±0.57 195.66 ±0.47 TABLE II - Behavioral observations of rats on EESS administration Parameters Dose Behavioural observations 30 min 2 H 4 H 8 H 24 H 7th Day 14th Day Salivation Control Normal Normal Normal Normal Normal Normal Normal EESS 300 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 1000 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 2000 mg/kg Normal Normal Normal Normal Normal Normal Normal Sleep Control Normal Normal Normal Normal Normal Normal Normal EESS 300 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 1000 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 2000 mg/kg Normal Normal Normal Normal Normal Normal Normal Mortality Control Not found Not found Not found Not found Not found Not found Not found EESS 300 mg/kg Not found Not found Not found Not found Not found Not found Not found EESS 1000 mg/kg Not found Not found Not found Not found Not found Not found Not found EESS 2000 mg/kg Not found Not found Not found Not found Not found Not found Not found TABLE I - Effect of extract on body weight of rats in Acute toxicity study TABLE II - Behavioral observations of rats on EESS administration Parameters Dose Behavioural observations 30 min 2 H 4 H 8 H 24 H 7th Day 14th Day Salivation Control Normal Normal Normal Normal Normal Normal Normal EESS 300 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 1000 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 2000 mg/kg Normal Normal Normal Normal Normal Normal Normal Sleep Control Normal Normal Normal Normal Normal Normal Normal EESS 300 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 1000 mg/kg Normal Normal Normal Normal Normal Normal Normal EESS 2000 mg/kg Normal Normal Normal Normal Normal Normal Normal Mortality Control Not found Not found Not found Not found Not found Not found Not found EESS 300 mg/kg Not found Not found Not found Not found Not found Not found Not found EESS 1000 mg/kg Not found Not found Not found Not found Not found Not found Not found EESS 2000 mg/kg Not found Not found Not found Not found Not found Not found Not found TABLE II - Behavioral observations of rats on EESS administration Hypoglycemic action of SS extract on Oral Glucose loaded rats Hypoglycemic action of SS extract on Oral Glucose loaded rats glucose level was observed after administration of glucose orally. The EESS at concentration of 200 mg/kg and 400 mg/kg evoked significant (p<0.01) hypoglycemic effects after 120 min as compared to control. glucose level was observed after administration of glucose orally. The EESS at concentration of 200 mg/kg and 400 mg/kg evoked significant (p<0.01) hypoglycemic effects after 120 min as compared to control. The effect of EESS on glucose tolerance in rats is depicted in (Table III). Considerable change in blood Page 10/17 Braz. J. Pharm. Sci. 2023;59: e21820 TABLE III - Hypoglycemic action of S. Suaveolens on Oral Glucose loaded rats Group No. Treatment Blood glucose level (mg/dL) 0 min 30 min 60 min 120 min Control 72.83± 1.47 133.16± 1.16 127.33± 1.21 111.33± 1.21 Std (Glibenkamide 2mg/kg) 82.16± 2.13* 121.16± 1.16* 97.33± 1.36 89.16± 1.47 TABLE III - Hypoglycemic action of S. Suaveolens on Oral Glucose loaded rats Page 10/17 Braz. J. Pharm. Sci. 2023;59: e21820 Braz. J. Pharm. Sci. 2023;59: e21820 tioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats TABLE III - Hypoglycemic action of S. Suaveolens on Oral Glucose loaded rats Group No. Treatment Blood glucose level (mg/dL) 0 min 30 min 60 min 120 min EESS 200 mg/kg 88.5± 2.07* 122.16 ± 1.60* 92.33 ± 1.75* 89.66± 1.21** EESS 400mg/kg 90.83± 1.16* 136.66± 1.63* 95.83 ± 1.72* 93.5 ± 1.37* TABLE III - Hypoglycemic action of S. Suaveolens on Oral Glucose loaded rats rats (group II) gained significant (p<0.05) lower weight than normal control rats (group I). Both groups (IV and V) treated with EESS at (200 mg/kg and 400 mg/kg) showed elevated body weight as compared to diabetic control however, it is less when compared to standard (Table IV). Thus, the diabetic rat displayed frequently decrease in the body weight. An oral glucose tolerance test identifies how your body handles glucose before and after the meal. OGTT is considered a ‘24 carat gold sensitive’ test for screening and diagnosis of glucose utilization those other tests miss. In the current OGTT study, the EESS shut down the blood glucose from 60 min as compared to control. This confirms the appropriate utilization of glucose by rats. The capability of EESS to reduce postprandial glucose may be attributed to a decrease in glycogen split up and synthesis of glucose, which shoot up the glucose absorption. Hypoglycemic action of SS extract on Oral Glucose loaded rats The obtained results clearly revealed the capability of the extract to minimize hyperglycemia- related problems in diabetes (Kifle, Yesuf, Atnafie, 2020). Glucose is the body’s carbohydrate currency. In diabetes mellitus, body cells fail to exploit this currency for the generation of energy instead of nutrient material used as an energy source. Ultimately reduction in protein storage leads to a reduction in body weight. Throughout the experimental process, STZ-induced diabetic rats showed a consistent drop in body weight. On the other hand, the oral administration of the ethanolic extract to diabetic rats marked upgrading in body weight as compared to others. Thus, obtained results proved the utilization of glucose for the generation of energy instead of protein (Pedro et al., 2004). Change in body weight The body weights of all groups of rats treated with different formulations was measured. Diabetic control TABLE IV - Effect of S. suaveolens extract on body weight TABLE IV - Effect of S. suaveolens extract on body weight Group No. Administration/ Treatment Body weight (g) 1st 10th 20th 1 Normal 200.16 ± 15.62 187.5 ± 11.48 180.83 ± 11.83 2 Diabetic control 198.5 ± 14.78 187.5 ± 12.51 174.66 ± 12.69 3 Standard 198.33 ± 15.81* 196.83 ± 17.42 197.66 ± 11.84 4 EESS 200 mg/kg 200.33 ±12.42* 181.33 ± 11.03 186.33 ± 10.57* 5 EESS 400mg/kg 195.16 ± 2.60* 184.16 ± 3.93* 185.66 ± 8.18** TABLE IV - Effect of S. suaveolens extract on body weight Page 11/17 Page 11/17 Braz. J. Pharm. Sci. 2023;59: e21820 Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma Change in blood glucose level after ethanolic extract treatment. Change in blood glucose level after ethanolic extract treatment. EESS displayed remarkably reduced blood glucose levels compared to diabetic control. After 20 days of study, STZ-induced diabetic control group showed significantly high blood glucose level (224.5±5.68 mg/dL) compared to normal control rats (84.16± 2.31 mg/dL). The blood sugar levels were analysed on the first day of the study. The STZ-induced rats’ group (II, III, IV and V) showed momentous blood sugar concentration when compared to the normal control group (I) (Table V). On the 10th day after induction, the diabetic control group exhibited a marked augment in blood glucose level. In contrast, the group treated with standard showed a considerably lower blood sugar concentration than EESS. Furthermore, groups IV and V treated with On the other hand, the group treated with standard and EESS extract at both concentrations demonstrated a noteworthy reduction in the blood glucose level compared diabetic control group. Besides, the antidiabetic activity of EESS at 400mg/kg dose was equivalent to the standard treatment. Thus, the obtained results revealed the antidiabetic potential of SS. TABLE V - Hypoglycemic activity of S. Suaveolens in diabetic animals Group No. Administration/ Treatment Blood glucose (mg/dL) 1st 10th 20th 1 Normal 81.83 ± 3.12 83.66 ± 3.26 84.16 ± 2.31 2 Diabetic control 209.83 ± 7.49 216.83 ± 7.54 224.5 ± 5.68 3 Standard 159.33 ± 25.37* 143.16 ± 18.36* 128 ± 9.95* 4 EESS 200 mg/kg 182.66 ± 3.98* 167.33 ± 3.20* 148.66 ± 3.01** 5 EESS 400mg/kg 179.66 ± 5.60* 153.5 ± 4.67* 137.66 ± 3.14* TABLE V - Hypoglycemic activity of S. Suaveolens in diabetic animals Change in the serum lipid levels the regular group. The groups treated with standard EESS at both concentrations demonstrated a remarkable reduction in total cholesterol, triglycerides, HDL- cholesterol, VLDL-cholesterol, and LDL-cholesterol. EESS showed antihyperlipidemic activity in a dose- dependent manner. The antihyperlipidemic activity of EESS at 400 mg/kg dose was at par with standard treatment. The HDL, LDL, VLDL, triglycerides, and total cholesterol levels in the serum of different groups on the last day of the experiment were measured and noted in Table VI. The diabetic control group exhibited elevated total cholesterol, triglycerides, HDL-cholesterol, VLDL-cholesterol, and LDL-cholesterol levels than Page 12/17 Braz. J. Pharm. Sci. 2023;59: e21820 TABLE VI - Effects of S. Suaveolens extract on lipid profile parameters Group No. Administration/ Treatment Concentration in mg/dL Total Cholesterol Triglycerides LDL -cholesterol VLDL- cholesterol HDL -cholesterol 1 Normal 78.33 ± 1.44 105.86 ± 1.06 41.35 ± 1.42 21.72 ± 0.21 12.78 ± 1.00 2 Diabetic control 162.56 ± 1.53 177.44 ± 1.13 87.38 ± 1.68 35.48 ± 0.22 87.56 ± 2.66 TABLE VI - Effects of S. Suaveolens extract on lipid profile parameters TABLE VI - Effects of S. Suaveolens extract on lipid profile parameters Concentration in mg/dL Braz. J. Pharm. Sci. 2023;59: e21820 tioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats TABLE VI - Effects of S. Suaveolens extract on lipid profile parameters Group No. Administration/ Treatment Concentration in mg/dL Total Cholesterol Triglycerides LDL -cholesterol VLDL- cholesterol HDL -cholesterol 3 Standard 92.62 ± 0.76** 85.63 ± 1.27* 51.33 ± 1.72* 17.12 ± 0.25** 24.15 ± 2.09* 4 Ethanol extract 200 mg/kg 121.97 ± 1.00* 140.33 ± 1.24* 64.98 ± 2.37* 28.06 ± 0.24 ** 28.92 ± 2.08* 5 Ethanol extract 400mg/kg 97.79 ± 1.17* 121.08 ± 0.96** 51.53 ± 1.95* 24.21 ± 0.19** 22.03 ± 2.86* TABLE VI - Effects of S. Suaveolens extract on lipid profile parameters Diabetes mellitus is characterised by metabolic manifestation; among these lipid metabolism affects the most. Dyslipidemia is a condition of elevated triglycerides and cholesterol levels, a warning sign for clinical manifestation of a coronary heart disease (Elberry et al., 2015). Triglycerides serves as a fuel reserve of the body and are stored in the adipose tissue. Under normal physiological conditions, triglycerides hydrolyse to free fatty acids through lipase. However, in diabetes, triglyceride level is higher than normal due to the inactivation of lipase triggered by insulin. Change in the serum lipid levels The serum TC, TG and LDL concentrations rose significantly, and HDL levels decreased. In this study, lipid biomarkers of diabetic rats showed a higher level of TC, TG and LDL, and a reduced HDL. Administration of EESS led to a decreased level of TC, TG, and LDL, and an increased level of HDL. The effect of EESS proved effective in lowering lipid biomarkers by improving their metabolism rate close to the standard metabolism rate. This finding suggests that the extract is capable of reducing dyslipidemia condition. (Seedevi et al., 2020). Defensive In-vivo antioxidant action In STZ-induced diabetic rats the level of SOD, CAT, and GSH are decreased (32.92 ± 2.45, 4.32 ± 0.08 and 7.74 ± 0.34), while the level of MDA (1.43 ± 0.04) increased in liver tissue, when compared to normal rats (14.17 ± 3.41, 3.92 ± 0.05, 5.78 ± 0.24 and 2.91 ± 0.11). Treatment of diabetes using EESS with a loading dose 200 mg and 400 mg/kg body weight revitalised the levels of all these biochemicals (SOD, CAT and GSH) to that of normal. The treatment using EESS with a 400 mg/kg dosage was found to be niftier to revitalise the levels of biochemicals, than the 200 mg/kg dosage. (Table VII). Braz. J. Pharm. Sci. 2023;59: e21820 Page 13/17 TABLE VII - Protective outcomes of EESS on antioxidant enzymes Group No. Administration/ Treatment MDA (moles MDA/mg proteins/mL) SOD (unit/ mg proteins) Catalase μmol/ mg pr.min GSH (nmol per minute per milligram) 1 Normal 1.43 ± 0.04 32.92 ± 2.45 4.32 ± 0.08 7.74 ± 0.34 2 Diabetic control 2.91 ± 0.11 14.17 ± 3.41 3.92 ± 0.05 5.78 ± 0.24 3 Standard 1.54 ± 0.04** 32.92 ± 2.92* 4.29 ± 0.06** 7.12 ± 0.49* 4 Ethanol extract 200 mg/kg 1.70 ± 0.07** 32.50 ± 4.18* 3.74 ± 0.06** 6.78 ± 0.18* 5 Ethanol extract 400mg/kg 1.57 ± 0.04** 32.92 ± 1.88* 3.89 ± 0.06** 7.29 ± 0.20* TABLE VII - Protective outcomes of EESS on antioxidant enzymes Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma Defensive effect on the pancreas Persistent hyperglycemia results in the generation of free radicals that cause a rise in oxidative stress leading damage to biomolecules, cells, and tissues, due to decreased antioxidant shielding system (Khaled et al., 2011; Dahech et al., 2011). Oxidation of lipids causes augments in the production of peroxides and free radicals due to a weakening defensive mechanism. In diabetes auto-oxidation and non-enzymatic glycation, resulting in an elevated level of oxygen free radicals, cripple SOD. Also, high glucose levels capable of inactivating glutathione peroxidase (GPx) due to glycation and severe oxidative stress are manifestations of GSH. Hydrogen peroxide (H2O2) generated by the action of SOD on superoxide is metabolized by CAT. MDA is an unstable byproduct of lipid peroxidation. Oxidative stress is responsible for elevated MDA levels, which is an indicator for the measurement of oxidative stress. The current statistics show that STZ administration caused a significant increase in oxidative stress, as seen by a reduction in antioxidant enzyme activity. In this study, EESS displayed a marked increase in MDA and a decrease in SOD, CAT, and GSH in STZ-induced diabetic rats. The antioxidant repairing mechanism is attributed to the free radical scavenging action of EESS (Sajid et al., 2020; Dworzański et al., 2020). Histo-architecture examination of the pancreas is depicted (Figure 4). The architecture of normal pancreas was well organised, while the diabetic pancreatic tissue shows degeneration in the islet due to necrosis. Also, reduction and devastation of β-cells was observed in the diabetic pancreatic cells. The standard drug glibenclamide when administered in diabetic rats repaired the normal structure of the islet. Furthermore, the administration of EESS at a dose of 200 mg/kg and 400 mg/kg to diabetic rats showed backtracked pathological changes to normal and degeneration of some islet cells. Furthermore, the histological observations of the pancreatic tissue conclude that it is protected by the SS root extract. Recovery of the Langerhans’ islet count to normal in diabetic rats with EESS may reveal the regeneration potential of the plant extract. There were findings of insulinogenic effects from different extracts of medicinal plants leading to the modulation of β-cells. An enhanced secretion of insulin from the β-cells or by rejuvenation of β-cells could be the possible cause for an antihyperglycemic effect (Prasad, Prabhu, 2012; Xue et al., 2009). Page 14/17 Braz. J. Pharm. Sci. 2023;59: e21820 FIGURE 4 - Histo- architecture of Pancreas. FIGURE 4 - Histo- architecture of Pancreas. Pancreas histopathology (A) Normal control rat pancreas showing normal islets of Langerhans with pale rounded and ovoid β-cells in the center (arrow), embedded in exocrine portion of pancreas. (B) Diabetic control rat pancreas showing shrinkage of islets of Langerhans with degeneration and necrosis of components cells where its nucleus appeared densely basophilic and karyolysis is evident (arrow) (C) Standard treated rat pancreas showing normal islets of Langerhans with pale rounded and ovoid β-cells in the center (arrow) (D) Pancreas of diabetic rat treated Ethanolic 200 mg/kg extract showing normal islets of Langerhans with its normal pale large round to ovoid shaped containing cells (arrow) that embedded in exocrine portion of pancreas. (E) Pancreas of diabetic rat treated with Ethanolic 400 mg/ kg extract showing normal sized islets of Langerhans but some degeneration of the β cell in the center were noticed (arrow). The authors wish to thank Tatyasaheb Kore College of Pharmacy and Genesis Institute of Pharmacy for extending their support throughout the research work. REFERENCES Adedayo AO, Rachel VS, Richard LB, Jessica ME, Ingrid G, Gail MW, et al. Type 2 diabetes mellitus, glycemic control, and cancer risk. Eur J Cancer Prev. 2014;23(2):134–140. Ajaib M, Khalid S, Hanif U. Spermadictyon suaveolens. A potential natural and antioxidant source. Int J Phytomed. 2016;6(2):256-267. Alimi AA, Ashafa AO. An in vitro evaluation of the antioxidant and antidiabetic potential of Sutherlandia montana E. Phillips and R.A. Dyer leaf extracts. Asian Pac J Trop Biomed. 2017;7(9):765–772. Asif M, Saleem M, Yousaf S, Saadullah M, Zafar M, Khan RU, et al. Antidiabetic activity of aqueous extract of Sigesbeckia orientalis (St. Paul’s Wort) in alloxan-induced diabetes model. Braz J Pharm Sci. 2019;55:1-10. In conclusion, in the present study we evaluated the in vitro and in vivo antioxidant and antidiabetic potential of the ethanolic extract of Spermadicyton suaveolens using biochemical parameters. The in vitro antioxidants and antidiabetic activities were investigated for all extracts, among them ethanolic extract showed more dominancy over other. The EESS exerts anti-hyperglycemic potential in STZ-induced diabetic rats, with considerable effect compared to standard glibenclamide. And the serum investigation proved the anti-hyperlipidemic potential of EESS by restoring all the parameters to normal. However, future work is needed for isolation and identification of phytoconstituents using various spectral techniques. Moreover, systemic work would be needed for exploring the molecular mechanism responsible for antidiabetic action. Bancroft JD, Gamble M. Theory and practice of histological techniques. Neuropathol Exp Neurol. 2008;67. Birdee GS, Yeh G. Complementary and alternative medicine therapies for diabetes: a clinical review. Clin Diabetes. 2018;28:147-155. Boora F, Chirisa E, Mukanganyama S. Evaluation of Nitrite radical scavenging properties of selected Zimbabwean plant extracts and their phytoconstituents. J Food Process. 2014:1-7. Cho NH, Shaw JE, Karuranga S, Huang Y, Rocha Fernandes JD, Ohlrogge AW, et al. IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract. 2018;138:271-281. Dahech I, Belghith KS, Khaled H, Fekib A, Belghith H, Mejdoub H. Antidiabetic activity of levan polysaccharide in alloxan-induced diabetic rats. Int J Biol Macromol. 2011;49(4):742-746. Diabetes facts and figures. International Diabetes Federation. IDF Diabetes Atlas 10th ed. 2021. Diabetes facts and figures. International Diabetes Federation. IDF Diabetes Atlas 10th ed. 2021. Defensive effect on the pancreas FIGURE 4 - Histo- architecture of Pancreas. Braz. J. Pharm. Sci. 2023;59: e21820 Page 14/17 Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats Ethics approval and consent to participate Dworzański J, Strycharz-Dudziak M, Kliszczewska E, Kiełczykowska M, Dworzanska A, Drop B. Glutathione peroxidase (GPx) and superoxide dismutase (SOD) activity in patients with diabetes mellitus type 2 infected with Epstein-Barr virus. Plos One. 2020;15(3):1-10. The study was permitted by the Institution Animal Ethical Committee and Rajarambapu College of Pharmacy, Kasegaon (1290/PO/Re/S/09/CPCSEA, Protocol No. RCP/P-18/18-19 Dated: 16/03/2019). Braz. J. Pharm. Sci. 2023;59: e21820 Page 15/17 Pratik Prakash Maske, Popat Sonappa Kumbhar, Ashok Gurulingappa Wali, John Intru Disouza, Maya Sharma Merghem M, Dahamna S, Khennouf S. In Vivo Antioxidant activity of Ruta montana L. Extracts. J Mater Environ Sci. 2019;10(5):470-477. Merghem M, Dahamna S, Khennouf S. In Vivo Antioxidant activity of Ruta montana L. Extracts. J Mater Environ Sci. 2019;10(5):470-477. Elberry AA, Harraz FM, Ghareib SA, Gabr SA, Nagy AA, Sattar EA. Methanolic extract of Marrubium vulgare ameliorates hyperglycemia and dyslipidemia in streptozotocin-induced diabetic rats. Int J Diabetes Mellitus. 2015;3(1):37-44. Musmade K, Rakshe A, Mokat D. Pharmacognostic studies of drug Spermadictyon Suaveolens Roxb. Int J Pharmacogn. 2016;3(5):234-239. Falguni MFZ, Islam MA, Hasan MM, Mousum SMM, Ashraduzzaman M, Khatun S. Antioxidant and antidiabetic properties of Vitex nigundo L. Leaves. Am J Life Sci. 2017;5(1):21-26. Papitha R, Ravi L, Selvaraj CI. Phytochemical studies and Gc-Ms analysis of Spermadictyon Suaveolens Rox. Int J Pharm Pharm Sci. 2017;9(3):143-149. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without the use of preparative centrifuge. Clin Chem. 1972;18(6):499–502. Patel A, Patel A, Patel A, Patel NM. Determination of polyphenols and free radical scavenging activity of Tephrosia purpurea linn leaves (Leguminosae). Pharmacogn Res. 2010;2(3):152–158. Furman BL. Streptozotocin-Induced diabetic models in mice and rats. Current Protocols. 2021;78(1):1-21. Pedro M, Montserrat B, Munoz MC, Munoz-Castaneda JR, Bujalance I, Tunez I. Protective effect of Montilla- Moriles appellation red wine on oxidative stress induced by streptozotocin in the rat. J Nutr Biochem. 2004;15(11):688– 693. Gudise V, Chowdhury B, Manjappa AS. In vitro free radical scavenging and antidiabetic activity of aqueous and ethanolic leaf extracts: a comparative evaluation of Argyreia pierreana and Matelea denticulate. Future J Pharm Sci. 2019;5(1):1-11. Prabhakar PK. Pathophysiology of Secondary Complications of Diabetes Mellitus. Asian J Pharm Clin Res. 2016;9(1):32- 36. Habu JB, Ibeh BO. In vitro antioxidant capacity and free radical scavenging evaluation of active metabolite constituents of Newbouldia laevis ethanolic leaf extract. Biol Res. 2015;48(1):16. Prasad K, Prabhu GK. Braz. J. Pharm. Sci. 2023;59: e21820 Ethics approval and consent to participate Image analysis tools for evaluation of microscopic views of immunohistochemically stained specimen in medical research–a review. J Med Systems. 2012;36(4):2621–2631. Harput US, Genc Y, Khan N, Saracoglu I . Radical scavenging effects of different Veronica Species. Rec Nat Prod. 2011;5(2):100-107. Rai PK, Srivastava AK, Sharma B, Dhar P, Mishra AK, Watal G. Use of Laser-Induced breakdown spectroscopy for the detection of glycemic elements in indian medicinal plants. Evid Based Complementary Altern Med. 2013;1-9. Hazra B, Biswas S, Mandal N. Antioxidant and free radical scavenging activity of Spondias pinnat. BMC Complement Altern Med. 2008,8:63. Sajid M, Khan MR, Ismail H, Latif S, Rahim AA, Mehboob R, et al. Antidiabetic and antioxidant potential of Alnus nitida leaves in alloxan induced diabetic rats. J Ethanopharmacol. 2020;251:112544. Khaled H, Bassem J, Serge C, Abdallah A, Abdelfattah E. Therapeutic effects of soy isoflavones on α-amylase activity, insulin deficiency, liver–kidney function and metabolic disorders in diabetic rats. Nat Prod Res. 2011;25(3):244-255. Khan SA, Kiyumi AR, Sheidi MS, Khusaibi TS, Shehhi NM, Alam T. In vitro inhibitory effects on a-glucosidase and a-amylase level and antioxidant potential of seeds of Phoenix dactylifera L. Asian Pac J Trop Biomed. 2016;6(4):322–329. Sani MF, Kouhsari SM, Moradabadi L. Effects of three medicinal plants extracts in experimental diabetes: Antioxidant enzymes activities and plasma lipids profiles in comparison with metformin. Iran J Pharm Sci. 2012;11(3):897-90. Kifle ZD, Yesuf JS, Atnafie SA. Evaluation of in vitro and in vivo Anti-Diabetic, Anti-Hyperlipidemic and Anti-Oxidant Activity of Flower Crude Extract and Solvent Fractions of Hagenia Abyssinica (Rosaceae). J Exp Pharmacol. 2020;12:151–167. Satyanarayana U, Chakrapani U, Biochemistry, fourth ed., Elsevier, 2012. Seedevi P, Ganesan AR, Moovendhan M, Mohan K, Sivasankar P, Loganathan S, et al. Anti-diabetic activity of crude polysaccharide and rhamnoseenriched polysaccharide from G. lithophila on Streptozotocin (STZ)-induced in Wistar rats. Sci Rep. 2020;10:556. Kulkarni MG, Sathe PS. Phytochemical and GC-MS analysis of Hamiltonia suaveolens (Roxb). Int J Chemtech Res. 2013;5(1):212-219. Test N0. 423; Acute Oral Toxicity- Actute Toxic Class Method. Organisation for Economic Co-operation and Developmement. OECD Publicating, Paris. 2002. Kunwar A, Priyadarsini KI .Free radicals, oxidative stress and importance of antioxidant in human health. J MedALLied Sci. 2011;1(2):53-60. Braz. J. Pharm. Sci. 2023;59: e21820 Page 16/17 This is an open-access article distributed under the terms of the Creative Commons Attribution License. Received for publication on 13th December 2021 Accepted for publication on 05th July 2022 Ethics approval and consent to participate Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic Rats Antioxidant, Antidiabetic and Lipid Profiling of Spermadicyton Suaveolens in Streptozotocin (STZ) Induced Diabetic R Telagari M, Hullatti K. In-vitro α-amylase and α-glucosidase inhibitory activity of Adiantum caudatum Linn. and Celosia argentea Linn. extracts and fractions. Indian J Pharmacol. 2015;47(4):425–429. Telagari M, Hullatti K. In-vitro α-amylase and α-glucosidase inhibitory activity of Adiantum caudatum Linn. and Celosia argentea Linn. extracts and fractions. Indian J Pharmacol. 2015;47(4):425–429. Watal G, Dhar P, Srivastava S, Sharma B. Herbal medicine as an alternative medicine for treating diabetes: The global burden. Evid Based Complementary Altern Med. 2014;1-2. Xue SX, Chen XM, Lu JX, Jin Q. Protective effect of sulphated Achyran thes bidentata polysaccharides on streptozotocin induced oxidative stress in rats. Carbohydr Polym. 2009;75(3):415–419. Uddin MS, Hossain MS, Mamun AA, Tewari D, Asaduzzaman M, Islam MS, et al. Phytochemical analysis and antioxidant profile of methanolic extract of seed, pulp and peel of Baccaurea ramiflora Lour. Asian Pac J Trop Med. 2018;11(7):443-450. Received for publication on 13th December 2021 Accepted for publication on 05th July 2022 Page 17/17 Braz. J. Pharm. Sci. 2023;59: e21820
https://openalex.org/W4387418261	https://jelectrochem.xmu.edu.cn/cgi/viewcontent.cgi?article=1303&context=journal	English	null	Development of high active aluminum alloy sacrificail anode	Deleted Journal	1,995	cc-by	117	Recommended Citation Recommended Citation Zhaobo Wei, Jianhua Wu, Guangzhang Chen, Songrao Jiang. Development of high active aluminum alloy sacrificail anode[J]. Journal of Electrochemistry, 1995 , 1(3): 339-341. DOI: 10.61558/2993-074X.1303 Available at: https://jelectrochem xmu edu cn/journal/vol1/iss3/15 Journal of Electrochemistry Journal of Electrochemistry Journal of Electrochemistry Journal of Electrochemistry Journal of Electrochemistry Journal of Electrochemistry Volume 1 Issue 3 Development of high active aluminum alloy sacrificail anode Development of high active aluminum alloy sacrificail anode Zhaobo Wei Jianhua Wu Guangzhang Chen Songrao Jiang Available at: https://jelectrochem.xmu.edu.cn/journal/vol1/iss3/15 This Article is brought to you for free and open access by Journal of Electrochemistry. It has been accepted for inclusion in Journal of Electrochemistry by an authorized editor of Journal of Electrochemistry.
https://openalex.org/W2290963622	https://ora.ox.ac.uk/objects/uuid:9bcafc29-32da-4fff-b27d-2b05d432d28a/files/m4cbe7f0223156f89c787e7540f0d93ac	English	null	Super-Resolved Traction Force Microscopy (STFM)	Nano letters	2,016	cc-by	6,072	Letter pubs.acs.org/NanoLett Super-Resolved Traction Force Microscopy (STFM) Huw Colin-York,† Dilip Shrestha,† James H. Felce,† Dominic Waithe,‡ Emad Moeendarbary,§,⊥ Simon J. Davis,† Christian Eggeling,,†,‡ and Marco Fritzsche,† †MRC Human Immunology Unit and ‡Wolfson Imaging Centre Oxford, Weatherall Institute of Molecular University of Oxford, Headley Way, OX3 9DS Oxford, United Kingdom §Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02142, United States ⊥Department of Mechanical Engineering, University College London, WC1E 7JE London, United Kingdom * S Supporting Information §Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02142, United States ⊥Department of Mechanical Engineering, University College London, WC1E 7JE London, United Kingdom * S Supporting Information * S Supporting Information ABSTRACT: Measuring small forces is a major challenge in cell bio- logy. Here we improve the spatial resolution and accuracy of force reconstruction of the well-established technique of traction force micro- scopy (TFM) using STED microscopy. The increased spatial resolution of STED-TFM (STFM) allows a greater than 5-fold higher sampling of the forces generated by the cell than conventional TFM, accessing the nano instead of the micron scale. This improvement is highlighted by computer simulations and an activating RBL cell model system. KEYWORDS: Super-resolution microscopy, traction force microscopy, mechanobiology, actin cytoskeleton ABSTRACT: Measuring small forces is a major challenge in cell bio- logy. Here we improve the spatial resolution and accuracy of force reconstruction of the well-established technique of traction force micro- scopy (TFM) using STED microscopy. The increased spatial resolution of STED-TFM (STFM) allows a greater than 5-fold higher sampling of the forces generated by the cell than conventional TFM, accessing the nano instead of the micron scale. This improvement is highlighted by computer simulations and an activating RBL cell model system. KEYWORDS: Super-resolution microscopy, traction force microscopy, mechanobiology, actin cytoskeleton KEYWORDS: Super-resolution microscopy, traction force microscopy, mechanobiology, actin cytoskeleton I reconstruction in Fourier space allowed for greatly decreased computation time.10 Owing to the inverse nature of the problem, work has shown the need for regularization to control the inﬂuence of experimental noise in the measured displacements on the traction solution.11 In addition to FTTC, other methods of traction reconstruction have been developed whereby experimental knowledge of the traction locations is used to aid the traction recovery, for example, traction reconstruction with point forces (TRPF)11 and more recently, model based traction force microscopy (MBTFM).12 These methods are only appli- cable in cases where the traction location can be inferred from ﬂuorescent data, as in the case of ﬂuorescently labeled focal adhesions. In the more general case, where no knowledge of the traction location is assumed, FTTC is more suitable and is the methodology used in this work. I t is becoming increasingly clear that mechanical force plays a crucial role in many biological processes, including adhesion, migration, and cell signaling.1−3 Forces act across many length scales, from tissue to the single cell and ultimately down to the molecular level, as is true for cells of the immune system where individual cell−cell and receptor−ligand interactions are crucial.4−6 In order to understand the role of forces within a given biological system, it is important that we have the appropriate tools and techniques that allow quantiﬁcation of mechanical forces at the relevant length scales. A commonly used technique to measure forces on the micron-scale in cell biological systems is traction force microscopy (TFM) (or simply traction microscopy). py p y py Beginning with the pioneering work of Harris et al., ﬂexible substrates, such as polyacrylamide (PAA) gels, have been used to investigate cellular tractions and forces for over 30 years.7 In a typical TFM experiment, a thin (20−30 μm) elastic gel is formed on a glass coverslip onto which proteins facilitating cell adher- ence can be attached (Figure 1a).8 Within the gel, ﬂuorescent beads serve as ﬁducial markers and imaging of the bead positions over time during the application of cellular tractions allows the displacement of the gel to be quantiﬁed. By combining the measured displacement ﬁeld with knowledge of the mechanical properties of the gel, the tractions responsible for the displace- ments can be calculated. DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 © 2016 American Chemical Society While improved analysis tools might introduce some advancements in resolving force ﬁelds (e.g., using TRPF involving knowledge from additional ﬂuorescence data of the sample, as outlined above), these approaches are experimentally still limited by the ﬁnite size of the PSF as given by diﬀraction for conventional optical microscopes (i.e., they only push the TFM read-out to its ultimate limit as given by diﬀraction). To over- come these challenges, here we improved the spatial resolution and accuracy of force reconstruction of TFM by using super- resolution optical STED microscopy.14 To examine the eﬀects of the sampling density on traction force recovery, we ﬁrst conducted computer simulations in which a gel of deﬁned stiﬀness (3 kPa) was exposed to a uniform circular traction ﬁeld (T) (0.3 kPa) of varying spatial sizes (0.1− 4.0 μm). The resulting displacement ﬁeld was then calculated using the mathematical framework provided by FTTC (Supporting Information, eq S3) (Figure 2a). To simulate the discrete nature of bead sampling, the displacement ﬁeld was subsampled at random points, with sampling densities corre- sponding to those attainable by confocal or STED microscopy: 15 beads per μm2 for high, theoretically achievable STED reso- lution (40 nm), 3 beads per μm2 for STED resolution achievable in the current experiments (80 nm), and 0.5 beads per μm2 for confocal. The subsampled displacement ﬁeld was then trans- formed back into a traction ﬁeld (Supporting Information, eq S5). The recovered and simulated traction ﬁeld were then In addition to the sampling density, another important factor in determining the accuracy of TFM is the method used to recover the displacement of the beads from the ﬂuorescent images. The most common methods of extracting bead dis- placements are those based on single particle tracking (SPT), where each individual bead must be localized, and those methods based on statistical comparisons of ﬂuorescent images, such as particle image velocimetry (PIV). To this end, the image is divided into a grid, and each grid element is spatially correlated between frames to assess the degree of movement. PIV does not require localization of each bead but is limited spatially by the size of the grid elements required to give accurate correlations. It can be shown that the increased resolution of STED allows for more accurate recovery of the displacement ﬁeld and hence a more accurate force ﬁeld in both SPT and PIV (Supporting Information, Figure S3 and Table S1). The optimal theoretical treatment of the traction solution has been the subject of much research.9−11 Dembo et al. provided a rigorous mathematical framework for the use of elastic materials to measure traction forces,9 which was further developed with the introduction of Fourier transform traction cytometry (FTTC) whose treatment of the force The greatest shortcoming of classical TFM is its limited sensitivity due to the ﬁnite density at which the displacement ﬁeld can be sampled within the gel.13 The density of ﬁducial markers must be high enough to reﬂect the complexity of the traction ﬁeld that is applied by the cell. If the bead density is too low, areas of the gel will move without being reported by any bead movement and the traction information is lost. This can be thought of as a sampling problem, where to meet the Nyquist criteria the spatial sampling frequency of the displacement ﬁeld Received: January 21, 2016 Revised: February 25, 2016 Published: February 29, 2016 2633 Nano Letters Figure 1. Theoretical characterization of STFM. (a) Schematic representation of a typical TFM setup. An elastic polyacrylamide gel ﬁlled with ﬂuorescent marker beads is covalently attached to a glass coverslip and functionalized with proteins that facilitate cell adherence. Traction forces applied by the cell to the top surface of the gel results in lateral displacements of the gel which can be quantiﬁed by imaging the displacement of the beads within the gel. (b) Theoretical relationship between the sampling density and the Nyquist limit (dashed line), with three diﬀerent bead densities highlighted (red, blue, green as labeled), exemplifying that a bead density of 15 μm−2 would allow the recovery of tractions 500 nm in size. Crosses show the smallest recoverable tractions from simulations performed at the three bead densities shown in Figure 2. Open circles show the smallest recoverable traction from simulations where noise is added and regularization used. Nano Letters Letter Figure 1. Theoretical characterization of STFM. (a) Schematic representation of a typical TFM setup. An elastic polyacrylamide gel ﬁlled with ﬂuorescent marker beads is covalently attached to a glass coverslip and functionalized with proteins that facilitate cell adherence. Traction forces applied by the cell to the top surface of the gel results in lateral displacements of the gel which can be quantiﬁed by imaging the displacement of the beads within the gel. To over- come these challenges, here we improved the spatial resolution and accuracy of force reconstruction of TFM by using super- resolution optical STED microscopy.14 = \|\| \|\| −\|\| \|\| \|\| \|\| + \|\| \|\| T T T T DTM recovered simulated recovered simulated = \|\| \|\| −\|\| \|\| \|\| \|\| + \|\| \|\| T T T T DTM recovered simulated recovered simulated By calculating the DTM for circular traction zones of varying diameters at the three diﬀerent sampling densities, it is evident that increasing the sampling density allows for the successful recovery of spatially more conﬁned tractions (Figure 2b). By adding artiﬁcial Gaussian distributed noise to the displacement ﬁeld at a level consistent with the experiment (10% of the maximum) and using the regularized solution, it is clear that the relationship between sampling density and traction recovery is maintained (Supporting Information, eq S7) (Figure 2c). For each sampling density, deﬁning a DTM of −0.2 as the minimum required to recover a traction, it is possible to plot the corre- sponding traction size on Figure 1b, which displayed good agreement between simulation and the Nyquist limit. To further demonstrate this improvement, the simulation and recovery is shown for a circular traction (0.3 kPa) of 1 μm diameter at the three diﬀerent sampling densities (Figure 2d). We also show the simulation and corresponding recovery of a more complex periodic traction ﬁeld (0−0.3 kPa) with a wavelength of 1 μm (Figure 2e). In all cases, we ﬁnd that the higher the sampling density, the more detail is recovered in the traction ﬁeld. However, due to its reliance on the spectral separation of the beads, this technique is ultimately limited by the spectral range of the microscope. Because the beads must be imaged at the top surface of the gel, this requires a microscope technique that can operate away from the coverslip, meaning TIRF or near-ﬁeld microscopy are not suitable. Cellular traction ﬁelds are typically on the nanoscale range rather than on the micron-scale. Conse- quently, there remains a need to improve the spatial resolution of TFM. (b) Theoretical relationship between the sampling density and the Nyquist limit (dashed line), with three diﬀerent bead densities highlighted (red, blue, green as labeled), exemplifying that a bead density of 15 μm−2 would allow the recovery of tractions 500 nm in size. Crosses show the smallest recoverable tractions from simulations performed at the three bead densities shown in Figure 2. Open circles show the smallest recoverable traction from simulations where noise is added and regularization used. cterization of STFM. (a) Schematic representation of a typical TFM setup. An elastic polyacrylamide gel ﬁlled with compared, and the diﬀerence quantiﬁed via a metric known as the deviation of traction magnitude (DTM), where a DTM of −1 represents a complete underestimation and 0 represents a perfect recovery of the traction.13 must be twice that of any details that may be resolved in the displacement ﬁeld (Figure 1b).13 Experimentally, this limit is imposed by the ﬁnite size point spread function (PSF) that results from each marker bead. At high densities, the PSF of each individual bead begin to overlap, meaning nearby beads can no longer be resolved individually, obscuring details of their relative displacement. A ﬁrst attempt to overcome this limitation involved the use of two diﬀerent colors of marker beads which proved that the recovery of micron sized tractions are feasible.13 However, due to its reliance on the spectral separation of the beads, this technique is ultimately limited by the spectral range of the microscope. Because the beads must be imaged at the top surface of the gel, this requires a microscope technique that can operate away from the coverslip, meaning TIRF or near-ﬁeld microscopy are not suitable. Cellular traction ﬁelds are typically on the nanoscale range rather than on the micron-scale. Conse- quently, there remains a need to improve the spatial resolution of TFM. While improved analysis tools might introduce some advancements in resolving force ﬁelds (e.g., using TRPF involving knowledge from additional ﬂuorescence data of the sample, as outlined above), these approaches are experimentally still limited by the ﬁnite size of the PSF as given by diﬀraction for conventional optical microscopes (i.e., they only push the TFM read-out to its ultimate limit as given by diﬀraction). To over- come these challenges, here we improved the spatial resolution and accuracy of force reconstruction of TFM by using super- resolution optical STED microscopy.14 must be twice that of any details that may be resolved in the displacement ﬁeld (Figure 1b).13 Experimentally, this limit is imposed by the ﬁnite size point spread function (PSF) that results from each marker bead. At high densities, the PSF of each individual bead begin to overlap, meaning nearby beads can no longer be resolved individually, obscuring details of their relative displacement. A ﬁrst attempt to overcome this limitation involved the use of two diﬀerent colors of marker beads which proved that the recovery of micron sized tractions are feasible.13 must be twice that of any details that may be resolved in the displacement ﬁeld (Figure 1b).13 Experimentally, this limit is imposed by the ﬁnite size point spread function (PSF) that results from each marker bead. At high densities, the PSF of each individual bead begin to overlap, meaning nearby beads can no longer be resolved individually, obscuring details of their relative displacement. A ﬁrst attempt to overcome this limitation involved the use of two diﬀerent colors of marker beads which proved that the recovery of micron sized tractions are feasible.13 However, due to its reliance on the spectral separation of the beads, this technique is ultimately limited by the spectral range of the microscope. Because the beads must be imaged at the top surface of the gel, this requires a microscope technique that can operate away from the coverslip, meaning TIRF or near-ﬁeld microscopy are not suitable. Cellular traction ﬁelds are typically on the nanoscale range rather than on the micron-scale. Conse- quently, there remains a need to improve the spatial resolution of TFM. While improved analysis tools might introduce some advancements in resolving force ﬁelds (e.g., using TRPF involving knowledge from additional ﬂuorescence data of the sample, as outlined above), these approaches are experimentally still limited by the ﬁnite size of the PSF as given by diﬀraction for conventional optical microscopes (i.e., they only push the TFM read-out to its ultimate limit as given by diﬀraction). DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 Note, the magnitude of 2634 DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 Letter Nano Letters . Outline of the simulation process. (a) A uniform circular traction ﬁeld Tsimulated(x) is simulated and the corresponding displacement ﬁeld d (heat map; high traction magnitude warm colors, low traction magnitudes cold colors, white arrows: traction direction). The displacem hen subsampled at a confocal and STED density (red dots: bead positions, black arrows: bead displacements), the traction ﬁeld recov (x) and the simulation and recovery compared by the deviation of traction magnitude (DTM). Scale bar 1 μm. (b) DTM for varying tra s at three sampling densities, confocal (red), medium STED (blue), and maximum STED (green). A DTM of 0 represents a perfect tra whereas a DTM of −1 represents a complete underestimation. Dotted line: DTM for no subsampling. Line deviates from zero at large trac tifacts introduced by the ﬁnite size of the simulated gel area. (c) Same as b with the addition of artiﬁcial noise and using the regularized solu very similar dependency as b except for the no subsampling case (dotted line), where regularization masks the recovered tractions at le atching that of the artiﬁcial noise. (d) Simulation and traction recovery for a 1 μm diameter circular traction zone (0.3 kPa). Scale bar 2 lation and traction recovery for a 1 μm wavelength periodic traction pattern (0−0.3 kPa). Scale bar 2 μm. Figure 2. Outline of the simulation process. (a) A uniform circular traction ﬁeld Tsimulated(x) is simulated and the corresponding displacement ﬁeld u(x) calculated (heat map; high traction magnitude warm colors, low traction magnitudes cold colors, white arrows: traction direction). The displacement ﬁeld is then subsampled at a confocal and STED density (red dots: bead positions, black arrows: bead displacements), the traction ﬁeld recovered Trecovered(x) and the simulation and recovery compared by the deviation of traction magnitude (DTM). Scale bar 1 μm. (b) DTM for varying traction diameters at three sampling densities, confocal (red), medium STED (blue), and maximum STED (green). A DTM of 0 represents a perfect traction recovery, whereas a DTM of −1 represents a complete underestimation. Dotted line: DTM for no subsampling. Line deviates from zero at large tractions due to artifacts introduced by the ﬁnite size of the simulated gel area. DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 For confocal at high bead density (lower left) no bead tracks could be resolved; instead a bar chart is shown, quantifying the ability to successfully locate and track beads in the high density confocal case compared to the high density STED case (total number of beads: 140 STED, 60 confocal). (d) Recovered traction ﬁeld for the high density STED tracking of c (left) and extrapolated low density eﬀective confocal tracking (right) with force color-coded in kPa. (e) Quantiﬁcation of the F-actin ﬂow from the high density STED recording of c by optical ﬂow (left) and correlation (color coded with 1.0 showing maximum correlation) with the bead displacement (right). Figure 3. Experimental demonstration of STFM. (a) Gel functionalization. (Left) Scheme: The roughly 30 μm thick PAA gel layer (light blue) was loaded with 40 nm-large red ﬂuorescent beads (red dots) and surface-coated with poly-L-lysine (light green) followed by attachment of IgE (green). (Middle, right) Confocal z−x proﬁle images of the gel cross-section showing concentration of Alexa488 labeled IgE (green, middle) and red ﬂuorescent beads (red, right) at the top surface of the gel. Scale bar 30 μm. (b) Representative confocal image of ﬂuorescent F-actin (Lifeact-citrine) expressing RBL cell (green) interacting with IgE coated 3 kPa PAA gel loaded with the red ﬂuorescent beads (red). Scale bar 10 μm. (c) Time-lapse imaging of the spreading cell edge results in the displacement of the beads within the gel, monitored for diﬀerent conditions as labeled. (Left panels) Confocal images of ﬂuorescent F-actin (green) and confocal or STED images of red ﬂuorescent beads (red) at a certain time point together with the temporal displacement tracks of the beads (time color-coded as labeled), for low (0.4 μm−2) and high (2.2 μm−2) bead density. Scale bar 2 μm. For confocal at high bead density (lower left) no bead tracks could be resolved; instead a bar chart is shown, quantifying the ability to successfully locate and track beads in the high density confocal case compared to the high density STED case (total number of beads: 140 STED, 60 confocal). (d) Recovered traction ﬁeld for the high density STED tracking of c (left) and extrapolated low density eﬀective confocal tracking (right) with force color-coded in kPa. (e) Quantiﬁcation of the F-actin ﬂow from the high density STED recording of c by optical ﬂow (left) and correlation (color coded with 1.0 showing maximum correlation) with the bead displacement (right). (c) Same as b with the addition of artiﬁcial noise and using the regularized solution, showing very similar dependency as b except for the no subsampling case (dotted line), where regularization masks the recovered tractions at length scales matching that of the artiﬁcial noise. (d) Simulation and traction recovery for a 1 μm diameter circular traction zone (0.3 kPa). Scale bar 2 μm. (e) Simulation and traction recovery for a 1 μm wavelength periodic traction pattern (0−0.3 kPa). Scale bar 2 μm. PIV recovered high bead density based force ﬁelds can be close to the simulated values but increased spatial accuracy necessitates STED. 3 kPa gel loaded with 40 nm red-ﬂuorescent beads and coated with the antibody, IgE (Figure 3a). The Fc portion of IgE binds with high aﬃnity (equilibrium constant Ka = 1010 M−1, oﬀ-rate koff = 10−5 s−1) to the FcεRI present on the RBL cell surface, which results in cell spreading and activation.15,16 By ﬂuo- rescently labeling actin ﬁlaments via Lifeact-citrine, we were able To observe the eﬀect of increased sampling experimentally in the single cell environment, RBL cells, which express high levels of the Fcε receptor-1 (FcεRI), were allowed to interact with a 2635 DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 Letter Figure 3. Experimental demonstration of STFM. (a) Gel functionalization. (Left) Scheme: The roughly 30 μm thick PAA gel layer (light blue) was loaded with 40 nm-large red ﬂuorescent beads (red dots) and surface-coated with poly-L-lysine (light green) followed by attachment of IgE (green). (Middle, right) Confocal z−x proﬁle images of the gel cross-section showing concentration of Alexa488 labeled IgE (green, middle) and red ﬂuorescent beads (red, right) at the top surface of the gel. Scale bar 30 μm. (b) Representative confocal image of ﬂuorescent F-actin (Lifeact-citrine) expressing RBL cell (green) interacting with IgE coated 3 kPa PAA gel loaded with the red ﬂuorescent beads (red). Scale bar 10 μm. (c) Time-lapse imaging of the spreading cell edge results in the displacement of the beads within the gel, monitored for diﬀerent conditions as labeled. (Left panels) Confocal images of ﬂuorescent F-actin (green) and confocal or STED images of red ﬂuorescent beads (red) at a certain time point together with the temporal displacement tracks of the beads (time color-coded as labeled), for low (0.4 μm−2) and high (2.2 μm−2) bead density. Scale bar 2 μm. DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 ■ASSOCIATED CONTENT * S Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b00273. Materials and methods, a description of the bead tracking software, additional simulations comparing bead tracking algorithms and force recovery, and a description of the ﬂuorescence-force correlation analysis (PDF) Finally, to identify the origin of mechanical force generation in RBL cells, we combined ﬂuorescent imaging and STFM. The technique of optical ﬂow enables the spatial change in pixel intensity to be quantiﬁed over time and is commonly applied in computer vision to assess the spatial shift between image frames. Here, we apply this technique to extract the retrograde ﬂow vector ﬁeld of ﬂuorescently labeled ﬁlamentous actin (Lifeact- citrine). This vector ﬁeld was then correlated via a dot product with the displacement ﬁeld of the beads in the gel, yielding a spatial correlation of actin and bead displacement (Figure 3e).17 The correlation highlights that areas of the cell showing the most dynamic actin coincide with the areas of greatest bead displace- ment, highlighting that it was indeed the ﬂow of actin within the cell that was responsible for the observed tractions. As in the case of traction forces, the higher bead density leads to a higher information content in the displacement ﬁeld and hence a more reliable correlation between the two vector ﬁelds. Representative large ﬁeld-of-view confocal time-lapse movie of an RBL cell expressing Lifeact-citrine (green) spreading on an IgE coated 3 kPa PAA gel loaded with 40 nm red ﬂuorescent beads (red) at a low density (0.4 μm−2). The beads are seen to move beneath the cell, indicating that forces are being transferred by the cell to the gel. Scale bar 10 μm. Frame rate 0.2 s−1. Total acquisition time 200 s (AVI). Representative large ﬁeld-of-view confocal time-lapse movie of an RBL cell expressing Lifeact-citrine (green) spreading on an IgE coated 3 kPa PAA gel loaded with 40 nm red ﬂuorescent beads (red) at a low density (0.4 μm−2). The beads are seen to move beneath the cell, indicating that forces are being transferred by the cell to the gel. Scale bar 10 μm. Frame rate 0.2 s−1. Total acquisition time 200 s (AVI). Representative close-up confocal time-lapse movie of a RBL cell expressing Lifeact-citrine (green) spreading on an IgE coated 3 kPa PAA gel loaded with 40 nm red ﬂuorescent beads (red) at a low density (0.4 μm−2). ■ASSOCIATED CONTENT Again, the beads are seen to move beneath the cell, indicating that forces are being transferred by the cell to the gel. Scale bar 2 μm. Frame rate 0.2 s−1. Total acquisition time 120 s (AVI). In summary, the increase in accuracy of STFM is important when considering cellular forces on small length scales, as is the case for receptor-antagonist interactions in immune cells.18 Using STFM, we are now better able to make links between the forces generated by the cell, and those molecules which are responsible for force generation. This is particularly valuable when force measurements are coupled to ﬂuorescent data, as is shown in Figure 3e. Equivalent time-lapse movie to that shown in Supple- mentary Movie S2, this time imaged with STED microscopy in the case of the beads. Scale bar 2 μm. Frame rate 0.05 s−1. Total acquisition time 120 s (AVI). Representative close-up confocal time-lapse movie of a RBL cell expressing Lifeact-citrine (green) spreading on an IgE coated 3 kPa PAA gel loaded with 40 nm red ﬂuorescent beads (red) at a high bead density (2.2μm−2). A signicant number of beads are seen to overlap, preventing their reliable tracking. Scale bar 2 μm. Frame rate 0.06 s−1. Total acquisition time 48 s (AVI). Representative close-up confocal time-lapse movie of a RBL cell expressing Lifeact-citrine (green) spreading on an IgE coated 3 kPa PAA gel loaded with 40 nm red ﬂuorescent beads (red) at a high bead density (2.2μm−2). A signicant number of beads are seen to overlap, preventing their reliable tracking. Scale bar 2 μm. Frame rate 0.06 s−1. Total acquisition time 48 s (AVI). We have focused on the experimental aspects of TFM. Moreover, recent work has suggested that theoretical aspects may be equally important in optimizing the accuracy of force reconstruction.19 For example, in the case of sparse focal adhesions, it has been demonstrated that the L1-norm is favorable to the L2-norm used to assess to degree of regula- rization,20,21 owing to a greater retention of the high resolution detail in the force maps. We had no prior knowledge of the traction ﬁeld induced by the RBL cells, therefore we choose L2-norm regularization as the more general case. However, further work should focus on optimizing both experimental and computational aspects of the technique to further increase the accuracy of STFM. confocal and STED (2.2 beads μm−2) (Figure 3c) (Movies S4 and S5). Here we deﬁne low density as the maximum trackable density by confocal, and high density as the maximum trackable density of our current STED experiments. Note, bead sampling values in the scenario of high density STED were a moderate and robust choice considering the experimental optical conditions and needs of the biological specimen. However, they were below the computationally predicted possible advances of STFM. Speciﬁ- cally, the bead densities are a function of the microscopes PSF size and could be improved in future work by optimizing the imaging conditions, e.g., minimizing optical aberrations (see discussion). to visualize the dynamically spreading cell edge as the cell-gel contact area increased (Figure 3b). On visual inspection (Movie S1), the beads within this area are seen to move elastically in a directed manner toward the direction of cell motion, indicating forces being applied by the cell to the compliant gel beneath via the FcεRI-IgE interaction. By using a water immersion objective, optical aberrations from imaging through the gel layer were low, allowing STED imaging with a spatial resolution of around 80 nm (Figure S1), and we have observed no visible sign of cell degradation (such as cell contraction) during the around 120 s long recordings. To assess the displacement of the beads we chose to use SPT as it allows the movements of each individual bead to be cap- tured. PIV is generally best suited to tractions where collective Next, to demonstrate the inﬂuence of the bead sampling density, we introduce four diﬀerent scenarios; low density confocal and STED (0.4 beads μm−2) (Movies S2 and S3), and high density 2636 DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 Nano Letters Letter the beads, and ﬂuorescence light sensitivity of the biological specimen. Moreover, the nature of the (S)TFM setup requires all imaging to be done at the top surface of the gel, meaning imaging is subject to aberrations induced by the mismatch in refractive index of the gel and the immersion media. Improvements in aberration correction, for example using adaptive optics would reduce the eﬀect of these aberrations and would result in an improved STED resolution, possibly along all three spatial dimensions,22 allowing even higher bead densities to be used and the accuracy of experiments to approach those shown possible by simulations. This also presents the opportunity of performing 3D-TFM in high resolution. In the same way that 2D-STED can increase the accuracy of the tangential force reconstruction, using 3D-STED would allow for a greater sampling of the forces perpendicular to the gel surface. bead movements are expected, for example focal adhesions. For RBL cells, forces may arise from localized receptor−ligand inter- actions and may be spatially complex. In the low density case, applying a custom written MATLAB SPT algorithm allowed the displacements of all beads within the ﬁeld of view to be measured in both the confocal and STED image sequences (Supporting Information and Figure S2). In the high density case, confocal imaging resulted in a signiﬁcant number of overlapping PSFs, preventing reliable bead tracking. However, on applying the STED beam, beads were resolved individually and the tracking was successful (Figure 3c). In all cases, bead tracks were inter- polated onto a regular mesh and the corresponding traction ﬁeld calculated using the appropriate degree of regularization (Figure S4). Obviously, in the high density case only the STED imaging yielded a traction ﬁeld (Figure 3d). To directly compare the eﬀect of sampling density on the ability to accurately recover the traction ﬁeld of the same cell, beads in the high density STED case were randomly deleted until the bead density was equal to that attainable by confocal tracking (Figure 3d). It is clear that this reduces the information content present in the displacement ﬁeld, and hence reduces the ﬁne detail in the traction ﬁeld. ■ACKNOWLEDGMENTS We greatly acknowledge support from the Wolfson Imaging Centre (Christoﬀer Lagerholm and Esther Garcia), general lab support by Sumita Ganguly, and funding by the Medical Research Council (grant number MC_UU_12010/unit pro- grammes G0902418 and MC_UU_12025), MRC/BBSRC/ EPSRC (grant number MR/K01577X/1), Wellcome Trust (grant ref 104924/14/Z/14), and Wolfson Foundation. Notes Notes The authors declare no competing ﬁnancial interest. ■REFERENCES (1) Trepat, X.; Wasserman, M. R.; Angelini, T. E.; Millet, E.; Weitz, D. A.; Butler, J. P.; Fredberg, J. J. Nat. Phys. 2009, 5 (6), 426−430. (1) Trepat, X.; Wasserman, M. R.; Angelini, T. E.; Millet, E.; Weitz, D. A.; Butler, J. P.; Fredberg, J. J. Nat. Phys. 2009, 5 (6), 426−430. (2) Plotnikov, S. V.; Pasapera, A. M.; Sabass, B.; Waterman, C. M. Cell 2012, 151 (7), 1513−1527. (2) Plotnikov, S. V.; Pasapera, A. M.; Sabass, B.; Waterman, C. M. Cell 2012, 151 (7), 1513−1527. (3) Iskratsch, T.; Wolfenson, H.; Sheetz, M. P. Nat. Rev. Mol. Cell Biol. 2014, 15 (12), 825−833. (3) Iskratsch, T.; Wolfenson, H.; Sheetz, M. P. Nat. Rev. Mol. Cell Biol. 2014, 15 (12), 825−833. (4) Nelson, C. M.; Jean, R. P.; Tan, J. L.; Liu, W. F.; Sniadecki, N. J.; Spector, A. A.; Chen, C. S. Proc. Natl. Acad. Sci. U. S. A. 2005, 102 (33), 11594−11599. (5) Vogel, V.; Sheetz, M. Nat. Rev. Mol. Cell Biol. 2006, 7 (4), 265−275. (6) Tabdanov, E.; Gondarenko, S.; Kumari, S.; Liapis, A.; Dustin, M. L.; Sheetz, M. P.; Kam, L. C.; Iskratsch, T. Integr. Biol. 2015, 7, 1272. (7) Harris, A.; Wild, P.; Stopak, D. Science 1980, 208 (4440), 177−179. (8) M k l R Ki h ß N C C M H ff B Bi h J (5) Vogel, V.; Sheetz, M. Nat. Rev. Mol. Cell Biol. 2006, 7 (4), 265−275. (6) Tabdanov, E.; Gondarenko, S.; Kumari, S.; Liapis, A.; Dustin, M. L.; Sheetz, M. P.; Kam, L. C.; Iskratsch, T. Integr. Biol. 2015, 7, 1272. ( ) ( ) , ; , ; , g , , (7) Harris, A.; Wild, P.; Stopak, D. Science 1980, 208 (4440), 177−179. ( ) (8) Merkel, R.; Kirchgeßner, N.; Cesa, C. M.; Hoffmann, B. Biophys. J. 2007, 93 (9), 3314−3323. (9) Dembo, M.; Wang, Y. L. Biophys. J. 1999, 76 (4), 2307−2316. (9) Dembo, M.; Wang, Y. L. Biophys. J. 1999, 76 (4), 2307−2316. (10) Butler, J. P.; Tolić-Nørrelykke, I. M.; Fabry, B.; Fredberg, J. J. Am. h l ll h l ( ) ( ) g p y J ( ) (10) Butler, J. P.; Tolić-Nørrelykke, I. M.; Fabry, B.; Fredberg, J. J. Am. J. Physiol Cell Physiol. 2002, 282 (3), C595−C605. ( ) J y y g J J J. Physiol Cell Physiol. 2002, 282 (3), C595−C605. J. Physiol Cell Physiol. 2002, 282 (3), C595−C605. DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 ■ASSOCIATED CONTENT q Equivalent time-lapse movie to that shown in Supple- mentary Movie S4, this time imaged with STED microscopy in the case of the beads. The improved spatial resolution made possible using STED now means that all beads in the ﬁeld of view are trackable. Scale bar 2 μm. Frame rate 0.05 s−1. Total acquisition time 120 s (AVI). Notably, increasing the location accuracy of STFM is theoretically not limited since it scales with the applied STED power.14 This needs to be balanced with other optical factors such as maximal bead density within the gel, the ability to track E-mail: christian.eggeling@rdm.ox.ac.uk. E-mail: marco.fritzsche@rdm.ox.ac.uk. DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638 2637 Letter Letter 52 (2), 489−509. ■REFERENCES (11) Schwarz, U. S.; Balaban, N. Q.; Riveline, D.; Bershadsky, A.; Geiger, B.; Safran, S. A. Biophys. J. 2002, 83 (3), 1380−1394. g (12) Soiné, J. R. D.; Brand, C. A.; Stricker, J.; Oakes, P. W.; Gardel, M. (12) Soiné, J. R. D.; Brand, C. A.; Stricker, J.; Oakes, P. W.; Gardel, M. L.; Schwarz, U. S. PLoS Comput. Biol. 2015, 11 (3), e1004076. (12) Soine, J. R. D.; Brand, C. A.; Stricker, J.; Oakes, P. W.; Gard L.; Schwarz, U. S. PLoS Comput. Biol. 2015, 11 (3), e1004076. L.; Schwarz, U. S. PLoS Comput. Biol. 2015, 11 (3), e10040 (13) Sabass, B.; Gardel, M. L.; Waterman, C. M.; Schwarz, U. S. Biophys. J. 2008, 94 (1), 207−220. p y J ( ) (14) Hell, S. W.; Wichmann, J. Opt. Lett. 1994, 19 (11), 780−782. ( ) ( ) (15) Gould, H. J.; Sutton, B. J. Nat. Rev. Immunol. 2008, 8 (3), 205− 217. (16) Kitaura, J.; Song, J.; Tsai, M.; Asai, K.; Maeda-Yamamoto, M.; Mocsai, A.; Kawakami, Y.; Liu, F.-T.; Lowell, C. A.; Barisas, B. G.; Galli, (16) Kitaura, J.; Song, J.; Tsai, M.; Asai, K.; Maeda-Yamamoto, M.; Mocsai, A.; Kawakami, Y.; Liu, F.-T.; Lowell, C. A.; Barisas, B. G.; Galli, S. J.; Kawakami, T. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (22), 12911− 12916. S. J.; Kawakami, T. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (22), 12911− 12916. (17) Liu, C. Beyond Pixels: Exploring New Representations and Applications for Motion Analysis; MIT, 2009. pp f y (18) Kumari, S.; Dustin, M. L. Curr. Biol. 25, (10), R413−R (19) Schwarz, U. S.; Soiné, J. R. D. Biochim. Biophys. Acta, Mol. Cell Res. 2015, 1853 (11, Part B), 3095−3104. (20) Han, S. J.; Oak, Y.; Groisman, A.; Danuser, G. Nat. Methods 2015, 12 (7), 653−656. (21) Candes, E. J.; Romberg, J.; Tao, T. IEEE Trans. Inf. Theory 2006, 52 (2), 489−509. (22) Gould, T. J.; Kromann, E. B.; Burke, D.; Booth, M. J.; Bewersdorf, J. Opt. Lett. 2013, 38 (11), 1860−1862. 2638 DOI: 10.1021/acs.nanolett.6b00273 Nano Lett. 2016, 16, 2633−2638
https://openalex.org/W2806899758	https://europepmc.org/articles/pmc6025552?pdf=render	English	null	Seebeck Coefficient of Thermocouples from Nickel-Coated Carbon Fibers: Theory and Experiment	Materials	2,018	cc-by	7,967	Received: 18 April 2018; Accepted: 25 May 2018; Published: 30 May 2018 Abstract: Thermocouples made of etched and non-etched nickel-coated carbon yarn (NiCCY) were investigated. Theoretic Seebeck coefﬁcients were compared to experimental results from measurements of generated electric voltage by these thermocouples. The etching process for making thermocouples was performed by immersion of NiCCY in the solution containing a mixture of hydrochloric acid (HCl) (37% of concentration), and hydrogen peroxide (H2O2) in three different concentrations—3%, 6%, and 10%. Thirty minutes of etching to remove Ni from NiCCY was followed by washing and drying. Next, the ability to generate electrical voltage by the thermocouples (being a junction of the etched and the non-etched NiCCY) was measured in different ranges of temperatures, both a cold junction (291.15–293.15 K) and a hot junction (293.15–325.15 K). A formula predicting the Seebeck coefﬁcient of this thermocouple was elaborated, taking into consideration resistance values of the tested samples. It was proven that there is a good agreement between the theoretical and experimental data, especially for the yarns etched with 6% and 10% peroxide (both were mixed with HCl). The electrical resistance of non-fully etched nickel remaining on the carbon ﬁber surface (R1) can have a signiﬁcant effect on the thermocouples’ characteristics. Keywords: thermocouple; Seebeck coefﬁcient; conductive yarn; nickel-coated carbon ﬁber Seebeck Coefﬁcient of Thermocouples from Nickel-Coated Carbon Fibers: Theory and Experiment H di t H di t 1 2 * ID Gilb t D M 3 I b l Ci i l k W ób l 1 C l H tl 1 d 1 Department of Materials, Textiles and Chemical Engineering, Ghent University, Technologiepark 907, 9052 Zwijnaarde, Belgium; budysiowka@gmail.com (I.C.-W.); carla.hertleer@gmail.com (C.H.); Lieva.VanLangenhove@UGent.be (L.V.L.) 1 Department of Materials, Textiles and Chemical Engineering, Ghent University, Technologiepark 907, 9052 Zwijnaarde, Belgium; budysiowka@gmail.com (I.C.-W.); carla.hertleer@gmail.com (C.H.); Lieva.VanLangenhove@UGent.be (L.V.L.) 1 Department of Materials, Textiles and Chemical Engineering, Ghent University, Technologiepark 907, 9052 Zwijnaarde, Belgium; budysiowka@gmail.com (I.C.-W.); carla.hertleer@gmail.com (C.H.); Lieva.VanLangenhove@UGent.be (L.V.L.) 1 Department of Materials, Textiles and Chemical Engineering, Ghent University, Technologiepark 907, 9052 Zwijnaarde, Belgium; budysiowka@gmail.com (I.C.-W.); carla.hertleer@gmail.com (C.H.); Lieva.VanLangenhove@UGent.be (L.V.L.) 2 Department of Textile Chemistry, Politeknik STTT Bandung, Jalan Jakarta 31, Bandung 40272, Indonesia 3 Department of Electronics and Information Systems, Ghent University, Technologiepark 15, 9052 Zwijnaarde, Belgium; Gilbert.DeMey@UGent.be * C d h di h di @ b 2 Department of Textile Chemistry, Politeknik STTT Bandung, Jalan Jakarta 31, Bandung 40272, Indonesia 3 Department of Electronics and Information Systems, Ghent University, Technologiepark 15, 9052 Zwijnaarde, Belgium; Gilbert.DeMey@UGent.be 2 Department of Textile Chemistry, Politeknik STTT Bandung, Jalan Jakarta 31, Bandung 40272, Indo 3 Department of Electronics and Information Systems, Ghent University, Technologiepark 15, 9052 Zwijnaarde, Belgium; Gilbert.DeMey@UGent.be 2 Department of Textile Chemistry, Politeknik STTT Bandung, Jalan Jakarta 31, Bandung 40272, Indonesia 3 Department of Electronics and Information Systems, Ghent University, Technologiepark 15, 9052 Zwijnaarde, Belgium; Gilbert.DeMey@UGent.be * Correspondence: hardiant.hardianto@ugent.be materials materials materials Materials 2018, 11, 922; doi:10.3390/ma11060922 1. Introduction The toxicity of carbon nanotubes is also a problem. Therefore, nickel-coated carbon yarn (NiCCY) can be a good candidate material for fabricating a textile-based thermoelectric generator. From our preliminary experiment, a thermocouple from two different conductive textile yarns i.e., carbon ﬁber and NiCCY demonstrated a Seebeck coefﬁcient of about 18 µV/K. According to this result, there is a possibility to create a thermopile using NiCCY. Carbon ﬁber coated with nickel is also available on the market. This led us to an idea to create a textile-based thermopile from a single NiCCY, provided that the nickel can be removed selectively to form a series of C-Ni junctions along the NiCCY forming a thermopile to achieve higher voltage output since the voltage generated by a thermopile is proportional to the number of thermocouple junctions. The previous report showed that mixture of peroxide and hydrochloric acid can signiﬁcantly remove the nickel from nickel-coated carbon ﬁber [13]. Textile-based conductive yarns are sought for their integration into the textile structures to act as thermocouples or thermopiles, to create more ﬂexible textile-based temperature- or heat ﬂux sensors. thermocouples or thermopiles, to create more ﬂexible textile-based temperature- or heat ﬂux sensors. There is a great potential for making thermocouples based on existing carbon yarns, especially with Tenax®-J HTS40 A23 12K 1420tex provided by Toho Tenax Europe GmbH, Germany, which has a Ni coating [13]. In this nickel-coated carbon yarn (NiCCY), both C and Ni are electric conductors. Tenax® refers to a group of high-performance carbon ﬁbers. The high tenacity (HT) ﬁbers provide excellent mechanical laminate properties. This ﬁber is manufactured from a polyacrylonitrile (PAN) precursor and is surface-treated in order to promote adhesion to organic matrix polymers [14]. In this paper, we are going to study the effect of the electrical resistance of the etched NiCCY on the Seebeck coefﬁcient of the thermocouple made of the etched NiCCY and non-etched because it is easier to measure the electrical resistance than to measure the thickness of the nickel layer on NiCCY which is less than 0.25 µm. A formula for predicting the Seebeck coefﬁcient of this thermocouple is going to be elaborated by taking into consideration the resistance values of the yarns. This theoretical formula will be compared to the experimental results. The electric conductivity σ of the nickel is much higher compared to carbon (σNi = 14.3 × 106 S/m vs. 1. Introduction Smart textiles development has entered a new stage where all the electronic components previously incorporated with textile elements are gradually replaced by electronic-like textiles components, e.g., textile capacitors [1], highly ﬂexible textile antennas [2]. This means that electronic components are usually replaced by pliable and limp ﬁlms, yarns, ﬁbers, conductive coatings, printed electronics, etc., to make these new smart textiles stretchable, washable, durable and lasting. These last three features have been playing an important role in smart textiles release to the market and the fact that these three conditions have not been fulﬁlled is exactly why one may not observe all the interesting innovative smart textiles solutions on the market. This research work is aimed at building a reliable textile thermocouple on a linear textile product (yarn) that could be incorporated as weft, with a thin fabric being a stratum. This stratum that separates two zones provides information on measurements of temperature differences between these two zones. In order to create a textile thermocouple, a junction of two conductive yarns has to be created. In fact, any thermocouple can be created by making a proper circuit with two different conductors that Materials 2018, 11, 922; doi:10.3390/ma11060922 www.mdpi.com/journal/materials Materials 2018, 11, 922 2 of 10 allow the generation of an electric voltage. This electric voltage is known as the Seebeck effect (S). Its efﬁciency depends on a ratio of the generated electric voltage and a temperature difference between environments where each of the conductors’ join [3]. There are several known solutions attempting textile application of the thermocouple principal presented by other authors, e.g., using constantan with steel to create a thermocouple [4], and constantan with copper for the same purpose [5]. Thermocouples were integrated into textile substrates by researchers as a temperature sensor [6] and heat ﬂux sensor [7–9]. However, there is a disadvantage to the application of these sorts of materials because they are brittle and consequently break easily. Additionally, metal wires integrated into textiles as thermocouples affect the ﬂexibility of these textiles. Therefore, ﬁnding textile-based conductive yarns for fabricating a thermocouple and a thermopile for wearable textiles is of great interest. Recently, huge improvements of the Seebeck coefﬁcient have been found by using other forms of carbon such as carbon nanotube and graphene [10–12]. However, these promising materials are not available in textile yarns which can be inserted into wearable textiles. 1. Introduction Theoretical diagram of etched and non-etched segment of one filament of nickel-coated carbon fiber as a thermocouple where C is carbon fiber; Ni is nickel coating on the surface of C fiber; S, S1, and S2 are the cross-section areas of C, Ni on etched segment and Ni on non-etched segment, i l T i T d T ld d h i l ϕ i l i , , Materials 2018, 11, x FOR PEER REVIEW 3 of 10 Figure 1. Theoretical diagram of etched and non-etched segment of one filament of nickel-coated carbon fiber as a thermocouple where C is carbon fiber; Ni is nickel coating on the surface of C fiber; S, S1, and S2 are the cross-section areas of C, Ni on etched segment and Ni on non-etched segment, respectively; T is temperature; TC and TH are cold and hot temperatures, respectively; ϕ is electric potential; Vj is a voltage at junction (x = 0); V0 is a voltage at a single end of nickel-coated carbon fiber (x = b); x is axis along the thermocouple. Theoretically the most efficient thermocouple is created if a significant part or the whole of the Figure 1. Theoretical diagram of etched and non-etched segment of one ﬁlament of nickel-coated carbon ﬁber as a thermocouple where C is carbon ﬁber; Ni is nickel coating on the surface of C ﬁber; S, S1, and S2 are the cross-section areas of C, Ni on etched segment and Ni on non-etched segment, respectively; T is temperature; TC and TH are cold and hot temperatures, respectively; ϕ is electric potential; Vj is a voltage at junction (x = 0); V0 is a voltage at a single end of nickel-coated carbon ﬁber (x = b); x is axis along the thermocouple. Figure 1. Theoretical diagram of etched and non-etched segment of one filament of nickel-coated carbon fiber as a thermocouple where C is carbon fiber; Ni is nickel coating on the surface of C fiber; S, S1, and S2 are the cross-section areas of C, Ni on etched segment and Ni on non-etched segment, layer of Ni is removed from the surface of CF during the etching process. 1. Introduction During the etching process, it was possible to remove an unknown amount of Ni so that when the periodically etched yarn and the non-etched yarn were connected to the voltage meter, the junction allowed for the generation of an output voltage of about 18 µV/K. The etching process and the subsequent measurements were performed on the whole yarn (all the fibers in the yarn were treated at the same time). It would be difficult to treat a single fiber individually. Due to the higher electrical conductivity of Ni, the right part of the fiber section might seem to be made from pure nickel and the presence of carbon may be overlooked. Occasionally, it might happen that not all the nickel has been removed from the part that was intentionally subject to etching so that the performance of the thermocouple might be reduced. 2. Materials and Methods 2.1. Material Theoretically, the most efﬁcient thermocouple is created if a signiﬁcant part or the whole of the layer of Ni is removed from the surface of CF during the etching process. During the etching process, it was possible to remove an unknown amount of Ni so that when the periodically etched yarn and the non-etched yarn were connected to the voltage meter, the junction allowed for the generation of an output voltage of about 18 µV/K. The etching process and the subsequent measurements were performed on the whole yarn (all the ﬁbers in the yarn were treated at the same time). It would be difﬁcult to treat a single ﬁber individually. Due to the higher electrical conductivity of Ni, the right part of the ﬁber section might seem to be made from pure nickel and the presence of carbon may be overlooked. Occasionally, it might happen that not all the nickel has been removed from the part that was intentionally subject to etching so that the performance of the thermocouple might be reduced. potential; Vj is a voltage at junction (x = 0); V0 is a voltage at a single end of nickel-coated carbon fiber (x = b); x is axis along the thermocouple. Theoretically, the most efficient thermocouple is created if a significant part or the whole of the layer of Ni is removed from the surface of CF during the etching process. 1. Introduction σC = 5.9 × 106 S/m) [15]. In order to make a thermocouple, we have to use two different materials. In this work, this will be performed by etching the Ni layers from the part of the NiCCY, as schematically represented in Figure 1. Both ends (x = −a and x = b) are at the “cold” temperature, TC, which is normally the ambient room temperature. The junction (x = 0) between the etched and the non-etched part is at a higher temperature, TH. At both cold ends, metallic contacts are provided to measure the generated voltage, V0. Figure 1 demonstrates an idealised situation in which a single ﬁber is partially etched, thus having two different amounts of Ni on its surface. The idealised and simpliﬁed situation where two different thicknesses of Ni on carbon ﬁber (CF) are presented, makes the mathematical modeling of a theoretical approach transparent. 3 of 10 Materials 2018, 11, 922 Materials 2018, 11, x FOR PEER REVIEW 3 of 10 Figure 1. Theoretical diagram of etched and non-etched segment of one filament of nickel-coated carbon fiber as a thermocouple where C is carbon fiber; Ni is nickel coating on the surface of C fiber; S, S1, and S2 are the cross-section areas of C, Ni on etched segment and Ni on non-etched segment, respectively; T is temperature; TC and TH are cold and hot temperatures, respectively; ϕ is electric potential; Vj is a voltage at junction (x = 0); V0 is a voltage at a single end of nickel-coated carbon fiber (x = b); x is axis along the thermocouple. Theoretically, the most efficient thermocouple is created if a significant part or the whole of the Figure 1. Theoretical diagram of etched and non-etched segment of one ﬁlament of nickel-coated carbon ﬁber as a thermocouple where C is carbon ﬁber; Ni is nickel coating on the surface of C ﬁber; S, S1, and S2 are the cross-section areas of C, Ni on etched segment and Ni on non-etched segment, respectively; T is temperature; TC and TH are cold and hot temperatures, respectively; ϕ is electric potential; Vj is a voltage at junction (x = 0); V0 is a voltage at a single end of nickel-coated carbon ﬁber (x = b); x is axis along the thermocouple. Materials 2018, 11, x FOR PEER REVIEW 3 of 10 Figure 1. 1. Introduction During the etching process, it was possible to remove an unknown amount of Ni so that when the periodically etched yarn and the non-etched yarn were connected to the voltage meter, the junction allowed for the generation of an output voltage of about 18 µV/K. The etching process and the subsequent measurements were performed on the whole yarn (all the fibers in the yarn were treated at the same time). It would be difficult to treat a single fiber individually. Due to the higher electrical conductivity of Ni, the right part of the fiber section might seem to be made from pure nickel and the presence of carbon may be overlooked. Occasionally, it might happen that not all the nickel has been removed from the part that Germa 2.1. Material 2. Mate 2 1 M Materials used in this experiment were similar to previous work [13]. The NiCCY used in this experiment is Tenax®-J HTS40 A23 12K 1420tex MC from Toho Tenax Europe GmbH, Wuppertal, Germany. Figure 2 shows the yarn and Table 1 presents its parameters. 2.1. Material Materials used in this experiment were similar to previous work [13]. The NiCCY used in this experiment is Tenax®-J HTS40 A23 12K 1420tex MC from Toho Tenax Europe GmbH, Wuppertal, Germany. Figure 2 shows the yarn and Table 1 presents its parameters. Figure 2. An image of Tenax®-J HTS40 1420tex NiCCY provided by Toho Tenax Europe GmbH, Wuppertal, Germany. The image was taken with OneBird Smart 5M 300X USB Digital Microscope Camera Video with MicroCapture. The diameter of 6.906 mm was measured without any pretension [14]. Figure 2. An image of Tenax®-J HTS40 1420tex NiCCY provided by Toho Tenax Europe GmbH, Wuppertal, Germany. The image was taken with OneBird Smart 5M 300X USB Digital Microscope Camera Video with MicroCapture. The diameter of 6.906 mm was measured without any pretension [14]. Figure 2. An image of Tenax®-J HTS40 1420tex NiCCY provided by Toho Tenax Europe GmbH, Wuppertal, Germany. The image was taken with OneBird Smart 5M 300X USB Digital Microscope Camera Video with MicroCapture. The diameter of 6.906 mm was measured without any pretension [14]. Figure 2. An image of Tenax®-J HTS40 1420tex NiCCY provided by Toho Tenax Europe GmbH, Wuppertal, Germany. The image was taken with OneBird Smart 5M 300X USB Digital Microscope Camera Video with MicroCapture. The diameter of 6.906 mm was measured without any pretension [14]. Figure 2. An image of Tenax®-J HTS40 1420tex NiCCY provided by Toho Tenax Europe GmbH, Wuppertal, Germany. The image was taken with OneBird Smart 5M 300X USB Digital Microscope Camera Video with MicroCapture. The diameter of 6.906 mm was measured without any pretension [14]. 4 of 10 4 of 10 Materials 2018, 11, 922 Materials 2018, 11, x FOR Table 1. Characteristics of nickel-coated carbon yarn (NiCCY) [14]. Parameter Description Table 1. Characteristics of nickel-coated carbon yarn (NiCCY) [14]. Parameter Description Table 1. Characteristics of nickel-coated carbon yarn (NiCCY) [14]. Parameter Description Raw material Carbon Liner density [tex] 1420 tex Coating 0.25 µm of Ni No. of ﬁlaments 1200 Filament diameter [µm] 7.5 incl. 2.3. Seebeck Coefﬁcient Measurement 2.3. Seebeck Coefficient Measurement The electric voltage of the thermocouple samples was measured using a nanovoltmeter Ampliﬁcateur NV 724 from Setaram, Lyon, France. A Fluke 52 digital thermometer (Fluke, Everett, Washington, DC, USA) was used to measure the temperature during voltage measurement. The higher temperature range for the hot junction was controlled by an electric hot plate, while the lower temperature range for the cold junction was dependent on the external conditions of the laboratory, which were controlled and stable. Figure 3 shows the voltage measurement set up of etched and non-etched NiCCY. The electric voltage of the thermocouple samples was measured using a nanovoltmeter Amplificateur NV 724 from Setaram, Lyon, France. A Fluke 52 digital thermometer (Fluke, Everett, Washington, DC, USA) was used to measure the temperature during voltage measurement. The higher temperature range for the hot junction was controlled by an electric hot plate, while the lower temperature range for the cold junction was dependent on the external conditions of the laboratory, which were controlled and stable. Figure 3 shows the voltage measurement set up of etched and non- etched NiCCY. Figure 3. Illustration of the voltage measurement set up. The junction was placed on the hot plate that had been covered with a piece of paper and a wood weight was placed on the junction and thermometer probe. Figure 3. Illustration of the voltage measurement set up. The junction was placed on the hot plate that had been covered with a piece of paper and a wood weight was placed on the junction and thermometer probe. Figure 3. Illustration of the voltage measurement set up. The junction was placed on the hot plate that had been covered with a piece of paper and a wood weight was placed on the junction and thermometer probe. Figure 3. Illustration of the voltage measurement set up. The junction was placed on the hot plate that had been covered with a piece of paper and a wood weight was placed on the junction and thermometer probe. The efficiency of the thermocouple was measured through different ranges of temperatures for a cold junction (291.15–293.15 K) and for a hot junction (293.15 K–325.15 K). These temperature ranges were the only ones enabling the performance of stable and repeatable measurements. Germa 2.1. Material 2. Mate 2 1 M Coating Density [g/cm3] 2.70 Twist [tpm, type] 0 Linear electrical resistance [Ω/m] 2.2667 Commercial name Tenax®-J HTS40 Parameter Description Raw material Carbon Liner density [tex] 1420 tex Coating 0.25 µm of Ni No. of filaments 1200 Filament diameter [µm] 7.5 incl. Coating Density [g/cm3] 2.70 Twist [tpm, type] 0 Linear electrical resistance [Ω/m] 2.2667 Commercial name Tenax®-J HTS40 2.2. Etching Process 2.2. Etching Process The process of The process of removing Ni in this experiment is called an etching process. Combination of HCl and H2O2 solution was utilized to remove Ni from NiCCY and is then called etching solution. The H2O2 concentration varied from 3%, to 6% up, and to 10%, while HCl was kept constant at 37%. The ratio between H2O2 and HCl was 1:1. The samples were immersed in the etching solutions for 30 min. Next, the sample was washed with water to remove the remaining chemicals from the yarn. The excess of water remaining on the sample was absorbed by blotting paper several times and all the samples were air dried at room temperature for a minimum of 24 h prior to testing [13]. The process of removing Ni in this experiment is called an etching process. Combination of HCl and H2O2 solution was utilized to remove Ni from NiCCY and is then called etching solution. The H2O2 concentration varied from 3%, to 6% up, and to 10%, while HCl was kept constant at 37%. The ratio between H2O2 and HCl was 1:1. The samples were immersed in the etching solutions for 30 min. Next, the sample was washed with water to remove the remaining chemicals from the yarn. The excess of water remaining on the sample was absorbed by blotting paper several times and all the samples were air dried at room temperature for a minimum of 24 h prior to testing [13]. 2.3. Seebeck Coefﬁcient Measurement 2.3. Seebeck Coefficient Measurement 2 4 Th i l A l i The efﬁciency of the thermocouple was measured through different ranges of temperatures for a cold junction (291.15–293.15 K) and for a hot junction (293.15–325.15 K). These temperature ranges were the only ones enabling the performance of stable and repeatable measurements. 2.4. Theoretical Analysis The following theo 2.4. Theoretical Analysis The following theoretical calculations are formulated according to the schematic diagram of NiCCY that has been shown in Figure 1 earlier. Generally, the temperature varies along the thermocouple wire: T(x). We know that T(−a) = T(b) = TC and T(0) = TH. Similarly, the electric potential The following theoretical calculations are formulated according to the schematic diagram of NiCCY that has been shown in Figure 1 earlier. Generally, the temperature varies along the 5 of 10 Materials 2018, 11, 922 thermocouple wire: T(x). We know that T(−a) = T(b) = TC and T(0) = TH. Similarly, the electric potential φ (x) depends on x. The generated voltage can be calculated as: thermocouple wire: T(x). We know that T(−a) = T(b) = TC and T(0) = TH. Similarly, the electric potential φ (x) depends on x. The generated voltage can be calculated as: V0 = φ (b) −φ (−a) (1) (1) Inside each conductor, the current density J (expressed in A/m2) is given from literature [16]: J = −σ dφ dx + εdT dx (2) (2) where σ is the electric conductivity and ε the Seebeck coefﬁcient. From the literature, the numerical values are εNi = −14.8 µV/K and εC = +3.0 µV/K [17]. Hence, the value 14.8 + 3 = 17.8 µV/K is the highest value one can obtain with a carbon-nickel thermocouple. The electric current I1 ﬂowing through the left part is then: where σ is the electric conductivity and ε the Seebeck coefﬁcient. From the literature, the numerical values are εNi = −14.8 µV/K and εC = +3.0 µV/K [17]. Hence, the value 14.8 + 3 = 17.8 µV/K is the highest value one can obtain with a carbon-nickel thermocouple. The electric current I1 ﬂowing through the left part is then: I1 = −σC dφ dx + εC dT dx S −σNi dφ dx + εNi dT dx S1 (3) (3) Similarly, for the right part, we have to replace S1 by S2 to get I2: Similarly, for the right part, we have to replace S1 by S2 to get I2: I2 = −σC dφ dx + εC dT dx S −σNi dφ dx + εNi dT dx S2 (4) (4) The thermocouple voltage is measured with a meter having a high input impedance. 3.1. Microscopic Observation after Etching Process 3.1. Microscopic Observation after Etching Process 3.1. Microscopic Observation after Etching Process 3.1. Microscopic Observation after Etching Process After the chemical treatment of the samples of NiCCY, one observed the effect of the etching process excreted on the treated samples through scanning electron microscope (SEM) image. An untreated sample of NiCCY in Figure 4a was compared to the treated samples in Figure 4b–d where the treatment was 37% HCl and 3%, 6% and 10% of H2O2, respectively. After the chemical treatment of the samples of NiCCY, one observed the effect of the etching process excreted on the treated samples through scanning electron microscope (SEM) image. An untreated sample of NiCCY in Figure 4a was compared to the treated samples in Figure 4b–d where the treatment was 37% HCl and 3%, 6% and 10% of H2O2, respectively. Figure 4. Images from scanning electron microscope (Jeol JSM-7600F) taken with 5000x magnification: (a) Untreated nickel-coated carbon fiber; (b) nickel-coated carbon fiber after treatment of 3% H2O2 + 37% HCl; (c) nickel-coated carbon fiber after treatment of 6 % H2O2 + 37% HCl; (d) nickel-coated carbon fiber after treatment of 10% H2O2 + 37% HCl. Figure 4. Images from scanning electron microscope (Jeol JSM-7600F) taken with 5000x magniﬁcation: (a) Untreated nickel-coated carbon ﬁber; (b) nickel-coated carbon ﬁber after treatment of 3% H2O2 + 37% HCl; (c) nickel-coated carbon ﬁber after treatment of 6 % H2O2 + 37% HCl; (d) nickel-coated carbon ﬁber after treatment of 10% H2O2 + 37% HCl. Figure 4. Images from scanning electron microscope (Jeol JSM-7600F) taken with 5000x magnification: (a) Untreated nickel-coated carbon fiber; (b) nickel-coated carbon fiber after treatment of 3% H2O2 + 37% HCl; (c) nickel-coated carbon fiber after treatment of 6 % H2O2 + 37% HCl; (d) nickel-coated carbon fiber after treatment of 10% H2O2 + 37% HCl. Figure 4. Images from scanning electron microscope (Jeol JSM-7600F) taken with 5000x magniﬁcation: (a) Untreated nickel-coated carbon ﬁber; (b) nickel-coated carbon ﬁber after treatment of 3% H2O2 + 37% HCl; (c) nickel-coated carbon ﬁber after treatment of 6 % H2O2 + 37% HCl; (d) nickel-coated carbon ﬁber after treatment of 10% H2O2 + 37% HCl. Based on the application of different chemical concentration treatments of H2O2 to the NiCCY and the microstructure (topography) of untreated and treated NiCCY observed via SEM as shown in Figure 4a–d, one may draw a conclusion concerning the potential impact of the treatment. 2.4. Theoretical Analysis The following theo 2.4. Theoretical Analysis Materials 2018, 11, x FOR PEER REVIEW 6 of 10 which is the formula we are most familiar with. Note that the term (12) is also the highest value one can obtain with a C-Ni thermocouple. Materials 2018, 11, x FOR PEER REVIEW 6 of 10 2.4. Theoretical Analysis The following theo 2.4. Theoretical Analysis Hence, almost no current can ﬂow or: The thermocouple voltage is measured with a meter having a high input impedance. Hence, almost no current can ﬂow or: I1 = I2 = 0 (5) (5) Integrating (3) with respect to x gives then: Integrating (3) with respect to x gives then: σC(φ + εCT)S + σNi(φ + εNiT)S1 = A (6) σC(φ + εCT)S + σNi(φ + εNiT)S1 = A (6) (6) where A is an integration constant. Applying (6) in the point x = −a and x = 0 gives, after subtraction: σC Vj + εCTH S + σNi Vj + εNiTH S1 −σCεCTCS −σNiεNiTCS1 = 0 (7) σC Vj + εCTH S + σNi Vj + εNiTH S1 −σCεCTCS −σNiεNiTCS1 = 0 (7) (7) or: or: (σCS + σNiS1)Vj + (σCεCS + σNiεNiS1)(TH −TC) = 0 (8) lation for the right part (0 < x < b) gives us: (σCS + σNiS1)Vj + (σCεCS + σNiεNiS1)(TH −TC) = 0 (8) (8) A similar calculation for the right part (0 < x < b) gives us: (σCS + σNiS2) Vj −V0 + (σCεCS + σNiεNiS2)(TH −TC) = 0 (9) (9) Elimination of the junction potential Vj from (8) and (9) gives us ﬁnally: V0 = (TH −TC) σCεCS + σNiεNiS2 σCS + σNiS2 −σCεCS + σNiεNiS1 σCS + σNiS1 (10) (10) Rewriting (10) gives: V0 = (TH −TC) σCσNiS(S2 −S1) (σCS + σNiS2)(σCS + σNiS1)(εNi −εC) (11) (11) As expected, we obtain an output voltage V0 proportional to the temperature difference TH −TC. Obviously, the result of Equation (11) is also proportional to the difference of the two Seebeck coefﬁcients, εNi −εC. As expected, we obtain an output voltage V0 proportional to the temperature difference TH −TC. Obviously, the result of Equation (11) is also proportional to the difference of the two Seebeck coefﬁcients, εNi −εC. Materials 2018, 11, 922 6 of 10 If S1 = 0, or Ni has been completely removed from the left part, and σNiS2 ≫σCS, the relation (11) can be simpliﬁed to: If S1 = 0, or Ni has been completely removed from the left part, and σNiS2 ≫σCS, the relation (11) can be simpliﬁed to: (12) V0 = (TH −TC)(εNi −εC) (12) which is the formula we are most familiar with. Note that the term (12) is also the highest value one can obtain with a C-Ni thermocouple. 3.1. Microscopic Observation after Etching Process 3.1. Microscopic Observation after Etching Process Namely, the higher the concentration of H2O2, the more intense the etching process was, with less Ni remaining on the surface of C. An obvious difference may be noticed between untreated NiCCY (Figure 4a) and the treated one (Figure 4b), where the H2O2 concentration was only 3%. In the case of this treatment, the surface of Ni was visibly incised, generating a porous-like layer on the surface of CF. Additionally, some zones of notches on the surface of NiCCY are also visible. This image confirms that in the initial phase, the etching is not homogeneous along the entire yarns, which will also influence the electrical resistance and thermocouple characteristics. Based on the application of different chemical concentration treatments of H2O2 to the NiCCY and the microstructure (topography) of untreated and treated NiCCY observed via SEM as shown in Figure 4a–d, one may draw a conclusion concerning the potential impact of the treatment. Namely, the higher the concentration of H2O2, the more intense the etching process was, with less Ni remaining on the surface of C. An obvious difference may be noticed between untreated NiCCY (Figure 4a) and the treated one (Figure 4b), where the H2O2 concentration was only 3%. In the case of this treatment, the surface of Ni was visibly incised, generating a porous-like layer on the surface of CF. Additionally, some zones of notches on the surface of NiCCY are also visible. This image conﬁrms that in the initial phase, the etching is not homogeneous along the entire yarns, which will also inﬂuence the electrical resistance and thermocouple characteristics. In Figure 4c, one may observe a further etch on the Ni layer due to an increased concentration of H2O2 from 3% up to 6%. Although the concentration of H2O2 is higher in the case of the treatment presented in Figure 4c, there are still some larger islands of Ni clearly visible. Nevertheless, a large In Figure 4c, one may observe a further etch on the Ni layer due to an increased concentration of H2O2 from 3% up to 6%. Although the concentration of H2O2 is higher in the case of the treatment presented in Figure 4c, there are still some larger islands of Ni clearly visible. Nevertheless, a large zone 7 of 10 Materials 2018, 11, 922 of clean and grooved CF is also visible. 3.1. Microscopic Observation after Etching Process 3.1. Microscopic Observation after Etching Process In the case of the highest utilized concentration of H2O2 in this experiment, only some small dust, sparsely distributed, are present on the surface of CF (Figure 4d). Therefore, this conﬁrms that the sample treated in 10% H2O2 + 37% HCl can be considered as a fully etched yarn. 3.2. Comparison between Experiment and Theory In order to check the theoretical analysis, it is more convenient to use resistance values of the yarns to be inserted in (11). The reason is obvious; resistance can be easily measured whereas a cross section like S, S1 or S2 are hard to obtain. The following resistances (expressed in Ω/m) are deﬁned: The following resistances (expressed in Ω/m) are deﬁned: RC = 1 σCS (13) R1 = 1 σNiS1 (14) R2 = 1 σNiS2 (15) RC = 1 σCS (13) R1 = 1 σNiS1 (14) R2 = 1 σNiS2 (15) Equation (11) is then rewritten as: (13) (14) (15) Ni 2 Equation (11) is then rewritten as: Equation (11) is then rewritten as: V0 = (TH −TC) RC(R1 −R2) (RC + R1)(RC + R2)(εNi −εC) (16) (16) In order to verify the theoretical formula (16), one has to measure the resistance of the yarns involved in our experiment. The results are presented in Table 2 [13]. Table 2. Linear electrical resistance of etched and non-etched yarns [13]. Table 2. Linear electrical resistance of etched and non-etched yarns [13]. Etching Condition Linear Electrical Resistance [Ω/m] Non-etched 2.2667 3% H2O2 + 37% HCl (1:1) 2.8667 6% H2O2 + 37% HCl (1:1) 31.533 10% H2O2 + 37% HCl (1:1) 45.933 From the SEM image, it can be considered that Ni was completely removed in the etching solution containing 10% H2O2 + 37% HCl (1:1). The electric resistance is then just the resistance of the carbon or: RC = 45.933 Ω/m (17) (17) The non-etched part has a resistance of 2.267 Ω/m, which is the parallel connection of RC and R2, or: 1 1 1 1 RC + 1 R2 = 1 2.2667 or R2 = 2.3844 Ω/m (18) (18) Similarly, one can calculate the value of R1 for 6% H2O2 and 37% HCl (1:1) etching: Similarly, one can calculate the value of R1 for 6% H2O2 and 37% HCl (1:1) etching: 1 RC + 1 R1 = 1 31.533 or R1(6%) = 100.584 Ω/m (19) (19) and for 3% H2O2 and 37% HCl (1:1) etching: 1 RC + 1 R1 = 1 2.867 or R1(3%) = 3.0575 Ω/m (20) (20) 8 of 10 Materials 2018, 11, 922 Inserting all the known values into the Equations (12) and (16), one can plot the theoretical graphs as shown in Figure 5 (with lines). 3.2. Comparison between Experiment and Theory The graph for the sample treated with 10% H2O2 + 37% HCl was calculated with Equation (12) because S1 was considered equal to 0. The graphs for the 3% H2O2 + 37% HCl and 6% H2O2 + 37% HCl were calculated with Equation (16). In Figure 5, the graph of the experimental data (with markers) was also presented as a comparison to the theoretical one. Materials 2018, 11, x FOR PEER REVIEW 8 of 10 Values inside the boxes are linear trend line equations for the corresponding data of the 3–10% H2O2 with which each was mixed with 37% HCl in 1:1 ratio. The orange and blue boxes are attributed to the experimental and theoretical trend lines, respectively. The Seebeck coefﬁcient values were taken from the slope of voltage vs. temperature difference. Values inside the boxes are linear trend line equations for the corresponding data of the 3–10% H2O2 with which each was mixed with 37% HCl in 1:1 ratio. The orange and blue boxes are attributed to the experimental and theoretical trend lines, respectively. The Seebeck coefficient values were taken from the slope of voltage vs temperature difference It is clear from Figure 5 that the agreement between the theoretical and experimental data is quite good for the yarns etched with 6% and 10% H2O2. The case with 3% H2O2 shows a poor agreement. It must be pointed out, however, that 3% H2O2 was also a very poor etching. This is proved by the resistance values presented in Table 2. The linear electrical resistance only changed from 2.2667 to 2.8667 Ω/m or 26%. taken from the slope of voltage vs. temperature difference. It is clear from Figure 5 that the agreement between the theoretical and experimental data is quite good for the yarns etched with 6% and 10% H2O2. The case with 3% H2O2 shows a poor agreement. It must be pointed out, however, that 3% H2O2 was also a very poor etching. This is proved by the resistance values presented in Table 2. The linear electrical resistance only changed from 2.2667 to 2.8667 Ω/m or 26%. Figure 5. Plot of voltage vs. temperature difference of the samples (theoretical and experimental). Th E ti (16) b d Figure 5. Plot of voltage vs. temperature difference of the samples (theoretical and experimental). Figure 5. Plot of voltage vs. temperature difference of the samples (theoretical and experimental). In thi 4. Conclusions used as the conductive materials for creating textile-based thermocouple. After performing the stripping process in three different concentration of stripping solutions, it is obvious that the higher the concentration of H2O2, the more intense the etching process was, with less Ni remaining on the surface of CF. From the theoretical calculation and experimental data, it is proved that there is a good agreement between the theoretical and experimental data, especially for the yarns etched with 6% and 10% H2O2 (both are mixed with 37% HCl). The yarn etched with 3% H2O2 shows a poor agreement due to the very poor etching action of the chemicals at 3% H2O2 and 37% HCl. The 10% H2O2 + 37% HCl was efficient enough to etch the Ni from NiCCY. We can conclude that the more efficient the etching process, the better the Seebeck coefficient created from the etched and non-etched NiCCY is. The R-squares of the experimental graphs are all more than 0.99, showing that the data are very close to the fitted regression line. The value of 𝑅1 has a great influence on the whole characteristics of this thermocouple. Overall, we are successful in developing a theoretical formula to calculate the Seebeck coefficient of thermocouples made of etched and non-etched NiCCY based on the resistance value of the samples due to the simplicity of electrical resistance measurement. Author Contributions: H.H., G.D.M., C.H., and L.V.L. conceived and designed the experiments; H.H. performed the experiments; H.H., G.D.M. and I.C.-W. analyzed the data; H.H., G.D.M. and I.C.-W. wrote the paper; H.H., G.D.M., I.C.-W. and L.V.L. revised the paper. In this work, a combination of etched and non-etched nickel-coated carbon yarn (NiCCY) was used as the conductive materials for creating textile-based thermocouple. After performing the stripping process in three different concentration of stripping solutions, it is obvious that the higher the concentration of H2O2, the more intense the etching process was, with less Ni remaining on the surface of CF. From the theoretical calculation and experimental data, it is proved that there is a good agreement between the theoretical and experimental data, especially for the yarns etched with 6% and 10% H2O2 (both are mixed with 37% HCl). The yarn etched with 3% H2O2 shows a poor agreement due to the very poor etching action of the chemicals at 3% H2O2 and 37% HCl. 3.2. Comparison between Experiment and Theory h E b d Figure 5. Plot of voltage vs. temperature difference of the samples (theoretical and experimental). The Equation (16) can be rearranged as: 𝑉= (𝑇−𝑇)(𝜀 − The Equation (16) can be rearranged as: The Equation (16) can be rearranged as: 𝑉= (𝑇−𝑇)(𝜀௜− The Equation (16) can be rearranged as: The Equation (16) can be rearranged as: 𝑉 (𝑇 𝑇)( ) 𝑅஼ 𝑅ଶቀ𝑅ଵ 𝑅ଶ−1ቁ (21) The Equation (16) can be rearranged as: ଴ ( ு ஼)( ே௜ ஼) ቀ𝑅஼ 𝑅ଶ+ 𝑅ଵ 𝑅ଶቁቀ𝑅஼ 𝑅ଶ+ 1ቁ ( ) values from Equations (17)–(20) into the theoretical formula (21), we can make a d voltage 𝑉଴ versus 𝑅ଵ/𝑅ଶ as shown in Figure 6. One observes that the value of atic influence on the overall performance of this thermocouple. V0 = (TH −TC)(εNi −εC) RC R2 R1 R2 −1 RC R2 + R1 R2 RC R2 + 1 (21) (21) If we insert the values from Equations (17)–(20) into the theoretical formula (21), we can make a plot of the generated voltage V0 versus R1/R2 as shown in Figure 6. One observes that the value of R1 can have a dramatic inﬂuence on the overall performance of this thermocouple. If we insert the values from Equations (17)–(20) into the theoretical formula (21), we can make a plot of the generated voltage V0 versus R1/R2 as shown in Figure 6. One observes that the value of R1 can have a dramatic inﬂuence on the overall performance of this thermocouple. 9 of 10 Materials 2018, 11, 922 Figure 6. The generated voltage 𝑉଴ versus 𝑅ଵ/𝑅ଶ plotted from Equation (21) with a temperature difference of 1 K. 4 Conclusions Figure 6. The generated voltage V0 versus R1/R2 plotted from Equation (21) with a temperature difference of 1 K. Figure 6. The generated voltage 𝑉଴ versus 𝑅ଵ/𝑅ଶ plotted from Equation (21) with a temperature difference of 1 K. Figure 6. The generated voltage V0 versus R1/R2 plotted from Equation (21) with a temperature difference of 1 K. In thi 4. Conclusions The 10% H2O2 + 37% HCl was efﬁcient enough to etch the Ni from NiCCY. We can conclude that the more efﬁcient the etching process, the better the Seebeck coefﬁcient created from the etched and non-etched NiCCY is. The R-squares of the experimental graphs are all more than 0.99, showing that the data are very close to the ﬁtted regression line. The value of R1 has a great inﬂuence on the whole characteristics of this thermocouple. Overall, we are successful in developing a theoretical formula to calculate the Seebeck coefﬁcient of thermocouples made of etched and non-etched NiCCY based on the resistance value of the samples due to the simplicity of electrical resistance measurement. Funding: This research was funded by the Government of Indonesia through the Indonesian Endowment Fund for Education (LPDP). Acknowledgments: Hardianto on leave from Politeknik STTT Bandung wants to thank LPDP for the financial Author Contributions: H.H., G.D.M., C.H. and L.V.L. conceived and designed the experiments; H.H. performed the experiments; H.H., G.D.M. and I.C.-W. analyzed the data; H.H., G.D.M. and I.C.-W. wrote the paper; H.H., G.D.M., I.C.-W. and L.V.L. revised the paper. support for his stay at Ghent University as a PhD student. Conflicts of Interest: The authors declare no conflict of interest. Funding: This research was funded by the Government of Indonesia through the Indonesian Endowment Fund for Education (LPDP). References Acknowledgments: Hardianto on leave from Politeknik STTT Bandung wants to thank LPDP for the ﬁnancial support for his stay at Ghent University as a PhD student. References Acknowledgments: Hardianto on leave from Politeknik STTT Bandung wants to thank LPDP for the ﬁnancial support for his stay at Ghent University as a PhD student. 1. Choi, C.; Kim, K.M.; Kim, K.J.; Lepró, X.; Spinks, G.M.; Baugh capacitance via weavable superelastic biscrolled yarn sup Conﬂicts of Interest: The authors declare no conﬂict of interest. 3. Uchida, K.; Takahashi, S.; Harii, K.; Ieda, J.; Koshibae, W.; Ando, K.; Maekawa, S.; Saitoh, E. Observation of the spin Seebeck effect. Nature 2008, 455, 778–781. [CrossRef] [PubMed] of the spin Seebeck effect. Nature 2008, 455, 778 781, doi:10.1038/nature07321. 2. Huang, X.; Leng, T.; Zhu, M.; Zhang, X.; Chen, J.; Chang, K.; Aqeeli, M.; Geim, A.K.; Novoselov, K.S.; Hu, Z. Highly Flexible and Conductive printed graphene for wireless wearable communications applications. Sci. Rep. 2015, 5, 18298. [CrossRef] [PubMed] d 2. H References Z. Highly Flexible and Conductive printed graphene for wireless wearable communications applications. Sci. Rep. 2015, 5, 18298, doi:10.1038/srep18298. 3. Uchida, K.; Takahashi, S.; Harii, K.; Ieda, J.; Koshibae, W.; Ando, K.; Maekawa, S.; Saitoh, E. Observation of the spin Seebeck effect. Nature 2008, 455, 778–781, doi:10.1038/nature07321. 1. Choi, C.; Kim, K.M.; Kim, K.J.; Lepró, X.; Spinks, G.M.; Baughman, R.H.; Kim, S.J. Improvement of system capacitance via weavable superelastic biscrolled yarn supercapacitors. Nat. Commun. 2016, 7, 13811. [CrossRef] [PubMed] of the spin Seebeck effect. Nature 2008, 455, 778 781, doi:10.1038/nature07321. 2. Huang, X.; Leng, T.; Zhu, M.; Zhang, X.; Chen, J.; Chang, K.; Aqeeli, M.; Geim, A.K.; Novoselov, K.S.; Hu, Z. Highly Flexible and Conductive printed graphene for wireless wearable communications applications. Sci. Rep. 2015, 5, 18298. [CrossRef] [PubMed] o t e spi Seebeck e ect. Nature 008, 55, 8 8 , doi: 0. 038/ atu e0 3 . 2. Huang, X.; Leng, T.; Zhu, M.; Zhang, X.; Chen, J.; Chang, K.; Aqeeli, M.; Geim, A.K.; Novoselov, K.S.; Hu, Z. Highly Flexible and Conductive printed graphene for wireless wearable communications applications. Sci. Rep. 2015, 5, 18298. [CrossRef] [PubMed] p [ ] [ ] 3. Uchida, K.; Takahashi, S.; Harii, K.; Ieda, J.; Koshibae, W.; Ando, K.; Maekawa, S.; Saitoh, E. Observation of the spin Seebeck effect. Nature 2008, 455, 778–781. [CrossRef] [PubMed] Materials 2018, 11, 922 10 of 10 10 of 10 4. Ziegler, S.; Frydrysiak, M. Initial research into the structure and working conditions of textile thermocouples. Fibres Text. East. Eur. 2009, 77, 84–88. 5. Gidik, H.; Bedek, G.; Dupont, D. Developing thermophysical sensors with textile auxiliary wall. In Smart Textiles and Their Applications; Woodhead Publishing: Cambridge, MA, USA, 2016; pp. 423–453. 6. Alexander, R. Jones The Application of Temperature Sensors Into Fabric Substrates. Master’s Thesis, Kansas State University, Manhattan, KS, USA, 2011. 7. Dupont, D.; Godts, P.; Leclercq, D. Design of textile heat ﬂowmeter combining evaporation phenomena. Text. Res. J. 2006, 76, 772–776. [CrossRef] 8. Gidik, H.; Bedek, G.; Dupont, D.; Codau, C. Impact of the textile substrate on the heat transfer of a textile heat ﬂux sensor. Sens. Actuators A Phys. 2015, 230, 25–32. [CrossRef] 9. Hertleer, C.; Bedek, G.; Dupont, D. A Textile-Based Heat Flux Sensor. In Proceedings of the VI International Technical Textiles Congress, Izmir, Turkey, 14–16 October 2015. 10. Meng, C.; Liu, C.; Fan, S. d 2. H References A promising approach to enhanced thermoelectric properties using carbon nanotube networks. Adv. Mater. 2010, 22, 535–539. [CrossRef] [PubMed] 11. Esfarjani, K.; Zebarjadi, M.; Kawazoe, Y. Thermoelectric properties of a nanocontact made of two-capped single-wall carbon nanotubes calculated within the tight-binding approximation. Phys. Rev. B Condens. Matter Mater. Phys. 2006, 73. [CrossRef] 12. Mahmoud, L.; Abdul Samad, Y.; Alhawari, M.; Mohammad, B.; Liao, K.; Ismail, M. Combination of PVA with graphene to improve the seebeck coefﬁcient for thermoelectric generator applications. J. Electron. Mater. 2015, 44, 420–424. [CrossRef] 13. Hardianto, A.; Hertleer, C.; Mey, G.; De Langenhove, L. Van Removing nickel from nickel-coated carbon ﬁbers. IOP Conf. Ser. Mater. Sci. Eng. 2017, 254, 72010. [CrossRef] 14. Wróbel, I.C.; Knockaert, J.; Mey, G. De Shielding the electromagnetic waves by inserting conductive lightweight materials into woven curtains. J. Fash. Technol. Text. Eng. 2017, 1–4. [CrossRef] 15. TIBTECH Innovations Properties Table of Stainless Steel, Metals and Other Conductive Materials. Available online: http://www.tibtech.com/conductivity.php (accessed on 1 February 2018). 16. Bejan, A. Advanced Engineering Thermodynamics, 3rd ed.; Wiley: Hoboken, NJ, USA, 2006; ISBN 0471677639. 17. McGhee, J.; Korczynski, J.; Henderson, I.; Kulesza, W. Scientiﬁc Metrology; ACGM Lodart S.A.: Lódz, Poland, 1998; ISBN 83-904299-9-3. © 2018 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/). © 2018 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/W3207701353	https://europepmc.org/articles/pmc8547277?pdf=render	English	null	Prognostic value of D-dimer/fibrinogen ratio on in-hospital outcomes of patients with heart failure and COVID-19	Biomarkers in medicine	2,021	cc-by	7,015	Prognostic value of D-dimer/fibrinogen ratio on in-hospital outcomes of patients with heart failure and COVID-19 Selda Murat,1 , Bektas Murat2 , Muhammet Dural1 , Gurbet Ozge Mert1 & Yukse 1 Selda Murat,1 , Bektas Murat2 , Muhammet Dural1 , Gurbet Ozge Me Cavusoglu1 1Medical Faculty Department of Cardiology, Eskisehir Osmangazi University, Eskisehir, 26040, Turkey 2Department of Cardiology, Eskisehir City Hospital, Eskisehir, 26080, Turkey Author for correspondence: Tel.: +90 222 239 2979; selda.eraslan@hotmail.com Selda Murat,1 , Bektas Murat2 , Muhammet Dural1 , Gurbet Ozge Me Cavusoglu1 1Medical Faculty Department of Cardiology, Eskisehir Osmangazi University, Eskisehir, 26040, Turkey 2Department of Cardiology, Eskisehir City Hospital, Eskisehir, 26080, Turkey Author for correspondence: Tel.: +90 222 239 2979; selda.eraslan@hotmail.com g 1Medical Faculty Department of Cardiology, Eskisehir Osmangazi University, Eskisehir, 26040, Turkey 2Department of Cardiology, Eskisehir City Hospital, Eskisehir, 26080, Turkey Author for correspondence: Tel.: +90 222 239 2979; selda.eraslan@hotmail.com Aim: In the present study, the relationship between D-dimer/fibrinogen ratio (DFR) and in-hospital outcomes was evaluated in patients with COVID-19 and a diagnosis of heart failure (HF). Materials & methods: In-hospital outcomes were compared in patients with high and low DFR values. Results: With regard to in-hospital outcomes, patients in the third tertile of DFR had a higher rate of mechanical ventilation, cardiogenic shock and death (p < 0.001). The length of ICU stay was longer in the third tertile group (p < 0.001). When evaluated together with infection markers, DFR was found to be an independent predictor of outcomes. Conclusion: DFR can be used as a prognostic marker in patients with COVID-19 with a diagnosis of HF, and perhaps more valuable than other infection markers. First draft submitted: 28 April 2021; Accepted for publication: 17 August 2021; Published online: 20 October 2021 Keywords: coronavirus disease • D-dimer/fibrinogen ratio • heart failure COVID-19, caused by SARS-CoV-2, is a rapidly spreading pandemic associated with high morbidity and mor- tality worldwide, with the number of new cases and deaths continuing to rise [1,2]. Although COVID-19 is transmitted primarily through respiratory droplets/aerosols, initially causing pneumonia in the lung, it may affect multiple organs, including the cardiovascular system [3–6]. Previous studies have shown that comorbidities such as hypertension (HT), diabetis mellitus (DM) and cardiovascular disease including heart failure (HF) are associated with poor prognosis [1,7,8]. Some recent studies have shown the role of HF both as a risk factor for a poor clinical course and for increased mortality and as a possible consequence of COVID-19-related myocardial damage [6,9,10]. Research Article For reprint orders, please contact: reprints@futuremedicine.com Prognostic value of D-dimer/fibrinogen ratio on in-hospital outcomes of patients with heart failure and COVID-19 COVID-19 has been described as a thrombo-inﬂammatory syndrome [11,12]. Among severely ill patients with mortality, diffuse endothelial dysfunction, widespread coagulopathy and complement-induced thrombosis have re- sulted in the development of thromboembolism [13]. Several studies show that ﬁbrinogen, D-dimer and ﬁbrinogen degradation products are associated with mortality through coagulation impairment and inﬂammation in patients with COVID-19. Although some previous studies reported that D-dimer elevation is associated with the severity of COVID-19 [1,14,15], a recent study conducted by Emmanuel et al. reported confusion and potential for misinfor- mation regarding D-dimer [16]. Fibrinogen is also well studied in patients with COVID-19. Although it has been shown in some previous studies that ﬁbrinogen is associated with the severity of COVID-19 disease, other stud- ies have reported that ﬁbrinogen alone is not signiﬁcant and should be evaluated together with D-dimer [17–19]. Some studies have evaluated the value of plasma D-dimer/ﬁbrinogen ratio (DFR) in the diagnosis of pulmonary thromboembolism and lower extremity venous thrombosis [20–22]. DFR also shows a unique value in determining the pathophysiological mechanism of stroke [23] and can predict the prognosis of gastrointestinal stromal tumors in patients hospitalized with HF [24]; however, the prognostic values of DFR in patients with COVID-19 with a history of HF is unclear. Therefore, the present study was designed to evaluate the relationship between DFR and in-hospital prognosis for patients with COVID-19 with a history of HF. 10.2217/bmm-2021-0341 C⃝2021 Future Medicine Ltd Biomark. Med. (Epub ahead of print) ISSN 1752-0363 Research Article Murat, Murat, Dural, Mert & Cavusoglu Data collection All data were checked by two researchers to ascertain accuracy. Material & methods Study subjects & design The study enrolled 240 consecutive patients admitted to Eskisehir Osmangazi University, Medical Faculty Depart- ment of Cardiology and Eskisehir City Hospital Department of Cardiology with laboratory-conﬁrmed SARS-CoV-2 infection from 15 March 2020 to 1 December 2020. The diagnosis of COVID-19 was conﬁrmed by RNA reverse- transcriptase polymerase chain reaction (RT-PCR) detection of the SARS-CoV-2 in the clinical laboratory of both centers. The exclusion criteria were age under 18 years old, outpatients who were referred to another hospital during their hospitalization, acute myocardial infarction, patients with a history of cerebral infarct (n = 5), malignancy (n = 2) and patients with acute deep venous thrombosis and/or pulmonary thromboembolism (n = 1). Heart failure was identiﬁed from the International Classiﬁcation of Diseases, tenth revision, clinical modiﬁcation (ICD-10-CM) codes in patient electronic medical records (EMR), including I50 (heart failure), I50.1 (left ventricular failure, unspeciﬁed), I50.2 (systolic [congestive] heart failure) and I50.9 (heart failure, unspeciﬁed) diagnostic codes. A total of 232 patients with HF were included in the study. Data collection Demographic characteristics (age and sex); cardiovascular history and chronic diseases including heart failure, arterial hypertension, diabetes mellitus, ischemic heart disease, atrial ﬁbrillation, chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), devices (pacemaker and implantable cardioverter-deﬁbrillator); clinical data (vital signs, laboratory ﬁndings) and therapy were collected from EMRs. Heart rate, oxygen saturation and blood pressure were recorded at the time of admission. The length of stay in the ICU and ward was obtained from discharge records. Data on counts or levels of hemoglobin (Hb), white blood cells (WBCs), lymphocytes (L), neutrophils (N), sodium, potassium, glucose, alanine transaminase (ALT), aspartate aminotransferase (AST), serum albumin (ALB), serum creatinine (sCr), blood urea nitrogen (BUN) and the highest values of inﬂammatory markers such as C-reactive protein (CRP), ferritin, lactate and procalcitonin were included. From the D-dimer and ﬁbrinogen values, two values were taken into consideration: The ﬁrst within 24 hours of admission to hospital and the highest values. D-dimer levels (normal range 0–0.50 mg/l) were measured by immunoturbidimetry and ﬁbrinogen levels (normal range 170–420 mg/dl) by clotting method using the Sysmex CS5100 automated coagulation analyzer. DFR was calculated using the following formula: DFR = D −dimer μg/ml Fibrinogen mg/dl × 100 DFR = D −dimer μg/ml Fibrinogen mg/dl × 100 In-hospital mortality, respiratory failure requiring noninvasive mechanical ventilation and orotracheal intubation, duration of ICU stay, acute kidney injury treated with renal replacement therapy, need for blood transfusion and use of intravenous vasopressor drug were evaluated. The clinical outcomes were deﬁned as all-cause death, respiratory failure requiring mechanical ventilation and cardiogenic shock during hospitalization. Cardiogenic shock was deﬁned as hypotension (SBP <90 mmHg) with evidence of hypoperfusion and end-organ dysfunction [25]. The left ventricular ejection fraction (LVEF) before or during index COVID-19 admission was recorded, if available. All data were checked by two researchers to ascertain accuracy. In-hospital mortality, respiratory failure requiring noninvasive mechanical ventilation and orotracheal intubation, duration of ICU stay, acute kidney injury treated with renal replacement therapy, need for blood transfusion and use of intravenous vasopressor drug were evaluated. The clinical outcomes were deﬁned as all-cause death, respiratory failure requiring mechanical ventilation and cardiogenic shock during hospitalization. Cardiogenic shock was deﬁned as hypotension (SBP <90 mmHg) with evidence of hypoperfusion and end-organ dysfunction [25]. The left ventricular ejection fraction (LVEF) before or during index COVID-19 admission was recorded, if available. Statistical analysis All variables that were analyzed in the univariate model were included in the multivariate model to evaluate the comprehensive effects of DFR on the end point event. The hazard ratio (HR) and its 95% CI were calculated. All comparisons were two-tailed, with p < 0.05 considered statistically signiﬁcant. IBM SPSS Statistics 21.0 (IBM Corp. Released 2012. IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp.) was used for the analyses. Statistical analysis Normally distributed continuous variables are reported as mean ± standard deviation (SD), while skewed variables are expressed as medians and interquartile ranges (IQRs). Shapiro-Wilk tests were performed and density maps were drawn to determine the normality of the distribution of the continuous variables. Differences between groups were compared by Chi-square test or Fisher’s exact test. Comparisons of differences between groups were made by analysis of variance (one-way ANOVA). The Kruskal–Wallis H test was used to compare the groups that did not conform to the normal distribution. Patients with a history of HF with COVID-19 were grouped into tertiles 1–3. The ﬁrst tertile included DFR values <0.37, the second tertile included DFR values ranging 0.38–1–1.13 and the third tertile included values higher than 1.13. Patients were also stratiﬁed according to receiver operating characteristic (ROC) curve analysis. The optimal cut-off value for serum D-dimer/ﬁbrinogen ratio was calculated as 0.61; therefore, a Kaplan-Meier survival curve was constructed according to the cut-off value for DFR. The log-rank test was conducted to compare differences. The relationship between each variable and the composite of mechanical ventilation, death Biomark. Med. (Epub ahead of print) 10.2217/bmm-2021-0341 future science group DFR in heart failure with COVID-19 Research Article 0 100 – specificity Sensitivity 0 100 100 20 40 60 80 20 40 80 60 D dimer/fibrinogen ratio First CRP, mg/l First D-dimer, ng/ml First ferritin, ng/dl First fibrinogen, mg/dl First procalcitonin, ng/ml Figure 1. Receiver operating characteristic curve for significant markers in the prediction of clinical outcomes. 0 100 – specificity Sensitivity 0 100 100 20 40 60 80 20 40 80 60 D dimer/fibrinogen ratio First CRP, mg/l First D-dimer, ng/ml First ferritin, ng/dl First fibrinogen, mg/dl First procalcitonin, ng/ml ure 1. Receiver operating characteristic curve for significant markers in the prediction of clinical outcomes and cardiogenic shock events was analyzed by Cox proportional hazard regression. All variables that were analyzed in the univariate model were included in the multivariate model to evaluate the comprehensive effects of DFR on the end point event. The hazard ratio (HR) and its 95% CI were calculated. All comparisons were two-tailed, with p < 0.05 considered statistically signiﬁcant. IBM SPSS Statistics 21.0 (IBM Corp. Released 2012. IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp.) was used for the analyses. and cardiogenic shock events was analyzed by Cox proportional hazard regression. Results In this two-center retrospective study, after exclusion and inclusion criteria, 232 patients with a history of HF with laboratory-conﬁrmed COVID-19 were included in the ﬁnal analysis. In ROC curve analysis, DFR >0.61 predicted poor outcomes in patients with HF and COVID-19 with a sensitivity of 47.6% and speciﬁcity of 90.7%. Patients were categorized in DFR tertiles, ﬁrst (<0.37), second (0.38–1–1.13) and third (>1.13). The mean age of the study population was 73.2 ± 10 years, with males predominant (158; 68.1%). The presence of DM, HT, CAD, prior history of AF and COPD was similar among both groups. With regard to in-hospital outcomes, patients in the third tertile had a higher rate of requiring mechanical ventilation, cardiogenic shock and death (p < 0.001; Table 1). ROC curve analyses were performed to compare the predictive performances among DFR, D-dimer, ﬁbrinogen and some traditional inﬂammatory markers, such as ferritin, CRP and procalcitonin (Figure 1). As presented in Table 2, cut-off values and the area under the ROC curve also were calculated. On the whole, DFR had a better predictive value than other markers. Based on cox proportional analyses; SBP (HR: 0.98; 95% CI: 0.97–0.99; p < 0.001), heart rate (HR: 1.01; 95% CI: 1.00–1.01; p = 0.013), admission oxygen saturation (HR: 0.96; 95% CI: 0.94–0.98; p < 0.001), highest procalcitonin (HR: 1.01; 95% CI: 1.00–1.03; p = 0.022), ferritin (HR: 1.00; 95% CI: 1.00–1.00; p = 0.001), lactate (HR: 1.06; 95% CI: 1.01–1.12; p = 0.020), CRP (HR: 1.00; 95% CI: 1.00–1.00; p = 0.045), total protein (HR: 0.79; 95% CI: 0.64–0.99; p = 0.041) and DFR (HR: 1.03; 95% CI: 1.00–1.07; p = 0.032) predicted clinical outcomes including death, cardiogenic shock and mechanical ventilation events in all populations (Figure 2). Multivariate cox analyses showed that the following were independent prognostic factors of outcomes: SBP (HR: future science group 10.2217/bmm-2021-0341 Research Article Murat, Murat, Dural, Mert & Cavusoglu Table 1. Baseline characteristics, medications, vital signs and laboratory data in patients with COVID-19 and heart Table 1. Baseline characteristics, medications, vital signs and laboratory data in patients with C characteristics, medications, vital signs and laboratory data in patients with COVID-19 and heart Table 1. Baseline characteristics, medications, vital signs and laboratory data in patients with COVID-19 and heart failure. Results Variables Total (n = 232) Tertile 1 (<0.37) n = 79 Tertile 2 (0.38–1.13) n = 76 Tertile 3 (>1.13) n = 77 p-value Age, years, mean ± SD 73.3 ± 10.1 72.9 ± 9.7 74.8 ± 10.4 73.3 ± 10.1 0.297 Male sex, n (%) 158 (68.1) 54 (68.4) 51 (67.1) 53 (68.8) 0.275 Hypertension, n (%) 198 (85.3) 68 (86.1) 62 (81.6) 68 (88.3) 0.529 Diabetes, n (%) 112 (48.3) 35 (44.3) 35 (46.1) 42 (54.5) 0.341 CAD, n (%) 196 (84.5) 66 (83.5) 65 (85.5) 65 (84.4) 0.561 Atrial fibrillation/flutter, n (%) 78 (33.6) 30 (38) 24 (31.6) 24 (31.2) 0.174 COPD, n (%) 63 (27.2) 26 (32.9) 19 (25) 18 (23.4) 0.171 LVEF, %, ±SD 36.71 ± 8.45 38.5 ± 7.85 35.8 ± 8.62 35.7 ± 8.72 0.115 CKD, n (%) 52 (22.4) 16 (20.3) 16 (21.1) 20 (26.0) 0.037 Vital signs at hospital admission, mean ± SD SBP (mmHg) 112.5 ± 22.0 116.2 ± 23.2 114.0 ± 24.1 107.7 ± 18.6 0.021 Heart rate (bpm) 96.6 ± 20.7 92.3 ± 20.2 97.1 ± 21.7 100.6 ± 19.7 0.023 First oxygen saturation (%) 86.3 ± 8.34 88.2 ± 6.97 86.6 ± 9.22 83.9 ± 8.27 0.005 Laboratory data, mean ± SD Haemoglobin (g/dl) 12.2 ± 2.27 12.3 ± 2.30 12.7 ± 2.34 11.8 ± 2.11 0.039 WBC (10∧3/μl) 11.0 ± 5.31 9.83 ± 4.50 10.6 ± 4.73 12.8 ± 6.18 0.004 Platelet (10∧3/μl) 214.7 ± 100.9 224.6 ± 100.5 206.3 ± 75.4 212.7 ± 121.6 0.223 Neutrophil (10∧3/μl) 8.8 ± 5.10 8.00 ± 4.52 8.71 ± 4.44 9.95 ± 6.12 0.078 Lymphocyte (10∧3/μl) 0.95 ± 0.84 1.04 ± 0.79 1.04 ± 1.04 0.75 ± 0.64 0.001 Glucose (mg/dl) 174.6 ± 78.01 156.5 ± 68.0 163.3 ± 65.0 204.3 ± 90.6 0.001 eGFR (ml/min/1.73 m2) 46.0 (27.0–73.0) 52.0 (29.0–75.0) 50.0 (28.0–71.0) 39.0 (20.0–66.5) 0.148 Sodium (mmol/l) 136.2 ± 6.60 136.1 ± 5.7 136.0 ± 5.8 136.3 ± 8.0 0.599 Potassium (mmol/l) 4.52 ± 0.74 4.55 ± 0.64 4.54 ± 0.83 4.46 ± 0.76 0.608 Total protein (g/dl) 6.25 ± 0.93 6.57 ± 0.94 6.27 ± 0.84 5.89 ± 0.87 0.001 Albumin (g/dl) 3.4 (3.1–3.90) 3.7 (3.2–4.0) 3.5 (3.1–3.9) 3.2 (3.0–3.6) 0.001 First fibrinogen (mg/dl) 467.8 ± 178.7 454.4 ± 178.0 500.4 ± 177.4 449.3 ± 178.5 0.261 First procalsitonin (ng/ml) 0.26 (0.1–2.07) 0.15 (0.05–0.28) 0.29 (0.11–3.30) 0.61 (0.20–5.07) 0.001 First D-dimer (ng/ml) 1.1 (0.56–3.6) 0.71 (0.43–1.11) 1.53 (0.61–3.26) 3.67 (0.72–8.60) 0.001 First ferritin (ng/dl) 287.0 (114.5–682.0) 173.9 (78.5–505.8) 295.0 (104.0–678.0) 423.4 (196.4–842.0) 0.001 First LDH (u/l) 288.5 (214.0–474.5) 254.0 (205.0–365.0) 296.0 (212.0–491.0) 366.0 (251.2–567.5) 0.001 First CRP (mg/l) 62.6 (16.32–133.0) 36.9 (11.1–86.0) 74.3 (16.5–150.7) 88.5 (35.2–158.4) 0.001 Highest CRP (mg/l) 117.15 (58.25–190.75) 83.1 (40.0–151.9) 132.9 (77.4–191.9) 146.0 (82.4–222.0) 0.001 Highest troponin (ng/l) 135.9 (31.6–383.3) 68.0 (14.7–270.1) 157.5 (44.2–393.5) 177.2 (59.0–487.3) 0.002 Highest ferritin (ng/dl) 628.0 (362.25–1474.2) 448.5 (254.7–769.0) 625.0 (386.0–1501.0) 1002.0 (481.0–2329) 0.001 Highest D-dimer (ng/ml) 3.2 (1.17–7.99) 0.89 (0.66–1.61) 3.21 (2.61–4.65) 9.14 (6.86–16.27) 0.001 Highest fibrinogen (mg/dl) 529.5 ± 219.4 530.0 ± 223.1 537.4 ± 198.7 521.1 ± 237.6 0.809 Highest LDH, (u/l) 450.0 (293.0–632.5) 356.0 (240.5–491.0) 430.0 (346.0–602.0) 575.0 (403.5–788.0) 0.001 Highest procalsitonin (ng/ml) 1.4 (0.22–7.84) 0.43 (0.13–1.63) 1.60 (0.33–8.50) 4.72 (0.61–16.6) 0.001 Highest lactate (mmol/l) 3.2 (2.07–5.2) 2.50 (1.70–3.80) 2.4 (2.1–4.2) 4.25 (2.45–7.42) 0.001 Clinical outcomes Mechanical ventilation, n (%) 98 (42.2%) 23 (29.1%) 27 (35.5%) 48 (62.3%) 0.001 Death, n (%) 91 (39.2%) 20 (25.3%) 25 (32.9%) 46 (59.7%) 0.001 Cardiogenic shock, n (%) 114 (49.1%) 23 (29.1%) 37 (48.7%) 54 (70.1%) 0.001 Blood transfusion, n (%) 28 (12.1%) 9 (11.4%) 11 (14.5%) 8 (10.4%) 0.297 CAD: Coronary artery disease; CKD: Chronic kidney disease; COPD: Chronic obstructive pulmonary disease; CRP: C-reactive protein; eGFR: Estimated glomerular filtration rate; ICU: Intensive care unit; LDH: Lactate dehydrogenase; SBP: Systolic blood pressure; WBC: White blood cell. Results Age Sex Hypertension Diabetes Coronary artery disease Prior history of atrial fibrillation/flutter LVEF SBP Heart rate First oxygen saturation Platelet Lynphocyte Uric acid Total protein Albumin Highest CRP Highest troponin Highest ferritin D-dimer/fibrinogen ratio Highest LDH Highest procalcitonin Highest lactate HR 0.3548 3.5481 1.122 Figure 2. Forest plot of the univariate Cox proportional analyses for the composite of death, cardiogenic shock and intubation events in the total population (n = 232). 0.97; 95% CI: 0.96–0.99; p = 0.023), highest LDH (HR: 0.99; 95% CI: 0.99–1.00; p = 0.038) and DFR (HR: 1.07; 95% CI: 1.00–1.13; p = 0.027; Table 3). Table 2. The cut-off values and area under the curve for D-dimer/fibrinogen ratio, D-dimer, fibrinogen, ferritin, C-reactive protein and procalcitonin. Variables cut-off values AUC Sen (%) Spe (%) D-dimer/fibrinogen ratio 0.61 0.741 47.6 90.7 D-dimer (ng/ml) 2.89 0.791 75.8 71.4 Fibrinogen (mg/dl) 546 0.682 62.4 73.4 Ferritin (ng/dl) 215 0.714 75.6 62 CRP (mg/l) 88.5 0.724 54.8 82.1 Procalcitonin (mg/ml) 0.202 0.704 70.5 61.9 AUC: Area under the curve; CRP: C-reactive protein. Age Sex Hypertension Diabetes Coronary artery disease Prior history of atrial fibrillation/flutter LVEF SBP Heart rate First oxygen saturation Platelet Lynphocyte Uric acid Total protein Albumin Highest CRP Highest troponin Highest ferritin D-dimer/fibrinogen ratio Highest LDH Highest procalcitonin Highest lactate HR 0.3548 3.5481 1.122 Figure 2. Forest plot of the univariate Cox proportional analyses for the composite of death, cardiogenic shock and intubation events in the total population (n = 232). 0.97; 95% CI: 0.96–0.99; p = 0.023), highest LDH (HR: 0.99; 95% CI: 0.99–1.00; p = 0.038) and DFR (HR: 1.07; 95% CI: 1.00–1.13; p = 0.027; Table 3). Table 2. The cut-off values and area under the curve for D-dimer/fibrinogen ratio, D-dimer, fibrinogen, ferritin, C-reactive protein and procalcitonin. Variables cut-off values AUC Sen (%) Spe (%) D-dimer/fibrinogen ratio 0.61 0.741 47.6 90.7 D-dimer (ng/ml) 2.89 0.791 75.8 71.4 Fibrinogen (mg/dl) 546 0.682 62.4 73.4 Ferritin (ng/dl) 215 0.714 75.6 62 CRP (mg/l) 88.5 0.724 54.8 82.1 Procalcitonin (mg/ml) 0.202 0.704 70.5 61.9 AUC: Area under the curve; CRP: C-reactive protein. ff values and area under the curve for D-dimer/fibrinogen ratio, D-dimer, fibrinogen, ferritin, and procalcitonin. Table 2. The cut-off values and area under the curve for D-dimer/fibrinogen ratio, D-dimer, f C-reactive protein and procalcitonin. Results p Biomark. Med. (Epub ahead of print) 10.2217/bmm-2021-0341 future science group DFR in heart failure with COVID-19 Research Article Table 1. Baseline characteristics, medications, vital signs and laboratory data in patients with COVID-19 and heart failure (cont.). Variables Total (n = 232) Tertile 1 (<0.37) n = 79 Tertile 2 (0.38–1.13) n = 76 Tertile 3 (>1.13) n = 77 p-value Dialyses, n (%) 34 (14.7%) 8 (10.1%) 6 (7.9%) 20 (26.0%) 0.529 Composite outcome, n (%) 124 (53.4%) 26 (32.9%) 41 (53.9%) 57 (74.0%) 0.001 Total length of hospital stay, days 12.71 ± 10.37 9.60 ± 6.40 12.86 ± 8.17 16.40 ± 14.15 0.001 Length of stay in ward (days) 6.0 (2.0–8.0) 6.0 (3.0–8.0) 6.0 (3.25–9.0) 5.0 (0.0–8.0) 0.432 Length of stay in ICU (days) 3.0 (0.0–9.0) 0.0 (0.0–4.0) 3.5 (0.0–11.0) 7.0 (3.0–13.0) 0.001 CAD: Coronary artery disease; CKD: Chronic kidney disease; COPD: Chronic obstructive pulmonary disease; CRP: C-reactive protein; eGFR: Estimated glomerular filtration rate; ICU: Intensive care unit; LDH: Lactate dehydrogenase; SBP: Systolic blood pressure; WBC: White blood cell. Table 1. Baseline characteristics, medications, vital signs and laboratory data in patients with COVID-19 and heart failure (cont.). Variables Total Tertile 1 (<0 37) Tertile 2 (0 38 1 13) Tertile 3 (>1 13) p value racteristics, medications, vital signs and laboratory data in patients with COVID-19 and heart Table 1. Baseline characteristics, medications, vital signs and laboratory data in patients with C failure (cont ) CAD: Coronary artery disease; CKD: Chronic kidney disease; COPD: Chronic obstructive pulmonary disease; CRP: C-reactive protein; eGFR: Estim ICU: Intensive care unit; LDH: Lactate dehydrogenase; SBP: Systolic blood pressure; WBC: White blood cell. CAD: Coronary artery disease; CKD: Chronic kidney disease; COPD: Chronic obstructive pulmonary disease; CRP: C-reactive protein; eGFR: Estimated glomerular filtration rate; ICU: Intensive care unit; LDH: Lactate dehydrogenase; SBP: Systolic blood pressure; WBC: White blood cell. Table 2. The cut-off values and area under the curve for D-dimer/fibrinogen ratio, D-dimer, fibrinogen, ferritin, C-reactive protein and procalcitonin. Variables cut-off values AUC Sen (%) Spe (%) D-dimer/fibrinogen ratio 0.61 0.741 47.6 90.7 D-dimer (ng/ml) 2.89 0.791 75.8 71.4 Fibrinogen (mg/dl) 546 0.682 62.4 73.4 Ferritin (ng/dl) 215 0.714 75.6 62 CRP (mg/l) 88.5 0.724 54.8 82.1 Procalcitonin (mg/ml) 0.202 0.704 70.5 61.9 AUC: Area under the curve; CRP: C-reactive protein. Results Variable p-value Hazard ratio 95% CI Lower Upper SBP, mmHg, mean ± SD 0.023 0.97 0.96 0.99 D-dimer/fibrinogen ratio 0.027 1.07 1.00 1.13 Highest LDH (u/l) 0.038 0.99 0.99 1.00 LDH: Lactate dehydrogenase; SBP: Systolic blood pressure. Table 3. Enter methods of multivariate Cox proportional analyses for composite of death, cardiogenic shock and intubation events in the total population (n = 232). 0.0 Length of stay hospital Cum survival 0.0 100.0 0.2 0.4 0.6 0.8 1.0 20.0 40.0 60.0 80.0 Log-rank test: 0.029 D-dimer/fibrinogen ratio ≤0.61 D-dimer/fibrinogen ratio >0.61 Figure 3. Kaplan-Meier survival curves for clinical outcomes in patients diagnosed with COVID-19 with a history of heart failure according to higher and lower D-dimer/fibrinogen ratio. 0.0 Length of stay hospital Cum survival 0.0 100.0 0.2 0.4 0.6 0.8 1.0 20.0 40.0 60.0 80.0 Log-rank test: 0.029 D-dimer/fibrinogen ratio ≤0.61 D-dimer/fibrinogen ratio >0.61 Figure 3. Kaplan-Meier survival curves for clinical outcomes in patients diagnosed with COVID-19 with a history of heart failure according to higher and lower D-dimer/fibrinogen ratio. Figure 3. Kaplan-Meier survival curves for clinical outcomes in patients diagnosed with COVID-19 with a history of heart failure according to higher and lower D-dimer/fibrinogen ratio. Results Age Sex Hypertension Diabetes Coronary artery disease Prior history of atrial fibrillation/flutter LVEF SBP Heart rate First oxygen saturation Platelet Lynphocyte Uric acid Total protein Albumin Highest CRP Highest troponin Highest ferritin D-dimer/fibrinogen ratio Highest LDH Highest procalcitonin Highest lactate HR 0.3548 3.5481 1.122 Figure 2. Forest plot of the univariate Cox proportional analyses for the composite of death, cardiogenic shock and intubation events in the total population (n = 232). Age Sex Hypertension Diabetes Coronary artery disease Prior history of atrial fibrillation/flutter LVEF SBP Heart rate First oxygen saturation Platelet Lynphocyte Uric acid Total protein Albumin Highest CRP Highest troponin Highest ferritin D-dimer/fibrinogen ratio Highest LDH Highest procalcitonin Highest lactate HR 0.3548 3.5481 1.122 Figure 2. Forest plot of the univariate Cox proportional analyses for the composite of death, cardiogenic shock and intubation events in the total population (n = 232). Figure 2. Forest plot of the univariate Cox proportional analyses for the composite of death, cardiogenic shock and intubation events in the total population (n = 232). 0.97; 95% CI: 0.96–0.99; p = 0.023), highest LDH (HR: 0.99; 95% CI: 0.99–1.00; p = 0.038) and DFR (HR: 1.07; 95% CI: 1.00–1.13; p = 0.027; Table 3). 0.97; 95% CI: 0.96–0.99; p = 0.023), highest LDH (HR: 0.99; 95% CI: 0.99–1.00; p = 0.038) and DFR (HR: 1.07; 95% CI: 1.00–1.13; p = 0.027; Table 3). future science group 10.2217/bmm-2021-0341 Research Article Murat, Murat, Dural, Mert & Cavusoglu Table 3. Enter methods of multivariate Cox proportional analyses for composite of death, cardiogenic shock and intubation events in the total population (n = 232). Variable p-value Hazard ratio 95% CI Lower Upper SBP, mmHg, mean ± SD 0.023 0.97 0.96 0.99 D-dimer/fibrinogen ratio 0.027 1.07 1.00 1.13 Highest LDH (u/l) 0.038 0.99 0.99 1.00 LDH: Lactate dehydrogenase; SBP: Systolic blood pressure. 0.0 Length of stay hospital Cum survival 0.0 100.0 0.2 0.4 0.6 0.8 1.0 20.0 40.0 60.0 80.0 Log-rank test: 0.029 D-dimer/fibrinogen ratio ≤0.61 D-dimer/fibrinogen ratio >0.61 Figure 3. Kaplan-Meier survival curves for clinical outcomes in patients diagnosed with COVID-19 with a history of heart failure according to higher and lower D-dimer/fibrinogen ratio. Discussion Table 3. Enter methods of multivariate Cox proportional analyses for composite of death, cardiogenic shock and intubation events in the total population (n = 232). Discussion In this study, 27.5% of HF patients diagnosed with COVID-19 had a high DFR (>0.61) at admission and in patients with a history of HF diagnosed with COVID-19, DFR elevation upon admission was associated with poor clinical outcomes, longer hospital stays, length of stay in the ICU and in-hospital mortality (Figure 3). Previous studies have shown the correlation between plasma D-dimer and ﬁbrinogen concentrations and the severity of COVID-19. Fibrinogen is an acute-phase protein, which is synthesized by the liver in response to IL-1- and IL-6-derived stimulation, and is involved in ﬁbrin formation as the last step of a triggered coagulation activity [26]. Fibrinogen has become an important biomarker in the course of COVID-19 disease, as it is associated with both inﬂammation and coagulopathy. Han et al. investigated the changes in blood coagulation of patients infected with COVID-19 by comparing them with healthy controls. They reported that ﬁbrinogen levels were higher in both mild and severely ill patients than healthy patients [17]. In another study, the difference in ﬁbrinogen was reported to be nonsigniﬁcant between surviving and nonsurviving patients with COVID-19 in a different cohort (5.16 vs 4.51 g/l, p = 0.149) [18]. Hayıroglu et al. reported that ﬁbrinogen might not have a predictive value for mortality in patients with COVID-19 and should be evaluated together with D-dimer for proper prognostic Biomark. Med. (Epub ahead of print) 10.2217/bmm-2021-0341 future science group DFR in heart failure with COVID-19 Research Article DFR in heart failure with COVID-19 Research Article Research Article assumption [19]. Although the mortality rate in the third tertile DFR group was 59.7%, there was no difference in ﬁbrinogen value between the groups. In addition, ﬁbrinogen alone was not found to be predictive in terms of mortality and clinical outcomes. y D-dimer, which is produced by the breakdown of ﬁbrin by plasmin, is another biomarker closely related to thrombotic and ﬁbrinolytic processes. Apart from the diagnosis of coagulation abnormalities, D-dimer increases in many conditions such as inﬂammation, vasculitis, pregnancy, cancer and HF. Several studies have shown that D- dimer levels are associated with the severity and clinical outcomes of community-acquired pneumonia [27]. Among adults admitted to the emergency room, infection,s rather than venous thromboembolism (VTE)/pulmonary embolism (PE), are the most common reason for D-dimer elevation [28]. Discussion In addition, some studies have reported that a high D-dimer level is a highly nonspeciﬁc marker of VTE and may be a sign of inﬂammation rather than thrombosis [29,30]. Many studies have previously evaluated the relationship between COVID-19 and D-dimer. In some studies, the level of D-dimer reﬂects the severity of the disease, while in others it is useful in predicting in-hospital prognosis. A large study of 1065 hospitalized patients with COVID-19 reported that higher D-dimer at admission was associated with a greater risk of all-cause mortality, need for mechanical ventilation and VTE. The investigators also concluded that D-dimer at admission, as an isolated measure, did not appear to be a reliable prognostic test for outcomes among patients with COVID-19 [31]. Although D-dimer has been well studied, the predictive effect of DFR in COVID-19 disease has not been examined. Furthermore, despite the results of these studies, the prognostic and predictive value of D-dimer has not been studied in patients with COVID-19 and speciﬁc diseases. To the best of our knowledge, this study is the ﬁrst to investigate the association of DFR with the outcomes of patients with HF and COVID-19. p In this study, the AUC was found to be 0.74 for the ROC of DFR at admission, which is considered a predictor of in-hospital mortality. The optimal DFR cut-off value of 0.61 provided sensitivity and speciﬁcity for the prediction of in-hospital outcomes. Unlike previous studies in which D-dimer was evaluated, this study was conducted in a speciﬁc patient population with high in-hospital poor outcomes. A study of 343 inpatients with COVID-19 from Wuhan reported that AUC of 0.89 for a ROC of D-dimer at admission as a predictor of in-hospital mortality. They also reported that the optimal D-dimer cutoff of 2 μg/ml for prediction of death. However, these ﬁndings were based on only 13 death events in the cohort – hardly sufﬁcient to construct a rigorous and reliable ROC [32]. In another study of 138 consecutive patients with COVID-19 indicated that ‘D-dimer was higher in non-survivors than in survivors’, however, this was based on a subgroup analysis that included 33 patients, only ﬁve of whom were nonsurvivors [33]. Discussion In the present study of patients with HF, (which is an important group in terms of COVID- 19 mortality), high DFR at admission was more speciﬁc than D-dimer and ﬁbrinogen in predicting clinical outcomes, such as requiring mechanical ventilation, cardiogenic shock, in-hospital death, and length of hospital and ICU stay. y Severe COVID-19 is commonly complicated with coagulopathy, but the elevation of D-dimer seen in patients with COVID-19 may be associated with inﬂammation rather than venous thromboembolism [33,34]. In a study that included 449 patients with COVID-19, there was no difference in D-dimer at admission as compared with 104 patients with non-COVID pneumonia [34]. Likewise, the last two studies showed that D-dimer correlates with inﬂammatory markers, such as CRP and procalcitonin, and there is no deﬁnitive conclusion about its direct relationship with VTE [31,35]. In the present study, a high DFR value was signiﬁcantly correlated with prognosis- determinant biomarkers of both inﬂammation and disease, such as CRP, LDH, procalcitonin, lactate, troponin and ferritin. When evaluated together with these parameters, DFR was found to be associated with in-hospital outcomes in both univariate and multivariate analyses. Inﬂammatory activation may be the underlying cause of worse clinical outcomes in both patients with HF and those with COVID-19. Limitations The present study has several limitations. First, it is a retrospective, observational study. Second, there was a lack of evaluation of conditions such as in-hospital undiagnosed PE and VTE. This may have contributed to the dynamic changes in ﬁbrinogen and D-dimer levels observed in this study. Further studies are needed to conﬁrm these ﬁndings. References Papers of special note have been highlighted as: • of interest; •• of considerable interest Papers of special note have been highlighted as: • of interest; •• of considerable interest Papers of special note have been highlighted as: • of interest; •• of considerable interest 1. Huang C, Wang Y, Li X et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet 395(10223), 497–506 (2020). 1. Huang C, Wang Y, Li X et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet 395(10223), 497–506 (2020). 2. Yonas E, Alwi I, Pranata R et al. Effect of heart failure on the outcome of COVID-19 – a meta analysis and systematic review Emerg. Med. 6, 204–211 (2021). 2. Yonas E, Alwi I, Pranata R et al. Effect of heart failure on the outcome of COVID-19 – a meta analysis and systematic review. Am. J. Emerg. Med. 6, 204–211 (2021). 3. Zheng Y-Y, Ma Y-T, Zhang J-Y, Xie X. COVID-19 and the cardiovascular system. Nat. Rev. Cardiol. 17(5), 259–260 (2020 4. Madjid M, Safavi-Naeini P, Solomon SD, Vardeny O. Potential effects of coronaviruses on the cardiovascular system: a review. JAMA Cardiol. 5(7), 831–840 (2020). 4. Madjid M, Safavi-Naeini P, Solomon SD, Vardeny O. Potential effects of coronaviruses on the cardiovascular system: a review. JAMA Cardiol. 5(7), 831–840 (2020). 5. Clerkin KJ, Fried JA, Raikhelkar J et al. COVID-19 and cardiovascular disease. Circulation 141(20), 1648–1655 (2020). 6. Tomasoni D, Italia L, Adamo M et al. COVID-19 and heart failure: from infection to inﬂammation and angiotensin II stim Searching for evidence from a new disease. Eur. J. Heart Failure 22(6), 957–966 (2020). •• Although COVID-19 is transmitted through breathing, primarily causing pneumonia, this paper states that it signiﬁcantly affects all organs, including the heart. 7. Wang D, Hu B, Hu C et al. Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus – infected pneumonia in Wuhan, China. JAMA 323(11), 1061–1069 (2020). 8. Pranata R, Huang I, Lim MA, Wahjoepramono EJ, July J. Impact of cerebrovascular and cardiovascular diseases on mortality and severity of COVID-19 – systematic review, meta-analysis, and meta-regression. J. Stroke Cerebrovasc. Dis. 29(8), 104949 (2020). 9. Yang X, Yu Y, Xu J et al. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir. Med. 8(5), 475–481 (2020). 10. Ethical conduct of research This study was carried out in line with research regulations, including the approval of the Ethics Committee of Eskisehir Osmangazi University, dated 30/03/2021 and numbered 2021-03/10 and according to the principles of the ‘World Medical Association Helsinki Declaration’. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or finan- cial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. No writing assistance was utilized in the production of this manuscript. Ethical conduct of research Conclusion The DFR value at admission may guide physicians in determining the outcomes, length of hospital stay, treatment protocol and regimen for patients with a history of HF who are admitted with a diagnosis of COVID-19. f future science group 10.2217/bmm-2021-0341 Research Article Murat, Murat, Dural, Mert & Cavusoglu rt failure is one of the most important causes of mortality in patients with COVID-19. • Previous studies have shown that fibrinogen and D-dimer are associated with the severity of COVID-19 disease; however, the prognostic value of DFR in patients with COVID-19 with a history of heart failure is unclear. • Previous studies have shown that fibrinogen and D-dimer are associated with the severity of COVID-19 disease; however, the prognostic value of DFR in patients with COVID-19 with a history of heart failure is unclear. • Elevated DFR values may define prothrombotic activity in conditions with excessive fibrinogen consumption and D dimer formation independent of the absolute values of both • Elevated DFR values may define prothrombotic activity in conditions with excessive fibrinogen consumption and D-dimer formation, independent of the absolute values of both. p • Laboratory parameters of patients with COVID-19 have shown a prothrombotic diathesis with significantly high fibrinogen levels in critically ill patients. However, in the late stages, thrombolysis decreases fibrinogen levels and increases D-dimer. • Laboratory parameters of patients with COVID-19 have shown a prothrombotic diathesis with significantly high fibrinogen levels in critically ill patients. However, in the late stages, thrombolysis decreases fibrinogen levels and increases D-dimer. • Multivariate Cox regression analysis showed that DFR was significantly associated with in-hospital outcomes. • DFR can be used as a biomarker in patients with COVID-19 and heart failure, which is associated with increased mortality and worse outcomes. • Multivariate Cox regression analysis showed that DFR was significantly associated with in-hospital outcomes. • DFR can be used as a biomarker in patients with COVID-19 and heart failure, which is associated with increased mortality and worse outcomes. Author contributions Constructing an idea or hypothesis for research: S Murat. Planning methodology: S Murat and Y Cavusoglu. Data collection: B Murat, M Dural and GO Mert. Statistical analysis: S Murat and B Murat. Writing: S Murat and B Murat. Financial & competing interests disclosure DFR in heart failure with COVID-19 Research Article DFR in heart failure with COVID-19 Research Article DFR in heart failure with COVID-19 Research Article 12. Henry BM, Vikse J, Benoit S, Favaloro EJ, Lippi G. Hyperinﬂammation and derangement of renin-angiotensin-aldosterone system in COVID-19: a novel hypothesis for clinically suspected hypercoagulopathy and microvascular immunothrombosis. Clin. Chim. Acta 507, 167–173 (2020). 12. Henry BM, Vikse J, Benoit S, Favaloro EJ, Lippi G. Hyperinﬂammation and derangement of renin-angiotensin-aldosterone system in COVID-19: a novel hypothesis for clinically suspected hypercoagulopathy and microvascular immunothrombosis. Clin. Chim. Acta 507, 167–173 (2020). 13. Perico L, Benigni A, Casiraghi F, Ng LF, Renia L, Remuzzi G. Immunity, endothelial injury and complement-induced coagulopathy in COVID-19. Nat. Rev. Nephrol. 17(1), 46–64 (2021). •• States that COVID-19-positive patients have abnormal blood test results. •• States that COVID-19-positive patients have abnormal blood test results. 14. Arachchillage DR, Laffan M. Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia. J. Thromb. Haemostas. 18(5), 1233–1234 (2020). 15. Bai Y, Yao L, Wei T et al. Presumed asymptomatic carrier transmission of COVID-19. JAMA 323(14), 1406–1407 (2020). •• This paper discusses the confusion regarding the measurement method and reference ranges for the use of D-dimer as a biomarker. 16. Favaloro EJ, Thachil J. Reporting of D-dimer data in COVID-19: some confusion and potential for misinformation. Clin. Chem. Lab. Med. 58(8), 1191–1199 (2020). 17. Han H, Yang L, Liu R et al. Prominent changes in blood coagulation of patients with SARS-CoV-2 infection. Clin. Chem. Lab. Med. 58(7), 1116–1120 (2020). 18. Tang N, Li D, Wang X, Sun Z. Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia. J. Thromb. Haemost. 18(4), 844–847 (2020). • Fibrinogen and D-dimer variants and anticoagulant therapy have been employed in COVID-19 disease. In addition, the authors emphasize that ﬁbrinogen should not be evaluated alone. 19. Hayıro˘glu M, C¸ ınar T, Tekkes¸in A. Fibrinogen and D-dimer variances and anticoagulation recommendations in Covid-19: literature review. Rev. Assoc. Med. Bras. 66(6), 842–848 (2020). 20. Kara H, Bayir A, Degirmenci S et al. D-dimer and D-dimer/ﬁbrinogen ratio in predicting pulmonary embolism in patients a hospital emergency department. Acta Clin. Belg. 69(4), 240–245 (2014). 21. Hajsadeghi S, Kerman SR, Khojandi M et al. Accuracy of D-dimer: ﬁbrinogen ratio to diagnose pulmonary thromboembolism in patients admitted to intensive care units. Cardiovasc. J. Afr. 23(8), 446–456 (2012). 22. Kucher N, Kohler HP, Dornh¨ofer T, Wallmann D, L¨ammle B. DFR in heart failure with COVID-19 Research Article Accuracy of D-dimer/ﬁbrinogen ratio to predict pulmonary embolism: a prospective diagnostic study. J. Thromb. Haemost. 1(4), 708–713 (2003). 23. Alvarez-Perez FJ, Castelo-Branco M, Alvarez-Sabin J. Usefulness of measurement of ﬁbrinogen, D-dimer, D-dimer/ﬁbrinogen ratio, C reactive protein and erythrocyte sedimentation rate to assess the pathophysiology and mechanism of ischaemic stroke. J. Neurol. Neurosurg. Psychiatry 82(9), 986–992 (2011). • States the prognostic value of D-dimer in hospitalized patients with heart failure. 24. Zhao TJ, Yang QK, Tan CY, Bi LD, Li J, Miao ZL. Prognostic value of D-dimer/ﬁbrinogen ratio in the adverse outcomes of patients hospitalized for heart failure. Biomark. Med. 14(18), 1733–1745 (2020). 25. Ponikowski P, Voors AA, Anker SD et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC) developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur. Heart J. 37(27), 2129–2200 (2016). 26. Levi M, Toh CH, Thachil J, Watson HG. Guidelines for the diagnosis and management of disseminated intravascular coagulation. British Committee for Standards in Haematology. Br. J. Haematol. 145(1), 24–33 (2009). • Indicates that D-dimer is an important biomarker for predicting disease severity and mortality in patients with COVID-19. 27. Yao Y, Cao J, Wang Q et al. D-dimer as a biomarker for disease severity and mortality in COVID-19 patients: a case control study. J. Intensive Care 8, 49 (2020). 28. Lippi G, Bonfanti L, Saccenti C, Cervellin G. Causes of elevated D-dimer in patients admitted to a large urban emergency department. Eur. J. Intern. Med. 25(1), 45–48 (2014). 29. Borowiec A, D˛abrowski R, Kowalik I et al. Elevated levels of d-dimer are associated with inﬂammation and disease activity rather than risk of venous thromboembolism in patients with granulomatosis with polyangiitis in long term observation. Adv. Med. Sci. 65(1), 97–101 (2020). 30. Shorr AF, Thomas SJ, Alkins SA, Fitzpatrick TM, Ling GS. D-dimer correlates with proinﬂammatory cytokine levels and outcomes in critically ill patients. Chest 121(4), 1262–1268 (2002). 31. Naymagon L, Zubizarreta N, Feld J et al. Admission D-dimer levels, D-dimer trends, and outcomes in COVID-19. Thromb. Res. 196, 99–105 (2020). 32. Zhang L, Yan X, Fan Q et al. D-dimer levels on admission to predict in-hospital mortality in patients with Covid-19. J. Thromb. Haemost. 18(6), 1324–1329 (2020). 33. References Dalia T, Lahan S, Ranka S et al. Impact of congestive heart failure and role of cardiac biomarkers in COVID-19 patients: a systematic review and meta-analysis. Indian Heart J. 73(1), 91–98 (2021). 11. Ciceri F, Beretta L, Scandroglio AM et al. Microvascular COVID-19 lung vessels obstructive thromboinﬂammatory syndrome (MicroCLOTS): an atypical acute respiratory distress syndrome working hypothesis. Crit. Care Resus. 22(2), 95 (2020). Biomark. Med. (Epub ahead of print) 10.2217/bmm-2021-0341 future science group Research Article Murat, Murat, Dural, Mert & Cavusoglu 34. Yin S, Huang M, Li D, Tang N. Difference of coagulation features between severe pneumonia induced by SARS-CoV2 and non-SARS-CoV2. J. Thromb. Thrombolysis doi:10.1007/s11239-020-02105-8 (2020) (Epub ahead of print). 35. Al-Samkari H, Karp Leaf RS, Dzik WH et al. COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infection. Blood 136(4), 489–500 (2020). DFR in heart failure with COVID-19 Research Article Wang D, Hu B, Hu C et al. Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China. JAMA 323(11), 1061–1069 (2020). future science group 10.2217/bmm-2021-0341 Research Article Murat, Murat, Dural, Mert & Cavusoglu 35. Al-Samkari H, Karp Leaf RS, Dzik WH et al. COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infection. Blood 136(4), 489–500 (2020). 35. Al-Samkari H, Karp Leaf RS, Dzik WH et al. COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infection. Blood 136(4), 489–500 (2020). p Biomark. Med. (Epub ahead of print) 10.2217/bmm-2021-0341 future science group
https://openalex.org/W2984930179	https://www.aclweb.org/anthology/D19-5205.pdf	English	null	English to Hindi Multi-modal Neural Machine Translation and Hindi Image Captioning	null	2,019	cc-by	3,700	Abstract source-target text to translate sentences (Shah et al., 2016). Interestingly, multi-modal concept improved the translation quality of generating the captions of the images (Dash et al., 2019) as well as signiﬁcant improvement over text-only NMT system (Huang et al., 2016). In text-only NMT system, the encoder-decoder framework of NMT is a widely accepted technique used in the task of MT. Because it handles sequence to sequence learning problem for variable length source and target sentences and also, handles long term de- pendency problem using long short term memory (LSTM) (Sutskever et al., 2014). The demer- its of basic encoder-decoder model is that it fails to encode all necessary information into the con- text vector when the sentence is too long. Hence, to handle such problem attention-based encoder- decoder model is introduced, which allows the decoder to focus on different parts of the source sequence at different decoding steps (Bahdanau et al., 2015). (Luong et al., 2015) enhanced the attention model that merges global, accompany- ing to all source words and local, only pay atten- tion to a part of source words. The attention-based NMT system shows a promising outcome in vari- ous languages (Pathak and Pakray, 2018; Pathak et al., 2018; Laskar et al., 2019). Current work has been investigated for English to Hindi translation. There are three different tracks, namely, multi- modal translation, Hindi-only image captioning and text-only translation using NMT system and participated in WAT2019 multi-modal translation task. With the widespread use of Machine Trans- lation (MT) techniques, attempt to minimize communication gap among people from di- verse linguistic backgrounds. We have par- ticipated in Workshop on Asian Transla- tion 2019 (WAT2019) multi-modal translation task. There are three types of submission track namely, multi-modal translation, Hindi- only image captioning and text-only transla- tion for English to Hindi translation. The main challenge is to provide a precise MT output. The multi-modal concept incorporates textual and visual features in the translation task. In this work, multi-modal translation track re- lies on pre-trained convolutional neural net- works (CNN) with Visual Geometry Group having 19 layered (VGG19) to extract image features and attention-based Neural Machine Translation (NMT) system for translation. The merge-model of recurrent neural network (RNN) and CNN is used for the Hindi-only image captioning. The text-only translation track is based on the transformer model of the NMT system. English to Hindi Multi-modal Neural Machine Translation and Hindi Image Captioning Sahinur Rahman Laskar, Rohit Pratap Singh, Partha Pakray and Sivaji Bandyopadhya Department of Computer Science and Engineering National Institute of Technology Silchar Assam, India {sahinurlaskar.nits,rohitkako,parthapakray,sivaji.cse.ju}@gmail. Sahinur Rahman Laskar, Rohit Pratap Singh, Partha Pakray and Sivaji Bandyopadhyay Department of Computer Science and Engineering National Institute of Technology Silchar Assam, India {sahinurlaskar.nits,rohitkako,parthapakray,sivaji.cse.ju}@gmail.com Abstract The ofﬁcial results evaluated at WAT2019 translation task, which shows that our multi-modal NMT system achieved Bilin- gual Evaluation Understudy (BLEU) score 20.37, Rank-based Intuitive Bilingual Eval- uation Score (RIBES) 0.642838, Adequacy- Fluency Metrics (AMFM) score 0.668260 for challenge test data and BLEU score 40.55, RIBES 0.760080, AMFM score 0.770860 for evaluation test data in English to Hindi multi- modal translation respectively. Proceedings of the 6th Workshop on Asian Translation, pages 62–67 Hong Kong, China, November 4, 2019. c⃝2019 Association for Computational Linguistics 1 2 Related Works The multi-modal translation is an emerging task of the MT community, where visual features of image combine with textual features of parallel Literature survey mainly focused on multimodal based NMT works, where multimodal informa- 62 Proceedings of the 6th Workshop on Asian Translation, pages 62–67 Hong Kong, China, November 4, 2019. c⃝2019 Association for Computational Linguistics 1 have used global features of the images. tion (text and image) integrating into the attention- based encoder-decoder architecture. (Huang et al., 2016), proposed a model using attention based NMT, where regional and global visual fea- tures are attached in parallel with multiple encod- ing threads and each thread is followed by the text sequence. They obtained BLEU score 36.5, which outperformed the text-only baseline model BLEU score 34.5. (Calixto and Liu, 2017) used bi- directional recurrent neural network (RNN) with gated recurrent unit (GRU) in the encoding phase instead of single-layer unidirectional LSTM in (Huang et al., 2016) and also, used image features separately either as a word in the source sentence or directly for encoder or decoder initialization unlike word only in (Huang et al., 2016), achieved BLEU score 38.5, 43.9 in English to German and German to English translation respectively. (Cal- ixto et al., 2017), introduced two independent at- tention mechanisms over source language words and visual features in a single decoder RNN, which signiﬁcantly improve over the models used in (Huang et al., 2016), obtained BLEU score 39.0, 43.2 in English to German and German to English translation respectively. (Dutta Chowd- hury et al., 2018), investigated multimodel NMT following settings of (Calixto and Liu, 2017) for Hindi to English translation and acquired BLEU score 24.2. 3 System Description The primary steps of the system operations are data preprocessing, system training and system testing and the same have been illustrated in fol- lowing subsections. The multimodal NMT toolkit (Calixto and Liu, 2017; Calixto et al., 2017) is em- ployed to build the multimodal NMT system for multimodal translation task, which are based on the pytorch port of OpenNMT (Klein et al., 2017). For text-only translation task, OpenNMT is de- ployed to build the NMT system and in the case of Hindi-only image captioning track, publicly avail- able VGG16 and LSTM in Keras library, are used to build the system (Simonyan and Zisserman, 2015; Tanti et al., 2018). We have used Hindi visual genome dataset in each track of WAT2019 multi-modal translation task provided by the or- ganizer (Nakazawa et al., 2019). We have not used image coordinates (Width, Height) provided in the dataset to indicate the rectangular region in the image described by the caption. Because, we 3.1 Data Preprocessing The data preprocessing steps of each track are car- ried out separately. In the multi-modal translation track, ﬁrstly, image features for training, valida- tion and test data are extracted from the image data set as mentioned in Table 1. We have used publicly available pre-trained CNN with VGG19 via batch normalization for extraction of both global and local visual features from the image dataset as shown in Table 1. Secondly, pri- mary functions of preprocessing step, tokeniza- tion, lowercasing and applying byte pair encod- ing (BPE) model of source and target sentences. For this purpose, OpenNMT toolkit is used to make a dictionary of vocabulary size of dimension 8300, 7984 for English-Hindi parallel sentence pairs, which indexes the words during the train- ing process. In the text-only translation track, we have considered only source-target corresponding sentences as shown in Table 1 to build the dic- tionary, vocabulary size of dimension 8300, 7984 using the OpenNMT toolkit. In the Hindi-only im- age captioning track, image features are extracted via pre-trained CNN with VGG16 from the image data set as shown in Table 1. The image extracted features are 1-dimensional 4,096 element vector. The text input sequences, maximum description length of 22 words, are cleaned to get the vocabu- lary size of 5605. 3.3 System Testing System training is followed by the system test- ing process in each track separately. This process is required for predicting translations of test in- stances/items as shown in Table 1. 3.2 System Training After preprocessing of data, the system training process is performed in each track separately in Multiple Graphics Processing Units (GPU) envi- ronment to boost the performance of training. In the multi-modal translation track, the source (En- glish) and target (Hindi) sentences are fed into encoder-decoder RNN. The multi-modal NMT system is trained using doubly-attentive decoder following settings of (Calixto et al., 2017), where the multi-modal NMT incorporates two different attention mechanism across the source-language words and visual features in a single decoder RNN. Both encoder and decoder consists of a two- layer network of LSTM nodes, which contains 500 units in each layer. The multi-modal NMT system is trained up to 100 epoch. The default set- tings drop out of 0.3, batch size 40 and layer nor- malization are used for a stable training run. In the 63 2 Nature of corpus Name of Corpus Number of instances/items Training Englsih-Hindi 28,929 (Text data) Image data 28,929 Test (Evaluation Set) English to Hindi 1595 (Text data) Image data 1595 Test (Challenge Set) English to Hindi 1400 (Text data) Image data 1400 Validation English-Hindi 998 (Text data) Image data 998 Table 1: Corpus Statistics (Nakazawa et al., 2019). Nature of corpus Name of Corpus Number of instances/items Table 1: Corpus Statistics (Nakazawa et al., 2019). 2002), RIBES (Isozaki et al., 2010) and AMFM (Banchs et al., 2015) are used to measure perfor- mance of predicted translations. We have partic- ipated in all the track of the multi-modal trans- lation task and our team name is 683. In multi- modal translation track, a total of three teams, in- cluding our team participated for both challenge and evaluation test data in English to Hindi trans- lation. We have acquired BLEU, RIBES, AMFM score 20.37, 0.642838, 0.668260 for challenge test set and BLEU, RIBES, AMFM score 40.55, 0.760080, 0.770860 for evaluation test set respec- tively, higher than other teams as shown in Ta- ble 2. However, we have attained lower BLEU, RIBES and AMFM scores than other teams in text-only and Hindi-only image captioning trans- lation track as shown in Table 3 and 4 respectively. Moreover, from Table 2, 3 and 4, it is observed that when translating English to Hindi our multi- modal translation outperforms our text only trans- lation as well as our Hindi-only image caption- ing. 3.2 System Training To further analyze the best and worst per- formance of multi-modal translation in compari- son to text-only and Hindi-only image captioning, sample predicted sentences on challenge test data, reference target sentences and Google translation on same test data are considered in Table 5, 6. In Table 5, our multi-modal NMT system provides perfect prediction like reference target sentence, Google translation and close to text-only trans- lation but wrong translation in Hindi-only image captioning. However, in Table 6, prediction of source word “court” is inappropriate like Google translation, text-only translation and wrong trans- lation in Hindi-only image captioning. training process of text-only translation track, the NMT system is trained up to 25,000 epoch to build the train models by transformer model of NMT system. For a small dataset in text-only transla- tion, it is not required up to 25,000 epoch. But in this dataset, we need to trained up to 25,000 be- cause of learning curve grows up to 24,000 then falls. Hence, we have chosen predicted translation at an optimum point on 24,000 epoch. In the train- ing process of Hindi-only image captioning track, we have used merge-model following settings of (Tanti et al., 2018). The preprocessed image fea- ture vector of 4096 elements are processed by a dense layer to provide 256 elements for represen- tation of the image. Afterward, the input text se- quence of 22 words length are fed into a word em- bedding layer to convert it into vector form which is followed by LSTM based RNN layer contains 256 nodes. Both the ﬁxed-length vectors (Image and text) generated are merged together and pro- cessed by a dense layer to build the train models up to 20 epoch. 1http://lotus.kuee.kyoto-u.ac.jp/WAT/evaluation/index.html 4 Result and Analysis 5 Conclusion and Future Work Current work participates in three different trans- lation tracks at WAT2019 namely, multi-modal, text-only and Hindi-only image captioning for 6 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 System BLEU Challenge Test Set Evaluation Test Set System-1 (Our system) 20.37 40.55 System-2 12.58 28.45 System-3 11.77 28.27 System-4 10.19 27.39 RIBES Challenge Test Set Evaluation Test Set System-1 (Our system) 0.642838 0.760080 System-2 0.507192 0.692880 System-3 0.487897 0.676444 System-4 0.482373 0.634567 AMFM Challenge Test Set Evaluation Test Set System-1 (Our system) 0.668260 0.770860 System-2 0.659840 0.722110 System-3 0.632060 0.707520 System-4 0.559990 0.682060 Table 2: BLEU, RIBES and AMFM scores result of participated teams for multi-modal translation track. System BLEU Challenge Test Set Evaluation Test Set System-1 30.94 41.32 System-2 30.34 - System-3 - 38.95 System-4 (Our system) 15.85 38.19 System-5 (Our system) 14.69 25.34 System-6 5.56 20.13 RIBES Challenge Test Set Evaluation Test Set System-1 0.734435 0.770754 System-2 0.726998 - System-3 - 0.749535 System-4 (Our system) 0.550964 0.744158 System-5 (Our system) 0.550568 0.636152 System-6 0.373560 0.574366 AMFM Challenge Test Set Evaluation Test Set System-1 0.775890 0.784950 System-2 0.773260 - System-3 (Our system) 0.632910 0.763940 System-4 - 0.762180 System-5 (Our system) 0.578930 0.656370 System-6 0.461110 0.615290 Table 3: BLEU, RIBES and AMFM scores result of participated teams for text-only translation track. Table 3: BLEU, RIBES and AMFM scores result of participated teams for text-only translation track. System Challenge Test Set RIBES AMFM System-1 0.080028 0.385960 System-2 (Our system) 0.034482 0.335390 Table 4: RIBES, AMFM scores result of participated teams for Hindi-only image captioning track. Table 4: RIBES, AMFM scores result of participated teams for Hindi-only image captioning track. 4 Result and Analysis The ofﬁcial evaluation results of the competition for English to Hindi multi-modal translation task are reported by the organizer 1. Automatic eval- uation metrics namely, BLEU (Papineni et al., 64 3 65 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 System BLEU Challenge Test Set Evaluation Test Set System-1 (Our system) 20.37 40.55 System-2 12.58 28.45 System-3 11.77 28.27 System-4 10.19 27.39 RIBES Challenge Test Set Evaluation Test Set System-1 (Our system) 0.642838 0.760080 System-2 0.507192 0.692880 System-3 0.487897 0.676444 System-4 0.482373 0.634567 AMFM Challenge Test Set Evaluation Test Set System-1 (Our system) 0.668260 0.770860 System-2 0.659840 0.722110 System-3 0.632060 0.707520 System-4 0.559990 0.682060 Table 2: BLEU, RIBES and AMFM scores result of participated teams for multi-modal translation track. System BLEU Challenge Test Set Evaluation Test Set System-1 30.94 41.32 System-2 30.34 - System-3 - 38.95 System-4 (Our system) 15.85 38.19 System-5 (Our system) 14.69 25.34 System-6 5.56 20.13 RIBES Challenge Test Set Evaluation Test Set System-1 0.734435 0.770754 System-2 0.726998 - System-3 - 0.749535 System-4 (Our system) 0.550964 0.744158 System-5 (Our system) 0.550568 0.636152 System-6 0.373560 0.574366 AMFM Challenge Test Set Evaluation Test Set System-1 0.775890 0.784950 System-2 0.773260 - System-3 (Our system) 0.632910 0.763940 System-4 - 0.762180 System-5 (Our system) 0.578930 0.656370 System-6 0.461110 0.615290 Table 3: BLEU, RIBES and AMFM scores result of participated teams for text-only translation track. System Challenge Test Set RIBES AMFM System-1 0.080028 0.385960 System-2 (Our system) 0.034482 0.335390 Table 4: RIBES, AMFM scores result of participated teams for Hindi-only image captioning track. 5 Conclusion and Future Work Current work participates in three different trans- lation tracks at WAT2019 namely, multi-modal, text-only and Hindi-only image captioning for 65 4 Table 6: Worst performance example in English to Hindi multi-modal translation. Acknowledgement Authors would like to thank WAT2019 Transla- tion task organizers for organizing this competi- tion and also, thank Centre for Natural Language Processing (CNLP) and Department of Computer Science and Engineering at National Institute of Technology, Silchar for providing the requisite support and infrastructure to execute this work. References Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Ben- gio. 2015. Neural machine translation by jointly learning to align and translate. In 3rd Inter- national Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings. Rafael E. Banchs, Luis F. D’Haro, and Haizhou Li. 2015. Adequacy-ﬂuency metrics: Evaluating mt in the continuous space model framework. IEEE/ACM English to Hindi translation. In the current competition, our multi-modal NMT system ob- tained higher BLEU scores than other participants in case of challenge as well as evaluation test data. The multi-modal NMT system is based on a doubly-attentive decoder to predict sentences, which shows better performance than text-only as well as Hindi-only image captioning. The com- bination of textual as well as visual features rea- sons about multi-modal translation outperforms text-only translation as well as Hindi-only image captioning tasks. However, close analysis of pre- dicted sentences on the given test data remarks that more experiment and analysis are needed in future work to enhance the performance of multi- modal NMT system. Table 6: Worst performance example in English to Hindi multi-modal translation. Table 5: Best performance examples in English to Hindi multi modal translation Table 5: Best performance examples in English to Hindi multi-modal translation. Table 6: Worst performance example in English to Hindi multi-modal translation. Acknowledgement Authors would like to thank WAT2019 Transla- tion task organizers for organizing this competi- tion and also, thank Centre for Natural Language Processing (CNLP) and Department of Computer Science and Engineering at National Institute of Technology, Silchar for providing the requisite support and infrastructure to execute this work. Rafael E. Banchs, Luis F. D’Haro, and Haizhou Li. 2015. Adequacy-ﬂuency metrics: Evaluating mt in the continuous space model framework. IEEE/ACM References Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Ben- gio. 2015. Neural machine translation by jointly learning to align and translate. In 3rd Inter- national Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings. Table 5: Best performance examples in English to Hindi multi-modal translation. 66 5 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 Trans. Audio, Speech and Lang. Proc., 23(3):472– 482. Toshiaki Nakazawa, Chenchen Ding, Raj Dabre, Hideya Mino, Isao Goto, Win Pa Pa, Nobushige Doi, Yusuke Oda, Anoop Kunchukuttan, Shan- tipriya Parida, Ondej Bojar, and Sadao Kurohashi. 2019. Overview of the 6th workshop on Asian translation. In Proceedings of the 6th Workshop on Asian Translation, Hong Kong. Association for Computational Linguistics. Iacer Calixto and Qun Liu. 2017. Incorporating global visual features into attention-based neural machine translation. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Pro- cessing, pages 992–1003, Copenhagen, Denmark. Association for Computational Linguistics. Kishore Papineni, Salim Roukos, Todd Ward, and Wei- Jing Zhu. 2002. Bleu: A method for automatic eval- uation of machine translation. In Proceedings of the 40th Annual Meeting on Association for Com- putational Linguistics, ACL ’02, pages 311–318, Stroudsburg, PA, USA. Association for Computa- tional Linguistics. Iacer Calixto, Qun Liu, and Nick Campbell. 2017. Doubly-attentive decoder for multi-modal neural machine translation. In Proceedings of the 55th An- nual Meeting of the Association for Computational Linguistics, ACL 2017, Vancouver, Canada, July 30 - August 4, Volume 1: Long Papers, pages 1913– 1924. Amarnath Pathak and Partha Pakray. 2018. Neural ma- chine translation for indian languages. Journal of Intelligent Systems, pages 1–13. Sandeep Kumar Dash, Saurav Saha, Partha Pakray, and Alexander Gelbukh. 2019. Generating image cap- tions through multimodal embedding. Journal of Intelligent and Fuzzy Systems, 36(5):4787–4796. Amarnath Pathak, Partha Pakray, and Jereemi Ben- tham. 2018. English–mizo machine translation us- ing neural and statistical approaches. References Neural Com- puting and Applications, 30:1–17. Koel Dutta Chowdhury, Mohammed Hasanuzzaman, and Qun Liu. 2018. Multimodal neural machine translation for low-resource language pairs using synthetic data. pages 33–42. Kashif Shah, Josiah Wang, and Lucia Specia. 2016. SHEF-multimodal: Grounding machine translation on images. In Proceedings of the First Conference on Machine Translation: Volume 2, Shared Task Pa- pers, pages 660–665, Berlin, Germany. Association for Computational Linguistics. Po-Yao Huang, Frederick Liu, Sz-Rung Shiang, Jean Oh, and Chris Dyer. 2016. Attention-based mul- timodal neural machine translation. In Proceed- ings of the First Conference on Machine Transla- tion: Volume 2, Shared Task Papers, pages 639– 645, Berlin, Germany. Association for Computa- tional Linguistics. Karen Simonyan and Andrew Zisserman. 2015. Very deep convolutional networks for large-scale image recognition. In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Pro- ceedings. Hideki Isozaki, Tsutomu Hirao, Kevin Duh, Katsuhito Sudoh, and Hajime Tsukada. 2010. Automatic eval- uation of translation quality for distant language pairs. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Process- ing, pages 944–952, Cambridge, MA. Association for Computational Linguistics. Ilya Sutskever, Oriol Vinyals, and Quoc V. Le. 2014. Sequence to sequence learning with neural net- works. In Proceedings of the 27th International Conference on Neural Information Processing Sys- tems - Volume 2, NIPS’14, pages 3104–3112, Cam- bridge, MA, USA. MIT Press. Guillaume Klein, Yoon Kim, Yuntian Deng, Jean Senellart, and Alexander Rush. 2017. Opennmt: Open-source toolkit for neural machine translation. In Proceedings of ACL 2017, System Demonstra- tions, pages 67–72, Vancouver, Canada. Association for Computational Linguistics. Marc Tanti, Albert Gatt, and Kenneth P. Camilleri. 2018. Where to put the image in an image cap- tion generator. Natural Language Engineering, 24(3):467–489. Sahinur Rahman Laskar, Partha Pakray, and Sivaji Bandyopadhyay. 2019. Neural machine translation: Hindi-Nepali. In Proceedings of the Fourth Con- ference on Machine Translation (Volume 3: Shared Task Papers, Day 2), pages 202–207, Florence, Italy. Association for Computational Linguistics. Thang Luong, Hieu Pham, and Christopher D. Man- ning. 2015. Effective approaches to attention-based neural machine translation. In Proceedings of the 2015 Conference on Empirical Methods in Natu- ral Language Processing, pages 1412–1421, Lis- bon, Portugal. Association for Computational Lin- guistics. 67 6
https://openalex.org/W3166358520	https://aclanthology.org/2021.bionlp-1.8.pdf	English	null	Overview of the MEDIQA 2021 Shared Task on Summarization in the Medical Domain	null	2,021	cc-by	8,194	1www-nlpir.nist.gov/projects/duc Abstract challenges from 2000 to 2007 and the Text Anal- ysis Conference2 (TAC) ran four shared tasks (2008-2011) on news summarization. The last TAC 2014 summarization task tackled biomedi- cal article summarization with referring sentences from external citations. Recent efforts in sum- marization have focused on neural methods (See et al., 2017; Gehrmann et al., 2018) using bench- mark datasets compiled from news articles, such as the CNN-DailyMail dataset (CNN-DM) (Her- mann et al., 2015). However, despite its impor- tance, fewer efforts have tackled text summariza- tion in the biomedical domain for both consumer and clinical text and its applications in Question Answering (QA) (Afantenos et al., 2005; Mishra et al., 2014; Afzal et al., 2020). The MEDIQA 2021 shared tasks at the BioNLP 2021 workshop addressed three tasks on summarization for medical text: (i) a question summarization task aimed at explor- ing new approaches to understanding com- plex real-world consumer health queries, (ii) a multi-answer summarization task that targeted aggregation of multiple relevant answers to a biomedical question into one concise and rel- evant answer, and (iii) a radiology report sum- marization task addressing the development of clinically relevant impressions from radiol- ogy report ﬁndings. Thirty-ﬁve teams partici- pated in these shared tasks with sixteen work- ing notes submitted (ﬁfteen accepted) describ- ing a wide variety of models developed and tested on the shared and external datasets. In this paper, we describe the tasks, the datasets, the models and techniques developed by vari- ous teams, the results of the evaluation, and a study of correlations among various summa- rization evaluation measures. We hope that these shared tasks will bring new research and insights in biomedical text summarization and evaluation. While the 2019 BioNLP-MEDIQA3 edition fo- cused on question entailment and textual infer- ence and their applications in medical Question Answering (Ben Abacha et al., 2019), MEDIQA 20214 addresses the gap in medical text summa- rization by promoting research on summarization for consumer health QA and clinical text. Three shared tasks are proposed for the summarization of (i) consumer health questions, (ii) multiple an- swers extracted from reliable medical sources to create one answer for each question, and (iii) tex- tual clinical ﬁndings in radiology reports to gener- ate radiology impression statements. Asma Ben Abacha NLM/NIH benabachaa@nih.gov Yassine Mrabet NLM/NIH mrabety@mail.nih.gov Yuhao Zhang Stanford University zyh@stanford.edu Asma Ben Abacha NLM/NIH benabachaa@nih.gov Yassine Mrabet NLM/NIH mrabety@mail.nih.gov Yuhao Zhang Stanford University zyh@stanford.edu Asma Ben Abacha NLM/NIH benabachaa@nih.gov Chaitanya Shivade Amazon shivadc@amazon.com Curtis Langlotz Stanford University langlotz@stanford.edu Dina Demner-Fushman NLM/NIH ddemner@mail.nih.gov Chaitanya Shivade Amazon shivadc@amazon.com Curtis Langlotz Stanford University langlotz@stanford.edu Dina Demner-Fushman NLM/NIH ddemner@mail.nih.gov Proceedings of the BioNLP 2021 workshop, pages 74–85 June 11, 2021. ©2021 Association for Computational Linguistics 4 2tac.nist.gov/tracks 3sites.google.com/view/mediqa2019 4sites.google.com/view/mediqa2021 5chiqa.nlm.nih.gov 1 Introduction Text summarization aims to create natural lan- guage summaries that represent the most impor- tant information in a given text. Extractive sum- marization approaches tackle the task by selecting content from the original text without any modiﬁ- cation (Nallapati et al., 2017; Xiao and Carenini, 2019; Zhong et al., 2020), while abstractive ap- proaches extend the summaries’ vocabulary to out-of-text words (Rush et al., 2015; Gehrmann et al., 2018; Chen and Bansal, 2018). For the ﬁrst two tasks, we created new test sets for the ofﬁcial evaluation using consumer health questions received by the U.S. National Library of Medicine (NLM) and answers retrieved from re- liable sources using the Consumer Health Ques- tion Answering system CHiQA5. For the third task, we created a new test set by combining public radiology reports in the Indiana Univer- Several past challenges and shared tasks have focused on summarization. The Document Un- derstanding Conference1 (DUC) organized seven 74 multiple relevant answers to a medical question (Savery et al., 2020). sity dataset (Demner-Fushman et al., 2016) and newly released chest x-ray reports from the Stan- ford Health Care. 2.1 Consumer Health Question Summarization (QS) The MeQSum dataset of consumer health ques- tions and their summaries (Ben Abacha and Demner-Fushman, 2019b) was suggested as a training dataset. It consists of 1,000 consumer health questions and their associated summaries. Participants were encouraged to use available ex- ternal resources including, but not limited to, med- ical QA datasets and question focus and type recognition datasets. For instance, the Consumer Health Questions dataset (Kilicoglu et al., 2018) contains annotations of medical entities, focus, and type of the MeQSum questions and additional NLM questions6. Consumer health questions tend to contain pe- ripheral information that hinders automatic Ques- tion Answering (QA). Empirical studies based on manual expert summarization of these questions showed a substantial improvement of 58% in QA performance (Ben Abacha and Demner-Fushman, 2019a). Effective automatic summarization meth- ods for consumer health questions could therefore play a key role in enhancing medical question an- swering. The goal of this task is to promote the de- velopment of new summarization approaches that address speciﬁcally the challenges of long and po- tentially complex consumer health questions. Rel- evant approaches should be able to generate a con- densed question expressing the minimum informa- tion required to ﬁnd correct answers to the origi- nal question (Ben Abacha and Demner-Fushman, 2019b). The new QS validation and test sets7 cover a wide range of topics and question types such as Treatment, Information, Side effects, Cause, Ef- fect, Person-Organization, Diet-Lifestyle, Compli- cations, Contraindications, Diagnosis, Usage, In- teraction, Ingredients, Prognosis, Susceptibility, Transmission, and Toxicity. They consist of man- ually de-identiﬁed consumer health questions re- ceived by the U.S. National Library of Medicine and gold summaries created by medical experts. The validation set includes 50 NLM questions and 2.3 Radiology Report Summarization (RRS) Through these tasks, we focus on studying: The task of radiology report summarization aims to promote the development of clinical summa- rization models that are able to generate the con- cise impression section (i.e., summary) of a radi- ology report conditioned on the free-text ﬁndings and background sections (Zhang et al., 2018). The resulting systems have signiﬁcant potential to im- prove the efﬁciency of clinical communications and accelerate the radiology workﬂow. While state-of-the-art techniques in language generation have enabled the generation of ﬂuent summaries, these models occasionally generate spurious facts limiting the clinical validity of the generated sum- maries (Zhang et al., 2020b). It is therefore impor- tant to develop systems that are able to summarize the radiology ﬁndings in a consistent manner. • The best approaches according to the summarization task objective and the language/vocabulary (consumers’ ques- tions, patient-oriented medical text, and professional clinical reports); • The impact of medical data scarcity on the development and performance of summa- rization methods in comparison with open- domain summarization; • The effects of different summary evaluation measures including lexical metrics such as ROUGE (Lin, 2004), embedding-based met- rics such as BERTScore (Zhang et al., 2019), and hybrid ensemble-oriented metrics such as HOLMS (Mrabet and Demner-Fushman, 2020). 6https://bionlp.nlm.nih.gov/ CHIQAcollections/CHQA-Corpus-1.0.zip 7https://github.com/abachaa/ MEDIQA2021/tree/main/Task1 NLM Question: Question summary used in answer retrieval: Question summary used in answer retrieval: What tests are used to diagnose dementia? Question summary used in answer retrieval: What tests are used to diagnose dementia? covid-19 how long to quarantine after being positive how long are you contagious if i tested positive for covid-19. how long before i can safely return to work after a positive covid 19 test CHiQA’s Answer #1: Dementia is not a speciﬁc dis- ease. It is a descriptive term for a collection of symp- toms that can be caused by a number of disorders that affect the brain. People with dementia have signiﬁ- cantly impaired intellectual functioning that interferes with normal activities and relationships. They also lose their ability to solve problems and maintain emotional control, and they may experience personality changes and behavioral problems, such as agitation, delusions, and hallucinations (...). CHiQA’s Answer #1: Dementia is not a speciﬁc dis- ease. It is a descriptive term for a collection of symp- toms that can be caused by a number of disorders that affect the brain. People with dementia have signiﬁ- cantly impaired intellectual functioning that interferes with normal activities and relationships. They also lose their ability to solve problems and maintain emotional control, and they may experience personality changes and behavioral problems, such as agitation, delusions, and hallucinations (...). Question Summary: How long will I remain conta- gious after testing positive for COVID-19? Question Summary: How long will I remain conta- gious after testing positive for COVID-19? Table 1: Test set examples for the QS task. Table 1: Test set examples for the QS task. their summaries with additional annotations of the question focus and type. The test set contains 80 consumer health questions. Table 1 presents two examples from the QS test set. CHiQA’s Answer #2: To diagnose dementia, doctors ﬁrst assess whether a person has an underlying treat- able condition such as depression, abnormal thyroid function, normal pressure hydrocephalus, or vitamin B12 deﬁciency. Early diagnosis is important, as some causes for symptoms can be treated. In many cases, the speciﬁc type of dementia a person has may not be con- ﬁrmed until after the person has died and the brain is examined.A medical assessment for dementia generally includes:- Patient history (...) - Physical exam (...) - Neurological tests (...). 8https://github.com/abachaa/ MEDIQA2021/tree/main/Task2 Question Summary: How can I get rid of pain caused by foraminal stenosis and nerve compression? Question Summary: How can I get rid of pain caused by foraminal stenosis and nerve compression? Question Summary: How can I get rid of pain caused by foraminal stenosis and nerve compression? 3.2 MAS Datasets The MEDIQA-AnS dataset (Savery et al., 2020) was suggested as a training set for the MAS task. Participants were allowed to use available exter- nal resources (e.g. existing medical QA datasets) as well as data creation, selection, and augmenta- tion methods. To create the MAS validation and test sets8, we used 130 consumer health questions received by NLM. In order to retrieve more ac- curate answers, we created question summaries that we used to query the medical QA system CHiQA that searches for answers from only trust- worthy medical information sources (Ben Abacha and Demner-Fushman, 2019c; Demner-Fushman et al., 2020). Reference Extractive Summary: Dementia is not a speciﬁc disease. It is a descriptive term for a collection of symptoms that can be caused by a number of disor- ders that affect the brain. Doctors diagnose dementia only if two or more brain functions - such as memory and language skills – are signiﬁcantly impaired without loss of consciousness. To diagnose dementia, doctors ﬁrst assess whether a person has an underlying treat- able condition such as depression, abnormal thyroid function, normal pressure hydrocephalus, or vitamin B12 deﬁciency. Early diagnosis is important, as some causes for symptoms can be treated. In many cases, the speciﬁc type of dementia a person has may not be con- ﬁrmed until after the person has died and the brain is examined.A medical assessment for dementia generally includes:- Patient history (...) - Physical exam (...) - Neurological tests (...). The answer summaries were manually created by medical experts. We provided both extractive and abstractive gold summaries, and encouraged the use of all types of summarization approaches (extractive, abstractive, and hybrid). The MAS validation set contains 192 answers to 50 medi- cal questions. The test set contains 303 answers to 80 medical questions. Each question has at least two answers, one extractive multi-answer sum- mary, and one abstractive multi-answer summary. Table 2 presents an example from the test set. Reference Abstractive Summary: Dementia could be caused by many different diseases of the brain. it is di- agnosed if at least two brain functions are effected, for example, if people experience memory loss and changes in behavior and personality. Diagnostic tests for de- mentia include family history, physical examination, and neurological tests to asses balance, sensory func- tions, reﬂexes, vision, eye movements, and cognitive functions. In many cases, the type of dementia is con- ﬁrmed after the person dies. Table 2: Test set example for the MAS task (QID:105). 2.2 Multi-Answer Summarization (MAS) Different answers can bring complementary per- spectives that are likely to beneﬁt the users of QA systems. The goal of this task is to promote the development of multi-answer summarization ap- proaches that could solve simultaneously the ag- gregation and summarization problems posed by 75 Example 1 (QID: 139) Example 1 (QID: 139) Original NLM question: I have dementia like symp- toms and wanted to know where is the best source to be tested for diagnosis? I have been prescribed An- ticholinergic medicine since 2008...since I have been diagnosed with, Celiac disease and Obstructive Sleep Apnea. I think I have Frontal Temporal lobe atrophy. I’m going to try to get tested...any references on which process is easiest would be much appreciated. I can’t take my Nasalcrom allergy spay any more nor, valium or prozac, benadryl and glutamate additives in meats because it sends me straight into cognitive emergency state and irrational thinking NLM Question: did anyone have this and does it re- quire surgery? my mri says forminal stenosis from bone spurs c4,5,6. my nerve test shows severe nerve com- pression c7,8. i’m in so much pain, mostly my arm and shoulder and leg. waiting to see the pain specialist to see what’s next. would love to know what you guys think is required. 3.2 MAS Datasets Table 2: Test set example for the MAS task (QID:105). Table 2: Test set example for the MAS task (QID:105). 76 3.3 RRS Datasets ized similarity and a lexical ROUGE component through a multi-dimensional Gaussian function (Mrabet and Demner-Fushman, 2020). HOLMS was evaluated on multiple DUC and TAC datasets, and three correlation factors (Pearson’s, Spear- man’s, and Kendall’s), and was shown to ben- eﬁt from the complementary strengths of lexi- cal and language model-based similarity measure- ments for evaluating summarization systems. We focus on the summarization of chest radiogra- phy reports for the RRS task, since chest radiog- raphy represents the most common study type in radiology, and public resources for chest studies are easily accessible. For training, we sampled a collection of 91,544 reports from the MIMIC- CXR chest X-ray report dataset9 based on simple criteria such as the acceptable length of each sec- tion. For validation, we combined another 2,000 reports from the MIMIC-CXR dataset and 2,000 reports from the Indiana University chest X-ray dataset10(Demner-Fushman et al., 2016). We sam- pled the reports such that there is no overlapping patients in the validation and training sets. In this shared task, we chose ROUGE-2 as our ofﬁcial ranking metric following its superiority observed by Owczarzak et al. (2012) on multi- ple TAC summarization datasets, and by Bhandari et al. (2020c) on the CNN-DM dataset. We chose two additional metrics for the three tasks: (1) BERTScore for its wider adoption as a language model-based text generation metric, and (2) HOLMS for its hybrid and ensemble-oriented approach. For the RRS task we also considered an additional evaluation metric based on the ham- ming similarity on the labels produced by the CheXbert labeler (Smit et al., 2020) when applied to both the system and reference summaries, sim- ilar to the approach by Zhang et al. (2020b). For the ofﬁcial test set, we used a combination of 300 reports from the Indiana dataset and 300 newly released chest X-ray reports drawn from the Stanford Health Care system. We intentionally de- signed the test set to be partially from a hospi- tal system different from the training set (out-of- domain) to test the generalizability of the partici- pating systems. 9https://physionet.org/content/ mimic-cxr/2.0.0/ 10openi.nlm.nih.gov/faq#collection 4.1 Evaluation Measures Our baseline system for the QS task relied on a distilled PEGASUS model (Zhang et al., 2020a) trained on the CNN-DM dataset and ﬁne-tuned on a combination of biomedical answer-to-question data and question summarization data from MeQ- Sum, LiveQA-Med data (Ben Abacha et al., 2017), a collection of clinical questions (Ely et al., 2000), and Quora question pairs dataset (Iyer et al., 2017). For the Quora and clinical questions datasets, we extracted only the question pairs with a minimum token reduction ratio of 33%. Several new metrics for evaluating text genera- tion systems were studied in recent years (Mao et al., 2020; Bhandari et al., 2020a,b; Zhang et al., 2019; Sellam et al., 2020), with a focus on eval- uating text generation based on deep and contex- tualized representations. To understand these met- rics in the context of summarization, Fabbri et al. (2020) have compared 34 traditional and recent model-based metrics on a manually annotated sub- set from the CNN-DM dataset. Although the study relied only on one correlation factor (Kendall’s Tau) and one dataset, it highlighted the (contin- ued) general relevance of ROUGE variants (Lin, 2004) and the challenge of designing or determin- ing the best measure to use. Speciﬁcally, the study found that a different measure obtained the best score in each of the four considered evaluation dimensions: coherence, consistency, ﬂuency, and relevance, with substantial discrepancies in rank- ings. Our extractive baseline for the MAS task relied on sentence clustering and selection. We used our ﬁne-tuned question summarization model to gen- erate a short question from each sentence, and then clustered the sentences using a word-based cosine distance between the generated questions and a distance threshold set to 0.7. Intersecting clusters were merged. For each cluster, we selected the sentence that was the best cumulative TF-IDF an- swer to all other sentences as a representative. In parallel, HOLMS was recently proposed as an ensemble measure combining both contextual- For the RRS task, we prepared three baselines: a base pointer-generator model without modeling the background section of a radiology report, a full pointer-generator model with background model- 77 ing (Zhang et al., 2018), and a zero-shot T5-base summarization model (Raffel et al., 2020). of PEGASUS, T5, and BART models according to hand-picked features based on the contents of the input question and lengths of the outputs. 5.2 Summarization Approaches & Results A vast majority of the approaches submitted to the QS and RRS tasks were abstractive and relied on ﬁne-tuning of pre-trained generative language models and encoders-decoders architectures. For the MAS task, most submitted approaches were extractive and used a wide spectrum of sentence selection techniques. 1. a relevant learning-based ensemble method that could rely either on the textual outputs or the logits of single models. 2. a more systemic way to select the most rel- evant datasets for both pretraining and ﬁne tuning. Question Summarization. Table 4 presents the ofﬁcial results of the teams with accepted working notes papers from the 22 teams that participated in the QS task. Multi-Answer Summarization. Both extractive and abstractive approaches were used by the 17 teams that submitted runs to MAS task (Zhu et al., 2021; Can et al., 2021; Xu et al., 2021; Mrini et al., 2021; Yadav et al., 2021; Le et al., 2021; Lee et al., 2021a). Table 5 and Table 6 present ofﬁcial results of the teams with extractive and abstractive sys- tems when evaluated, respectively, on extractive gold summaries and abstractive gold summaries. All approaches submitted to the question sum- marization task were abstractive methods relying on the ﬁne-tuning of pretrained transformer mod- els (Vaswani et al., 2017). A wide variety of ﬁne tuning, knowledge-based, and ensemble methods was investigated by the participating teams to achieve higher performance (Mrini et al., 2021; Xu et al., 2021; Zhu et al., 2021; S¨anger et al., 2021; Lee et al., 2021b; Balumuri et al., 2021; Yadav et al., 2021; He et al., 2021; Lee et al., 2021a). A ﬁrst interesting insight from the overview is that building ensemble models with deep neural networks such as discriminators is not a trivial task, and achieves results that stay on par with the best single model (S¨anger et al., 2021). In contrast, heuristic, downstream ensembles of the models outputs led to substantial improvements when compared to its components/single models (He et al., 2021). The best performing approach relied on such an ensemble by ranking the outputs The best MAS run (Zhu et al., 2021) relied on an ensemble method and a recent multi-document summarization approach (Xu and Lapata, 2020) using a Roberta model to rank locally the can- didate sentences and a Markov chain to evaluate them globally. 4.1 Evaluation Measures Spell checking was also a performance boost factor in the question summarization task with some teams using a knowledge base to correct misspelling er- rors in the original long questions (He et al., 2021), and others relying on third party tools such as CSpell (Yadav et al., 2021; Lu et al., 2019). The datasets used for transfer learning or ﬁne-tuning also played a major role in the achieved perfor- mance as demonstrated, for instance, by the com- bination of datasets from HealthCareMagic, ques- tion entailment recognition and question summa- rization in (Mrini et al., 2021). Moving forward, we think that the overview of the question sum- marization task revealed two key challenges that need to be addressed to enhance the relevance and performance of existing systems: 5.1 Participating Teams In total, 35 teams participated in the MEDIQA shared tasks and submitted 310 individual runs (with a limit of ten runs per team per task). Ta- ble 3 presents the participating teams with ac- cepted working notes papers. The results of all 35 teams are available on AIcrowd and on the MEDIQA 2021 website. 11www.aicrowd.com/challenges/ mediqa-2021 5 Ofﬁcial Results We published three AIcrowd projects (one for each task) to release the datasets and manage team reg- istration, submission, and leaderboard ranking11. 5.2 Summarization Approaches & Results A similar approach was also used by the ChicHealth team (Xu et al., 2021) with- out a downstream ensemble method. Participat- ing teams used transfer learning (e.g. (Mrini et al., 2021)) as well as answer sentence selection meth- ods. Sentence selection was used in building ex- tractive summaries (e.g. (Can et al., 2021)) and as an intermediate step in abstractive summarization to provide more concise inputs to generative mod- els (e.g. (Le et al., 2021)). Different models, such 78 Team Institution QS MAS RRS BDKG (Dai et al., 2021) Baidu, Inc ✓ ChicHealth (Xu et al., 2021) Chic Health ✓ ✓ damo nlp (He et al., 2021) Alibaba Group ✓ ✓ IBMResearch (Mahajan et al., 2021) IBM Research ✓ MNLP (Lee et al., 2021a) George Mason University ✓ ✓ NCUEE-NLP (Lee et al., 2021b) National Central University ✓ NLM (Yadav et al., 2021) U.S. National Library of Medicine ✓ ✓ optumize (Kondadadi et al., 2021) Optum ✓ paht nlp (Zhu et al., 2021) ECNU & Pingan Health Tech ✓ ✓ ✓ QIAI (Delbrouck et al., 2021) Stanford University ✓ ✓ SB NITK (Balumuri et al., 2021) National Institute of Technology Karnataka ✓ UCSD-Adobe (Mrini et al., 2021) UC San Diego & Adobe Research ✓ ✓ UETﬁshes (Le et al., 2021) VNU University of Engineering and Technology ✓ UETrice (Can et al., 2021) VNU University of Engineering and Technology ✓ WBI (S¨anger et al., 2021) Humboldt University of Berlin ✓ Table 3: Participating teams with accepted working notes papers at BioNLP-MEDIQA 2021 Table 3: Participating teams with accepted working notes papers at BioNLP-MEDIQA 2021 Table 3: Participating teams with accepted working notes papers at BioNLP-MEDIQA 2021 Rank Team ROUGE-2 ROUGE-1 ROUGE-L HOLMS BERTScore 1 damo nlp 0.1608 0.3514 0.3131 0.5677 0.6898 2 WBI 0.1599 0.3340 0.3149 0.5767 0.6996 3 NCUEE-NLP 0.1597 0.3352 0.3090 0.5787 0.6960 4 NLM 0.1514 0.3556 0.3110 0.5649 0.6892 5 UCSD-Adobe 0.1414 0.3463 0.3065 0.5586 0.6942 6 ChicHealth 0.1398 0.3403 0.2962 0.5551 0.6810 7 SB NITK 0.1393 0.3331 0.3077 0.5663 0.7025 – QS Baseline 0.1373 0.3203 0.2962 0.5672 0.6277 8 MNLP 0.1114 0.2840 0.2587 0.5455 0.6732 9 paht nlp 0.0935 0.2486 0.2331 0.5428 0.6591 10 QIAI 0.0385 0.1514 0.1356 0.4898 0.5101 Table 4: Ofﬁcial results of the MEDIQA-QS task. Team ROUGE-2 ROUGE-1 ROUGE-L HOLMS BERTScore Table 4: Ofﬁcial results of the MEDIQA-QS task. Table 6: Ofﬁcial results of the MEDIQA-MAS task (2): Abstractive Approaches. Ranks in bold and in parenthe- sis correspond to evaluation on extractive gold summaries and on abstractive gold summaries, respectively. 5.2 Summarization Approaches & Results Rank Team ROUGE-2 CheXbert Stanford Indiana Stanford Indiana 1 BDKG 0.2768 0.5955 0.6547 0.7052 2 ChicHealth 0.2690 0.3781 0.6291 0.5873 3 damo nlp 0.2687 0.2839 0.6645 0.5517 4 optumize 0.2654 0.5182 0.6474 0.6592 5 QIAI 0.2516 0.5039 0.5508 0.4970 6 paht nlp 0.2491 0.1483 0.6834 0.6148 – baseline (PG-full) 0.2414 0.3054 0.6216 0.5466 – baseline (PG-base) 0.2408 0.2870 0.5892 0.4754 7 IBMResearch 0.2283 0.5880 0.6472 0.6937 – baseline (T5) 0.1280 0.0610 0.5067 0.5609 Table 8: Ofﬁcial results of the MEDIQA-RRS task on the Stanford and Indiana test splits. Table 8: Ofﬁcial results of the MEDIQA-RRS task on the Stanford and Indiana test splits. as BART and T5, and datasets (e.g. MEDIQA- AnS, MSMARCO, MEDIQA-2019) have been used for single and multiple answer summariza- tion (Yadav et al., 2021; Mrini et al., 2021; Zhu et al., 2021; Can et al., 2021). Radiology Report Summarization. 14 teams participated in the RRS task. Table 7 presents the ofﬁcial results of the teams (with accepted papers) on the full test set, and Table 8 presents the results on the Stanford and Indiana subsets of the test set. Similar to the previous tasks, participating teams for the RRS task have extensively used pre- trained transformer models: out of the 7 teams that submitted papers describing their systems, 6 re- ported the use of pretrained language models such as BART or PEGASUS in their submissions (Xu et al., 2021; Zhu et al., 2021; Kondadadi et al., 2021; Dai et al., 2021; Mahajan et al., 2021; He et al., 2021). Among them, Xu et al. (2021); Zhu et al. (2021); Dai et al. (2021) reported that best results were achieved with pretrained PEGASUS models, while Kondadadi et al. (2021) reported better results from BART. Xu et al. (2021) and Zhu et al. (2021) reported that using PEGASUS models pretrained on the PubMed corpus yielded worse results than using the general PEGASUS models, potentially due to the domain difference of the RRS task with the PubMed text. as BART and T5, and datasets (e.g. MEDIQA- AnS, MSMARCO, MEDIQA-2019) have been used for single and multiple answer summariza- tion (Yadav et al., 2021; Mrini et al., 2021; Zhu et al., 2021; Can et al., 2021). In addition to the use of pretrained models, the highest-ranked systems from Dai et al. (2021) made effective use of a dedicated domain adapta- tion module, an ensemble module, and text nor- malization heuristics. 5.2 Summarization Approaches & Results Rank Team ROUGE-2 ROUGE-1 ROUGE-L HOLMS BERTScore 1 paht nlp 0.5076 0.5848 0.4354 0.7047 0.8038 2 UETrice 0.5040 0.6110 0.4412 0.7383 0.7958 3 ChicHealth 0.4893 0.5776 0.4261 0.7033 0.7916 4 UCSD-Adobe 0.4720 0.6073 0.4289 0.7612 0.7753 5 NLM 0.4677 0.5470 0.3276 0.6575 0.7645 Table 5: Ofﬁcial results of the MEDIQA-MAS task (1): Extractive Approaches. Team ROUGE-2 ROUGE-1 ROUGE-L HOLMS BERTScore Table 5: Ofﬁcial results of the MEDIQA-MAS task (1): Extractive Approaches. Team Rank ROUGE-2 ROUGE-1 ROUGE-L HOLMS BERTScore paht nlp 1 0.5076 0.5848 0.4354 0.7047 0.8038 (1) 0.1621 0.3215 0.1910 0.4220 0.6528 UETﬁshes 2 0.4698 0.5720 0.4001 0.6970 0.7821 (3) 0.1495 0.3124 0.1885 0.4213 0.6466 UCSD-Adobe 3 0.4595 0.5921 0.4170 0.7502 0.7689 (2) 0.1604 0.3843 0.2117 0.4937 0.6326 MNLP 4 0.2594 0.4220 0.2954 0.6568 0.6479 (4) 0.1167 0.3490 0.2047 0.5269 0.5763 Table 6: Ofﬁcial results of the MEDIQA-MAS task (2): Abstractive Approaches. Ranks in bold and in parenthe is correspond to evaluation on extractive gold summaries and on abstractive gold summaries, respectively. Table 6: Ofﬁcial results of the MEDIQA-MAS task (2): Abstractive Approaches. Ranks in bold and in parenthe- sis correspond to evaluation on extractive gold summaries and on abstractive gold summaries, respectively. 79 Rank Team R-2 R-1 R-L HOLMS BERTScore CheXbert 1 BDKG 0.4362 0.5572 0.5365 0.7402 0.7184 0.6927 2 IBMResearch 0.4082 0.5328 0.5134 0.7185 0.7115 0.6774 3 optumize 0.3918 0.5185 0.4957 0.7087 0.6975 0.6773 4 QIAI 0.3778 0.4954 0.4793 0.7132 0.5328 0.5565 5 ChicHealth 0.3236 0.4606 0.4410 0.6822 0.6768 0.6261 6 damo nlp 0.2763 0.4329 0.4115 0.6604 0.6576 0.6343 – baseline (PG-full) 0.2734 0.4182 0.4041 0.6647 0.6194 0.6014 – baseline (PG-base) 0.2639 0.4026 0.3885 0.6553 0.6103 0.5537 7 paht nlp 0.1987 0.3400 0.3053 0.5915 0.5985 0.6705 – baseline (T5) 0.0945 0.2108 0.1831 0.4432 0.4921 0.5245 Table 7: Ofﬁcial results of the MEDIQA-RRS task on the full test set. Table 7: Ofﬁcial results of the MEDIQA-RRS task on the full test set. 5.2 Summarization Approaches & Results Zhu et al. (2021) reported that freezing the embedding layer in the pre- trained models helps the model generalize at test time. Kondadadi et al. (2021) reported that adding the background section as input improves perfor- mance at validation time, but not test time, sug- gesting that the model performance is sensitive to the different text styles of the background sections from different splits. Mahajan et al. (2021) fo- cused their study on the factual consistency of gen- erated summaries, and proposed a specialized fact- aware re-ranking approach based on the predicted disease values from the ﬁndings section with a transformer model. As a result, their submissions Radiology Report Summarization. 14 teams participated in the RRS task. Table 7 presents the ofﬁcial results of the teams (with accepted papers) on the full test set, and Table 8 presents the results on the Stanford and Indiana subsets of the test set. Similar to the previous tasks, participating teams for the RRS task have extensively used pre- trained transformer models: out of the 7 teams that submitted papers describing their systems, 6 re- ported the use of pretrained language models such as BART or PEGASUS in their submissions (Xu et al., 2021; Zhu et al., 2021; Kondadadi et al., 2021; Dai et al., 2021; Mahajan et al., 2021; He et al., 2021). Among them, Xu et al. (2021); Zhu et al. (2021); Dai et al. (2021) reported that best results were achieved with pretrained PEGASUS models, while Kondadadi et al. (2021) reported better results from BART. Xu et al. (2021) and 80 from a Pearson coefﬁcient range between 0.663 and 0.958 to a range between 0.193 and 0.372. As all submitted QS runs were described as abstrac- tive or hybrid approaches, this discrepancy might be due to a stronger disagreement on summary as- sessment due to semantically-close but lexically distant summaries. It is also likely that the lex- ical distance between paraphrases was more pro- nounced due to the lengths of the question sum- maries, which are shorter than the summaries in the MAS task. achieved competitive rankings under the CheXbert metric. Lastly, Delbrouck et al. (2021) studied the use of image features for the RSS task: they re- trieved and linked images for each study to the re- port at training and validation time, and combined a visual encoder with a text encoder for the sum- marization task. 5.2 Summarization Approaches & Results They found that at validation time the multi-modal setting is beneﬁcial to the summa- rization of MIMIC reports, but not to the Indiana reports, potentially due to the distribution shift in the images. 6 Correlations among the Evaluation Measures We presented an overview of the MEDIQA 2021 shared tasks on summarization in the medical do- main. We presented the results for the three tasks on Question Summarization, Multi-Answer Summarization and Radiology Reports Summa- rization, and discussed the impact of summariza- tion approaches and automatic evaluation meth- ods. We ﬁnd that pre-trained transformer mod- els, ﬁne-tuning on the carefully selected domain- speciﬁc text and ensemble methods worked well for all three summarization tasks. The results en- courage future research to include in-depth ex- ploration of ensemble methods, systematic ap- proaches to selection of datasets for pre-training and ﬁne-tuning, as well as a thorough assessment of the quality and relevance of different evaluation measures for summarization. We hope that the MEDIQA 2021 shared tasks will encourage fur- ther research efforts in medical text summarization and evaluation. In this section, we discuss correlations between the different evaluation metrics that we used in the challenge. Table 9 shows Pearson correla- tions between the F1 scores of the three lexical measures (ROUGE-1, ROUGE-2, and ROUGE-L) and the two language model-based and ensemble- based measures (i.e., HOLMS and BERTScore). Over all three tasks the HOLMS metric had a better Pearson correlation with ROUGE, ranging from 0.734 to 0.755, while also maintaining a high correlation of 0.736 with BERTScore. This obser- vation supports the ﬁndings from the experiments in (Mrabet and Demner-Fushman, 2020), which suggested that lexical measures such as ROUGE and language model-based measures bring differ- ent and complementary perspectives to summary- evaluation. Table 10 shows Pearson correlations for the RRS task. HOLMS is substantially closer than CheXbert and BERTScore in its correlation with ROUGE for the RRS task, while maintaining high correlation of respectively 0.645 and 0.702 with CheXbert and BERTScore. Acknowledgments This research was partially supported by the In- tramural Research Program of the National Li- brary of Medicine, National Institutes of Health. We thank Anna Ripple (NLM/NIH) for her help with the manual annotation and Soumya Gayen (NLM/NIH) for his help with the summarization interface. In contrast, BERTScore is substantially closer than HOLMS in its correlation with the ROUGE metrics for both the MAS task (cf. table 11) and the QS task (see Table 12). Two factors that could explain these correlations are (i) the predominance of extractive runs in the MAS task and (ii) the se- quential n-gram-based modeling in HOLMS that takes into account the order of the n-grams, while BERTScore relies on a cosine distance between two given sets of token embeddings. References Stergos D. Afantenos, Vangelis Karkaletsis, and Pana- giotis Stamatopoulos. 2005. Summarization from medical documents: A survey. Artif Intell Med, 33(2):157–77. Both language model-based measures had pos- itive correlations with ROUGE for the QS task, but the level of correlation was substantially lower when compared to the MAS and RRS tasks, going Muhammad Afzal, Fakhare Alam, Khalid Mahmood Malik, and Ghaus M Malik. 2020. Clinical context– aware biomedical text summarization using deep 81 Measure ROUGE-1 ROUGE-2 ROUGE-L HOLMS BERTScore ROUGE-1 1.000 ROUGE-2 0.966 1.000 ROUGE-L 0.813 0.762 1.000 HOLMS 0.734 0.722 0.755 1.000 BERTScore 0.546 0.519 0.409 0.736 1.000 Table 9: Pearson correlations between metrics aggregated over all three tasks. For ROUGE and BERTScore we use their F1 scores. Best correlations with the ROUGE metrics are highlighted in bold. Measure ROUGE-1 ROUGE-2 ROUGE-L HOLMS BERTScore Table 9: Pearson correlations between metrics aggregated over all three tasks. For ROUGE and BERTScore we use their F1 scores. Best correlations with the ROUGE metrics are highlighted in bold. Measure ROUGE-1 ROUGE-2 ROUGE-L CheXbert HOLMS BERTScore ROUGE-1 1.000 ROUGE-2 0.970 1.000 ROUGE-L 0.998 0.975 1.000 CheXbert 0.777 0.667 0.749 1.000 HOLMS 0.951 0.938 0.958 0.645 1.000 BERTScore 0.752 0.663 0.743 0.719 0.702 1.000 Table 10: Pearson correlations between metrics for the RRS task. For ROUGE and BERTScore we used the F1 scores. Best correlations with the lexical measures are highlighted in bold. Table 10: Pearson correlations between metrics for the RRS task. For ROUGE and BERTScore we used the F1 scores. Best correlations with the lexical measures are highlighted in bold. Measure ROUGE-1 ROUGE-2 ROUGE-L HOLMS BERTScore ROUGE-1 1.000 ROUGE-2 0.960 1.000 ROUGE-L 0.951 0.946 1.000 HOLMS 0.812 0.823 0.873 1.000 BERTSCore 0.913 0.924 0.889 0.784 1.000 Table 11: Pearson correlations between metrics for the MAS task. Extractive runs were evaluated on extractive gold summaries. Abstractive runs were evaluated on both extractive and abstractive gold summaries. All evaluation scores were concatenated to compute correlations. For ROUGE and BERTScore we used the F1 scores. Best correlations with the lexical measures are highlighted in bold. Measure ROUGE-1 ROUGE-2 ROUGE-L HOLMS BERTScore ROUGE-1 1.000 ROUGE-2 0.960 1.000 ROUGE-L 0.951 0.946 1.000 HOLMS 0.812 0.823 0.873 1.000 BERTSCore 0.913 0.924 0.889 0.784 1.000 Table 11: Pearson correlations between metrics for the MAS task. Extractive runs were evaluated on extractive gold summaries. Abstractive runs were evaluated on both extractive and abstractive gold summaries. References All evaluation scores were concatenated to compute correlations. For ROUGE and BERTScore we used the F1 scores. Best correlations with the lexical measures are highlighted in bold. Table 11: Pearson correlations between metrics for the MAS task. Extractive runs were evaluated on extractive gold summaries. Abstractive runs were evaluated on both extractive and abstractive gold summaries. All evaluation scores were concatenated to compute correlations. For ROUGE and BERTScore we used the F1 scores. Best correlations with the lexical measures are highlighted in bold. Measure ROUGE-1 ROUGE-2 ROUGE-L HOLMS BERTScore ROUGE-1 1.000 ROUGE-2 0.951 1.000 ROUGE-L 0.944 0.981 1.000 HOLMS 0.193 0.204 0.259 1.000 BERTSCore 0.292 0.332 0.372 0.972 1.000 Table 12: Pearson correlations between metrics for the QS task. For ROUGE and BERTScore we used the F1 scores. Best correlations with the lexical measures are highlighted in bold. Table 12: Pearson correlations between metrics for the QS task. For ROUGE and BERTScore we used the F1 scores. Best correlations with the lexical measures are highlighted in bold. 82 neural network: Model development and validation. J Med Internet Res, 22(10):e19810. Duy-Cat Can, Quoc-An Nguyen, Quoc-Hung Duong, Minh-Quang Nguyen, Huy-Son Nguyen, Cam- Van Thi Nguyen, Quang-Thuy Ha, and Mai-Vu Tran. 2021. Uetrice at mediqa 2021: A prosper-thy- neighbour extractive multi-document summariza- tion model. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Language Processing, NAACL-BioNLP 2021. Association for Computa- tional Linguistics. Spandana Balumuri, Sony Bachina, and Sowmya Ka- math S. 2021. Sb nitk at mediqa 2021: Lever- aging transfer learning for question summariza- tion in medical domain. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Yen-Chun Chen and Mohit Bansal. 2018. Fast abstrac- tive summarization with reinforce-selected sentence rewriting. In Proceedings of the 56th Annual Meet- ing of the Association for Computational Linguis- tics, ACL 2018, Melbourne, Australia, July 15-20, 2018, Volume 1: Long Papers, pages 675–686. As- sociation for Computational Linguistics. Asma Ben Abacha, Eugene Agichtein, Yuval Pinter, and Dina Demner-Fushman. 2017. Overview of the medical question answering task at trec 2017 liveqa. In TREC 2017. Asma Ben Abacha and Dina Demner-Fushman. 2019a. On the role of question summarization and informa- tion source restriction in consumer health question answering. In Proceedings of the AMIA 2019 Infor- matics Summit, San Francisco, CA, USA, 2019. Songtai Dai, Quan Wang, Yajuan Lyu, and Yong Zhu. 2021. References Bdkg at mediqa 2021: System report for the radiology report summarization task. In Proceed- ings of the 20th SIGBioMed Workshop on Biomed- ical Language Processing, NAACL-BioNLP 2021. Association for Computational Linguistics. Asma Ben Abacha and Dina Demner-Fushman. 2019b. On the summarization of consumer health questions. In Proceedings of the 57th Conference of the Asso- ciation for Computational Linguistics, ACL 2019, Florence, Italy, July 28- August 2, 2019, Volume 1: Long Papers, pages 2228–2234. Association for Computational Linguistics. Jean-Benoit Delbrouck, Cassie Zhang, and Daniel Ru- bin. 2021. Qiai at mediqa 2021: Multimodal ra- diology report summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Asma Ben Abacha and Dina Demner-Fushman. 2019c. A question-entailment approach to question answer- ing. BMC Bioinform., 20(1):511:1–511:23. Dina Demner-Fushman, Marc D Kohli, Marc B Rosen- man, Sonya E Shooshan, Laritza Rodriguez, Sameer Antani, George R Thoma, and Clement J McDon- ald. 2016. Preparing a collection of radiology ex- aminations for distribution and retrieval. Journal of the American Medical Informatics Association, 23(2):304–310. Asma Ben Abacha, Chaitanya Shivade, and Dina Demner-Fushman. 2019. Overview of the MEDIQA 2019 shared task on textual inference, question entailment and question answering. In Proceed- ings of the 18th BioNLP Workshop and Shared Task, BioNLP@ACL 2019, Florence, Italy, August 1, 2019, pages 370–379. Association for Computa- tional Linguistics. Dina Demner-Fushman, Yassine Mrabet, and Asma Ben Abacha. 2020. Consumer health information and question answering: helping consumers ﬁnd an- swers to their health-related information needs. J. Am. Medical Informatics Assoc., 27(2):194–201. Manik Bhandari, Pranav Narayan Gour, Atabak Ash- faq, and Pengfei Liu. 2020a. Metrics also disagree in the low scoring range: Revisiting summarization evaluation metrics. In Proceedings of the 28th In- ternational Conference on Computational Linguis- tics, COLING 2020, Barcelona, Spain (Online), De- cember 8-13, 2020, pages 5702–5711. International Committee on Computational Linguistics. John W. Ely, Jerome A. Osheroff, Paul N. Gorman, Mark H. Ebell, M. Lee Chambliss, Eric A. Pifer, and P. Zoe Stavri. 2000. A taxonomy of generic clini- cal questions: classiﬁcation study. British Medical Journal, 321:429–432. Manik Bhandari, Pranav Narayan Gour, Atabak Ash- faq, Pengfei Liu, and Graham Neubig. 2020b. Re- evaluating evaluation in text summarization. In Pro- ceedings of the 2020 Conference on Empirical Meth- ods in Natural Language Processing, EMNLP 2020, Online, November 16-20, 2020, pages 9347–9359. Association for Computational Linguistics. References Alexander R Fabbri, Wojciech Kry´sci´nski, Bryan McCann, Caiming Xiong, Richard Socher, and Dragomir Radev. 2020. Summeval: Re- evaluating summarization evaluation. arXiv preprint arXiv:2007.12626. Sebastian Gehrmann, Yuntian Deng, and Alexander M. Rush. 2018. Bottom-up abstractive summarization. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, Brussels, Belgium, October 31 - November 4, 2018, pages 4098–4109. Manik Bhandari, Pranav Narayan Gour, Atabak Ash- faq, Pengfei Liu, and Graham Neubig. 2020c. Re- evaluating evaluation in text summarization. In Pro- ceedings of the 2020 Conference on Empirical Meth- ods in Natural Language Processing (EMNLP). 83 consumer language (cspell). Journal of the Amer- ican Medical Informatics Association, 26(3):211– 218. Yifan He, Mosha Chen, and Songfang Huang. 2021. damo nlp at mediqa 2021: Knowledge-base prepro- cessing and coverage-oriented reranking for med- ical question summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Diwakar Mahajan, Ching-Huei Tsou, and Jennifer J Liang. 2021. Ibmresearch at mediqa 2021: To- ward improving factual correctness of radiology re- port abstractive summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Karl Moritz Hermann, Tom´as Kocisk´y, Edward Grefenstette, Lasse Espeholt, Will Kay, Mustafa Su- leyman, and Phil Blunsom. 2015. Teaching ma- chines to read and comprehend. In NIPS, pages 1693–1701. Yuning Mao, Liyuan Liu, Qi Zhu, Xiang Ren, and Jiawei Han. 2020. Facet-aware evaluation for ex- tractive summarization. In Proceedings of the 58th Annual Meeting of the Association for Computa- tional Linguistics, ACL 2020, Online, July 5-10, 2020, pages 4941–4957. Association for Computa- tional Linguistics. Shankar Iyer, Nikhil Dandekar, and Korn´el Csernai. 2017. First quora dataset release: Question pairs. Halil Kilicoglu, Asma Ben Abacha, Yassine Mrabet, Sonya E. Shooshan, Laritza Rodriguez, Kate Mas- terton, and Dina Demner-Fushman. 2018. Seman- tic annotation of consumer health questions. BMC Bioinform., 19(1):34:1–34:28. Rashmi Mishra, Jiantao Bian, Marcelo Fiszman, Charlene R. Weir, Siddhartha Jonnalagadda, Javed Mostafa, and Guilherme Del Fiol. 2014. Text sum- marization in the biomedical domain: A systematic review of recent research. Journal of Biomedical In- formatics, 52:457–467. Special Section: Methods in Clinical Research Informatics. Ravi Kondadadi, Sahil Manchanda, Jason Ngo, and Ronan McCormack. 2021. Optum at mediqa 2021: Abstractive summarization of radiology reports us- ing simple bart ﬁnetuning. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. References for medical summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. for medical summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Alexander M. Rush, Sumit Chopra, and Jason Weston. 2015. A neural attention model for abstractive sen- tence summarization. In Proceedings of the 2015 Conference on Empirical Methods in Natural Lan- guage Processing, EMNLP 2015, Lisbon, Portugal, September 17-21, 2015, pages 379–389. Yumo Xu and Mirella Lapata. 2020. Coarse-to- ﬁne query focused multi-document summarization. In Proceedings of the 2020 Conference on Em- pirical Methods in Natural Language Processing (EMNLP), pages 3632–3645. Max E. Savery, Asma Ben Abachaand Soumya Gayen, and Dina Demner-Fushman. 2020. Question-driven summarization of answers to consumer health ques- tions. Scientiﬁc Data, Nature, 7. Shweta Yadav, Mourad Sarrouti, and Deepak Gupta. 2021. Nlm at mediqa 2021: Transfer learning- based approaches for consumer question and multi- answer summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Abigail See, Peter J. Liu, and Christopher D. Manning. 2017. Get to the point: Summarization with pointer- generator networks. In Proceedings of the 55th An- nual Meeting of the Association for Computational Linguistics, ACL 2017, Vancouver, Canada, July 30 - August 4, Volume 1: Long Papers, pages 1073– 1083. Jingqing Zhang, Yao Zhao, Mohammad Saleh, and Pe- ter Liu. 2020a. PEGASUS: Pre-training with ex- tracted gap-sentences for abstractive summarization. In International Conference on Machine Learning. Thibault Sellam, Dipanjan Das, and Ankur P Parikh. 2020. Bleurt: Learning robust metrics for text gen- eration. arXiv preprint arXiv:2004.04696. Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q Weinberger, and Yoav Artzi. 2019. Bertscore: Eval- uating text generation with bert. arXiv preprint arXiv:1904.09675. Akshay Smit, Saahil Jain, Pranav Rajpurkar, Anuj Pa- reek, Andrew Ng, and Matthew Lungren. 2020. Combining automatic labelers and expert annota- tions for accurate radiology report labeling using BERT. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). Yuhao Zhang, Daisy Yi Ding, Tianpei Qian, Christo- pher D. Manning, and Curtis P. Langlotz. 2018. Learning to summarize radiology ﬁndings. In EMNLP 2018 Workshop on Health Text Mining and Information Analysis. Mario S¨anger, Leon Weber, and Ulf Leser. 2021. Wbi at mediqa 2021: Summarizing consumer health questions with generative transformers. References Yassine Mrabet and Dina Demner-Fushman. 2020. HOLMS: alternative summary evaluation with large language models. In Proceedings of the 28th In- ternational Conference on Computational Linguis- tics, COLING 2020, Barcelona, Spain (Online), De- cember 8-13, 2020, pages 5679–5688. International Committee on Computational Linguistics. Hoang-Quynh Le, Quoc-An Nguyen, Quoc-Hung Duong, Minh-Quang Nguyen, Huy-Son Nguyen, Tam Doan Thanh, Hai-Yen Thi Vuong, and Trang M. Nguyen. 2021. Uetﬁshes at mediqa 2021: Standing-on-the-shoulders-of-giants model for ab- stractive multi-answer summarization. In Proceed- ings of the 20th SIGBioMed Workshop on Biomed- ical Language Processing, NAACL-BioNLP 2021. Association for Computational Linguistics. Khalil Mrini, Franck Dernoncourt, Seunghyun Yoon, Trung Bui, Walter Chang, Emilias Farcas, and Ndapa Nakashole. 2021. Ucsd-adobe at mediqa 2021: Transfer learning and answer sentence selec- tion for medical summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Jooyeon Lee, Huong Dang, Ozlem Uzuner, and Sam Henry. 2021a. Mnlp at mediqa 2021: Fine-tuning pegasus for consumer health question summariza- tion. In Proceedings of the 20th SIGBioMed Work- shop on Biomedical Language Processing, NAACL- BioNLP 2021. Association for Computational Lin- guistics. Ramesh Nallapati, Feifei Zhai, and Bowen Zhou. 2017. Summarunner: A recurrent neural network based se- quence model for extractive summarization of doc- uments. In Proceedings of the Thirty-First AAAI Conference on Artiﬁcial Intelligence, February 4-9, 2017, San Francisco, CA, USA., pages 3075–3081. Lung-Hao Lee, Po-Han Chen, Yu-Xiang Zeng, Po- Lei Lee, and Kuo-Kai Shyu. 2021b. Ncuee-nlp at mediqa 2021: Health question summarization us- ing pegasus transformers. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Lan- guage Processing, NAACL-BioNLP 2021. Associa- tion for Computational Linguistics. Karolina Owczarzak, John M. Conroy, Hoa Trang Dang, and Ani Nenkova. 2012. An assess- ment of the accuracy of automatic evaluation in summarization. In Proceedings of Workshop on Evaluation Metrics and System Comparison for Automatic Summarization@NACCL-HLT 2012, Montr`eal, Canada, June 2012, 2012, pages 1–9. As- sociation for Computational Linguistics. Chin-Yew Lin. 2004. ROUGE: A package for auto- matic evaluation of summaries. In Text Summariza- tion Branches Out, pages 74–81, Barcelona, Spain. Association for Computational Linguistics. Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J Liu. 2020. Exploring the limits Chris J Lu, Alan R Aronson, Sonya E Shooshan, and Dina Demner-Fushman. 2019. Spell checker for 84 of transfer learning with a uniﬁed text-to-text trans- former. Journal of Machine Learning Research, 21:1–67. References In Proceed- ings of the 20th SIGBioMed Workshop on Biomed- ical Language Processing, NAACL-BioNLP 2021. Association for Computational Linguistics. Yu Yuhao Zhang, Derek Merck, Emily Bao Tsai, Christo- pher D. Manning, and Curtis Langlotz. 2020b. Op- timizing the factual correctness of a summary: A study of summarizing radiology reports. In Pro- ceedings of the 58th Annual Meeting of the Asso- ciation for Computational Linguistics. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. arXiv preprint arXiv:1706.03762. Ming Zhong, Pengfei Liu, Yiran Chen, Danqing Wang, Xipeng Qiu, and Xuanjing Huang. 2020. Extractive summarization as text matching. In Proceedings of the 58th Annual Meeting of the Association for Com- putational Linguistics. Wen Xiao and Giuseppe Carenini. 2019. Extractive summarization of long documents by combining global and local context. In Proceedings of the 2019 Conference on Empirical Methods in Natu- ral Language Processing and the 9th International Joint Conference on Natural Language Processing, EMNLP-IJCNLP 2019, Hong Kong, China, Novem- ber 3-7, 2019, pages 3009–3019. Association for Computational Linguistics. Wei Zhu, Yilong He, Ling Chai, Yunxiao Fan, Yuan Ni, GUOTONG XIE, and Xiaoling Wang. 2021. paht nlp at mediqa 2021: Multi-grained query fo- cused multi-answer summarization. In Proceedings of the 20th SIGBioMed Workshop on Biomedical Language Processing, NAACL-BioNLP 2021. Asso- ciation for Computational Linguistics. Liwen Xu, Yan Zhang, Lei Hong, Yi Cai, and Szui Sung. 2021. Chichealth @ mediqa 2021: Ex- ploring the limits of pre-trained seq2seq models 85
https://openalex.org/W2923547902	https://www.zora.uzh.ch/id/eprint/171180/1/Barbosa_Celebrating_40_years_of_ironman.pdf	English	null	Celebrating 40 Years of Ironman: How the Champions Perform	International journal of environmental research and public health/International journal of environmental research and public health	2,019	cc-by	4,837	Zurich Open Repository and Archive Zurich Open Repository and Archive University of Zurich University Library Strickhofstrasse 39 CH-8057 Zurich www.zora.uzh.ch University of Zurich University Library Strickhofstrasse 39 CH-8057 Zurich www.zora.uzh.ch Received: 7 February 2019; Accepted: 16 March 2019; Published: 20 March 2019 Abstract: We aimed to determine which discipline had the greater performance improvements in the history of Ironman triathlon in Hawaii and also which discipline had the greater inﬂuence in overall race time. Data from 1983 to 2018 of the top three women and men of each year who competed in the Ironman World Championship were included. In addition to exploratory data analyses, linear regressions between split times and years of achievement were performed. Further, a stepwise multiple linear regression was applied using total race time as the dependent variable and split times as the independent variables. Both women and men signiﬁcantly improved their performances from 1983 to 2018 in the Ironman World Championship. Swimming had the largest difference in improvements between men and women (3.0% versus 12.1%, respectively). A negative and signiﬁcant decrease in each discipline was identiﬁed for both women and men, with cycling being the discipline with the greatest reduction. The results from the stepwise multiple regression indicated that cycling was the discipline with the highest inﬂuence on overall race time for both sexes. Based on the ﬁndings of this study, cycling seems to be the Ironman triathlon discipline that most improved overall race times and is also the discipline with the greatest inﬂuence on the overall race time of elite men and women in the Ironman World Championship. Keywords: triathlon; cycling; running; swimming; endurance Celebrating 40 Years of Ironman: How the Champions Perform Barbosa, Lucas Pinheiro ; Sousa, Caio Victor ; Sales, Marcelo Magalhães ; Olher, Rafael Dos Reis ; Aguiar, Samuel Silva ; Santos, Patrick Anderson ; Tiozzo, Eduard ; Simões, Herbert Gustavo ; Nikolaidis, Pantelis Theodoros ; Knechtle, Beat DOI: https://doi.org/10.3390/ijerph16061019 DOI: https://doi.org/10.3390/ijerph16061019 Posted at the Zurich Open Repository and Archive, University of Zurich ZORA URL: https://doi.org/10.5167/uzh-171180 Journal Article Published Version The following work is licensed under a Creative Commons: Attribution 4. Posted at the Zurich Open Repository and Archive, University of Zurich ZORA URL: https://doi.org/10.5167/uzh-171180 Journal Article Published Version ollowing work is licensed under a Creative Commons: Attribution 4.0 International (CC BY 4.0) License. Originally published at: Barbosa, Lucas Pinheiro; Sousa, Caio Victor; Sales, Marcelo Magalhães; Olher, Rafael Dos Reis; Aguiar, Samuel Silva; Santos, Patrick Anderson; Tiozzo, Eduard; Simões, Herbert Gustavo; Nikolaidis, Pantelis Theodoros; Knech- tle, Beat (2019). Celebrating 40 Years of Ironman: How the Champions Perform. International Journal of Envi- ronmental Research and Public Health, 16(6):1019. DOI: https://doi.org/10.3390/ijerph16061019 International Journal of Environmental Research and Public Health Celebrating 40 Years of Ironman: How the Champions Perform Lucas Pinheiro Barbosa 1,†, Caio Victor Sousa 1,2,† , Marcelo Magalhães Sales 3 , Rafael dos Reis Olher 1, Samuel Silva Aguiar 1 , Patrick Anderson Santos 1, Eduard Tiozzo 2, Herbert Gustavo Simões 1, Pantelis Theodoros Nikolaidis 4 and Beat Knechtle 5,6,* 1 Graduate Program in Physical Education, Catholic University of Brasília, 71966-700 Brasília, Brazil; lduarte.barbosa@gmail.com (L.P.B.); cvsousa89@gmail.com (C.V.S.); rﬂolher@gmail.com (R.d.R.O.); ssaguiar0@gmail.com (S.S.A.); patricksantospas@gmail.com (P.A.S.); hgsimoes@gmail.com (H.G.S.) 2 Miller School of Medicine, University of Miami, Coral Gables, FL 33124, USA; etiozzo@med.miami.edu 3 Physical Education Department, Goias State University, Quirinópolis, 75860-000 GO, Brazil; marcelomagalhaessales@gmail.com 1 Graduate Program in Physical Education, Catholic University of Brasília, 71966-700 Brasília, Brazil; lduarte.barbosa@gmail.com (L.P.B.); cvsousa89@gmail.com (C.V.S.); rﬂolher@gmail.com (R.d.R.O.); ssaguiar0@gmail.com (S.S.A.); patricksantospas@gmail.com (P.A.S.); hgsimoes@gmail.com (H.G.S.) 2 Miller School of Medicine, University of Miami, Coral Gables, FL 33124, USA; etiozzo@med miami edu 6 Institute of Primary Care, University of Zurich, 8006 Zurich, Switzerland * Correspondence: beat.knechtle@hispeed.ch; Tel.: +41-(0)71-226-93-00 † These authors contributed equally to this work. 1. Introduction The Ironman triathlon consists of swimming 2.4 miles (3.8 km), cycling 112 miles (180 km), and running 26.2 miles (42.2 km) and is considered as one of the most challenging ultra-endurance events worldwide [1,2]. Although triathlon started in San Diego, California, the history of Ironman triathlon started in 1978 in Hawaii, with the ﬁrst Ironman World Championship being held in Kailua-Kona (Big Island) three years later, also in Hawaii [1–3]. Int. J. Environ. Res. Public Health 2019, 16, 1019; doi:10.3390/ijerph16061019 www.mdpi.com/journal/ijerph www.mdpi.com/journal/ijerph 2 of 9 Int. J. Environ. Res. Public Health 2019, 16, 1019 Nowadays, the Ironman events take place all over the world, with amateur and professional athletes competing in these events to qualify for the World Championship in Kailua-Kona. Ironman Hawaii in considered as the toughest Ironman race in the world due to the course, the environmental conditions, and the competitiveness of the event [2,4]. The race itself is one of the most popular triathlon events in the world, with a growing competitiveness and performance improvement in non-elite triathletes [1,5,6]. In addition, it should be highlighted that the best professional triathletes in the world often achieve new records in Kailua-Kona [7]. In order to help coaches and athletes with both training plans and race strategy, performance trends have been analyzed in the past few years in many endurance sports [8–11]. Speciﬁcally in triathlon, relevant studies have been conducted for Olympic distance (1.5 km swim/40 km cycle/10 km run) [12,13], half-distance (half-Ironman: 1.9 km swim/90 km cycle/21 km run) [13,14], full-distance (3.8 km swim/180 km cycle/42.195 km run) [14,15], and ultra-triathlons (distance > Ironman) [16,17]. To date, two studies investigated the performance of amateur triathletes [2,5], but none of them included only elite women and men. Ofoghi et al. [18] investigated which discipline would have the greater inﬂuence on overall performance in an Olympic triathlon and concluded that running was the most decisive, followed by swimming and cycling. On the other hand, Sousa et al. [19] analyzed all sub-8-h performances in full-distance triathlon (i.e., Ironman) and reported that cycling was the discipline with the greatest inﬂuence on the overall result, followed by running and swimming. Additionally, it is noteworthy that in 2018 the female and male winners of the 2018 World Championship improved the course records, showing that the fastest Ironman triathletes worldwide can further improve their performances. 1. Introduction However, to the best of our knowledge, the only two studies analyzing the Ironman World Championship results concerned amateur athletes in the analysis, with one of the studies analyzing races up to 2007 [2] and the other analyzing races from 2002 to 2015 [5]. Therefore, we aimed to analyze only elite men and women competing in the Ironman World Championship from 1983 to 2018 in order to determine (i) which discipline had the greatest performance improvement in the last 35 years; (ii) which discipline had the greatest inﬂuence on overall result; and (iii) whether women were really closing the gap to men. 2. Methods 2.1. Ethical Approval 2.1. Ethical Approval All procedures used in the study were approved by the Institutional Review Board of Kanton St. Gallen, Switzerland, with a waiver of the requirement for informed consent of the participants given the fact that the study involved the analysis of publicly available data (1 June 2010). 2.2. Data 2.2. Data All data were obtained from a publicly available database (www.ironman.com). All ofﬁcial overall race and split times from the top three women’s and men’s ﬁnishers of the Ironman World Championship from 1983 to 2018 were included in the analysis. Table 1 presents the descriptive distribution of women’s races including the Ironman World Championship Race/Split Record (among top three ﬁnishers), whereas Table 2 presents men’s data. Table 1. Women’s total and split race times in the Ironman World Championship from 1983 to 2018. Race Time Median (25–75 Percentile) Mean (±SD) Ironman World Championship Race/Split Record * Overall 09:16:48 (09:03:51–09:26:18) 09:19:06 (00:26:14) 08:26:18 Swimming 00:57:00 (00:55:26–01:00:09) 00:57:34 (00:03:24) 00:48:14 Table 1. Women’s total and split race times in the Ironman World Championship from 1983 to 2018. Int. J. Environ. Res. Public Health 2019, 16, 1019 3 of 9 Table 1. Cont. Race Time Median (25–75 Percentile) Mean (±SD) Ironman World Championship Race/Split Record * Cycling 05:08:39 (05:00:14–05:17:50) 05:11:22 (00:18:06) 04:26:07 Running 03:08:10 (03:04:09–03:16:31) 03:10:10 (00:10:37) 02:50:26 * Within top three ﬁnishers from 1983 to 2018. Table 2. Men’s total and split race times in the Ironman World Championship from 1983 to 2018. Race Time Median (25–75 Percentile) Mean (±SD) Ironman World Championship Race/Split Record * Overall 08:22:02 (08:14:37–08:33:02) 08:26:28 (00:18:26) 07:52:39 Swimming 00:51:43 (00:51:00–00:53:02) 00:53:18 (00:06:07) 00:48:02 Cycling 04:37:47 (04:30:16–04:46:15) 04:42:15 (00:21:01) 04:12:25 Running 03:08:10 (02:46:42–02:57:00) 02:55:21 (± 00:18:57) 02:39:59 * Within top three ﬁnishers from 1983 to 2018. Table 1. Cont. Table 1. Cont. Race Time Median (25–75 Percentile) Mean (±SD) Ironman World Championship Race/Split Record * Cycling 05:08:39 (05:00:14–05:17:50) 05:11:22 (00:18:06) 04:26:07 Running 03:08:10 (03:04:09–03:16:31) 03:10:10 (00:10:37) 02:50:26 * Within top three ﬁnishers from 1983 to 2018. Table 2. Men’s total and split race times in the Ironman World Championship from 1983 to 2018. Race Time Median (25–75 Percentile) Mean (±SD) Ironman World Championship Race/Split Record * Overall 08:22:02 (08:14:37–08:33:02) 08:26:28 (00:18:26) 07:52:39 Swimming 00:51:43 (00:51:00–00:53:02) 00:53:18 (00:06:07) 00:48:02 Cycling 04:37:47 (04:30:16–04:46:15) 04:42:15 (00:21:01) 04:12:25 Running 03:08:10 (02:46:42–02:57:00) 02:55:21 (± 00:18:57) 02:39:59 * Within top three ﬁnishers from 1983 to 2018. Table 2. Men’s total and split race times in the Ironman World Championship from 1983 to 2018. * Within top three ﬁnishers from 1983 to 2018. 2.3. Statistical Analysis Initially, an exploratory analysis of the data was carried out, in which central tendency (median and mean), dispersion (interquartile ranges (25 and 75 percentiles and standard deviation), and extreme (lowest value) measures were calculated (Tables 1 and 2). Furthermore, all data were transformed in seconds and non-linear regressions (second order) were performed between each split time and year of achievement. Linear regressions were used for splits because the non-linear ﬁtting line was the same as the linear. The relative difference (percentage) between the ﬁrst (1983) and last (2018) World Championship’s top three performances was calculated for both women and men. Regarding overall race time, non-linear regression analyses were performed since the trend line had a better ﬁt than linear regression. A comparison of average race times between the top three athletes and the chasing group (4th to 10th place ﬁnishers) was performed. Finally, a stepwise multiple linear regression was performed using overall race time as the dependent variable and split times as independent variables. The signiﬁcance level was set as 5% (p < 0.05), and all procedures were performed using SPSS v21.0 (IBM SPSS Statistics for Windows. Armonk, NY: IBM Corp). 3. Results Men improved in overall race time by 13.3% from 1983 to 2018, whereas women improved by 20.8% (Table 3). Swimming showed the largest difference in improvements between men and women (3.0% versus 12.1%, respectively), and running showed the smallest difference (12.5% versus 15.5%, respectively) for the three split disciplines. Table 3. Women’s and men’s percentage performance improvements in the Ironman World Championship from 1983 to 2018. Table 3. Women’s and men’s percentage performance improvements in the Ironman World Championship from 1983 to 2018. Table 3. Women’s and men’s percentage performance improvements in the Ironman World Championship from 1983 to 2018. Total Total Difference Decade Average Decade Average Difference Overall Women 20.8% 7.5% 5.20% 1.87% Men 13.3% 3.33% Swimming Women 12.1% 9.1% 3.25% 2.50% Men 3% 0.75% Int. J. Environ. Res. Public Health 2019, 16, 1019 4 of 9 Table 3. Cont. Table 3. Cont. Total Total Difference Decade Average Decade Average Difference Cycling Women 26.4% 9.5% 6.60% 2.37% Men 16.9% 4.23% Running Women 15.5% 3% 3.88% 0.75% Men 12.5% 3.13% Total Total Difference Decade Average Decade Average Difference Both women and men signiﬁcantly improved their performances from 1983 to 2018 in the Ironman World Championship in Kona, Hawaii (Figures 1 and 2). The world record was improved almost every three years (see Supplementary Table S1 for accurate race time values from each year’s champions). Figure 1. Dispersion and non-linear regression of overall race time performances in the Ironman World Championship from 1983 to 2018 of women and men. Gold trophies represent the champion in each year. Figure 1. Dispersion and non-linear regression of overall race time performances in the Ironman World Championship from 1983 to 2018 of women and men. Gold trophies represent the champion in each year. 5 of 9 Int. J. Environ. Res. Public Health 2019, 16, 1019 t. J. Environ. Res. Public Health 2019, 16, 1019 5 of Figure 2. Dispersion and non-linear regression overall race time performances between the top three ﬁnishers and the chasing group (4th to 10th place ﬁnishers) from women and men in the Ironman World Championship from 1983 to 2018. Figure 2. Dispersion and non-linear regression overall race time performances between the top three ﬁnishers and the chasing group (4th to 10th place ﬁnishers) from women and men in the Ironman World Championship from 1983 to 2018. 3. Results The linear regression of split disciplines shows a negative and signiﬁcant slope for all disciplines for both women (swimming: −6.94 to 0.47; cycling: −71.06 to −36.98 ; running: −45.79 to −26.86 ; Figure 3) and men (swimming: −19.37 to −6.77 ; cycling: −96.01 to −60.52 ; running: −65.06 to −30.58 ; Figure 4) ( indicates p < 0.001). The greatest slope in both sexes was for cycling. − − − − − − − − − − − Figure 3. Dispersion and linear regression of split-times performances in the Ironman World Championship from 1983 to 2018 of women. Figure 3. Dispersion and linear regression of split-times performances in the Ironman World Championship from 1983 to 2018 of women. 6 of 9 Int. J. Environ. Res. Public Health 2019, 16, 1019 Figure 4. Dispersion and linear regression of split-times performances in the Ironman World Championship from 1983 to 2018 of men. Figure 4. Dispersion and linear regression of split-times performances in the Ironman World Championship from 1983 to 2018 of men. The best-ﬁtting model from the stepwise multiple regression included swimming, cycling, and running split times for both women and men (Table 4). Cycling was the discipline with the greatest standardized beta for both sexes. The swimming discipline resulted in a negative standardized coefﬁcient for the men. Table 4. Standardized coefﬁcient from stepwise multiple regression using total race time as the dependent variable of Ironman World Championship from 1983 to 2018. Table 4. Standardized coefﬁcient from stepwise multiple regression using total race time as the dependent variable of Ironman World Championship from 1983 to 2018. Standardized β Coefﬁcient R2 R2aj Swimming Cycling Running Women 0.129 0.690 0.405 0.857 0.856 Men −0.290 0.895 0.250 0.781 0.775 Standardized β Coefﬁcient R2 R2aj Swimming Cycling Running Women 0.129 0.690 0.405 0.857 0.856 Men −0.290 0.895 0.250 0.781 0.775 4. Discussion The main ﬁnding of this manuscript was that cycling has been the Ironman triathlon discipline with the greatest improvement rate throughout the years and also has had the greatest inﬂuence on overall race time for both women and men. However, apparently both women and men have improved their performances over the years in all triathlon disciplines. It is worth mentioning that women had a greater improvement than men in all triathlon disciplines and consequently in total race times. Jeukendrup and Martin [20] had previously reported that cycling in aero position and the use of lighter wheels (i.e., elbows on handlebars and carbon wheels, respectively, which had developed for use in time-trial and triathlon bicycles) makes an athlete signiﬁcantly faster. Thus, cycling performance also had new technologies that could inﬂuence the performance increase, from the outﬁt to the bicycle itself, all of which contributed to make the athlete more comfortable, aerodynamic, and consequently faster. Although cycling is the discipline that encompasses more time in comparison to swimming and running in Olympic distance and short distances, it does not have an inﬂuence in overall race time, being the least important of the three disciplines [21]. In Olympic distance and short distances, athletes normally swim really fast to be able to leave transition one with the ﬁrst pack of cyclists and stay within the leading and chasing peloton, thus saving the energy for the running [18]. However, 7 of 9 Int. J. Environ. Res. Public Health 2019, 16, 1019 in Ironman races drafting during cycling is not allowed, making cycling a more competitive discipline, which means that athletes have to apply some strategy in order to cycle fast enough to remain in a competitive position but still save energy for the running leg. Similarly, in an analysis only including top full-distance triathlon performances, the authors reported that cycling was the discipline that most inﬂuenced overall performance in elite men racing below 8 h of overall race time, followed by running and swimming [19]. A performance analysis on Ironman races investigated more than 340,000 triathletes racing in 253 different race locations and concluded that the race tactics in an Ironman triathlon should focus on saving energy during the ﬁrst two disciplines for the running split [22]. This conclusion is different from the ﬁndings of the present study, which suggest that athletes seem to apply greater effort in cycling than during running. 4. Discussion It is worth mentioning that this analysis was carried out with a majority of age groupers (non-elite), whereas the present study only considered the top three elite professionals from each year. It is noteworthy that 4th to 10th place ﬁnishers in the Ironman World Championship seemed to have a substantial performance improvement in the last decade of the event, with consistently much closer groupings in the top ten athletes for both men and women. With regard to performance throughout the race, the performance analysis of the Ironman World Championship with amateurs reported a performance increase in all disciplines for men and women [23]. However, the authors suggest that this improvement in performance may be due to an increased number of athletes and morphological changes [23]. The overall performance increase throughout the years can be mostly attributed to the development of new nutrition and training strategies [24–27]. A controversial result was the negative coefﬁcient for the swim split in men, which would mean that a slower swim could lead to a better overall race time. We believe that this statistical outcome is due to the speciﬁcity of the sample, as only the top three athletes in the overall race were considered, and these athletes are not always the best swimmers. For example, in the 2018 World Championship, none of the top three overall athletes were among the top 10 swimmers. Concerning the performance gap between men and women, it has markedly reduced in the last decades. At the 2018 Ironman World Championship, women improved by 21% while men improved by 13%; the absolute gap between them reduced from 1 h and 38 min to 33 min. Indeed, in the 2018 Championship, the female champion Daniela Ryf crossed the ﬁnish line ahead of 20 elite professional men who ﬁnished the race. Some previous studies have concluded that women have been closing the gap in swimming [28,29], in running [30,31], and even in triathlon [2]. We believe that women may still close the gap in an Ironman someday despite the body composition and physiological differences that exist between men and women. One of the possible explanations for this can be attributed to cultural changes that have favored a greater participation of women in all sports, including triathlon, thus increasing competitiveness and therefore performance [7,9,10,32]. References 1. Stiefel, M.; Rüst, C.A.; Rosemann, T.; Knechtle, B. A comparison of participation and performance in age-group ﬁnishers competing in and qualifying for Ironman Hawaii. Int. J. Gen. Med. 2013, 6, 67. 1. Stiefel, M.; Rüst, C.A.; Rosemann, T.; Knechtle, B. A comparison of participation and performance in age-group ﬁnishers competing in and qualifying for Ironman Hawaii. Int. J. Gen. Med. 2013, 6, 67. . Corporation, W.T. The Ironman Story. Available online: http://www.ironman.com/triathlon/history.as axzz2BrLoGbOE (accessed on 22 October 2018). 4. Sparks, S.; Cable, N.; Doran, D.; Maclaren, D. The inﬂuence of environmental temperature on duathlon performance. Ergonomics 2005, 48, 1558–1567. [CrossRef] [PubMed] 4. Sparks, S.; Cable, N.; Doran, D.; Maclaren, D. The inﬂuence of environmental temperature on duathlon performance. Ergonomics 2005, 48, 1558–1567. [CrossRef] [PubMed] 5. Käch, I.W.; Rüst, C.A.; Nikolaidis, P.T.; Rosemann, T.; Knechtle, B. The age-related performance decline in Ironman triathlon starts earlier in swimming than in cycling and running. J. Strength Cond. Res. 2018, 32, 379–395. [CrossRef] [PubMed] 6. Stiefel, M.; Knechtle, B.; Lepers, R. Master triathletes have not reached limits in their Ironman triathlon performance. Scand. J. Med. Sci. Sports 2014, 24, 89–97. [CrossRef] 7. Knechtle, B.; Zingg, M.A.; Rosemann, T.; Rüst, C.A. Sex difference in top performers from Ironman to Double Deca Iron ultra-triathlon. Open Access J. Sports Med. 2014, 5, 159. [CrossRef] 8. Knechtle, B.; Knechtle, P.; Rüst, C.A.; Rosemann, T. A comparison of anthropometric and training characteristics of Ironman triathletes and Triple Iron ultra-triathletes. J. Sports Sci. 2011, 29, 1373–1380. [CrossRef] 9. Etter, F.; Knechtle, B.; Bukowski, A.; Rüst, C.A.; Rosemann, T.; Lepers, R. Age and gender interactions in short distance triathlon performance. J. Sports Sci. 2013, 31, 996–1006. [CrossRef] 10. Le Meur, Y.; Hausswirth, C.; Dorel, S.; Bignet, F.; Brisswalter, J.; Bernard, T. Inﬂuence of gender on pacing adopted by elite triathletes during a competition. Eur. J. Appl. Physiol. 2009, 106, 535–545. [CrossRef] 11. Esteve-lanao, J.; San Juan, A.F.; Earnest, C.P.; Foster, C.; Lucia, A. How do endurance runners actually train? Relationship with competition performance. Med. Sci. Sports Exerc. 2005, 37, 496–504. [CrossRef] 12. Wonerow, M.; Rüst, C.A.; Nikolaidis, P.T.; Rosemann, T.; Knechtle, B. Performance Trends in Age Group Triathletes in the Olympic Distance Triathlon at the World Championships 2009-2014. Chin. J. Physiol. 2017, 60, 137–150. [CrossRef] 13. Wu, S.S.X.; Peiffer, J.J.; Brisswalter, J.; Nosaka, K.; Lau, W.Y.; Abbiss, C.R. 5. Conclusions In conclusion, cycling seems to be the triathlon discipline that most improved overall race times and is also the discipline that had the greatest inﬂuence on the overall race time in elite men and women in the Ironman World Championship. Furthermore, within the last 40 years of Ironman Hawaii, both men and women improved overall time performance, but women improved more, thereby closing the gap to men. Supplementary Materials: The following are available online at http://www.mdpi.com/1660-4601/16/ s1, Table S1: Overall times in the Ironman World Championship from 1983 to 2018 for women and men. Author Contributions: Conceptualization: C.V.S., L.P.B., P.T.N., and B.K.; methodology: C.V.S., L.P.B., M.M.S., R.d.R.O., S.S.A., P.A.S., P.T.N., and B.K.; formal analysis: C.V.S., L.P.B., and M.M.S.; writing—original draft preparation: C.V.S. and L.P.B.; writing—review and editing: C.V.S., L.P.B., M.M.S., P.A.S., E.T., H.G.S., P.T.N., and B.K.; visualization: C.V.S., L.P.B., M.M.S., P.A.S., E.T., H.G.S., P.T.N., and B.K.; supervision: P.T.N. and B.K.; project administration: P.T.N., and B.K. Author Contributions: Conceptualization: C.V.S., L.P.B., P.T.N., and B.K.; methodology: C.V.S., L.P.B., M.M.S., R.d.R.O., S.S.A., P.A.S., P.T.N., and B.K.; formal analysis: C.V.S., L.P.B., and M.M.S.; writing—original draft preparation: C.V.S. and L.P.B.; writing—review and editing: C.V.S., L.P.B., M.M.S., P.A.S., E.T., H.G.S., P.T.N., and B.K.; visualization: C.V.S., L.P.B., M.M.S., P.A.S., E.T., H.G.S., P.T.N., and B.K.; supervision: P.T.N. and B.K.; project administration: P.T.N., and B.K. Author Contributions: Conceptualization: C.V.S., L.P.B., P.T.N., and B.K.; methodology: C.V.S., L.P.B., M.M.S., R.d.R.O., S.S.A., P.A.S., P.T.N., and B.K.; formal analysis: C.V.S., L.P.B., and M.M.S.; writing—original draft preparation: C.V.S. and L.P.B.; writing—review and editing: C.V.S., L.P.B., M.M.S., P.A.S., E.T., H.G.S., P.T.N., and B.K.; visualization: C.V.S., L.P.B., M.M.S., P.A.S., E.T., H.G.S., P.T.N., and B.K.; supervision: P.T.N. and B.K.; project administration: P.T.N., and B.K. Funding: This research received no external funding. Funding: This research received no external funding. Conﬂicts of Interest: The authors report no conﬂicts of interest in this work. Conﬂicts of Interest: The authors report no conﬂicts of interest in this work. 4. Discussion When comparing performance with other endurance modalities such as ultra-triathlon and marathon running, the performance gaps between women and men are getting smaller each year. Knechtle et al. [33] investigated the performance trends of Double Iron ultra-triathlon (2I; 2x Ironman distance), Triple Iron ultra-triathlon (3I; 3x Ironman distance), and Deca Iron ultra-triathlon (10I; 10x Ironman distance) from 1985 to 2009 and reported a smaller sex difference in 2I and 3I. Conversely, Nikolaidis et al. [31] investigated the performance of male and female athletes running the marathon and concluded that men are still faster than women, but the performance gap remained unchanged for the past few years. Regarding the speciﬁc Ironman World Championship, Kailua-Kona is one of the toughest races within the entire Ironman circuit, which typically requires athletes to swim in choppy waters, cycle with a lot of wind, and run in hot and sunny weather [34,35]. The race course has not always been the same in Ironman Hawaii, with small changes every two or three years in order to accommodate safety precautions and/or local transit logistics. Although this may affect the overall race time, the distances remained standard and we believe that any small course changes affecting a speciﬁc split race time are diluted within the sample and do not represent a great confounder to the general results of this study. Int. J. Environ. Res. Public Health 2019, 16, 1019 8 of 9 References Pacing strategies during the swim, cycle and run disciplines of sprint, Olympic and half-Ironman triathlons. Eur. J. Appl. Physiol. 2015, 115, 1147–1154. [CrossRef] 14. Knechtle, R.; Rüst, C.A.; Rosemann, T.; Knechtle, B. The best triathletes are older in longer race distances–a comparison between Olympic, Half-Ironman and Ironman distance triathlon. Springerplus 2014, 3, 538. [CrossRef] [PubMed] 9 of 9 Int. J. Environ. Res. Public Health 2019, 16, 1019 15. Dallam, G.M.; Jonas, S.; Miller, T.K. Medical considerations in triathlon competition. Sports Med. 2005, 35, 143–161. [CrossRef] 16. Zaryski, C.; Smith, D.J. Training principles and issues for ultra-endurance athletes. Curr. Sports Med. Rep. 2005, 4, 165–170. [CrossRef] [PubMed] 17. Abbiss, C.R.; Quod, M.J.; Martin, D.T.; Netto, K.J.; Nosaka, K.; Lee, H.; Surriano, R.; Bishop, D.; Laursen, P.B. Dynamic pacing strategies during the cycle phase of an Ironman triathlon. Med. Sci. Sports Exerc. 2006, 38, 726–734. [CrossRef] 18. Ofoghi, B.; Zeleznikow, J.; Macmahon, C.; Rehula, J.; Dwyer, D.B. Performance analysis and prediction in triathlon. J. Sports Sci. 2016, 34, 607–612. [CrossRef] [PubMed] 19. Sousa, C.V.; Barbosa, L.P.; Sales, M.M.; Santos, P.A.; Tiozzo, E.; Simões, H.G.; Nikolaidis, P.T.; Knechtle, B. Cycling as the Best Sub-8-Hour Performance Predictor in Full Distance Triathlon. Sports 2019, 7, 24. [CrossRef] 20. Jeukendrup, A.E.; Martin, J. Improving cycling performance. Sports Med. 2001, 31, 559–569. [CrossRef] [PubMed] 21. Figueiredo, P.; Marques, E.A.; Lepers, R. Changes in contributions of swimming, cycling, and running performances on overall triathlon performance over a 26-year period. J. Strength Cond. Res. 2016, 30, 2406–2415. [CrossRef] [PubMed] 22. Knechtle, B.; Käch, I.; Rosemann, T.; Nikolaidis, P.T. The effect of sex, age and performance level on pacing of Ironman triathletes. Res. Sports Med. 2019, 27, 99–111. [CrossRef] [PubMed] 23. Lepers, R.; Rust, C.A.; Stapley, P.J.; Knechtle, B. Relative improvements in endurance performance with age: Evidence from 25 years of Hawaii Ironman racing. Age 2013, 35, 953–962. [CrossRef] [PubMed] 24. Issurin, V.B. Biological Background of Block Periodized Endurance Training: A Review. Sports Med. 2018. [CrossRef] 25. McCormick, A.; Meijen, C.; Marcora, S. Psychological Determinants of Whole-Body Endurance Performance. Sports Med. 2015, 45, 997–1015. [CrossRef] [PubMed] 26. Doma, K.; Deakin, G.B.; Bentley, D.J. Implications of Impaired Endurance Performance following Single Bouts of Resistance Training: An Alternate Concurrent Training Perspective. Sports Med. 2017, 47, 2187–2200. [CrossRef] 27. Burke, L.M.; Hawley, J.A.; Jeukendrup, A.; Morton, J.P.; Stellingwerff, T.; Maughan, R.J. © 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/). References Toward a Common Understanding of Diet-Exercise Strategies to Manipulate Fuel Availability for Training and Competition Preparation in Endurance Sport. Int. J. Sport Nutr. Exerc. Metab. 2018, 28, 451–463. [CrossRef] [PubMed] 28. Nikolaidis, P.T.; Di Gangi, S.; de Sousa, C.V.; Valeri, F.; Rosemann, T.; Knechtle, B. Sex difference in open-water swimming-The Triple Crown of Open Water Swimming 1875–2017. PLoS ONE 2018, 13, e0202003. [CrossRef] [PubMed] 29. Nikolaidis, P.T.; de Sousa, C.V.; Knechtle, B. Sex difference in long-distance open-water swimming races - does nationality play a role? Res. Sports Med. 2018, 26, 332–344. [CrossRef] [PubMed] 30. Hoffman, M.D.; Wegelin, J.A. The Western States 100-mile endurance run: Participation and performance trends. Med. Sci. Sports Exerc. 2009, 41, 2191. [CrossRef] 31. Nikolaidis, P.T.; Rosemann, T.; Knechtle, B. Sex Differences in the Age of Peak Marathon Race Time. Chin. J. Physiol. 2018, 61, 85–91. [CrossRef] Tarnopolsky, M.A. Gender differences in substrate metabolism during endurance exercise. Can. J Appl. Physiol. 2000, 25, 312–327. [CrossRef] pp y 33. Knechtle, B.; Knechtle, P.; Lepers, R. Participation and performance trends in ultra-triathlons from 1985 to 2009. Scand. J. Med. Sci. Sports 2011, 21, e82–e90. [CrossRef] 34. Sousa, C.; Aguiar, S.; Olher, R.; Sales, M.; Moraes, M.; Nikolaidis, P.; Rosemann, T.; Knechtle, B.; Simões, H. Hydration status after an Ironman triathlon: A meta-analysis. J. Hum. Kinet. 2018. [CrossRef] 35. Laird, R.H.; Johnson, D. The medical perspective of the Kona Ironman Triathlon. Sports Med. Arthrosc. Rev. 2012, 20, 239. [CrossRef] © 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/W3181355478	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0254754&type=printable	English	null	Impact of implementation of the national institute for health and clinical excellence (NICE) head injury guideline in a tertiary care center emergency department: A pre and post-intervention study	PloS one	2,021	cc-by	7,853	PLOS ONE PLOS ONE RESEARCH ARTICLE Methods We consecutively recruited 139 traumatic head injury (THI) patients in this prospective pre- post interventional study conducted in the ED of a tertiary care center. We implemented the NICE guideline into routine practice using multimodal intervention through physicians’ educa- tion sessions, information sheets and guideline-dissemination. The pre and post-implementa- tion CT head scan rates were compared. The post-implementation guideline adherence was assessed. Online Google form-questionnaires including 12 validated case scenarios were dis- tributed to the attending physicians at the end of both phases to assess their confidence levels. Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0254754 Copyright: © 2021 Pradhan 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. Editor: Belinda J. Gabbe, Monash University, AUSTRALIA Received: March 19, 2021 Accepted: July 4, 2021 Published: July 15, 2021 Received: March 19, 2021 Accepted: July 4, 2021 Published: July 15, 2021 Abstract Citation: Pradhan P, Pradhan A, Shrestha AP, Shrestha A, Paudel RC, Shrestha R (2021) Impact of implementation of the national institute for health and clinical excellence (NICE) head injury guideline in a tertiary care center emergency department: A pre and post-intervention study. PLoS ONE 16(7): e0254754. https://doi.org/ 10.1371/journal.pone.0254754 Pratisha PradhanID1☯, Alok Pradhan1‡, Anmol Purna Shrestha1‡, Abha Shrestha2‡, Ram Chandra Paudel3‡, Roshana ShresthaID1☯ Pratisha PradhanID1☯, Alok Pradhan1‡, Anmol Purna Shrestha1‡, Abha Shrestha2‡, Ram Chandra Paudel3‡, Roshana ShresthaID1☯ 1 Department of General Practice and Emergency Medicine, Kathmandu University School of Medical Sciences, Dhulikhel, Kavrepalanchok, Bagmati Province, Nepal, 2 Department of Community Medicine, Kathmandu University School of Medical Sciences, Dhulikhel, Kavrepalanchok, Bagmati Province, Nepal, 3 Department of Radiodiagnosis and Imaging, Kathmandu University School of Medical Sciences, Dhulikhel, Kavrepalanchok, Bagmati Province, Nepal ☯These authors contributed equally to this work. ‡ AP, APS, AS and RCP also contributed equally to this work. * pratisha.7@gmail.com Introduction Head injury, a common presentation to the emergency department (ED), is a substantial problem in developing countries like Nepal. The current national institute for health and clini- cal excellence (NICE) guideline revised in January 2014 focuses on effective clinical assessment and early management of head injuries according to their severity in all age groups. This study assessed the impact of implementing this guideline on the proportions of computed tomography (CT) head scans, guideline adherence, and confidence level of the attending physicians. Editor: Belinda J. Gabbe, Monash University, AUSTRALIA Background Traumatic head injury (THI) is a comprehensive term that describes injuries to the scalp, skull and/or underlying tissue in the head due to trauma other than superficial injuries to the face [1]. Worldwide, 69 million individuals are estimated to sustain THI every year with an inci- dence of mild THI of about 131 cases per 100,000 people, moderate THI of about 15 cases per 100,000 people, and severe THI approximately 14 cases per 100,000 people [2]. Due to the rapid surge in development, motorization and economical liberty, the risk of THI has increased in many Asian countries. Available data show that Asia has the highest percentage of THI due to falls (77.0%), unintentional injuries (57.0%), and road traffic accidents (RTA) (48.0%) [3]. In Nepal, the incidence of THI is estimated to be around 382 per 100,000 with a maximum number of patients presenting with mild THI [4, 5]. Patients with THI initially present to the emergency department (ED) and only a few patients require any intervention. Neurologic examinations and clinical history are a must, fol- lowed by a radiological description of THI lesions in imaging. The non-contrast computed tomography (NCCT) has become the investigation of choice for THI cases as it has both sensi- tivities and specificities approaching 100% for detecting any intracranial lesion [6–8]. In resource-limited settings like ours, CT scans must be ordered thoughtfully due to financial apprehensions. Patients’ increased medical expenditures, prolonged ED stays, and concern for adverse effects of radiation exposure and physician’s fear of any missed findings and desire to expedite a diagnosis, all affect the decision-making capacity immensely. A composed approach is required to ensure the ordering of head CTs when necessary while vindicating the potential disadvantages of over-imaging. For this, appropriate guidelines can help scrutinize cases [9, 10]. At hospitals in our country, the foremost attending physicians in the ED are the medical officers (MO) who are recent graduates. Their exposure to the ED is limited to about one month during their internship. The decision to order a CT head scan for THI patients is chal- lenging as their experience may not be sufficient in making a rational judgment regarding the use of CT imaging for such patients [9]. Results The implementation resulted in a statistically significant decrease in the proportion of CT head scan rates from 92.0% to 70.0% (p-value = 0.005). Following educational 1 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact interventions, improved guideline adherence of 20.3 percentage points (p-value = 0.001) was observed. Nine ED attending physicians were enrolled in the study who showed statisti- cally significant improvement in their confidence level following the intervention. The NICE guideline showed a sensitivity and specificity of 93.6% and 76.4% with 82.6% accuracy compared to that of clinical judgment (100%, 34.6%, and 58.1% respectively) in detecting intracranial lesions. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: The author(s) received no specific funding for this work. Competing interests: The authors have declared that no competing interests exist. Conclusion The implementation was successful in satisfying the aim of the NICE guideline by decreas- ing the proportion of CT head scans, improving guideline adherence and increasing the con- fidence of the attending physicians. PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 Study setting The study was conducted in the ED of Dhulikhel Hospital-Kathmandu University Hospital (DH-KUH), a community-based teaching hospital serving people from more than 50 out of 75 districts of the country. According to the ED audit of 2018, nearly 20,000 patients visited the ED, of which about 40% were trauma patients that year. It is a 30-bedded ED staffed by 5 facul- ties, 10–12 MOs and nurses/paramedics. The ED patients are first received by the on-duty tri- age-health personnel who take a short focused history and vitals to categorize them into trauma or non-trauma and further according to the severity in reference to Rapid Emergency Triage and Treatment System (RETTS) [12]. Following triage, the patient is taken to the respective zones in ED and the treatment is carried out by the attending MO/faculty. Background Several decision rules [11] are accessible to guide clini- cal decision-making for the use of CT scans for THI patients and on comparison of these vali- dated rules, the national institute for health and clinical excellence (NICE) criteria showed the highest specificity to identify any intracranial traumatic lesions. 2 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact The NICE head injury guideline [1] was developed in 2003 and updated in 2007 that ensured the replacement of skull radiography by CT scans as the prime imaging modality for evaluating THI cases. Then in 2014, the updated NICE guideline focused on earlier imaging and reporting after various systematic reviews. Its sensitivity and specificity for intracranial pathology range from 46.1 to 73.0% and 61% to 82.1% respectively [8, 11]. Study participants All patients who came to the ED of DH with THI within 24 hours of the incident were enrolled in the study. THI was defined as any injury with impact around the head and/or in the face due to trauma such as RTA, fall injury, physical assault and others (sports injuries, firearm, struck by object and occupation-related injuries). Patients with non-traumatic head injuries i.e. injuries to the brain that are not caused by an external physical force to the head, such as stroke, hypertensive subarachnoid hemorrhage, hypoxic injuries, were excluded from the study. Penetrating head injuries in which the dura mater is breached is an obvious tell-tale- sign of brain injury, and thus were not included in the study. Furthermore, brought dead patients, incomplete data, re-attendees for the same head injury, pregnant women and patient/ visitors denying consent were also excluded. For evaluation of self-reported confidence levels of the attending physicians, all the medical officers of the ED participated voluntarily. Study design This was a prospective pre-post intervention study. Objectives Despite such multiple validated guidelines and data demonstrating their validity and generaliz- ability, execution of these rules in our clinical settings has been suboptimal. The primary out- come of this study was to assess the impact of implementing the NICE guideline on the proportion of CT scans performed. Furthermore, this study evaluated the guideline adherence and decision-making capacity of the attending physicians, before and after the implementation of the guideline. Variables After ethical approval by the Institutional Review Committee (IRC number 155/18) on December 02, 2018, we developed a proforma for data collection which was validated by all five faculty members of the ED. Necessary amendments were made and finalized after the pilot testing in ten THI patients not included in the study. 3 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact The attending physician recorded data in the predesigned proforma which included hospi- tal identification number, demographic data {age (<16 years and 16 years), gender, (male or female)}, triage category (red, orange, yellow, green), mechanism of the injury (RTA, fall, assault, others), mode of transport (self or ambulance), any pre-hospital treatment, the pres- ence of influence of drugs/alcohol and any comorbidities. Glasgow coma scale (GCS) [13] on arrival (13–15, 9–12, 8 as mild, moderate and severe head injuries, respectively), physician’s clinical diagnosis and information on whether CT was sent or not were noted. If CT was sent, the indication of CT head according to the physician’s opinion and CT head findings as stated by a radiologist were recorded. Lastly, the disposition of the patients from the ED (discharge, admission, referral and mortality) was mentioned. The outcome of each patient who was dis- charged from the ED without a CT head or who refused the investigation was followed up on the 10th day and his/her health status was inquired and documented. An online Google form–questionnaire was sent via email to all the attending physicians that contained 12 case scenarios (four pediatric and eight adults) related to trauma for which they had to decide whether or not head CT was indicated. If they assumed indicated, they were asked to mention the indication for CT according to their opinion. Furthermore, the level of confidence using a Likert scale of 10 (1 = low and 10 = high) was used to determine their confi- dence level to order a head CT for each case. The 12 case scenarios were referenced from real cases visiting ED of DH and were validated by the ED faculties. PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 Study size This study considered a 95% confidence interval, 80% power, an equal-intervention/control ratio, margin of error of 0.3 and a drop rate of 25%. Since the proportions of CT head scans performed in the pre and post-implementation of the NICE guideline were not known, we considered it to be 50%. Using the standard statistical formula, the sample size to compare two proportions was calculated to be 46 in each group. Thus, consecutive sampling of 50 patients was done in each study period [14] as shown in Fig 1. Study procedure Standards for Reporting Diagnostic Accuracy studies (STARD) [15] flow diagrams for patients who presented with THI in pre and post-implementation periods respectively. https://doi.org/10.1371/journal.pone.0254754.g001 https://doi.org/10.1371/journal.pone.0254754.g001 Study procedure Pre-implementation period. The data collection was carried out from 2nd February, 2019 to 28th March, 2019, only after the initial stabilization of the patient and hence patient manage- ment was not delayed by the research. Informed written consent was taken from the patient or the patient’s key caretaker (in case the patient’s condition was not sound enough to give con- sent) meeting the inclusion criteria. The indication of the CT head was decided by the physi- cians according to their clinical judgment. We reviewed the electronic medical records of the samples to determine if the CT had been requested in reference to the NICE guideline. The online Google-questionnaire to assess the knowledge and confidence was sent to all the attend- ing physicians of the ED and their responses were collected. Intervention period. After completion of the pre-implementation period, a multimodal intervention was carried out to implement the NICE guideline into routine practice at ED. Components of the intervention included the clinical endorsement of the guideline in the ED, staff education and guideline dissemination. A didactic case-based teaching session was con- ducted for all the clinical staff of ED including faculties, resident doctors, attending physicians, nurses and paramedics. During the session, the current departmental pattern regarding head CTs and results of the pre-implementation phase was shared, followed by an explanation of the flowchart depicting the NICE guideline to enhance their clinical decision-making capabil- ity. Posters illustrating the NICE guideline flowchart were displayed in all zones of ED. Post-implementation period. In the post-implementation period (1st May, 2019 to 20th June, 2019), a checklist of the NICE guideline CT head indications was attached along with the proforma. The attending physicians of the ED collected the data in a similar fashion. We rein- forced the adherence to the guideline by working on-site for the first week of its implementa- tion. We collected such duly filled proforma from the ED and the electronic medical record search corroborated the patient information. After the post-implementation phase, the online Google form-questionnaire containing the same 12 case scenarios were distributed to the attending physicians and their confidence level was explored. 4 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact Fig 1. Standards for Reporting Diagnostic Accuracy studies (STARD) [15] flow diagrams for patients who presented with THI in pre and post-implementation periods respectively. Fig 1. PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 Performance of the NICE guideline In total, 81 CT head scans were performed during the study period of which 14 (six and eigh- teen for pre-and post-implementation phases respectively) were performed on patient request. On excluding those, in the pre-implementation period, 40 out of the 44 THI patients required a head CT scan on retrospective implementation of the guideline. In the post-implementation period, 27 out of the 42 THI patients required a head CT scan according to the NICE guide- line. The THI patients who did not undergo a CT head were followed up through phone calls and none had any danger signs or required re-visits. The diagnostic evaluation of the NICE guideline in comparison to the physician’s clinical judgment was done using a 22 table and calculated as shown in Tables 2 and 3, with the CT findings and final outcome kept as the gold standard. A 100% guideline adherence was not achieved in the post-implementation period in our study. Baseline characteristics The demography, incident details and clinical parameters of the study population are illus- trated in Table 1. A comparison of the baseline characteristics of both the study populations was made which showed that they were similar groups of samples. Results Following the implementation of the NICE guideline in this study, there was a decrease in the proportion of CT head scans performed, increased guideline adherence and an increase in the confidence levels of the attending physicians to indicate a CT head scan in cases of THI. The proportion of CT head scans performed decreased from 46 (92.0%) to 35 (70.0%) in our study. There was also a decrease in the proportion of CT head scans performed without indication from 17 (42.5%) to 6 (22.2%) and improved adherence in relation to CT scanning of the head from 57.5% to 77.8%. A total of nine ED attending physicians were enrolled in the study who showed statistically significant improvement in their confidence level post-intervention. Proportion of CT head scans and CT findings in the pre and post- implementation periods The proportion of CT head scans performed in the pre and the post-implementation periods are shown in Table 4. In comparison, there was a significant decrease in the proportion of CT scans performed after the implementation of the guideline (p-value = 0.005). There were how- ever 2 cases in the pre-implementation period who underwent CT head scan without indica- tion as per NICE guideline and the report was abnormal. Statistical methods The data from the patients were entered into a Microsoft Excel spreadsheet and the data from the Google forms were downloaded in the excel spreadsheet. The data analysis was done using Statistical Package for the Social Sciences (SPSS) version 21. For descriptive statistics, frequency (percentage) and mean with standard deviation (SD) or median with interquartile range (IQR) were used for continuous parameters whichever appli- cable. Graphical and tabular presentations were also plotted. The categorical variables were analyzed by cross-tabulation using chi-square. Due to a small sample group and asymmetri- cally distributed data, the t-test was complemented with a Mann-Whitney U-test to compare the medians of the continuous independent variables. Diagnostic tests were estimated using 22 tables. The performance of the NICE guideline was calculated using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy. The scans sent on the patient’s request were excluded for comparison of the number of CT scans performed with and without indication and guideline adherence in the pre and post-implementation peri- ods. The pre-and the post-implementation Likert scale ratings for the confidence level of the attending physicians were expressed as medians with IQR and compared with Wilcoxon signed-rank test. 5 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact PLOS ONE PLOS ONE NICE head injury guideline implementation and its impact Table 1. Demography, incident details and clinical parameters of the study population, n = 100. S. No Baseline Characteristics Pre implementation of NICE N1 = 50 Post implementation of NICE N2 = 50 p- value 1 Age Category (years) n (%) 0.334a 15 10 (20.0) 13 (26.0) 16–64 35 (70.0) 30 (60.0) > 65 5 (10.0) 7 (14.0) 2 Gender n (%) 0.296a Female 10 (20.0) 17 (34.0) 3 Triage Category n (%) 0.843a Red 22 (44.0) 6 (12.0.0) Orange 8 (16.0) 12 (24.0) Yellow 20 (40.0) 27 (54.0) Green 0 (0) 5 (10.0) 4 GCS on Arrival 0.059b Median (IQR) 15 (14.75–15.00) 15 (15.00–15.00) Mild 42 (84.0) 46 (92.0) Moderate 2 (4.0) 3 (6.0) Severe 6 (12.0) 1 (2.0) 5 Mechanism of Injury n (%) 0.379a Fall 24 (48.0) 23 (46.0) Physical Assault 10 (20.0) 7 (14.0) RTA 13 (26.0) 14 (28.0) Others 3 (6.0) 6 (12.0) 6 Mode of Transport n (%) Self 26 (52.0) 33 (66.0) 0.197a Ambulance 24 (48.0) 17 (34.0) 7 Taken pre- hospital treatment n (%) 2 (4.0) 6 (12.0) 0.594a 8 Under influence of Alcohol/ Drugs n (%) 4 (8.0) 6 (12.0) 0.441a 9 Presence of Comorbidities n (%) 2 (4.0) 2 (4.0) 0.768a 10 Disposition from the ED n (%) 0.088a Discharge 20 (40.0) 26 (52.0) Admission 14 (28.0) 5 (10.0) Referral 5 (10.0) 4 (8.0) DOR/LAMA 11 (22.0) 15 (30.0) Table 2. Performance of the NICE guideline in comparison to the clinical judgment, n = 86. Clinical judgment N = 86 NICE guideline N = 86 CT head scan not indicated CT head scan indicated CT head scan not indicated CT head scan indicated THI n (%) 0 (0) 31 (36.1) 2 (2.3) 29 (33.7) No THI n (%) 19 (22.1) 36 (41.8) 42 (48.8) 13 (15.2) NICE-National institute for health and clinical excellence; CT-Computed tomography; THI-Traumatic head injury (as indicated by CT head finding and outcome of the patient). https://doi.org/10.1371/journal.pone.0254754.t002 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 7 / 14 Table 2. Performance of the NICE guideline in comparison to the clinical judgment, n = 86. Guideline adherence If the NICE guideline were to be applied on the pre-implementation phase–study population, of the total 46 CT head scans performed, 50.0% were indicated while 50.0% were not indicated. Likewise in the post-implementation phase, of the 35 CT head scans performed, 60.0% were indicated while 40.0% were not indicated. Following exclusion of the CT head scans performed on patient request, the proportion of the CTs performed without indication decreased by 20.3 percentage points (from 42.5%-22.2%), while the proportion of indicated CT scans increased by 20.3 percentage points (from 57.5%-77.8%) which is a statistically significant improved adherence (chi-squared p = 0.001) following the implementation of the NICE guideline. (Table 5) 6 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 Chi square b-Mann-Whitney U test; NICE-National institute for health and clinical excellence; IQR-Interquartile range; GCS-Glasgow coma scale; DOR-Discharge on request; LAMA-Leave against medical advice. PLOS ONE CT performed/not performed according to Attending Physician’s Clinical Judgement Pre-implementation of NICE N1 = 44 Post-implementation of NICE N2 = 42 CT Head Scan not indicated CT head scan indicated CT Head Scan not indicated CT head scan indicated CT head not performed n = 19 n (%) 4 (100) 0 (0) 15 (100) 0 (0) CT head performed n = 67 n (%) 17 (42.5) 23 (57.5) 6 (22.2) 21 (77.8) NICE-National institute for health and clinical excellence; CT-Computed tomography. Table 5. Guideline adherence during different phases, n = 86. PLOS ONE NICE head injury guideline implementation and its impact Table 3. Diagnostic tests. CT Indication Sensitivity Specificity PPV NPV Accuracy Clinical judgement % (95% CI) 100 (88.8–100.0) 34.6 (22.2–48.6) 46.3 (41.6–51.1) 100 (82.4–100.0) 58.1 (47.0–68.7) NICE % (95% CI) 93.6 (78.6–99.2) 76.4 (62.9–86.8) 69.1 (57.9–78.4) 95.5 (84.5–98.8) 82.6 (72.9–89.9) CI-Confidence Interval; NICE-National institute for health and clinical excellence; PPV-Positive predictive value; NPV-Negative predictive value. https://doi.org/10.1371/journal.pone.0254754.t003 Table 4. Proportion of CT head scans and CT findings in the pre and post-implementation phases of the study, n = 100. Pre-implementation of NICE N1 = 50 Post-implementation of NICE N2 = 50 CT Performed n (%) 46 (92.0) 35 (70.0) Normal 31 (67.4) 19 (54.3) Abnormal 15 (32.6) 16 (45.7) Fractures 14 (93.3) 9 (52.9) With pneumocephalus 1 (6.7) 3 (17.7) Intracranial hemorrhages 10 (66.7) 10 (58.8) (EDH/ SDH/ SAH/ IPH) NICE-National institute for health and clinical excellence; CT-Computed tomography; EDH-Extradural hematoma; SDH-Subdural hematoma; SAH-Subarachnoid hemorrhage; IPH-Intraparenchymal hemorrhage. Table 3. Diagnostic tests. CT Indication Sensitivity Specificity PPV NPV Accuracy Clinical judgement % (95% CI) 100 (88.8–100.0) 34.6 (22.2–48.6) 46.3 (41.6–51.1) 100 (82.4–100.0) 58.1 (47.0–68.7) NICE % (95% CI) 93.6 (78.6–99.2) 76.4 (62.9–86.8) 69.1 (57.9–78.4) 95.5 (84.5–98.8) 82.6 (72.9–89.9) CI-Confidence Interval; NICE-National institute for health and clinical excellence; PPV-Positive predictive value; NPV-Negative predictive value. https://doi.org/10.1371/journal.pone.0254754.t003 Table 4. Proportion of CT head scans and CT findings in the pre and post-implementation phases of the study, n = 100. Pre-implementation of NICE N1 = 50 Post-implementation of NICE N2 = 50 CT Performed n (%) 46 (92.0) 35 (70.0) Normal 31 (67.4) 19 (54.3) Abnormal 15 (32.6) 16 (45.7) Fractures 14 (93.3) 9 (52.9) With pneumocephalus 1 (6.7) 3 (17.7) Intracranial hemorrhages 10 (66.7) 10 (58.8) (EDH/ SDH/ SAH/ IPH) NICE-National institute for health and clinical excellence; CT-Computed tomography; EDH-Extradural hematoma; SDH-Subdural hematoma; SAH-Subarachnoid hemorrhage; IPH-Intraparenchymal hemorrhage. https://doi.org/10.1371/journal.pone.0254754.t004 of CT head scans and CT findings in the pre and post-implementation phases of the study, n = 100. Table 5. Guideline adherence during different phases, n = 86. Evaluation of the ED attending physicians The median scores of the questionnaire were 11 in both pre-implementation (IQR: 9.0–11.5) and post-implementation (IQR: 11.0–12.0) phases. However, the score ranged from 8 to 12 and 10 to 12 in the pre and post-implementation periods respectively. The scores were com- pared using the Wilcoxon signed-rank test that showed a significant difference (p- value = 0.047) as shown in Fig 2. The level of confidence to indicate or not indicate the CT scan was evaluated using a Likert scale of 10 for each case. Table 6 shows that 9 out of 12 cases showed statistically significant dif- ferences in terms of improvement in the confidence level. In 3 case scenarios, there was no sta- tistically significant improvement. PLOS ONE Clinical judgment N = 86 NICE guideline N = 86 CT head scan not indicated CT head scan indicated CT head scan not indicated CT head scan indicated THI n (%) 0 (0) 31 (36.1) 2 (2.3) 29 (33.7) No THI n (%) 19 (22.1) 36 (41.8) 42 (48.8) 13 (15.2) NICE-National institute for health and clinical excellence; CT-Computed tomography; THI-Traumatic head injury (as indicated by CT head finding and outcome of the patient). https://doi org/10 1371/journal pone 0254754 t002 Table 2. Performance of the NICE guideline in comparison to the clinical judgment, n = 86. PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 7 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 Discussion The importance of the problem of THI is always underestimated due to the lack of research and good quality data. This is a study in Nepal to investigate the impact and adherence of a head injury guideline at DH-ED. Following the implementation of the NICE guideline in this 8 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 NICE head injury guideline implementation and its impact PLOS ONE study, there was a decreased proportion of CT head scans performed, increased guideline adherence and an increase in the confidence levels of the MOs to indicate a CT head scan in cases of THI. Fig 2. Box plot showing pre and post-implementation scores. https://doi.org/10.1371/journal.pone.0254754.g002 Table 6. Confidence of the ED attending physicians in relation to the 12 case scenarios. Cases Pre-implementation confidence Median (IQR) Post-implementation confidence Median (IQR) p-valuea 1 8 (5.0–9.5) 10 (10–10) 0.018 2 8 (4.0–8.0) 10 (9–10) 0.049 3 8 (4.5–9.0) 10 (9.5–10.0) 0.017 4 7 (4.0–7.0) 9 (7.5–10.0) 0.029 5 8 (4.5–8.5) 10 (10.0–10.0) 0.007 6 7 (5.5–8.5) 10 (9.5–10.0) 0.007 7 7 (6.0–9.0) 10 (7.0–10.0) 0.048 8 9 (6.0–10.0) 10 (9.5–10.0) 0.129 9 9 (5.5–10.0) 10 (9.5–10.0) 0.174 10 9 (5.5–10.0) 10 (9.5–10.0) 0.102 11 9 (7.0–10.0) 10 (10.0–10.0) 0.038 12 8 (4.5–9.0) 10 (10.0–10.0) 0.011 aWilcoxon signed-rank test; IQR-Interquartile range. https://doi.org/10.1371/journal.pone.0254754.t006 PLOS ONE NICE head injury guideline implementation and its impact study, there was a decreased proportion of CT head scans performed, increased guideline adherence and an increase in the confidence levels of the MOs to indicate a CT head scan in cases of THI. Fig 2. Box plot showing pre and post-implementation scores. https://doi.org/10.1371/journal.pone.0254754.g002 Table 6. Confidence of the ED attending physicians in relation to the 12 case scenarios. Cases Pre-implementation confidence Median (IQR) Post-implementation confidence Median (IQR) p-valuea 1 8 (5.0–9.5) 10 (10–10) 0.018 2 8 (4.0–8.0) 10 (9–10) 0.049 3 8 (4.5–9.0) 10 (9.5–10.0) 0.017 4 7 (4.0–7.0) 9 (7.5–10.0) 0.029 5 8 (4.5–8.5) 10 (10.0–10.0) 0.007 6 7 (5.5–8.5) 10 (9.5–10.0) 0.007 7 7 (6.0–9.0) 10 (7.0–10.0) 0.048 8 9 (6.0–10.0) 10 (9.5–10.0) 0.129 9 9 (5.5–10.0) 10 (9.5–10.0) 0.174 10 9 (5.5–10.0) 10 (9.5–10.0) 0.102 11 9 (7.0–10.0) 10 (10.0–10.0) 0.038 12 8 (4.5–9.0) 10 (10.0–10.0) 0.011 aWilcoxon signed-rank test; IQR-Interquartile range. https://doi.org/10.1371/journal.pone.0254754.t006 j y g p p PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 9 / 14 Fig 2. Box plot showing pre and post-implementation scores. aWilcoxon signed-rank test; IQR-Interquartile range. https://doi.org/10.1371/journal.pone.0254754.t006 aWilcoxon signed-rank test; IQR-Interquartile range. aWilcoxon signed-rank test; IQR-Interquartile range. Performance of the NICE guideline The CT finding was a contusion in the right frontal and parieto-occipital region with a depressed comminuted fracture in the frontal bone with orbital extensions, fractured nasal bones with overlying soft tissue swelling/hema- toma and orbital emphysema. The NICE guideline makes no recommendations for THI patients with a history of alcohol consumption in contrast to the New Orleans Criteria (NOC) which recommends head CT scans for THI patients under any drug or alcohol influence [25]. Th 12 l i h hi f f ll f h i h f b p y g [ ] The next case was 12 years, male with a history of fall from a height of about two meters, sustaining an injury on the right side of the head. Although there was no indication for CT head according to the NICE guideline in this case, when it comes to children, it is very difficult to decide. Thus the clinicians mostly interpret the criteria according to the clinical situation faced by them in the ED. Demography This study shows that the median age group sustaining head injuries was 27 years (IQR: 16.0– 42.5). The most common age group vulnerable for injury in this study was 16–65 years, fol- lowed by <16 years, 65years (65.0%, 23.0% and 12.0% respectively) which is similar to the findings of various studies [16–19]. A reason for this could be that the exposure of young peo- ple to accidents in traffic and unsafe working situations. The average life expectancy in Nepal is rather low (67.5 years) and the median age is 23.4 years which would explain the minimal number of old age people sustaining THI [20]. A significant gender disparity is noted in this study, with male (73.0%) being more prone to any injury. The preponderance of males over females is noted in all modalities of injuries in our study, as in many studies [16–23] conducted all around the world. Discussion study, there was a decreased proportion of CT head scans performed, increased guideline adherence and an increase in the confidence levels of the MOs to indicate a CT head scan in cases of THI. Table 6. Confidence of the ED attending physicians in relation to the 12 case scenarios. Table 6. Confidence of the ED attending physicians in relation to the 12 case scenarios. Cases Pre-implementation confidence Median (IQR) Post-implementation confidence Median (IQR) p-valuea 1 8 (5.0–9.5) 10 (10–10) 0.018 2 8 (4.0–8.0) 10 (9–10) 0.049 3 8 (4.5–9.0) 10 (9.5–10.0) 0.017 4 7 (4.0–7.0) 9 (7.5–10.0) 0.029 5 8 (4.5–8.5) 10 (10.0–10.0) 0.007 6 7 (5.5–8.5) 10 (9.5–10.0) 0.007 7 7 (6.0–9.0) 10 (7.0–10.0) 0.048 8 9 (6.0–10.0) 10 (9.5–10.0) 0.129 9 9 (5.5–10.0) 10 (9.5–10.0) 0.174 10 9 (5.5–10.0) 10 (9.5–10.0) 0.102 11 9 (7.0–10.0) 10 (10.0–10.0) 0.038 12 8 (4.5–9.0) 10 (10.0–10.0) 0.011 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 9 / 14 PLOS ONE NICE head injury guideline implementation and its impact Performance of the NICE guideline Compared to the traditional clinical judgment, the NICE guideline showed a sensitivity and specificity of 93.6% and 76.4% with 82.6% accuracy in this study. Had the adherence been bet- ter, the post-implementation period guideline specificity and PPV would have increased more. These values were similar to the systematic review [24] done in Italy. Lower sensitivities (82.1% and 72.5%) and specificities (46.1% and 60.9%) were shown in multicenter validation studies [8, 11] in Netherland and concluded that the lowest number of patients requiring scan for either of the outcomes was reached with the NICE criteria. In an audit in England and Wales [21], the NICE guideline resulted in 100% sensitivity and 93.8% specificity which was quite high compared to our results. In our study, there were two positive findings in the non-indicated CT scans during the pre-implementation period. The first case was a 24 years male who presented with a history of RTA sustaining injury mostly on the face. He was under the influence of alcohol and thus CT head scan was ordered as per the clinical judgment. The CT finding was a contusion in the right frontal and parieto-occipital region with a depressed comminuted fracture in the frontal bone with orbital extensions, fractured nasal bones with overlying soft tissue swelling/hema- toma and orbital emphysema. The NICE guideline makes no recommendations for THI patients with a history of alcohol consumption in contrast to the New Orleans Criteria (NOC) which recommends head CT scans for THI patients under any drug or alcohol influence [25]. The next case was 12 years, male with a history of fall from a height of about two meters, sustaining an injury on the right side of the head. Although there was no indication for CT head according to the NICE guideline in this case, when it comes to children, it is very difficult to decide. Thus the clinicians mostly interpret the criteria according to the clinical situation faced by them in the ED. In our study, there were two positive findings in the non-indicated CT scans during the pre-implementation period. The first case was a 24 years male who presented with a history of RTA sustaining injury mostly on the face. He was under the influence of alcohol and thus CT head scan was ordered as per the clinical judgment. Guideline adherence We found that the implementation of the NICE guideline was associated with improved adherence in relation to CT scan of the head from 57.5% to 77.8%. This improvement of 20.3% represents both a reduction in unnecessary CT usage and thus radiation exposure. This increment in adherence in our study could be explained by the multiple approaches: teachings during the implementation phase, posters of the guideline to help remind all the attending physicians of the indications for CT scans. We reinforced the adherence to the NICE guideline by supervising the MOs on-site in the first week of implementation. Similar findings were noted in the study conducted at a major children’s hospital ED which showed a significant increase in guideline adherence from 79.2% to 85.1% [22]. The study by Mooney et al. [26] showed compliance of 94.2% pre-implementation of NICE guideline which increased to 98.8% post-guideline implementation in adults. The studies by Harris et al. [27] and Dhungana et al. [17] also concluded that the adherence to the NICE guideline could be improved by interventions such as an educational program to the trauma team of the hospital, a repeated survey of the attending physicians and their feedback. Proportion of CT head scans performed The proportion of CT head scans performed decreased from 46 (92.0%) to 35 (70.0%) in our study. The result is contradictory in various previous studies [21]. Correspondingly there have been studies [26, 27] showing a similar decrement in the proportion of CT head scans per- formed following the implementation of the NICE guideline. A previous study [22] showed that the proportion of CT head scans performed decreased by a lesser rate than our study, from 40 (4.8%) to 23 (2.4%) and the proportion of CT head scans performed without indica- tion decreased from 14 (1.7%) to 4 (0.4%). On the contrary, the implementation of the NICE guideline in the ED of a teaching hospital and a District General Hospital led to a two to five- fold increase in the CT head scan rate [16]. 10 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact Following interventions for quality improvement to adhere to the NICE guideline in a study [27], there was a statistically significant decrease of 23.0 percentage points in the number of CT heads requested with no clear indication following intervention (p = 0.00027) which was similar to our results (20.3%, p = 0.001). This reduction in unnecessary CT head scans may result in considerable savings to numerous families while ensuring access for those really in need. In addition, improved triage of head injuries in the rural communities would help in needful referral for further imaging and earlier access to care. These changes could potentially reduce costs involved with transport, CT scanning, and also reduce the strain on emergency, neurology, and radiology departments. Evaluation of the ED attending physicians The decision to order a head CT for patients with THI is quite challenging, particularly in mild THIs. The actual degree of harm (radiation exposure, false-positive findings) or benefit (dis- covery of treatable intracranial injury) is not comprehensively known. Nonetheless, the attend- ing physicians must make this decision in almost every shift. Many clinical and non-clinical factors influence their decision to order a head CT [9]. The current research focused on their confidence level to make the decision with and without a guideline protocol. There was statisti- cally significant increased self-reported confidence after the guideline implementation. In three case scenarios, there was no statistically significant improvement in their level of confidence. The cases were a child with a dangerous mechanism of injury with four episodes of vomiting, an adult with one post-traumatic seizure and another adult with GCS <13 at pre- sentation. Since the indications for head CT were quite obvious, the MOs were confident to indicate the CT scan even before the implementation. S1 File. Pre-implementation questionnaire—Google form. (PDF) S1 File. Pre-implementation questionnaire—Google form. (PDF) S1 File. Pre-implementation questionnaire—Google form. (PDF) S2 File. Post-implementation questionnaire—Google form. (PDF) S2 File. Post-implementation questionnaire—Google form. (PDF) S3 File. Implementation of NICE guideline in ER. (PDF) S3 File. Implementation of NICE guideline in ER. (PDF) S4 File. Flowchart posters for ER. (PDF) Limitations These before and after type of studies have the strength of suggesting that the outcome is impacted by the intervention, however there are a number of limitations. Firstly, we did not have control over elements that are changing at the time of intervention such as joining of a new attending physician or start of neurosurgical department. Although, the pre and post- implementation evaluation of confidence levels of the attending physicians were done among the same set and no new physicians were included in between, this study did not take into account the likelihood of prior knowledge of the attending physicians about the NICE PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 11 / 14 PLOS ONE NICE head injury guideline implementation and its impact guideline or any other head injury guidelines prior to the intervention. Furthermore, conve- nient sampling method used in this study causes sampling bias and some sample have been missed in between. Randomized control trial (RCT) would have been a better study design but randomization was not feasible in our setting. Guideline adherence with respect to the timeline was not focused in this study and the reporting of the CT head scans done at off-hours were not available within 1 hour of the investigation. However, it was available by the first working hour of the next day. Even in patients with normal head CT results and no identified intracra- nial injuries, there may have been subtle brain injury which would have been better detected by an MRI [28] or perhaps a second CT scanning, neither of which were obtained in such patients. Conclusion The overall compliance with the NICE guideline was achieved and there was a reduction in unnecessary CT scans and thus exposure to radiation. No patient with THI was missed with the implementation of the guideline and the attending physicians felt more confident to decide on indications for CT head scans for THI cases after the implementation of the NICE guideline. g Furthermore, strategies leading to protocol-guided head CT ordering reduce the rates of CT scans and thus reduce the costs. This leaves another space for research regarding the cost- benefit as an impact of the implementation of the NICE guideline in a low-middle income country like Nepal. Acknowledgments Medical officers (Anuradha Pradhan, Sareen Shrestha, Manoj Dawadi, Suyog Shrestha, Hema Joshi, Prem Waiba, Rahul Neupane, Ambika Belbase and Vishnu Khadka) of the Department of General Practice and Emergency Medicine, DH, for assisting in data collection. ED faculties for their expert opinions and guidance. All the participants for their enthusiastic participation. Supporting information S1 File. Pre-implementation questionnaire—Google form. (PDF) S2 File. Post-implementation questionnaire—Google form. (PDF) S3 File. Implementation of NICE guideline in ER. (PDF) S4 File. Flowchart posters for ER. (PDF) S1 File. Pre-implementation questionnaire—Google form. (PDF) References 1. Head injury: assessment and early management [Internet]. National Institute for Health and Care Excel- lence (UK); 2019 [cited 2021 Jan 19]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK552670/ 2. Dewan MC, Rattani A, Gupta S, Baticulon RE, Hung Y-C, Punchak M, et al. Estimating the global inci- dence of traumatic brain injury. Journal of Neurosurgery [Internet]. 2019; 130:1080–97. Available from: http://dx.doi.org/10.3171/2017.10.jns17352 3. Puvanachandra P, Hyder AA. The burden of traumatic brain injury in Asia: A call for Research. Pak J Neurol Sci. 2009; 4:27–32. 4. Murray CJL, Lopez AD, Organization WH, Bank W, Health HS of P. The Global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020: summary [Internet]. World Health Organization; 1996 [cited 2020 Jun 22]. Avail- able from: https://apps.who.int/iris/handle/10665/41864 5. Paudel SS, Luitel R, Bista A, Baniya A, Panta DJ, Shrestha B, et al. Scenario of Head Injury Patients in Tertiary Care Hospital of Nepal. J Nepal Health Res Counc [Internet]. 2020 [cited 2020 Sep 16]; 18:112–5. Available from: http://www.jnhrc.com.np/index.php/jnhrc/article/view/2276 https://doi.org/10. 33314/jnhrc.v18i1.2276 PMID: 32335604 6. Shetty VS, Reis MN, Aulino JM, Berger KL, Broder J, Choudhri AF, et al. ACR Appropriateness Criteria Head Trauma. J Am Coll Radiol [Internet]. 2016; 13:668–79. Available from: https://doi.org/10.1016/j. jacr.2016.02.023 PMID: 27262056 7. Wintermark M, Sanelli PC, Anzai Y, Tsiouris AJ, Whitlow CT, ACR Head Injury Institute, et al. Imaging evidence and recommendations for traumatic brain injury: conventional neuroimaging techniques. J Am Coll Radiol [Internet]. 2015; 12:e1–14. Available from: https://doi.org/10.1016/j.jacr.2014.10.014 PMID: 25456317 8. Smits M, Dippel DWJ, de Haan GG, Dekker HM, Vos PE, Kool DR, et al. Minor head injury: guidelines for the use of CT—a multicenter validation study. Radiology [Internet]. 2007; 245:831–8. Available from: https://doi.org/10.1148/radiol.2452061509 PMID: 17911536 9. Probst MA, Kanzaria HK, Schriger DL. A conceptual model of emergency physician decision making for head computed tomography in mild head injury. Am J Emerg Med [Internet]. 2014; 32:645–50. Available from: https://doi.org/10.1016/j.ajem.2014.01.003 PMID: 24560384 10. DeAngelis J, Lou V, Li T, Tran H, Bremjit P, McCann M, et al. Head CT for Minor Head Injury Presenting to the Emergency Department in the Era of Choosing Wisely. Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health [Internet]. 2017 [cited 2020 Jan 20];18. Available from: https://escholarship.org/uc/item/2p02k1rf https://doi.org/10.5811/westjem.2017.6.33685 PMID: 28874933 11. Foks KA, van den Brand CL, Lingsma HF, van der Naalt J, Jacobs B, de Jong E, et al. Author Contributions Conceptualization: Pratisha Pradhan, Roshana Shrestha. Data curation: Roshana Shrestha. Formal analysis: Pratisha Pradhan, Anmol Purna Shrestha, Abha Shrestha, Roshana Shrestha. Investigation: Pratisha Pradhan, Ram Chandra Paudel. 12 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact Methodology: Pratisha Pradhan, Alok Pradhan, Anmol Purna Shrestha, Abha Shrestha, Roshana Shrestha. Software: Alok Pradhan, Abha Shrestha, Roshana Shrestha. Software: Alok Pradhan, Abha Shrestha, Roshana Shrestha. Supervision: Anmol Purna Shrestha, Roshana Shrestha. Writing – original draft: Pratisha Pradhan. Writing – original draft: Pratisha Pradhan. Writing – review & editing: Pratisha Pradhan, Alok Pradhan, Anmol Purna Shrestha, Abha Shrestha, Ram Chandra Paudel, Roshana Shrestha. References External valida- tion of computed tomography decision rules for minor head injury: prospective, multicentre cohort study in the Netherlands. BMJ [Internet]. 2018 [cited 2020 Jun 17];k3527. Available from: http://www.bmj. com/lookup/doi/10.1136/bmj.k3527 PMID: 30143521 12. Henning B, Lydersen S, Døllner H. A reliability study of the rapid emergency triage and treatment sys- tem for children. Scand J Trauma Resusc Emerg Med [Internet]. 2016 [cited 2020 Jun 20];24. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4766636/ https://doi.org/10.1186/s13049-016- 0207-6 PMID: 26911367 13. Teasdale G, Jennett B. Assessment and prognosis of coma after head injury. Acta Neurochir (Wien). 1976; 34:45–55. Available from: https://doi.org/10.1007/BF01405862 PMID: 961490 14. Sakpal TV. Sample Size Estimation in Clinical Trial. Perspect Clin Res [Internet]. 2010 [cited 2020 Nov 25]; 1:67–9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3148614/ PMID: 21829786 13 / 14 PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 PLOS ONE NICE head injury guideline implementation and its impact 15. Korevaar DA, Cohen JF, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, et al. Updating standards for reporting diagnostic accuracy: the development of STARD 2015. Res Integr Peer Rev [Internet]. 2016 [cited 2020 Sep 25]; 1:7. Available from: http://researchintegrityjournal.biomedcentral.com/ articles/10.1186/s41073-016-0014-7 PMID: 29451535 16. Hassan Z, Smith M, Littlewood S, Bouamra O, Hughes D, Biggin C, et al. Head injuries: a study evaluat- ing the impact of the NICE head injury guidelines. Emerg Med J [Internet]. 2005; 22:845–9. Available from: https://doi.org/10.1136/emj.2004.021717 PMID: 16299190 17. Dhungana S, Shrestha MK, Ghartimagar D, Ghosh A. Emergency imaging of head and cranio-facial injuries: Implementing NICE guidelines–a cross sectional analysis from western region of Nepal. ms [Internet]. 2015 [cited 2021 Jun 8]; 3:218–24. Available from: https://cmrasociety.org/journal/index.php/ ms/article/view/54 18. Onwuchekwa RC, Echem RC. An epidemiologic study of traumatic head injuries in the emergency department of a tertiary health institution. 2018 [cited 2020 Sep 16]; 20:24–9. Available from: http:// www.jmedtropics.org/article.asp?issn=2276-7096;year=2018;volume=20;issue=1;spage=24;epage= 29;aulast=Onwuchekwa 19. Khadka NK, Karmacharya BG, Jha R, Adhikary DR, Sharma GR. An Audit of Head Injury at Bir Hospital \| Nepal journal of Neuroscience \| I. Nepal journal of Neuroscience [Internet]. 2013 [cited 2020 Sep 18]; 10:68–71. Available from: https://journals.indexcopernicus.com/search/article?articleId=1776810 20. Stenholm E. Analysis of Traumatic Head Injury in Kathmandu, Nepal. 2016;14–41. 21. Shravat BP, Huseyin TS, Hynes KA. NICE guideline for the management of head injury: an audit dem- onstrating its impact on a district general hospital, with a cost analysis for England and Wales. Emerg Med J [Internet]. 2006; 23:109–13. Available from: https://doi.org/10.1136/emj.2004.022327 PMID: 16439738 22. Ramjeeawon N, Lecky F, Burke DP, Ramlakhan S. PLOS ONE \| https://doi.org/10.1371/journal.pone.0254754 July 15, 2021 References Implementing the National Institute for Health and Clinical Excellence Head Injury 2014 Guidelines in a major children’s hospital emergency department. Eur J Emerg Med [Internet]. 2019; 26:158–62. Available from: https://doi.org/10.1097/MEJ. 0000000000000512 PMID: 29112522 23. Adhikari K, Gupta MK, Pant AR, Rauniyar RK. Clinical Patterns and Computed Tomography Findings in Patients with Cranio-Cerebral Trauma in Tertiary Hospital in Eastern Nepal. Journal of Nepal Health Research Council [Internet]. 2019 [cited 2020 Sep 18]; 17:56–60. Available from: http://nepmed.nhrc. gov.np/index.php/jnhrc/article/view/494 https://doi.org/10.33314/jnhrc.1269 PMID: 31110378 24. Fabbri A, Servadei F, Marchesini G, Dente M, Iervese T, Spada M, et al. Clinical performance of NICE recommendations versus NCWFNS proposal in patients with mild head injury. J Neurotrauma [Internet]. 2005; 22:1419–27. Available from: https://doi.org/10.1089/neu.2005.22.1419 PMID: 16379580 25. Kavalci C, Aksel G, Salt O, Yilmaz MS, Demir A, Kavalci G, et al. Comparison of the Canadian CT head rule and the new orleans criteria in patients with minor head injury. World J Emerg Surg [Internet]. 2014 [cited 2021 Jun 8]; 9:31. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3997198/ https://doi.org/10.1186/1749-7922-9-31 PMID: 24742359 26. Mooney JS, Yates A, Sellar L, Shipway T, Roberts C, Parris R, et al. Emergency head injury imaging: implementing NICE 2007 in a tertiary neurosciences centre and a busy district general hospital. Emerg Med J [Internet]. 2011; 28:778–82. Available from: https://doi.org/10.1136/emj.2009.083360 PMID: 21030548 27. Harris L, Axinte L, Campbell P, Amin N. Computer Tomography (CT) for head injury: adherence to the National Institute for Health and Care Excellence (NICE) criteria. Brain Injury [Internet]. 2019 [cited 2020 Jun 17]; 33:1539–44. Available from: https://www.tandfonline.com/doi/full/10.1080/02699052. 2019.1658130 PMID: 31442088 28. Rivara FP, Kuppermann N, Ellenbogen RG. Use of Clinical Prediction Rules for Guiding Use of Com- puted Tomography in Adults With Head Trauma. JAMA [Internet]. 2015; 314:2629–31. Available from: https://doi.org/10.1001/jama.2015.17298 PMID: 26717029 14 / 14
W4237479326.txt	https://zenodo.org/records/2264092/files/article.pdf	de		Herzschallstudien	Pflügers Archiv für die gesamte Physiologie des Menschen und der Tiere/Pflügers Archiv	1,910	public-domain	20,936	509 (Aus dem poliklinischen Institut fiir innere Medizin der Universitht Berlin. Geh.-Rat Prof. Dr. Senator.) Herzsehallstudien 1). Von Heinrieh Gerhartz. # (Mit 12 Textfiguren.) Wer jemals den gutgespielten ,Militi~rmarsch" yon B o e r gehi)rt hat, kennt die heftige Vibration des Thorax bei den Paukenschli~gen. Wir haben hier den demonstrierendsten Beweis fiir die grundlegende Tatsache, dass die Schallschwingungen nach ihrer Fortleitung durch die Luft imstande sind, KSrper in Schwingungen zu versetzen, Im angef~hrten Beispiele sind diese Schwiugungen deutlich zu filhlen. Es ist nicht im mindesten zweifelhaft, dass sie auch leicht aufzuzeichnen sind; denn es gelingt, Luftvibrationen yon ausserordentlich viel geringerer Amplitude sichtbar zu machen. Der Herzschall steht nahe der Grenze dessen, was das Ohr zu hSren vermag. Wird, wie es allgemein iiblich ist, zu seiner Aufzeichnung eine Membran benutzt, so ist es notwendig, sich nach MSglichkeit klarzumachen, um welche Membrandurchbiegungen es sich hier handelt, um einen ungefi~hren Anhalt filr die Leistungen eines Herzschallschreibers zu gewinnen. Den Grenzwert des hSrbaren Schalles festzustellen, ist mehrfach durch Rechnung oder Experiment (Lord R a y 1e i g h, F r a n k e, 1) Abgeschlossen 15. Oktober~ilg09. Dieser Arbeit waren noch 48 Abbildungen zugedacht gewesen. Es musste aber der hohen Herstellungskosten wegen yon einer Reproduktion abgesehen werden. Ich bedauere, aus diesem Grunde nicht in der Lage zu sein~ iiberzeugender an getreuen Kopien der Originalkurven, die allein in Betracht kommen kSnnen, den Beweis fiir meine Angaben fiibren zu kSnnen. Ich bin gern bereit, Interessenten die Abbildungen zur Verfiigung zu stellen~ und habe zur Erleichterung im Text einen Teil derselben mit riimischen Ziffern bezeichnet. 510 Heinrich Gerhartz: C r o s s und M a n s f i e l d , T o e p l e r und B o l t z m a n n ) v e r s u c h t worden. Die Angaben der Forscher gehen trotz der verschiedenen Wege nicht allzu weit auseinander; sie schwanken zwischen 0,05 und 1,27 ~ . Wirklich einwandfreie Messungen sind kaum vorhanden. Die besten diirften die yon S h a w i) sein. Sie wurden mit einem Instrument ausgeft~hrt, yon dem sein Erfinder behauptet, class es imstande sei, wegen ungenfigender Intensiti~t nicht mehr h0rbare hmplituden noch zu messen. Aus diesem Grunde diirfte es angebracht sein, die Methode kurz anzudeuten. S h a w misst den elektriscben Strom, der den geringsten im Telephon (580 Schwingungen des Grundtons) noch hSrbaren Schall gibt. Es wird zuni~chst mit einer Mikrometerschraube. deren Exkursionen noch durch ein Reduzierhebelsystem ausserordentlich verkleinert werden, variabele Entfernung zwischen der Telephonmembran und einem verste]lbaren Kontakt hergestellt. Daraufhin misst man den Gleichstrom, der nStig ist, die Membran wieder bis zur Beriihrung des Kontaktknopfes durehzubiegen. Ich habe den Eindruck gewonnen, dass diese Methode exakt genug ist, um die angegebenen Zahlen als gent~gend zuverli~ssig erscheinen zu lassen. Die Leistungsfi~higkeit dieses Apparates geht nach Angabe S h a w's herab bis auf 0,4 ~/~. T a b e l l e 1. S haw's Skala der L a u t h e i t m i t den zugeh~rigen S c h a l l a m p l i t u d e n . Noch hSrbar . . . . . . . . . . . . . . . :Noch bequem laut . . . . . . . . . . . . . Untere Grenze des unangenehm Lauten (,,just uncomfortably loud") . . . . . . . . . . . Untere Grenze des ausserordentlich Lauten (,:just overpowering") . . . . . . . . . . . Telephonmembranamplituden amplituden Luf~- 0,7 M~ 50 ~ 0,14 pp 10 p ~ 1000 ~ 200 M/~ 5000 ~ 1000 pt~ S h a w fund filr den momentanen Schall ungefi~hr die oben genannten Werte, niimlich 0,7 f~,~ far das getibtere rechte, 0,9 /~/~ fQr 1) P. E. S h a w , The amplitude of the minimum audible impulse sound. Proceed. of the Roy. Soc. of London Ser. A. t. 76 p. 360--366. 1905. - - The improved electric micrometer. Proceed. Of the Roy. Soc. of London Ser. A. p. 350--359. 1905. Herzschallstudien. 5 11 das linke Ohr. Davon ausgehend, hat S h a w eine Skala der Lautheit aufgestellt, welche bier ebenfalls yon Interesse ist (Tabelle 1). ~eben die ftir die Membranexkursionen geltenden Zahlen sind die entsprechenden fiir die Luftvibrationen gesetzt, indem mit Lord Ra y I e i g h angenommen wurde, dass die letzteren sich zu den ersteren wie 1 : 5 verhalten. In Reihe 1 der Tabelle ist das i~usserste Minimum: so wie es unter den gtinstigsten Bedingungen far ein scharf horchendes Ohr gefunden wird; denn die Versuche wurden in einem vollkommen ruhigen unterirdischen GewSlbe nachts zwischen 12 und 4 Uhr ausgefiihrt. Die zweite Reihe bezieht sich auf Schall, der an der Grenze der HSrbarkeit far ein Ohr, das nicht auf ihn horcht, steht. Man kann also sagen, dass es das ,,physiologische" Minimum ist. Anforderungen an die Empfindlichkeit eines Herzschallschreibers. Das physiologische Minimum dtirfte schon tiber das Ausserste dessen hinausgehen, was ein Herzschallapparat leisten muss. Halten wir aber der Einfachheit wegen an dem oben stehenden Werte yon 1 0 / ~ lest. Mit dieser Angabe ist jedoch die Empfindlichkeit nicht eindeutig bestimmt, da die Schwingungszahl unberticksichtigt geblieben ist. Es wird angenommenl), dass die Intensiti~t eines 1) M. W i e n (l~ber Telephonplatten mit hohen EigentSnen. Ann. d. Physik Bd. 18 S. 1049--1050. 1905. u auch E. W i e r s c h , Ann. d. Physik Bd. 17. S. 999. 1905) schreibt z. B. an einer Stelle, dass es fiir die Ubertragung der Sprache vorteilhafter ist, die iiblichen Telephonplatten mit etwa 700 Schwingungen durch solche yon 7000 Sehwingungen zu ersetzen, weil die letzteren den Schwingungszahlen der Sprache, hauptsachlieh der Zischlaute; nigher li~gen. Dabei, sehreibt er, wird die Intensiti~t ,fiir tiefe TSne unter sonst gleichen Urnsti~nden 10 000 real kleiner als bei den gebrauchlichen Telephonen. ~immt man den Eigenton noch eine Oktave hSher, so sinkt die Intensitat auf 1/16oooo!~' Wie man sieht, ist bier vorausgesetzt, dass die Intensit~t im Quadrat des Verhaltnis~,es der Schwingungsfrequenzen abnimmt. In dieser Weise taxiert auch W. E i n t h o v e n (Ein dritter Herzton. P f l t i g e r ' s Arch. Bd. 120. S. 33--34. 1907) die Intensiti~t der TOne nach der geachriebenen Amplitude. E i n t h o v e n land bei dem 1907 publizierten Falle fiir die beiden ersten TSne eine Amplitude 14 mm, fiir den dritten 2 ram, also ein Amplitudenverhi~ltnis 7; die Schwingungszahl war bei den beiden ersten TSnen ~ 100, bei dem dritten ~ 50; hier war also das Verhi~ltnis ~ [2. ~ennen wir das erstere q, das letztere b, so ist ~" ~ a ~ ' b u ~ 72-2 ~ ~ 196. , D e r dritte Ton ist also noch ungefi~hr 200mal schwiicher als der erste oder der zweite." 512 Heinrich G e r h a r t z : Schalles dem Quadrat der Schwingungszahl und dem Quadrat der Amplitude proportional ist. Nun gelten die oben genannten S h a w ' s c h e n Zahlen far eine Telephonmembran und die Sprache. Der Mittelwert der Schwingungszahlen der menschlichen Stimme soil bei etwa 500--5000 Schwingungen liegen ( W i e n , l.c.). Rechnea wir einmal rund ftir die Herztiine mit 70, fiir die Sprache mit 700 Schwingungen, so ergibt die Rechnung mit gleichgesetztem Intensiti~tsminimum und einer Amplitude yon 10 gtt (nach S h a w ) Fig. 1. Schall- und Spitzenstosskurveyon Aorten- und Mitralinsuffizienz.(Schema.) fflr die Sprache, dass die Minimalamplitude der Herzti)ne 1,0 ~ ist, 500lathe VergrSsserung wtirde also noch deutlich sicht- und analysierbare Exkursionen (0,5 ram) ergeben. Handelte es sich dagegen um TSne der doppelten Sehwingungszahl (140), so wfire eine 2000lathe VergrSsserung erfbrderlich, um gleichgrosse Exkursionen zu erhalten. Die Herzgeri~uschzacken wiesen in einer mit meinem Registrierapparat aufgenommenen Mitralinsuffizienz-Sehallkurve (Fig. 1) eine Amplitude von etwa 0~5 ram, auf. Die Frequenz betrug 77 Doppelschwingungen pro Sekunde. Da der Apparatl), mit dem diese Kurve 1) H. G e r h a r t z ~ Die Aufzeichnung von Schallerscheinungen, insbesondere die des Herzschalles. Zeitschr. f. exper. Pathol. u. Ther. Bd. 5 S. 5 ft. 1908. 513 Herzschallstudien. aufgenommen wurde, 400 fach vergr0sserte, waren die Amplituden 0,5 ~ der Membran 40~ ~--- 0,00125 mm gross. 'Dies e , Zahl gibt einen weiteren Anhalt daftir, dass man es in diesen. Za~ken tatsiichlich mit dem Abbild eines Herzgeri~usches zu tun hat. ~' In einer Gesangkurve eines Bassisten, der alas grosse e (----- 80 Schwingungen) mit mittlerer Kraft auf den Vokal A sang (Nr. I), Fig. 2. Tonleiter (Orgel). Amplituden bei konstante~ Intensit~t. waren die Amplituden 4,25 2 - - 2,125 mm gross, was einer Membran2,125 durchbiegung yon 400 - - 0~005 mm entspricht. In diesem Zusammenhange dtirften folgende eigene Erfahrungen tiber die Beziehungen yon Schwingungszahl zu Amplitude interessieren. Von den TSnen einer Kirchenorgel wurden sowohl yon einer Pfeife E. Pflfiger, Archiv ffir Physiologie. Bd. 131. ~5 5!4 Heinrich Gerhartz: (Posaune) .wie~ yore gekoppelten Register bei gleichem Anblasestrom Tiine versehiedener Sehwingungszahl gesehriebenl). Fig. 2 (Nr. II) zeigt ein~Seh~ema der Amplituden. Man gewahrt, wie mit steigender Tonleiter d.ie: Amplitude progressiv abnimmt. Versueht man am Klavier den Versuch zu wiederholen, so misslingt er. Man erh/ilt dann eine u - n r e g e l m / i s s i g verlaufende Skala, aueh wenn man sieh bemtiht, mit ~mSglichst gleicher~ Intensiti~t die Tonleiter zu spielen. Dieser Versuch ist ein instruktives Beispiel far die relative Unf/~higkeit unseres GehSrorganes~ geringe Intensit/~tsunterschiede wahrzunehmen. Dass nichts anderes bier im Spiele ist, geht daraus hervoL dass es beim Klavier! nieht gelang, bei Wiederholung der Versuche Kurven gleicher hmi)litude zu erhalten. E s wi~re interessant, diese Versuche unter Behutzung eines Klavierspiel-Vorsatzapparates zu I wiederholen. Die oben zugrunde gelegten Anschauungen sind allgemein verbreitet, kSnnen aber nur einen gewissen Anhalt bieten, da die Grundlagen, auf die sich die Reehnung aufbaut, wohl zu unsicher sind. Sie gentigen jedoeh als Unterlage for die Konstruktion eines Herzschallsehreibers. Wie man sieht, sind die hnforderungen, die an die VergrSsserungsleistung gestellt werden, techniseh sehr hohe. St6rungen beim Herzschallschreiben. Bevori (lie einzelnen Methoden an der Hand ihrer Ero'ebnisse zur Spracl~e kommed, ist es wfinschenswert, einige etwaige Fehlerquellen allgemeiner Natur zu besprechen. E i n t h o v e n, dem wir die ersten Erfolge auf unserem Gebiete verdanken, hatte hauptsiiehlich aueh darauf Bedaeht genommen, Ersehiitterungen, die etwa durch das Schallzuleitungsrohr verursacht werden k6nnten, zu vermeiden. Er hat sich davon iiberzeugt, dass dies mSglieh ist. Ieh muss ibm darin vollkommen beistimmen, dass eine einigermaassen dickwandige kurze Gummisehlauehverbindung, auch wenn ~ie ~ hin und her bewegt wird. nicht stOrt. Wird sie seitlieh wenig komprimiert, so treten allerdings Naehteile auf. Es werden da~n Membranschwingungen beobachtet, die langsam an1) Durch diesen Versuch wird gleichzeitig der Beweis geliefert7 dass mein Registrierapp'arat auf Schatl der S'chwingungszahlen im Bereiche der HerzschallhShe (am Instrument bes'timmt) koTrekt, d. h. mit charakteris~ischen Figuren antwortet. , Herzschallstadien. 515 steigen und wieder abfallen, Der Versuch zeigt, wie schwer eine Membran, die vor dem einen Ende der Zuleitung angebracht ist, bei seitlicher Kompression der letzteren reagiert gegenfiber elner Luftpulsation in der Richtung der Achse der Zuleitungsri~hre. Die bei dem Versuch entstehenden Schallschwingungen sind so ausserorden~lich klein, dass sie ausser acht gelassen werden k~nnen. Noch leichter zu erkennen a l s die Kompressionsbewegungen sind die z. B. durch den Spitzenstoss in einer li~ngeren elastischen Schlauchverbindung verursachten Bewegungen. Dutch starkwandige Zuleitungsri)hren werden die besprochenen Feblerquellen geniigend vermieden. Eine grSssere Fehlerquelle, deren Beseitigung Schwierigkeiten machen kann, ist die Obertragung konstanter oder periodischer Vibrationen yon Teilen des Registrierapparates selbst. In erster Reilie kommt hier das Triebwerk zur Bewegung des Films in Betracht! Die Vibrationen des Uh~'werks lassen nicht zu, dass dieses auf demselben Tisch befestigt wird, aaf dem sich die tibrige Aufnahmeapparatur befindet. Am besten werden die Schwierigkeiten so gehoben, dass die Filmachse dutch eine Cardani'sche Gelenkverbindung mit tier Uhrwerksachse gekuppelt und das Uhrwerk auf einem anderen freistehenden Tische, von .dem aus keine Erschfitterungen zu dem eigentlichen Apparattische dutch den Boden fibertragen werden kSnnen, aufgestellt wird. Dass diese Anordnung, die auch den mitunter recht ffihlbaren Vorteil der freien Beweglichkeit des Uhrwerktisches besitzt, sehr zweckmassig ist, habe ich durch sehr zahlreiche Aufnahmen, die vor jeder Herzschallregistrierung oder gleichzeitig mit ihr mittelst der zweiten Aufnahmeapparatur vorgenommen wurden, erwiesen, so dass ich die volle Garantie daffir, dass in dieser Hinsicht keine Fehler in meine Kurven sich eingeschlichen haben, fibernehmen kann. Ich glaube auch nicht, dass bisher irgendein anderer, der fiber Herzschallregistrierung berichtet hat, der bier vorliegenden Schwierigkeiten nicht He~r geworden ist; es ergibt sich wenigstens aus den verSffentlichten Kurven fiir eine solche Anschauung kein hnhaltspunkt. Dagegen spricht manche Erscheinung daftir, dass in der Regel nicht die Luftamplituden des Herzschalles geschrieben wurden, sondern a u f d i e M e m b r a n f o r t g e p f l a n z t e Vibrationen r die Luftschallschwingungen erzeugenden Teile. Wenn ein Beispiel das, was ich sagen Will, l~esser,zu erlautern imstande ist~ mSchte ich an die Stimmgabel erinnern. Sie macht, 35~.. 516 Heinrich Gerhartz: wenn sie angesehlagen wird, eine hnzahl Vibrationen, die ihrer Schwingungszahl entsprechen. Verbindet man die Gabeln mit einem Scbreibstift, so zeichnet bei geeigneter bekannter Vorrichtung der Stilt diese Gabelvibrationen ohne Schwierigkeit auf. Das ist das Nteste primitive Modell eines Schallschreibers. Die Luftamplituden haben hier auf die Bewegungen des Schreibstiftes keinen Einfiuss; denn sie sind zu schwach, um ihn in Bewegung zu setzen. Warden sie es tun~ so wiirde man Oszillationen erhalten, die~ obwohl an Amplitude geringer, durchaus Frequenz und Charakter der Stimmgabelschwingungen besassen. So ist es auch verschieden, ob man die beim Sprechen oder Singen erfolgenden Vibrationen des Kehlkopfes scbreibt oder ihren Effekt, Sprache und Gesang. Das letztere ist ausserordentlich viel sehwieriger. Dieser letztere Modus ist abet beim Herzseball nicht t~berfii]ssig; denn bier suchen wir aus der Bewegung des Organs die dem Scball entsprechenden Vibrationen zu isolieren, da dem H e r z s c h a l l der wesentliehe Weft zukommt, abet Oszillationen ganz verscbiedener Frequenz in der Summe der Organvibration untergebracht sind. Gelingt es nun, sie zu analysieren, so steht nichts dem im Wege, sieh mit tier Erf•llung des techniscb einfacheren Problems zu begniigen. Diese Analyse ist zuni~ebst dutch die 0 f f n u n g des geschlossenen lufterffillten Systemes, das das Zuleitungsrohr bis zur Aufnabmemembraa umfasst, versucht worden, wie E i n t h o v e n und W e i s s annebmen, mit Erfolg. Ich sehe meine Aufgabe zuniiehst darin, die Berechtigung zu dieser Auffassung einer Kritik zu unterziehen, wobei ich reich auf meine eigenen Experimente, die ich zur Erledigung der far die Beurteilung des bis jetzt vorliegenden Materiales wichtigsten Frage vorgenommen habe, stiitze. Besprechung der bisher verifffentlichten Herzschallkurven. Bedient man sich zur Aufnahme der Pulsationen, die durch die Herzarbeit der Herzflliehe der Brust mitgeteilt werden, eines Zuleitungssystemes, wie es E i n t h 0 v e n wiederholt beschrieb 2), d. h. eines gebogenen Rohres, das an der konvexen Seite der Biegung 1) W. Einthoven und M. A. J. Geluk~ Die Registrierung der HerztSne. Pfltiger's Arch. Bd. 57 S. 617. 1894. -- W. Einthoven, Die Registrierung der menschlichen HerztOne mittels des Saitengalvanometers. Pfltiger's Arch. Bd. 117 S. 461--472. 1907. -- Vgl. auch H. Gerhartz, 1. c. S. 512. Herzschallstudien. 517 einen mit einem Hahn versehenen Rohransatz tragt, so werden bei fester Verbindung zwischen Schallaufnahmeapparat und Brustwand die Vibrationen der letzteren der Aufnahmemembran je nach dem Grade ihrer Intensitat und der Empfind!ichkeit des Schreibapparates mehr 0der weniger intensiv ~bertragen. Man erhalt so im wesentlichen die Spitzenstosskurve. Die Stossbewegung des Herzens kann im Experiment ad libitum (lurch Impulse gemischter Frequenz ersetzt werden. Wird allmhhlich der Hahn des Seitenrohres geSffnet~ so sinkt die Amplitude der Oszillationen in demselben Maasse, wie das ,,geschlossene System" (lurch das ,offene" ersetzt wird (Nr. III). Die letztgewonnene Kurve stellt dann die entsprechende reduzierte Kurve derselben Vibration dar. Man erkennt in ihr die wesentlichen Linienz~ge wieder. Beide Kurven sind natt~r]ieh nicht vollkommen analog, da die geschriebenen Exkursionen etwas variierten Trotzdem ist die Charakteristik der Kurve dieselbe; die sekundare Kurve unterscheidet sich yon der primaren eben nur durch ibre Amplitude, so dass die in der prim~tren Kurve an und f~r sich schon kleinen Amplituden in der anderen nicht mehr zu erkennen sind. Es fragt sich, ob sich in den als Schallkurven publizierten Kurven Anhaltspunkte finden .lassen, welche fi~r die beschriebene Entstehungsart sprechen. Das ist ia der Tat der Fall. E i n t h o v e n bildet in Fig. 10 der auf S. 51(~ zweitzitierten Arbeit eine Kurve ab, yon d e r e r (S. 470) schreibt, dass ihre Erklarung ihm ~rosse Schwierigkeiten dargeboten hat. Die interessante Kurve stammt von einem Menschen mit Mitralinsuffizienz. Sie muss also das systolische Ger~usch dieses Herzfehlers wiedergeben. Die Figur aber (S. 470) ,zeigt for jede Herzperiode ausser der Pause ft~nf deutlich voneinander zu unterscheidende SchSlle" . . . . ,Welche Bedeuturl~' jeder dieser Sch~lle hat, und in welcher Reihenfolge sie in der Periode vorkommen, k/)nnen wir nicht mit Gewissheit entscheiden"... ,,F~r die Deutung... wagen wir vorltiufig die folgende Hypothese : a ~ prasystolisches Ger/~usch, .~, so, s3 ~ drei Phasen des systolisehen GerSusches, 2 ~ diastolischer Ton, p ~ Pause, w/~hrend dann die in tier Figur angebrachte Klammer die ganze Periode begrenzen mag. Eine der Schwierigkeiten, die Figur zu deuten, liegt in der Dauer des schallfreien Stadiums jeder Herzperiode. Dieses Stadium seheint bei der Untersuchung mittels des Stethoskopes ziemlich lange zu dauern, w~hrend es im Photogramm nur ausserst kurz zum Vorschein kommt. WOnscht man die mangelhafte t)bereinstimmung 513 Heinlich G e r h a r t z : zwischen den beiden Beobachtungsmethoden den Fehlern des Registrierens zuzuschreiben, so m~sste man annehmen, class ein zufi~lliger Schall die schreibende Saite gerade im schallfreien Stadium tier Herzwirkung in Schwingung versetzt hatte. Aber diese Annahme muss als sehr unwahrscheinlich verworfen werden, Weft ja die mehr als sieben nacheinanderfolgenden Herzperioden des Photogrammes an Gleichf~rmigkeit nut wenig zu wansehen t~brig lassen". Einthoven ist, wie so oft, geneigt, den Miingeln des Ohres schuld zu geben und seiner Registriermethode die Ste]lung einer oberen Instanz zuzumessen. Ich teile die optimistische Anschauung, dass ,nur darum so regelmhssi~ eine Pause gehbrt wird, weft d~s Ohr unmittelbar vorher dutch den ziemlieh starken Schatl des diastolischen Tones getroffen wird, wodurch es temporar ermfidet wird", durchaus nicht, schon weft meine nun ftinfj~hrige Erfahrung in der Registrierung yon Schallerseheinungen reich zu oft yon der Superioritat unseres GehSrorganes t~berzeugt hat. Ich sehe in der yon den iibrigen Darste]lungen abweichenden Kurve lediglich eine durch das Zwischentreten des Spitzenstosses modifizierte Schallkurve und sehe demgemass das systolisehe Ger~tusch im Abschnitte sl und s~ in einer Form, wie sie den anderweitigen Bildern der E i n t h o v e n ' s e h e n Methodik entspricht, a, s 8 und p sind die Spitzenstossanteile, 2 ist der diastolische Ton in Ubereinstimmung mit E i n t h o v e n . Die Ausrechnung, die durch die ausserordentlich bequeme Darstellung der E i n t h o v e n ' s c h e n Figuren ermSglicht ist, gibt mir in dieser Auffassung recht; denn die Analyse der Kurve ergibt folgendes: ,~ ~ - 3 Impulse ~ 5,5 9 0,02" ~- 0,11 " ~ 28 Sehwingungen pro Sek. sl ~ 18 ,, -~- 7,4 9 0,02" == 0,148" = 122 . . . . . . ss -~-- 5 ,, ~ 3,6 9 0,02" --~ 0,072" z 69 ,, .... .s~ ---:- 2 ,, ~- 5 , 0 . 0 , 0 2 " ~ 0,10 " ~ 20 ,, .... . A (12" " 2 3 ,, -~/2,5 v,~" ~ 0 , 0 5 ~ 60 ,, ,, ,, Da man sich durch die Auskultation und Vergleich des Geh0rten am Harmonium leicht aberzeugen kann, dass die HerztSne selten unter 50 Schwingungen pro Sekunde liegen, darfen Schwingungen von einer Frequenz unter 30 pro Sekunde wohl k~um als Anteile eines Herztones oder -gerausches angesprochen werden. Zu dem kommt, class sich unter Eliminierung tier frequenten Sch:wingungen durch Mittelung der Exkursionen eine Kurve Herzschallstudiem 519 rekonstruieren li~sst, die einer SpitzenstosskurVe' in ih,ren~charakteristischen ~Ziigen durchaus geniigend i~hnlich sieht; ~m' :die~igei~etisehe BeXiehung vollkommen ptausibel erscheinen Zu'laSsen; E s :scheint mir in dieser Beziehung auch nicht iiberflgssig'zu sein, noch anzuftihren, dass E i n t h o v e n ausdrilcklich eine ,starke Hypertrophie tier beiden Herzh~ilften" erwi~hnt (S. 470) - - eine Angabe durchaus in unserem Sinne. , "! Uber die Zuleitungsmethodik yon W e is s 1) kann man sich nach seinen Angaben kaum ein ausreichendes Bild machen. ~I,ch finde auf S. 349 der ausfuhrlichen Publikation die Angabe: ,,~ach dem Gesagten ist es natiirlich, class die eigenen Versuche darauf ausgingen, die Registrierung der Herzt0ne vorzunehmen, ohne dass eine feste Verbindung des Registrierapparates auf tier BrUstwand besteht." Das ,,Gesagte" bezieht sich erstens auf eine Bemerkung Ew al d's 2), die akzeptiert wird, dass Gefahren da sintl, die ,,in der mechanischen ErschiRterung des Apparates durch die Herzbewegung:und hierdurch erzeugten Eigenschwingungen liegen" (S. 348), ferner auf dem eventuellen Yorhandensein yon Fehlerquellen, die dadurch gegeben seien, dass 1. Volumpulse der Brustmusku]atur, 2. im Tempo des Herzschlages der haltenden Person erfolgende Bewegungen stSrend einwirken. Es ist mir aus der Literatur nichts dartlber bekannt geworden, dass tatsi~ehlich solche Fehler gemacht worden sind. Ich habe versucht, sie kilnstlich zu erzeugen - - es ist mir hie gelungen. W e i s s zitiert die Bemerkung y o n Hotow:in:ski'sS), dass in demselben Momente, in dem dig HerztSne laut werden, Pulsationen existieren, die der Spannung der Klappen ihre Entstehung verdanken. Haben solche Pulse dieselbe Frequenz wie die HerztSne, so sind sie ihr Ausdruck. Die Pulsationen, yon denen y o n H o l o w i n s k i redet, besassen aber eine sehr niedrige Frequenz, so niedrig, dass sie ,,directement insensibles ~ l'ouie h cause de leur petite fr6quence" waren. Da Telephone und Mikrophone einen Eigenton yon in tier Regel mehr als 500 Schwingungen pro Sekunde haben; kann yon 1) O. Weiss and G. Joachim, Registrierung und Reproduktion menschlicher Herzt6ne und Herzgeri~usche. Pflfiger's Arch. Bd. 123. S. 841--386. 1908. 2) R. Ewald, Referat fiber K. Hfirthle's Arbeit: Uber die Erkl~rung des Kardiogramms usw. Centralbl. f. Physiol. Bd. 7. S. 52r--53. i893. 3) A. de Holowinski, Sur la photographie ale's bruits'rib cceur. Arch. de physiol, norm. et path. 1896 p. 893--897. 520 Heinrich C~erhartz: Eigenschwingungen bier keine Rede sein, sondern die ,,secousses m~caniques" sind Spitzenstossanteile, die zu den HerztOnen unmOglich in der Art in Beziehung stehen kiinnen, dass die Vibrationen identisch sind mit den die TOne erzeugenden, wie v o n H o 1 o w i n s k i annimmt denn solche haben die gleiche Frequenz. Weiss aussert sich hierzu auf S. 363 seiner Arbeit in P f l i i g e r ' s Archiv Bd. 123 folgendermassen: ,,Vermutlich sind die langsamen Schwingungen (sc. seiner Kurve 14) die secousses de petite fr4quence de H o l o w i n k i ' s , die nach diesem Autor den Herztiinen isocbroli sein sollen. Das sie es wirklich sind, zeigen meine Kurven. Vielleicht ist auch das~ was E i n t h o v e n als Herzti~ne registriert hat, identisch mit diesen langsamen Schwingungen. Das kSnnte wohl sein; denn die Schwingungen seiner Kurven sind sehr gering." Da Schall yon etwa 16 Schwingungen an als sol,her vom Ohr perzipiert wird, miissen die Impulse, die v. H o l o w i n s k i beschrieben hat, als dumpfer Stoss empfunden worden sein und unter der Schallschwelle gelegen haben; denn er schreibt doch als Cbarakteristikum ,,insensibles h l'ouie h cause de leur petite fr~quence". Die Schwingungszahlen E i n t h o v e n ' s liegen weir oberhalb der Schwelle (vgl. weiter unten S. 531), kOnnen also nicht mit dem identisch sein, was v. H o l o w i n s k i beschreibt. W~ts es tats~ichlicb gewesen i,~t, ist schwer zu sagen. Ich vermute, dass es die beiden grossen Erhebungen in den prim~ren Spitzenstossschwingungea sind: die ich weiter unten (S. 536) im Anscblusse an die Erwahnung der F r a n k ' s c h e n Scballkurve bespreche. Mag dem aber sein, wie ihm wolle, jedenfalls hat W e i s s darin recht, dass er die verschiedenfrequenten Pulsationen differenzieren will, Die ()ffnung des Systemes nach dem Vorbilde E i n t h o v e n ' s ist hierzu das eine Mittel, und dessert bat sich auch W e i s s bedient. ,,Es ist wichtig, dass der Trichter (Kautschuk) luftdicht an die Brustwand anschliesst." Die Zuleitung ist also vor der Membran often. Diese Methode hat den Vorteil, die Scballwellen von einer griisseren Aufnahmefli~che aufzufangen und sie konzentriert der kleinen Membran zuzufiihren. Im wesentlichen liegen also.die Dinge vor, wie sie oben scbon far das Einthoven'sche Verfahren auseinandergesetzt wurden. Ich finde auch in den Kurven, die W e i s s mitgibt, Anhaltspunkte dafilr in den Kardiophonogrammen 12, 13, 18, 25, 27 und 29. Die Beurteilung ist hier ausserordentlich erschwert dutch die grosse Ausziehung der Herzschallstudien. 521 Kurvenl). Es ist leicht zu verstehen, wie dieser Faktor wirkt. Bei schwacher Pulsation des Herzstosses kann schon die einfache Ausziehung der langsamen Impulse sie verschwinden machen. Es dtirfte sich also in den Weiss'schen Kurven lediglich um R e d u k t i o n des Spitzenstosses handeln nach dem Muster, wie Ei n t h o v e n es zuerst benutzt hat - - im Prinzip ist es g]eichgiiltig, ob man einen Hahn 5ffnet oder dadurch das Loch macht, dass man alas Zuleitungsrohr etwas yon dem Apparat abhalt. Ein zweiter Weg, ausser der Analysierung der gemischten Kurven, die Eliminierung des Spitzenstosses zu erreichen, ist die Erschwerung der Impulsleitung. Ich habe Kurven (Nr. IV und V) in der Weise aufgenommen, dass zwischen Brustwand und Telephon eine Lage Tuch gelegt wurde. Die Registrierung geschah mit einem E d e 1m a n n'schen Saitengalvanometer, in dessen Stromkreis ein Kondensator eingeschaltet worden war, damit tier Faden durch die konstanten Str0me nicht abgelenkt und eine bequemere Einstellung des Fadens in die Mitre des Films ermSglicht wurde. Die Aufnahme des Mitralstenosengeri~usches bei einer 31ji~hrigen Frau wurde yon der Stelle des Spitzenstosses aus vorgenommen. Infolgedessen sieht man in der Kurve die Wellenform des Kardiogrammes (Nr. VI) angedeutet, da die Telephonmembran gri~sserer Durehbiegungen nicht ffthig ist. Mitunter sieht man die Exkursionen niedriger Frequenz vollkommen eliminiert~ obwohl die Geri~uschoszillationen noch wiederzufinden, allerdin~s auch infolge der sehlecht leitenden Zwischenschicht erheblich reduziert sind. Man kann sich davon dutch die Ausmessung der Oszillationen leicht iiberzeugen. Ieh besitze yon .denselben Kranken Spitzenstosskurven, auf die die Schalloszillationen superponiert sind. In den bei offenem System geschriebenen Kurven (Nr. VII) sieht man, class auch hier die tats~ehlich .dureh die Sehwingungszahl als Sehalloszillationen charakterisierten Schwingungen getreu in recht gut erkenn- und studierbarer Weise wiedergegeben sind, zugleich aber auch, dass es mi)glich ist, sie yon den Exkursionen des Spitzenstosses zu eliminieren und gesondert darzustellen, dass aber in dieser Methode hier kein besonderer u erkannt werden kann~ denn an der Kardiogrammschallkurve (Nr. VIII) sind die schnellen Oszillationen ebenso leicht auszuzahlen als in den Kurven, die in Horizontallinie geschrieben wurden. 1) 1. c. Anm. 1) auf S. 519. -- S. 353: ,,DieGeschwindigkeitder Bewegung l~isst sich dur:q~ geeignete Einrichtungen auf 25 cm/sec-I regulieren." 522 Heinrich Gerhartz: In den schematischen Fig. 1, 3 und 4 (Nr. IX, X und XI), die nach dem Gesagten leicht verstiindlich sind, bringe ich einige weitere Beispiele: Meine Kurven lehren unter anderm, dass die Reaktionsfahigkeit yon Schallmembranen sich Schallschwingungen nicht so wie grossen Impulsen gegeniiber verhi~lt; denn aueh die grbssten Schallintensiti~ten maehen nut relativ winzige Membranexkursionen, d. h. eine Membran kann auf Schall jeder Intensitat korrekt reagieren und dennoch grobe Impulse falsch bzw. iiberhaupt nicht wiedergeben. Es scheint mir, class die Dicke der Membran beziiglich der Reaktionsfahigkeit auf Fig. 3. Mitr~lstenosen-Schall- und -Spitzenstosskurve. (Schema.) grobe Impulse eine grSssere Rolle spielt als hinsichtlich der Reaktionsleistung auf Schallimpulse. In letzterem Falle treten die an der Reagierfhhigkeit noch beteiligten Faktoren, wie Elastizit~it und Schwingungsfreiheit, Steifigkeit usw., durchaus in den Vordergrund. So erklart es sich, dass eine Telephonmembran dutch wenig frequente StSsse fast nicht durchgebogen werden kann, obwohl eine dtinne Schallmembran auch in dieser Hinsicht charakteristische Formen zum Ausdrucke bringt. Ftir beide Zwecke bestehen verschiedene Optima der Membranbeschaffenheit. Aus dem Gesagtea folgt, dass es bei der Aufzeichnung des Schalles nicht n ur darauf ankommt, die Masse der Membran zu reduzieren, sondern wesentlich auch auf Herzschallstudiea. 523 andere Dinge. Trotzdem besteht kein Zweifel, dass das Gewicht und demnach auch die Tri~gheit der Membran bzw. des ganzen schwingenden Systems um so geringer sein muss, je geringer die Intensiti~t des aufzunehmenden Schalles ist. Dieses Gesetz habe ich bei Versuchen mit meinem Apparat immer besti~tigt gefunden, indem ich Membranen .verschiedenster Beschaffenheit unter Beobachtung aller t~brigen Verhi~ltnisse durchprafte. So sprach z. B. eine relativ dicke Kollodiummembran auf Schall schon sehr gut an, wenn mit einer diinnen Holzmembran noeh keine siehtbaren Exkursionen beobachtet werden konnten. ~ f ? Fig. 4. Aorteninsuffizienz-Schall- und -Spitzenstosskurve. (Schema.) Membranen geringer Masse haben, was ihre Verwendung ft~r die Aufzeichnung des Herzschalles angeht, daneben noch den Vorzug, die Schallschrift mit der Aufzeichnung der Pulse kombinieren zu k6nnen. Ich halte das yon vornherein far eine zweckmiissige Kombination, weil sie die Lage tier akustischen Schwingungen, die sich. auf die Pulsationen superponieren, in der exaktesten Weise festlegen l~sst. E~ fragt sich, ob aus dieser Interferenz keine Fi~lschungen entstehen. In der Tat ist das mSglich, und bei nahbenachbarter Frequenz beider Impulse kann es unmi~glich sein, die Kurve zu differenzieren. Hier weiss man dann nicht, welche Welle dem eineD, welche dem 524 Heinrich Gerhartz: anderen Impuls angehi~rt. Es kann auch bei einem gewissen Grade tier Eliminierung der groben Pulsationsbewegungen unm(iglich werden, die differenten Wellen zu identifizieren. In solchen Fallen ]iisst sich eventuell Aufschluss dadurch gewinnen, dass man progressiv die eine Bewegung dutch 0ffnung des Luftsystems zum Schwinden bringt und die Wellen quantitativ verfolgt. V(illig e i n w a n d f r e i i s t a b e r a l l e i n , w i e ich s c h o n frt~her n a c h d r i i c k l i c h b e t o n t h a b e , d a s V e r f a b r e n , d i e grobschlitgioen Pulsationsbewegungen dadurch aufz u h a l t e n , dass m a n s i e e i n e o ' r o s s m a s s i g e M e m b r a n im o b e n g e n a n n t e n S i n n e p a s s i e r e n l a s s t . In gfinstigen F~llen reicht die Luftamplitude der Schallwellen tiir diesen Modus aus. Es ist mir bisher nur gelungen, von der F1clstersprache (30 cm Entfernung), yon leisest angesehlagenen Klaviertbnen und besoMers piano gespielten Orgelt0nen (3 m Entfernung) Aufnahmen durch eine 1 cm dicke Holzwand hindarch herzustellen; yore Herzschall besitze ich bisher, ausgenommen die Aufnahme des Herzschalles einer Mitralinsuffizienz, keine vtillio einwandfreie Kurve, die lediglich auf diese Weise erhaltene Luftamplituden des Schalles wiedergibt. Es mag alas zum Teil daran liegen, dass ich keine Menschen mit Herzt0nen sehr grosser Intensitiit zuj diesen Aufnahmen babe erhalten k0nnen und die Vergrbsserung des bisher benutzten Registrierapparates noch nicht far die anderen Fi~lle ausreieht. In dem erwi~hnten einen Falle (Mitralinsuffizienz) ist es mir meines Erachtens einmal gelungen, das systolisehe Gerauseh durcb eine 4 mm dieke Tannenbolzmembran, die in die Mitre eines 17 cm langen, glattwandigen, ko~ischen Ahornholztrichters eingeleimt war, hindurch aufzunehmen. Wie die Auskultation erg~b, nahm dieser Trichter den Herzschall korrekt auf. In diesem F a l l e - meines Wissens bisher der einzige -- ist also die Trennung der Spitzenstosselemente yon den Schalloszillationen einwandfrei durchgefiihrt. Leider sind die Oszillationen yon so geringer Amplitude, dass ihre Reproduktion unmbglich ist. Mit der Lupe betrachtet, stellen sie sich dem Auge so dar, wie die Schallkurven, die bei offenem System geschrieben wurden (Nr. XII), zeigen. In den ersten Kurven, die bei grosser ()ffnung des Zuleitungssystemes geschrieben wurden, waren die Spitzenstosselemente ~och sichtbar. In tier Kurve, bei deren Aufnahme das Zuleitungsrohr, 35 cm yon tier Brustwand und ca. 5 cm yon der Membran, allseitig vollst~tndig often war: ist die reine Schall- Herzschallstudien. 525 kurve repri~sentiert. Die Korrespondenz zwischen den letzteren und der durch den Holztrichter aufgenommenen Kurve wurde durch die Ausrechnung der Dauer, der Schwingungsfrequenz und die Identiti~t des tibrigen Charakters bewiesen~ Die Daten der hier besprochenen Schallkurve bringe ich weiter unten (S. 566). Ich stehe nun nach wie vor auf dem Standpunkte, d a s s es das Z i e l d e r H e r z s c h a l l . r e g i s t r i e r u n g sein muss, in a l l e n F a l l e n r e i n e L u f t s c h a l l a m p l i t u d e n in der verlangten Weise schreiben zu kSnnen. Bisheuteistdasnoch mit keinem der angegebenen Apparate erzielt wordenl). Da aber durchweg angenommen wird, dass die in der letzten Zeit publizierten Herzschall-Registriermethoden dies zu leisten vermiigen, muss es unsere hufgabe sein, unsere Behauptung im einzelnen noch weiter zu beweisen. Die Aufzeichnung des Herzschalles ist die dringendste und wichtigste Forderung der Physiologie und Klinik des Kreislaufes; sie muss einen strengen kritischen Maassstab vertragen kSnnen und auf solidester theoretischer und technischer Basis ruhen. Ich wiinsche die nachfolgenden Ausfiihrungen nur yon diesem Gesichtspunkte aus aufgefasst zu sehen. Es scheint, dass manche Autoren sich haben verleiten lassen, in jedem Falle, wo zwei Gruppen yon Oszillationen in einer Herzperiode auftreten, auf den Doppelschall des Herzens zu schliessen. In dieser Folgerung liegt aber ein Trugschluss. Die Spitzenstosskurve des normalen und kranken Individuums setzt sich aus mehreren Impulsen zusammen. Diesen Impulsen kommt eine verschiedene Digniti~t zu. Sie bringt es mit sich, dass die 1) Sc. soweit aus den Angaben der Literatur zu ersehen ist. Mein eigener bisheriger Apparat vermag es bis jetzt nicht lediglich wegen ungeniigender VergrOsserung. Dieser techniscbe Fehler kann mit Leichtigkeit dadurch behoben werden, dass dem Apparat grSssere Abmessungen gegeben werden. Ich babe reich dutch entsprechende Aufnahmen iiberzeugt, dass die benutzte Lichtquelle auch tiir mehrfache starkere VergrSsserung noch bequem ausreicht, also hier keine Schwierigkeiten entstehen. Dass die benutzten Membranen Schall yon der Schwingungszahl der HerztSne in charakteristischer Form aufnehmen, hube ich oben an den Orgelkurven gezeigt. Ich hoffe, bald die technisch relativ leichte Umi~nderung des Apparates vornehmen zu kSnnen. -- Es kann hier nicht meine Aufgabe sein zu priifen, was die weitei-e Ausarbeitung der yon anderen Autoren angegebenen Schallregistriersysteme zu leisten :imstande ist. 526 Heinrich Gerhartz: relative Beziehung zueinander variiert, d. h. bei vergrSsserter Intensiti~t des Spitzenstosses braucht nieht notwendig die Beziehung der einzelnen Komponenten zueinander gewahrt zu bleiben. Ich besitze sowohl in Kurven yore Menschen wie vom Hund beweisendes Material hierfiir (Nr. XIII). Es werden im wesentlichen zwei Impulse beobachtet, ein erster, der langsam ansteigt und wieder in sich gegliedert ist~ und ein zweiter, der schnellere Oszillationen einleitet und mit einer Erhebung endet, die etwa die Frequenz der ersten UntergruFpen der primaren Erhebung besitzt. Der letzte Impuls der Periode findet sich in der dutch 0ffnung des Zuleitungssystemes reduzierten Kurve in normaler GrSsse vor. yon allen anderen, dem Gipfel des ersteren abgesehen, ist nichts zu sehen. Die Kurve enth~dt dann lediglich zwei Zackengruppen. Diese Erhebungen habe;n nichts mit den HerztSnen zu tun. Das beweist 1. ihre Genese, 2. der Umstand, dass diese, wie mich andere Kurven, in denen sie geschrieben wurden, gelehrt haben, eine h0here Frequenz besitzen. Es ist klar, dass in einem relativ offenen System bzw. in einer weitgeschriebenen Kurve~ falls die Vorbedingungen zu Eigenschwingungen in der Apparatur gegeben sind, diese beiden Impulse, die auch beim Menschen auftreten, Anlass zur Ausl5sung you Eigenschwingungen geben ki~nnen. Eigenschwingungen st5ren am ehesten bei grosset Spannung des Aufnahmeapparates, well hier die Di~mpfung am schwierigsten ist. Da in den Versuchen E i n t h o v e n ' s der Quarzfaden des Saitengalvanometers 3230 Schwingungen in der Sekunde machte, ist man versucht, die Methodik Ei n t h o v e n ~s zuniichst auf diese Fehlerquelle hin zu untersuchen. Ungeniigende Di~mpfung zeigt sich vor allem in steter gleichfSrmiger Unruhe des Fadens. Es gibt keine von den yon E i n t h o v e n verbffentlichten Kurveu, in denen die Oszillation des Fadens auch in den Pausen nicht sichtbar w~re. Sie steigert sich mitunter so welt, class die Abgrenzung tier tibrigen Impulse darunter leidet. Die genauere Beurteilung ist auch erschwert, weil in den meisten Reproduktionen der Kurven die exakte Auszi~hlung der Schwingungen kaum mSglich erscbeint. Ich will abet davon absehen und lege mehr Wert auf den Charakter der grSsseren Impulse. Alle Schalloszillationen E i n t h o v e n ' s enden in typischer Sinusschwingung mit relativ langsamem Dekrement. (Vgl. besonders Piliiger's Archiv Bd. 120. Tar. I Kurve 5 und E i n t h o v e n ' s Fig. 3 und 4 aus Bd. 117.) Herzsehallstudien. 527 Sinusschwingungen sind aber typisch far freie Schwingungen. ~atarlich kann nicht behauptet werden, dass nicht auch' der Herzschall mit Sinuswellen abfallen kann. Meine Erfahrung ~ r i c h t jedoch entschieden dagegen, dass es oft der Fall ist, wovon man sich 'ja auch durch das GehSr aberzeugen kann. Sinuswellen bedarfen zu ihrer Entstehung eines Impulses. In der letzten Arbeit E i n t h o v e n ' s findet man nun bTotizen~ die vielleicht nach dieser Richtung hin Aufschluss geben. E i n t h o y e n teilt flort in Fig. 1 auf S. 36 Kurven mit, we]che die Beziehung des Kar0tispulses zum Elektrokardiogramm ~'eranschaulichen. Daraus berechne ich, dass die Ventrikelzacke des Elektrokardiogrammes 0,168 Sek. vor dem Anstieg der Karotis sich erhebt. Die Beziehung zwischen Karotis und HerztSnen lasst sich aus Fig. 4 auf Tafel I in P fl tige r ' s Archiv 120 entnehmen. Diese Schallkurve , --~ i t i . : ,. ~ i i i i q , O'1." Fig. 5. Beziehung zwischen Herzschall, Ventrikeldruck und Elektrokardiogramm beim Hunde. Elektrokardiogramm und Ventrikeldruck naeh Kahn. Latenz zwischen Druekanstieg und I. Ton naeh eigenen Messungen zu 0,012" (Mensch) angenommem Tondauer ~ Mittel aus eigenen und F r a n k ' s Beobachtungen [I ~ 0,072" (Frank) und 0,076" (Eig.), I I = 0,044" bzw. 0~045"; Intervall zwischen Ende des I. Tones und Anfang des II. ~ 0,093" bzw. 0,085", im Mittel also 0,089".] bezieht sich allerdings auf die AortentSne. Schon E i n t h o v e n hat aber berechnet, dass der II. Spitzenton und der II. Aortenton in demselben Augenblicke beginnen. Aus den Reproduktionen der Tafel I E i n t h 0 v e n ' s l~sst sich dutch Ausmessung des Intervalles zwischen Anfang des I. und des II. Tones dasselbe far den I. Ton zeigen. Daraus geht hervor, dass der I. Herzton kurz hinter die Ventrikelzacke des Elektrokardiogrammes anzusetzen ist. Nach K a h n 1) erfolgt nun ebenfa]ls kurz nach der Ventrikelzacke (0,05 Sek. spater beim Hunde) tier systolische Druckanstieg. Letzterer koinzidiert aber nach C h a u v e a u und M a r e y 2) wiederum mit dem 1) R. [-I. Kahn, Beitr~ge zur Kenntnis des Elektrokardiogramms. Pfliiger's Arch. Bd. 126. S. 197--225. 1909. 2) Chauveau und Marey, Tray. du lab. de Marey t. 1 p. 25. 1875. 528 Gerhartz: Heinrich Spitzenstoss. Es erfolgt also genau zu der Zeit, wo die tlerztSne in Ei n t h o v e n' s Kurven angesetzt sind, die systolische Erhebung des Spitzenstosses, woraus so viel hervorgeht, dass eine AuslSsung der ,I-Ton:'-Schwingungen durch den Spitzenstoss, der des II. Tones durch die Inzisurerhebung, plausibel erscheinen muss. Ein wesentliches Charakteristikum mtissen die Herzschallschriften, wenn man auch auf die iibrigen (Intensitiit, Dauer usw.) verzichten will, wenigstens zeigen, d. i. die Identitat der Schwingungsfrequenz zwischen Schall und seiner Schrift. Ich babe die Einthoven'schen Kurven gesunder Menschen daraufhin untersucht und bin zu folgendem Ergebnisse gekommen: Pflt~ger's Archiv Bd. 117 T a f e l XVIII. F i g u r 1. I. Ton : 8ekunde, 3 Schwingurlgen in 2,5"0,04" = 0,10% Also 30,0 Schwingungen pro 4 :, ,, 3 , 0 - 0 , 0 4 " ~ 0,12". , 33,3 ,, ,, ,, 4 . . . . 4,0.0,04" ~ 0,16". , 25,0 ,, :~ ,, Mittlere 2 II. Ton : Schwingungen in 1 , 0 - 0 , 0 4 " Dauer 0,38 : 3 ~ , 0 , 1 3 '. Also 50,0 Schwingungen pro Sekunde. ~ 0,04". 3 ,, ., 3 , 0 . 0 , 0 4 " ~ 0,12", ,, 25,0 , ,, , 2 ,, ,, 1,7 9 0 , 0 4 " = 0,07", ,, 28,6 ,: ,, ,, 1 ,, , = 0,03". ,, 33,3 ,, ,, ,, 0,8-0,04" Mittlere Dauer 0,26 : 4 ~ Figur 0,065" 2. I. Ton : 7 Schwi~gungen in 4 , 0 " 0 , 0 4 " ~ 0,16". 9 , , 5,3-0,04" ~ 0,212". , 42,4 ,, 7 ., , 4,5 9 0,04" ~ 0,18". ,, 39,0 ,, :, , 7 :, , 4,0-0,04" ~ 0,16". ,, 43,7.5 , , ;, 8 :, :, 4 , 0 9 0 , 0 4 " ~ 0~16". ,, 50,0 :, ,, ,, 6 ,, ;, 4 , 0 . 0 , 0 4 " ~ 0,16". ,, "37,5 ,, ., ,, 5 ,, ,, 5 , 0 . 0 0 4 " ~ 0,20". ,, 25,0 , ,, Mittlere Dauer Also 43,8 Schwingungen pro Sekunde. ~ 1,232 : 7 ~ .... 0,176". II. Ton" 4 Schwingungen in 2,6. 004" ~ 0,104". Also 4 ,, ,, 2,5.0,04" ~ 0,10". ,, 40,0 :, ., , 5 ,, ., 3,0-0,04" ~ 0,12". , 41,7 ,, , ,, 5 ,, , 2,8.0,04" ~- 0,112". , 44,7 ., ,, ,, 4 :, ,, 2,5.0,04" ~ 0,10". , 40,0 ,, ,, ,, ,, ,,. 2 , 8 . 0 , 0 4 " -~ 0,112". :, gi,7 ,, ,, ,. 5 ~ 38,5 Schwingungen pro Sekunde. 529 Herzschallstudien. 6 S c h w i n g u n g e n in 3 , 0 - 0 , 0 4 " = 0,12". 4 ,, ,, 2 , 2 . 0 , 0 4 " = 0,088". ,, 45,5 ,, ,, ,, 3 ,, ,, 2 , 5 . 0 , 0 4 " = 0,10". ,, 30,0 , ,, , 4 ,, ,, 2 , 0 . 0 , 0 4 " = 0,08". ,, 49,9 , ,, Mittlere Dauer = Also 50,0 Schwingungen pro Sekunde. 1,036 : 10 = Figur 0 , 1 0 3 6 ". " 3. I. T o n : 6 Schwingungen in 5 ,, ,, 6 , 5 . 0 , 0 2 " 5,5 9 0 , 0 2 " = = 0,13". 0,11". A l s o 54,6 S c h w i n g u n g e n p r o S e k u n d e . ,, 38,5 ~, ,, ,, 6 ,, ,, 7,0.0,02" = 0,14". ,, 42,9 ~, ,, ,, 3 ,, ,, 4 , 0 . 0 , 0 2 " = 0,08". ,, 37,5 ~' 1' ,, 3 ,, ,, 3 , 0 - 0 , 0 2 " = 0,06". ,, 50,0 ,, 5 ,, ,, 6 , 0 " 0 , 0 2 " ~ 0,12". ,, 41,7 ,; 5 ,, ,, 6 , 0 . 0 , 0 2 " = 0,12". , 41,7 ,, 6 , 0 . 0 , 0 2 " = 0,12". ,, 41,7 Mittlere Dauer = 0,88": 8 = ,, 0,11". II. T o n : 3 S c h w i n g u n g e n in 2 , 5 . 0 , 0 2 " ~ 0,05". 3 = 0,06". Also 60 Schwingungen pro Sekunde. ,, 50 ,, , ,, ,, ,, 3 , 0 . 0 , 0 2 " 3 , ,, 3 ; 0 . 0 , 0 2 " = 0,06". , 61) ,, ,, 4 , 0 . 0 , 0 2 " : 0,08". ,, 75 8 ~) ,, ,, 5 , 0 - 0 , 0 2 " ' = 0,10". ,, 80 ,, 3 ,, ,, 3 , 0 . 0 , 0 2 " 0,06". ,, 50 . 0,41": 6 = 0,07". = Mittlere Dauer = Figur 5O , . . . ,, . , 4. I. T o n 9 6 S c h w i n g u n g e n in 6 ; 5 - 0 , 0 2 " = 0,13". 6 ,, , 6,5.0,02" = 0,13". Also 46,2 Schwingungen pro Sekunde. ,, 46,2 ,, , ,, 5 ,, ,, 7 , 0 . 0 , 0 2 " ~ 0,14". ,, 35,7 ,, ,, ,, 6 ,, ,, 7 , 0 . 0 , 0 2 " = 0,14"i ,, 42,9 ,, , ,, 6 ,, ,, 6 , 8 . 0 , 0 2 " = 0,136". ,, 44,1 ,, ,, ,, 7 ,, ,, 7 . . . . 7 , 7,3.0,02" = 0'146". ,, 48,0 , , ,, 8,0.0,02" = 0,16". ,, 43,75 ,, ,, ,, ,, 7 , 4 . 0 , 0 2 " -- 0,148". 47,3 , ,, ,, :8 = 0,141". Mittlere Dauer = , - p! 1,130 II. T o n : 6 S c h w i n g u n g e n in 7 , 8 . 0 , 0 2 " = 0,156". 5 ,, ,, 5 , 0 . 0 , 0 2 " ~ 0,10". , 50,0 , ,, ,, 3 ,, ,, 3 , 3 . 0 , 0 2 " = 0,066". , 45,5 ,, ,, ,, 3 ,, ,, 3,0.0,02" ~ 0,06". ,, 50,0 3 ,, ,, 3 , 0 - 0 , 0 2 " = 0,06".. , 50,0 4 ,, ,, 3 , 0 . 0 , 0 2 " 0,06". ,, 66,6 = Also 38,5 Schwingungen pro Sekunde. 1) S e h r f r e q u e n t e S c h w i n g u n g e n m i t g e z ~ h l t . E. P f l f i g e r , Archly ffir l)hysiologie. Bd. 131. . ,, ,, ,, , , ;, , ,; , 38 530 Heinrich Gerhartz: 3 Schwingungen in 3,0.0,02" = 0,06". 4 ,, 3,0 9 0 , 0 2 " ~ 0,06 % ,, Mittlere Dauer husserdem Archly Bd. sind 120 66,6 ,, pro Sekunde. ,, ,, Pfl i1 g e r ' s 0,622" : 8 = 0,078" eines Gesunden in Die vorkommenden TaM I abgebildet. sind nacb meiner Berechnung: dort Figur 1. Toll : 5 Schwingungen 5 ,, Herzspitzent6ne Schwingungszahlen ]. = A l s o 50,0 8 c h w i n g u n g e n ,, in 4,8 9 0 , 0 2 " = 0,096". 4 , 0 90 , 0 2 " = 0,08". A l s o 52,1 8 c h w i u g u n g e n ,, 75 ,, pro 8ekunde. ,, ., ,~ 3 , 3 . 0 , 0 2 " = 0,066". ,, 75,8 ,, ,, ,, 3,2 - 0 , 0 2 " = 0,064". ,, 94 ,, ,, ,, 3,2- 0,02" ~ 0,064". ,, 94 ,, ,, ,, g,6.0,02" = 0,072". ,, 97,2 ,, , ,, 3,3- 0,02" = 0,066 ". ,, 60,6 ,, ,, ,, Mittlere D~uer = 0,488":7 = 0,0697". 2,5 - 0 , 0 2 " = 0,050". 2,3 0,02" ~ 0,046". ,, 65,2 ,, II. Ton : 4 8ehwingungen in Also 80 Sehwingungen pro Sekunde. ,, ,, 2 ~, ,, 2,2 0,02" = 0,044';. ,, 45,5 ,, ,, ,, 3 ,, ~, 2,1 0,02" ~ 0,042'2 ,, 71,4 ., ,, ,, 1,8 0,02" ~ 0,036". , 55,5 ,, ,, ,, 2,0 0,02" = 0,04'2 ,, 75 ,, ,, ,, 2,5 0,02" = 0,05". ,, 60 ,, Dauer -- 0,308" : 7 = 0,044". Figur 2. 3 ,, ,, Mittlere I. Toll : 4 Schwingungen in 1 , 8 . 0 , 0 2 ' ' ' ~ 0,036". 6 ,, , 3,7.0,02" ~ 0,074".. ,, 81,1 , ,, ,, 5 ,, , 2,6.0,02" = 0,052". ,, 96,2 ,, ,, ,, 6 ,, ,, a , 4 . 0 , 0 2 " = 0,068". ,, 88,2 ,, ,, ,, 6 ,, ,, 3,0 9 0 , 0 2 " = 0,06". ,, 100 , ,, ,, 5 ,, ,, = 0,058". ,, 86,2 ,, ,, ,, 7 ,, ,, .3,0 9 0 , 0 2 " 0,06". ,, 116,7 ,, ,. ,, 2,9.0,02" Mittlere Dauer = A l s o 111,1 8 c h w i n g u n g e n 0,408":7 = pro Sekunde. 0,058". II. Ton: 3 SchMngungen in 1,8-0,02" -- 0,036% 2 ~ ,., ,, 2~0 9 0 , 0 2 " ~ 0i04". ,, 50 ,, ,, ,, 5 , , , 2,7 - 0 , 0 2 " ~ 0;054'~'. ,, 92,6 ,, ,, ,, 3 ,, ,, 2 , 0 . 0 , 0 2 " = 0,04". , 75 ,, ,, ,, 3 ,, ,, 2 , 0 - 0 , 0 2 ~' ~ 0,04". ' ,, 75 ,, ,, ,, 4 ,, ,, 2,0.0,02" ~ 0,04". ,, 100 . ,, ,, 3 ,, ,, 1,8- 0,02" = 0,036". , 83,3 Mittlere Dauer ~ Also 83,3 8chwingungen ..0,286" : 7 = 0,041" pro Sekunde. Herzschallstudien. 531 Hieraus berechne ich folgende M i t t e 1z a h 1 e n: I. Ton : Pfliiger's Arch. Bd. 117 F i g . S c h w i n g u n g e n p r o Sek. / 1:29,4 ,. ~, ,, 117 ,, 2 : 4 0 , 2 ,, ,, ,, ,, ,, ,, 117 ,, 3:43,6 , ,, ,, 117 ,, 4:44,3 ,, ,, Schwingungen Mittel: 89,4 S c h w i n g u n g e n . Pfltiger's pro Arch. Bd. 120 Fig. 1 : 7 8 , 4 S c h w i n g u n g e n p r o Sek. 120 Im Mittel I 2 : 97,1 ,, Sekunde. ,, Mittel: 87,7 S c h w i n g u n g e n . !I. Ton : Pfliiger's ,, A r c h . Bd. 117 Fig. 1:34~,2 S c h w i n g u n g e n p r o Sek. ! ;, , 117 ,, 2 : 4 2 , 5 , ,, ,, ~, ,, ,, 117 ,, 3:61 ,, ,, , I] Im Mittel ,, ,, ,, 117 ,, 4:52,15 ,, ,, ,, I. bb,7 Mittel: 47,b S c h w i n g u n g e n . Pfltiger's ,, Arch. Bd. 120 Fig. t : 04,7 S c h w i n g u n g e n p r o Sek. ,, ,, 120 ,, 2:79,9 ,, ,, Mittel: 72,3 S c h w i n g u n g e n . ,, : I Schwingungen pro Sekunde. I[ t Hieraus geht hervor, dass der II. Herzton, ausgenommen den anscheinend doch abnormen Fall, in dem ein dritter Herzton registriert wurde ( P f l u g e r ' s Archiv Bd. 120) hSher ist als der I. Ton. Ich time so far den I. Ton 39,4, fiir den II. 47,5 Sehwingungen pro Sekunde. Die bier sich zeigende hfhere Frequenz des II. Herztones entspricht nun durchaus der Norm. Auch die relative Karze des II. Tones stimmt mit dem normalen Verhalten tiberein. Ich babe die Dauer aus E i n t h o v e n ' s Figuren berechnet (Tabelle 2) und finde hier den II. Ton z u . 5 0 : - 7 1 % , im Mittel zu 60 % der Dauer des I. Tones. Man k0nnte annehmen, dass dies dagegen spricht, class EigenTabelle Mittlere Dauer 2. der Einthoven'schen Herzt~ne. (Nach meiner Berechnung.) I. H e r z t o n II. H e r z t o n 0,13" ~ 0 , 1 7 6 " I,t 0 1 8 9 " 0,11-" t\| ' 00141 " i t 0,070 "1) I-'a aa~ ,, 0 , 0 5 8 " /J , " ' ~ ' ~ 0,065" , 0,1036" ~0079" : 57% 0,07aa~",, -('[ des L T o n e s ' ,v.~ . ..... r l Pfliiger's ,, ,,~ ,, s Arch. Bd. 117 Fig. 1 ,, ,, 117 ,, 2 - ~, ::~117 ,, ,:3 . , . ,, 117. ,, 4 ,, ,, ,,, 120 ,,. 120 ,, 1 ,, 2 0,044 "- [ 0 , 0 t 2 " ~ 66 % 0 , 0 4 1 " ;/ des I. T o n e s I 36* 532 Heinrich Gerhartz: schwingungen des Systemes wesentlich im Spiele sind, da man dann eine Proportionalitat zwischen Dauer und Frequenz der Schwingungen erwartet, d. h. um so niedrigere Frequenz vorhanden sein wiirde, je langer die Phase des betreffenden Tones dauerte, was hier fehlt. Man muss aber in Betracht ziehen, dass diese Annahme nur richtig ist bei Voraussetzung gleicher Intensitiit. Diese Voraussetzung trifft aber nicht zu. Ich schliesse hieran die Diskussion der Ergebnisse, die mit der Weiss'schen Methodik gewonnen wurden. Ich habe mich auch in dieser Betrachtung wieder an die Reproduktionen gehalten in der Annahme, dass sie hierfiir geniigend korrekt sind. W e i s s schreibt allerdings, dass diese ,,sehr mangelhaft und vielfach inkorrekt" sind (1. c. S. 519. S. 360). Man darf deshalb abe, doch wohl annehmen, dass die wesentlichsten Dingey Schwingungszahl und Dauer, geniigend zuverlassig angegeben sind, da doch sonst die Darstellung der mit Zeitlinie versehenen Kurven unn0tig gewesen ware. Ich bin bei de, huszi~hlung so vorgegangen, dass ich die Reichbreite der Zackengruppen in Millimetern ablas, die gefundene Millimeterzahl dann in der direkt unter der Kurve befindlichen 1/loo-Sekundenmarkierung in Sekunden umsetzte und die Anzahl der Schwingungen zahlte. Ich beschriinke reich hier wegen tier relativen Sparlichkeit der hngaben nicht nur auf die SpitzentSne. Seite 360 Fig. 8 (PulmonaltSne). 2 Schwingungen in 0,03". Also 66,7 pro Sekunde. ,, a) 5 , , 0,08". ,~ 62,5 ,~ , ~ b) 4 ,, ,, 0,055". , 72 , ,, ~ I. Ton. II. Seite 360 Fig. 9 (PulmonaltSne). a) 4 Schwingungen in 0,09". b) 5 ,, ,, 0,06". I. Ton. II. , 63,8. ,, ,, 0,055". , 72,7 ~ , ~ Mittel b) 2 , ,, 0,025". ,, 80 ,~ ,, ! 76,3. Also 87,5 pro Sekunde. ~ Mittel Seite 361 Fig. 10 (Aortent6ne). a) 7 Schwingungen in 0,08". ,, Mittel a) 4 I. Ton. IL Also 44,4 pro Sekunde. [ J ,, 83,3 , ,, Mittel 67,2. b) 5 ,, ,, 0,08". ,, 62,5 , ,, ] 75. a) 6 ,, , 0,065". , 92,3 ,, , [ Mittel b) 8 ,, ,, 0,085". , 94,1 ,, ,, ] 93,2. 1) Ich komme hier zu etwas anderen Zahlen, als E i n t h o v e n (S. 33) aDgibt Das mag damit gentigende Erkltirung finden~ dass E i n t h o v e n zur Berechnung die Originale benutzt hat, ich die Reproduktionen zugrunde gelegt habe. An de, Sache selbst wird hierdurch nichts geiindert, da es sich um Yergleiche zwischen den Zahlen derselben Person handelt. (I. T o n ~ 0,08" Dauer, II. T o n ~ - 0,05"). 533 Herzschallstudien. Seite 361 Fig. 11 (AortentSne). I. Ton. a) 4 Schwingungen in 0~05". Also 80 pro Sekunde. b) 6 , , 0,063". , 95,2 . . . . II. , a) 8 ,, , 0,09". , 88,9, ;, b) 3 , , 0,035". ,, 85,7 , ,, [ J t J Mittel 87,6. Mittel 87,3. Seite 361 Fig. 12 (SpitzentSne). I. Ton. a) 8 Schwingungen in 0,09". Also 88,8 pro Sekunde. \| b) 5 ,, , 0,05". ,, 100 , ,, ] II. ,, a) 10 , ,, 0,10". , 100 ;, ,, I b) 10 ,, ,, 0,11 ". ,, 90,9 , , I Mittel 94,4. Mittel 95,4. Seite 361 Fig. 13 (SpitzentSne). I. Ton. a) 7 Schwingungen in 0,08". Also 87,5 pro Sekunde. ~ Mittel b) 10 , , 0,10". , 100 , ,: -~ 93,7. II. , a) 5 , , 0,056". , 89,3 , , ~ Mittel b) 5 , , 0,057". , 87,7 ,, , -f 88,5. Seite 364 Fig. 14 (SpitzentSne). I. Ton. 4 Schwingungen in 0,06". Also 66,6 pro Sekunde. II. ,, 8 , ,, 0,085". ,, 94,1 . . . . Seite 364 Fig. 15 (Spitzent5ne). I. Ton. 7 Schwingungen in 0,083% Also 84,3 pro Sekunde. II. , 14 , , 0,13". , 107,7 , ,, Hieraus ergeben sich als Durchschnitt die Werte der folgenden Tabelle 3; d . h . die II. TOne sind, im Durchschnitt betrachtet, hSher als die ersten; dagegen sind die II. SpitzentSne und Aortent0ne entweder gleichlang oder langerdauernd als die ersten, was, da es sich in den Reproduktionen um Kurven gesunder Individuen handelt, kaum mit der Wirklichkeit iibereinstimmen kann. Zusammengehalten mit den in typischen Sinuskurven endenden Kurvenbildern Figur 8, 10, 11, 13 (II. Ton) u. a. wiirde es einen berechtigten Verdacht auf Eigenschwingungen wachrufen, wenn nicht yon W e i s s eine Reihe yon Kurven beigebracht worden w~tren, welche dieser Deutung widersprechen und unzweifelhaft ,erzwungene" und Schall-Schwingungen darstellen. Eine Aufkliirung ist aber hiermit i~ber den genannten Punkt k a u m gegeben. Merkwiirdig ist auch, dass in Figur 8, 9, 11, 12 und 13 I. und II. Ton gleiche, bzw. sogar der II. Ton niedrigere HShe besitzt als der I. T o n , die durchschnittliche hi)here Schwingungsfrequenz also auf einigen wenigen sehr hoch liegenden Schwingungszahlen des II. Tones beruht. Der II. Ton besitzt bei W e i s s regelmiissig grSssere Amplitude als der I. Heinrich Gerhartz: OO'-x Tabelle 3. Znsammenstellung der Ergebnisse, der W e i s s ' schen Methodik. (Nach meiner Berechnung der verSffentlichten Kurven.) Dauer Fre~ uenz I. Ton. Spitzenton . . Pulmonalton . Aortenton . . II. Ton. Spitzenton . . Pulmonalton . ~kortenton . . 84,8 64,3 ~1,3 77,0 . 0,077" 0,06" 0,068" ]\| 0,068" ~ (96 ~ des \| 1I. Tones) 96,4 71,7 90,2 $6,1 0,090" 0,054" 0,069" ~ f 0,071 " Vergleieht man die bisher bereehneten Daten, so ergibt sieh folgende Zusammenstellung 4: T a b e l l , e 4. Yergleichende Zusammenstelhmg tier aus den Kurven yon E i n t h o y e n und W e i s s z u berechnenden Charakteristika tier HerztSne. Einthoven Weiss (Pflf~ger's Archiv Bd. 117) I. Herzton II. Herzton Sehwingungszahl pro Sekunde Dauer Schwingungszahl pro 5ekunde Dauer 39,4 47,5 0,139" 0,079" 77,0 86,1 0,068" 0,071 " Das heisst, wenn es gestattet ist, die Daten der verSffentlichten Kurven dem Vergleich der Systeme zugrunde zu legen, was natt~rlich nieht streng richtig, aber aueh nieht zu umgehen ist: D i e Schwingungszahl d e s I. u n d II. H e r z t o n e s i s t b e i W e i s s f a s t g e n a u d o p p e l t so h o e h als bei E i n t h o v e n , f e r n e r i s t die D a u e r d e s I. T o n e s bei W e i s s h a l b so g r o s s a l s b e i E i n t h o v e n . D i e II. T S n e h a b e n bei b e i d e n A u t o r e n gleiehe Dauer. Die Erklarung dieser frappanten Differenzen hat ihre Sehwierigkeit. Man wird auch hier am ehesten geneigt sein, an Eigenschwingungen bei beiden Systemen zu denken: wachst die Frequenz, so nimmt, gleiehe einwirkende Impulse vorausgesetzt, die Dauer ab. Das kann fi]r den I. Ton ohne Sehwierigkeit die Situation erklaren. Aber wie steht es mit dem II. Ton? Hier seheint mir die Erklarung in folgendem zu liegen: Bei E i n t h o v e n ist die Amplitude des II. Tones niedriger als die des I. ; bei W ei s s ist das umgekehrte Herzschallstudien. 535 Verhalten vorhanden. Das spricht wohl entschieden dafiir, dass die Ifitensitiit des auftreffenden Impulses-A gleichgfiltig, was es gewesen sein mag - - d a s eine Mal ( W e i s s ) gross, das andere Mal niedrig gewesen "isL Bei W e i s s wiirde also die Verkiirzung eIer II. Ton= periode~ die infolge der hSheren Frequenz hiitte eintreten mi]ssen, d u t c h die ErhShung des das II. Tonbild verursachenden Impulses kompensiert worden sein. ~ 9 Diese Erkli~rung erscheint mir annehmbarer ais die~ dass bei E i n t h o v e n der Grundton, bei W e i s s die Oktave dazu registriert wurde infolge der differierenden Anspruchsfahigkeit tier Registriersysteme, Man wird sich erinnern, dass ja E_~ 40 Sghwingungen, "~--1 80 (.E8_ 1 = 76,8); G_~ 48, G_ 1 96 (F_I z,85i3)Schwingungen entspricht. Die Differenzen zwischen den beobacht'eten :und den Normalzahlen liegen innerhalb tier Fehlergrenzen, Bei der Untersuchung der Ergebnisse der F r a n k'schen Methode ist auch vor allem zu finden, dass es sich ebenso wenig wie bei allen anderen Verfahren um die Aufzeichnung yon Luftschallamplituden handeln kann. i Es folgt das aus folgenden WortenJ): ,Ferner li~sst es sieh ohne weiteres einsehen, dass man keine oder nur unmessbare Schwingungen erhalten wird, wenn man die Membran nicht durch eine geschlossene Luftsi~ule mit der Brustwand in Verbindung bringt. Muss ja doch auch das Ohr entweder direkt an die Brustwand angelegt oder durch ein festes oder ein Schlauchstethoskop mit tier Brustwand verbunden werden, wenn man die tterztSne hSren will." Die Korrektheit der Schallaufzeichnungen wird niit folgendem begriindet: ,Man kann so die Schwingungen der Membran, die durch die HerztSne erregt wird, scharf und bestimmt u D d - - wie ich glaube - - korrekt erhalten. Die Kurve verliiuft im allgemeinen so, dass yon der eigentlichen Herzstosskurve nichts zu entdecken ist: die Schwingungen erheben sich aus einer fast gerade Veflaufenden Linie. Sie wird nur auf eine kurze Zeit tier Herzrevolution beschri~nkt." F r a n k ist dabei ein Trugschluss unterlaufen, da es sich bei tier von ihm als Herzschallbild angegebenen :Kurve 2) i j 1) O. F r a n k , Die unmittelbare Registrierang der HerztSne. Miinch. reed.. Wochenschr. Bd. 51 S. 953--954. 1904. 2) O. F r a n k , Der Puls in den Arterien. Zeitschr. f. BioL Bd. 46 (28) S. 524. 1905. (Fig. 30.) -- Meines Wissens die einzige yon Frank veriiffentlichte Schallkurve. 536 Heinrich Gerhartz: lediglich u m eine stark geditmpfte Spitzenstosskurve handeln kann. Ich besitze eine der F r a n k ' s c h e n Kurve vSllig entspreehende, die yon einem Menschen aufgenommen wurde, dessen Herztiine infolge eines reichlichen Fettpolsters so leise waren, dass yon einer Aufzeichnung der Schallschwingungen keine Rede sein konnte. (Nr. XIV.) Um die Analogie beider Kurven nigher erliiutern zu kSnnen, schliesse ich einige Daten (Tabelle 5) an, die die Ausrechnung der Kurve ergeben hat, und stelle noch die Charakteristika einer anderen, vSllig entsprechenden Kurve eines itlteren Mannes und einer jungen weiblichen Person hinzu. Gerade aus der zweiten Angabe ersieht man, dass die ~on F r a n k als Herzt0ne angesprochenen Schwingungen hier keine Herzschallschwingungen sein ki~nnen ; denn die Schwingungszahl liegt ganz abnorm fief unter der Frequenzschwelle. Zum Uberfiuss babe ich noch am Harmonium den dem I. Herzton entsprechenden Ton aufgesucht und - - kontrolliert dutch eine andere musikalische Person - - G_~ ~---96 gefunden. Tabelle 5. ~ H e r z t o n ~ c h a r a k t e r i s t i k a yon Kurven ~ die dem F r a n k ' schen K a r d i o . phonogramm entsprechen. Herztondauer I. Ton III. Ton Differenz zwischen I. TonBegiun un4 KarotispulsAns~ieg I. Ton III. Ton Herz- 55,5 45,4 38,3 47,8 2514 39,3 54 50 0,41" 1,034" 0,857" 0,857" Herztonfrequenz periodendauer l F r a n k : Hund') . . . . G e r h a r t z : Junger Mann Alter Mann. Junge Frau. 0,072" 0,131" 0,105" 0,108" 0,014" 0,042" 0~053" 0,062" 0,059"(Aorta) 0,133" --- f Es liisst sich der Beweis auch auf andere Weise ftthren an der Hand der Skizzen Fig. 6 (Nr. XV), welche die Schwingungen, welche dem I. Herzton entsprechen sollen, getrennnt und genetisch wiedergeben. Man wird daraus ohne Schwierigkeit ersehen, dass die erste, kleine Zacke immer akzidentell ist in der ersten Hauptzacke 2, ferner, dass 3 mit 4, 5 und 6 sich zu eineL' Gruppe zusammensetzt, d. h. 3, 4 und 5 drei auf die Haupterhebung 6 superponierte Zacken sind. Es kSnnen also zwei Erhebungen der ganzen Gruppe keine Schallzacken sein; sie gehSren der Grundexkursion des Spitzenstosses an und sind, wie Skizze e zeigt, in seinen Verlauf eingeschaltet2). 1) Eigene Ausrechnungen! 2) ,,Secousses ~' yon v. H o l o w i n s k i ? 537 Herzschallstudien. :Nur die Erhebungen 1, 3, 4 und 5 sind akzidentell. Es ist nun allerdings miiglich, dass ebensolche akzidentelle Schwingungen mit den Erhebungen 2 und 6 zusammenfallen. Es wtirde sich also gewissermaassen dabei um eine Zufiilligkeit handeln. Ein anderes Beispiel, welches zeigt, wie leicht u. U. Teile einer Spitzenstosskurve ffir Schalloszillationen angesehen werden kiinnen, mSchte ich noch erwahnen. Ich babe einmal den Spitzenstoss so aufgenommen, dass ein Aufnahmetrichter mittels T-Stiiek die Erschiitterungen in zwei Zuleitungsrohren den beiden verschieden stark gedi~mpften Membranen des Apparates zuftihrte. In der Kurve (Nr. XVI) J 2 ~~ ~ ,~ A ~.~ Fig. 6. Modifikationen der primi~ren Sehwingungen. stimmte die Frequenz der ,,primiiren ~' Schwingungen mit der des I. Herztones der jungen weiblichen Person, yon der die Aufnahmen stammen, ilberein. Ich babe sie in den Kurven zu 54 Schwingungen pro Sekunde berechnet. Das Bild der ,primi~ren" Impulse stimmte ferner vollkommen mit dem F r a n k ' s und dem der eben besprochenen Kurve iiberein. Somit lag es nahe, in tier Kurve eine recht schSne Schallkurve zu sehen. Es musste aber stutzig machen, dass die Frequenz des ,,II. Herztones" mit 50,3 Schwingungen pro Sekunde niedriger war als die der Schwingungen des ,,I. Herztones". Das Riitsel 15ste sich, als ich in der gleichen Weise, als die Herzperiode li~nger geworden war, aufnahm. Die Herzperiode hatte bier yon 0,857 Sek. bei der ersten Messung auf 0,8955 Sek. sich erhSht. Die ,,Ton"dauer nahm ebenfalls zu, die Schwingungszahl aber 538 Heinrich Gerhartz: ab und zwar unter einen annehmbaren Wert. Wie man aus der~ Einzelwerten, die ich in Tabelle 6 zusammengestellt habe, ablesen kann, gingen Dauer und Schwingungszahl umgekehrt proportional. Das heisst: die ,primaren Exkursionen" wurden bei der Verti~ngerung der Periode auseinander gezogen, ihr Charakter blieb gewahrt und insbesondere die Anzahl der Einzelexkursionen. Es kann sich also keineswegs hier um interponierte Schalloszillationen handeln, sondern wir haben es lediglieh mit Spitzenstossantei!en zu tun, die ihrem Charakter nach die ersteren t~uschend imitieren. Tabelle Dauer I. Au~hme . . . II. Au~ahme . . . ProzenmaleZu-bzw. Abnahme . . . . G. ISchwingungs-I ~ ]Schwingungs-Ipe~'iode zahl ] ~aue~ I zahl I 0,1085" 0,144" 54 39,9 0,062 " 020797" 50,3 37,6 (33 ~ (26 %) (28 ~ (2.50/o) 0,857" 0,8955" Nach solchen Erfahrungen habe ich reich geht~tet, solche Gebilde als Herzschallschwingungen zu verwerten, und ich stiitze reich in den spateren Ausft~hrmlgen nur auf Kurven, in denen die frequenten und durch die Kontrolle am Instrument als S c h a l l ausreichend charakterisierten Exkursionen so auf die Spitzenstosskurve superponiert sind, dass eine Verwechselung ausgeschlossen ist. Eine Tauschung kann mitunter sehr leicht erfolgen, ja ich bin i;~berzeugt, wie ich schon oben angegeben habe, dass die ,primaren Schwingungen" gelegentlich in den Kurven direkt den I. Herzton darstellen. Ich besitze einige in dieser Hinsicht lehrreiche Erfahrungen aus tier Pathologie. In Fig. 1 hatte ich eine Kurve mitgeteilt, die den Spitzenstoss und das systolische Geri~usch eines Kranken, der an Mitralinsuflizienz litt, betraf. Ich habe zahlreiche Kurven yon diesem Kranken aufgenommen, aber keine erhalten, welche die besprochenen ,primaren" und ,sekundaren" Zacken besessen hatte. Der junge Mensch besass ein systolisches Geri~usch yon typischem Charakter. Die Kurve lehrt, dass das Gerausch in dem Moment anhebt, wo bei normalen Fallen der I. Herzton beginnt. Das Ge- Herzschallstudien. 539 ritusch selbst hat unregelmassigen Charakter, wie man leicht erkennen kann, auch ohne dass die Schallschwingungen in die ttorizontale projiziert werden. Ich erinnere reich einer Mitralstenose, wo das prasystolische Gerausch in der typischen Weise mit einem paukenden I. Ton endete und in der Kurve diese Schallverhaltnisse charakteristiseh wiedergegeben waren: der L Ton entspracb einer Schwingungsgruppe, die in jeder Hinsicht den oben diskutierten primaren Schwingungen entsprach. Klinische und Registriererfahrungen gingen also in diesen Fallen vollkommen parallel, und, worauf es bier ankommt, zwischen dem Charakter der ,,primaren Oszillationen" und denen des I. Herztones ist kein Unterschied zu finden. In diesen letzten Fallen wurden die direkt mit meiner Methode aufgezeichneten Oszillationen auch mit der Mikrophon-Saitengalvanometer-Methode E i n t h o v e n ' s verglichen. Es besteht auch bier eine vollkommene Identitat. Dafar, dass nun auch eine'Koinzidenz zwischen der W e i s s ' s c h e n Schwingungsgruppe des I. Tones und der hier diskutierten Zackengruppe bestehen kann, spricht, dass die zeitliehe Distanz zwischen dem Beginn der Erhebungen bis zum Anstieg des Karotispulses - - wie Weiss annimmt, eine konstante Beziehung - - bei F r a n k (beim Hund) 0,059 Sek., bei Weiss 0,07 Sek. ist, also gut iibereinstimmt. Ich selbst babe allerdings individuelle Differenzen dieser Distanz gefunden. Mail sieht, dass grosse Erfahrung und Vorsicht dazu geh0rt, um mit Sicherheit die Kurven zu deuten. Bei den Schallbildern der M a rb e'schen Methodik 1) ist es schwer, sich fiber ihre zeitlicben Charakteristika zu orientieren. Man sieht in den Bildern, die R o o s s) yore Herzen des Menschen aufgenommen hat, allerdings doch Figuren, die sich analysieren ]assert. Zur Berechnung 9 der Schwingungszahl hat man die Ellipsenbogen abzuzi~hlen. Diese Differenzierung ist aber schwierig, zum Tell unmSglicb, 1) Karl Mar b e, Registrierung der HerztSne mittels russender Flammem 9P f l a g e r ' s Arch. Bd. 120 S. 205--209. 1907. 2) E. R oos, l~ber objektive Aufzeichnung tier Schallerscheinungen des Herzens. Deutsches Arch. f. klin. Med. Bd. 92 S. 314--335. 1908. -- Dort weitere Literatur. 540 Heinrich Gerhartz: und auch R oos, der doch viel Erfahrung in der Deutung der Bilder haben muss, scheint sich t~ber die Anzahl der Ringe mitunter nicht schlUssig werden zu kSnnen, wie aus einer diesbezt~glichen Bemerkung hervorgeht. Die Ursache ist klar; je frequenter die Impulse sind, desto mehr werden die Bogen abgeflacht und desto weniger heben sie sich voneinander ab. Wahrscheinlich werden sie auch unscb~rfer. Schon deswillen und wegen der Schwierigkeit, die die Kombination mit anderen wichtigen Erscheinungen tier Herzt~itigkeit bietet, scheint mir die Methode, die erforderliche VergrSsserung der Oszillationen vorausgesetzt, praktisch far unsere Zwecke recht wenig brauchbar zu sein. Schwerer wiegend sind aber andere Bedenken. M a r b e und R oos verwenden eine Kapsel, die luftdicht adaptiert wird. Es ,_'st also keine u flit die Eliminierung der Spitzenstosspulsationen getroflen. R oos gibt an, dass er sich im Laufe der Untersuchungen aberzeugt hat, dass der Spitzenstoss ,,sich im Flammenbild nicht ausdrtickt und die Aufnahmen nicht st0rt". Ein Beweis daffir oder eine Erkl~rung, weshalb es so ist, wird nicht gegeben. R oos .schreibt bezt~glich dieses wichfigsteu Punktes lediglich: , , F r a n k , dessen Aufnahmeapparat im Prinzip der gleiche ist, erhielt auch nichts von einer Herzstosskurve, ebensowenig M a r b e " (S. 319), w~hrend M a r b e selbst sich vorsichtiger ausspricht: ,,Alle yon mir hergestellten Russbilder s c h e i n e n ~brigens durch den Herzspitzenstoss nicht beeinflusst zu sein, wie auch F r a n k mitteilt, dass in seinen Kurven nichts yon tier eigentlichen Herzstosskurve zu entdeeken war" (S. 208). Da die Russbilder ,,nicht ohne Bert'lcksichtigung der Zeitregistrierung miteinander verglichen werden kSnnen", ist es nicht mSglich, Stellung zu dem Marbe'schen Registrierverfahren zu nehmen. Es ergibt sich aus dem vergleichenden Studium der mit den verschiedenen Methoden geschriebenen ,,Schall"kurven, dass in der Spitzenstosskurve des Herzens Oszillationen nachweisbar sind, die an der Stelle des Kardiogrammes liegen, an der, soweit die Lokalisierfi~higkeit. des Ohres eine solche gestattet, die Herzt~ne gehSrt werden, dass im allgemeinen auch die vom GehSr wahrzunehmenden Charakteristika des Herzschalles, relative Ktirze und hShere Frequenz des II. Tones mit der Deutung, es handele sich hier um die Schallfigur der HerztSne, harmonieren. Allerdings sind Herzschallstudien. 541 im einzelnen hbweichungen zwischen den verschiedenen Untersuchera vorhanden, die tiber die sicherlich bestehenden individuellen und subjektiven Differenzen hinausgehen, und an denen nur die differierende nicht gentigende Methodik schuld sein kann. Es hat sich aber auch gezeigt, dass in sehr vielen Fallen eine exakte Differenzierung an den Kurven vorzunehmen auf uniiberwindliche Schwierigkeiten stSsst, so dass als einzig erstrebenswertes Endziel bleibt, durch Einschaltung einer starren Zwischenwand in das Schallzuleitungssystem reine Luftschallamplituden zu schreiben und auf diesem Wege die Kurven zu analysieren. Solange dies nicht erreicht ist, ist es notwendig, die Entscheidung in der theoretischen Durcharbeitung der Methodik zu suchen. Ihr fhllt hier eine nicht geringe hufgabe zu; denn die Herzschallschrift steht an der Grenze des Erreichbaren. An Schwierigkeit iibertrifft sie durchaus die weitgehendsten Anforderungen, die man sonst an optische Einrichtungen zu ste]len gewohnt ist, da sie Priizision des Lichtpunktes mit grosser Vergri%serungsfiihigkeit kombinieren muss. Andererseits fordern die minimalen Durchbiegungen der Membran eine Labilitiit und Pri~zision der Abnahmevorrichtungen, wie sie in dieser Vollendung bisher nie erstrebt zu werden brauchte. Diese Schwierigkeiten, die ich in meiner friiheren Arbeit erschiipfend besprochen babe, entschuldigen es, dass die husbeute aus den Verfahren nur Schritt auf Schritt weiter gehen kann. Ich habe friiher schon angegeben, welche Gri~nde reich veranlassen, dem yon mir angegebenen jederzeit gebrauchsfertigen hpparat, der auf dem Prinzip der direkten 1) Registrierung beruht, grosse Vorzt~ge hinsichtlich der Korrektheit der hufzeichnungen und praktischeu Brauchbarkeit zuzumessen. Soweit man die bis zu meiner friiheren Publikation verSffentlichten Verfahren heranziehen will, findet man dort die genauere Beweisfahrung. Zur Weiss'schen Methodik babe ich in einer spi~ter in diesem Archiv erschienenen Publikation Stellung genommen. Ich verweise auf das frtiher Gesagte und rage hier nur die in hussicht gestellte genauere Beschreibung der hpparatur ein. 1) W. Einthoven: ,Es braucht jedoch nicht hervorgehoben zu werden, class eine strenge graphische Darstellung den direkt registrierten kapillar-elektrometrischen Kurven vorzuziehen ist usw.~ P flii g e r ~s Arch. Bd. 117 S. 469. 1907. 54-2 Heinrich Gerhartz: Beschreibung meines Schallsehreibers. Die Leistungsfiihigkeit meines Apparates beruht auf der theoretisch beliebig welt zu treibenden VergrSsserung korrekter Exkursionen einer Membran, deren Bewegungen photographisch registriert werden. i " Fig. 7. Schema meines Schallschreibers. a ~ Membran. b - - Ubertragungssti~bci~en, c ~ Spiegel. d ~ Lichtstrahl. e ~ Lichtquelle. f ~ Film. Das Prinzip der Apparatur (siehe Schema Fig. 7) besteht darin, dass ein leichtes, drehbar aufgehiingtes Spiegelchen mittels eines kurzen Stabchens mit der Mitte der Metal)ran so verbunden ist, dass einer Membrandurchbiegung eine entsprechende Spiegeldrehung zukommt. Analog dem P o g g e n d o r f- System ' fi~llt auf die ses Spiegelchen ein Lichtstrahl, der nach dem Film bin refiektiert wird Herzschallstudien. 543 -and dessert Bewegungen also die Membrandurchbiegungen vergrSssert wiedergeben. Der hpparat (Fig. 8) setzt sich somit aus folgenden Teilen zusammen: I . Membranbefestigung, II. Spiegelvorrichtung und Di~mpfung, III. Lichtfiihrung, IV. Kamera mit Kassette, V. Z~it, registrie~ung. I. M e ~ m b r a n b e f e s t i g u n g . Die ausserst dtinne Kollodiummembran ist in einem Gehi~use vor Licht und sonstigen Einflt~ssen (LuftstrSmungen usw.) geschatzt untergebracht. Dieser Teil des Schallschreibers stellt somit eine photographische Kamera dar. Da die Membran lotrecht steht und aus die Schneide eines Halteringes aufgelegt istl), ist eine miiglichst weite Ausnutzung garantiert. Die Membran hat einen Durchmesser yon 20 ram. Der Schall wird aber nur in einer Rohrbreite von 6 mm Durchmesser zugefiihrt, so dass der Schallimpuls konzentrisch auf die Membran auftrifft. Dieses Rohr triigt aussen ein Gewinde, welches gestattet, die Membran ~'or und zuriick zu schieben. Das ist voa Nutzen, weil so bequem die Spiegelstellung, indem man eine an Stelle der Kassette eingeschobene Mattscheibe beobachtet, beurteilt und reguliert werden kann. Um zu verhfiten, dass wi~hl~end der Aufnahme die Membran dutch Bewegungen in den besprochenen Gewind~n beeinfiusst wird, wird nach der Einstellung kurz vor der Aufnahme das i~ussere Gewinde durch eine aus Blei hergestellte Schraube festgestellt. Nach der Membran zu verjilngt sich das Lumen des Zuleitungszylinders etwas; es fallt yon einem Durchmesser yon 9,0 allmi~hlich in einer Strecke yon 25 mm auf 8 mm Durchmes~ser~.ab. " Einer Rechtsdrehung der Schraube um 0,1 mm (am Anfang der Schraube gemessen) entspricht eine Vorschiebung der Membran um 3 ~.' Es .ist demnach ermSglicht, was mir fi~r theoretische Studien recht wertvoll erscheint, die G r u n d s p a n n u n g der Membran bzw. die Labiliti~t der Einstellung in weiten Grenzen Zu v a r i i e r e n und bei labilster und gespannter Membran den Effekt, der dutch die auf die Membran auftreffenden Impulse ausgeli!st wird, zu .studieren. :1) Vgl. auch S. 19 der friiheren Arbeit und die dortigen Abbildungen 2 a ~44 Heinrich Gerhartz: Ich beobachte diese Durchbiegungen und Einstellungen an einer Millimeterskala, die auf der Mattscheibe der Kassette angebracht ist, und berechne die Membranexkursionen aus den messbaren Verschiebungen des Bildpunktes, da die Entfernung Spiegel--Mattscheibe (bzw. Film) und der Winkel, in dem das Licht auffallt und refiektiert wird, bekannt ist. Ich bringe, um die Elastiziti~t der Membran unbeeinfiusst zu halten, die Membran nur wi~hrend der hufnahmen Fig. 9. Membranteil des Aufnahmeapparates. (Ansicht yon oben.) mit dem Spiegel in Bertihrung, nehme also vor jeder Aufnahme die ausserordentlich leicht herzustellende Verbindung zwischen Membran und Spiegel vor. Das geschieht so, dass an der i~usseren Schraube der Zuleitung so lange nach rechts gedreht wird, bis eine weitere Einschiebung der Membran gegen den Spiegel hin Verschiebung des Lichtpunktes auf der Mattscheibe bewirkt. Durch weitere Rechtsdrehung ist jeder gewiinschte Grad der Durchbiegung und Membranspannung herstellbar. Das leichte und starre Sti~bchen, das den Zweck hat, die Membran mit dem Spiegel in Verbindung zu bringen, wird auf die 545 Herzschallstudien. Mitte der Membran aufgeleimt, was durch eine einfacbe Zentriervorrichtang in i~usserst bequemer Weise durchzuftihren ~st. (Siehe Fig. 9, in der dasSti~bchen sichtbar ist.) Nahe dem anderen Ende wird das Stabchen durch den Stiibchentriiger geft~hrt. II. S p i e g e l v o r r i c h t u n g u n d Di~mpfung. Ein i~usserst feines Eisenplattchen triigt auf der einen Seite das Spiegelchen, auf der anderen an einer Kante (vgl. Schema Fig. 10) zwei sehr kurze Nadelspitzen, die in zwei entsprechenden Kiirnungen des Polschuhes eines Magneten sich aufsetzen, wodurch sicb dieses Eisenpli~ttchen in die Richtung der magnetiscben Kraftlinien einstellt und fast reibungsfrei J I schwingen kann. Um die Kraft, mit der das i~usserst leichte Glasspiegelchen durch Magnetismus gehalten wird~ und die Richtung, in der ~ es sich befindet, -r zu kOnnen, sind die Pole in mehr oder weniger weite Distanz zu bringen. Das ist dadurch erreicht worden, dass die Pole in Schlittenffihrung verschiebbar sind. Da die Polschuhe die Tendenz haben, das Eisenplattchen stets wieder in seine Ruhelage zurilckzubringen, wird eine vorzilgliche ~ Diimpfung erreicht, die, falls ein Elektromagnet gewi~hlt wird, veranderlich ist. Soll der Spiegel in den 'Apparat eingesetzt Fig. 10. ' Schema der Spiegelaufhangunz. (Anwerden, was z. B. dann notwendig werden sichtvon 9ben.).N---S~ kann, wenn dutch eine ungeschickte Bewegung Pole des Magneten. a Eisenplattchen. b der zu untersuchenden Person die Exkursionen Spiegelchen. des Spiegels so gross werden, class er aus den Kerben gleitet, so wird er mit einer Hornpinzette, die planparallele Branchen besitzt, in der Weise gefasst, dass die ~Nadeln senkrecht fibereinander stehen. Hi~lt man ihn dann fiber die Kerben, so schnappt er ein und wird nun mit einem entfetteten Haarpinsel angedrfickt. Da die Kerben gleicbe Tiefe besitzen u n d die Nadeln des Spiegels gleich lang sind, steht der richtig eingesetzte Spiegel vollkommen senkrecht, und gleichzeitig ist auch das reflektierte Lichtbiindel in die gewtinschte HShe ei~lgestellt und eine richtige Arbeit des Spiegels garantiert. E. P f l f i g e r , Archiv fiir Physiologie. Bd. 131. ~7 546 Heinrich Gerhartz: Die Polschuhe sind verschiebbar. Man hat es also in der Hand, die Distanz des hngriffspunktes des Sti~bchens am Spiegel yon der Drehkante zu veriindern. Es ist klar, dass dadurch ein ausserordentlich wichtiges Hilfsmittel gewonnen ist, die Vergri~sserung der Spiegelexkursionen zu variieren. In Fig. 11 ist die Wirkungsweise dieser Einrichtung schematisch tibertrieben dargestellt. In Fig. 11a ist die Distanz zwischen hngriffs- und Drehpunkt halb so gross wie in Fig. 11b. Infolgedessen ist der Verdrehungswinkel des Spiegels griisser bei gleicher Membrandurchbiegung. / /~r,...~..._j~ /., Sr jJ RL--, // i//I 1 Fig, lla. Fig. 11b. Fig. 11. Schema der Lichtstrahlexkursiou bei variabeler Distanz Stiihchen--~adel (Drehpunkt). S$. ~ Stabchen. S p . ~ Spiegel 5/. ~ ~adeln. a ~ Membrandurc~biegung, a,, ~ Wiukelahweichung bei doppeltem Hebelarm des St~bchens wie bei a. ,Ya ~ Verg~aderungder WinkelvergrSsserung. Eine Grenze ist der Variation der Lichtexkursion mittels der Polschuhverschiebung dadurch gezogen; dass die Verschiebung des Poles die magnetische Kraft, mit der der Spiegel festgehalten wird, andert. Das Optimum fi~r den Grad der hierdurch bewirkten Dampfung des Systemes muss durch Erfahrung gefunden werden, da die schnelle Messung auf Schwierigkeiten stBsst. Eine exakte Dosierung derselben ist auch so ohne weiteres mOglich, wenn man eiumhl die Spiegelexkursionen bei verschiedener Einstellung geschrieben hat. III. L i c h t f i i h r u n g . Da (ter hpparat zur gegenseitigen Kontrolle der Kurven zwei gleiche Aufnahmevorrichtm~gen besitzt, ist die Lichtfiihrung in folgender Weise ausgebildet worden. Es erschien zweckm~ssig, for beide Kurven die gleiche Lichtquelle zu benutzen, um dutch AuslBschen der Lampe beide Kurven Herzschallstudien. 547 in Koinzidenz zu bringen. Als Lichtquelle wird eine kleine elektrische Lampe (Osramlampe yon 2--4 Volt), die durch einen Akkumulator gespeist wird, verwendet. Ihre Helligkeit wird dutch einen Vorschaltwiderstand veri~ndert. Die Lampe ist in einem Holzzehi~use untergebracht. Dieses Gehause ist dutch zwei Tuben mit der Aufnahmekamera verbunden. Diese Tuben sind ebenso wie die Gebause des Apparates zur Vermeidung yon Reflexen und Lichtverlusten innen schwarz mattiert. Seitlich yon der Lampe sitzt auf diesen Rohren je ein Prisma, alas die yon der Lampe auf dieses fallenden Lichtstrahlen in das Robr refiektiert. Am anderen Ende der Tuben sind Sammellinsen angebracht, die das Strahlenbiindel auf den Spiegel dirigieren und einen scbarfen Bildpunkt auf der Einstellseheibe der Kamera entwerfen. Zwischen der Lampe und den beiden Prismen sind Spaltblenden angeordnet. StaDd Starkstrom zur Verfi~gung, so habe ich auch eine Nernstlampe verwendet. In meiner frt~heren Arbeit schrieb ieh, dass Osramlampen eigens hergestellt werden mussten. Das ist nicht mehr niitig. Die Einrichtung ist jetzt so getroffen, dass eine kleine ki~ufliche Osramlampe im hinteren Apparatgebause so senkrecht steht, dass die beiden Fiiden in gleicher Richtung mit den den Tuben anliegenden Prismakatheten lie~zen. Es fit]lt also ein nur schmaler Lichtfaden auf die Prismen. Die erwahnten S p a l t b l e n d e n - vor der Lampe in der Vertikalen im Scharnier bewegliche schwarzgestrichene sehmale Metallplatten - schneiden das Lieht des Fadens yon oben und unten ab, so dass in praxi nut ein feiner Lichtpunkt - - in Wirklichkeit natiirlich ein viereckiges Lichtbtindel - - auf Prisma und Spiegel fitllt. Auf diese Weise lassen sich mit dem Apparat Kurven von soleher Feinheit schreiben, als waren sie mit einer spitzen Nadel gezogen. Es ist das notwendig, wenn mSglichst viel aus den Kurven herausgeholt werden muss. Ein dick oder bandartig zeichnendes Lichtbtindel wtirde die wertvollsten Exkursionen verwischen bzw. sogar ganzlich verdecken. IV. K a m e r a mit Kassette. Die Kamera, in die man auf Fig. 8 v0n oben hinein sehen kann, tri~gt einen aufklappbaren Deckel. Der Boden ist durchbohrt, um die Membran beim Schliessen des Deckels nieht dutch den Luftiiberdruck zu zerreissen. Die der Membran gegent~berliezende Seite tr~gt die ausweehselbaren Kassetten. Zmn Einstellen wird eine Mattscheibe benutzt, die an den in Frage kommenden Stellen Millimeter37* 548 Heinrich Gerhartz: Teilung besitzt, um die Durehbiegung der Membran beurteilen zu k~nnen. Ist die Einstellung erfolgt, so wird die Mattscheibe gegea eine Filmkassette ausgewechselt. Ein R~hmchen dient dazu, den ablaufenden Film in der Ebene der Mattscheibe zu halten. Die benutzten Films sind 6 cm breit und bis zu 1 m lang (gebrauchliche Kodakfilms). Der Antrieb ~reift an dem Aehsenfortsatz der oberen Rolle an. Um die Holzrollen in die Kassette einsetzen zu kSnnen, sitzen an einer Schmalseite der Kassette die Zapfen an Flachfedern. Sie werden yon diesen ffir gewShnlieh gegen die Rolle gedrackt gehalten; nur beim Einsetzen tier Rollen werden sie nach aussen gespannt. Die Rilekwand der Kassette ist abnehmbar zweeks leichten Einsetzens der Filmrollen. Die obere Fihnrolle wurde friiher dutch eine biegsame Welle angetrieben. Diese Einrichtung hat sich aber nicht bewhhrt. Es ist deshalb und mn gleichzeitig eine bewe~liche Verbindung mit dem auf einem anderen Tische stehenden Uhrwerk zu haben, dieseMethode (lurch eine neue, bessere Konstruktion ersetzt worden. Sie besteht dari~, dass die Antriebswelle dutch zwei C a r d a n i ' sche Gelenke in drei Teile zerlegt ist. Das eine Gelenk ist nahe der Kamera, alas andere liegt in der Nahe der Antriebsachse am Uhrwerk. An den Zapfen der C a r d a n i ' sehen Gelenke sind kleine Filzplattehen zwisehengelegt. Diese Einrichtung hat sieh ausserordentlieh bew~hlt, insbesondere in tier Hinsicht, dass so ein absolut sicherer Schutz gegen Ubertragung yon Erschfitterungen vom rotierenden Uhrwerk her gew~hrleistet wird. Die Grundplatte des Apparates sowie des Uhrwerkes ist noch dureh dicke Filzscheiben gegen Erschfitterungen gesichert. Da die Membran senkrecht steht, reagiert sie wenig auf Vibrationen, die ihr vom Boden her zugefii~hrt werden. Sie verhNt sich in dieser ~Beziehung ganstiger als z. B. der Saitengalvanometerfaden. Das Uhrwerk wird durch einen Zentrifugalregulator in mSglichst gleichm~tssigem Gange erhalten. Die Umdrehungszahl des Uhrwerkes kann reguliert Werden. V. Z e i t r e g i s t r i e r u n g . Die einfachste Zeitfeststellung kann mittelbar erfolgen, indem bei der Aufnahme die Pulsfrequenz festgestellt wird. Man kennt dadureh die Zeitdauer einer tterzperiode auf dem Film. Herzschallstudien: 549 Wtlnscht man eine unmittelbare Zeitmarkierung, so ist die allgemein Qbliche Methode der gleichzeitigen Photographierung der Bewegungen einer schwingenden Metallfeder, die z. B. gerade 100 Schwingungen pro Sekunde haben kann, angebracht. Nati~rlich darf die durch deren Bewegung hervorgerufene Erschntterung die Registrierkurve nicht beeinflussen. Bei meiner hpparatur erfolgt die Zeitregistrierung dadurch. dass bei jeder Umdrehung des Uhrwerkes zweimal ft~r einen Augenblick eine Blende den einen registrierenden Lichtstrahl auslSscht. Die Bewegung der Blende geschieht dadurch, dass ein Schnepper in zwei Aussparun~,en einer auf der hntriebsachse sitzenden Scheibe durch Federdruck einschnappt. Die Achse d~s Schneppers geht durch die Kamerawand hindurch und tri~gt innen an einem Hebelarm~ eine kleine Blende, die sich entsprechend dem Schnepper bewegt. Durch Vergleichen mit tier Uhr wurde die Zeit zwischen zwei Marken festgestel\|t. Meistens wurde die Membran, zu der der zeitregistrierende Lichtstrahl gehSrt, unbeeinflusst gelassen, um StOrungen zu erkennen. Die Zeitmarkierung mit Blende ist wohl einfacher als die mit einer schwingenden Stahlzunge, aber nicht so exakt wie letztere. Ergel)nisse. Die Herzgerausche sind ein Ausdruck der StrSmungsverhaltnisse im Herzen, und da die letzteren einerseits yon tier pressenden Kraft, ~ndererseits yon den Widersti~nden abhiingen, ergeben sich direkte Beziehungen vor allem zu den Phasen der Kontraktion des Herzmuskels. Insofern wird aus dem analytischen Studium, namentlich tier Pathologie des Herzschalles, sicherlich ein wichtiges Hilfsmittel fur die FOrderung der Physiologie der Herzarbeit erwachsen. Ist doch alas StrOmen des Blutes das sine qua non, der wirksame und gewollte prinzipielle Effekt der Herzarbeit, yon dessert zeitlichen Verhaltnissen die Bedeutung des Herzens ft~r den KSrper reguliert wird. I c h postuliere deshalb far die Herzschallschrift eine Bedeutung, wie sie der hktionsstrom- und Pulsschrift nicht entfernt zukommen kann. Nun erkennt man ohne weiteres, class die Beziehuno'en des Herzschalles zur Kontraktion des Herzens, die mit den beiden letztgenannten Methoden einer exakten Messung wohl zug~mglich ist, der wichtigste Ausgangspunkt ft~r das Studium der sekundaren Funktionen, 550 Heinrich Gerhartz: das Klappenspiel usw.i sind. Es ergibt sich daraus, dass das kombinierte Studium von Herzschall und Elektrokardiogramm zuni~chst im u der Er0rterung, die dem Fortschritt dienen soll, stehen muss. Die Deutu.ng dieser Relation ist relativ leicht. Verwickelter scheinen mir yon vornherein die Beziehungen des Herzschalles zu den wahrnehmbaren Form~nderungen des Herzens, zum Spitzenstoss und zum peripheren Puls. Es ist den letztgenannten Spiegelbildern der Herzfunktion zum u gemacht worden, dass sie ein wenig getreues Bild yon der Herzarbeit entwerfen, ja, dass man nicht eigentlich weiss, was sie sind und bedeuten. I~un kana aber wohl kein Zweifel sein, dass die Erledigung dieser i~tiologischen Frage nicht notwendig VorbediDgung f[lr einen eventuellen Wert der genannten u ist. Es kommt hinzu, dass unsere jetzige Methodik, wie sie yon F r a n k inauguriert ist, gestattet, sich auf Kurvenbilder zu sttltzen, die yon technischen Fehlern frei sind. Ich stehe nicht an, der Identitat der Kurvenbilder wegen in dieser Hinsicht dem eben beschriebenen Apparat gleiche Vorzfige zuzumessen, wie sie F r a n k seiner Konstruktion zurt~hmt. Die Kombination der verschiedenen Untersuchungsmethoden hat in gewissem Sinne ihre Schwierigkeiten. Dass es gelingt, Herzschall und Spitzenstoss in einer Kurve aufzuzeichnen, habe ich oben gezeigt. Auch die gleichzeitige Schrift yon peripherem Pnls, Spitzenstoss und Herzschall ist nicht schwierig. Wie oben erwahnt, ist die z e i t 1i c h e K o i n z i d e n z der yon zwei Aufnahmestellen aus aufzunehmenden Kurven dadurch zu gewinnen, dass die Lampe einen Augenblick ausgeschaltet wird. Es entsteht dann auf dem Film in beiden Kurven eine Li~cke. Es kann vorkommen, dass in der einen Kurve der Ausfall grSsser als in der anderen ist. Dies beruht darauf, dass der den Lichtstrahl primi~r in seiner Hiihe einstellende Horizontalblendenspalt beiderseits ungleich hoch gestanden hat. Der eine Spiegel erupting also Lieht yore oberen~ der andere yore unteren Lampenfaden. Die HShe der Lichtabnahmestelle bedingt aber die Dauer der Intermittenz; denn die Spitze des Fadens erlischt schneller als die Basis; felner be~innt sie sphter zu gliihen, weil die Basis beim Wiederanziinden hShere Temperatur besitzt. Als Vergleichspunkte der beiden AuslSschmarken gelten also hier Mittelpunkt der einen und Mittelpunkt der anderen Intermittenz, d. h. die Koinzidenzkorrektur wird ausgedrtickt durch die i~rtliche und zeitliche Differenz beider genannten Punkte. Herzschallstudien. 55 1 Schwieriger in technischer Hinsicht ist die H e r s t e l l u n g ~'on Koinzidenzmarken auf Aktionsstromkurve einer-und Schal1-1) und P u l s k u r v e a n d e r e r s e i t s . Bisher ist keine Methode bekannt gegeben worden, die das in tier bier erforderlichen Weise ermi~glichte. N i k o l a i ~) half sich im Experiment so, dass er dutch Vagusreizung Herzstillstand erzeugte und nun das Herz reizte. Die Methode ist relativ roh und nur beschri~nkter An wendung fahig. Besser ist eine Methode, die S a m o j l o f f und Hahn benutzt haben. Sie zeichneten das Schattenbild des Hebels ihres die mechanischen Bewegungen aufzeichnenden Apparates auf dem Saitengalvanometerfilm gieichzeitig mit dem Elektrokardiogramm auf. Das liisst sich bei den modernen, wohl allein korrekt aufzeiehnenden Bewegungsschreibern nicht machen, da sie den Lichtstrahl als Hebel benutzen. Das Saitengalvanometer schreibt aber umgekehrt ein Schattenbild auf hellbelichtetem Film. Die Methoden schliessen sich also in der jetzigen Form aus. Eine Kombination ist aber dennoch mSglich. Ich bin aus finanziellen Grtinden genStigt gewesen, vorliiufig eine Einrichtung zu treffen, die nicht wie die angedeutete den hSchsten Anforderungen an Korrektheit gerecht wird, aber billio.:en Ansprachen gentigen di~rfte. Sie beruht auf der B e n u t z u n g d e r Zeitregistrierungseinrichtung des Schallschreibers zur Herbeifahrung von Koinzidenzmarken. Bei den Aufnahmen mit gleicbzeitiger Elektrokardiogrammregistriernng drackte der auf Seite 549 besehriebene Blendenschnepper, wenn er nicht in einer der beiden Scheibenkerben sass, gegen ein Kontaktsti~bchen eines Stromkreises, in den die zwischen Saitengalvanometer und Saitengalvanometerfilm befindliche, also auf dem Elektrokardiogramm markierende zweite Blende eingeschaltet war. Diese Blende sass an dem leichtbeweglichen Anker eines Elektromagneten. Sie wurde, solange Strom in dem System Eiektromagnet--hkkumulator--Schnepper floss, festgehalten und sank, wenn 1) Direkte Registrierungsmethodik ! Die zeitliche Kombination von Elektrokardiogramm und Schallregistrierung nach E i n t h o v e n ' s Methode ist nattirlich einfach zu bewerkstelligen. 2) Fr. K r a u s und G. F. N i k o l a i ~ Uber das Elektrokardiogramm unter normalen und pathologischen Verhhltnissen. Verhandl. d. Berl. reed. Gesellsch. Bd. 2 S. 228. 1907. 552 Heinric.h Gerhartz: der Anker herabfiel, ebenfalls nach unten. Dabei deckte.dann die Blende das Lichtbiindel der Saitengalvanometerlampe ab, so dass auf der Kurve eine Intermittenz entstand. Solange der Schnepper gegen das Kontaktsti~bchen dri~ckte und dieses mit einem Kontakt des Stromkreises in Verbindung blieb, war die Blende ausser Aktion. Schnappte aber der Schnepper in die Scheibenkerbe ein, so war sofort die (')ffnung des Kontaktes geschehen und die Koinzidenzmarke da. Auf diese Weise war es also mSglich~ auf beiden Films entsprechende Markierungen zu erhalten; und zwar musste der Anfang der Marke als Kontrolle gelten. In diesen Markierungen h~ttte man sich nicht zurecht finden kSnnen, wenn nicht Zeichen vorhanden gewesen wi~ren, welche als Anhaltspunkte gedient batten. Diese Markierungen wurden dutch den Ausfall von ein, zwei oder mehreren Marken erzielt. Um eine Koinzidenzmarke zum Wegfall zu bringen, war es nur nStig, den Einschnapphebel, bevor er an die Kerbe kam, festzuhalten, so class er nicht in die Kerbe gleiten konnte. Es fragt sich nun, ob tatsi~chlieh keine Zeitdifferenz zwischen dem Beginn der beiden Marken besteht. Um dies zu untersuchen, wurde auf die Filmachse, paralIel zu tier mit den beiden Zeitmarkierungskerben versehenen Scheibe, eine grosse kreisrunde Scheibe aufgesetzt und nun langsam die Achse gedreht. Es wurde dann der Moment beobachtet, warm der Kontakt verloren ging~ wann der erste Lichtschimmer der SchMlregistrierapparatmarke erschien, und wann die volle Lichtmarke bei dem letzteren Apparat auftrat. Es zeigte sich dabei, dass der erste Lichtschimmer der Zeitmarke und der Kontaktverlust gleichzeitig erfolgten. Infolgedessert wurde der Ausrechnung der Kurven des Schallschreibers die maximale matte Liehtmarke zugrunde gelegt. Durch solche Marken wird der ganze Film in einzelne Zeitperioden, die yon Beginn der erscheinenden Marke bis zum Wiederbeginn der ni~chstfolgenden Marke reichen, zerlegt. Die Ausziihlung yon 33 solchen Zeitperioden ergab far 32,91 mm mittlere Periode einen Zeitwert yon 1,3364 Sek.; d. h. 1 mm entspricht 0,0406 Sek. Bei dieser untersuchung fii,llt in die huaen, dass die ersten Zeitperioden karzer sind als die sp~teren. Dies hangt damit zusammen, dass der Film sich zfi Anfang auf eine leere Rolle aufwickelt, nach und nach aber auf eine durch aie Filmauflagerung Herzschallstudien. 553 immer dickere. Es ist deshalb, wenn man Zeitperioden vergleichen will, notwendig, diese auf einen einzigen Wert zu beziehen 1). Nun entsprechen aber auch die Elektrokardiogrammperioden, der immer vorhandenen geringen Unregelmi~ssigkeit der Motorumdrehung wegen, weder einander, noch, da sie auf eine andersgestaltete Filmrolle aufgewickelt wurden, den Zeitperioden des anderen hufnahmeapparates. Es ist also eine doppelte Reduktion auf einen Normalwert notwendig. Diese Reduktion habe ich zeichnerisch vorgenommen. In jeder Kurve wurden yon jedem beachtenswerten KurvenpuDkte die Ordinate zur Kurvenabszisse, die dem Filmrand parallel lief, gezogen. Dann wurde auf einem Zeichenbrett zwischen beide Films und parallel zu der auf ihnen aufgezeichneten Abszisse, bzw. zu ihren Randern, ein Streifen Millimeterpapier befestigt, auf welches, parallel zu den Filmri~ndern, eine doppelte Normalzeitperiode - - doppelt, um eine grSssere Genauigkeit beim Zeichnen zu erzielen - - ( = 2-32,91 65,82 mm ~- 1,3364 Sek.) abgetragen wurde. Wurden nun die Anfangspunkte der Kurvenabszissen mit dem Anfang der Normalperiode verbunden und wurde ebenso mit den Endpunkten entsprechend verfahren, so waren zwei Proportionensysteme vorhanden, in denen die Kurvenpunkte auf der Normallinie so gefunden wurden, dass von den entsprechenden Kurvenabszissenpunkten zum Schnittpunkte der Anfangs- und Endlinien Gerade gezogen wurden. Auf diese Weise wurden auf eine Linie, auf der 1 mm ~-0,0203 Sek. entsprach, alle markanten [unkte der beiden Kurven eingetragen und konnten so direkt verglichen werden, falls nicht zwischen Zeitund Sehalllinie des Schallaufzeichnungsapparates eine Koinzidenzkorrektur vorzunehmen war. Die U n t e r s u c h u n g e n , die mit dem oben beschriebenen Schallregistrierer angestellt wurden, und bei denen ich wiederum von Herrn R u h m e r in der tatkri~ftigsten Weise unterstatzt wurde, ergaben, wie zu erwarten stand, far die einzelnen untersuchten Individuen voneinanJer abweichende Resultate. Bei der Persbn, 1) Diese etwas unvollkommene, aber hier wegen der Notwendigkeit einer doppelten Reduktion mittels Zeichnung nicht besonders lfistige Methodik kann mit Leichtigkeit dutch ein Band ohne Ende ersetzt werden. Dabei fi~llen die Perioden gleich aus. 554 Heinrich Gerhartz: die am griindlichsten yon mir untersucht wurde, und auf die sich deshalb, wo nichts anderes gesagt ist, die folgenden hngaben beziehen, bei einem 30jahrigen gesunden Manne A., wurde als Mittelwert yon insgesamt 22 Ausz~hlungen filr den I. Herzspitzenton eine Schwingungszahl yon 57,4 Schwingungenl), fi:lr den II. eine solche yon 69 Schwingungen pro Sekunde erhalten. Der II. Ton war also, wie auch die Auskultation des Herzens deutlich erkennen liess, etwas hOher. In die Sprache der Musik umgesetzt~ entsprieht der I. Spitzenton am ehesten Contra-B. [57,6 Schwingungen~)], der II. dem grossen Des (69,1 Schwingungen). Differenzen zwischen den einzelnen Messungen kSnnen deshalb nicht vermieden werden, weil die geringen Febler, die bei der Auswertung der Kurven unterlaufen, sich bei der Aufrechnung auf den Sekundenwert vielfach multiplizieren. Die Differenzen, die bei verschiedenen Personen in der Frequenz gefunden werden, sind recht erheblich. Ich selbst verfiige bei verscbiedenen Personen bisher tiber Messungen, die yon 34 bis 74 Schwingungen for den I., yon 35 bis 82 Schwingungen far den II. Spitzenton reichen. Die Zahlen bezieben sich auf ein Dutzend Fi~lle, far die ich als Mittelwert 55,2 Sehwingungen ftir den I., 62~4 fiir den II. Ton time. Musikalisch ausgedrfickt, finde ich also for die beiden HerztSne folgende mittlere Beziehung: Bei derselben Person ist der I. Herzton an der ..... Spitze recht konstant. Die oben erwi~hnten, an einem ~)~- 1 - ~ . ~ recht grossen Material gewonnenen Zahlen bezogen ~ ~T sich auf den 6. M~trz 1909. Am 19. November 1908 I Ton II. Ton Co,tr~-Ai~Co,~r~-Hi~ war der I. Herzton desselben Individuums yon 55,5 02,5 mehreren Personen auskultiert und unabhangig auf A_~ am Harmonium festgelegt worden. Es war also sowohl Identiti~t mit dem registrierten Schall vorhanden, wie die Frequenz zu den versehiedenen Zeiten dieselbe geblieben. Diese Konstanz des I. Herztones erklart sich leicht, wenn wir in ihm die Muskelkomponente als den zu seinem Zustandekommen wesentlichsten Faktor ansehen. Ich habe versucht, in der A ufz e i c h n u n g d e s M u s k e l t o n e s einen weiteren Anhalt fiir diese Anschauung zu gewinnen. Zusammen mit A. L o e w y bin ieh so vorgegangen, dass ein Trichter auf die Beuger des Unterarmes auf1) Doppelscltwingungen, wie tiblich. 2) Physikalische 8timmung. Herzschallstudien. 555 gesetzt und mittels Schlauch mit der Zuleitung des Schallregistrierers verbunden wurde. Es werden also bei dieser Methode alle durch, die Kontraktion tier uater dem Trichter liegenden Muskelmasse entstehenden Bewegungen dem Apparat zugeftihrt. Dabei entsteht eine Kurve aus Impulsen verschiedener Frequenz, und zwar superponieren sich recht frequente Schwingungen auf sehr lantzsam verlaufende, so dass eine Differenzierung keine Schwierigkeiten bietet. Man wird wohl kaum fehlgehen, wenn man in den frequenten Oszillationen, die mit dem Anstieg der Kontraktionskurve beginnen und mit ihrem Ende abschliessen, Muskelschwingungen sieht, die dem Muskelton zugrunde liegen. Damit stimmt ibre Frequenz durchaus iibereinl denn ich finde ffir den von den Unterarmbeugern abgenommenen Ton im Mittel 56,3 Schwingungen pro Sekunde. Die Differenzen gehen ffir verschiedene Muskeln (sechs Aufnahmen) von 48--60 Schwingungen; man wird hierauf keinen Wert legen dfirfen, zumal sie sich nicht auf dieselbe Person beziehen. Ich babe mich fiberdies yon der Richtigkeit der mittleren Schwingungszahl durch die Kontrolle am Instrument fiberzeugt. Ich selbst finde in meinen Kurven auch eine leidliche Obereinstimmung zwiscben dem I. Spitzenton und dem I. Arterienton. Abgesehen davon, dass diese Ubereinstimmung allein keinen Beweis f[lr eine kausale Beziehung abgeben kSnnte, verfiige ich in dieser Hinsicht noch fiber zu wenig Material, um die Identitat der Frequenz der Schwingungen ft~r eine allgemeine Erscheinung halten zu dfirfen. Untersuche ich die bisher in der Literatur fiber diesen Punkt gebrachten hngaben, so finde ich bei E i n t h o v e n ( P f l [ l g e r ' s Arch. Bd. 120) ffir den I. Spitzenton (vgl. S. 531) 87,7 Schwingungen, I. Aortenton . . . . 45,4 ,, (I. Pulmonalton 35,6 , ), also eine totale Differenz, bei W e i s s (1. c. Anm. 1 S. 519) dagegen ffir den I. Spitzenton (Fig. 12 S. 360 ft., 1. c.) 94,0 Schwingungen, I. Aortenton . . . . . . . . . 81,3 I. Pulmonalton . . . . . . . . 64,8 ,, also Zahlen, die schon eher einander sich niihern. Ich maehe noch darauf aufmerksam, dass, wie ich S. 527 berechnet babe, bei E i n t h o v e n I. Spitzenton und I. Aortenton in 556 Heinrich Gerhartz: demselben Augenblicke beginnen; allerdings stimmt diese Angabe nicht zu dem mit der frtiheren Methodik E i n t h o v e n' s erhaltenen Befund. Es dtirfte verfraht sein, schon jetzt die Frage nach der Genese des I. Arterientones entscheiden zu wollen. Die Frequenz der Sehwingungen des II. H e r z t o n e s ist gr(isserem Wechsel unterworfen als die des ersten. Diese Tatsaehe, die ffir die Erkli~rung seiner Entstehung nieht unwiehtig ist, wird am versfitndlichsten werden durch die Untersuchung seiner Veranderung infolge energischer Arbeitsleistung. Ich habe diese Beziehungen an einem Mteren gesunden Manne D. studiert und finale aus einem Material yon 18 Auszithlungen als Mittelwerte Fflr die R u h e p e r i o d e ...... Far die Z e i t g l e i c h n a c h Arbeitsleistung ........ tier I. Ton IL Ton 48,7 Schwingungen 60,1 Schwingungen 46~7 75,5 j , , Wiihrend die Frequenz des I. Spitzentones also gleichblieb, stieg die des II. nicht unerheblich. Die Frequenzsteigerung des II. Tones erfolgte um 25,6 % der Schwingungszahl; die ErhShung der Herzperiodenfi'equenz betrug zufMlig genau ebe~soviel, 26,3 %. Die Schallschwingungen des II. Tones sind also in der Kurve prim~tr vorbanden, da sie sieh mit ihrer Lange ungefi~br parallel verschieben. Die Verkflrzung der Herzperiode betraf in dem in Rede stehenden Falle die Diastole des Herzens, wie es ja die R egel ist. Rechnen wir die Spanne vom Beginn der Vorhofzacke des Elektrokardiogrammes bis zum Ende der Finalschwankung als Systole, den Rest als Diastole, die Ventrikelsystole yon dem Beginn der Kammerzacke bis zum Ende der Finalschwankung, so stehen sich zeitlieh gegentiber: Ruhe Arbeit 0,48 Sek. 0,32 , 0,19 ,~ 0~46 Sek. 0,31 ,, 0,07 , \| Gesamtsystole . . . . . . . . . . Kammersystole . . . . . . . . . . Diastole . . . . . . . . . . . . . I ] Der I. Spitzenton hiflt sich also an die Systole, der II. an die Diastole des Herzens, und zwar erscheint der IL Herzton abhitngig in seiner Frequenz yon der Geschwindigkeit, mit der sich die Herzschallstudien. 557 Schwingungen des elastischen Systems des hnfangsteiles der Aorta vollziehen. Das erkli~rt auch die vollkommene Identitgt der Schwingungszahl vom II. Spitzen- und II. Karotis-bzw. Aortenton, die sowohl W e i s s als ich selbst gefunden habe. Bei E i n t h o v e n finde ich eine diesbezagliche Ubereinstimmung allerdings nicht. In seiner letzten Arbeit ( P f l i l g e r ' s Arch. Bd. 120) teilt er Spitzenton, Aorten- und Pulmonalton derselben Person mit. Ich babe aus den dort mitgegebenen Tafeln die Schwingungszahlen selbst berechnet and finde: II. Spitzenton 72,3 Schwingungen pro Sekunde, II. Aortenton 46,9 ,, ,, ,, II. Pulmonalton 48,4 :, ,, , Bei W e i s s Zahlen far den dagegen (Fig. 10--13) sind die entsprechenden II. Spitzenton 91,9 Schwingungen pro Sekunde,. II. Aortenton 90,2 . . . . . , II. Pulmonalton 71,7 , ,, ,, II. Spitzenton und Aortenton besitzen also vollkommen identische Schwingungsfrequenz. Ich finde das auch wohl schon durch die blosse Auskultation far genfigend festgestellt. Die D a u e r d e r H e r z t 6 n e land ich ftir den am genauesten untersuchten Fall A i m Mittel yon 16 Einzelauszi~hlungen zu 0,094 Sek. ftir den I., und 0,075 Sek. far den II. Spitzenton. Der II. machte also 79,5 % des I. aus. Als Mittel aller yon mir berechneten F~lle finde ich einen etwas hSheren Wert far den I. Ton: 0,1t Sek., for den II. 0,072 Sek. ~ 65,3 % des I. Tones. In Tabelle 7 habe ich T a b e l l e 7. tterzschallcharakteristika. Dauer der Herzperiode Einthoven (Pfl~ger's Arch. Bd. 117) Frank (Hund) Weiss ~ e r h a r t z (Mensch) ,, (Hund) Mittel (Mensch) . . . . . . . . i Herztondauer Schwingungszahl der Herzt6ne I. Ton ]II. Ton I. Ton III. Ton 0,139" (0,064") 0,41 " 0,072" 0,068 " 0,835 " 0,110" 0,328" 0,076 { " 0,107 " t 0,079" (0,042") 0,044" 55,5 i 45,4 0,071" 77,0 i 86,1 0,072,, 55,2 62,4 0,045"i 105,0 111,0 0,074" 57,2 65,3 558 Heinrich Gerhartz: die Zahlen, die ich naeh den Kurven der iibrigen Autoren berechnet habe, daneben gestellt. Im ,Mittel" aller hier beriicksichtizten, allerdings wohl kaum vergleichbaren Zahlen ergibt sich der II. Herzton (0,074 Sek.) zu 69,2 ~ des I. (0,107 Sek.) beim Menschen. Beim H u n d babe ich einmal die Dauer des I. Spitzentones zu 0,076 Sek., die des II. zu 0,045 Sek. festgestellt; die Herzperiode dauerte 0,328 Sek. Der II. Ton dauerte also nur 5 9 % des I. an. Die Schwingungszahlen waren: fi~r den I. Ton 105 Schwingungen, ffir den II. Ton 111. Wie man aus der Tabelle 7 ersieht, entspricht die yon mir gemessene Dauer vollkommen derjenigen, die F r a n k ange~eben hat. Bei einer Steigerung der Herzfrequenz~ wobei die Periode yon 0,43 Sek. auf die oben beriicksichtigte yon 0,328 Sek. herabging, fiel gleichzeitig die Schallphase, wie ich die Spanne yore Beginn des L bis zum Ende des II. Tones nennen will, yon 01202 Sek. auf 0,16t Sek. herab. Eine gleiche Verkiirzung der Tonphase habe ich auch helm Menschen nach der A r b e i t s 1e i s t u n g gefunden. Ieh land z. B. in der R u h e ftir den I. Spitzenton eine Dauer yon 0,115 Sek., ,, , ,, , ,, II. ,, ,, , i, 0,064 , naeh der A r b e i t ,, ,, I. ,I 1, , I, 0,132 ,, , ,, .... ,, II. , ,i ,, , 0,057 I, Bei der bestuntersuchten Person A wurde der S p i t z e n s t o s s yon der linken Kammer gebildet, wovon man sich mittels der ROntgendurchleuchtung bequem tiberzeugen konnte. Die Spitzen-stosskurve wurde hier also yon den Formveriinderungen der linken Herzkammer hervorgerufen. Fiir solche FMle diirfte sicher das zutreffen, was C h a u v e a u und M a r e y bei Tieren feststellten: dass der Beginn des systolisehen Anteiles des Spitzenstosses und des Kammerdruckes genau gleiehzeitig erfolgen. Wit sind also bier in tier Lage, in der kardiographischen Kurve den Beginn der Kammermuskelzusammenziehung zu finden. Die Feststellung des Systolebeginnes ist mitunter recht schwierig; und ich schiebe es vor allem hierauf, wenn die Kardiographie heute in Misskredit gekommen ist. Die Schwierigkeit der Analysierung ist in der Variabilitiit der 'kardiographischen Kurve, die auf ver~tnderte Aufnahmebedingungen zurt~ekzufuhren ist, gelegen. Kontrolliert man die Spitzenstosskurve mit der Elektrokardiogrammoder Karotiskurve, so wird es kaum vorkommen, dass die Einzelerhebungen der Spitzenstosskurve falsch gedeutet werden; man ent- 559 Herzsehallstudien. deckt dann unter der Maske die Teile, die man sucht, und ist iiberrascht yon der Konstanz ihrer Charakteristika. Es ist hier nicht der Ort, den Ursachen der hier vorliegenden Schwierigkeiten nachzusp~ren, sondern ich begni~ge mich damit, die B e z i e h u n g e n der w i c h t i g s t e n P u n k t e d e s S p i t z e n s t o s s e s zum H e r z s c h a l l festzulegen. Ich finde im Mittel zahlreicher~ sich auf die oben wiederholt genannte Person A beziehender Messungen den Beginn des I. Spitzentones 0,012 Sek. nach dem Beginn des systolischen Anstiegs des Kardiogrammes, den Beginn des II. Tones 0,251 Sek. nach dieser Marke. Das Ende liegt also nach dem oben Gesagten ft~r den I. Ton 0,106 Sek., far den II. 0,326 Sek. nach dem Anfang der Kammererhebung des Spitzenstosses. Da der Abstand des letzten Punktes bis zur Inzisur 0,232 Sek. ausmachte, beginnt der II. Herzton 0,019 Sek. nach der Inzisur des absteigenden Sehenkels des Spitzenstosses. Kurze Zeit nach den Aufnahmen, bei denen die hier o:enaDnten Werte ermittelt wurden, wurde ffir eine etwas andere Herzperiode das Intervall zwisehen Anstieg der Spitzenstoss-Kammersystole und Karotispulsanstieg zu 0,107 Sek. gemessen. Es ergibt sich daraus ein Weft yon 0,095 Sek. far die Spanne zwischen Beginn des I. Spitzentones und Karotispulsanstieg. Tabelle 8. Bezieliungen tier Herztiine zu Karotispuls und Spitzenstoss. Herzperiodendauer E i n t h o v e n u. Geluk Einthoven (Pfltiger's Arch. Bd. 120) Weiss . . . . . . . F r a n k (Hund) . . . . Gerhartz: A .... ~, ,~ B . . . . C . . . . 0,78" Beginn des I. Tones vor der Kammererhebung des Spitzenst0~ses 0,014" naeh der ] Kammer- ] vor dem erhebung ] Karotisd~s I pulsSloitzenanstieg stossos -- 0,149" 0,78" 0,16" 0,41 " 0,721 " 1,034" 0,75" 0.085 "1 ) L0,059 "j 0,095" 0,133" 0,042" --- 0,012" I 0,02" gleichzeltig 1) Unkorrigierter Weft (vgl. Originalarbeit). Beginn d. 1I. Tones nach der Karamererhebung des Spltzenstosses 0,315" 0,25" 0,517" 0,257" nach dem Karotispuls"anstieg 0,18" 0712" 0,20" [0,106 "] 0,21 " 0,497 " 072 I4 " 560 Heinrich Gerhartz: Die genannten Werte sind keine Konstanten. Die wenigen Fiille, ilber die mir bisher einwandfreie Zahlen zu Gebote stehen, Weichen nicht unerheblich voneinander ab, und auch die Differenzen, die sich in den Angaben der Literatur finden, sind recht gross. Ich stelle alle Zahlen, filr deren Berecbnung sich in den Arbeiten Anderer Unterlagen finden, in Tabelle 8 mit den yon mir gefundenen Zahlen zusammen. Will man Wert auf eine Mittelung der dort berechneten Zahlen legen, so wfirde alas Resultat lauten: Der I. Herzton beginnt im Moment der systolischen Kontraktion, und zwar 0,111 Sek. vor dem Anstieg der Karotis. Der II. Spitzenton folgt 0,237 Sek. nach dem Karotisanstieg. Diese Werte hatten filr den Mensehen Geltung. Ieh selbst halte es nicht ft'lr richtig, so divergente Ergebnisse zu mitteln. Ich gehe nun ilber zur Besprechung der B e z i e h u n g e n d e s H e r z s c h a l l e s zum h k t i o n s s t r o m des Herzens. Zu der Aufzeichnung der Elektrokardiogramme, die in der welter oben schon erwhhnten Weise in Koinzidenz mit der Schallkurve gebracht wurden, diente ein E d e l m a n n ' s c h e s Saitengalvanometer (Fadenspannung 12 o). In den Stromkreis des Saitengalvanometers war ein Kondensator eingeschaltet. Als Elektroden dienten zwei in Wannen~ die mit lauwarmem Wasser gefffilt waren, tauchende Kupferbleche. Es wurde yon linker und rechter Hand abgeleitet. Die Erhebungen im Elektrokardiogramm des Falles A, die hier in erster Reihe interessieren, waren relati~ niedrig. Im Mittel sehr zahlreieher Aufnahmen wurde die Ordinate der Vorhoferhebung zu 41% , die der Finalschwankung zu 5 9 % der Ventrikelzackenhiihe gefunden. Diese Proportionen waren jedoch nicht konstant. Bei 2[ hintereinander folgenden Aufnahmen variierten die Zahlen fiir die Vorhofzacke yon 17 % - - 5 4 % tier zugehSrigen Kammerzacke, die fiir die Nachschwankung yon 50% --78% der KammerhShe. Die H0he dei~ letzten Erhebung war also viel konstanter als die der ersten; denn die der u schwankte hier bei derselben Person um das Dreifache. Wesentlicher sind die Abszissenwerte. Die Herzperiode yon 9 VentrikelhOhe zu VentrikelhShe gemessen, betrug 0,7213 Sek. In dieser Periode machte der Abstand des Vorhoferhebungsbeginnes bis zum Anstieg der Kammerzacke 0,115 Sek. aus. Die Vorhoferhebung dauerte 0,07.1 Sek. Von ihrem Beginn Herzschallstudieu. 561 bis zur H~)henordinate wurden 0,04 8ek. gemessen. Die beiden Anteile tier Vorhoferhebung verhielten sieh also wie 56,7 % :43,3 %. Die Ventrikelerhebung dauerte 0,057 Sek., ihr erster Tell bis zur HShe 0,025 Sek. Demnaeh ist alas Verh~ltnis der beiden Ano teile (0,025 8ek. und 0,0325 8ek.) ~ 43,26~ : 56,74~ Es verh~lt sieh hier also genau umgekehrt wie beim Vorhof. Die Finalsehwankung wurde zu 0,097 Sek. Dauer ermittelt. Ihre erste Periode mass 0,055 Sek., ihre zweite 0~ Sek., so dass also die h~ehste Erhebung der Naehsehwankung diese im Verh~ltnis 56,8~ (I):43,2 ~ (lI) teilte. Wie man erkennt, ist die Verteilung der beiden Absehnitte bei der Nachsehwankung und beim Vorhof genau die gleiehe. Es mtissen also wohl bei der Bildung der beiden Zaeken aueh gleiehe Gesetze obwalten, und zwar nur bei ihnen; denn bei der Ventrikelzaeke verh~lt es sich anders. Zuf~lligkeiten k~nnen kaum eine Rolle spielen; denn die Zahlen sind Mit~elzahlen aus 21 Einzelausmessungen, und zu versehiedenen Zeiten gemaehte Aufnahmen lassen die gleiehen interessanten Beziehungen erkennen. Ventrikelzaekenbeginn und Anstieg der Finalsehwankung waren 0,1547 Sek. auseinander. Da die Marken in den beiden zeitlieh in Koinzidenz zu bringenden Kurven zusammenfallen, ist es nur nStig, die Verz6gerung, die die Spitzenstoss- (und ,Sehall")-Impulse im Zuleitungssehlaueh erleiden~ in Ansehlag zu bringen. Diese betrug in dem 39,5 em langen und 0,75 em weiten Rohr 0,0177 Sek. Sie wurde in der Weise ermittelt, dass der Spitzenstoss gleiehzeitig beiden Membranen des Apparates zugeleitet wurde, der einen Aufnahmemembran aber in einem 1 m l~ngeren Sehlaueh (T-Stack hinter dem Aufnahtnetriehter). Der angegebene Wert stellt das Mittel aus zahlreiehen Messungen dar. Auf diesen Unterlagen baut sieh die Beziehung zwisehen Elektrokardiogramm, Spitzenstoss und tterzschall so auf, wie es in Fig. 12 zur etwas sehematischen Darstellung gekommen ist. Wie man sieht, ist die Ventrikelzaeke vSllig abgelaufen, ehe die Latenzperiode zu Ende ist, d.h. es verh~lt sieh genau wie beim quergestreiften Musket. Ieh finde also hier f~r den Mensehen nieht alas, was S a m oj I o f f 1) auf Grund seiner kombiniert-kapillarelektrometrisehen trod Suspensionsexperimente yore Frosehherzen ausgesa~t hat~ 1) A. S amoj t off, Beitr~ge zur Elektrophysiologie-des Herzens. Arch f. Anat. u.) Physiol. 1906 (Suppl.) S. 207---229. E. P f l f i . g e r , Archly fiir Physiologie. Bd. 131. 38 Heinrich Gerhartz: 562 Beginn der Vorhofzacke des Elektrokardiogrammes und Beginn der Vorhoferhebung des Spitzenstosses sind 0,058 Sek. auseinander. F[~r die Kammerzusammenziehung wurde eine Latenz yon 0,049 Sek. gefunden. Die Finalschrwankung beginnt 0,046 Sek. vor der SpitzenstosshShe; ihre ttShe fi~llt mit der der Finalschwankung zusammen. Das Ende der Finalschwankung liegt 0,029 Sek. vor der Inzisur. Der I. Herzto~ f~ngt 0,061 Sek. nach dem Beginn der Ventrikelzacke an; sein Ende flillt mit dem Beginn der Finalschwankung zusammen und liegt kurz vor der HOhe der Spitzenstossexkursion. Die Finalschwankung ist abgelaufen, wenn der II. Ton erschallt; er liegt 0,048 Sek. spi~ter. ,J I I ; ! , ~o ~; -.1 , ,I L , {o, ~21 ~. i t ",, - f" . ,r i , t , , I, ' I [ f I ~ ,, ,~ h Fig. 12. Beziehungzwischen Herzschall-, Spitzenstoss-, Karotiskur~e und Elektrokardiogramm, .schematisch gezeichnet nach eigenen Untersuchungen an demselben Individuum. Der wichtigste Wert ist unstreitig die Latenzdauer zwischen den beiderlei Kammererhebungem Die Vorhoferhebung des Spitzenstosses kann unm0glich die Vorhofsystole bedeutea; denn wird in das Spitzenstoss-Elektrokardiogrammschema die Druckkurve K a h n ' s (1. c. Anm. 1 S. 527) in der Weise, dass das EIektrokardiogramm als Unterlao~e filr die Berechauug dient (vgl. Fig. 5) eingefilhrt, so zeigt sich, dass der V0rhofdruck das Maximum erreicht hat, wenn im Kardiogramm die Erhebung anfi~ngt. Unter der Voraussetzung, class die Formveri~nderung des Herzens so welt vorgeschritten ist, dass das linke Herz der Stelle, yon wo aus die Pulsati0neu geschrieben werden, anliegt - - falls es das nicht schon immer t u t - - , kann nur der Eintritt der Blutmasse in die Kammer die Erhebung veranlassen. Es kann hier also yon einer Latenzbestimmung keiae Retie sein. Herzschallstudien. 5(}3 Gibt nun wirklich die Kammererhebung des Spitzenstosses den Beginn der Kammersystole an, so miissen die erstere Erhebung und der Beginn des Kammerdruckanstieges yon der Erhebung zur Elektrokardiogramm-Ventrikelzacke gleiehweit entfernt sein. Beim Menschen ist eine derartige Feststellung unmi~glich. Es li~sst sich abet am Tierversuch meine Latenzdauer auf ihre Richtigkeit hin kontrollieren. Kahn registrierte beim Hunde gleichzeitig die Druckanderungen im rechten Ventrikel und das E!ektrokardiogramm. Der Druck wurde mit dem G ad'sehen Blutwellenschreiber aufgezeichnet und die LeitungsverzOgerung entsprechend in Rechnung gesetzt. K a h n land folgendes: ,,Die Vorhofzaeke E i n t h o v e n ' s fi~llt mit der Vorhofzacke der Ventrikeldruckkurve zusammen oder geht ihr etwas voran. Die Ventrikelzacke lauft vor der Tiitigkeit des Ventrikels fast vollsti~ndig ab. Sie verschwindet wahrend der Anspannungszeit des Ventrikels und ist zur Zeit des Beginnes der Austreibungszeit voriiber. Die Nachschwankun~ fallt noch in die Austreibungszeit; sie findet sich regelmi~ssig innerhalb jenes Zeitabschnittes, in welehem so hiiufig die Ventrikeldruckkurve ein Plateau aufweist." Die genauen zeitlichen Verhi~ltnisse habe ich sowohl naeh K a h n ' s photographiseher Darstellung wie nach seiner Koinzidenzfigur 16 berechnet. Sie sind fiir beide Abbildungen identiseh, und zwar bemisst sich das lntervall zwischen Vorhofdruckbeginn und Vorhoferhebung im Elektrokardiogramm zu 0,038 Sek., die Spanne zwischen der Kammererhebung in beiderlei Kurven zu 0,0518 Sek. 1), ein Wert, der mit dem oben fiir den Menschen von mir gefundenen (0,049 Sek.) vollkommen geniigende Ubereinstimmung besitzt. E s ist also mSglich, beim Menschen mit Sicherheit die L a t e n z d e s H e r z m u s k e l s zu b e s t i m m e n . Ich mSchte diese Feststellung namentlich fiir die Pathologie des Herzens ftir einen grossen Gewinn halten, da wohl anzunehmen ist, dass der Herzmuskel yon pathologischen Zustiinden ebenso stark in seiner Latenz beeinfiusst wird wie tier periphere Muskel. In Figur 5, auf die ich bier noch einmal verweise, sind die Angaben K a h n ' s dazu benutzt worden, meine Erfahrungen iiber den Herzschall des Hundes in die tibrigen Erscheinungen des Ablaufes der Herztatigkeit einzureihen. Elektrokardiogramm, Druck1) Nikolai (1. c. S. 228) gibt 0,06 SeE far den Hund an. 38 * 564 Heinrich Gerhartz: kurve und Latenz der beiden sind von K a h n entlehnt; der Druckanstieg der Kammer ist mit der entspreehenden Spitzenstosserhebung identifiziert, alles iibrige naeh meinen eigenen Erfahrungen eingezeichnet. Meine Untersuchungen erweisen eklatant die Richtigkeit der Martius'schen 1) Deutung des Spitzenstosses, d.h. ,,der ansteigende Schenkel des Systolenanteiles entspricht genau der u der Systole [d. h. der Zeit, welche der sich kontrahierende Herzmuskel braucht, ,,urn seinen Inhalt unter den geforderten Druck zu bringen, bzw. welche vergeht zwischen dem Beginn der Systole und der ErSffnung der Semilunarklappen(~], der absteigende Schenkel desselben fi~llt mit der Austreibungszeit des Blutes zusammen% Die Systole dauert yore Anstieg tier zweiten Erhebung his zur Inzisur; sie sehliesst ab mit dem Ende der Finalschwankung. Sie ist die Resultante der Kraft, mit der die Kammer sich zusammen.zieht, und dem Widerstande, den der Aortendruek leistet. Infolgedessen steht aueh zu erwarten, dass sich direkte Beziehungen ergeben werden zu der Dauer der Strecke, die yon der Inzisur his zum Beginn des II. Tones reicht. Die Ubereinstimmung mit dem Befunde K a h n ' s ist so gross, dass man die Beschreibung, die dieser Autor von dem Koinzidenzbilde des Druekes und des Elektrokardiogrammes gibt, kurzweg auf Spitzenstoss und Elektrokardiogramm, was die Kammer angeht, iibertragen kann: ,,Die Ventrikelzacke lhuft vor der Tatigkeit des u fast vollsti~ndig ab. Sie verschwindet wi~hrend der Anspannungszeit des u und ist zur Zeit des Beginnes der Austreihungszeit vori~ber. Die Nachschwankung f~llt noch in die Austreibungszeit." Ich time es auffallend, dass die Latenz bei der Vorhofkontraktion eine andere ist als bei dem Ventrikel, wenn man nach deu herrschenden Anschauungen in der Vorhoferhebung ein Analogon tier Ventrikelzaeke sieht, wie ich es ja oben - - mangels besserer, sieherer Erkli~rung- selbst dargestellt h a b e - - : auffallend einesteils, weil meines Wissens so grosse Differenzen in der Latenzdauer zwisehen gleiehartigen Muskelteilen sonst nirgends vorkommen, dam~ auch, weil mitunter nach K a h n ' s Beobachtungen die ,,Vorhofzacke Ei n t h o v e n's mit der Vorhofzacke der Ventrikeldruckkurve 1) Mar tiu s ~ Graphische Untersuchungen i~ber die Herzbewegung. I. Zei~schr. f. klin. Med. Bd. 13 S. 346. 1888. Herzschallstudien. 565 zusammenfallt", also der Charakter tier Latenz verloren geht. Ich erinnere im Anschlusse hieran, dass die Formgestaltung der Vorhoferhebung, wie ich oben berechnet habe, im Elektrokardiogramm durchaus den Charakter tier Fiualschwankung tragt, mit der der Kammerzacke aber nicht im miudesten Ahnlichkeit aufweist. Nun koinzidiert die Finalschwankung. wie K a h n ' s und meine Erfahrungen tibereinstimmend bekunden, mit der Zeit, in der das Blut in die grossen Gefasse abstrSmt; far die Vorhofzacke ist aber auch die Koinzidenz mit der StrOmung aus den VorhSfen in die Kammern ausserordentlich wahrscheinlich. Sollte in diesen Beziehungen nicht die Genese der beiden Elektrokardiogrammerhebungen zu suchen sein? Weitere Beweise fiir die Richtigkeit meiner obigen Darstellung lasseu sich aus der Pathologie schSpfen. Fig. 1 zeigt das Schallbild einer Mitralinsuffizienz. Man erkennt ohne Schwierigkeit, dass der Beginn des Gerausches in den Beginn des systolischen Kardiogrammanstieges bzw. einen sehr geringen Zeitabschnitt sparer fatlt, auf denselben Moment, in dem der I. Spitzenton beim normalen Menschen erschallt, obwohl die Genese beider Phhnomene durchaus verschieden ist. Das prasystolische Gerausch der MitralsteDose dagegen f~ngt mit dem Anstieg der vorhergehenden Erhebung, wo diese auso~epragt ist, an [No. VI~)]. Wilrde dieser Punkt des Spitzenstosses den Beginn der Kontraktion des Vorhofes markieren, so w~re das unverst~ndlich. Das Gerausch koinzidiert mit dem Momente, in dem in der Norm das aus dem Vorhof in den Ventrikel einfliessende Blur die erste Spitzenstosserhebung veranlasst Das ~)bersichtsbild Fig. 12 lehrt, class sich das Intervall zwischen den beiden HerztSnen nicht mit tier Systole im hergebrachten Sinne deckt. Die Systole kommt im Spitzenstoss zum Vorsehein an einer eindeutigen Stelle und reicht bis zur Inzisur, dauert also 0,232 Sek.; der Zeitraum zwischen den Anf~ngen der beiden T6ne dauert abet 0,238 Sek. Die im Spitzenstoss zutage tretende ,,Systole" stimmt in der Dauer auch nicht mit tier Elektrokardiogrammsystole aberein; denn diese macht 0,252 Sek. aus. Ich schlage vor, um die Verwirrung, die infolge tier verschiedenen Auslegung des Ausdruckes ,,Systole" nun schon seit langem herrscht, nicht 1 In Fig. 3 nicht sichtbar. 56(~ Heinrich Gerhartz: noch griisser zu machen, fiir die Schatlkurve neue, nichts praejudizierende kusdriicke zu nehmen, und zwar so zu benennen: Schallphase = Zeit yore Beginn des I. Herztones his zmn Ende des II. Tones; I n t e r v a l l ~--- Zeit yore Ende des I. Herztones his zum Anfang des II. ; P a u s e----~ Zeitabsehnitt yore Ende des IL Tones his zum Beginn des niichsten I. Tones. Schallphase und Pause setzen die H e r z p e r i o d e zusammen. Die sehr variabele Beziehung der beiden Zeitabschnitte Schallphase Pause ist tier P h a s e n i n d e x . Auf die oben (Fall A) mitgeteilten eigenen Erfahrungen angewandt, wiirde sich also alas Kardiophonogramm zeitlich so zusammensetzen: Junger ~Iensch Hun(]_ Sehallphase . . . . . . 0,313" 0A61 " Intervall . . . . . . . 0,144" 0,040" Pause . . . . . . . . 0,408" 0,171 " 0.313" 1 0,161" 1 Phasenindex . . . . . . 0.408" 1,3 0,171 " 1,1 Schallphase 0,313" 0,161 " Herzperiode . Pause q- 0,408" q- 0,171 " 0,721 " 0,332 " In dem weiter oben (S. 524) erwi~hnten Falle von I n s u f f i z i e n z d e r M i t r a li s war die Zusammensetzung im Mittel : Systolisches Gerausch . . . . . . Intervall . . . . . . . II. Ton . . . . . . . . . . . Schallphase . . . . . . . . . Pause . . . . . . . . . . . 0,285" 0,13 " 0,068" 0,484" 0,165" Phasenindex . . . . . . . . . 0,484" __ 2,9 0,165" 1 0,484" 0,165" ~ Herzperiode . . . . / Schallphase [Pause . . 0,649 " Die Differenz zwischen dem Beginn des systolischen Geriiusches und dem Anstieg des Karotispulses betrug 0,13" Herzschallstudien.. 567 Das Gerausch besass im Mittel eineFrequenz yon 48,4 Schwingungen, der II. Ton eine yon 62,1 Schwingungen pro Sekunde. Mitunter war bei der Kranken, die eine sehr irregulare Herzaktion besass, der II. Ton gespalten. Ich habe in einem solchen Falle folgende Werte gefunden: Systolisches Gerausch 0,256" 1. Intervall . . . . . . 0,063" 2. Ton . . . . . . . 0,028" 2. Intervall . . . . . 9,040" 3. Ton . . . . . . . 0,103" (Getrennt in zwei Abschnitte a ~ 0 , 0 2 8 und b----:0,074 mit kaum messbarem Intervall.) Sehallphase . . . . . . 0,490" Pause . . . . . . . 0,279" Phasenindex Herzperiode . . . . . 0,490" __ 1,8 0,279" 1 / Schallphase 0,490" \| Pause . . 0,279" 0,769"
https://openalex.org/W3009019141	https://scindeks-clanci.ceon.rs/data/pdf/2683-4693/2019/2683-46931902113J.pdf	English	null	Competitiveness of sauerkraut production	Western Balkan Journal of Agricultural Economics and Rural Development	2,019	cc-by-sa	4,644	Abstract Vegetable production represents important segment of the Serbian agricultural output. Within the structure of produced vegetable, cabbage has significant share. Cabbage conservation, primarily for the winter period, in a form of sauerkraut (for preparation of salads and hot dishes) has long tradition both at national and international level. Sauerkraut production is usually in function of value added products generation at the farms oriented to vegetable production that boosts their long-term sustainability. The main goal of this article is to compare contribution margin gained at the farm involved in sauerkraut production made from conventionally or organically produced cabbage. One production year is observed (gained data refers to 2018). It could be concluded that pickling of organic cabbage allows creation of higher farm’s income and contribution margin. Vegetable production represents important segment of the Serbian agricultural output. Within the structure of produced vegetable, cabbage has significant share. Cabbage conservation, primarily for the winter period, in a form of sauerkraut (for preparation of salads and hot dishes) has long tradition both at national and international level. Sauerkraut production is usually in function of value added products generation at the farms oriented to vegetable production that boosts their long-term sustainability. g p g y The main goal of this article is to compare contribution margin gained at the farm involved in sauerkraut production made from conventionally or organically produced cabbage. One production year is observed (gained data refers to 2018). It could be concluded that pickling of organic cabbage allows creation of higher farm’s income and contribution margin. Key words: conventional and organic cabbage, sauerkraut production, contribution margin. JEL5: Q12, Q19 Marko Jeločnik2, Jonel Subić3, Vlado Kovačević4 Marko Jeločnik2, Jonel Subić3, Vlado Kovačević4 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 3 Jonel Subić, Ph.D., Principal Research Fellow, Institute of Agricultural Economics, Volgina Street no. 15, 11060 Belgrade, Serbia, Phone: +381 11 69 72 863, E-mail: jonel_s@iep.bg.ac.rs 2 Marko Jeločnik, Ph.D., Research Associate, Institute of Agricultural Economics, Volgina Street no. 15, 11060 Belgrade, Serbia, Phone: +381 11 69 72 852, E-mail: marko_j@iep.bg.ac.rs 4 Vlado Kovačević, Ph.D., Research Associate, Institute of Agricultural Economics, Volgina Street no. 15, 11060 Belgrade, Serbia, Phone: +381 11 69 72 858, E-mail: vlado_k@iep.bg.ac.rs 1 Paper is a part of research at the project no. III 46006, financed by the Ministry of Education, Science and Technological Development of the Republic of Serbia, as well as the Project Advancement of knowledge transfer towards the approaching to safer and more competitive agricultural products gained by processing at small farms within the sector of milk, meat, fruit and vegetable, financed by the Ministry of Agriculture, Forestry and Water management of the Republic of Serbia. 5 Article info: Original Article, Received: 1st October 2019., Accepted: 14th October 2019. Introduction Cabbage (Brassica oleracea var. capitata) is perennial, cultivated, herbaceous plant gained through the selection of wild cabbage over a very long period (into the human nutrition and medicine was introduced during the ancient Greece in the 4th century BC). Cabbage has a pronounced nutritional and medicinal value, as well as low calorie value. It is contained in the availability of certain useful phytochemicals, 1 Paper is a part of research at the project no. III 46006, financed by the Ministry of Education, Science and Technological Development of the Republic of Serbia, as well as the Project Advancement of knowledge transfer towards the approaching to safer and more competitive agricultural products gained by processing at small farms within the sector of milk, meat, fruit and vegetable, financed by the Ministry of Agriculture, Forestry and Water management of the Republic of Serbia. 1 Paper is a part of research at the project no. III 46006, financed by the Ministry of Education, Science and Technological Development of the Republic of Serbia, as well as the Project Advancement of knowledge transfer towards the approaching to safer and more competitive agricultural products gained by processing at small farms within the sector of milk, meat, fruit and vegetable, financed by the Ministry of Agriculture, Forestry and Water management of the Republic of Serbia. 2 Marko Jeločnik, Ph.D., Research Associate, Institute of Agricultural Economics, Volgina Street no. 15, 11060 Belgrade, Serbia, Phone: +381 11 69 72 852, E-mail: marko_j@iep.bg.ac.rs 3 Jonel Subić, Ph.D., Principal Research Fellow, Institute of Agricultural Economics, Volgina Street no. 15, 11060 Belgrade, Serbia, Phone: +381 11 69 72 863, E-mail: jonel_s@iep.bg.ac.rs 4 Vlado Kovačević, Ph.D., Research Associate, Institute of Agricultural Economics, Volgina Street no. 15, 11060 Belgrade, Serbia, Phone: +381 11 69 72 858, E-mail: vlado_k@iep.bg.ac.rs 5 Article info: Original Article, Received: 1st October 2019., Accepted: 14th October 2019. 113 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 such as indole-3-carbinol (I3C), sulforaphane, phenolic compounds, carotenoids or glucosinolates (antioxidants desirable for detoxification), as well as the most of vitamins (primarily C, or A, B1, B2, PP, B6, B12, P, K and U), minerals (Ca, K, S, etc.) and carbohydrates (mostly free sugars, starch, cellulose and hemicellulose), (Cvetković, 2014). Cabbage significantly appears within the structure of vegetable production in Serbia. After potatoes, it occupies the largest production areas (Červenski, Medić Pap, 2018). Introduction They cover around 9,550 ha (Milić et al., 2018), where within the sowing structure are dominant cereals and fruit, while vegetables are present at small areas, up to couple hundred hectares (Sojić, 2017). Besides, some estimations show that organically produced cabbage covers just several hectares. On the other hand, traditionally sauerkraut has played a significant role in the dietary habits of the European population, especially during the winter. It is estimated that in certain countries it involves more than 70% of total fresh cabbage production, as well as consumption per capita of around 4 kg (in Eastern Europe region). Besides production, in Serbia the most of fresh cabbage is processed (fermented) at small family farms and households. Within the volume of market-oriented sauerkraut, 50% ends up in retail, while the rest is realized through the green market or direct sales at the farm gate (Simović, 2010). Usually, sustainable agriculture is turned to the ability of agricultural systems to handle the high crop or animal productivity over the long period, while they are flexible enough to adapt to expecting changes. In essence, sustainable agriculture is mainly characterized by maintaining of environmental quality, stable crop or animal productivity and social acceptability (Đokić, 2019). Besides, there is a need for reconsideration of long-term farm sustainability, as well as certain factors that affect it (D’souza et al., 1993; Baccar et al., 2019): farms’ size and economic strength; production orientation; level of specialization; age and number of farm members; availability of natural and production resources; level of farmers’ education and readiness for tech-tech transfer, knowledge and skills improvement, or added value creation; level of awareness linked to environmental issues; etc. In function of strengthening the farm sustainability is also the availability of specific agricultural and/or food products, or services that carry within themselves certain level of added value, and that can be successfully realized at the local or regional market. They can be the result of either production specialization or diversification of farm activities. During the previous decades, farmers seriously fight to ﬁnd the right way to increase their income throughout the process of “adding value” to raw agricultural products they produce. There are a number of methods for that, such are: cleaning, cooling, packag- ing, labelling, processing, distributing, churning, grinding, hulling, extracting, drying, smoking, handcrafting, etc. Introduction It is most commonly used in human nutrition, as fresh or processed (usually as sauerkraut), for the preparation of various salads and meals. In the diet are used the head, leaf or juice of cabbage. Cabbage is mainly cultivated at the small family farms, which dispose with at maximum 5 ha of production area. Production is dominantly turned to direct or natural consumption and industrial processing. On the other side, in the total volume of produced vegetables, cabbage is generally positioned around the 10th place, where, as with other vegetables, annual production is insufficient to cover domestic demand (GRS, 2019). The SORS collects and presents summarized production data for all brassica plants (cabbage and kale). From the data presented in the Table 1., it could be noticed that during the last 7 years (2012-2018), there has been a negative trend in the total harvested areas under the brassica plants. Oscillations related to yields and total production is primarily caused by the climatic factor (increase in deficit or bad spatial and temporal dispersion of rainfalls, or increase in average temperatures within the growing season), as cabbage production requires significant volume of water and frequent watering. Table 1. Harvested surfaces, total production and average yields of cabbage and kale at national level (period 2012-2018) Table 1. Harvested surfaces, total production and average yields of cabbage and kale at national level (period 2012-2018) Year Harvested area (in ha) Total production (in t) Yield (in t/ha) 2012. 11,890 303,893 25.6 2013. 11,246 303,893 27.0 2014. 11,116 261,240 23.5 2015. 11,039 288,698 26.2 2016. 10,804 290,001 26.8 2017. 10,213 262,545 25.7 2018. 8,251 209,353 25.4 Average 10,651 274,232 25.7 Source: SORS 2019 114 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 Although surfaces under organic production are still limited (less than 0.5% of utilized agricultural area), their growth has positive trend at national level. They cover around 9,550 ha (Milić et al., 2018), where within the sowing structure are dominant cereals and fruit, while vegetables are present at small areas, up to couple hundred hectares (Sojić, 2017). Besides, some estimations show that organically produced cabbage covers just several hectares. Although surfaces under organic production are still limited (less than 0.5% of utilized agricultural area), their growth has positive trend at national level. Introduction So contemporary farm business recognizes the adding value as “selling the sizzle, not the steak”, where the sizzle is the product of information, edu- cation, entertainment, image, and other intangible attributes (Born, Bachmann, 2006). 115 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 In support of previously mentioned is the business philosophy of “niche marketing”, which refers to the realization of a specific product at the relatively small but well- defined market segment, such are certain geographical territory, specialized sector of industry, or certain ethnic, religious, age, gender or some other consumers’ group. As example could be noted local traditionally produced products or products with geographical origin, organic products, or foodstuff intended for vegetarians or sportsmen (Anzaku, Salau, 2017). In support of previously mentioned is the business philosophy of “niche marketing”, which refers to the realization of a specific product at the relatively small but well- defined market segment, such are certain geographical territory, specialized sector of industry, or certain ethnic, religious, age, gender or some other consumers’ group. As example could be noted local traditionally produced products or products with geographical origin, organic products, or foodstuff intended for vegetarians or sportsmen (Anzaku, Salau, 2017). From the farmer’s perspective, value added represents a portfolio of agricultural activities and practices by which they want to maximally adjust their agro-food products and services with the preferences of end-consumers. Value-added is available to farmers who are improving their position within the supply chain, moving closer to the end-consumers, or making changes or advancements within the existing production process in order to change, point up or preserve certain characteristics of their products (Coltrain et al., 2000). Results with discussion The analysis of economic justification of sauerkraut production is based on the data obtained from a registered family farm that has been active in the vegetable sector for many years. The farm is located in the South-Banat District, at the territory of the Pančevo city. Production is dominantly turned to ecologically oriented growing (in line to strict use of the GAP principles) of cucumbers and tomatoes in greenhouse, as well as production of cabbage and other Brassicaceae plants in the open field. The most of the produced vegetables is selling as fresh throughout the local green market or at farm gate. Besides, certain volume of vegetables is transformed through the processing into the higher value products (such as sauerkraut, pasteurized tomato juice, pickles, hot ketchup, torshi, etc.). They are selling throughout the local retail, green market, or at farm gate. In the focus of this paper is a comparative analysis of the economic justification of processing (pickling) of cabbage produced within the conventional or organic system of production. Since the farm has many years of experience in cabbage processing (sauerkraut production), according the fact that processing of cabbage produced in any production system requires the same equipment, technological approach and norms of used inputs, it would be interesting to consider economic effects initiated by changing of the basic input (cabbage) within the process of sauerkraut production. Starting the process of cabbage processing at the farm in the volume of around 16 t of fresh cabbage per year (split into the two cycles), requires certain level of investments in infrastructure equipment of processing facility (electrification, access to fresh water and drainage of technical water, thermal insulation, lining the ceramic tiles, etc.), as well as purchase of required equipment. Total sum needed for that purpose is almost 2,115,000 RSD, or approximately 18,000 EUR (1 EUR = 117,5 RSD). Organizing the cabbage processing at the farm involves a production facility of approximately 65 m2 consisted from at least three separate rooms (with space for pickling, for manipulation under the inputs and final product, and for cold storing of the packed final product). Used methodology and data sources All necessary data used for analytical calculations are obtained from family farm traditionally oriented to vegetable production and vegetable processing, located in South-Banat District. It has to be noted that farm is involved in conventional vegetable production and further processing, while large share in produced vegetable has cabbage production. Collected data refers to production year 2018. Besides, it’s also used the secondary data of SORS, scientific and professional literature oriented to the researched field of production and processing. Main research goal is comparison of economic effects (incomes and variable costs) gained in sauerkraut production during one year based on the use of conventionally and organi- cally produced cabbage. Economic effects are reconsidered and compared throughout the preparation of analytical calculations based on variable costs (contribution margin). Calculation of contribution margin linked to the production of certain crop culture (including vegetables) or food product (food processing) considers subtracting of variable costs from incomes generated within the observed production line (Nastić et al., 2018). Character of variable costs in food (vegetable) processing mainly have used inputs and services, such are fresh vegetable, salt or certain additives, water, energy, external services and engaged labour, etc. Better comparison is enabled by the tabular presentation of all data and economic results (in RSD) for the total volume of sauerkraut production (volume of processed cabbage at the farm is unified and limited by the capacity of production space and used equipment). 116 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 Results with discussion Required equipment includes: 15 barrels (from industrial plastic, with the capacity of 1,000 l); worktable (acid-resistant); vacuum machine (two-chambers); cabbage cutting machines (with working parts of stainless steel); set of knifes; 30 plastic water weights (5 kg each); 5 plastic containers (with the capacity of up to 20 kg); 2 digital weighing scales (up to 40 kg); 4 fine polished wooden pallets for sauerkraut squeezing; cart; etc. (Subić, Tomić, 2019). 117 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 Short description of used technology in the sauerkraut production, as well as incomes, costs and contribution margin gained throughout the pickling of organically and conventionally produced cabbage, are given in the following table (Table 2.). Table 2. Contribution margin in sauerkraut production Element U.M. Sauerkraut from conventionally produced cabbage Sauerkraut from organically produced cabbage Quantity Price/UM (in RSD) Total (in RSD) Quantity Price/UM (in RSD) Total (in RSD) I Incomes Sauerkraut kg 12,000 70 840,000 12,000 160 1,920,000 Subsidies pack. - - Value of production (total I) 840,000 1,920,000 II Variable costs Cabbage kg 15,750 17.5 275,625 15,750 75 1,181,250 Salt kg 600 140 84,000 600 140 84,000 Foil pcs 30 275 8,250 30 275 8,250 Vacuum bag pcs 6,000 7.5 45,000 6,000 8,5 51,000 Transport box pcs 400 55 22,000 400 55 22,000 Labour h 408 270 110,160 408 270 110,160 Energy - - - 27,000 - - 27,000 Water - - - 1,800 - - 1,800 Dez & Der - - - 9,500 - - 9,500 Quality control - - - 8,500 - - 10,500 Certification - - - - - - 118,000 Taxes - - - 6,750 - - 6,750 Other costs - - - 6,250 - - 6,250 Sum of variable costs (total II) 604,835 1,636,460 III Contribution margin (I-II) 235,165 283,540 Source: IAE, 2019. Table 2. Contribution margin in sauerkraut production Source: IAE, 2019. Source: IAE, 2019. Hereinafter, it will be given certain clarifications regarding the data shown in previous table (Table 2.). The season of production and sale of sauerkraut is framed by the period September - January. Technological process of sauerkraut production lasts for 20-25 days. As was previously mentioned, farm has on disposal 15 barrels, where each barrel is used twice during the season of sauerkraut production (annual production is technically limited to 30 barrels). For processing is used self-produced or from local vegetable growers purchased first class fresh cabbage. Results with discussion It’s shown the average wholesale price of fresh cabbage 118 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 (conventionally and organically produced) that includes its delivery to farm. One season of cabbage pickling requires the purchase of 15,750 kg of fresh cabbage. After the purchase, fresh cabbage is mechanically cleaned (stripping the outer leaves), and preparing for pickling (cutting the root or chopping the heads into the small pieces). With this activity, the volume of fresh cabbage that enters the pickling process is reduced for 5%. Each barrel could receive 500 kg of fresh cabbage, where during the pickling process volume of cabbage is additionally reduced for 20%. So after the draining around 12,000 kg of final product could be packaged (in vacuum bags and shipping boxes) and sold usually to local retailer. In addition to fresh cabbage, a solution of water and table salt is adding into the barrel. Then, the barrel is tucked in well by the foil and additionally loaded with plastic water weights. Share of salt within the structure of production mass is about 2%, i.e. each barrel contains 20 kg of salt. Salt is buying in bulk in 5 kg bags. As was previously mentioned, the process of pickling requires the use of a foil for the barrel sealing, in order to preserve necessary conditions for cabbage fermentation (new foil is placed during the each filling of the barrel). Sauerkraut is delivering at the market in vacuum bags (with printed data about the producer and products’ declaration) in volume of 1-2 kg/bag. Vacuum bags are additionally packaged in plasticized cardboard boxes (15 bags/box). Labour costs consider the occasional engagement of two persons for the realization of several processing activities. Although the labour costs (the most often only costs of external labour are used (Subic, Jelocnik, 2019)) may or may not burden the analytical calculation of contribution margin, in order to better perceive the effects that arise from cabbage processing, they are fully shown within the structure of variable costs . Working staff is represented by a farmer and external worker, where the gross wage per working hour accounts around 270 RSD. The process of filling the barrels with fresh cabbage (cleaning, cutting, chopping and stacking of the cabbage, pouring the brine and sealing the barrels) requires in total the fund of 168 working hours. Results with discussion - Under the same processing capacity, the contribution margin achieved in processing of organically produced cabbage is for more than 20% higher than achieved one in processing of conventionally produced cabbage. - Under the same processing capacity, the contribution margin achieved in processing of organically produced cabbage is for more than 20% higher than achieved one in processing of conventionally produced cabbage. - Initiation of processing of organically produced cabbage needs much higher amount of financial assets (2.7 times higher). Besides, achieved incomes are 2.3 times higher. - In both lines, in the structure of variable costs dominate the costs of fresh cabbage (in processing of organically produced cabbage even 72%). - In both lines, in the structure of variable costs dominate the costs of fresh cabbage (in processing of organically produced cabbage even 72%). Results with discussion Taking out the sauerkraut from the barrels and its stacking at the wooden pallets for further draining (before packing the sauerkraut is pyramidally stacked at the pallets and left to drain for next 24 hours) requires in total the 120 working hours. Packing (vacuuming) of sauerkraut and its transfer to cold storage requires additional 120 working hours. So, the total labour (for the whole season) requires the fund of 408 working hours. Production of sauerkraut is additionally burdened with the costs of electricity (for lighting, heating of the working facility, for operation of the cold storage 119 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 and used equipment, etc.), water costs (for the use of fresh water and disposal of wastewater), costs of disinfection and pest control within the facility, costs of final product quality control (verification of the product declaration), costs of taxes and fees, and other costs. In order to mark the final product of processing with the tag representing it as a organic, production of sauerkraut based on organic cabbage has to be previously certified (certification of technological process, i.e. used facility and equipment) by authorized certification body. Besides, processor has to proof that the input he uses in pickling is produced in organic system. So, within the processing of organic products are used just certified inputs. If parallel with the processing of conventional inputs, processing line of organic inputs wants to be established, the entire technological process must be separated in time or space. Also, before the beginning of organic products processing, all equipment must be pre- cleaned with allowed cleanser. In addition, mixing of organically and conventionally produced inputs and final products of processing is not allowed. It can be used only by the regulative on organic production allowed additives, excipients, minerals, etc. Domestic processed certified organic products may be labelled with the national organic product sign and logo of the inspection body that previously certified the primary product (input) if the final product of processing contains minimum 95% of ingredients of agricultural origin produced in line to principles of organic production (Serbia Organica, 2019). According to previously mentioned next could be concluded: According to previously mentioned next could be concluded: - In both processing lines are gained positive contribution margin. Conclusion Cabbage production has long tradition at the territory of Serbia. As primary product gained from cabbage processing, sauerkraut is unavoidable food product, especially 120 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 during the winter time. On the other hand, areas under organically produced vegetable, specifically cabbage, currently are below production capabilities and market demand. In paper were tested and compared contribution margins gained in sauerkraut production based on the use of organically or conventionally produced cabbage, as mentioned productions require identical production process and equipment. Obtained results show that for certain farm, processing of organic cabbage could be for 20% more profitable, although starting of its conduction (purchasing of inputs) requires much higher amount of financial assets. This could be such important information given the assumption that the needs at national niche market for organic sauerkraut are much greater than currently possible production. This is crucial for two reasons. Firstly, mentioned can pull organic cabbage producers to expand their production areas and offer fresh cabbage to farms that are already active in cabbage processing. By this processors could additionally engaged their available processing capacity to maximal level, reaching at that way the higher grade of profitability and economic sustainability. Secondly, producers of organic cabbage can divers their production activity with introduction of cabbage processing at the farm. By that they could contribute to maintaining and increase of farm income, especially in years with low sales price of fresh organic cabbage. Thirdly, any farm active only in the segment of cabbage processing (whether organic or conventional cabbage), with presented capacities cannot provide its own, primarily financial sustainability (gained incomes and contribution margins are too low for maintaining of decent life for farm members). In line to the volume of produced final product, required work engagement and possible incomes, cabbage processing could be just in function of secondary (additional) activity at the farm. References 1. Anzaku, T. A. K., Salau, E. S. (2017). Niche marketing potentials for farm entrepreneurs in Nigeria. Journal of Agricultural Extension, 21(3):136-142. 2. Baccar, M., Bouaziz, A., Dugué, P., Gafsi, M., Le Gal, P. Y. (2019). The determining factors of farm sustainability in a context of growing agricultural intensification. Agroecology and Sustainable Food Systems, 43(4):386-408. 3. Born, H., Bachmann, J. (2006). Adding value to farm products: An overview. ATTRA - National Sustainable Agriculture Information Service, Fayetteville, USA, pp. 1-12. 4. Červenski, J., Medić-Pap, S. (2018). Proizvodnja kupusa. monografija, Institut za ratarstvo i povrtarstvo, Novi Sad, Srbija. 121 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 5. Coltrain, D., Barton, D., Boland, M. (2000). Value Added: Opportunities and Strategies. Arthur Capper Cooperative Center, Department of Agricultural Economics, Kansas State University, Manhattan, USA. 6. Cvetković, B. (2014). Primena tehnoloških postupaka spontane fermentaciji i osmotske dehidratacije za unapređenje nutritivnog profila, senzornih svojstava i održivosti kupusa, doktorska disertacija, Tehnološki fakultet, Univerziteta u Novom Sadu, Novi Sad, Srbija. 7. D’souza, G., Cyphers, D., Phipps, T. (1993). Factors affecting the adoption of sustainable agricultural practices. Agricultural and Resource Economics Review, 22(2):159-165. 8. Đokić, M. (2019). Sustainable agricultural and rural development in the European Union. Economics of Sustainable Development, 3(1):29-43. 9. GRS (2019). Zaklјučak o usvajanju IPARD programa za Republiku Srbiju za period 2014–2020. Godine. Vlada Republike Srbije, Beograd, Srbija, Sluzbeni glasnik Republike Srbije, br. 30/2016-3, 84/2017-30, 20/2019-22, 55/2019-55. 10. IAE (2019). Field data related to cabbage processing (sauerkraut production). Internal data, Institute of Agricultural Economics (IAE), Belgrade, Serbia. 11. Milić, D., Lukač Bulatović, M., Milić, D. (2018). Uporedna analiza organske proizvodnje voća u Evropskoj uniji i Srbiji. Agroekonomika, 47(80):13-22. 12. Nastić, L., Jeločnik, M., Subić, J. (2018). Contribution margin in silage maze production. Ekonomika, 64(4):71-80. 13. Serbia Organica (2019). Prerada organskih proizvoda. Portal Serbia Organica (National Association for Organic Production), Belgrade, Serbia, available at: https://serbiaorganica.info/prerada-organskih-proizvoda/#, retrieved at: 28.9.2019. 14. Simović, Đ. (2010). Kvalitetan sortiment za kiseli kupus. portal Poljoprivreda. info, available at: https://poljoprivreda.info/tekst/kvalitetan-sortiment-za- kiseli-kupus 15. Sojić, S. (2017). Kreiranje brenda organskih poljoprivredno-prehrambenih proizvoda u Republici Srbiji. doktorska disertacija, Univerzitet u Novom Sadu, Poljoprivredni fakultet, Novi Sad, Srbija. 16. SORS (2019). Data related to cabbage and kale production for the period 2012-2018. portal of the Statistical Office of the Republic of Serbia (SORS), Belgrade, Serbia, available at: https://data.stat.gov.rs/Home/ Result/130102?languageCode=sr-Cyrl 122 WBJAERD, Vol. 1, No. 2 (85-160), July - December, 2019 17. Subić, J., Jeločnik, M. References (2019). Economic Effectiveness of Ecologically Acceptable Production of Vegetables in Protected Area. In: Sustainable Agriculture and Rural Development in Terms of the Republic of Serbia Strategic Goals Realization within the Danube Region: Sustainability and Multifunctionality, thematic proceedings, (Eds.) Subić, J., Kuzman, B., Jeločnik, M., Vasile, A. Institute of Agricultural Economics, Belgrade, Serbia, pp. 333-352. 18. Subić, J., Tomić, V. (2019). Programi investicija u preradu bezbedne hrane na malim poljoprivrednim gazdinstvima za mleko, meso, voće i povrće. In: Unapređenje transfera znanja radi dobijanja bezbednih i konkurentnih poljoprivrednih proizvoda, koji su dobijeni preradom na malim gazdinstvima u sektorima mleka, mesa, voća i povrća, (Edt.) Kovačević, V., monograph, Institute of Agricultural Economics, Belgrade, Serbia, pp. 93-158. 123
https://openalex.org/W4224984810	https://bmccancer.biomedcentral.com/track/pdf/10.1186/s12885-022-09569-2	English	null	Tryptophan hydroxylase 1 drives glioma progression by modulating the serotonin/L1CAM/NF-κB signaling pathway	BMC cancer	2,022	cc-by	8,598	Zhang et al. BMC Cancer (2022) 22:457 https://doi.org/10.1186/s12885-022-09569-2 © 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, visithttp://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. Abstract Tumor relapse and the development of chem- oresistance represent the predominant cause of treat- ment failure, leading to a dismal five-year survival rate of only 5% and a median survival duration of < 15 months [3, 4]. To date, the molecular mechanisms underlying glioma development remain unidentified. Thus, further Abstract Background: Glioma is one of the main causes of cancer-related mortality worldwide and is associated with high heterogeneity. However, the key players determining the fate of glioma remain obscure. In the present study, we shed light on tumor metabolism and aimed to investigate the role of tryptophan hydroxylase 1 (TPH-1) in the advance- ment of glioma. Method: Herein, the levels of TPH-1 expression in glioma tissues were evaluated using The Cancer Genome Atlas (TCGA) database. Further, the proliferative characteristics and migration ability of TPH-1 overexpressing LN229/T98G cells were evaluated. Additionally, we performed a cytotoxicity analysis using temozolomide (TMZ) in these cells. We also examined the tumor growth and survival time in a mouse model of glioma treated with chemotherapeutic agents and a TPH-1 inhibitor. Results: The results of both clinical and experimental data showed that excess TPH-1 expression resulted in sustained glioma progression and a dismal overall survival in these patients. Mechanistically, TPH-1 increased the production of serotonin in glioma cells. The elevated serotonin levels then augmented the NF-κB signaling pathway through the upregulation of the L1-cell adhesion molecule (L1CAM), thereby contributing to cellular proliferation, invasive migra- tion, and drug resistance. In vivo experiments demonstrated potent antitumor effects, which benefited further from the synergistic combination of TMZ and LX-1031. Conclusion: Taken together, these data suggested that TPH-1 facilitated cellular proliferation, migration, and chem- oresistance in glioma through the serotonin/L1CAM/NF-κB pathway. By demonstrating the link of amino acid meta- bolic enzymes with tumor development, our findings may provide a potentially viable target for therapeutic manipu- lation aimed at eradicating glioma. Keywords: TPH-1, Glioma, Serotonin, L1CAM Keywords: TPH-1, Glioma, Serotonin, L1CAM [1]. The mainstay of glioma treatment consists of surgi- cal resection alone or in combination with chemother- apy and radiation therapy, however, the efficacy remains modest [2]. Tumor relapse and the development of chem- oresistance represent the predominant cause of treat- ment failure, leading to a dismal five-year survival rate of only 5% and a median survival duration of < 15 months [3, 4]. To date, the molecular mechanisms underlying glioma development remain unidentified. Thus, further [1]. The mainstay of glioma treatment consists of surgi- cal resection alone or in combination with chemother- apy and radiation therapy, however, the efficacy remains modest [2]. Background Glioma is among the tumor types that show the worst prognoses and is characterized by low differentiation, extensive angiogenesis, and a high invasive potential Correspondence: zhangchaoguoedu@outlook.com †Jie Zhang and Zhangchao Guo contributed equally to this work. 1 Department of Neurosurgery, Ya’ an people’s Hospital, Ya’ an 625000, People’s Republic of China Full list of author information is available at the end of the article Correspondence: zhangchaoguoedu@outlook.com †Jie Zhang and Zhangchao Guo contributed equally to this work. 1 Department of Neurosurgery, Ya’ an people’s Hospital, Ya’ an 625000, People’s Republic of China Full list of author information is available at the end of the article © 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, visithttp://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. Zhang et al. BMC Cancer (2022) 22:457 Zhang et al. BMC Cancer (2022) 22:457 Page 2 of 11 can promote carcinogenesis, including IL-1β, BCL2, and VEGF [21, 22]. However, the correlation between trypto- phan metabolism and NF-κB during tumor progression remains poorly understood.h investigations are warranted to gain insight into gli- oma progression and provide a new avenue for tumor elimination. Tumor development is a complicated and multistage process that involves both genetic and environmental factors [5]. Cellular metabolism is extensively reported to function as a master regulator of tumor behaviors [6]. To cope with the high proliferation rate, cancer cells may undergo metabolic reprogramming and utilize high glucose for sufficient ATP supply. Recent findings indi- cate that in addition to an abundance of glucose, amino acids are exploited by tumors to fulfill the requirements of energy and biosynthesis in cells [7]. Tryptophan, an essential amino acid, serves a significant role in malignant conversion and tumor advancement [8]. The majority of Trp degradation occurs through the kynurenine (Kyn) pathway, resulting in the generation of several metabo- lites with diverse biological functions. The rate-limiting step of the Kyn pathway is catalyzed by indoleamine-2, 3-dioxygenase (IDO), and tryptophan-2, 3-dioxygenase (TDO), which are intimately linked to the modulation of immune responses and offer promising targets for cancer therapy [9]. While a small fraction of Trp is converted to 5-hydroxytryptophan by tryptophan hydroxylase-1 (TPH-1) and provides precursors for the production of serotonin, a classical neurotransmitter with multiple roles in the cardiovascular system, endocrinology, gut motil- ity, reproductive function, and carcinogenesis [10, 11]. In fact, TPH-1 is up-regulated in several tumor types [12]. Jaya et al. report that TPH-1 is preferentially expressed in triple-negative breast cancer (TNBC) and its silencing markedly suppresses cellular proliferation and invasion [13]. They further suggest that TPH-1 facilitates tumor progression via autocrine serotonin signaling. Moreo- ver, TPH-1 knockdown or 4-chloro-DL-phenylalanine (a TPH-1 inhibitor) treatment can retard tumor growth in mice models of colorectal cancer [14]. At present, tel- otristat, a tryptophan hydroxylase inhibitor, is being eval- uated in clinical trials for the treatment of patients with metastatic neuroendocrine tumors [15]. Despite the sub- stantial research efforts, a clear understanding of the role of TPH-1 in tumor progression is lacking. Using The Cancer Genome Atlas (TCGA) database, we observed that TPH-1 expression was markedly aug- mented in glioma tissues and that patients with a high TPH-1 expression exhibited poorer overall survival rates. In addition, data from the in vitro experiments showed that TPH-1 facilitated glioma cell proliferation and decreased the efficacy of chemotherapy by catalyz- ing the production of serotonin. Our study further indi- cated that TPH-1 drove glioma development in an L1-cell adhesion molecule (L1-CAM)/NF-κB dependent man- ner. Importantly, a combination of chemotherapeutic drugs and a TPH-1 inhibitor yielded excellent treatment outcomes in the xenograft mouse models. These results are expected to motivate extensive research for targeting TPH-1 signals in pursuit of optimal therapeutic strategies for glioma. Cellular proliferation assay Cellular proliferation in LN229 and T98G cells was deter- mined using the Cell Counting Kit-8 (CCK8) Assay Kit (Solarbio, China). Briefly, 2 × 103 tumor cells were resus- pended and seeded per well in a 96-well plate. The cell numbers were monitored daily using the CCK-8 solution. A microplate reader (Thermo Fisher, USA) was used to measure the sample absorbance (OD) at 450 nm. Three independent experiments were performed. The NF-κB transcription factor plays an essential role in inflammation and innate immunity [16, 17]. More importantly, recent studies suggest that NF-κB serves as a crucial player in different steps of tumor progression [18]. NF-κB cooperates with multiple signaling molecules and pathways, such as STAT3, P53, or the adhesion mol- ecule, LCAM. The crosstalk between NF-κB and pro- survival signals can also be mediated by different kinases, including AKT and P38, which facilitate NF-κB tran- scriptional activity or affect upstream signaling [19, 20]. Additionally, many of the genes transcribed by NF-κB Materials and methods Cell culture and reagents g LN229 and T98G cell lines were purchased from Ameri- can Type Culture Collection (ATCC) and maintained in RPMI 1640 complete culture medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA). TPH-1 overexpressing LN229 and T98G cells were generated by Cyagen (China) and the expressions were determined by western blotting. Temozolomide (TMZ) and serotonin were purchased from Sigma (USA). The NF-κB inhibitor, QNZ (EVP4593), was obtained from Selleck (USA). TPH-1 inhibitor, LX-1031, was obtained from Abcam (UK). TUNEL assay Cell apoptosis in tumor tissues was determined by ter- minal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) staining (Solarbio, China). Briefly, mice were injected with subcutaneously 1 × 106 vector or TPH1 overexpressing LN229 cells on day 0. After 12 days, mice were treated with serotonin (1 μg in 100 ml PBS) or PBS by intratumor injection. On day 13, mice were treated with TMZ (0.05 μg in 100 ml PBS). On day 15, tumor tissues were harvested and frozen sections were permeabilized with 0.1% Triton X-100, and incubated with the TUNEL reaction mixture according to guidance of Kit. The percentage of apoptotic cells was determined as TUNEL positive cells/total number of cells. Western blotting Glioma cells were lysed using 1% NP40 buffer containing a protease inhibitor cocktail (Solarbio, China). The Pro- tein Quantitative Analysis Kit (Solarbio, China) was used for protein quantification. 25 μg protein were separated using a 10% SDS gel and the separated proteins were transferred onto polyvinylidene fluoride immobilon- membranes. Subsequently, cropped (or not) immobilon- membranes were blocked using 5% nonfat dried milk and incubated with the following primary antibodies: anti-TPH-1 (Abcam, UK), anti-L1CAM (Abcam, UK), anti-NF-κB (Abcam, UK), and anti-β-actin (Abcam, UK). Next, the membranes were incubated using an HRP-con- jugated secondary antibody (Abcam, UK) for one hour at room temperature. Proteins were visualized using the ECL detection kit (Thermo Fisher, USA). Transwell assay 5 1 × 105 T98G or 5 × 104 LN229 cells were seeded in a Transwell insert (8 μm, Corning, USA) for assessing the cellular migration ability. After 48 h, the migrating cells were fixed with paraformaldehyde and stained using crystal violet. The migrated cells were counted and each experiment was performed thrice independently. Zhang et al. BMC Cancer (2022) 22:457 Zhang et al. BMC Cancer (2022) 22:457 Page 3 of 11 Page 3 of 11 Immunofluorescence assay l Paraffin sections of glioma tissues were de-waxed and treated with sodium citrate for antigen retrieval. Subse- quently, the samples were blocked with 5% bovine serum albumin for 30 min at room temperature, and incu- bated with the following primary antibodies: anti-TPH-1 (Abcam, UK), anti- serotonin antibody (Abcam, UK), and anti-L1CAM (Abcam, UK), and anti-NF-κB (Abcam, UK), overnight at 4 °C. These samples were then incu- bated with goat anti-rabbit secondary antibody (Abcam, UK) and the nuclei were stained with DAPI (Solarbio, China). The intensity of protein expression was analyzed using the Image J 6.0 software (USA). Paraffin sections of human glioma tissues were collected from the pathology department of Ya’ an People’s Hospi- tal. Glioma tissues were divided into the TPH-1high and TPH-1low groups based on the median expression of TPH-1, which was determined by immunohistochem- istry or immunofluorescence assays. All glioma patients were informed and provided written consent to par- ticipate in the study. The clinical experiments were per- formed according to the guidelines in the Declaration of Helsinki and approved by the Ethics Committee of Ya’ an People’s Hospital. Cytotoxicity assay Cell apoptosis of LN229/T98G cells induced by TMZ or inhibitors was determined using the FITC-Annexin V/ PE-PI apoptosis detection kit (Becton Dickinson, USA) according to the guidance of manufacturer. Briefly, LN229 and T98G cells were treated with TMZ (1 μg/ml) for 48 h. Then cells were stained with FITC-Annexin V/ PE-PI staining solution for 20 min at room tempera- ture. After that, cell apoptosis was detected by a C6 flow cytometer (Becton Dickinson, USA). RNA interference For L1CAM silencing, LN229 and T98G cells were treated using siRNA oligonucleotides at a concentra- tion of 100 nM using Oligofectamine (Thermo, USA). The siRNA sequences used were as follows: 5′-AGG GAUGGUGUCCACUUCAAATT-3 and 5′-UGAAGU CGAGCGAUCCGUAG-3′. The silencing efficiencies for LICAM in LN229 and T98G cells were examined by real- time PCR. Enzyme‑linked immunosorbent assay (Elisa) assay Patient information, including clinical and gene-expres- sion data, was obtained from TCGA database, which included data for 99 glioma patients (survival and RNA expression data, https://www.cbioportal.org/), 5 normal tissue and 156 glioma tissues (TPH1 RNA expression data, http://ualcan.path.uab.edu/index.html). Overall survival in the two groups were analyzed and compared by the Kaplan–Meier method. Differences in gene expression were tested for statistical significance by the Student’s t-test using GraphPad Prism software. Gene analysis were performed with the use of the open-source R software (2.1.0). y y y Elisa assay was performed according to the guidance of manufacturer. Human Tryptophan Elisa Kits were pur- chased from Guidechem (China). Human Serotonin Elisa Kits were purchased from Abcam (UK). Three independ- ent experiments were performed. Animal protocols Female 6 ~ 8 weeks old NOD-SCID mice were purchased from Huafukang (China) and maintained in Specific Pathogen Free (SPF) room. For tumor volume assay, mice were subcutaneously injected with 106 LN229 cells (n = 6 in each group). After 10 days, mice were treated Zhang et al. BMC Cancer (2022) 22:457 Page 4 of 11 Page 4 of 11 them to those in the normal tissues using TCGA data- base. Intriguingly, an elevated expression of TPH-1 was observed in glioma tissues relative to the normal tissues (Fig. 1A), suggesting the potential role of TPH-1 in pro- moting glioma development. To validate our results, we evaluated the overall survival of glioma patients. The top 10% of the patients with the highest TPH-1 expression were categorized into the high TPH-1 group (n = 49) and those with the lowest 10% expression were divided into the low TPH-1 group (n = 50). Indeed, patients with a high TPH-1 expression exhibited a poor overall survival relative to those in the low TPH-1 group (Fig. 1B). Given the poor prognosis of glioma patients with a high TPH-1 expression, we reasonably speculated whether TPH-1 expression could influence the cellular proliferation/ migration in glioma. Thus, TPH-1 was overexpressed in glioma cancer cell lines, LN229 and T98G (Fig. 1C). The cellular proliferation/migration abilities were determined using the CCK8 and Transwell assays. Notably, TPH-1 overexpression significantly promoted cellular prolif- eration (Fig. 1D) and migration (Fig. 1E) of glioma cells. In addition, TPH-1 overexpressing LN229/T98G cells exhibited enhanced resistance to the chemotherapeu- tic agent, TMZ (Fig. 1F), suggesting that a high level of TPH-1 resulted in enhanced tumor growth and reduced chemo-sensitivity in glioma. To further confirm the role with PBS, TMZ (5 mg/kg), LX-1031(5 mg/kg) or combin- ing therapy twice a week. Tumor volume and survival of mice was recorded every day. The calculation formula of tumor volume is: tumor volume = length × width 2/2. The animal studies were conducted in accordance with the Public Health Service Policy and complied with the WHO guidelines for the humane use and care of animals. All animal protocols were monitored by the Animal Eth- ics Committee of Ya’ an people’s Hospital. with PBS, TMZ (5 mg/kg), LX-1031(5 mg/kg) or combin- ing therapy twice a week. Tumor volume and survival of mice was recorded every day. The calculation formula of tumor volume is: tumor volume = length × width 2/2. Animal protocols The animal studies were conducted in accordance with the Public Health Service Policy and complied with the WHO guidelines for the humane use and care of animals. All animal protocols were monitored by the Animal Eth- ics Committee of Ya’ an people’s Hospital. Statistical analysis All data were presented as the mean ± SEM and analyzed using GraphPad 7.0. The statistical significance between the two groups was calculated using the Student’s t-test or a one-way ANOVA for three or more groups. Kaplan–Meier curves were used for survival analysis. All experiments in our study were performed in independ- ent triplicates. , p < 0.05; *, p < 0.01; n.s, no significant difference. TPH‑1 promotes glioma progression and correlates with a poor progression To assess the potential relevance of TPH-1 in glioma development, first, we evaluated the transcriptomic expression of TPH-1 in 156 glioma tissues and compared Fig. 1 TPH-1 promoted glioma progression. A the transcriptome expression of TPH-1 in 156 glioma tissues and 5 normal tissues. B overall survival of glioma patients divided into high TPH-1 (n = 49) and low TPH-1 (n = 50) groups. C western blotting of TPH-1 in LN229/T98G (VEC) and TPH-1 overexpressing (OE) LN229/T98G cells. D cell proliferation of LN229/T98G (VEC) and TPH-1 overexpressing LN229/T98G cells (OE) determined by CCK8 assay. E relative migrating cells of LN229/T98G (VEC) and TPH-1 overexpressing LN229/T98G cells (OE) determined by transwell assay. F cell apoptosis of LN229/T98G (VEC) and TPH-1 overexpressing LN229/T98G cells (OE) treated with TMZ (48 h) Fig. 1 TPH-1 promoted glioma progression. A the transcriptome expression of TPH-1 in 156 glioma tissues and 5 normal tissues. B overall survival of glioma patients divided into high TPH-1 (n = 49) and low TPH-1 (n = 50) groups. C western blotting of TPH-1 in LN229/T98G (VEC) and TPH-1 overexpressing (OE) LN229/T98G cells. D cell proliferation of LN229/T98G (VEC) and TPH-1 overexpressing LN229/T98G cells (OE) determined by CCK8 assay. E relative migrating cells of LN229/T98G (VEC) and TPH-1 overexpressing LN229/T98G cells (OE) determined by transwell assay. F cell apoptosis of LN229/T98G (VEC) and TPH-1 overexpressing LN229/T98G cells (OE) treated with TMZ (48 h) Zhang et al. BMC Cancer (2022) 22:457 Zhang et al. BMC Cancer (2022) 22:457 Page 5 of 11 the cellular proliferative characteristics (Fig. 2C) and migration phenotypes (Fig. 2D) of glioma cells. Moreo- ver, serotonin treatment enhanced the TMZ resistance both in vitro and in vivo (Fig. 2E and S1B). These results suggested that TPH-1 mediated serotonin production could promote glioma progression. In an attempt to fur- ther confirm the correlation between TPH-1 expression and serotonin production in vivo, we collected 20 tumor tissues from glioma patients and determined the expres- sion of TPH-1 and serotonin production by immunofluo- rescence and ELISA. The intensity of TPH-1 expression was quantified using the Image-J software. Increased ser- otonin levels were observed in TPH-1high glioma tissues relative to the TPH-1low glioma tissues (Fig. 2F and G). The correlational analysis between serotonin production and TPH-1 expression in Fig. 2F and G was performed in glioma tissues, wherein serotonin was found to be posi- tively correlated with THP1 (Fig. 2H). TPH‑1 promotes glioma progression and correlates with a poor progression Collectively, these findings highlighted that TPH-1 mediated tryptophan hydroxylation could produce serotonin, thereby promot- ing glioma progression. of TPH1 in regulating the responses to TMZ, mice were subcutaneously injected with 1 × 106 vector or TPH1 overexpressing LN229 cells, and treated with TMZ on day 13. A TUNEL assay was performed to evaluate the cellular cytotoxicity in vivo. Indeed, TPH1 overexpres- sion suppressed the cytotoxicity of TMZ in vivo (Fig. S1A). Taken together, these results implied that TPH-1 played a role in promoting glioma progression and was correlated with poor progression. of TPH1 in regulating the responses to TMZ, mice were subcutaneously injected with 1 × 106 vector or TPH1 overexpressing LN229 cells, and treated with TMZ on day 13. A TUNEL assay was performed to evaluate the cellular cytotoxicity in vivo. Indeed, TPH1 overexpres- sion suppressed the cytotoxicity of TMZ in vivo (Fig. S1A). Taken together, these results implied that TPH-1 played a role in promoting glioma progression and was correlated with poor progression. Serotonin derived from tryptophan hydroxylation drives proliferation and invasion of glioma cells We aimed to elucidate the mechanism underlying TPH- 1-induced glioma progression. TPH-1 is reportedly involved in amino acid metabolism and can catalyze the hydroxylation of L-Trp to serotonin. Recent studies sug- gest that serotonin can promote cell proliferation or inva- sion in breast/liver cancers [11]. This led us to reasonably speculate that TPH-1 may mediate the Trp hydroxylation to produce serotonin, thereby promoting glioma develop- ment. To confirm our hypothesis, we cultured tumor cells for 48 h, and examined the Trp consumption and sero- tonin production using the supernatant. Enhanced Trp consumption (Fig. 2A) and serotonin production (Fig. 2B) were observed in TPH-1 overexpressing LN229/T98G cells relative to LN229/T98G cells, suggesting that TPH-1 could catalyze the hydroxylation of L-Trp to generate ser- otonin. Subsequently, we treated LN229/T98G cells with serotonin and evaluated their proliferation/migration abilities. Serotonin treatment significantly strengthened Serotonin upregulates L1CAM signaling to regulate glioma progression To clarify the mechanism underlying serotonin promot- ing glioma development, we examined the mRNA expres- sion profile of glioma cancer patients as shown in Fig. 1B. The top 30 overexpressed genes in the high TPH-1 group were identified (Fig. 3A). Notably, an increased mRNA expression of L1CAM was observed in the TPH-1 high Fig. 2 Serotonin drove glioma cells proliferation and invasion. A and B 105 LN229/T98G (VEC) or TPH-1 overexpressing LN229/T98G cells (OE) were cultured in DMEM culture medium for 48 h. Trp consumption (A) and serotonin production (B) was determined by Elisa assay. C cell proliferation of LN229/T98G cells treated with PBS or serotonin (10 nM). D relative migrating cells of LN229/T98G cells treated with PBS or serotonin (10 nM). E LN229 or T98G cells were pre-treated with PBS or serotonin (10 nM) for 48 h. Then cell apoptosis of LN229/T98G treated with PBS or TMZ (48 h) were determined. F immunofluorescence staining of TPH-1 in TPH-1high and TPH-1low glioma tissues from patients. The scale bar was 25 μm. G serotonin secretion in TPH-1high and TPH-1low glioma tissues from patients, detected by Elisa analysis. H the correlation analysis between TPH-1 and serotonin in 20 glioma tissues (R2 = 0.466) Fig. 2 Serotonin drove glioma cells proliferation and invasion. A and B 105 LN229/T98G (VEC) or TPH-1 overexpressing LN229/T98G cells (OE) were cultured in DMEM culture medium for 48 h. Trp consumption (A) and serotonin production (B) was determined by Elisa assay. C cell proliferation of LN229/T98G cells treated with PBS or serotonin (10 nM). D relative migrating cells of LN229/T98G cells treated with PBS or serotonin (10 nM). E LN229 or T98G cells were pre-treated with PBS or serotonin (10 nM) for 48 h. Then cell apoptosis of LN229/T98G treated with PBS or TMZ (48 h) were determined. F immunofluorescence staining of TPH-1 in TPH-1high and TPH-1low glioma tissues from patients. The scale bar was 25 μm. G serotonin secretion in TPH-1high and TPH-1low glioma tissues from patients, detected by Elisa analysis. H the correlation analysis between TPH-1 and serotonin in 20 glioma tissues (R2 = 0.466) Fig. 2 Serotonin drove glioma cells proliferation and invasion. A and B 105 LN229/T98G (VEC) or TPH-1 overexpressing LN229/T98G cells (OE) were cultured in DMEM culture medium for 48 h. Trp consumption (A) and serotonin production (B) was determined by Elisa assay. Serotonin upregulates L1CAM signaling to regulate glioma progression E cell proliferation of PBS cultured LN229/T98G (scrambled siRNA treatment), serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. F relative migrating cells of PBS cultured LN229/T98G (scrambled siRNA treatment), serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. G PBS cultured LN229/T98G (scrambled siRNA treatment) was collected, and serotonin (10 nM) cultured LN229/T98G cells were pre-treated with scrambled and L1CAM siRNA. Then cell apoptosis of LN229/T98G treated with PBS or TMZ (1 μg/ml, 48 h) were determined. H overall survival of glioma patients with high L1CAM (n = 49) and low L1CAM (n = 50) expression ignaling to regulate glioma progression. A Heatmap of top 30 overexpressing genes in TPH-1high glioma Fig. 3 Serotonin upregulated L1CAM signaling to regulate glioma progression. A Heatmap of top 30 overexpressing genes in TPH-1high glioma patients (n = 49) in comparison with TPH-1low glioma patients (n = 50). B GO enrichment analysis in TPH-1high glioma patients (n = 49) in comparison with TPH-1low glioma patients (n = 50). C western blotting of L1CAM in LN229/T98G cells treated with PBS or serotonin (10 nM), and TPH-1 overexpressing LN229/T98G cells. D relative expression of L1CAM in LN229/T98G cells treated with scrambled and L1CAM siRNA, determined by qPCR. E cell proliferation of PBS cultured LN229/T98G (scrambled siRNA treatment), serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. F relative migrating cells of PBS cultured LN229/T98G (scrambled siRNA treatment), serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. G PBS cultured LN229/T98G (scrambled siRNA treatment) was collected, and serotonin (10 nM) cultured LN229/T98G cells were pre-treated with scrambled and L1CAM siRNA. Then cell apoptosis of LN229/T98G treated with PBS or TMZ (1 μg/ml, 48 h) were determined. H overall survival of glioma patients with high L1CAM (n = 49) and low L1CAM (n = 50) expression relationship between serotonin production and L1CAM signaling in glioma. To further confirm the role of L1CAM in the regulation of glioma progression, siRNA constructs were used to deplete L1CAM in LN229/T98G cells along with serotonin treatment (Fig. 3D). In these cells, L1CAM silencing significantly suppressed cellular proliferation (Fig. 3E) and migration (Fig. 3F) induced by serotonin. Consistently, the suppressive effects were also observed in cellular apoptosis analysis (Fig. 3G), suggest- ing that L1CAM played a role in glioma development. Serotonin upregulates L1CAM signaling to regulate glioma progression C cell proliferation of LN229/T98G cells treated with PBS or serotonin (10 nM). D relative migrating cells of LN229/T98G cells treated with PBS or serotonin (10 nM). E LN229 or T98G cells were pre-treated with PBS or serotonin (10 nM) for 48 h. Then cell apoptosis of LN229/T98G treated with PBS or TMZ (48 h) were determined. F immunofluorescence staining of TPH-1 in TPH-1high and TPH-1low glioma tissues from patients. The scale bar was 25 μm. G serotonin secretion in TPH-1high and TPH-1low glioma tissues from patients, detected by Elisa analysis. H the correlation analysis between TPH-1 and serotonin in 20 glioma tissues (R2 = 0.466) Zhang et al. BMC Cancer (2022) 22:457 Page 6 of 11 Fig. 3 Serotonin upregulated L1CAM signaling to regulate glioma progression. A Heatmap of top 30 overexpressing genes in TPH-1high glioma patients (n = 49) in comparison with TPH-1low glioma patients (n = 50). B GO enrichment analysis in TPH-1high glioma patients (n = 49) in comparison with TPH-1low glioma patients (n = 50). C western blotting of L1CAM in LN229/T98G cells treated with PBS or serotonin (10 nM), and TPH-1 overexpressing LN229/T98G cells. D relative expression of L1CAM in LN229/T98G cells treated with scrambled and L1CAM siRNA, determined by qPCR. E cell proliferation of PBS cultured LN229/T98G (scrambled siRNA treatment), serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. F relative migrating cells of PBS cultured LN229/T98G (scrambled siRNA treatment), serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. G PBS cultured LN229/T98G (scrambled siRNA treatment) was collected, and serotonin (10 nM) cultured LN229/T98G cells were pre-treated with scrambled and L1CAM siRNA. Then cell apoptosis of LN229/T98G treated with PBS or TMZ (1 μg/ml, 48 h) were determined. H overall survival of glioma patients with high L1CAM (n = 49) and low L1CAM (n = 50) expression Fig. 3 Serotonin upregulated L1CAM signaling to regulate glioma progression. A Heatmap of top 30 overexpressing genes in TPH-1high glioma patients (n = 49) in comparison with TPH-1low glioma patients (n = 50). B GO enrichment analysis in TPH-1high glioma patients (n = 49) in comparison with TPH-1low glioma patients (n = 50). C western blotting of L1CAM in LN229/T98G cells treated with PBS or serotonin (10 nM), and TPH-1 overexpressing LN229/T98G cells. D relative expression of L1CAM in LN229/T98G cells treated with scrambled and L1CAM siRNA, determined by qPCR. Serotonin upregulates L1CAM signaling to regulate glioma progression Additionally, L1CAM silencing also suppressed the pro- tumor effects induced due to TPH1 overexpression (Fig. glioma cancer tissues, and L1CAM reportedly serves as an L1 cell adhesion molecule, thereby participating in the process of cellular migration of tumor cells. Consist- ently, the GO enrichment analysis also indicated that the cell adhesion molecule signaling was involved in TPH-1 associated tumor progression (Fig. 3B). Therefore, we sought to examine the role of L1CAM in regulating glioma development. Thus, the expression of L1CAM was determined in THP1 overexpressing and serotonin- treated tumor cells. THP1 overexpression and seroto- nin treatment resulted in the upregulation of L1CAM in LN229 and T98G cells (Fig. 3C), which suggested a direct Zhang et al. BMC Cancer (2022) 22:457 Zhang et al. BMC Cancer (2022) 22:457 Page 7 of 11 Page 7 of 11 NF-κB levels in tumor cells, whereas L1CAM silenc- ing suppressed the upregulation of NF-κB (Fig. 4A). Blockade of the NF-κB signaling cascade by QNZ effi- ciently suppressed the proliferative properties (Fig. 4B) and migratory potential (Fig. 4C) induced by seroto- nin. Similarly, QNZ treatment suppressed the TMZ resistance in serotonin-treated LN229 and T98G cells (Fig. 4D). These results showed that L1CAM upregu- lated NF-κB expression to regulate glioma devel- opment. Next, we examined the expressions of the downstream factors of NF-κB, including c-Myc, IL-8, IL-1β, VEGF, CDK2, and COX2. Serotonin treatment mediated the upregulation of c-Myc and IL-1β in LN229 and T98G cells, whereas QNZ treatment sup- pressed the gene upregulation induced by serotonin (Fig. S1F). Subsequently, we evaluated the influence of NFKB1 expression on the prognosis of glioma patients. A poor overall survival was observed in patients with high NFKB1 expression (Fig. 4E). Collectively, these results suggested that serotonin upregulated NF-κB expression through L1CAM signaling, thus resulting in tumor progression and a poor prognosis in glioma patients. S1C, D, and E). Furthermore, we observed that L1CAM expression was positively correlated with a poor prog- nosis of glioma patients (Fig. 3H). Taken together, these results suggested that serotonin facilitated glioma devel- opment by regulating L1CAM expression. Serotonin upregulates NF‑κB expression through L1CAM signaling g g Next, we aimed to elucidate the mechanism underly- ing L1CAM-controlled tumor progression in glioma. As shown in Fig. 3B, the results of the GO enrichment analysis indicated that NF-κB signaling was substan- tially involved in TPH-1 associated biological activities. In fact, accumulating evidence suggests that L1CAM can induce constitutive NF-κB activation in pancreatic adenocarcinoma cells. The transcription factor, NF-κB, plays an essential role as a stressor in the cellular envi- ronment and controls the expression of various regula- tory genes, that directly regulate cellular proliferation, migration, and drug resistance in several tumor types. To determine the role of NF-κB in glioma, we examined the protein expression of NF-κB in serotonin-treated LN229/T98G cells. Serotonin treatment increased the Fig. 4 Serotonin upregulated NF-κB expression through LICAM signaling. A Western blotting of NF-κB in LN229/T98G cells, serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. B Cell proliferation of serotonin (10 nM) cultured LN229/T98G cells treated with PBS and QNZ (10 nM). C Relative migrating cells of serotonin (10 nM) cultured LN229/T98G cells treated with PBS and QNZ (10 nM). D Serotonin (10 nM) cultured LN229/T98G cells were pre-treated with QNZ (10 nM, 48 h). Then cell apoptosis of LN229/T98G treated with PBS or TMZ (1 μg/ml, 48 h) were determined. E Overall survival of glioma patients with high NFKB1 (n = 49) and low NFKB1 (n = 50) expression Fig. 4 Serotonin upregulated NF-κB expression through LICAM signaling. A Western blotting of NF-κB in LN229/T98G cells, serotonin (10 nM) cultured LN229/T98G cells treated with scrambled and L1CAM siRNA. B Cell proliferation of serotonin (10 nM) cultured LN229/T98G cells treated with PBS and QNZ (10 nM). C Relative migrating cells of serotonin (10 nM) cultured LN229/T98G cells treated with PBS and QNZ (10 nM). D Serotonin (10 nM) cultured LN229/T98G cells were pre-treated with QNZ (10 nM, 48 h). Then cell apoptosis of LN229/T98G treated with PBS or TMZ (1 μg/ml, 48 h) were determined. E Overall survival of glioma patients with high NFKB1 (n = 49) and low NFKB1 (n = 50) expression Zhang et al. BMC Cancer (2022) 22:457 Page 8 of 11 Zhang et al. BMC Cancer (2022) 22:457 TMZ resistance in LN229 cells. Moreover, inhibition of TPH-1 by LX-1031 significantly suppressed the tumor growth and improved the outcomes of TMZ treatment (Fig. 5D). Serotonin upregulates NF‑κB expression through L1CAM signaling A similar result was recorded in the survival analysis for tumor-bearing mice (Fig. 5E). Those results suggested that TPH-1 may serve as a novel indicator for tumor diagnoses and provided innovative targets for gli- oma therapy. Interruption of TPH‑1 signals attenuates the aggressiveness of glioma in vivo Our previous results suggested the role of TPH-1 and serotonin in promoting glioma cell proliferation. Therefore, we were interested in evaluating the influ- ence of TPH-1 on glioma tumor growth in vivo. Thus, TPH-1 overexpressing LN229 cells were subcutaneously implanted into immunodeficient mice, and the tumor volumes were recorded. Consistent with in vitro results, a rapid tumor growth curve was observed in TPH-1 over- expressing LN229 bearing mice relative to the vec-LN229 bearing mice (Fig. 5A). In line with the high expression of TPH-1, higher protein levels of serotonin (Fig. 5B), L1CAM, and NF-κB (Fig. 5C) were found in TPH-1 over- expressing LN229 bearing mice (H&E staining in Fig. S1H). To retard the sustained tumor growth induced by TPH-1, we combined TMZ with the TPH-1 inhibitor, LX-1031, and treated the TPH-1 overexpressing LN229 bearing mice. Intriguingly, limited anti-cancerous effects were observed in the TMZ groups, consistent with our in vitro findings which suggested that TPH-1 promoted Discussion Mechanistically, L1CAM employed several downstream pathways through its interac- tion with different cell surface receptors. Steve et al. report that ectopic L1CAM expression in 3 T3 and K1735-C11 cells induces sustained extracellular signal- regulated kinase (ERK) activation, which may require the participation of growth factor [37]. L1CAM can also be cleaved by the metalloproteinases, ADAM10 and ADAM17, which further results in the release of a 200 kDa soluble ectodomain fulfilling diverse functions in both tumor and immune cells [38]. In the present study, we found that serotonin treatment could signifi- cantly elevate LICAM expression, which further con- tributed to the activation of NF-κB, a crucial player in human neoplasms as it empowers several key attributes of cancer cells. Consistently, GO enrichment analysis indicated the engagement of NF-κB signaling in TPH- 1-related biological activities. Therefore, targeting TPH-1 may be a feasible modality for curing glioma. Our in vivo experiments demonstrated that LX-1031 co-operated with TMZ to suppress tumor growth and prolong survival, thus eventually resulting in favorable outcomes in glioma-bearing mice. Based on the above results, our findings demonstrated that TPH-1 con- tributed to glioma development through the serotonin/ L1CAM/NF-κB signaling pathway and offered possibil- ities for the application of TPH-1 inhibitors in glioma therapy. Recent advances in energy metabolism link amino acid metabolism to the malignant etiologies of cancer [23]. Trp catabolism is among the most researched topics, and several enzymes, as well as metabolites involved in tryp- tophan degradation, contribute to the aggressive traits of tumor cells [24]. Apart from the Kyn pathway, accumu- lating evidence shows the carcinogenic properties of the serotonergic pathway [25, 26]. During tumor progression, breast cancer tissues display elevated TPH-1 expression corresponding to enhanced serotonin synthesis [27]. Gianfranco et al. report that serotonin treatment causes a significant increase in the proliferative capacity of the cholangiocarcinoma cells and this change can be reversed by TPH-1 blockade [28]. Our study focused on the role of TPH-1 in glioma and the findings indicated that TPH-1 hydroxylated Trp led to serotonin production, thereby driving robust tumor growth and progression. Moreo- ver, our data linked TPH-1 to the chemotherapeutic sen- sitivity of glioma, demonstrating that TMZ resistance was aggravated in the LN229/T98G cells overexpress- ing TPH-1. Discussion Glioma, which exhibits strong invasiveness and a poor response to treatment, is generally difficult to cure due to the lack of effective pharmacological targets. Herein, we extended the prior work, and our findings under- lined TPH-1 participation in regulating the proliferative properties, migratory capacity, and chemotherapeutic sensitivity of glioma cells. Moreover, TPH-1 blockade produced a substantial suppression of tumor growth in a mouse model of glioma when administered in con- junction with chemotherapeutic agents. These data may Fig. 5 Interrupt of TPH-1 signals attenuated the aggressiveness of glioma in vivo. A LN229 (VEC) and TPH-1 overexpressing LN229 cells (OE) were subcutaneously injected into mice. The tumor volume was recorded. B-C The expression of serotonin (B), L1CAM and NF-κB (C) in tumor tissues from (A) was determined by Elisa assay or immunofluorescence. The scale bar was 25 μm. D Tumor volume of TPH-1 overexpressing LN229 bearing mice treated with PBS, TMZ, LX-1031 and TMZ combining LX-1031. E Survival time of TPH-1 overexpressing LN229 bearing mice treated with PBS, TMZ, LX-1031 and TMZ combining LX-1031 Fig. 5 Interrupt of TPH-1 signals attenuated the aggressiveness of glioma in vivo. A LN229 (VEC) and TPH-1 overexpressing LN229 cells (OE) were subcutaneously injected into mice. The tumor volume was recorded. B-C The expression of serotonin (B), L1CAM and NF-κB (C) in tumor tissues from (A) was determined by Elisa assay or immunofluorescence. The scale bar was 25 μm. D Tumor volume of TPH-1 overexpressing LN229 bearing mice treated with PBS, TMZ, LX-1031 and TMZ combining LX-1031. E Survival time of TPH-1 overexpressing LN229 bearing mice treated with PBS, TMZ, LX-1031 and TMZ combining LX-1031 Zhang et al. BMC Cancer (2022) 22:457 Page 9 of 11 Page 9 of 11 broaden the understanding of the complexity of glioma and aid in the development of therapeutic strategies. expression causes complete cessation of cellular migra- tion [36]. In line with previous findings, we confirmed the growth-promoting effects of L1CAM and addi- tionally found that L1CAM knockdown increased the chemotherapeutic sensitivity of LN229 and T98G cells. Importantly, we have highlighted the involvement of L1CAM in TPH-1 mediated glioma progression, as indicated by the upregulation of L1CAM in THP1 over- expressing glioma cells. Moreover, clinical data analysis demonstrated that L1CAM expression was a risk fac- tor for glioma patients, with higher values pertaining to a poor prognosis. Abbreviations 2. Xu S, Tang L, Li X, et al. Immunotherapy for glioma: current management and future application. Cancer Lett. 2020;476. https://doi.org/10.1016/j. canlet.2020.02.002. 2. Xu S, Tang L, Li X, et al. Immunotherapy for glioma: current management and future application. Cancer Lett. 2020;476. https://doi.org/10.1016/j. canlet.2020.02.002. 2. Xu S, Tang L, Li X, et al. Immunotherapy for glioma: current management and future application. Cancer Lett. 2020;476. https://doi.org/10.1016/j. canlet.2020.02.002. TCGA: The Cancer Genome Atlas; TPH-1: Tryptophan hydroxylase 1; TMZ: Temozolomide; L1CAM: L1-cell adhesion molecule; Kyn: Kynurenine; IDO: Indoleamine-2, 3-dioxygenase; TDO: Tryptophan-2, 3-dioxygenase; TNBC: Triple-negative breast cancer; TGF-β1: Transforming growth factor-beta1; ERK: Extracellular signal-regulated kinase. 3. Wang J, Yan L, Ai P, et al. Observation versus radiotherapy with or without temozolomide in postoperative WHO grade II high-risk low-grade glioma: a retrospective cohort study. Neurosurg Rev. 2021;44(3):1447–55. https:// doi.org/10.1007/s10143-020-01326-y. 4. Wang J, Wang Y, He Y, et al. Radiotherapy versus radiotherapy combined with temozolomide in high-risk low-grade gliomas after surgery: study protocol for a randomized controlled clinical trial. Trials. 2019;20(1):641. https://doi.org/10.1186/s13063-019-3741-5. Conclusion Received: 3 January 2022 Accepted: 13 April 2022 Received: 3 January 2022 Accepted: 13 April 2022 In conclusion, we demonstrated that TPH-1 functioned through the serotonin/L1CAM/NF-κB pathway, thus facilitating glioma cell proliferation, migration, and chemoresistance. Targeting TPH-1 may open new ave- nues for the clinical treatment of glioma. Page 10 of 11 Page 10 of 11 Page 10 of 11 Availability of data and materials 12. Balakrishna P, George S, Hatoum H, et al. Serotonin pathway in Cancer. Int J Mol Sci. 2021;22(3). https://doi.org/10.3390/ijms22031268. 12. Balakrishna P, George S, Hatoum H, et al. Serotonin pathway in Cancer. Int J Mol Sci. 2021;22(3). https://doi.org/10.3390/ijms22031268. The datasets generated and/or analysed during the current study are available on https://figshare.com/s/829e7f54b49e7f6d3713. 13. Gautam J, Banskota S, Regmi SC, et al. Tryptophan hydroxylase 1 and 5-HT receptor preferentially expressed in triple-negative breast cancer promote cancer progression through autocrine serotonin signaling. Mol Cancer. 2016;15(1):75. Consent for publication Not applicable. Not applicable. 17. Taniguchi K, Karin M. NF-κB, inflammation, immunity and cancer: coming of age. Nat Rev Immunol. 2018;18(5):309–24. https://doi.org/10.1038/nri. 2017.142. References 1. Reifenberger G, Wirsching H-G, Knobbe-Thomsen CB, et al. Advances in the molecular genetics of gliomas - implications for classification and therapy. Nat Rev Clin Oncol. 2017;14(7):434–52. https://doi.org/10.1038/ nrclinonc.2016.204. Competing interests The authors declare that they have no competing interests. 18. Li F, Zhang J, Arfuso F, et al. NF-κB in cancer therapy. Arch Toxicol. 2015;89(5):711–31. https://doi.org/10.1007/s00204-015-1470-4. Acknowledgements Not applicable. 8. Platten M, Nollen EAA, Röhrig UF, et al. Tryptophan metabolism as a common therapeutic target in cancer, neurodegeneration and beyond. Nat Rev Drug Discov. 2019;18(5):379–401. https://doi.org/10.1038/ s41573-019-0016-5. 8. Platten M, Nollen EAA, Röhrig UF, et al. Tryptophan metabolism as a common therapeutic target in cancer, neurodegeneration and beyond. Nat Rev Drug Discov. 2019;18(5):379–401. https://doi.org/10.1038/ s41573-019-0016-5. Funding None. 11. Sarrouilhe D, Mesnil M. Serotonin and human cancer: a critical view. Biochimie. 2019;161:46–50. https://doi.org/10.1016/j.biochi.2018.06.016. 11. Sarrouilhe D, Mesnil M. Serotonin and human cancer: a critical view. Biochimie. 2019;161:46–50. https://doi.org/10.1016/j.biochi.2018.06.016. Ethics approval and consent to participate 14. Li T, Fu B, Zhang X, et al. Overproduction of gastrointestinal 5-HT pro- motes colitis-associated colorectal Cancer progression via enhancing NLRP3 Inflammasome activation. Cancer Immunol Res. 2021;9(9):1008– 23. https://doi.org/10.1158/2326-6066.CIR-20-1043. This study was approved by the Ethics Committee of Ya’ an People’s Hospital, strictly following the Declaration of Helsinki. All patients agreed to participate in this research, and provided informed consent. The animal experiments were approved by the Institutional Animal Care and Use Committee of Ya’ an people’s Hospital and conducted in accordance with the ARRIVE guidelines as well as the National Institutes of Health guide for the care and use of Labora- tory animals. 15. Masab M, Saif MW. Telotristat ethyl: proof of principle and the first oral agent in the management of well-differentiated metastatic neuroendo- crine tumor and carcinoid syndrome diarrhea. Cancer Chemother Phar- macol. 2017;80(6):1055–62. https://doi.org/10.1007/s00280-017-3462-y.l 16. Hoesel B, Schmid JA. The complexity of NF-κB signaling in inflam- mation and cancer. Mol Cancer. 2013;12:86. https://doi.org/10.1186/ 1476-4598-12-86. Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12885-022-09569-2. g 5. McCawley LJ, Matrisian LM. Tumor progression: defining the soil round the tumor seed. Curr Biol. 2001;11(1):R25–7. the tumor seed. Curr Biol. 2001;11(1):R25–7. 6. Pavlova NN, Thompson CB. The emerging hallmarks of Cancer metabo- lism. Cell Metab. 2016;23(1):27–47. https://doi.org/10.1016/j.cmet.2015. 12.006. 7. Lieu EL, Nguyen T, Rhyne S, et al. Amino acids in cancer. Exp Mol Med. 2020;52(1):15–30. https://doi.org/10.1038/s12276-020-0375-3. 7. Lieu EL, Nguyen T, Rhyne S, et al. Amino acids in cancer. Exp Mol Med. 2020;52(1):15–30. https://doi.org/10.1038/s12276-020-0375-3. Authors’ contributions Zhangchao Guo conceived the project and wrote the manuscript. Jie Zhang, Zhangchao Guo and Qiangli Xie performed the experiments. Chuanhong Zhong, Xiangyu Gao, and Qiumei Yang carried out analysis and interpretation of data. All authors reviewed and approved the manuscript. 9. Opitz CA, Somarribas Patterson LF, Mohapatra SR, et al. The therapeutic potential of targeting tryptophan catabolism in cancer. Br J Cancer. 2020;122(1):30–44. https://doi.org/10.1038/s41416-019-0664-6. 9. Opitz CA, Somarribas Patterson LF, Mohapatra SR, et al. The therapeutic potential of targeting tryptophan catabolism in cancer. Br J Cancer. 2020;122(1):30–44. https://doi.org/10.1038/s41416-019-0664-6. 10. Berger M, Gray JA, Roth BL. The expanded biology of serotonin. Annu Rev Med. 2009;60:355–66. https://doi.org/10.1146/annurev.med.60.042307. 110802. 10. Berger M, Gray JA, Roth BL. The expanded biology of serotonin. Annu Rev Med. 2009;60:355–66. https://doi.org/10.1146/annurev.med.60.042307. 110802. Discussion To further investigate the clinical implica- tions of the serotonergic pathway, we performed analysis using TCGA dataset, which suggested that a high TPH-1 expression was associated with a poor prognosis in gli- oma patients. Therefore, it was imperative to disentangle the mechanisms underlying TPH-1-mediated regulation of glioma development. L1CAM, a transmembrane glycoprotein of the immu- noglobulin superfamily, was originally described in the nervous system, whereby it plays a role in brain devel- opment and functions [29]. Subsequent research attests to the presence of L1CAM in several cancer cell types, and a high L1CAM expression correlates with advanced tumor stages and grave prognoses [30]. L1CAM endows cancer cells with enhanced tumorigenic properties and motility, which can be reversed by gene silencing as well as antibodies [31]. L1CAM has been exploited as a diagnostic marker and more importantly as a prom- ising therapeutic target for the treatment of malignan- cies. The control of L1CAM expression in cancer has therefore drawn widespread research attention. Previ- ous studies indicate that L1CAM, situated specifically at the invasive front of tumor tissues, serves as a target for activation by β-catenin-TCF signaling in colorectal cancer [32]. Subsequent studies have provided evidence that the transforming growth factor-beta1 (TGF-β1) augments L1CAM expression in pancreatic and endo- metrial cancer cells, which is dependent on Slug, a tran- scription factor that modulates epithelial-mesenchymal transition [33–35]. In glioma, shutting down of L1CAM In light of the limitations to previous studies, herein, we highlight the relevance of TPH-1 in glioma advance- ment. We showed that (1) TPH-1 played a stimulatory function for glioma cell proliferation, motility, and drug resistance. (2) The tumor-promoting effects of TPH-1 were dependent on the serotonin/L1CAM/NF-κB signaling pathway. (3) The cessation of glioma growth could be accomplished by TMZ in combination with an L1CAM inhibitor, hinting at a potential target for ther- apeutic intervention. (4) The levels of TPH-1, L1CAM, and NF-κB in tumor tissues may serve as potential bio- markers for monitoring glioma progression and pre- dicting prognoses. Zhang et al. BMC Cancer (2022) 22:457 Zhang et al. BMC Cancer (2022) 22:457 Author details 1 Altevogt P, Ben-Ze’ev A, Gavert N, et al. Recent insights into the role of L1CAM in cancer initiation and progression. Int J Cancer. 2020;147(12):3292–6. https://doi.org/10.1002/ijc.33177. 31. Altevogt P, Doberstein K, Fogel M. L1CAM in human cancer. Int J Cancer. 2016;138(7):1565–76. https://doi.org/10.1002/ijc.29658. 32. Gavert N, Conacci-Sorrell M, Gast D, et al. L1, a novel target of beta- catenin signaling, transforms cells and is expressed at the invasive front of colon cancers. J Cell Biol. 2005;168(4):633–42. 33. Pfeifer M, Schirmer U, Geismann C, et al. L1CAM expression in endome- trial carcinomas is regulated by usage of two different promoter regions. BMC Mol Biol. 2010;11:64. https://doi.org/10.1186/1471-2199-11-64. 34. Geismann C, Morscheck M, Koch D, et al. Up-regulation of L1CAM in pan- creatic duct cells is transforming growth factor beta1- and slug-depend- ent: role in malignant transformation of pancreatic cancer. Cancer Res. 2009;69(10):4517–26. https://doi.org/10.1158/0008-5472.CAN-08-3493. 35. Huszar M, Pfeifer M, Schirmer U, et al. Up-regulation of L1CAM is linked to loss of hormone receptors and E-cadherin in aggressive subtypes of endometrial carcinomas. J Pathol. 2010;220(5):551–61. https://doi.org/10. 1002/path.2673. 36. Mohanan V, Temburni MK, Kappes JC, et al. L1CAM stimulates glioma cell motility and proliferation through the fibroblast growth factor receptor. Clin Exp Metastasis. 2013;30(4):507–20. https://doi.org/10.1007/ s10585-012-9555-4. 37. Silletti S, Yebra M, Perez B, et al. Extracellular signal-regulated kinase (ERK)- dependent gene expression contributes to L1 cell adhesion molecule- dependent motility and invasion. J Biol Chem. 2004;279(28):28880–8. 38. Riedle S, Kiefel H, Gast D, et al. Nuclear translocation and signalling of L1-CAM in human carcinoma cells requires ADAM10 and presenilin/ gamma-secretase activity. Biochem J. 2009;420(3):391–402. https://doi. org/10.1042/BJ20081625. 38. Riedle S, Kiefel H, Gast D, et al. Nuclear translocation and signalling of L1-CAM in human carcinoma cells requires ADAM10 and presenilin/ gamma-secretase activity. Biochem J. 2009;420(3):391–402. https://doi. org/10.1042/BJ20081625. Author details 1 1 Department of Neurosurgery, Ya’ an people’s Hospital, Ya’ an 625000, People’s Republic of China. 2 Department of Neurosurgery, Affiliated Hospital of South- west Medical University, Luzhou, People’s Republic of China. 3 Department of Cardiology, Chengdu Qingbaijiang District People’s Hospital, Chengdu, People’s Republic of China. 4 Pediatrics of Ya’ an people’s Hospital, Ya’ an, Peo- ple’s Republic of China. 5 Department of Geriatrics, Luzhou people’s Hospital, Luzhou, People’s Republic of China. 19. Oeckinghaus A, Hayden MS, Ghosh S. Crosstalk in NF-κB signaling path- ways. Nat Immunol. 2011;12(8):695–708. https://doi.org/10.1038/ni.2065. 20. Akıncılar SC, Wu L, Ng QF, et al. NAIL: an evolutionarily conserved lncRNA essential for licensing coordinated activation of p38 and NFκB in colitis. Gut. 2021;70(10):1857–71. https://doi.org/10.1136/gutjnl-2020-322980. 21. Mohan S, Abdelwahab SI, Kamalidehghan B, et al. Involvement of NF-κB and Bcl2/Bax signaling pathways in the apoptosis of MCF7 cells induced by a xanthone compound Pyranocycloartobiloxanthone a. Phytomedi- cine. 2012;19(11):1007–15. https://doi.org/10.1016/j.phymed.2012.05.012. Page 11 of 11 Zhang et al. BMC Cancer (2022) 22:457 Zhang et al. BMC Cancer (2022) 22:457 22. Carrero R, Cerrada I, Lledó E, et al. IL1β induces mesenchymal stem cells migration and leucocyte chemotaxis through NF-κB. Stem Cell Rev Rep. 2012;8(3):905–16. https://doi.org/10.1007/s12015-012-9364-9. 23. Sivanand S, Vander Heiden MG. Emerging roles for branched-chain amino acid metabolism in Cancer. Cancer Cell. 2020;37(2):147–56. https:// doi.org/10.1016/j.ccell.2019.12.011. 24. Liu X-H, Zhai X-Y. Role of tryptophan metabolism in cancers and thera- peutic implications. Biochimie. 2021;182:131–9. https://doi.org/10.1016/j. biochi.2021.01.005. 24. Liu X-H, Zhai X-Y. Role of tryptophan metabolism in cancers and thera- peutic implications. Biochimie. 2021;182:131–9. https://doi.org/10.1016/j. biochi.2021.01.005. 25. Gwynne WD, Shakeel MS, Girgis-Gabardo A, et al. The role of serotonin in breast Cancer stem cells. Molecules. 2021;26(11). https://doi.org/10.3390/ molecules26113171. 25. Gwynne WD, Shakeel MS, Girgis-Gabardo A, et al. The role of serotonin in breast Cancer stem cells. Molecules. 2021;26(11). https://doi.org/10.3390/ molecules26113171. 26. Kannen V, Bader M, Sakita JY, et al. The dual role of serotonin in colorectal Cancer. Trends Endocrinol Metab. 2020;31(8):611–25. https://doi.org/10. 1016/j.tem.2020.04.008. 27. Pai VP, Marshall AM, Hernandez LL, et al. Altered serotonin physiology in human breast cancers favors paradoxical growth and cell survival. Breast Cancer Res. 2009;11(6):R81. https://doi.org/10.1186/bcr2448. 28. Alpini G, Invernizzi P, Gaudio E, et al. Serotonin metabolism is dysregu- lated in cholangiocarcinoma, which has implications for tumor growth. Cancer Res. 2008;68(22):9184–93. https://doi.org/10.1158/0008-5472. CAN-08-2133. 29. Schäfer MKE, Altevogt P. L1CAM malfunction in the nervous system and human carcinomas. Cell Mol Life Sci. 2010;67(14):2425–37. https://doi. org/10.1007/s00018-010-0339-1. 30. • 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: Publisher’s Note S i N i • 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: Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.
https://openalex.org/W4319868379	https://bmcpediatr.biomedcentral.com/counter/pdf/10.1186/s12887-023-03862-0	English	null	Survey on human milk feeding and enteral feeding practices for very-low-birth-weight infants in NICUs in China Neonatal Network	BMC pediatrics	2,023	cc-by	5,138	Survey on human milk feeding and enteral feeding practices for very‑low‑birth‑weight infants in NICUs in China Neonatal Network Xiaoshan Hu1†, Junjie Lu1†, Jun Zhang1, Min Zhang1, Zhangbin Yu1, Shoo K. Lee2, Shuping H Xiaohui Chen1* BMC Pediatrics BMC Pediatrics Hu et al. BMC Pediatrics (2023) 23:75 https://doi.org/10.1186/s12887-023-03862-0 Open Access Abstract Correspondence: Shuping Han shupinghan@njmu.edu.cn Xiaohui Chen chenxiaohui@njmu.edu.cn 1 Department of Pediatrics, Maternity Hospital Affiliated to Nanjing Medical University/ Nanjing Maternal and Child Health Hospital, Nanjing, Jiangsu, China Correspondence: Shuping Han shupinghan@njmu.edu.cn Xiaohui Chen chenxiaohui@njmu.edu.cn 1 Department of Pediatrics, Maternity Hospital Affiliated to Nanjing Medical University/ Nanjing Maternal and Child Health Hospital, Nanjing, Jiangsu, China *Correspondence: Shuping Han shupinghan@njmu.edu.cn Xiaohui Chen chenxiaohui@njmu.edu.cn 1 Department of Pediatrics, Maternity Hospital Affiliated to Nanjing Medical University/ Nanjing Maternal and Child Health Hospital, Nanjing, Jiangsu, China Abstract Background The breastfeeding rate in China is lower than that in many other countries and the extent of adop- tion of the “Feeding Recommendations for Preterm Infants and Low Birth Weight Infants” guideline in NICUs remains unclear. Method A web-based survey about the current status of human milk feeding and enteral feeding practices at NICUs was sent to all China Neonatal Network’s cooperation units on September 7, 2021, and the respondents were given a month to send their responses. Results All sixty NICUs responded to the survey, the reply rate was 100%. All units encouraged breastfeeding and provided regular breastfeeding education. Thirty-six units (60.0%) had a dedicated breastfeeding/pumping room, 55 (91.7%) provided kangaroo care, 20 (33.3%) had family rooms, and 33 (55.0%) routinely provided family integrated care. Twenty hospitals (33.3%) had their own human milk banks, and only 13 (21.7%) used donor human milk. Eight units (13.3%) did not have written standard nutrition management guidelines for infants with body weight < 1500 g. Most units initiated minimal enteral nutrition with mother’s milk for infants with birth weight ˂1500 g within 24 h after birth. Fifty NICUs (83.3%) increased the volume of enteral feeding at 10–20 ml/kg daily. Thirty-one NICUs (51.7%) assessed gastric residual content before every feeding session. Forty-one NICUs (68.3%) did not change the course of enteral nutrition management during drug treatment for patent ductus arteriosus, and 29 NICUs (48.3%) instated NPO for 1 or 2 feeds during blood transfusion. Conclusion There were significant differences in human milk feeding and enteral feeding strategies between the NICUs in CHNN, but also similarities. The data obtained would be useful in the establishment of national enteral feed- ing guidelines for preterm infants and quality improvement of cooperation at the national level. Keywords Human milk feeding, Enteral feeding practice, Neonatal intensive care units, Survey †Xiaoshan Hu and Junjie Lu contributed equally to this work and they should be regarded as joint first authors. 2 Department of Pediatrics, University of Toronto Faculty of Medicine, Toronto, ON, Canada © 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://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. Background The survey contained questions about the current sta- tus of HM feeding and enteral feeding strategies,and the unit director or representative was responsible for com- pleting the survey and was instructed to provide answers based on common unit-level practice rather than their personal opinions/practice alone. The questionnaire, which consisted of single-choice, multiple-choice, and open-ended questions, gathered information about the demographic features of the unit, HM feeding character- istics, presence of a milk bank or access to DHM, initia- tion and advancement of enteral feeding, and indications for and the use of human milk fortifier (HMF), among other data(see supplementary materials). The deadline for submitting the filled-in surveys was October 7, 2021. g Human milk (HM) is the most ideal source of nutrition for newborns, especially preterm infants, on account of its short-term benefits, such as reducing the risk of necrotizing enterocolitis (NEC), sepsis, chronic lung disease(CLD), and mortality, and its long-term benefits, such as improving long-term neurodevelopment [1]. Apart from its medical advantages, HM feeding also has economic advantages [2]. Lactation onset is often delayed after premature delivery, and as a result, preterm infants receive an insufficient amount of milk during the first few critical days of life [3]. Donor human milk (DHM) and preterm formula are used as supplements when MOM is insufficient, and DHM is preferred over preterm formula as it is better in terms of nutritional composition and bio- logical value [4, 5]. . Open Access T d d Hu et al. BMC Pediatrics (2023) 23:75 Hu et al. BMC Pediatrics (2023) 23:75 Page 2 of 7 Data collection and analysis g Therefore, the World Health Organization and the American Society for Parenteral Nutrition have empha- sized the use of HM in preterm infants in their guidelines for the nutritional management of preterm/low-birth- weight infants [6]. Similarly, China issued “Feeding Rec- ommendations for Preterm Infants and Low Birth Weight Infants” in 2009, according to which mother’s own milk (MOM) was recommended as the first choice for preterm infants [7]. Following these guidelines, the breastfeed- ing rate in China’s NICUs increased from 23% in 2009 to 58.2% in 2018 [8]. However, the breastfeeding rate is still far lower than that in other countries [9]. Furthermore, the situation of implementation of “Feeding Recom- mendations for Preterm Infants and Low Birth Weight Infants” for each NICU was still unclear. Therefore, the purpose of this study was to examine and compare the status of HM feeding and enteral feeding strategies in China’s level 3 NICUs. This is the first national-level sur- vey on this topic in China, so the data obtained would be useful in the establishment of national enteral feeding guidelines for preterm infants and quality improvement of cooperation at the national level. Participants were enrolled between September 7 and October 7,2021. A link to the online questionnaire was sent through WeChat. Participants only needed to fill in the basic information for the NICUs and no personal information was attached to the data. All questions had to be completed before submission, duplicate responses were verified and eliminated.h i The survey results were collected online and all analyses were conducted using Microsoft Excel. The demographic information was described by median (interquartile range,1–3). Descriptive results are expressed as numbers and percentage (%). Results Sixty NICUs responded to the survey: 14 (23.3%) were from children’s hospitals; 23 (38.3%) maternity hospitals; and 23 (38.3%) general hospitals. The median number of beds in these units was 100, and the median number of neonates hospitalized in 2020 was 3000, including 907 premature infants, 140 very-low-birth weight (VLBW) infants, and 148 infants of gestational age < 32 weeks (Table 1). Methods All 60 units encouraged breastfeeding at their NICUs. Parents were educated about breastfeeding before deliv- ery at 26 units (43%), at the time of neonate admission at 57 units (95%), or during hospitalization at 57 units (95%). The staff involved in breastfeeding education included neonatal physicians at 51 units (85%), obste- tricians at 21 units (36%), nurses in neonatal units at 60 units (100%), and nurses at human milk banks (HMBs) at 17 units (28%). The content and mode of communica- tion of breastfeeding education for parents are shown in Table 2. Most of the units would repeat the breastfeed- ing instructions to the parents before neonatal discharge from the hospital (Table 2). Feeding with mother’ own milk Only 13 units (21.7%) were licensed to use DHM. The main reasons for using DHM were to reduce the risk of NEC and support mother’s breastfeeding. The most com- mon reason for not using DHM was the lack of a license. DHM was most frequently used in cases of extremely/ very low birth weight infants, feeding intolerance, and intestinal malabsorption (Table 3). Out of the 60 units, 36 (60.0%) had a dedicated breast- feeding/pumping room for mothers. Only 2 units (3.3%) did not accept MOM during neonatal hospitalization, and 36 units (61.7%) did not routinely pasteurize MOM after collection. Further, 55 units (91.7%) provided kan- garoo care (KC), 20 units (33.3%) had family rooms, and 33 units (55.0%) routinely carried out family integrated care (FICare, parents spent ≥ 6 h per day in the NICU and provided non-medical care for their infants under the close supervision of the FICare nurse). Enteral feeding strategies Eight units (13.3%) did not have any standardized pro- tocol for nutrition management for infants with body weight < 1500 g. Twenty-nine units (48.3%) initiated min- imal enteral nutrition (MEN,involves feeding with low volumes of 12–24 ml/kg/day in the first week of life) for infants with birth weight < 1500 g within 24 h after birth, and used fresh mother’ milk. However, 13 units (21.7%) used < 1250 g body weight as the criterion for initiating MEN, 13 units (21.7%) used < 1000 g and 5 units (8.3%) used < 750 g (Fig. 1A and B). The duration of MEN ranged from 24 to 144 h (Fig. 1C). Survey methods b On September 7, 2021, a prospective, cross-sectional, web-based survey was sent to the director or repre- sentative of 60 level 3 NICUs participating in the Chi- nese Neonatal Network (CHNN).The CHNN is the first national neonatal network and has the largest geographi- cally representative cohort from NICUs in China. It was officially put into operation on January 1, 2019. CHNN hospitals are tertiary referral hospitals with Grade A level 3 NICUs authorized by the Health Administration of China and have recognized expertise in caring for high- risk neonates. Page 3 of 7 Hu et al. BMC Pediatrics (2023) 23:75 Table 1 Demographic information M means median, IQR means interquartile range, IQR1 means the 25th percentile and IQR3 means the 75th percentile (N = 60); VLBW very low birth weight; GA gestational age Number of total beds in 2020 Number of neonates hospitalized in 2020 Number of premature infants hospitalized in 2020 Number of VLBW infants hospitalized in 2020 Number of GA < 32w infants hospitalized in 2020 M (IQR,1–3) 100(63–150) 3000(1995–4153) 907(590–1505) 140 (89–251) 148 (87–273) Table 2 Education about breastfeeding Content NO Mode NO Repeat breastfeeding instructions before discharge NO benefits of breastfeeding 60 oral instruction 59 importance of breastfeeding 56 collection,storage and trans- port of breast milk 60 Video 37 how to maintain lactation 50 kangaroo care 52 paper materials 57 feeding advises after discharge 58 how to maintain lactation 59 bedside education 39 the phone number of Breastfeeding counselling clinic 44 Feedback of breastfeeding 52 WeChat or SMS platform 40 Fellow-up after discharge 59 Discussion Nutrition is important to standardize among different NICUs, even in terms of long-term outcomes. For exam- ple, the prevention of extra-uterine growth retardation (EUGR) in preterm infants through nutritional strategies is of extreme importance, as the achievement of adequate growth has been associated with a better neurodevel- opmental outcome through childhood [10], although there is still no consensus regarding the best definition of EUGR to use to predict neurological outcome [11].hi p g This is the first comprehensive study to investigate human milk feeding and enteral feeding practices in NICUs from all districts in mainland China. As evident in the survey responses, we found significant variations but also similarities in the status of human milk feeding and enteral feeding strategies among the examined NICUs. In general, all the NICUs encouraged breastfeeding, and breastfeeding education was provided by nurses in the neonatal units. The content of breastfeeding education covered benefits of breastfeeding, collection,storage,and transport of breast milk,kangaroo care, how to keep lactation,feedback of breastfeeding. Breastfeeding infor- mation was provided via channels such as oral instruc- tion, video, paper materials,bedside education,WeChat or SMS platform and others. During the recent COVID- 19 pandemic, parental visitation practices were sus- pended at many NICUs in China but once the pandemic was brought under control, most NICUs restarted their previous parental visitation policies. At most NICUs, communication about breastfeeding with parents was implemented online or through face-to-face meetings at least once a week. Further, at the majority of the hospitals (58 NICUs, 96.7%), MOM was obtained during neonatal hospitalization. Fig. 1 Minimum enteral nutrition practices at NICUs. A Infants that require minimal enteral nutrition after birth; B Time at which minimal enteral nutrition was started after birth for infants with body weight < 1500 g; C Duration of minimal enteral nutrition for premature infants The present findings showed significant variations between the NICUs with regard to the provision of family rooms and FICare, the use of DHM, and enteral feeding strategies.55 NICUs (91.7%) provided KC. Accordingly, previous studies have documented that most NICUs in China were aware of the benefits of KC [12, 13]. Com- pared to the high rates of implementation of KC, family rooms were provided in only 20 NICUs (33.3%) and only 33 units (55.0%) routinely provided FICare. Donor human milk and human milk bank Twenty of the hospitals (33.3%) established their own HMB, and the sources of operating funds for the HMB were social funding at 4 hospitals (20%), hospital fund- ing at 15 (75%), and own unit funding at 14 (70%). All the HMBs provided their services for free, for both donors and recipients of DHM. Table 3 Donor human milk and human milk bank Reasons for providing donor human milk NO Reasons for not providing donor human milk NO Situations in which donor milk is used NO reduce NEC 13 Expensive 7 ELBW/VLBW 13 Improve feeding tolerance 11 Patents not receptive 14 Intestinal malabsorption 7 Reduce allergies 9 Milk bank guidelines inadequate 20 feeding intolerance 12 Family request 1 No beneficial 0 immunologic deficiency disease 0 Support mother’ breastfeeding 12 Inadequate growth 1 Malnutrition after surgery 5 Provide exclusively human milk feeding 8 Reduce mother’ milk production 1 Severe infection 1 Reduce nosocomial infection 7 No license 38 Table 3 Donor human milk and human milk bank Hu et al. BMC Pediatrics (2023) 23:75 Page 4 of 7 Page 4 of 7 Hu et al. BMC Pediatrics Hu et al. BMC Pediatrics (2023) 23:75 Fig. 1 Minimum enteral nutrition practices at NICUs. A Infants that require minimal enteral nutrition after birth; B Time at which minimal enteral nutrition was started after birth for infants with body weight < 1500 g; C Duration of minimal enteral nutrition for premature infants At 31 NICUs (51.7%), gastric residual content was examined before every feeding episode (Fig. 3A). At 41 NICUs (68.3%), the course of enteral nutrition man- agement was not changed during treatment for patent ductus arteriosus(PDA)(Fig. 3B). At 29 NICUs (48.3%), enteral feeding was stopped for 1 or 2 feeds during blood transfusion to prevent NEC (Fig. 3C). Discussion Family rooms are known to be beneficial for both infants and parents, as they improve weight gain, promote breastfeeding Most of the NICUs increased the volume of enteral feeding at the rate of 10–20 ml/kg daily (Fig. 2A) and discontinued parenteral nutrition when the enteral vol- ume reached ≥ 120 ml/kg daily (Fig. 2B). HMF was added when enteral volume reached 80–100 ml/kg daily in most units(80%), at the ratio of 1:50 (HMF to human milk) as initial dosage(58.3%) (Fig. 2C and D). Hu et al. BMC Pediatrics (2023) 23:75 Page 5 of 7 Hu et al. BMC Pediatrics Fig. 2 Enteral feeding and human milk fortifier at NICUs. A Rate at which the volume of enteral nutrition was increased in premature infants; (B) Volume of enteral nutrition required to discontinue parenteral nutrition; (C) Volume of enteral nutrition before the use of human milk fortifier was initiated; (D) Ratio of human milk fortifier at the time of initiation after NICU discharge, reduce parental stress and anxi- ety, as well as reduce the incidence of health-assistance related infections [14]. Additionally, FICare in Chinese NICUs was found to be associated with reduced hospital length of stay, medical expenditures, and rates of adverse outcomes [15]. Therefore, in the future, NICUs in China should promote FICare vigorously. g y Since the first HMB was established in Guangzhou in 2003, 20 HMBs have been established in China so far. All these HMBs provide their services free of charge to both donors and recipients of DHM. However, the cost of run- ning HMBs is quite high [16] and is almost entirely borne by the local hospitals, as shown by our findings. This is also one of the reasons for the small number of HMBs in China. Therefore, it is necessary to find alternate sources of funds for HMBs in China. In China, both parents and medical workers’ attitude about DHM is positive [17–19], but in our survey, only 13 units (21.7%) used DHM. The most common reason for not using DHM was the lack of a license. This highlights the need to improve the process of using DHM in China.In the 7 units that established an HMB but did not use DHM, the reason was because the banks were awaiting for their licensing.Therefore, they could only manage and use MOM in the HMB. Discussion y g Our study highlights the diversity of recommended feeding practices in NICUs across China. While most of the units initiated MEN for infants with birth weight < 1500 g (48.3%), the criteria for initiating MEN varied considerably across the remaining units. The opti- mal duration of MEN is still under debate [20, 21], and this was also reflected in the results of our study. With regard to feed volume, we found that most units (83.3%) increased enteral feeding intake by 10 to 20 ml/kg daily. A Cochrane review compared daily feed intake increments of 15 to 20 ml/kg versus 30 to 35 ml/kg and concluded that more rapid advancement did not increase the risk of NEC, mortality, or interruption of feeds [22]. Further, previous studies reported that early fortification (< 40 ml/ kg daily) was safe and well tolerated [23]. In the present study, we found a significant variation in both the recom- mended timing of initiation and the recommended ini- tial volume of HMF. Another international survey-based study also reported large variations in the timing of ini- tiation and volume of HMF at initiation [24]. Given these variations in enteral feeding strategies and the previous findings, national-level guidelines should recommend Hu et al. BMC Pediatrics (2023) 23:75 Page 6 of 7 Hu et al. BMC Pediatrics Fig. 3 Gastric residual content and changes in feeding strategy. A Time point for assessment of gastric residual content; B Changes in enteral feeding strategies during drug treatment for patent ductus arteriosus; C Enteral feeding changes during blood transfusion residuals; nonetheless, feeding management of this population should be carefully evaluated [26]. Our study showed that 41 units (68.3%) did not make any changes in their enteral feeding regimen during drug treatment for PDA. Further, feeding during blood trans- fusions may increase the risk for mesenteric ischemia and the development of transfusion-related NEC in pre- term infants [27]. In the present study, 29 units (48.3%) instated stop feeding for 1 or 2 feeds during blood transfusion, but 15 units (25.0%) reported that did not introduce any changes and only closely monitored the situation.h There are some limitations to this study. As with other surveys, a responder bias could not be ruled out, and the responses may, potentially, not be representative of the practices of the unit. Conclusion Th There were significant differences in human milk feed- ing and enteral feeding strategies between the NICUs in CHNN, but also similarities. Establishing national- level feeding guidelines for preterm and low birth weight infants and quality improvement of cooperation at the national level are needed. This is the first national-level survey on this topic in China, so the data obtained would be useful in the establishment of national enteral feeding guidelines for preterm infants and quality improvement of cooperation at the national level. Acknowledgements h h f ll The authors gratefully acknowledge all site investigators of the China Neonatal Network (CHNN) for their participation in this survey. Additional file 1. Fig. 3 Gastric residual content and changes in feeding strategy. A Time point for assessment of gastric residual content; B Changes in enteral feeding strategies during drug treatment for patent ductus arteriosus; C Enteral feeding changes during blood transfusion Authors’ contributions standardized enteral feeding and human milk fortifica- tion strategies for NICUs in China. Xiaoshan Hu and Junjie Lu conceptualized and designed the study, drafted the initial manuscript,and approved the final manuscript as submitted. Shuping Han and Xiaohui Chen conceptualized and designed the study, conducted data analysis, drafted the initial manuscript. Jun Zhang, Min Zhang and Zhangbin Yu participated in data collection and sorting, and reviewed and revised the manuscript. Shoo K. Lee prepared the data and initial analysis, critically reviewed the manuscript and approved the final manuscript as submitted. All the authors have read and approved the manuscript. Routine assessment of gastric residual content was reported by 31 units (51.7%).A previous study indi- cated that the omission of gastric residual evaluation increased the delivery of enteral nutrition as well as improved weight gain and led to earlier hospital dis- charge [25]. This suggests that the NICUs in China should probably change the previous practice.A previ- ous study reported that feeding during pharmacologi- cal PDA closure in preterm neonates was not associated with delay to reach full feeds, NEC incidence, or gastric Discussion We also did not evaluate whether the variation in feeding practices are associated with infant growth and outcomes such as NEC. Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12887-023-03862-0. References 1. Reena OF, Rashmi K, James D. A quality improvement project to decrease human milk errors in the NICU. Pediatrics. 2017;139(2):e20154451. 1. Reena OF, Rashmi K, James D. A quality improvement project to decrease human milk errors in the NICU. Pediatrics. 2017;139(2):e20154451. 1. Reena OF, Rashmi K, James D. A quality improvement project to decrease human milk errors in the NICU. Pediatrics. 2017;139(2):e20154451. 23. Beth G, Carmel TC, Cathie H. Does early compared to late fortification of human milk for preterm infants improve clinical outcomes? J Paediatr Child Health. 2019;55(7):867–72. 2. Tricia JJ, Aloka LP, Harold RB. Economic benefits and costs of human milk feedings: a strategy to reduce the risk of prematurity-related morbidities in very-low-birth-weight infants. Adv Nutr. 2014;5(2):207–12. 24 Klingenberg C, Embleton ND, Jacobs SE. Enteral feeding practices in very preterm infants: An international survey. Arch Dis Child Fetal Neonatal Ed. 2012;97:F56-61. 24 Klingenberg C, Embleton ND, Jacobs SE. Enteral feeding practices in very preterm infants: An international survey. Arch Dis Child Fetal Neonatal Ed. 2012;97:F56-61. y g 3. Corpeleijn WE, Kouwenhoven SM, Paap MC. Intake of own mother’s milk during the first days of life is associated with decreased morbidity and mortality in very low birth weight infants during the first 60 days of life. Neonatology. 2012;102(4):276–81. 25. Leslie AP, Michael W, Roberto MT. Effect of Gastric Residual Evaluation on Enteral Intake in Extremely Preterm Infants: A Randomized Clinical Trial. JAMA Pediatr. 2019;173(6):534–43. 25. Leslie AP, Michael W, Roberto MT. Effect of Gastric Residual Evaluation on Enteral Intake in Extremely Preterm Infants: A Randomized Clinical Trial. JAMA Pediatr. 2019;173(6):534–43. gy 4. Capriati T, Goffredo BM, Argentieri M, De Vivo L, Bernaschi P, Cairoli S, Laureti F, Reposi MP, Marino D, Benedetti S, Diamanti A. A Modified Holder Pasteurization Method for Donor Human Milk: Preliminary Data. Nutrients. 2019;11(5):1139. 26. Martini S, Aceti A, Galletti S, Beghetti I, Faldella G, Corvaglia L. To Feed or Not to Feed: A Critical Overview of Enteral Feeding Management and Gastrointestinal Complications in Preterm Neonates with a Patent Ductus Arteriosus. Nutrients. 2019;12(1):83. 26. Martini S, Aceti A, Galletti S, Beghetti I, Faldella G, Corvaglia L. To Feed or Not to Feed: A Critical Overview of Enteral Feeding Management and Gastrointestinal Complications in Preterm Neonates with a Patent Ductus Arteriosus. Nutrients. 2019;12(1):83. 5. Committee on Nutrition, Section on Breastfeeding, Committee on Fetus and Newborn. Consent for publication 19. Rui Y, Danqi C, Qingqi D. The effect of donor human milk on the length of hospital stay in very low birthweight infants: a systematic review and meta-analysis. Int Breastfeed J. 2020;15(1):89. Funding h d This study was supported by the Nanjing Medical Science and Technol- ogy Development Foundation (grant number ZKX19045); the "Specialized Diseases Cohort" Program of Nanjing Medical University (grant number NMUC2020037); and the Six Talent Peaks Project in Jiangsu Province (grant number LGY2019008). Page 7 of 7 Page 7 of 7 Hu et al. BMC Pediatrics (2023) 23:75 Hu et al. BMC Pediatrics References Donor human milk for the high-risk infant: Preparation, safety, and usage options in the United States. Pediatrics. 2017;139(1):e20163440. 27. Terri M, Cassandra DJ, Niki K. Feeding preterm infants during red blood cell transfusion is associated with a decline in postprandial mesenteric oxygenation. J Pediatr. 2014;165(3):464–71. 6. Agostoni C, Buonocore G, Carnielli VP. Enteral nutrient supply for preterm infants: Commentary from the European Society of Paediatric Gastroen- terology, Hepatology and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr. 2010;50(1):85–91. Publisher’s Note 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. 7. Editorial Board of Chinese Journal of Pediatrics, Sub specialty Groups of Neonatology and Child Health Care, & The Society of Pediatrics Chinese Medical Association. Feeding recommendations for preterm infants and low birth weight infants. Zhonghua Er Ke Za Zhi. 2009;47(7):508–10. 8. Yuanyuan Y, Hong L. Breastfeeding in hospitalised preterm infants: A sur- vey from 18 tertiary neonatal intensive care units across mainland China. J Paediatr Child Health. 2020;56(9):1432–7. 9. Henry CL, Paul SK, Nancy EW. A quality improvement project to increase breast milk use in very low birth weight infants. Pediatrics. 2012;130(6):e1679–87. • 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: 10. Giannì ML, Roggero P, Garbarino F, Bracco B, Fumagalli M, Agosti M, Mosca F. Nutrition and growth in infants born preterm from birth to adulthood. Early Hum Dev. 2013;89(Suppl 2):S41–4. 10. Giannì ML, Roggero P, Garbarino F, Bracco B, Fumagalli M, Agosti M, Mosca F. Nutrition and growth in infants born preterm from birth to adulthood. Early Hum Dev. 2013;89(Suppl 2):S41–4. 11. De Rose DU, Cota F, Gallini F, Bottoni A, Fabrizio GC, Ricci D, Romeo DM, Mercuri E, Vento G, Maggio L. Extra-uterine growth restriction in preterm infants: Neurodevelopmental outcomes according to different defini- tions. Eur J Paediatr Neurol. 2021;33:135–45. 11. Availability of data and materials 13. Bo Z, Zhiying D, Yingxi Z. Intermittent kangaroo mother care and the practice of breastfeeding late preterm infants: results from four hospitals in different provinces of China. Int Breastfeed J. 2020;15(1):64. The datasets used and analysed during the current study are available from the corresponding author upon request. 14. O’Brien K, Robson K, Bracht M. Effectiveness of Family Integrated Care in neonatal intensive care units on infant and parent outcomes: a mul- ticentre, multinational, cluster-randomised controlled trial. Lancet Child Adolesc Health. 2018;2(4):245–54. Received: 8 September 2022 Accepted: 24 January 2023 Received: 8 September 2022 Accepted: 24 January 2023 21. Ozlem B, Evrim AD, Duygu B. Prolonged minimal enteral nutrition versus early feeding advancements in preterm infants with birth weight ≤1250 g: a prospective randomized trial. J Matern Fetal Neonatal Med. 2022;35(2):341–7. 22. Oddie SJ, Young L, McGuire W. Slow advancement of enteral feed vol- umes to prevent necrotising enterocolitis in very low birth weight infants. Cochrane Database Syst Rev. 2017;8(8):CD001241. Competing interests The authors declare no competing interests. 20. Jessie M, Lauren Y, William M. Delayed introduction of progressive enteral feeds to prevent necrotising enterocolitis in very low birth weight infants. Cochrane Database Syst Rev. 2014;12:CD001970. 20. Jessie M, Lauren Y, William M. Delayed introduction of progressive enteral feeds to prevent necrotising enterocolitis in very low birth weight infants. Cochrane Database Syst Rev. 2014;12:CD001970. Ethics approval and consent to participate 15. Mingyan H, Xiangyu G, Ying L. Family Integrated Care for Preterm Infants in China: A Cluster Randomized Controlled Trial. J Pediatr. 2021;228:36–43 The ethical approval was waived by Medical Ethics Committee of Nanjing Maternity and Child Health Hospital as the study did not involve access to private patient data or alterations of feeding practices.Every participant was inquired the willingness to join this survey and informed consent was obtained. All methods were carried out in accordance with relevant guidelines and regulations. 16. Daili C, Kunkun Z, Guangjun Y. Cost Analysis of Operating a Human Milk Bank in China. J Hum Lact. 2020;36(2):264–72. 17. Ce T, Yamin L, Lee S. Lactating Women’s Knowledge and Attitudes About Donor Human Milk in China. J Hum Lact. 2021;37(1):52–61. 18. Na Z, Junyan L, Xinwen L. Factors associated with postpartum women’s knowledge, attitude and practice regarding human milk banks and milk donation: a cross-sectional survey. Midwifery. 2020;91: 102837. Publisher’s Note De Rose DU, Cota F, Gallini F, Bottoni A, Fabrizio GC, Ricci D, Romeo DM, Mercuri E, Vento G, Maggio L. Extra-uterine growth restriction in preterm infants: Neurodevelopmental outcomes according to different defini- tions. Eur J Paediatr Neurol. 2021;33:135–45. 12. Yao Z, Qingqi D, Binghua Z. Neonatal intensive care nurses’ knowledge and beliefs regarding kangaroo care in China: a national survey. BMJ Open. 2018;8(8):e021740. 12. Yao Z, Qingqi D, Binghua Z. Neonatal intensive care nurses’ knowledge and beliefs regarding kangaroo care in China: a national survey. BMJ Open. 2018;8(8):e021740.
https://openalex.org/W2126637736	https://cris.maastrichtuniversity.nl/ws/files/52823790/e12222.full.pdf	English	null	Deficient recovery response and adaptive feedback potential in dynamic gait stability in unilateral peripheral vestibular disorder patients	Physiological reports	2,014	cc-by	9,404	Document license: CC BY Document license: CC BY Document license: CC BY Citation for published version (APA): Citation for published version (APA): Citation for published version (APA): cCrum, C., Eysel-Gosepath, K., Epro, G., Meijer, K., Savelberg, H. H., Bruggemann, G. P., & aramanidis K (2014) Deficient recovery response and adaptive feedback potential in dynamic McCrum, C., Eysel-Gosepath, K., Epro, G., Meijer, K., Savelberg, H. H., Bruggemann, G. P., & Karamanidis, K. (2014). Deficient recovery response and adaptive feedback potential in dynamic gait stability in unilateral peripheral vestibular disorder patients. Physiological Reports, 2(12), Article e12222. https://doi.org/10.14814/phy2.12222 McCrum, C., Eysel-Gosepath, K., Epro, G., Meijer, K., Savelberg, H. H., Bruggemann, G. P., & Karamanidis, K. (2014). Deficient recovery response and adaptive feedback potential in dynamic gait stability in unilateral peripheral vestibular disorder patients. Physiological Reports, 2(12), Article e12222. https://doi.org/10.14814/phy2.12222 DOI: 10.14814/phy2.12222 DOI: 10.14814/phy2.12222 Document Version: Publisher's PDF, also known as Version of record Document status and date: Published: 01/01/2014 Document status and date: Published: 01/01/2014 Please check the document version of this publication: • A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. p • The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page numbers. Link to publication General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyrigh owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain Y f l di t ib t th URL id tif i th bli ti i th bli t l y y p g y You may freely distribute the URL identifying the publication in the public portal. If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement: www.umlib.nl/taverne-license Take down policy If b li th t th Take down policy If you believe that this document breaches copyright please contact us at: repository@maastrichtuniversity.nl providing details and we will investigate your claim. Download date: 24 Oct. 2024 Download date: 24 Oct. 2024 Funding information Funding information Financial support from the Forschungsservicestelle, German Sport University Cologne (Hochschulinterne Forschungsf€orderung) is greatly appreciated. Received: 29 October 2014; Accepted: 3 November 2014 doi: 10.14814/phy2.12222 Received: 29 October 2014; Accepted: 3 November 2014 doi: 10.14814/phy2.12222 Physiol Rep, 2(12), 2014, e12222, doi: 10.14814/phy2.12222 trips and falls while walking, effective postural adjust- ments and locomotor adaptations are required. Motor adaptations to environmental changes can be immediate reactive responses that rely on ongoing afferent feedback information, or feedforward predictive adjustments (Marigold and Patla 2002; Pai et al. 2003; Pavol et al. 2004; Bhatt et al. 2006; Lam et al. 2006). It is suggested that central pattern generators may be responsible for ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. 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. Physiological Reports ISSN 2051-817X Physiological Reports ISSN 2051-817X Deﬁcient recovery response and adaptive feedback potential in dynamic gait stability in unilateral peripheral vestibular disorder patients Deﬁcient recovery response and adaptive feedback potential in dynamic gait stability in unilateral peripheral vestibular disorder patients Christopher McCrum1, Katrin Eysel-Gosepath2, Gaspar Epro3,4, Kenneth Meijer1, Hans H. C. M. Savelberg1, Gert-Peter Br€uggemann3,5 & Kiros Karamanidis3,4,6 1 Human Movement Science, NUTRIM, School for Nutrition, Toxicology and Metabolism, Maastricht University, Maastricht, The Netherlands 2 Department of Otolaryngology, Head and Neck Surgery, Heinrich Heine University of D€usseldorf, D€usseldorf, Germany 3 Institute of Biomechanics and Orthopaedics, German Sport University Cologne, Cologne, Germany 4 Institute of Movement and Sport Gerontology, German Sport University Cologne, Cologne, Germany 5 Cologne Center for Musculoskeletal Biomechanics, Medical Faculty, University of Cologne, Cologne, Germany 6 Department of Mathematics and Technology, University of Applied Sciences Koblenz, RheinAhrCampus Remagen, Remagen, Germany Keywords Fall risk, gait adaptation, margin of stability, vestibular dysfunction. Unilateral peripheral vestibular disorder (UPVD) causes deﬁcient locomotor responses to novel environments due to a lack of accurate vestibular sensory information, increasing fall risk. This study aimed to examine recovery response (stability recovery actions) and adaptive feedback potential in dynamic stability of UPVD-patients and healthy control subjects during per- turbed walking. 17 UPVD-patients (>6 months since onset) and 17 matched healthy control participants walked on a treadmill and were subjected to eight unexpected perturbations during the swing phase of the right leg. For each perturbation, the margin of stability (MS; state of body’s centre of mass in relation to the base of support), was determined at touchdown of the per- turbed leg and during the following six recovery steps. The ﬁrst perturbation caused a reduced MS at touchdown for the perturbed leg compared to base- line, indicating an unstable position, with controls requiring ﬁve recovery steps to return to MS baseline and UPVD-patients not returning to baseline level within the analyzed six recovery steps. By the eighth perturbation, con- trol subjects needed two steps, and UPVD-patients required three recovery steps, both thereby improving their recovery response with practice. However, MS at touchdown of the perturbed leg increased only for the controls after repeated perturbations, indicating adaptive feedback-driven locomotor improvements for the controls, but not for the UPVD-patients. We concluded that UPVD-patients have a diminished ability to control dynamic gait stability during unexpected perturbations, increasing their fall risk, and that vestibular dysfunction may inhibit the neuromotor system adapting the reactive motor response to perturbations. Correspondence Kiros Karamanidis, Institute of Movement and Sport Gerontology, German Sport University Cologne, Am Sportpark M€ungersdorf 6, 50933 Cologne, Germany. Tel:+49-(0)221-4982-6144 Fax: +49-(0)221-4982-6143 E-mail: karamanidis@dshs-koeln.de ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. 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. 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 1 ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of Introduction A number of recent studies have demonstrated an improvement in the effectiveness of such feedforward adjustments (i.e., adjustments made in preparation for a given perturbation based on prior knowledge and experi- ence of the constraint) and/or reactive responses (i.e., adjustments made as a direct response following a pertur- bation) after repeated practice of different task constraints (Marigold and Patla 2002; Pai et al. 2003; Pavol et al. 2004; Bhatt et al. 2006; Bierbaum et al. 2010, 2011). For example, Bierbaum et al. (2010, 2011) reported favorable adaptations in feedforward adjustments and reactive responses in dynamic gait stability after repeated exposure to changes in surface compliance while walking in older and younger adults. The adaptations were observed as increases in margin of stability indicating a reduced fall risk (Bierbaum et al. 2010, 2011). The margin of stability (Hof et al. 2005) quantiﬁes the stability of the body con- ﬁguration using the difference between the base of sup- port and the extrapolated center of mass (calculated using the position and velocity of the center of mass). Negative margin of stability values indicate an unstable body con- ﬁguration, whereas positive margin of stability values indicate a stable body conﬁguration (i.e., no further motor actions are needed to preserve postural stability). The degree to which the reactive response can improve, known as adaptive feedback potential, describes the reac- tive, feedback-driven adaptations to unexpected perturba- tions that occur over time (Pavol et al. 2004; Bierbaum et al. 2011). Bierbaum et al. (2011) analyzed margin of stability at touchdown of the ﬁrst recovery step which was expected to reveal a predominantly reactive-driven motor response, due to the sudden and unexpected nat- ure of the perturbation. As knowledge and experience gathered over time for speciﬁc task constraints can posi- tively inﬂuence recovery performance, the integration of accurate sensory feedback information during the pertur- bations may be signiﬁcant for the success of such move- ment corrections and adaptations during locomotion. Therefore, this study aimed to examine the recovery responses and adaptive feedback potential in dynamic stability of UPVD patients and matched healthy control subjects during perturbed treadmill walking, by repeat- edly applying unilateral resistance unexpectedly to the swing phase of the right leg. In order to predominantly examine the reactive response of the participants, we determined the recovery stepping behavior of the partici- pants at touchdown of the perturbed leg. Introduction Human biped locomotion is a mechanically intricate motor task due to the need to produce effective and safe gait patterns during daily life. Such complexity is required to cope with changing environmental demands such as steps, slopes, or uneven terrain. To safely negotiate such environmental conditions, maintain balance, and avoid 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 1 Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. showing greater variability during artiﬁcial vestibular perturbations (Kennedy et al. 2003). Therefore, deﬁcien- cies in stability control during gait could be expected if the vestibular system is dysfunctional. spinal structures controlling basic gait patterns during reactive responses, with the cerebellum controlling feed- forward, predictive adjustments (Morton and Bastian 2006). A speciﬁc group that experiences vestibular dysfunction is unilateral peripheral vestibular disorder (UPVD) patients. UPVD includes disorders of the inner ear on one side such as vestibular neuritis (Hillier and McDon- nell 2011). UPVD is associated with imbalance and falls (Neuhauser et al. 2005; Homann et al. 2013) and is the leading cause of dizziness (L€uscher et al. 2014). It has been shown that UPVD patients struggle to maintain consistent arm actions during trunk movement without visual information (Raptis et al. 2007) and demonstrate deﬁciencies during goal-directed reaching tasks while both seated and standing (Borello-France et al. 1999). As upper body motor tasks during standing and sitting are nega- tively affected by UPVD, dynamic stability control during perturbed walking may also be disrupted by UPVD. Con- sequently, adaptations of such motor behaviors over time may also be diminished. Some level of motor response and adaptation may be possible through spinal structures driving reactive responses, as has been reported in human infants (Lam et al. 2003; Pang et al. 2003). There is evi- dence to suggest that vestibular and somatosensory infor- mation interact at the spinal level during reactive postural control (Horak et al. 2001), but it has not, to our knowl- edge, been investigated if recovery performance and reac- tive locomotor responses during perturbed walking are affected by UPVD. ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Subjects Gait perturbation was accomplished by using an unexpected (subjects were not warned) application of resistance using a custom built electronically driven magnet system during the swing phase of the right leg. For each perturbation block, dynamic stability parameters were examined at touchdown (TD) of the contralateral step of the left leg prior to perturbation (preL), TD of the perturbed step (pertbR) and TD of the six recovery steps following each perturbation (post1L–post6R). Washout periods between blocks (typically 1.5–2 min) were given, so that postural adjustments made due to the applied resistance dissipated prior to the next block. Figure 1. Experimental protocol of the gait perturbation task: The protocol began with a baseline period (nonperturbed walking), followed by eight unexpected perturbation blocks separated by washout periods. Baseline was deﬁned by averaging 12 consecutive steps of nonperturbed walking. Gait perturbation was accomplished by using an unexpected (subjects were not warned) application of resistance using a custom built electronically driven magnet system during the swing phase of the right leg. For each perturbation block, dynamic stability parameters were examined at touchdown (TD) of the contralateral step of the left leg prior to perturbation (preL), TD of the perturbed step (pertbR) and TD of the six recovery steps following each perturbation (post1L–post6R). Washout periods between blocks (typically 1.5–2 min) were given, so that postural adjustments made due to the applied resistance dissipated prior to the next block. Figure 1. Experimental protocol of the gait perturbation task: The protocol began with a baseline period (nonperturbed walking), followed by eight unexpected perturbation blocks separated by washout periods. Baseline was deﬁned by averaging 12 consecutive steps of nonperturbed walking. Gait perturbation was accomplished by using an unexpected (subjects were not warned) application of resistance using a custom built electronically driven magnet system during the swing phase of the right leg. For each perturbation block, dynamic stability parameters were examined at touchdown (TD) of the contralateral step of the left leg prior to perturbation (preL), TD of the perturbed step (pertbR) and TD of the six recovery steps following each perturbation (post1L–post6R). Washout periods between blocks (typically 1.5–2 min) were given, so that postural adjustments made due to the applied resistance dissipated prior to the next block. Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. role of the vestibular system in dynamic stability control during locomotion. prior to data collection. To achieve natural movement patterns, all participants were asked to wear their own regular sports shoes. Subjects always wore a safety harness connected to an overhead track. No instructions were given to the subjects regarding gaze ﬁxation. Subjects Seventeen adult patients diagnosed with UPVD recruited from a medical clinic (10 females and 7 males; age: 49 years (SD9); body height: 171.4 cm (SD7.3); body mass: 73.8 kg (SD14.1)) and 17 matched healthy adults [10 females and 7 males; 51 years (SD8); body height: 172.5 cm (SD8.2); body mass: 75.1 kg (SD15.2)] partici- pated in this study. Healthy adults were selected as con- trol subjects based on gender, age, anthropometric measures, and physical activity level (frequency of partici- pation in physical activity per week, determined by ques- tionnaire), to create matched pairs with the UPVD patients. Patients with diagnosed vestibular neuritis, that were at least 6 months since onset, were recruited. The patients were experiencing rotational vertigo and eight suffered from feelings of instability and unsteadiness. Four patients had fallen in the previous 6 months. Patients were assessed for inclusion and had their diagno- ses conﬁrmed by a specialist otolaryngologist during a clinical examination that included examining for sponta- neous and induced vestibular nystagmus, bithermal calo- ric tests under videostagmography, head impulse tests, rotating chair tests, and the examination of balance and coordination using Romberg and Unterberger tests. Fif- teen patients had a right side deﬁcit and two patients had a deﬁcit on the left side. The subjects of the healthy con- trol group were also medically screened by the same oto- laryngologist using identical tests to conﬁrm that they did not have UPVD. Patients were excluded if their time since UPVD onset was less than 6 months or if their diagnosis was not conﬁrmed. Further inclusion criteria for the patients and the control subjects were that they did not participate in sport or physical exercise more than once per week and had no other health problems, including cardiovascular disease, musculoskeletal injuries, or any other locomotor dysfunction. The study was approved by the university’s ethical board, the procedures of the study were explained to the participants, and written informed consent was obtained prior to the testing in accordance with the Declaration of Helsinki. Figure 1. Experimental protocol of the gait perturbation task: The protocol began with a baseline period (nonperturbed walking), followed by eight unexpected perturbation blocks separated by washout periods. Baseline was deﬁned by averaging 12 consecutive steps of nonperturbed walking. ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Materials and Methods The exact gait perturbation task and protocol has been described previously by S€uptitz et al. (2013). Brieﬂy, on the day of the measurements, the gait protocol started with ﬁve minutes of walking at 1.4 ms1 to allow the subjects to get accustomed once more to the treadmill and gait velocity. Following that, a Teﬂon rope was con- nected with Velcro straps to the subjects’ right leg above the ankle joint. The rope was connected to a custom built device which was used to apply and remove unexpectedly a unilateral resistance of 2.1 kg during the swing phase of the right leg using an electronically driven magnet system Introduction We deﬁned the reactive response using the above described method as the perturbation was unexpected with no prior warn- ing given, and the duration between onset of perturba- tion and touchdown was short (also see S€uptitz et al. 2013). It was hypothesized that recovery stepping behav- ior would be less effective and margin of stability at touchdown of the perturbed leg would demonstrate less improvement with practice in UPVD patients compared to matched healthy controls after repeated exposure to the perturbation, indicating a diminished adaptive feed- back potential in dynamic stability during perturbed walking in UPVD patients. Such ﬁndings could be rele- vant for fall risk reduction interventions in UPVD patients and could enhance our understanding of the Sensory feedback is important for successful and safe locomotion, as it informs the central nervous system about the actual state of the musculoskeletal system and the environmental conditions (Wolpert et al. 1995; Sainburg et al. 1999; Rossignol et al. 2006). In particular, the vestib- ular system plays an important role in gait, providing information regarding the position, velocity and direction of the head in space (Kennedy et al. 2003; Rossignol et al. 2006) and contributing to lower limb control (Bent et al. 2005). Additionally, gait trajectory is reported to be nega- tively affected when vestibular sensory input is disturbed, 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 2 Adaptive Adjustments in Dynamic Gait Stability in UPVD ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Experimental setup The washout was controlled for each individual using a live observation of the antero- posterior displacement of the toe markers while walking, using a motion capture system (Vicon Motion Systems, Oxford, UK), to determine the step length compared to baseline values (nonperturbed gait). In cases where clear differences (more than 10%) of the anteroposterior dis- placement of the toe markers compared with baseline existed longer than the typical 1.5–2 min washout period, the washout period was extended as long as was necessary for the values to return to baseline. However, this was not required for any subjects, with most returning to baseline within 1 min. Subjects were never warned about the application or removal of the resistance. The margin of stability in the anteroposterior direction during walking was determined using the extrapolated centre of mass concept (Hof et al. 2005; see also Fig. 2). Margin of stability was calculated as the difference between the anterior boundary of the base of support (horizontal component of the projection of the toe from the corresponding limb to the ground) and the extrapo- L g PCoM VCoM BSUmax CoM XCoM MS (positive values) L g PCoM MS (negative values) XCoM VCoM CoM BSUmax A B A B Figure 2. Schematic illustration of the inverted pendulum model during locomotion (Hof et al. 2005). PCoM represents the horizontal (anteroposterior) component of the projection of the center of mass (CoM) to the ground, VCoM represents the horizontal velocity of the CoM (anteroposterior), g is acceleration due to gravity and L is the pendulum length (i.e., distance between the CoM and the centre of the ankle joint in the sagittal plane). Margin of stability (MS) in the anterior direction is the instantaneous difference between the anterior boundary of the base of support (BSUmax) and the extrapolated center of mass (XCoM). Positive MS values (A) indicate a stable body conﬁguration (during unperturbed walking this would mean that additional motor actions would be required to continue walking, such as leading leg hip extensor and/or trailing leg ankle plantarﬂexor action), whereas negative margin of stability values (B) indicate an unstable body conﬁguration (i.e., subjects must make additional motor actions to preserve stability and to avoid a fall, e.g., by stepping). Experimental setup The washout was controlled for each individual using a live observation of the antero- posterior displacement of the toe markers while walking, using a motion capture system (Vicon Motion Systems, Oxford, UK), to determine the step length compared to baseline values (nonperturbed gait). In cases where clear differences (more than 10%) of the anteroposterior dis- placement of the toe markers compared with baseline existed longer than the typical 1.5–2 min washout period, the washout period was extended as long as was necessary for the values to return to baseline. However, this was not required for any subjects, with most returning to baseline within 1 min. Subjects were never warned about the application or removal of the resistance. (see S€uptitz et al. 2013). Participants then walked at 1.4 ms1for 3–4 min without perturbation and a baseline period (nonperturbed walking) of about 20 sec of walking was recorded at the end of this period (Fig. 1). Following the baseline period, the resistance was turned on for the entire duration of the swing phase of the right leg for one step and was subsequently turned off (S€uptitz et al. 2013). This one-time application of resistance was unex- pected for all participants. The instant of foot touchdown at the contralateral step of the left leg prior to perturba- tion (preL), the perturbed step (pertbR) and the six recov- ery steps following each perturbation (post1L–post6R), collectively deﬁned as unexpected perturbation block one (block1), were of interest for the analysis of the dynamic stability parameters (see Fig. 1; exact parameters are described below). These speciﬁc gait events were analyzed in order to determine if any alterations in gait occurred preperturbation due to preparation behavior or anxiety (preL) and to examine the response behavior of the par- ticipants during (pertbR) and postperturbation (post1L– post6R). The above described unexpected perturbation application and analysis procedure was repeated for a total of eight unexpected perturbation blocks (block1 to block8; see Fig. 1) within a period of approximately 25 min of walking. Between perturbation blocks sufﬁcient time (typically 1.5–2 min) was given as a “washout” per- iod, so that postural adjustments made due to the applied resistance dissipated (Fig. 1). Experimental setup All subjects walked on a motor-driven treadmill (pulsar 4.0, h/p/cosmos; Nussdorf-Traunstein, Germany) with a belt speed of 1.4 ms1. Familiarization with the treadmill was carried out at least once for each subject 4–7 days 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 3 Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. the markers were recorded by the Vicon Nexus motion capture system. The motion capture system was com- prised of eight infrared cameras operating at 120 Hz and the 3D coordinates of the markers were smoothed using a Woltring ﬁlter routine with a mean squared error of ﬁve (Woltring 1986). Segmental masses and their location were calculated based on the data reported by Dempster et al. (1959). (see S€uptitz et al. 2013). Participants then walked at 1.4 ms1for 3–4 min without perturbation and a baseline period (nonperturbed walking) of about 20 sec of walking was recorded at the end of this period (Fig. 1). Following the baseline period, the resistance was turned on for the entire duration of the swing phase of the right leg for one step and was subsequently turned off (S€uptitz et al. 2013). This one-time application of resistance was unex- pected for all participants. The instant of foot touchdown at the contralateral step of the left leg prior to perturba- tion (preL), the perturbed step (pertbR) and the six recov- ery steps following each perturbation (post1L–post6R), collectively deﬁned as unexpected perturbation block one (block1), were of interest for the analysis of the dynamic stability parameters (see Fig. 1; exact parameters are described below). These speciﬁc gait events were analyzed in order to determine if any alterations in gait occurred preperturbation due to preparation behavior or anxiety (preL) and to examine the response behavior of the par- ticipants during (pertbR) and postperturbation (post1L– post6R). The above described unexpected perturbation application and analysis procedure was repeated for a total of eight unexpected perturbation blocks (block1 to block8; see Fig. 1) within a period of approximately 25 min of walking. Between perturbation blocks sufﬁcient time (typically 1.5–2 min) was given as a “washout” per- iod, so that postural adjustments made due to the applied resistance dissipated (Fig. 1). Statistics A three-way repeated measures analysis of variance (ANOVA), with subject group (UPVD patients and healthy controls), gait event (base, pertbR, post1L–post6R; dependent variable) and perturbation block (block1 and block8; dependent variable) as factors was used to deter- mine differences in the margin of stability, extrapolated center of mass and base of support and, hence, to exam- ine the recovery response of the participants before and after repeated practice of the unexpected perturbation task. For the analysis of the adaptive feedback potential of the UPVD patients and the controls, the perturbed right leg (pertbR) of all eight examined perturbation blocks (block1 to block8) were considered. A two-way repeated measures ANOVA, with subject group (UPVD patients and healthy controls) and perturbation block (block1 to block8) as factors was used in order to examine and iden- tify any subject group or perturbation block related differ- ences in margin of stability and postural corrections during the course of the perturbations. Possible locomo- tor adjustments of the participants prior to each pertur- bation were checked by examining the margin of stability, extrapolated center of mass and base of support for each step preperturbation (preL in block1 to block8) in relation to baseline by using a two-way ANOVA with subject group and perturbation block as factors. For each signiﬁ- cant result, we applied simple contrasts to further investi- gate whether the outcome measures at certain gait events (preL, pertbR, post1L–post6R) differed from baseline or whether differences between subject groups or perturba- tion blocks existed. The adaptation magnitude of the UPVD patients and healthy controls was checked for dif- ferences using an independent samples t-test. The level of signiﬁcance for all tests was set at a = 0.05. Before apply- ing the statistical analyses, the distribution normality of our results for each variable was checked using the Kol- mogorov–Smirnov Test, which revealed normal distribu- tions (P values > 0.05). Statistical analyses were carried out with STATISTICA 7.1 (StatSoft Inc., Tulsa, OK, USA). All results are presented as mean and standard deviation (mean and SD). Statistics XCoM ¼ PCoM þ VCoM þ VBSj j ð Þ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ g L1 p where PCoM represents the horizontal (anteroposterior) component of the projection of the CoM to the ground, VCoM represents the horizontal velocity of the CoM (ante- roposterior), VBS represents the horizontal (anteroposteri- or) velocity of the anterior boundary of the base of support (calculated by the average velocity of the forefoot markers during the stance phase, approximately equal to the speed of the treadmill belt), g is acceleration due to gravity, and L is the distance between the CoM and the centre of the ankle joint in the sagittal plane (Fig. 2). The base of support was deﬁned as the distance between the anterior boundaries of the leading and trailing foot at touchdown, using the distance between the vertical projec- tions of the respective toe markers and the extrapolated center of mass at touchdown was calculated with reference to the anterior boundary of the base of support (leading leg; Fig. 2). The touchdown was determined using the acceleration of the tibia, measured by two-dimensional accelerometers (50 g; ADXL250; Analog Devices, Nor- wood, MA, USA) attached with tape to the midpoint of the tibias (S€uptitz et al. 2012). Baseline values for margin of stability, extrapolated center of mass and base of sup- port were calculated by averaging 12 consecutive steps (left and right; base, see Fig. 1) of nonperturbed walking for each subject. For the analysis of the locomotor response to the perturbation and the postural corrections before and after repeated practice of the perturbation task, the per- turbed right leg and at the six recovery steps of the ﬁrst and ﬁnal perturbation blocks (perturbation block1 and perturbation block8) were considered in the statistics. The extent of the adaptation of the reactive response due to the adaptive feedback potential of the participants was calculated using margin of stability at touchdown as follows: Adaptation Magnitude ¼ 1 MSBlock8 MSBase MSBlock1 MSBase 100 where MSBlock1 and MSBlock8 are margin of stability at touchdown of the perturbed leg of the ﬁrst and eighth blocks (pertbR in block1 and pertbR in block8), respec- tively, with margin of stability baseline represented by MSBase. Four patients with UPVD were unable to cope with the gait task, and were only able to prevent a fall after the sudden perturbation by grasping the treadmill handrails. Statistics Those four patients and their matched healthy control subjects were excluded from the analysis and, hence, only 13 patients and 13 control subjects were included in the statistics. Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. lated center of mass (Fig. 2). Extrapolated center of mass (XCoM) was deﬁned as follows: ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Dynamic stability analysis during walking The method used to analyze dynamic stability during treadmill walking has been described in detail in two pre- vious studies (S€uptitz et al. 2012, 2013). Brieﬂy, to track a twelve-segment, full body kinematic model (left and right foot, left and right lower leg, left and right thigh, trunk, left and right lower arms, left and right upper arms, head) and to examine gait kinematics, 26 retro reﬂective markers (radius 16 mm) were attached to ana- tomical landmarks on the skin and the 3D coordinates of 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 4 Adaptive Adjustments in Dynamic Gait Stability in UPVD Adaptive Adjustments in Dynamic Gait Stability in UPVD Recovery locomotor behavior before and after experience of eight perturbations The analysis of dynamic stability control during walking (nonperturbed gait and the ﬁrst and last unexpected perturbation blocks) revealed a signiﬁcant (P < 0.05) per- turbation block 9 gait event 9 subject group interaction for base of support and margin of stability, meaning the effect of subject group on dynamic stability control was Figure 4. Base of support and extrapolated centre of mass at touchdown (TD) during nonperturbed walking (base: average values of 12 consecutive steps), at TD of the perturbed leg (pertbR), and at TD of the following six consecutive steps to an unexpected perturbation (post1L–post6R) for unilateral peripheral vestibular disorder (UPVD; n = 13) patients and matched healthy controls (CONT; n = 13) during the ﬁrst (block1) and the ﬁnal (block8) unexpected perturbation during treadmill walking (set speed of 1.4 ms1; mean and SD). In order to maintain a stable body conﬁguration (i.e., positive margin of stability) the base of support must exceed the extrapolated center of mass. Please note that for the base of support there were tendencies (P < 0.1) for differences from baseline for the UPVD patients in block1 for post2R and post5L. ba: Statistically signiﬁcant difference to base for both subjects groups (P < 0.05). baC: Statistically signiﬁcant difference to base for the control group (P < 0.05). baP: Statistically signiﬁcant difference to base for the UPVD group (P < 0.05). blC: Statistically signiﬁcant difference to block1 for the control group (P < 0.05). blP: Statistically signiﬁcant difference to block1 for the UPVD group (P < 0.05). Statistically signiﬁcant difference between the UPVD and control groups (P < 0.05). Figure 3. Margin of stability at touchdown (TD) during nonperturbed walking (base: average values of 12 consecutive steps), at TD of the perturbed leg (pertbR), and at TD of the following six consecutive steps to an unexpected perturbation (post1L–post6R) for unilateral peripheral vestibular disorder (UPVD; n = 13) patients and matched healthy controls (CONT; n = 13) during the ﬁrst (block1) and the ﬁnal (block8) unexpected perturbation during treadmill walking (set speed of 1.4 ms1; mean and SD). Negative margin of stability values indicate an unstable body conﬁguration (i.e., subjects must make additional motor actions to preserve stability and to avoid a fall, e.g., by stepping), whereas positive margin of stability values indicate a stable body conﬁguration (i.e., no additional motor actions are needed to preserve stability). Clinical examination All examined participants of the UPVD patient group showed clear balance and coordination deﬁcits (16 failed the Unterberger test, 12 failed the straight line walking with eyes closed test, nine showed deﬁcits in the head impulse tests) and six patients had forms of nystagmus (one case of spontaneous nystagmus, ﬁve showed positive 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 5 Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. gait event and perturbation block speciﬁc. A simple con- trast test revealed no signiﬁcant differences in base of support and margin of stability at touchdown between results for head-shaking nystagmus). The clinical examin- ations of the control group conﬁrmed that all but one subject had any form of nystagmus, and no clear balance and coordination problems were identiﬁed. Only one control subject failed the Unterberger test, but this may be attributed to a lack of concentration, and as the sub- ject showed no other indications or symptoms of UPVD, the subject was not excluded from the study. ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Recovery locomotor behavior before and after experience of eight perturbations ba: Statistically signiﬁcant difference to base for both subjects groups (P < 0.05). baP: Statistically signiﬁcant difference to base for the UPVD group (P < 0.05). bl: Statistically signiﬁcant difference to block1 for both subjects groups (P < 0.05). blC: Statistically signiﬁcant difference to block1 for the control group (P < 0.05). blP: Statistically signiﬁcant difference to block1 for the UPVD group (P < 0.05). Statistically signiﬁcant difference between the UPVD and control groups (P < 0.05). 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 6 Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. UPVD patients and controls during nonperturbed walk- ing (base; Figs. 3 and 4). lower margin of stability values for the steps following the perturbation compared to baseline walking (for per- turbation block1: post1L–post6R; block8: post1L–post2R) with a signiﬁcant increase in margin of stability from perturbation block1 to perturbation block8 (post1L– post5L; P < 0.05). Hence, for perturbation block1 and perturbation block8, the UPVD patients required more than six and three recovery steps, respectively, to return to the margin of stability baseline level (P < 0.05; Fig. 3). Comparatively, the healthy control subjects needed for perturbation block1 and perturbation block8 ﬁve and two recovery steps to return to margin of sta- bility baseline, respectively, (P < 0.05; Fig. 3), with a sig- niﬁcant (P < 0.05) increase in margin of stability from block1 to block8 for the ﬁrst four steps postperturbation (post1L–post4R; Fig. 3). The application of unexpected unilateral resistance to the right leg while walking caused the margin of stability to signiﬁcantly decrease (P < 0.05) at touchdown of the perturbed right leg (pertbR) for both subject groups and for both perturbation block1 and perturbation block8, with no signiﬁcant differences between subject groups in the ﬁrst perturbation block (Fig. 3). However, the control group showed a signiﬁcant (P < 0.05) increase in margin of stability at touchdown of the perturbed leg in perturba- tion block8 in comparison to block1, resulting in signiﬁ- cantly lower margin of stability values for the UPVD patients compared to the control group in perturbation block8 (P < 0.05; Fig. 3). The extrapolated center of mass showed a signiﬁcant gait event 9 perturbation block inter- action (P < 0.05). Reactive locomotor responses across the eight perturbation blocks The analysis of the margin of stability at touchdown of the perturbed leg in each of the eight unexpected pertur- bation blocks (pertbR in block1 to pertbR in block8) dem- onstrated a signiﬁcant subject group x perturbation block interaction (P < 0.05; Fig. 5). The simple contrast test revealed no signiﬁcant differences in margin of stability at touchdown of the perturbed leg between subject groups within the ﬁrst four perturbation blocks (block1 to block4) but demonstrated signiﬁcantly higher margin of stability values for the control subjects compared to the UPVD patients for the last four perturbation blocks (block5 to block8; P < 0.05; Fig. 5). Accordingly, the control subjects showed signiﬁcantly (P < 0.05) higher margin of stability values for the last four unexpected perturbations (pertbR in block5 to block8) compared to the margin of stability at the ﬁrst unexpected perturbation (pertbR in block1; see Fig. 5). For the UPVD patients, the margin of stability at touchdown of the perturbed leg for any subsequent per- turbation block (pertbR in block2 to pertbR in block8) was not signiﬁcantly different from the ﬁrst unexpected per- turbation block (pertbR in block1; Fig. 5). Furthermore, the adaptation magnitude was signiﬁcantly (P < 0.05) greater for the healthy control participants compared with the UPVD patients, with controls and UPVD patients achieving mean magnitudes of 25.5% (SD30.2) and 0.4% (SD25.3), respectively. There was no signiﬁcant subject group effect, perturbation block effect or perturbation block x subject group interaction on base of support (range across the eight perturbation blocks and the two subject groups from 66.2 to 63.3 cm), margin of stability (7.1 to 4.9 cm) or extrapolated center of mass (59.5 to 57.5 cm) at touchdown of the contralateral leg left leg prior to perturbation. Hence, the base of support, margin At touchdown of the ﬁrst step following the ﬁrst and ﬁnal unexpected perturbations (post1L in block1 and block8), the base of support was signiﬁcantly (P < 0.05) higher compared to baseline walking for the control group, whereas no signiﬁcant differences were found for the UPVD patients (Fig. 4). Compared to the control group, the UPVD patients showed a lower base of sup- port at touchdown of the ﬁrst step following the pertur- bation independent of perturbation block (P < 0.05; Fig. 4). Recovery locomotor behavior before and after experience of eight perturbations Compared to baseline, the base of sup- port and the extrapolated center of mass were signiﬁcantly (P < 0.05) lower for both groups and both unexpected perturbation blocks at touchdown of the perturbed right leg (Fig. 4). However, the control group showed a signiﬁ- cant (P < 0.05) increase in base of support at touchdown of the perturbed leg in perturbation block8 in comparison to block1, whereas no signiﬁcant differences between blocks were found for the UPVD patients (Fig. 4). ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Reactive locomotor responses across the eight perturbation blocks During the following ﬁve consecutive steps (post2R–post6R) in perturbation block1, the UPVD patients showed lower base of support values compared to baseline (P values were not <0.05 for all steps; Fig. 4) and did not reach a steady state within the analyzed six steps following the unexpected perturbation (post1L– post6R). However, in the ﬁnal perturbation block (block8), the UPVD patients showed signiﬁcantly (P < 0.05) higher base of support values in comparison to perturbation block1 (Fig. 4), and reached baseline level base of support values within the third step following the perturbation (post3L). The extrapolated center of mass signiﬁcantly increased (P < 0.05) above baseline for both groups and both blocks at post1L, but showed no signiﬁ- cant subject group effect or subject group interaction (subject group 9 gait event, subject group 9 block, sub- ject group 9 block 9 gait event; Fig. 4). As a consequence of the above ﬁndings, the UPVD patient group demonstrated signiﬁcantly (P < 0.05) 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 7 e12222 Page 7 Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. mass at touchdown are differentially inﬂuenced by increased velocity, with the extrapolated center of mass affected to a greater extent (S€uptitz et al. 2012). of stability and extrapolated center of mass values for all preperturbation steps (preL in block1 to preL in block8) were not signiﬁcantly different from baseline (nonper- turbed walking) for either subject group. After the ﬁrst unexpected perturbation to the swing phase of the right leg in block1, margin of stability val- ues at touchdown of the perturbed leg (pertbR in block1) signiﬁcantly decreased (P < 0.05; Fig. 3) for both subject groups with mean values of about 15 cm, with no signiﬁcant differences between UPVD patients and controls. The appreciably negative margin of stability values demonstrates that the perturbation was appropri- ate to initiate an unstable body position at touchdown for the subjects. During the ﬁrst step following the per- turbation (post1L) in block1, both subject groups dem- onstrated an increased base of support at touchdown compared to at touchdown of the perturbed leg, but did not exceed the position of the extrapolated center of mass, resulting in an unstable body position (i.e., nega- tive margin of stability) at touchdown for both groups. Discussion This study aimed to examine the recovery responses (sta- bility recovery actions) and adaptive feedback potential in dynamic stability of UPVD patients and matched healthy control subjects during perturbed treadmill walking. In order to best isolate the reactive, feedback-driven response, the perturbations were unexpected, with no warning, the duration between onset of perturbation and touchdown was short, and the possible presence of prep- erturbation feedforward motor adjustments was checked. S€uptitz et al. (2013) recently demonstrated that such a method may be effective in analyzing feedback-driven locomotor corrections during perturbed gait. We hypoth- esized that recovery behavior would be less effective and adaptive feedback potential in dynamic stability during perturbed walking would be diminished in UPVD patients compared to matched healthy controls. It was found that, compared to matched healthy controls, UPVD patients required at least two more recovery steps to return to margin of stability baseline level after an unex- pected perturbation while walking. This clearly illustrates a UPVD patient related deﬁcit in dynamic stability con- trol during perturbed walking. Moreover, after repeated exposure to the perturbation task, signiﬁcant adaptive improvements in dynamic stability at touchdown of the perturbed leg were seen only in the healthy controls and the reactive adaptation magnitude was signiﬁcantly greater for the control participants compared to the UPVD patients. Therefore, the results support the hypothesis that UPVD patients have a diminished ability to effectively cope with unexpected perturbations while walking due to a slower recovery to margin of stability baseline values following perturbations and an apparent lack of feedback-driven locomotor adaptations. Figure 5. Margin of stability at touchdown of the perturbed leg (pertbR) for unilateral peripheral vestibular disorder (UPVD; n = 13) patients and matched healthy controls (CONT; n = 13) at the eight perturbation blocks following an unexpected perturbation during treadmill walking (set speed of 1.4 ms1; mean and SD). Negative margin of stability values indicate an unstable body conﬁguration (i.e., subjects must make additional motor actions to preserve stability and to avoid a fall, e.g., by stepping), whereas positive margin of stability values indicate a stable body conﬁguration (i.e., no additional motor actions are needed to preserve stability). Statistically signiﬁcant difference between the UPVD and control groups (P < 0.05). C1: Statistically signiﬁcant difference to pertbR1 for the control group (P < 0.05). Reactive locomotor responses across the eight perturbation blocks The controls were able to signiﬁcantly increase their base of support in comparison to their baseline (i.e., nonper- turbed walking) leading to higher (less negative) margin of stability values compared to the UPVD patients. By increasing the base of support more than the UPVD ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Adaptive Adjustments in Dynamic Gait Stability in UPVD Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. UPVD patients demonstrated an inability to increase their base of support greater than baseline level during the ﬁrst step postperturbation in the ﬁrst or ﬁnal block. This deﬁcient motor response resulted in a mechanically ineffective body position, negatively affecting the subse- quent recovery steps, delaying the return to baseline mar- gin of stability values for the UPVD patients. This delayed return demonstrates a persistent increased risk of falling in UPVD patients after repeated practice of the perturbation task, signifying the possible role of the ves- tibular system in the adaptation of the reactive response to repeated perturbations. patients, the control subjects created a more advanta- geous body position at the ﬁrst recovery step, positively inﬂuencing dynamic stability during the following recov- ery steps. A diminished ability to take a large anterior recovery step is a strong predictor for a higher fall risk (Maki and McIlroy 2006). Therefore, the above results suggest that the known increased fall risk in UPVD patients may be, in part, due to a diminished ability to effectively enlarge their base of support after an unex- pected perturbation while walking. In the course of several unexpected gait perturbations, control subjects showed signiﬁcant increases in margin of stability (less negative values) at touchdown of the per- turbed leg from the ﬁfth to eighth perturbation blocks compared to the ﬁrst block, attributable to a widened base of support in the eighth block. Due to the short duration between the onset of the perturbation and touchdown of the perturbed leg, and the lack of warning about the per- turbation, we can suggest that the observed motor response at the perturbed leg was, to a large extent, attributed to an improvement in the reactive, feedback-driven response. The current results conﬁrm previous ﬁndings (Pavol et al. 2004; Bierbaum et al. 2011) that healthy adults have the potential to adapt their recovery response to perturbations in a feedback-driven manner. We may speculate that, in this study, the internal representation of the postperturba- tion recovery steps and the motor response to the pertur- bation may have been continually updated with each perturbation, leading to a gradual improvement in recov- ery behavior in the healthy control subjects. Acknowledgments The ﬁnding that UPVD patients needed more steps to reach MS baseline in the ﬁnal block compared to controls may support the suggestion that the integration of vestib- ular and somatosensory feedback is necessary for postural adjustments to some degree (Horak et al. 2001). How- ever, the above-described improved recovery behavior may not have been solely caused by feedback-driven- reactive motor adjustments, due to task experience of the perturbation aiding cerebellar controlled feedforward adjustments (Morton and Bastian 2006). We thank Martin K€ussel-Feldker and Thomas F€orster and their teams for their technical assistance and support throughout this research project. Adaptive Adjustments in Dynamic Gait Stability in UPVD Analysis of the step prior to perturbation in all pertur- bation blocks (preL in block1 to block8) revealed no sig- niﬁcant differences in the margin of stability, base of support, or extrapolated center of mass compared to baseline for either subject group. This suggests that there were no clear predictive adjustments in gait or posture which inﬂuenced the recovery behavior following any of the perturbations. While anxiety may have caused the recovery behavior of four subjects that grasped the tread- mill handrails, this may have been due to insufﬁcient recovery responses to the perturbation. Having excluded these participants from the statistics, our results may underestimate the impact of UPVD on dynamic stability and could have been more pronounced had they been included in the analysis. In conclusion, our results demonstrate a deﬁciency in the ability of UPVD patients to effectively cope with unexpected perturbations while walking, before and after repeated practice of the perturbation task. UPVD patients showed a diminished recovery response, characterized by an inability to signiﬁcantly widen their base of support following the perturbations and an apparent absence of adaptive feedback potential in dynamic stability control. This suggests that a lack of accurate vestibular sensory feedback information may result in diminished correc- tions and adaptations of locomotor behavior during per- turbed walking, increasing the risk of falls during walking, and supports the notion that the vestibular system may be important for the adaptation of feedback-driven-reac- tive responses during locomotion. In contrast to the above ﬁndings, the examined UPVD patients showed no signiﬁcant changes in margin of stabil- ity for the perturbed leg across the eight perturbation blocks, demonstrating similar values in the ﬁrst and last block. Accordingly, the reactive adaptation magnitude after eight unexpected perturbations was signiﬁcantly higher for the control group than the UPVD patients (25.5% vs. 0.4%). The current results, therefore, indicate that unilater- ally disturbed vestibular sensory feedback information may negatively affect the accuracy of the updating of the internal representation of the internal and external mechanical envi- ronment during perturbed gait, which may result in a lack of adaptation of the reactive response to perturbations. ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Discussion Nonperturbed walking (baseline) revealed similar posi- tive margin of stability values at touchdown among both subject groups (on average about 6 cm). This indicates that walking on the treadmill at 1.4 ms1 had a similar dynamic stability demand for both subject groups. One might argue that the current positive margin of stability values may contradict some previous ﬁndings (Bierbaum et al. 2010, 2011; H€ohne et al. 2011) that showed negative margin of stability values at touchdown during nonper- turbed walking (range of 1.7 to 11.9 cm). However, the differences are likely related to the different gait veloc- ities of the studies (1.8–2.0 ms1 vs. 1.4 ms1 in this study) as the base of support and extrapolated center of Figure 5. Margin of stability at touchdown of the perturbed leg (pertbR) for unilateral peripheral vestibular disorder (UPVD; n = 13) patients and matched healthy controls (CONT; n = 13) at the eight perturbation blocks following an unexpected perturbation during treadmill walking (set speed of 1.4 ms1; mean and SD). Negative margin of stability values indicate an unstable body conﬁguration (i.e., subjects must make additional motor actions to preserve stability and to avoid a fall, e.g., by stepping), whereas positive margin of stability values indicate a stable body conﬁguration (i.e., no additional motor actions are needed to preserve stability). Statistically signiﬁcant difference between the UPVD and control groups (P < 0.05). C1: Statistically signiﬁcant difference to pertbR1 for the control group (P < 0.05). 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 8 References Maki, B. E., and W. E. McIlroy. 2006. Control of rapid limb movements for balance recovery: age-related changes and implications for fall prevention. Age Ageing 35:ii12– ii18. Bent, L. R., B. J. McFadyen, and J. T. Inglis. 2005. Vestibular contributions during human locomotor tasks. Exerc. Sport Sci. Rev. 33:107–113. Marigold, D. S., and A. E. Patla. 2002. Strategies for dynamic stability during locomotion on a slippery surface: effects of prior experience and knowledge. J. Neurophysiol. 88:339–353. Bhatt, T., J. D. Wening, and Y. C. Pai. 2006. Adaptive control of gait stability in reducing slip-related backward loss of balance. Exp. Brain Res. 170:61–73. Bierbaum, S., A. Peper, K. Karamanidis, and A. Arampatzis. 2010. Adaptational responses in dynamic stability during disturbed walking in the elderly. J. Biomech. 43:2362–2368. Morton, S. M., and A. J. Bastian. 2006. Cerebellar contributions to locomotor adaptations during splitbelt treadmill walking. J. Neurosci. 26:9107–9116. Bierbaum, S., A. Peper, K. Karamanidis, and A. Arampatzis. 2011. Adaptive feedback potential in dynamic stability during disturbed walking in the elderly. J. Biomech. 44:1921–1926. Neuhauser, H. K., M. Von Brevern, A. Radtke, F. Lezius, M. Feldmann, T. Ziese, et al. 2005. Epidemiology of vestibular vertigo: a neurotologic survey of the general population. Neurology 65:898–904. Borello-France, D. F., J. D. Gallagher, M. Redfern, J. M. Furman, and G. E. Carvell. 1999. Voluntary movement strategies of individuals with unilateral peripheral vestibular hypofunction. J. Vestib. Res. 9:265–275. Pai, Y. C., J. D. Wening, E. F. Runtz, K. Iqbal, and M. J. Pavol. 2003. Role of feedforward control of movement stability in reducing slip-related balance loss and falls among older adults. J. Neurophysiol. 90:755–762. Dempster, W. T., W. C. Gabel, and W. J. Felts. 1959. The anthropometry of the manual workspace for the seated subject. Am. J. Phys. Anthropol. 17:289–317. Pang, M. Y., T. Lam, and J. F. Yang. 2003. Infants adapt their stepping to repeated trip-inducing stimuli. J. Neurophysiol. 90:2731–2740. Hillier, S. L., and M. McDonnell. 2011. Vestibular rehabilitation for unilateral peripheral vestibular dysfunction. Cochrane Database Syst. Rev., 2:CD005397. Pavol, M. J., E. F. Runtz, and Y. C. Pai. 2004. Young and older adults exhibit proactive and reactive adaptations to repeated slip exposure. J. Gerontol. A Biol. Sci. Med. Sci. 59:494–502. Hof, A. L., M. G. Gazendam, and W. E. Sinke. 2005. The condition for dynamic stability. J. Biomech. 38:1–8. Raptis, H. A., E. Dannenbaum, N. Paquet, and A. G. Feldman. 2007. ª 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society. Conﬂict of interest No conﬂicts of interest, ﬁnancial or otherwise, are declared by the authors. 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 9 Adaptive Adjustments in Dynamic Gait Stability in UPVD C. McCrum et al. References Vestibular system may provide equivalent motor actions regardless of the number of body segments involved in the task. J. Neurophysiol. 97:4069–4078. H€ohne, A., C. Stark, G. P. Br€uggemann, and A. Arampatzis. 2011. Effects of reduced plantar cutaneous afferent feedback on locomotor adjustments in dynamic stability during perturbed walking. J. Biomech. 44:2194–2200. Rossignol, S., R. Dubuc, and J. P. Gossard. 2006. Dynamic sensorimotor interactions in locomotion. Physiol. Rev. 86:89–154. Homann, B., A. Plaschg, M. Grundner, A. Haubenhofer, T. Griedl, G. Ivanic, et al. 2013. The impact of neurological disorders on the risk for falls in the community dwelling elderly: a case-controlled study. BMJ Open 3:e003367. Sainburg, R. L., C. Ghez, and D. Kalakanis. 1999. Intersegmental dynamics are controlled by sequential anticipatory, error correction, and postural mechanisms. J. Neurophysiol. 81:1045–1056. Horak, F. B., G. M. Earhart, and V. Dietz. 2001. Postural responses to combinations of head and body displacements: vestibular-somatosensory interactions. Exp. Brain Res. 141:410–414. S€uptitz, F., K. Karamanidis, M. M. Catala, and G. P. Br€uggemann. 2012. Symmetry and reproducibility of the components of dynamic stability in young adults at different walking velocities on the treadmill. J. Electromyogr. Kinesiol. 22:301–307. Kennedy, P. M., A. N. Carlsen, J. T. Inglis, R. Chow, I. M. Franks, and R. Chua. 2003. Relative contributions of visual and vestibular information on the trajectory of human gait. Exp. Brain Res. 153:113–117. S€uptitz, F., M. M. Catala, G. P. Br€uggemann, and K. Karamanidis. 2013. Dynamic stability control during perturbed walking can be assessed by a reduced kinematic model across the adult female lifespan. Hum. Mov. Sci. 32:1404–1414. Lam, T., C. Wolstenholme, M. Van Der Linden, M. Y. C. Pang, and J. F. Yang. 2003. Stumbling corrective responses during treadmill-elicited stepping in human infants. J. Physiol. 553:319–331. Lam, T., M. Anderschitz, and V. Dietz. 2006. Contribution of feedback and feedforward strategies to locomotor adaptations. J. Neurophysiol. 95:766–773. Wolpert, D. M., Z. Ghahramani, and M. I. Jordan. 1995. An internal model for sensorimotor integration. Science 269:1880–1882. L€uscher, M., S. Theilgaard, and B. Edholm. 2014. Prevalence and characteristics of diagnostic groups amongst 1034 patients seen in ENT practices for dizziness. J. Laryngol. Otol. 128:128–133. Woltring, H. J. 1986. A Fortran package for generalized, cross-validatory spline smoothing and differentiation. Adv. Eng. Softw. 8:104–113. 2014 \| Vol. 2 \| Iss. 12 \| e12222 Page 10
https://openalex.org/W2020150330	http://openaccess.sgul.ac.uk/541/1/1471-2458-8-182.pdf	English	null	Less healthy, but more active: Opposing selection biases when recruiting older people to a physical activity study through primary care	BMC public health	2,008	cc-by	4,226	BioMed Central BioMed Central BMC Public Health Open Access Received: 11 January 2008 Accepted: 27 May 2008 MC Public Health 2008, 8:182 doi:10.1186/1471-2458-8-182 This article is available from: http://www.biomedcentral.com/1471-2458/8/182 © 2008 Harris et al; licensee BioMed Central Ltd. © 2008 Harris 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. © 2008 Harris et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creative which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cit Page 1 of 6 (page number not for citation purposes) Open Acce Research article Less healthy, but more active: Opposing selection biases when recruiting older people to a physical activity study through primary care Tess J Harris1,2, Christina R Victor3, Iain M Carey1, Rika Adams2 and Derek G Cook1 Address: 1Division of Community Health Sciences, St George's, University of London, Cranmer Terrace, Tooting, London, SW17 ORE, UK, 2Sonning Common Health Centre, Wood Lane, Sonning Common, Oxfordshire, RG4 9SW, UK and 3School of Health and Social Care, Reading University, Whiteknights Lane, Reading, Berkshire, UK Email: Tess J Harris - tharris@sgul.ac.uk; Christina R Victor - c.r.victor@reading.ac.uk; Iain M Carey - i.carey@sgul.ac.uk; Rika Adams - Rika.Adams@gp-K84020.nhs.uk; Derek G Cook - d.cook@sgul.ac.uk * Corresponding author Published: 27 May 2008 BMC Public Health 2008, 8:182 doi:10.1186/1471-2458-8-182 Received: 11 January 2008 Accepted: 27 May 2008 This article is available from: http://www.biomedcentral.com/1471-2458/8/182 © 2008 Harris 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. Received: 11 January 2008 Accepted: 27 May 2008 Abstract Background: Physical activity studies in older people experience poor recruitment. We wished to assess the influence of activity levels and health status on recruitment to a physical activity study in older people. Methods: Comparison of participants and non-participants to a physical activity study using accelerometers in patients aged ≥ 65 years registered with a UK primary care centre. Logistic regression was used to calculate odds ratios (OR) of participants in the accelerometer study with various adjustments. Analyses were initially adjusted for age, sex and household clustering; the health variables were then adjusted for physical activity levels and vice versa to look for independent effects. Results: 43%(240/560) participated in the physical activity study. Age had no effect but males were more likely to participate than females OR 1.4(1.1–1.8). 46% (76/164) of non-participants sent the questionnaire returned it. The 240 participants reported greater physical activity than the 76 non- participants on all measures, eg faster walking OR 3.2(1.4–7.7), or 10.4(3.2–33.3) after adjustment for health variables. Participants reported more health problems; this effect became statistically significant after controlling for physical activity, eg disability OR 2.4(1.1–5.1). Conclusion: Physical activity studies on older primary care patients may experience both a strong bias towards participants being more active and a weaker bias towards participants having more health problems and therefore primary care contact. The latter bias could be advantageous for physical activity intervention studies, where those with health problems need targeting. Ethical approval & informed consent Ethical approval & informed consent Ethics committee approval for the study was given by Oxfordshire REC A (reference no. 06/Q1604/94). The patient information sheets sent to all 560 individuals explained the study in detail, including the use of ques- tionnaire information provided by those returning ques- tionnaires, but not wanting to participate further. Full informed written consent was obtained from the 240 par- ticipating in the physical activity study. 88 did not return questionnaires Methods A f As part of a randomized controlled trial of different recruitment strategies to a physical activity study, 560 patients ≥ 65 years registered with a primary health care centre (general practice) in Oxfordshire, UK were ran- domly selected by household [8]. Those living in care homes, those with dementia, terminal illness, poorly con- trolled cardiac failure or unstable angina and those house- bound due to disability were first excluded by computer record search and by general (family) practitioner and dis- trict (community) nurse examination of registered patient lists. All 560 patients were invited to take part in a study measuring customary physical activity levels objectively for a 7-day period using motion sensors (accelerometers and pedometers). A random half (280) also received a 12- page questionnaire with their study information, asking details about physical health (general health, limiting longstanding illness [9], disability [10], pain [11], chronic disease [12], smoking status, weight and height), depres- sive symptoms [13], self-reported physical activity levels [14,15] and attitudes towards physical activity [16] (See Additional file 1). Subjects were encouraged to return the questionnaire, whether or not they participated in the physical activity study, thus allowing a comparison of par- ticipants and non-participants. Those participating who had not been randomized to receive a questionnaire, completed one at their baseline assessment. Background ticipation, both for surveys (46% [1], 57% [2]) and more markedly for intervention studies (26% [3], 32% [4], 35% [2]). Information about non-participants is lacking, but Background Physical activity studies on older people often recruit through primary care, such studies usually report low par- Page 1 of 6 (page number not for citation purposes) Page 1 of 6 (page number not for citation purposes) http://www.biomedcentral.com/1471-2458/8/182 BMC Public Health 2008, 8:182 the cluster option in STATA 9 [16]. For age and sex com- parisons the analyses were based on all 520 subjects. For analyses examining health and physical activity from the brief questionnaire, the comparison was based on all par- ticipants to the accelerometer study and those non-partic- ipants who received a questionnaire and returned it. Analyses were adjusted for age, sex and household cluster- ing. Health measures were adjusted additionally for the effect of physical activity (walking pace and number of hours walked in the last week) and physical activity meas- ures for the effect of health (limiting long-standing illness, number of chronic diseases, disability, falls and chronic pain) in order to see whether any effects were independ- ent. To check that it was reasonable to combine postal questionnaires on participants with those completed at baseline assessment, a comparison of participants and non-participants as above was repeated, restricted to those randomised to postal questionnaires. higher activity levels have been associated with increased recruitment [4,5], leading to potential selection bias and difficulties in generalizing results [6]. Unfortunately, these studies did not report on non-participants' health status. Whilst older people with higher activity levels tend to be healthier [7], recruiting through primary care could encourage those with more illnesses and primary care contact to respond, leading to bias in an opposing direc- tion. Our objective was to compare the self-reported health status and physical activity levels of participants and non-participants in a primary care based physical activity study. Results Recruitment rate to the physical activity study was 43%(240/560). Comparison overall of participants (n = 240) and non-participants (n = 320), adjusted for age, sex and household clustering, showed that males were more likely to participate OR 1.4(95% C.I. 1.1–1.8). There was no statistically significant association between age and participating, baseline age 65–69 OR = 1, age 70–79 OR 0.8(95% C.I. 0.5–1.1), age = 80 OR 0.8(95% C.I. 0.5– 1.4). Of the 280 people sent a postal questionnaire, 116 were recruited to the study and 164 were not, of these non-par- ticipants 76/164(46%) completed the questionnaire (see Figure 1). Table 1 shows the comparison of participants (n = 240) (116 from the postal questionnaire group, 124 who completed the questionnaire at baseline assessment) and non-participants (n = 76) who returned the question- naire. The results for age and sex are very similar to those for participants and non-participants overall. Of the 280 people sent a postal questionnaire, 116 were recruited to the study and 164 were not, of these non-par- ticipants 76/164(46%) completed the questionnaire (see Figure 1). Table 1 shows the comparison of participants (n = 240) (116 from the postal questionnaire group, 124 who completed the questionnaire at baseline assessment) and non-participants (n = 76) who returned the question- naire. The results for age and sex are very similar to those for participants and non-participants overall. Participant flow through the study Figure 1 Participant flow through the study. 560 randomly selected older people aged 65 280 randomly sent questionnaires 280 randomly not sent questionnaires 116 recruited into PA study 164 not recruited into PA study 124 recruited into PA study (questionnaires completed at recruitment) 156 not recruited into PA study 76 returned questionnaires 88 did not return questionnaires Analysis i i Participant flow through the study Figure 1 Participant flow through the study. Analysis i i Logistic regression was used to estimate odds ratios for participating in the accelerometer study, adjusting for age and sex as appropriate and household clustering, using Page 2 of 6 (page number not for citation purposes) BMC Public Health 2008, 8:182 http://www.biomedcentral.com/1471-2458/8/182 Table 1: Comparison of non-participants & participants of physical activity study amongst questionnaire responders Non-participants N = 76 n (%) Participants N = 240 n (%) Crude OR for participating (95% CI) OR (95% CI) adj for age, sex, household clustering1 DEMOGRAPHIC Sex Female 45 (59.2) 115 (47.9) 1 1 Male 31 (40.8) 125 (52.1) 1.6 (0.9–2.7) 1.6 (1.1–2.4) Age 65–69 23 (30.3) 87 (36.3) 1 1 70–79 40 (52.6) 111 (46.3) 0.7 (0.4–1.3) 0.8 (0.4–1.4) 80 or more 13 (17.1) 42 (17.5) 0.9 (0.4–1.9) 0.8 (0.3–2.0) SELF-REPORTED HEALTH OR (95% CI) adj for age, sex, household clustering & self- reported activity2 General Health Very good/Good 58 (84.1) 200 (85.1) 1 1 1 Fair/bad 11 (15.9) 35 (14.9) 0.9 (0.7–1.4) 1.1 (0.7–1.6) 1.3 (0.9–2.0) Limiting longstanding illness No 53(77.9) 176 (73.6) 1 1 1 Yes 15 (22.1) 65 (28.6) 1.4 (0.7–2.7) 1.6 (0.8–3.3) 2.9 (1.2–7.1) Disability No 40 (53.3) 125 (52.7) 1 1 1 Yes 35 (46.7 112 (47.3) 1.0 (0.6–1.7) 1.3 (0.7–2.5) 2.4 (1.1–5.1) Chronic pain No 46 (66.7) 146 (63.5) 1 1 1 Yes 23 (33.3) 84 (36.5) 1.2 (0.7–2.0) 1.4 (0.7–2.6) 2.1 (1.0–4.5) Chronic disease No 25 (32.9) 55 (22.9) 1 1 1 Yes 51 (67.1) 185 (77.1) 1.6 (0.9–2.9) 1.7 (0.9–3.0) 1.8 (1.0–3.2) Use a walking aid No 66 (89.2) 217 (91.9) 1 1 1 Yes 8 (10.8) 19 (8.1) 0.7 (0.3–1.7) 0.8 (0.3–2.1) 1.4 (0.5–3.8) Fallen in last year No 56 (76.7) 169 (71.9) 1 1 1 Yes 17 (23.3) 66 (28.1) 1.3 (0.7–2.4) 1.6 (0.8–2.9) 2.0 (1.0–4.0) Current smoker No 72 (96.0) 221 (94.0) 1 1 1 Yes 3 (4.0) 14 (6.0) 1.5 (0.4–5.4) 1.3 (0.3–5.1) 1.5 (0.4–5.1) Body Mass Index Normal weight 28 (46.7) 97 (43.1) 1 1 1 Overweight or obese 32 (53.3) 128 (56.9) 1.2 (0.7–2.0) 1.2 (0.7–2.4) 1.4 (0.7–2.6) Geriatric Depression Score3 <4 67 (88.2)) 213 (89.5) 1 1 1 4 or more 9 (11.8) 25 (10.5) 0.9 (0.4–2.0) 0.9 (0.4–2.0) 1.1 (0.5–2.5) SELF-REPORTED ACTIVITY OR (95% CI) adj for h h ld Table 1: Comparison of non-participants & participants of physical activity study amongst questionnaire responders mparison of non-participants & participants of physical activity study amongst questionnaire responders http://www.biomedcentral.com/1471-2458/8/182 BMC Public Health 2008, 8:182 Hours walked in last week? Analysis i i None 24 (31.6) 37 (15.4) 1 1 1 Up to 2 hours 25 (32.9) 87 (36.3) 2.3 (1.1–4.5) 2.3 (1.1–4.7) 2.5 (1.1–5.6) More than 2 hours 27 (35.5) 116 (48.3) 2.8 (1.4–5.4) 2.7 (1.3–5.5) 2.8 (1.3–6.1) Average hours gardening weekly None 19 (25.3) 23 (10.0) 1 1 1 Up to 3.5 hrs/week 30 (40.0) 102 (44.4) 2.8 (1.4–5.8) 2.7 (1.3–5.6) 2.7 (1.1–6.9) >3.5 hrs/week 26 (34.7) 105 (45.7) 3.3 (1.6–7.0) 2.8 (1.3–6.3) 3.3 (1.2–8.7) Activity compared to others Less active or the same 27 (35.5) 49 (20.6) 1 1 1 More active 33 (43.4) 116 (48.7) 1.9 (1.1–3.6) 1.9 (1.0–3.6) 2.2 (1.0–4.5) Far more active 16 (21.1) 73 (30.7) 2.5 (1.2–5.1) 2.3 (1.1–5.2) 3.1 (1.3–7.0) Do you cycle? No 63 (87.5) 187 (78.9) 1 1 1 Yes 9 (12.5) 50 (21.1) 1.9 (0.9–4.0) 1.7 (0.7–4.1) 2.0 (0.8–5.1) Do you walk a dog? No 67 (88.2) 184 (78.6) 1 1 1 Yes 9 (11.8) 50 (21.4) 2.0 (0.9–4.3) 2.0 (0.8–5.2) 2.0 (0.7–5.2) Do you do heavy housework? No 23 (34.3) 50 (22.7) 1 1 1 Yes 44 (65.7) 170 (77.3) 1.8 (1.0–3.2) 1.7 (1.0–3.2) 2.3 (1.2–4.5) Positive attitudes towards activity? Low 35 (53.9) 73 (31.6) 1 1 1 High 30 (46.2) 158 (68.4) 2.5 (1.4–4.4) 2.5 (1.4–4.5) 3.4 (1.8–6.6) 1Household clustering – adjusted for the clustering by 237 households. 2Self-reported physical activity- adjusted for effect of walking pace and hours walked in last week 3Geriatric Depression Score 15 items, using cut-off <4/≥4, which gives a sensitivity of 91% and specificity of 72% for detecting major depression [13]. 4Self-reported health – adjusted for effect of limiting longstanding illness, chronic disease, disability, falls & pain. Table 1: Comparison of non-participants & participants of physical activity study amongst questionnaire responders (Continued) on of non-participants & participants of physical activity study amongst questionnaire responders (Continued) Table 1: Comparison of non-participants & participants of physical activity study amongst questionna Table 1: Comparison of non-participants & participants of physical activity study [13]. 4Self-reported health – adjusted for effect of limiting longstanding illness, chronic disease, disability, falls & pain. After adjusting for age, sex and household clustering, par- ticipants tended to report more health problems than non-participants for most variables, (but no differences were statistically significant at p = 0.05). The exceptions were: use of a walking aid which showed a non-significant effect in the opposite direction; and depression score which was unrelated to participation. Analysis i i Adjusting for self- reported physical activity levels strengthened the associa- tions between poorer health and participating such that those reporting limiting longstanding illness, disability, chronic pain, chronic disease and a fall in the last year were more likely to participate. Analyses restricted to participants (116) and non-partici- pants (76) who returned the postal questionnaire showed very similar findings in terms of effect estimates to those presented, based on all questionnaire completers. Although the confidence intervals were wider due to smaller numbers, several associations with participation reached statistical significance at p = 0.05 for both health, e.g. disability, adjusted OR 2.4 (05% C.I. 1.1–5.5) and physical activity variables, e.g. higher activity levels than others, adjusted OR 3.1 (95% C.I. 1.1–8.7). Household clustering – adjusted for the clustering by 237 households. S f f ff f Authors' contributions likely to participate). This has been reported in other physical activity studies in older people [3-5]. Secondly, participants reported more health problems than non- participants. This effect opposes the physical activity effect, and was much clearer after controlling for self- reported physical activity. At first, these findings seem counter-intuitive. Other studies in older people have reported that participants have better health than non- participants [18-20], suggesting a "healthy volunteer" effect, although Ives et al found participants were more likely to have a disease history and use health services more [21]. However, it seems plausible that recruiting older people to studies through primary care may lead to those with increased contact with primary care (ie more illnesses) to be more likely to take part. This fits with other work showing that older primary care patients with poorer physical and psychological health and greater pri- mary care service use were more likely to give consent for their health records to be accessed for research [22]. Unfortunately, the other primary care based studies show- ing differences in physical activity levels between partici- pants and non-participants did not report on health or functional ability [3-5]. TJH, CRV and DGC designed the study. TJH and RA col- lected data. TJH, IMC and DGC were involved in analysis and interpretation of data. All authors were involved in drafting the manuscript and revising it critically for impor- tant intellectual content. All authors have given final approval of the version to be published. Conclusion p p J 6. Tai SS, Gould M, Iliffe S: Promoting healthy exercise among older people in general practice: issues in designing and eval- uating therapeutic interventions. British Journal of General Prac- tice 1997, 47:119-22. Physical activity studies on older people recruited from primary care settings may be biased by two opposing issues which need consideration when generalizing the results: a strong bias towards participants being more physically active, and a weaker bias towards participants having more health problems and therefore likely primary care contact. This latter bias could be used to advantage when considering interventions to increase physical activ- ity, where those who are least active and those with more physical health problems need targeting most. 7. Ettinger WH Jr: Physical activity and older people: a walk a day keeps the doctor away. J Am Geriatr Soc 1996, 44:207-8. 8. Harris TJ, Victor CR, Carey IM, Adams R, Cook DG: Optimising recruitment into a study of physical activity in older people: a randomised controlled trial of different approaches. Age & Ageing in press. 2008 7. Ettinger WH Jr: Physical activity and older people: a walk a day keeps the doctor away. J Am Geriatr Soc 1996, 44:207-8. 8. Harris TJ, Victor CR, Carey IM, Adams R, Cook DG: Optimising recruitment into a study of physical activity in older people: a randomised controlled trial of different approaches. Age & Ageing in press. 2008 9. Joint Health Surveys Unit: Health Survey for England 1998. In Methodology and Documentation Volume 2. London, The Stationery Office; 1999. g g p 9. Joint Health Surveys Unit: Health Survey for England 1998. In Methodology and Documentation Volume 2. London, The Stationery Office; 1999. 10. 10. McGee MA, Johnson A, Kay D: The Medical Research Council Cognitive Functioning and Ageing Study (MRC CFAS). The description of activities of daily living in five centres in Eng- land and Wales. Age & Ageing 1998, 27:605-13. land and Wales. Age & Ageing 1998, 27:605-13. References 1. Crombie IK, Irvine L, Williams B, et al.: Why older people do not participate in leisure time physical activity: a survey of activ- ity levels, beliefs and deterrents. Age Ageing 2004, 33:287-92. 1. Crombie IK, Irvine L, Williams B, et al.: Why older people do not participate in leisure time physical activity: a survey of activ- ity levels, beliefs and deterrents. Age Ageing 2004, 33:287-92. 1. Crombie IK, Irvine L, Williams B, et al.: Why older people do not participate in leisure time physical activity: a survey of activ- ity levels, beliefs and deterrents. Age Ageing 2004, 33:287-92. y g g g 2. Stevens W, Hillsdon M, Thorogood M, McArdle D: Cost-effective- ness of a primary care based physical activity intervention in 45–74 year old men & women: a randomised controlled trial. Br J Sports Med 1998, 32:236-41. J p 3. Munro JF, Nicholl JP, Brazier JE, Davey R, Cochrane T: Cost effec- tiveness of a community based exercise programme in over 65 year olds: cluster randomised trial. J Epidemiol Community Health 2004, 58:1004-10. 4. Halbert JA, Silagy CA, Finucane P, Withers RT, Hamdorf PA: Recruitment of older adults for a randomized, controlled trial of exercise advice in a general practice setting. J Am Ger- iatr Soc 1999, 47:477-81. 5. Crombie IK, McMurdo ME, Irvine L, Williams B: Overcoming bar- riers to recruitment in health research: concerns of poten- tial participants need to be dealt with. BMJ 2006, 333:398. Discussion Recruitment to our motion sensor activity study was 43%, higher than for physical activity intervention studies in this age group [2-4], but lower than for surveys [1,2]. With this recruitment level, estimation of potential non- response bias is important. After adjusting for age, sex and household clustering, par- ticipants reported more physical activity than non-partic- ipants for all measures and more positive attitudes towards physical activity. Apart from cycling and dog- walking, where numbers were small, these differences were all statistically significant and several showed dose- response effects. Adjusting these estimates for self- reported health strengthened these effects still further. Our findings suggest two separate issues leading to poten- tial bias. Firstly, participants were more physically active than non-participants (consistent with men being more Page 4 of 6 (page number not for citation purposes) (page number not for citation purposes) http://www.biomedcentral.com/1471-2458/8/182 http://www.biomedcentral.com/1471-2458/8/182 BMC Public Health 2008, 8:182 http://www.biomedcentral.com/1471-2458/8/182 Acknowledgements We are grateful to all the partners, staff and patients of Sonning Common Health Centre, Oxfordshire, UK for their support with this study. Funding for the study was provided by the Thames Valley Primary Care Research Partnership (WCRM03). The sponsors played no role in the design, execu- tion, analysis & interpretation of data or in the writing of the manuscript or the decision to submit the manuscript for publication. An important weakness of our study was that self-reported health and physical activity details were only available on those non-participants who completed the questionnaire. The similarities seen in age and sex comparisons between participants and non-participants overall and in those responding to the questionnaire is reassuring and suggests that the non-participant questionnaire responders are rep- resentative of non-participants, at least in terms of age and sex, although they could still differ in other important ways (such as activity level or health) that we lack infor- mation on. The similarities found when restricting analy- ses to only those sent postal questionnaires, confirms that it was reasonable to include participants in the analysis who completed the questionnaire at their baseline assess- ment. Additional material Additional file 1 supplementary questionnaire harris2008. Questionnaire on health and self-report physical activity levels. Click here for file [http://www.biomedcentral.com/content/supplementary/1471- 2458-8-182-S1.doc] Additional file 1 supplementary questionnaire harris2008. Questionnaire on health and self-report physical activity levels. Click here for file [http://www.biomedcentral.com/content/supplementary/1471- 2458-8-182-S1.doc] Competing interests 11. Ware JE, Sherbourne CD: The MOS 36-item short form health survey: conceptual framework and item selection. Med Care 1992, 30:473-83. p g The authors declare that they have no competing interests. Page 5 of 6 (page number not for citation purposes) Page 5 of 6 (page number not for citation purposes) BMC Public Health 2008, 8:182 http://www.biomedcentral.com/1471-2458/8/182 http://www.biomedcentral.com/1471-2458/8/182 http://www.biomedcentral.com/1471-2458/8/182 12. Roberts RE, Kaplan GA, Shema SJ, Strawbridge WJ: Prevalence and correlates of depression in an aging cohort: the Alameda County Study. J Gerontol B Psychol Sci Soc Sci 1997, 52:S252-S258. y y J y 13. D'Ath P, Katona P, Mullan E, Evans S, Katona C: Screening, detec- tion and management of depression in elderly primary care attenders. I: The acceptability and performance of the 15 item Geriatric Depression Scale (GDS15) & the develop- ment of short versions. Fam Pract 1994, 11:260-6. 14. Caspersen CJ, Bloemberg BP, Saris WH, Merritt RK, Kromhout D: The prevalence of selected physical activities and their rela- tion with coronary heart disease risk factors in elderly men: the Zutphen Study, 1985. Am J Epidemiol 1991, 133:1078-92. 15. Washburn RA, Smith KW, Jette AM, Janney CA: The Physical Activity Scale for the Elderly (PASE): development and eval- uation. J Clin Epidemiol 1993, 46:153-62. J p 16. Jette AM, Rooks D, Lachman M, Lin TH, Levenson C, Heislein D, et al.: Home-based resistance training: predictors of participa- tion and adherence. Gerontologist 1998, 38:412-21. g 17. STATA 9.0 Statistics/Data Analysis. College Station, Texas, USA, Statacorp; 2005. 18. van Heuvelen MJ, Hochstenbach JB, Brouwer WH, de Greef MH, Zijl- stra GA, van Jaarsveld E, et al.: Differences between participants and non-participants in an RCT on physical activity and psy- chological interventions for older persons. Aging Clin Exp Res 2005, 17:236-45. 19. Hebert R, Bravo G, Korner-Bitensky N, Voyer L: Refusal and infor- mation bias associated with postal questionnaires and face- to-face interviews in very elderly subjects. J Clin Epidemiol 1996, 49:373-81. 20. Wagner EH, Grothaus LC, Hecht JA, LaCroix AZ: Factors associ- ated with participation in a senior health promotion pro- gram. Gerontologist 1991, 31:598-602. g g 21. Ives DG, Traven ND, Kuller LH, Schulz R: Selection bias and non- response to health promotion in older adults. Epidemiology 1994, 5:456-61. 22. Harris T, Cook DG, Victor C, Beighton C, DeWilde S, Carey I: Link- ing questionnaires to primary care records: factors affecting consent in older people. Competing interests J Epidemiol Community Health 2005, 59:336-8. http://www.biomedcentral.com/1471-2458/8/182/pre pub http://www.biomedcentral.com/1471-2458/8/182/pre pub Publish with BioMed Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp BioMedcentral Page 6 of 6 (page number not for citation purposes) Publish with BioMed Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp BioMedcentral Publish with BioMed Central and every scientist can read your work free of charge Pre-publication history The pre-publication history for this paper can be accessed here:
https://openalex.org/W3178880622	https://www.researchsquare.com/article/rs-156307/v1.pdf?c=1631887365000	English	null	Correlation between full-thickness degenerative supraspinatus tear and radiographic parameters including the acromiohumeral centre edge angle and the greater tuberosity angle	BMC musculoskeletal disorders	2,021	cc-by	6,191	Do Acromiohumeral Centre Edge Angle and Greater Tuberosity Angle Correlate With the Full-Thickness Degenerative Supraspinatus Tear? Chaiyanun Vijittrakarnrung ( Mahidol University Praman Fuangfa Mahidol University Suphaneewan Jaovisidha Mahidol University Chusak Kijkunasathian Mahidol University Research Article Keywords: Rotator cuff tear, Acromiohumeral centre edge angle, Greater tuberosity angle, Subacromial impingement, Radiographic measurement Posted Date: February 3rd, 2021 DOI: https://doi.org/10.21203/rs.3.rs-156307/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 BMC Musculoskeletal Disorders on July 6th, 2021. See the published version at https://doi.org/10.1186/s12891-021-04489-x. Page 1/16 Page 1/16 Page 1/16 Abstract Background: Many radiographic parameters associated with the extrinsic cause of supraspinatus tear have been proposed. The aim of this study was to correlate the relationship between full-thickness degenerative supraspinatus tear (FTDST) and the patient’s radiographic parameters, including the acromiohumeral centre edge angle (ACEA) and the greater tuberosity angle (GTA). Methods: A retrospective study was conducted. We included 116 patients who had undergone shoulder arthroscopic surgery at our institute. The case group included FTDST patients, while the control group also included patients without evidence of supraspinatus tear. In each patient, the ACEA and GTA values were measured and analysed by two independent observers. Intra-inter observer reliability was assessed. Multivariate regression analysis was performed. Results: The ACEA values were significantly higher in FTDST, with a mean of 26.44°± 9.83° compared with 16.81° ± 7.72° in the control group (P < 0.001). Multivariate regression analysis also showed that higher ACEA values were associated with a FTDST (odds ratio 1.16 per degree, P = 0.01). Meanwhile, for GTA values, a statistically significant difference was found with a mean of 70.92° ± 6.64 compared with 67.84° ± 5.56 in the control group (P = 0.02). However, Stepwise regression analysis rejected GTA as a predictor for FTDST. Conclusions: Our study demonstrated that the presence of higher ACEA values is an independent significant risk factor for the presence of FTDST. Consequently, GTA values may be less helpful in assessing the risk of FTDST, especially in this specific population. Background Rotator cuff tear (RCT) is a common cause of chronic shoulder pain and disability (1). Extrinsic causes for RCT are usually associated with subacromial impingement (2), defined by the supraspinatus (SSP) tendon becoming entrapped between the acromion process and the greater tuberosity. As concern about SSP pathology increased, many previous studies focused on excessive lateral acromial coverage and confirmed that it is associated with a higher incidence of RCT (3-5). Regardless, the complex interlinkage of these acromial morphology parameters with RCT is still under extensive exploration. Recently, Singleton et al. (6) introduced the acromiohumeral centre edge angle (ACEA), a new measurement to be used with true AP shoulder radiography. The lateral projection of the acromion coverage humeral head, as a valid measurement with good reproducibility, has an accuracy comparable to measurement using a computed tomography scan. The results showed that the ACEA value was significantly higher in patients with RCT. However, because the Singleton et al. study only included acute traumatic RCT, the usefulness of ACEA in predicting degenerative RCT is still questionable. Furthermore, the subacromial impingement process necessarily comprises two sides of the bony structure, making evaluation of both bony sides, the acromion, and the greater tuberosity (GT) equally important. GT morphology still plays an important role in subacromial impingement. Some studies report that a displaced malunion GT was related to the worst outcome (7) and that the tuberosity procedure may provide a satisfactory result for irreparable massive rotator cuff tear (8). Additionally, the greater tuberosity angle (GTA), a new radiographic marker that evaluates the GT’s position proposed by Cunningham et al., has been advocated as a reliable predictor marker for RCT. They suggested that the development of degenerative RCT (9) was associated with a higher GTA values. However, their study group included both participants with partial-thickness RCT and participants with full-thickness RCT. According to a recent study (10), some acromion parameters are associated with full-thickness but not with partial-thickness RCT. For this Page 2/16 Page 2/16 reason, the specific full-thickness degenerative supraspinatus tear (FTDST) subgroup correlation with these parameters should be determined. Hence, no previous study has considered any of these values, including ACEA and GTA, as a specific tool to be used with patients with FTDST. Moreover, the literature contains limited information about these parameters, especially those pertaining to a Southeast Asian population. Sample A retrospective study was conducted in Ramathibodi Hospital, Mahidol University, Thailand. The medical records of all patients who had undergone shoulder arthroscopic surgery between April 2016 and July 2018 were collected. All pre-operative true AP shoulder radiographs were analysed. The patients selected were divided into two groups. The case group consisted of individuals presenting with a clinical diagnosis of FTDST, which was confirmed by history, preoperative-magnetic resonance imaging (MRI) scans, and intraoperative arthroscopic findings. The control group consisted of those with shoulder pain without any finding of RCT based on history; physical examination; preoperative-MRI scans; and arthroscopic findings, such as labral injury, shoulder instability, and primary adhesive capsulitis. Patient demographic data included age, gender, bicep pathology, fatty degeneration of the SSP by Goutallier staging (11), and SSP tear retraction in the frontal plane grading by Patte classification (12). Inclusion criteria&exclusion criteria The inclusion criteria were that the patients had a pre-operative true AP shoulder radiograph with the proximal humerus in an acceptable rotation of the affected shoulder. The exclusion criteria included partial-thickness RCT, any history of traumatic event to exclude any potential traumatic aetiology, previous surgery, fractures, and/or dislocation around the shoulder, congenital shoulder deformity, shoulder tumours, or infection. Patients with any evidence of osteoarthritis change in the glenohumeral joint were also excluded due to the possibility of producing outliers of these parameters. Patients meeting the inclusion and exclusion criteria were recruited. Therefore, the final sample comprised 116 patients who were enrolled into the study. The case group included 84 patients, and the control group included 32 patients. The studied population was Thai and exclusively Southeast Asian. This study was ethically approved by our hospital’s institutional research board committee (IRB number MURA2018/837). All methods in the study were carried out in accordance with the Helsinki guidelines and relevant CIOMS guidelines. Background Certain authors have attempted to evaluate the correlation of both parameters to the incidence of FTDST and analyse any association of both parameters with patients’ demographic data and arthroscopic findings. The primary objectives of our study were (1) to evaluate the presence of significant differences regarding the ACEA and the GTA values among patients with or without FTDST and (2) to assess the association between any of these parameters with other variables, such as age, sex, and SSP tear retraction. The hypothesis was that higher values for ACEA and GTA would be associated with a higher likelihood of detecting FTDST. Data collection & outcome measurement ACEA and GTA measurements were determined on true AP shoulder radiographs and described, as shown in Fig. 1. The ACEA is defined as the angle between a line drawn superiorly from the centre of the humeral head parallel to the glenoid and a line from the centre of the humeral head to the acromion’s outer edge (Fig. 1B) (6). The GTA represents the angle between a line parallel to the humerus diaphysis passing through the centre of rotation of the Page 3/16 Page 3/16 humeral head and a line connecting the upper edge of the humeral head to the most superolateral edge of the GT (Fig. 1C) (9). All measurements were performed by two independent assessors: an orthopaedic surgeon specializing in the shoulder and a radiologist specializing in musculoskeletal imaging) using a goniometer tool in the Picture Archiving and Communication System (PACS). Both were blinded from intra-operative findings. Then a repeated measurement was performed with one-month interval. Inter-observer and intra-observer reproducibility was determined. humeral head and a line connecting the upper edge of the humeral head to the most superolateral edge of the GT (Fig. 1C) (9). All measurements were performed by two independent assessors: an orthopaedic surgeon specializing in the shoulder and a radiologist specializing in musculoskeletal imaging) using a goniometer tool in the Picture Archiving and Communication System (PACS). Both were blinded from intra-operative findings. Then a repeated measurement was performed with one-month interval. Inter-observer and intra-observer reproducibility was determined. In an effort to avoid the effect of the rotation of the proximal humerus, the parameters were measured on the true AP shoulder radiograph. Each patient was placed in a supine position with the arm adducted to the side and in a neutral position. Rotation in the axial plane was accepted up to ± 20⁰. This protocol was due to unchanged ACEA and GTA parameters in this rotation range according to previous studies (6, 9). In an effort to avoid the effect of the rotation of the proximal humerus, the parameters were measured on the true AP shoulder radiograph. Each patient was placed in a supine position with the arm adducted to the side and in a neutral position. Rotation in the axial plane was accepted up to ± 20⁰. This protocol was due to unchanged ACEA and GTA parameters in this rotation range according to previous studies (6, 9). Statistical analysis Statistical analyses were calculated using Stata 16 software (StataCorp, College Station, TX, USA). The difference in the ACEA and GTA of the case and control groups was determined using independent t-tests. Correlations between parameters and age were assessed with the Pearson correlation coefficient method. Receiver operating characteristic (ROC) analysis curves were devised to determine the diagnostic ability of these parameters. Multivariate adjusted analysis by logistic regression analysis was used to determine each factor for the occurrence of FTDST with respective odds ratios. A Bonferroni post-hoc test was performed to evaluate if there was any significant difference among the subgroup analysis based on the SSP’s tear retraction grading. The limits of agreement between two assessors were examined with Bland-Altman plot analysis. The intra-rater reliability and inter-rater reliability were assessed using the intraclass correlation coefficient (ICC). The ICC was interpreted as follows: 0 to 0.40, poor; 0.41 to 0.60, moderate; 0.61 to 0.80, good; and 0.81 to 1.00, excellent (13). Patient demographic data The total number of recruited samples was 116 shoulders. The case group included 84 shoulders representing shoulders with FTDST, while the control group (without evidence of SSP tear) included 32 shoulders. The mean age was 64.19 ± 7.67 years (range, 46–81 years) for the case group and 35.81 ± 14.13 years (range, 15–65 years) for the control group. In every shoulder, we identified the presence of SSP tear, SSP tear retraction, and biceps pathology. In our series, 61.9% had biceps pathology in the case group compared to 12.5% in the control group. There were 42 patients with SSP retraction grade 1 (50%), 33 patients with SSP retraction grade 2 (39%), and 7 patients with SSP retraction grade 3 (8%). All patient characteristics are listed in Table 1. TABLE 1: Patient demographic characteristics Page 4/16 Demographic characteristics Study population Case group (n = 84) Control group (n = 32) Age (years)∞ Mean ± SD 64.19 ± 7.67 35.81 ± 14.13 Genderµ Male 33 (39.3%) 23 (71.9%) Female 51 (60.7%) 9 (28.1%) Bicep pathologyµ 52 (61.9%) 4 (12.5%) Goutallier(11) classificationµ Grade0 20 (23.8%) 32 (100%) Grade1 51 (60.7%) 0 (0%) Grade2 12 (14.3%) 0 (0%) Grade3 1 (1.19%) 0 (0%) Patte(12) classificationµ Normal 0 (0%) 32 (100%) Grade1 42 (50%) 0 (0%) Grade2 33 (39.3%) 0 (0%) Grade3 7 (8.3%) 0 (0%) ∞: value presented as mean ± standard deviation µ: value presented as the number of volunteers with that condition (percentage) Radiographic interpretation The ACEA and GTA were accessible in 116 shoulders. The average ACEA value was 23.79° ± 10.22°, and the average GTA value was 70.07° ± 6.49°. Comparing both parameters with the presence of FTDST, we found that both angles had a statistically significant association with the presence of SSP tear. The means of the ACEA and GTA variables in the case group were 26.44° ± 9.83° (95% CI, 24.31°–28.57°) and 70.92° ± 6.64 (95% CI, 69.48°–72.36°), respectively. In the control group, the means of the ACEA and GTA variables were 16.81° ± 7.72° (95% CI, 14.03°– 19.60°) and 67.84° ± 5.56 (95% CI, 65.84°–69.85°) respectively. Statistically significant associations between both variables and FTDST were found as shown in Table 2. TABLE 2: Comparison of the parameters between patients with or without full-thickness degenerative supraspinatus tear (FTDST). Page 5/16 Page 5/16 Page 5/16 Parameters Group Statistics N Mean (°) SD (°) p-value ACEA Case group 84 26.44 9.83 < 0.001** Control group 32 16.81 7.72 GTA Case group 84 70.92 6.64 0.02* Control group 32 67.84 5.56 Significant at level 0.05 Significant at level 0.01 Significant at level 0.05 For the assessment of any correlation between age and these parameters among total populations, a small positive strength of association (coefficient < 0.3) was found between age and both parameters. However, the results showed no statistical differences between these parameters within each group, as seen in Table 3. TABLE 3: Correlation analysis between parameters and age in the total population and in each of both groups. ABLE 3: Correlation analysis between parameters and age in the total population an Parameters Total (n = 116) Case group (n = 84) Control group (n = 32) Pearson correlation p-value Pearson correlation p- value Pearson correlation p-value ACEA 0.29 0.001 -0.22 0.045 0.18 0.324 GTA 0.25 0.006 0.27 0.014 -0.05 0.772 Significant at level 0.01 Comparing both parameters in correlation with patient gender, Table 4 showed that female gender had a statistically significant higher ACEA value compared to male gender in total populations. Meanwhile, no statistically significant difference in correlation with patient gender was found regarding GTA value. Comparing both parameters in correlation with patient gender, Table 4 showed that female gender had a statistically significant higher ACEA value compared to male gender in total populations. Meanwhile, no statistically significant difference in correlation with patient gender was found regarding GTA value. Radiographic interpretation TABLE 4: Comparison of parameters by gender among the total population and in each of both groups. TABLE 4: Comparison of parameters by gender among the total population and in each of both groups. Page 6/16 Page 6/16 Parameters Total (n = 116) Case group (n = 84) Control group (n = 32) N Mean (°) SD (°) p-value N Mean (°) SD (°) p- value N Mean (°) SD (°) p- value ACEA Female 60 25.83 10.78 0.025 51 27.24 10.69 0.36 9 17.89 7.69 0.63 Male 56 21.59 9.18 33 25.21 8.33 23 16.39 7.87 GTA Female 60 70.88 7.31 0.16 51 71.08 7.6 0.783 9 69.78 5.59 0.224 Male 56 69.19 5.4 33 70.67 4.89 23 67.09 5.49 Si ifi t t l l 0 01 The ROC curves were designed to evaluate the ability of both angles to predict FTDST. The curves showed that an ACEA was a good predictor for FTDST with an area under curve as 0.78. For a GTA value, the area under the curve was 0.67, which is interpreted as a fair predictor for FTDST (Figure 2). To determine the cut point, an ACEA value of 18° was a good predictor for full-thickness SSP tear, which had 85% sensitivity and 50% specificity, while a GTA value of 68° had 77% sensitivity and 44% specificity. This finding indicated that the ACEA value is a more accurate diagnostic test than is the GTA value. The differences in the cut-off values of ACEA and GTA are reported in Table 5. TABLE 5: Different ACEA and GTA cut-off values (PPV: positive predictive value, NPV: negative predictive value). Discussion RCT is one of the most common causes of chronic shoulder pain, leading to decreased functionality, declined quality of life and escalated utilization of health care resources (14). Due to its cost effectiveness and accessibility, the standard shoulder series is typically the first line of investigation for the patients suspected RCT (15) to provide additional information and needs of advanced images as MRI of the shoulder (16). This study aims to correlate between the radiographic parameters, as ACEA and GTA, from the standard shoulder radiographs and FTDST in those patients who presented with shoulder pain underwent arthroscopic surgery. Our results showed that the mean ACEA parameter was 26.44° ± 9.83° in the case group compared to 16.81° ± 7.72° in the control group, with a statistically significant difference between the groups. Female gender had a statistically significant higher ACEA value when comparing to male gender in the total population. Pearson correlation was used to identify whether any correlation was present between age and ACEA among this study’s total population. We found a negligible correlation between ACEA value and age (17). According to logistic regression analysis, an increased ACEA has an independent risk of FTDST, with an odds ratio of 1.13 per degree. In addition to higher ACEA values, the analysis manifested that increased age was a risk factor in FTDST. Our results regarding patient age aligned with those of prior studies, which revealed that the prevalence of RCT positively correlated with patient age (18). The results of in the present study with FTDST are comparable with those in the study by Singleton et al., which reported a positive correlation between ACEA and acute traumatic RCT (23.9 vs. 16.6, p < 0.001) (6). However, they did not consider RCT size and mentioned that implementation of ACEA parameter on the chronic degenerative tear is still doubtful. With our results, we certified that ACEA could be generalized to a FTDST population and used as a reliable measurement tool for detecting FTDST on standard shoulder plain radiographs. The association between higher ACEA values and rotator cuff pathology could be clarified in the same way that it has been for other parameters, such as the Acromial index (AI) and critical shoulder angle (CSA). Radiographic interpretation Interobserver reproducibility between both assessors was excellent for corresponding to previous reports (6, 9). The Bland–Altman plot of the mean difference between the repeated measurements is shown in Figure 4. TABLE 7: Summary of intra-rater and inter-rater reliability of ACEA, GTA (LOA: limits of agreement, ICC: intraclass correlation coefficient). TABLE 7: Summary of intra-rater and inter-rater reliability of ACEA, GTA (LOA: limits of agreement, ICC: intraclass correlation coefficient). Parameters Intra-observer reliability Inter-observer reliability Mean ± SD (°) 95% LOA (°) ICC (%) Mean ± SD (°) 95% LOA (°) ICC (%) ACEA -0.94 ± 3.2 -7.16 to 5.27 95 -0.27 ± 4.37 -8.84 to 8.283 91 GTA 0.60 ± 2.0 -3.34 to 4.53 94 -0.77 ± 2.75 -6.15 to 4.61 89 Radiographic interpretation ABLE 5: Different ACEA and GTA cut-off values (PPV: positive predictive value, NPV Page 7/16 Page 7/16 Cut-off value Sensitivity (%) Specificity (%) PPV (%) NPV (%) Accuracy (%) Cut-off value Sensitivity (%) Specificity (%) PPV (%) NPV (%) Accuracy (%) ACEA 15° 90.5 31.3 77.6 55.6 74 16° 86.9 40.6 79.3 54.2 74 17° 84.5 40.6 78.9 50.0 72 18° 84.5 50.0 81.6 55.2 75 19° 79.8 59.4 83.8 52.8 74 20° 77.4 59.4 83.3 50.0 72 21° 73.8 65.6 84.9 28.8 72 GTA 65° 83.3 28.1 75.3 39.1 68 66° 81.0 31.3 75.6 38.5 68 67° 79.8 37.5 77.0 41.4 68 68° 77.4 43.8 78.3 42.4 67 69° 73.8 43.8 77.5 38.9 66 70° 67.9 59.4 81.4 41.3 66 71° 60.7 71.9 85.0 41.1 64 The multivariate analysis showed that the ACEA parameter was the only parameter that was found to be statistically significant. A higher ACEA value indicated an increased risk of FTDST with an odd ratio of 1.16 per degree (P = 0.01). However, there was no statistical significance for the GTA parameter (P = 0.10) (Table 6). The risk factor for FTDST was increased age (odd ratio of 1.26 per year; p < 0.001). Our findings also showed that while the mean ACEA and GTA values of the FTDST group were larger than those of the control group, the means of the parameters among subgroups categorized by Patte classification did not show a significant difference. A comparison of both parameters is shown in Figure 3. TABLE 6: Multivariate analysis by logistic regression analysis for each factor associated with the presence of full- thickness degenerative supraspinatus tear (FTDST). TABLE 6: Multivariate analysis by logistic regression analysis for each factor associated with the presence of full- thickness degenerative supraspinatus tear (FTDST). TABLE 6: Multivariate analysis by logistic regression analysis for each factor associated with the presence of full- thickness degenerative supraspinatus tear (FTDST). Factor Odd ratio 95% Confidence interval p-value ACEA, per degree 1.16 1.04 -1.3 0.01 GTA, per degree 1.13 0.98 -1.32 0.10 Age, per year 1.26 1.13 -1.43 < 0.001** Gender, female to male 0.92 0.17 -5.01 0.93 Reliability testing for the ACEA and GTA values showed that the mean ACEA difference and the mean GTA difference were -0.94 ± 3.2 and 0.60 ± 2.0, respectively. The ICC values for the ACEA and GTA measurements were 0.95 and 0.94, respectively (Table 7). Discussion This was explained by the vertical force vector of the middle fibre of the deltoid muscle, where the pull, influenced by the lateral extension of the acromion, was directed upward, lead plausibly to SSP impingement and consequence tear due to a compression effect causing a degeneration tear of the rotator cuff, as stated in previous literature (19, 20). Page 9/16 Despite the high variability of GT morphology (21), the GTA parameter was first proposed in 2018 for a new reliable radiographic marker of degenerative RCT. However, the precise relationship between GTA's high values and the incidence of RCT is not yet understood. In the present study, GTA was found significantly between groups with a small size difference; the mean in the case group was 70.92° ± 6.64, compared to 67.84° ± 5.56 in the control group. However, multivariable analysis showed no statistically significant correlation between GTA and FTDST. Hence, stepwise multiple regression analysis rejected GTA as a predictor for FTDST. The result of the present study contrast with those reported by Cunningham et al., who noted that the mean GTA parameter was 72.5° ± 2.5° in the RCT group compared to 65.2° ± 4.1° in the control group and concluded that degenerative RCT in the European population was associated with GTA values of more than 70°(9). In addition, Yoo et al. reported that high GTA values accompanied RCT in the Korean population, based on MRI results (22). This discrepancy between results could be explained by the population-based anatomic variation that may exist in a Southeast Asian population, such as the Thai population. A recent study has documented that the Asian population exhibits a smaller humeral head compared to the Western population (23); these findings could possibly explain why the GTA effect is larger in Asians than in Europeans (22). Moreover, previous studies examined the GTA effect on individuals with partial- or full-thickness supraspinatus tears, but our study examined the GTA effect only on FTDST; partial tears were not taken into consideration. Recent studies (5, 10) demonstrated a dissimilarity association between acromial parameter and type of RCT tear (full thickness vs. partial thickness). Additionally, Seo et al. (24) reported that the mean GTA values for bursal-side partial RCT tended to be larger than those for full-thickness RCT. Consent for publication Not applicable. Discussion Although the findings were statistically significant, careful interpretation is essential to determining the clinical application of these parameters 9.6° and 3.1°, respectively; these results were within their SD range. Although the findings were statistically significant, careful interpretation is essential to determining the clinical application of these parameters List Of Abbreviations FTDST : full-thickness degenerative supraspinatus tear Ethics approval and consent to participate This retrospective study received the written approval by the Institutional Review Boards in Mahidol University (certificate of approval no. MURA2018/837). The Institutional Review Boards in Mahidol University waived the need to obtain consent for the collection, analysis and publication of the retrospectively obtained and anonymized data for this non-interventional study. Conclusion The presence of higher ACEA values is a significant independent risk factor for the presence of full-thickness, degenerative SSP tear. The measurement of ACEA could be a useful tool to determine additional advanced images such as MRI to confirm diagnosis of this condition in a patient with chronic shoulder pain. Consequently, GTA values may be less helpful in assessing the risk of full-thickness, degenerative SSP tear, especially in the Southeast Asian population. Discussion Therefore, the different populations’ characteristics could affect the magnitude of the parameter and may have produced different findings among studies as well This study was the first to evaluate these angles based on arthroscopic findings for their usefulness in determining the risk of FTDST. The most important finding was that ACEA was shown to have a statistically significant, high association with FTDST. Stepwise logistic regression analysis demonstrated that between ACEA and GTA, only ACEA appears to be valid in correlation with FTDST: GTA was rejected as a factor in predicting FTDST. No previous study has performed logistic regression to assess the effects of these parameters on FTDST. We summarized that ACEA could be used as an independent factor to assess the risk of FTDST in our study. As such, we also concluded that GTA could not be utilized to assess the risk of FTDST in the Southeast Asian population, particularly in the Thai population. Nevertheless, a positive correlation was not found between a higher value and SSP tear retraction grading. Thus, these parameters cannot be used to differentiate the severity of tear retraction of SSP. Although the possibility of type II errors should be considered, further biomechanics study is essential for specific assessment of the influence of these parameters on tear retraction. The present study had several limitations. First, due to limitation of retrospective study nature and arthroscopic based study design, some patient’s demographic data might have a difference between the groups. FTDST patients tend to be females who are older than patients with labral injury and shoulder instability, even adhesive capsulitis, which is mainly included in the control group. Despite this variation, the statistically significant difference still endured for the ACEA parameter after multivariable analysis. Second, our study has relatively small sample size compared to previous studies. However, a statistical power analysis was performed for sample size estimation, with alpha = .05 and power = 0.80. The sample size needed with this effect size was approximately 25 samples within each group. These were sufficient for detecting a difference of 5° between groups if the standard deviation of each group was 7.7., despite the fact that our proposed sample size was more than adequate for this study’s primary objective. Third, the difference of ACEA and GTA between the patients with and without FTDST was approximately Page 10/16 9.6° and 3.1°, respectively; these results were within their SD range. Availability of data and material The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. Page 11/16 Funding No external funding was obtained for this study. References 1. Yamamoto A, Takagishi K, Osawa T, Yanagawa T, Nakajima D, Shitara H, et al. Prevalence and risk factors of a rotator cuff tear in the general population. J Shoulder Elbow Surg. 2010;19(1):116-20. 1. Yamamoto A, Takagishi K, Osawa T, Yanagawa T, Nakajima D, Shitara H, et al. Prevalence and risk factors of a rotator cuff tear in the general population. J Shoulder Elbow Surg. 2010;19(1):116-20. 1. Yamamoto A, Takagishi K, Osawa T, Yanagawa T, Nakajima D, Shitara H, et al. Prevalence and risk factors of a rotator cuff tear in the general population. J Shoulder Elbow Surg. 2010;19(1):116-20. 2. Oh JH, Kim JY, Lee HK, Choi JA. Classification and clinical significance of acromial spur in rotator cuff tear: heel-type spur and rotator cuff tear. Clin Orthop Relat Res. 2010;468(6):1542-50. 2. Oh JH, Kim JY, Lee HK, Choi JA. Classification and clinical significance of acromial spur in rotator cuff tear: heel-type spur and rotator cuff tear. Clin Orthop Relat Res. 2010;468(6):1542-50. 3. Balke M, Schmidt C, Dedy N, Banerjee M, Bouillon B, Liem D. Correlation of acromial morphology with impingement syndrome and rotator cuff tears. Acta orthopaedica. 2013;84(2):178-83. 3. Balke M, Schmidt C, Dedy N, Banerjee M, Bouillon B, Liem D. Correlation of acromial morphology with impingement syndrome and rotator cuff tears. Acta orthopaedica. 2013;84(2):178-83. 4. Hamid N, Omid R, Yamaguchi K, Steger-May K, Stobbs G, Keener JD. Relationship of radiographic acromial characteristics and rotator cuff disease: a prospective investigation of clinical, radiographic, and sonographic findings. J Shoulder Elbow Surg. 2012;21(10):1289-98. 5. Kim JR, Ryu KJ, Hong IT, Kim BK, Kim JH. Can a high acromion index predict rotator cuff tears? Int Orthop. 2012;36(5):1019-24. 6. Singleton N, Agius L, Andrews S. The acromiohumeral centre edge angle: A new radiographic measurement and its association with rotator cuff pathology. Journal of Orthopaedic Surgery. 2017;25:230949901772795. 7. Rouleau DM, Laflamme GY, Mutch J. Fractures of the greater tuberosity of the humerus: a study of associated rotator cuff injury and atrophy. Shoulder Elbow. 2016;8(4):242-9. 7. Rouleau DM, Laflamme GY, Mutch J. Fractures of the greater tuberosity of the humerus: a study of associated rotator cuff injury and atrophy. Shoulder Elbow. 2016;8(4):242-9. 8. Park JG, Cho NS, Song JH, Baek JH, Rhee YG. Long-term outcome of tuberoplasty for irreparable massive rotator cuff tears: is tuberoplasty really applicable? J Shoulder Elbow Surg. 2016;25(2):224-31. 8. Park JG, Cho NS, Song JH, Baek JH, Rhee YG. Acknowledgements The authors would like to thank Department of Orthopedics, Faculty of Medicine Ramathibodi Hospital, Mahidol University, and Department of Radiology, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand for all of the kindly help and permission to carry out the study. Authors’ contributions CV and CK conceptualized and designed the study. CV and PF performed radiographic measurement and helped in the acquisition of data. CV and CK were responsible for analysis and interpretation of data. CV and CK helped to draft the manuscript. SJ and CK helped to revise the manuscript critically for important intellectual content. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. References Long-term outcome of tuberoplasty for irreparable massive rotator cuff tears: is tuberoplasty really applicable? J Shoulder Elbow Surg. 2016;25(2):224-31. 9. Cunningham G, Nicodeme-Paulin E, Smith MM, Holzer N, Cass B, Young AA. The greater tuberosity angle: a new predictor for rotator cuff tear. J Shoulder Elbow Surg. 2018;27(8):1415-21. 9. Cunningham G, Nicodeme-Paulin E, Smith MM, Holzer N, Cass B, Young AA. The greater tuberosity angle: a new predictor for rotator cuff tear. J Shoulder Elbow Surg. 2018;27(8):1415-21. 10. Pandey V, Vijayan D, Tapashetti S, Agarwal L, Kamath A, Acharya K, et al. Does scapular morphology affect the integrity of the rotator cuff? J Shoulder Elbow Surg. 2016;25(3):413-21. 10. Pandey V, Vijayan D, Tapashetti S, Agarwal L, Kamath A, Acharya K, et al. Does scapular morphology affect the integrity of the rotator cuff? J Shoulder Elbow Surg. 2016;25(3):413-21. Page 12/16 Page 12/16 11. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC. Fatty infiltration of disrupted rotator cuff muscles. Rev Rhum Engl Ed. 1995;62(6):415-22. 12. PATTE D. Classification of Rotator Cuff Lesions. Clinical Orthopaedics and Rela 13. Koo TK, Li MY. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J Chiropr Med. 2016;15(2):155-63. 14. Jeanfavre M, Husted S, Leff G. EXERCISE THERAPY IN THE NON-OPERATIVE TREATMENT OF FULL- THICKNESS ROTATOR CUFF TEARS: A SYSTEMATIC REVIEW. Int J Sports Phys Ther. 2018;13(3):335-78. 15. Hussain A, Muzzammil M, Butt F, Valsamis EM, Dwyer AJ. Effectiveness Of Plain Shoulder Radiograph In Detecting Degenerate Rotator Cuff Tears. J Ayub Med Coll Abbottabad. 2018;30(1):8-11. 16. Nazarian LN, Jacobson JA, Benson CB, Bancroft LW, Bedi A, McShane JM, et al. Imaging algorithms for evaluating suspected rotator cuff disease: Society of Radiologists in Ultrasound consensus conference statement. Radiology. 2013;267(2):589-95. 17. Mukaka MM. Statistics corner: A guide to appropriate use of correlation coefficient in medical research. Malawi Med J. 2012;24(3):69-71. 18. Yamamoto A, Takagishi K, Osawa T, Yanagawa T, Nakajima D, Shitara H, et al. Prevalence and risk factors of a rotator cuff tear in the general population. Journal of Shoulder and Elbow Surgery. 2010;19(1):116-20. 19. Gerber C, Catanzaro S, Betz M, Ernstbrunner L. Arthroscopic Correction of the Critical Shoulder Angle Through Lateral Acromioplasty: A Safe Adjunct to Rotator Cuff Repair. Arthroscopy. 2018;34(3):771-80. 20. Nyffeler RW, Werner CM, Sukthankar A, Schmid MR, Gerber C. Association of a large lateral extension of the acromion with rotator cuff tears. J Bone Joint Surg Am. 2006;88(4):800-5. 21. References Syed UAM, Davis DE, Ko JW, Lee BK, Huttman D, Seidl A, et al. Quantitative Anatomical Differences in the Shoulder. Orthopedics. 2017;40(3):155-60. 22. Yoo JS, Heo K, Yang JH, Seo JB. Greater tuberosity angle and critical shoulder angle according to the delamination patterns of rotator cuff tear. J Orthop. 2019;16(5):354-8. 22. Yoo JS, Heo K, Yang JH, Seo JB. Greater tuberosity angle and critical shoulder angle according to the delamination patterns of rotator cuff tear. J Orthop. 2019;16(5):354-8. 23. Cabezas AF, Krebes K, Hussey MM, Santoni BG, Kim HS, Frankle MA, et al. Morphologic Variability of the Shoulder between the Populations of North American and East Asian. Clin Orthop Surg. 2016;8(3):280-7. 23. Cabezas AF, Krebes K, Hussey MM, Santoni BG, Kim HS, Frankle MA, et al. Morphologic Variability of the Shoulder between the Populations of North American and East Asian. Clin Orthop Surg. 2016;8(3):280-7. 24. Seo J, Heo K, Kwon S, Yoo J. Critical shoulder angle and greater tuberosity angle according to the partial thickness rotator cuff tear patterns. Orthop Traumatol Surg Res. 2019;105(8):1543-8. 24. Seo J, Heo K, Kwon S, Yoo J. Critical shoulder angle and greater tuberosity angle according to the partial thickness rotator cuff tear patterns. Orthop Traumatol Surg Res. 2019;105(8):1543-8. Figures Page 13/16 Figure 1 On a true AP glenohumeral X-ray, (A) the humeral head is circled and marked centre of rotation. (B) Measuring the ACEA formed with the first drawn superiorly from the centre of the humerus parallel to the glenoid surface and the second drawn from the centre of the circle to the lateral-most aspect of the acromion edge. (C) Measuring the GTA angle formed by the first drawn parallel to the diaphyseal axis that passes through the humeral head centre of rotation; the second drawn connects the superior border of the humeral head to the superolateral edge of the greate tuberosity. Figure 1 On a true AP glenohumeral X-ray, (A) the humeral head is circled and marked centre of rotation. (B) Measuring the ACEA formed with the first drawn superiorly from the centre of the humerus parallel to the glenoid surface and the second drawn from the centre of the circle to the lateral-most aspect of the acromion edge. (C) Measuring the GTA angle formed by the first drawn parallel to the diaphyseal axis that passes through the humeral head centre of rotation; the second drawn connects the superior border of the humeral head to the superolateral edge of the greater tuberosity. Figure 2 (A) ACEA, (B) GTA. Receiver operating characteristic (ROC) curve, with an area under the ROC curve of 0.78 and 0.67, respectively. Figure 2 (A) ACEA, (B) GTA. Receiver operating characteristic (ROC) curve, with an area under the ROC curve of 0.78 and 0.67, respectively. Page 14/16 Page 14/16 Figure 3 Comparison of parameter (A) ACEA and (B) GTA between the groups graded by Patte classification. Error bars indicate the inter-quartile range (IQR) of the median. Black dots indicate values above the upper fence (1.5IQR). above the lines spanning between groups indicated P-values < 0.05. Grade 0 indicates no rotator cuff tear conditio Figure 3 Comparison of parameter (A) ACEA and (B) GTA between the groups graded by Patte classification. Error bars indicate the inter-quartile range (IQR) of the median. Black dots indicate values above the upper fence (1.5IQR). above the lines spanning between groups indicated P-values < 0.05. Grade 0 indicates no rotator cuff tear condition. Comparison of parameter (A) ACEA and (B) GTA between the groups graded by Patte classification. Error bars indicate the inter-quartile range (IQR) of the median. Black dots indicate values above the upper fence (1.5IQR). above the lines spanning between groups indicated P-values < 0.05. Grade 0 indicates no rotator cuff tear condition. Page 15/16 Figure 4 Figure 4 Page 15/16 Bland-Altman plot of the mean difference in ACEA measurements between one-month separate time points (A). The mean difference in ACEA measurements between two assessors (B). The mean difference between GTA measurements taken at one-month separate time points (C). The mean difference between GTA measurements of two assessors (D). Altman plot of the mean difference in ACEA measurements between one-month separate time points (A). The difference in ACEA measurements between two assessors (B). The mean difference between GTA rements taken at one-month separate time points (C). The mean difference between GTA measurements of sessors (D). Bland-Altman plot of the mean difference in ACEA measurements between one-month separate time points (A). The mean difference in ACEA measurements between two assessors (B). The mean difference between GTA measurements taken at one-month separate time points (C). The mean difference between GTA measurements of two assessors (D). Page 16/16
https://openalex.org/W2974918663	https://europepmc.org/articles/pmc6755627?pdf=render	English	null	Using a Studio-Academic Partnership to Advance Public Health Within a Pragmatic Yoga Setting	Journal of primary care & community health	2,019	cc-by	7,375	Original Research Original Research https://doi.org/10.1177/2150132719874621 Journal of Primary Care & Community Health Volume 10: 1–9 © The Author(s) 2019 DOI: 10.1177/2150132719874621 journals.sagepub.com/home/jpc https://doi.org/10.1177/2150132719874621 Journal of Primary Care & Community Health Volume 10: 1–9 © The Author(s) 2019 DOI: 10.1177/2150132719874621 journals.sagepub.com/home/jpc Keywords cohesion, community-based, participatory approaches, translation the past year.6 Second, because most yoga studies are epi- demiological, cross-sectional, or based on efficacy trials with explanatory participant samples,3,4,7-9 there is low generalizability and pragmaticism—that is, the ability of the evidence-base of yoga to translate into the real world.10 Third, many modern versions of yoga are not captured in yoga research, which has focused on clinical populations in controlled settings performing traditional styles of yoga such as hatha, Iyengar, and ashtanga.11 This may limit the match of the evidence-base for health-enhancing yoga ben- efits and real-world practice. Finally, there is a call for more context cognizant scholarship, moving beyond proof of concept yoga interventions to understanding real world Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). Abstract Objectives: To explore community-based yoga studio practitioners’ psychosocial variables, behaviors, and studio satisfaction. Methods: Concurrent mixed-methods study consisted of a survey for demographic variables and psychosocial variables of interest (e.g., mindfulness, self-compassion, physical activity participation) and interviews regarding reasons for participating at the yoga studio. Results: Participants (N = 138) were, on average, 35.58 ± 14.09 years old and predominantly female (91.3%), married (40.6%) or single (37%), Caucasian (75%), and college (25.4%) or graduate/medical school (45%) educated, with 54% meeting physical activity recommendations. On a 5-point Likert-type scale, participants reported being moderately cohesive (Msumscore = 3.87 ± 0.62), stressed (Msumscore = 3.2 ± 0.39), mindful (Msumscore = 3.4 ± 0.41), and self-compassionate (Msumscore = 3.26 ± 0.56). A rapid content analysis of interviews (n = 18), indicated that participants primarily practiced at the studio for the sense of community. Conclusions: Yoga practitioners reported positive perceptions and behaviors; however, opportunities remain for interventions to improve mental and physical health among individuals already attending a yoga studio. Through an academic-studio partnership, studio offerings may include low-dose evidence-based interventions to improve access to and uptake of a yoga practice. Using a Studio-Academic Partnership to Advance Public Health Within a Pragmatic Yoga Setting Samantha M. Harden1 , Abby M. Steketee1, Rachel Kelliher2, Keala A. Mason2, and Nicole Fitzwater Boyle2 Samantha M. Harden1 , Abby M. Steketee1, Rachel Kelliher2, Keala A. Mason2, and Nicole Fitzwater Boyle2 Introduction We need yoga more now (than ever) . . . human beings have always needed self-inquiry and presence and steadiness of mind . . . but because of the pace of our culture, the loneliness in our culture, how we live . . . this is a new phenomenon in the history of human beings . . . We have a lot of stress. We do not go out and work in the field . . . or go on a hunt . . . When people come to a yoga class, they have been very likely sitting too much, not eating well, drinking too much coffee. . . . trying to do too many things . . . When we have a yoga class, we are creating, whether we know it or not, community, states renowned yoga instructor, author, and physical ther- apist Judith Hanson Lasater.1 Within this quote, public health practitioners and researchers can identify complex, interrelated health concerns such as stress, isolation, and inactivity—all of which may be mitigated through the practice of yoga.2-4 1Virginia Tech, Blacksburg, VA, USA 2In Balance Yoga Studio, Blacksburg, VA, USA Corresponding Author: Samantha M. Harden, Physical Activity Research and Community Implementation Laboratory, Human Nutrition, Foods, and Exercise, Virginia Tech, 1981 Kraft Drive, Blacksburg, VA 24061, USA. Email: harden.samantha@vt.edu Studying the practice of yoga as a public health inter- vention2,5 is of particular interest for a number of reasons. First, 21 million Americans practiced yoga at least once in Journal of Primary Care & Community Health 2 and a balance from physical to spiritual expertise. Another example is that In Balance Yoga Studio offers annual stu- dio-centric engagement challenges. For three years, there was a 30-day challenge. Program components included salient strategies for behavior change16: (1) goal setting and friendly competition to attend 30 classes in 30 days, (2) self-monitoring on a physical board in the studio space, and (3) feedback and auditing through recognition on social media and the October e-newsletter. Approximately 70 cli- ents participated each year, for 3 years. External recognition of these efforts is seen in the studio’s receipt of the runner up for the wellness category of the “Best of the Blue Ridge” competition spanning from Maryland through Georgia along the East coast in the United States. Recruitment Researchers recruited a convenience sample of yoga class attendees in a single community-based studio in rural Southwest Virginia. Participants were recruited through flyers posted at the studio, 2 posts to each the studio’s Facebook and Instagram accounts, and announcements in the studio’s e-newsletter (N = 1400 recipients). Finally, and in alignment with the mission for cruelty-free eating, vegan food was avail- able from 9 am to 9 pm on a Sunday in October at the studio with members of the research team available to discuss the study. There are 8 classes every Sunday and there are typically 130 practitioners across a variety of class types (e.g., hot yoga, restorative, prenatal). The study was open to any individual who received the newsletter or attended yoga class at the stu- dio. Study participants were entered in a prize drawing to receive one of two $50 studio gift certificates. The Virginia Tech Institutional Review Board approved this study. At the time of the study, the community had 3 stand- alone yoga studios, and yoga classes at local gyms and through on-campus recreation. One studio of interest was the In Balance Yoga studio. The mission of the studio is to support and lead a yoga community that practices and provides a variety of yoga to all populations and is passionate about giving back to each other and our planet . . . to progress and expand our community we offer an array of studio classes, trainings, workshops, and retreats within our yoga studio and globally. The studio has a social entrepreneurship business model, which includes quarterly contributions to 3 nonprofit orga- nizations. The key values of the owner and management are “positivity, service, connection, and fun.” In alignment with this mission and values, the studio engages in a number of strategies to enhance perceptions of cohesion/inclusivity, to reach and retain more community members, and to provide an environment that facilitates a consistent yoga practice. Setting The college town of Blacksburg, Virginia is home to 42 000 residents of which 46% are female and 78% are Caucasian. The town has a large proportion of individuals who are 18 to 64 years old (82%), with only 13% and 5% of the popula- tion being younger than 18 or older than 65 years, respec- tively.15 This town is located in a southern state, a region of the nation with the lowest rates of yoga participation.6 Introduction yoga practices and adherence (including an individual’s regular practice, not adherence to an evidence-based intervention).12 Therefore, there is a prime opportunity to understand the behaviors and health factors of individuals within settings where people are naturally attending yoga classes: Their community-based yoga studio. There are over 6000 yoga studios nationwide and yoga represents a $16 billion indus- try.13 Yoga studio owners have a vested interest in under- standing how to keep clients engaged, and public health investigators are interested in matching evidence-based interventions with target audiences. All of this underscores the potential for an academic-community partnership14 to explore psychosocial variables, behaviors, and perceptions of yoga practitioners. Results identified in this exploratory study indicate areas for improvement—in terms of studio offerings and health promotion—as well as what is already working. Notably, for the purposes of this work, yoga is defined as a combination of breathing exercises (pranayama), meditation (dhyana), and postures (asana). Knowing the values and mission of the studio, members of the research team approached the studio to discuss a potential partnership to explore psychological factors, attendance, and exercise among studio participants. Notably, the studio owner declined similar partnership offers from other researchers in the past. However, the stu- dio owner perceived this research team to have buy-in and support of the studio and the desire to partner rather than simply access data.14 In addition, all members of the research team were also registered yoga instructors, 2 of whom were instructors at the partnering studio. Therefore, the owner accepted the invitation to explore these empirical and pragmatic outcomes of interest. Research Design This study used a concurrent, mixed-methods design.17 Cross- sectional surveys were provided for a 4-week period in online and paper-and-pencil formats. For the qualitative portion of the study, individuals attending the yoga studio during a research team–sponsored community event (where vegan food was offered) were randomly selected and asked if they would like Examples of these strategies include hosting approxi- mately 47 external/guest instructors or teachers per year. This provides practitioners in rural, Southwest Virginia with a diverse group of experts from different yoga lineages Harden et al 3 with kindness, understanding, and sense of shared humanity, instead of harshly judging oneself or ignoring distressing feelings. All items were on a 5-point Likert-type scale (1 = never, 5 = very often) and example items include, “When I’m feeling down, I tend to obsess and fixate on everything that’s wrong.” Self-compassion scores were obtained by reverse-coding responses that indicate an uncompassionate thought, feeling, or action. The higher the overall score, the higher the level of self-compassion. with kindness, understanding, and sense of shared humanity, instead of harshly judging oneself or ignoring distressing feelings. All items were on a 5-point Likert-type scale (1 = never, 5 = very often) and example items include, “When I’m feeling down, I tend to obsess and fixate on everything that’s wrong.” Self-compassion scores were obtained by reverse-coding responses that indicate an uncompassionate thought, feeling, or action. The higher the overall score, the higher the level of self-compassion. to participate in a video-recorded testimonial. Once written consent was obtained, the individuals who agreed to partici- pate in the study were interviewed by a researcher with one prompt: “Why do you practice at this studio?” There are several indicators that can be used as the denominator for this study: total number of individuals who have attended class in the past 3 months (N = 10 951); total number of individuals who subscribe to the newsletter (N = 1400); and total number of individuals who attended class during the community event (N = 133). Mindfulness. The 24-item Five Facet Mindfulness Ques- tionnaire: Short Form22 was used to assess (1) nonreactiv- ity to inner experience, (2) observing/noticing/attending to sensations/perceptions/thoughts/feelings, (3) acting with awareness/automatic pilot/concentration/nondistrac- tion, (4) describing/labeling with words, (5) nonjudging of experience (Baer et al., 2006).23 Each item is on a 5-point Likert-type scale (1 = never or very rarely true, 5 = very often or always true). Measures Attendance. The studio uses the MindBody business software to develop practitioner accounts and track studio attendance. With consent, practitioners’ attendance data were pulled from their accounts for the previous 12 months. These data were used to develop practice indicators of (1) total number of studio classes in the past 12 months, (2) average number of studio classes per week in the past 12 months, and (3) duration of membership. Cohesion. A modified version of the 21-item Physi- cal Activity Group Environment Questionnaire18 was employed. To be consistent with all other scales in the study, the items were converted from a 9-point Likert-type scale to a 5-point Likert-type scale. Items were modified from “I like the amount of physical activity I get with this group” to “I like the amount of physical activity I get from yoga classes here.” Studio perceptions. These items were developed in part- nership with the studio representatives. The items were related to cleanliness, trust, class offerings, and the envi- ronment. To match the scales in the rest of the survey, all items were on a 5-point Likert-type scale (1 = very strongly disagree, 5 = very strongly agree). Example items include “I feel like this studio offers services/classes that meet the needs of all age ranges” and “The environment of this stu- dio is calm and soothing.” These items, although not vali- dated, were chosen to help understand “what matters” to the studio management. Physical activity. The Stanford Leisure-Time Activity Categorical Item (L-Cat)19 is a single item survey with 6 descriptive categories ranging from inactive to very active, which was shown to have strong psychometric properties with high validity in overweight/obese women. According to the interpretation guide, if participants indicated a cat- egory 4 or higher, they were coded as meeting the physical activity guidelines for Americans (1 = yes, 0 = no). Research Design An example item is “I do jobs or tasks automatically without being aware of what I’m doing.” Fol- lowing the scoring recommendations, where appropriate, items were reverse coded, and the sum scores were used in analysis. The higher the overall score, the higher the level of mindfulness. Analytical Plan Quantitative data were descriptively summarized by pro- portions and means. Pearson correlation was used to detect associations between the psychometric variables in this study and membership status (number of classes attended), sex, teacher status, and each psychosocial variable (mind- fulness, cohesion, stress, self-compassion). One-way analy- sis of variance was conducted on years of membership and age with the key outcomes of interest. The a priori hypoth- eses were that duration of membership, number of classes attended, female sex, and teacher status would lead to greater perceptions of cohesion, stronger mindfulness and self-compassion ratings, and lower perceived stress. Perceived stress. The validated, 10-item Perceived Stress Scale (PSS)20 was used to assess stress. Each item is on a 5-point Likert-type scale (1 = never, 5 = very often) and an example item is “In the past month, how often have you felt that you were unable to control important things in your life?” PSS scores are obtained by reverse-coding responses to the 4 positively stated items (items 4, 5, 7, and 8) and subsequently summing across all scale items. The higher the overall score, the higher the perceived stress. Self-compassion. The validated, 12-item Self-Compassion Short form21 was used to assess the degree to which par- ticipants demonstrate self-compassion as defined as emo- tional regulation to appraise painful or distressing feelings Following the completion of the community event, qualita- tive data were transcribed verbatim by a trained undergradu- ate research assistant using Microsoft Word. The names of the individuals who participated in the testimonials, in addition to Journal of Primary Care & Community Health 4 Table 1. Sample Participants (n = 138) were aged 35.58 ± 14.09 years (range 19-72 years), predominantly female (91.3%), Caucasian (75%), and college (25.4%) or graduate/medi- cal school (45%) educated. In addition, many participants (29%) had been practicing yoga for 4 to 7 years, and 10% were certified instructors at the studio. The sociodemo- graphic characteristics of the sample are shown in Table 1. Thirty-one individuals were approached to provide testi- monials, and 18 provided data (58%). A majority of invited yoga practitioners declined to provide a testimonial because of not wanting to be on film or being sweaty from class (or a combination of the two). Other unique reasons for declining the invitation for a video recorded testimo- nial were that they had not been a member of the studio long enough (in their opinion), that they had suffered a recent loss in their family, or that they did not have time to provide a testimonial. Three of the testimonials (17%) were provided by instructors at the yoga studio that volun- teered and 83% (n = 15) were students at the yoga studio. Analytical Plan Summary of Sample Characteristics and Outcomes From Key Variables.a Demographic Characteristics Age, years (n = 125) 35.6 (±14) Race/Ethnicity (n = 129) Caucasian 108 (78.3) Asian 10 (7.75) African American 3 (2.2) Other 7 (5.1) Alaska Native 1 (0.7) Prefer not to answer 1 (0.7) Sex (n = 126) Male 12 (9.5) Female 114 (90.5) Education (n = 126) Grade 12 or GED 3 (2.4) College, 1-3 years 26 (20.2) College graduate 35 (27.8) Graduate/Medical school 62 (49.2) Yoga teacher status (n = 117) Not a teacher 61 (52.1) No, but would like to be a teacher someday 31 (26.5) Yes, 200 hours registered or higher 20 (17.1) Yes, not certified 5 (4.3) Distance from studio (n = 123) ≤10 minutes 74 (60.1) 11-20 minutes 37 (30.0) 21-30 minutes 3 (2.4) 31-40 minutes 2 (1.6) 41-50 minutes 3 (2.4) 51-60 minutes 1 (0.8) 240 minutes 2 (1.6) Varies 1 (0.8) Behaviors (n = 106) Meeting physical activity recommendations 75 (54) Yoga practice/Studio specific Duration of membership in years 3.08 (±2.69) Number of classes attended past 12 months 56.94 (±60.36) Class most frequently attended (n = 116) Hot flow-set 32 (27.6) Warm flow 19 (16.4) Flow 3 (9.5) Donations 7 (6.0) Basic flow 6 (5.2) Power flow 5 (4.3) Gentle yoga 5 (4.3) Yin 5 (4.3) Other/Not listed 5 (4.3) Hot Flow Mix 4 (3.4) (continued) Table 1. Summary of Sample Characteristics and Outcomes From Key Variables.a any sensitive information (name of teachers, occupation, name of studio, etc) that they provided in their responses, were excluded from the transcription. In order to protect the identity of the participants, coded labels were assigned to each participant and used throughout the transcription process. The same undergraduate assistant performed a content analysis24 on the transcribed videos to identify the meaning units. Meaning units were any word or phrase that represented a single idea. Both the transcription and meaning units were submitted to the lead author for content review. Following this review, the meaning units were categorized into themes. Any meaning that was conveyed by 8 or more participants was operationalized as a major emergent theme whereas a mean- ing unit that was mentioned by 7 or fewer participants was considered a minor emergent theme.25 Notably, studio instruc- tors and students were invited to provide testimonials, but the position of the participants at the yoga studio was disregarded in the content analysis of the transcriptions. (continued) Quantitative Of the 138 participants, 116 answered the question “What class do you attend the most (select one)?” Of the 29 class options, the top 3 most attended classes were Hot Flow-Set (23.2%), Warm Flow (13.8%), and Flow (8.0%). The least attended classes were Meditation (0.7%), Donations/Free Community (0.7%), and Pilates Sculpt (0.7%). Of the 138 participants, 116 also answered the question “What classes do you typically attend at this studio (select 5 Harden et al Demographic Characteristics Hot Flow2-Set 3 (2.6) Hot 26&2-Set 3 (2.6) Yoga Basics 2 (1.7) Prenatal yoga 2 (1.7) Aerial 2 (1.7) Restorative hammock yoga and meditation 2 (1.7) Meditation 1 (0.9) Donation/Community/Free class 1 (0.9) Pilates Sculpt 1 (0.9) Studio satisfaction (n = 116) on 5-point scale Studio satisfaction 4.53 (±0.67) Key Scale Sum Scores on 5-Point Scales Full Sample Mind-Body Approved Cohesion (n = 106) 3.87 (±0.62) 3.89 (±0.59) Perceived stress (n = 106) 3.20 (±0.39) 3.21 (±0.39) Self-compassion (n = 109) 3.27 (±0.56) 3.23 (±0.52) Mindfulness (n = 112) 3.42 (±0.41) 3.43 (±0.37) aAll data reported as means (±standard deviation) or number (percent). Table 1. (continued) Table 1. (continued) Key psychosocial variables. Four significant moderate associations were detected: Self-compassion and perceived stress had a moderate negative correlation (r = −.31, P = .000); as self-compassion increased, perceived stress decreased. Self-compassion and mindfulness had a moder- ate positive correlation (r = .61, P = .000); as self-compas- sion increased, mindfulness increased. In addition, number of days attended was significantly associated with positive perceptions of cohesion and positive self-compassion scores (P = .000). Please see Table 2 for full correlation matrix. Qualitative Across the 18 participants, the three most salient themes were: community, mental health benefits, and physical benefits. Most participants (n = 13, 72%) indicated that the community aspect was what brought them to practice at the yoga studio. Example meaning units from this theme included: “what really kept me coming was the sense of community” and “it feels like hOMe [tone inflection for OM].” In addition, 50% (n = 9) of participants indicated that the mental health benefits of yoga were what led them to practice yoga at the partnering yoga studio. Meaning units for this theme included statements such as: “getting back mentally” and “it makes me really happy.” aAll data reported as means (±standard deviation) or number (percent). all that apply)?” Of the 29 class options, the top 3 classes typically attended were Hot Flow-Set (49.3%), Warm Flow (43.5%), and Hot Flow Mix (31.9%). The 3 classes least typically attended were Restorative hammock yoga and meditation (2.2%), Fascae Freedom (0.7%), and Yoga for Seniors (0.7%). Based on a score of 4 or higher on the L-Cat, 54% (n = 75) of participants met physical activity recommendations. The third major theme was the physical benefits of yoga. Meaning units for physical benefits were provided by 50% (n = 9) of the participants. The 2 subthemes of physical benefits were yoga as a form of exercise or workout or yoga to balance out the other activities they performed. Five of the 9 (55%) participants who mentioned physical health benefits used yoga as a workout as expressed through mean- ing units such as “the asana, the practice and how my body feels,” “and what I love, especially about, [is] being fit and strong,” and “I practice here for the . . . Qualitative Positive perceptions of the studio itself were evident in the high scores for all variables of interest (the environment, teachers, etc.), and these positive percep- tions were also voiced by the individuals who shared testi- monials as to why they practiced at this location. The qualitative data contribute to the extant literature by high- lighting that, while people appreciated the mental and phys- ical benefits of yoga as well as teachers and the structure, the practitioners were most likely to attend the studio due to the sense of community. While many of these relationships were in the predicted directions, there are several empirical and pragmatic observations to consider. This study aimed to fill the gap in the literature around these 3 issues and found that yoga practitioners in this commu- nity-based studio have many positive perceptions (of yoga and the studio) as well as engage in positive health behav- iors. Overall, the quantitative data suggest that yoga studio clients who participated in this study were moderately stressed, mindful, self-compassionate, and cohesive. However, only half of the participants were meeting physi- cal activity recommendations. Self-compassion and per- ceived stress were negatively associated. This is unsurprising as self-compassion measures how individuals view them- selves during times of stress. In addition, self-compassion is a type of mindfulness in which an individual perceives his or her situation without judgment, so the measured associa- tion between self-compassion and mindfulness, although not previously explored in community settings, is in the expected direction. Positive perceptions of the studio itself were evident in the high scores for all variables of interest (the environment, teachers, etc.), and these positive percep- tions were also voiced by the individuals who shared testi- monials as to why they practiced at this location. The qualitative data contribute to the extant literature by high- lighting that, while people appreciated the mental and phys- ical benefits of yoga as well as teachers and the structure, the practitioners were most likely to attend the studio due to the sense of community. While many of these relationships were in the predicted directions, there are several empirical and pragmatic observations to consider. the day, week, month . . . the yoga studio supports that for me,” and “it grounds me”. Related to the minor theme of mindfulness is the holistic health benefits of yoga that was reported by 6% (n = 1) of the participants. Qualitative Journal of Primary Care & Community Health 6 Table 2. Correlation Matrix for Psychosocial Variables and Attendance. Table 2. Correlation Matrix for Psychosocial Variables and Attendance. Table 2. Correlation Matrix for Psychosocial Variables and Attendance. M (SD), N (%) 1 2 3 4 5 6 7 8 Female sex (1) 80 (87) — Yoga instructor status, affirmative (2) 14 (15) −.04 — Number of days attended past 12 monthsa (3) 59.94 (60.36) −.11 .15 — Meeting physical activity recommendations (L-Cat) (4) 75 (54) −.14 .16 .05 — PAGEQ sum score (5) 3.88 (0.59) −.10 .14 .43 −.00 — Self-compassion sum score (6) 3.22 (0.52) −.18 .05 .31 .02 .13 — Mindfulness sum score (7) 3.43 (0.37) −.18 .09 .12 .02 .13 .61 — Perceived stress sum score (8) 3.2 (0.39) .17 .00 −.09 .02 −.18 −.31 −.10 — aAttendance and duration of membership are only with the N = 92 who provided access to MindBody data. Correlation is significant at the .05 level (2-tailed). Correlation is significant at the .01 level (2-tailed). aAttendance and duration of membership are only with the N = 92 who provided access to MindBody data. Correlation is significant at the .05 level (2-tailed). *Correlation is significant at the .01 level (2-tailed). aAttendance and duration of membership are only with the N = 92 who provided access to MindBody data. Correlation is significant at the .05 level (2-tailed). **Correlation is significant at the .01 level (2-tailed). This study aimed to fill the gap in the literature around these 3 issues and found that yoga practitioners in this commu- nity-based studio have many positive perceptions (of yoga and the studio) as well as engage in positive health behav- iors. Overall, the quantitative data suggest that yoga studio clients who participated in this study were moderately stressed, mindful, self-compassionate, and cohesive. However, only half of the participants were meeting physi- cal activity recommendations. Self-compassion and per- ceived stress were negatively associated. This is unsurprising as self-compassion measures how individuals view them- selves during times of stress. In addition, self-compassion is a type of mindfulness in which an individual perceives his or her situation without judgment, so the measured associa- tion between self-compassion and mindfulness, although not previously explored in community settings, is in the expected direction. Qualitative self- growth in my own practice physically.” The other 4 of the 9 individuals who mentioned physical health benefits (44%) provided meaning units for the subtheme related to yoga as a balance to other workout regimens such as: “but I really wanted to get some strength in my muscles that I can’t get from weight lifting,” “yoga is the one thing to get all soreness out and bring balance [back into their legs after running long dis- tances],” and “I love this just kind [of] for a stretching and lengthening [after running and doing CrossFit].” Sum scores of the scales indicate that participants were sometimes stressed (Msumscore = 3.2 ± 0.39), often mindful Sum scores of the scales indicate that participants were sometimes stressed (Msumscore = 3.2 ± 0.39), often mindful (Msumscore = 3.4 ± 0.41), moderately self-compassionate (Msumscore = 3.26 ± 0.56), and moderately cohesive (Msumscore = 3.87 ± 0.62). ( sumscore ) (Msumscore = 3.4 ± 0.41), moderately self-compassionate (Msumscore = 3.26 ± 0.56), and moderately cohesive (Msumscore = 3.87 ± 0.62). There was a significant relationship between duration of membership and number of days attended in the past year (P = .008) and mindfulness (P = .04). There was also a significant relationship between age and the days attended in the past year (P = .025) and duration of mem- bership (P = .04). No other significant relationships were found between duration of membership or age with other key variables of interest (P > .05). Please see Table 2 for a full correlation matrix. Minor emergent themes identified in this data set were mindfulness, the variety of classes offered at the yoga stu- dio, the teachers employed by the yoga studio, the yoga stu- dio itself, and the holistic health benefits that yoga can provide. In this study, 22% (n = 4) of the participants reported that the mindfulness aspects of yoga were what brought them to practice at the yoga studio. Meaning units for this theme included: “meditation,” “set my attention for Finally, a subset of individuals consented to have their MindBody data shared with the research team to gather information on attendance. Participants (n = 92) were members for a range of less than 1 year to 8 years and attended 1 to 276 classes in the previous 12 months. Notably, 88% of this subset met physical activity recommendations according to their L-Cat response. Qualitative The meaning unit for this theme was “I personally do this [yoga] for holistic health.” The other 3 minor engagement themes pre- sented in the data set for this study were related to student perceptions of the yoga studio itself. Overall, 50% (n = 9) of the participants reported that they practice at the yoga studio because of the class struc- ture. Twenty-two percent (n = 4) of the participants prac- ticed at the yoga studio because of the yoga teachers who were employed by the studio owner. Meaning units for this theme included “the teachers are amazing,” and “the instruc- tors are very well-versed and make it such an open atmo- sphere.” In addition, the nonjudgmental teaching style and the client-centered emphasis that instructors integrated into their classes were highlighted in the data set and counted as meaning units. For example, one participant mentioned that instructors incorporated certain poses to treat injuries. A minor theme related to the teachers employed at the yoga studio was the variety in yoga classes offered: 17% (n = 3) of participants reported that the many different yoga classes offered at the yoga studio was what brought them to practice there. Examples of meaning units included: “I can do yoga every day, at different times and [a] variety of different classes,” and “I like the different types of yoga that they [the yoga studio] offer.” Finally, 11% (n = 2) reported that they practiced at the yoga studio because of the environment or atmosphere of the studio. Meaning units that contributed to this minor theme included: “the relaxed atmosphere,” and “a place, an atmosphere to escape my actual responsibilities . . . provides me (with) knowledge.” First, practitioners were rather cohesive and there was a significant positive correlation between the number of days individuals attend classes and positive perceptions of cohe- sion. Notably, this did not hold true for duration of member- ship and perceptions of cohesion. In other words, more frequent attendance was associated with higher perceptions of cohesion, but longer duration of membership was not associated with higher perceptions of cohesion. Simply hav- ing a membership did not provide the benefit of cohesion; individuals actually have to operationalize their membership through class attendance to experience cohesion. Relatedly, Discussion Limited research has been conducted within existing yoga studios and communities regarding psychosocial outcomes, physical activity behaviors, and yoga studio satisfaction. 7 7 Harden et al rather than inclusive. To meet the needs of these individuals, however, the studio owner and manager responded by offer- ing a series for these women with special considerations addressed in that the class was led by a female and the win- dows were covered to accommodate cultural preferences (i.e., practitioners were Muslim). Another strategy to pro- mote diversity in yoga participation (e.g., more diverse in terms of gender, race, ethnicity, religion, physical activities, and age) is to have more instructors from diverse back- grounds so that practitioners can identify themselves in their instructor.31 Future work is needed to understand what stu- dio offerings may attract more individuals from diverse populations. in a 1-year (6-month intervention, 6-month maintenance phase) stretching versus yoga intervention, preliminary results indicated that the stretching group significantly improved stress as measured by cortisol levels when com- pared to the yoga group.26 The researchers suggested that the key to this difference was that the stretching sessions were interpersonally dynamic whereas the yoga sessions did not prompt social interactions.26 This opportunity for engage- ment (i.e., facilitated by instructor, environment, or group norms) is the key of a “true group” rather than a group of individuals exercising at the same time.27 Therefore, there may be opportunities for low-dose group cohesion interventions at yoga studios. A group dynamics–based intervention features key principles of group environment, group structure, and group processes that may include interaction and communication, goal set- ting and progress updates, putting people in proximity to each other, and friendly competition.28,29 These principles are all represented in the studio’s 30-day challenge (described above) and may be a low-dose, high reward group dynamics–based intervention that may improve attendance, cohesion, inclusivity, and diversity. Future work is needed to explore the causal relationships between the 30-day challenge and outcomes of interest (attendance, cohesion, etc). Finally, it is notable that half of the participants self- reported that they were meeting the physical activity guide- lines for Americans. Discussion This may seem low considering that they engage in physical activity during a yoga class, but it is higher than the national average (e.g., 23.5% nationally vs 54% of studio members).32 Since the other half of the par- ticipants were not meeting recommendations, these results highlight opportunities to (1) test the physiological benefits of each yoga practice offered at this specific studio and (2) share these physiological benefits with practitioners at the studio. On the other hand, this low rate may also simply be reflective of the limitation that the L-Cat was not previously validated with this population. Future work is needed to determine the veracity of the L-Cat responses or to deter- mine a better fit measure of physical activity behaviors of yoga practitioners. , ) Our second observation is that tailored studio offerings may be beneficial to extend reach and representativeness. This aligns with other implementation research in that work is needed to understand for whom, under what conditions, and how a particular intervention or phenomenon works. For example, individuals from health disparate populations may need beginner level classes, high quality instructors, and explanation of yoga’s benefits. These key needs can be integrated into studio offerings or strategy but may not be impactful and well-received by those more familiar with yoga. Therefore, for individuals newer to the studio, an introductory series may be beneficial. In addition to build- ing knowledge and self-efficacy in a yoga practice, this introductory series could also launch a cohort of new prac- titioners to feel more cohesive, which could lead to greater adherence.30 The studio has offered a 4-week introductory series four times and limits these sessions to 15 individuals (to ensure individualized feedback and community) for the past 2 years. Future work is needed determine the feasibility and acceptability of offering an introductory series, and the mental and physical health benefits that may result from participation. Another area for future exploration is the fact that few participants (0.7%) reported attending the meditation-only classes despite stating that they attend yoga classes for men- tal health benefits. This may indicate that participants per- ceive receiving mental and physical health benefits in standard yoga-studio class practices, as opposed to needing to attend a stand-alone meditation class. Future work is needed to explore these relationships. These preliminary data and potential implications for future directions are an important first step for this aca- demic-studio partnership. Discussion However, this study has several limitations based on its design. First, as the survey por- tion was cross-sectional, it is impossible to make causal inferences between psychosocial factors and yoga class attendance or physical activity behaviors and yoga class attendance. Second, this study did not capture the con- founding variables of yoga self-efficacy33 and yoga acceptability.34 Third, there may have been selection bias in the sampling (i.e., convenience sampling), resulting in systematic bias that leads to skewed results. This selec- tion bias in the sampling might have been amplified by the availability of the survey during a community event where vegan snacks were offered at the studio; class attendees who chose to attend the event (and complete the survey) may systematically differ from class attendees While the participants in this study were reflective of nationwide yoga participants (e.g., predominantly Caucasian females)8,13 and the community in which the study was con- ducted, the studio values a focus on diversity and inclusivity. For example, a few community members requested a female- only yoga class. After careful consideration, studio manage- ment determined that a female-only class felt exclusive Journal of Primary Care & Community Health 8 practices [published online October 24, 2018]. Int J Yoga Therap. doi:10.17761/2019-00024 who did not attend the event. Fourth, both internal and external validity might have been affected by the study setting (i.e., administering the yoga-based survey within a yoga studio during a social event): Internal validity might have been decreased by social desirability effects as par- ticipants may have depicted themselves as consistent with yogic stereotypes, especially since at least some partici- pants completed the survey in close proximity to studio staff, teachers, and students. External validity might have been decreased by setting interaction (ecological valid- ity), making it inappropriate to generalize findings to other locations, occasions, and times. Fifth, criterion validity of the L-Cat might have been reduced by ambigu- ity over how yoga classes (i.e., especially different types of yoga classes) fit into the physical activity categories, or by ambiguity over whether participants should include yoga classes at all in their self-reported physical activity. Sixth, the sociodemographic characteristics of yoga prac- titioners at this studio are in line with national data sug- gesting that yoga is most commonly practiced by Caucasian, college-educated females. Discussion In fact, the propor- tion of females in this study was slightly higher than in other studies of yoga in community settings.8,35 Future work is needed to identify if perceptions and behaviors within the studio vary by key demographic variables or if the same positive perceptions are observed in other com- munity-based yoga studios. 3. Cramer H, Lauche R, Langhorst J, Dobos G. Is one yoga style better than another? A systematic review of associations of yoga style and conclusions in randomized yoga trials. Complement Ther Med. 2016;25:178-187. doi:10.1016/j.ctim.2016.02.015 4. Field T. Yoga research review. Complement Ther Clin Pract. 2016;24:145-161. doi:10.1016/j.ctcp.2016.06.005 5. Wang CC, Li K, Choudhury A, Gaylord S. Trends in yoga, tai chi, and qigong use among US adults, 2002-2017. Am J Public Health. 2019;109:755-761. doi:10.2105/AJPH.2019.304998 6. Cramer H, Ward L, Steel A, Lauche R, Dobos G, Zhang Y. Prevalence, patterns, and predictors of yoga use: results of a US nationally representative survey. Am J Prev Med. 2016;50:230-235. doi:10.1016/j.amepre.2015.07.037 7. Park CL, Braun T, Siegel T. Who practices yoga? A sys- tematic review of demographic, health-related, and psycho- social factors associated with yoga practice. J Behav Med. 2015;38:460-471. doi:10.1007/s10865-015-9618-5 8. Ayala SG, Wallson K, Birdee G. Characteristics of yoga prac- tice and predictors of practice frequency. Int J Yoga Therap. 2018;28:107-111. doi:10.17761/2018-00012R2 9. Govindaraj R, Karmani S, Varambally S, Gangadhar BN. Yoga and physical exercise—a review and comparison. Int Rev Psychiatry. 2016;28:242-253. doi:10.3109/09540261.20 16.1160878 10. Glasgow RE. What does it mean to be pragmatic? Pragmatic methods, measures, and models to facilitate research translation. Health Educ Behav. 2013;40:257-265. doi:10.1177/1090198113486805 Regardless of these limitations, this work provides an overview of the population, classes, and psychometric mea- surement of practitioners in a real-world setting. These data will be used to inform future decisions that will be empiri- cally and practically meaningful.36 11. Quilty MT, Saper RB, Goldstein R, Khalsa SBS. Yoga in the real world: perceptions, motivators, barriers, and patterns of use. Glob Adv Health Med. 2013;2:44-49. doi:10.7453/ gahmj.2013.2.1.008 12. Patwardhan AR. Yoga research and public health: is research aligned with the stakeholders’ needs? J Prim Care Community Health. 2017;8:31-36. doi:10.1177/2150131916664682 Declaration of Conflicting Interests 13. Ipsos Public Affairs. The 2016 Yoga in America Study Conducted by Yoga Journal and Yoga Alliance. https:// www.yogaalliance.org/Portals/0/2016%20Yoga%20in%20 America%20Study%20RESULTS.pdf. Published January 2016. Accessed August 22, 2019. The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Funding 14. Drahota A, Meza RD, Brikho B, et al. Community-academic partnerships: a systematic review of the state of the litera- ture and recommendations for future research. Milbank Q. 2016;94:163-214. doi:10.1111/1468-0009.12184 The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: They would like to thank the VT’s Open Access Subvention Fund for sponsoring the open access fee. 15. US Census Bureau. QuickFacts: Blacksburg town, Virginia. https://www.census.gov/quickfacts/fact/table/blacksburg- townvirginia/PST045218. Accessed April 21, 2019. ORCID iD 16. Michie S, van Stralen MM, West R. The behaviour change wheel: a new method for characterising and designing behaviour change interventions. Implement Sci. 2011;6:42. doi:10.1186/1748-5908-6-42 References 1. Lasater JH. The 3 keys to effective yoga teaching. https://www. yogauonline.com/yogau-product/6751. Accessed August 22, 2019. 1. Lasater JH. The 3 keys to effective yoga teaching. https://www. yogauonline.com/yogau-product/6751. Accessed August 22, 2019. 17. Creswell JW, Plano Clark VL. Designing and Conducting Mixed Methods Research. 2nd ed. Thousand Oaks, CA: Sage; 2010. 2. Smith BH, Lyons MD, Esat G. Yoga kernels: a public health model for developing and disseminating evidence-based yoga 18. Estabrooks PA, Carron AV. The Physical Activity Group Environment Questionnaire: An instrument for the assessment Harden et al 9 of cohesion in exercise classes. Group Dyn. 2000;4:230-243. doi:10.1037/1089-2699.4.3.230 the effectiveness of interventions to promote physical activ- ity. Sport Exerc Psychol Rev. 2006;2:19-35. 28. Carron AV, Spink KS. Team building in an exercise setting. Sport Psychol. 1993;7:8-18. doi:10.1123/tsp.7.1.8 19. Kiernan M, Schoffman DE, Lee K, et al. The Stanford Leisure-Time Activity Categorical Item (L-Cat): a single cat- egorical item sensitive to physical activity changes in over- weight/obese women. Int J Obes (Lond). 2013;37:1597-1602. doi:10.1038/ijo.2013.36 29. Estabrooks PA, Harden SM, Burke SM. Group dynamics in physical activity promotion: what works? Soc Personal Psychol Compass. 2012;6:18-40. doi:10.1111/j.1751-9004.2011 .00409.x 20. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24:385-396. 30. Estabrooks PA. Sustaining exercise participation through group cohesion. Exerc Sport Sci Rev. 2000;28:63-67. 21. Raes F, Pommier E, Neff KD, Van Gucht D. Construction and factorial validation of a short form of the Self-Compassion Scale. Clin Psychol Psychother. 2011;18:250-255. doi:10.1002 /cpp.702 31. Everette A, Brooks A, Ramalingam NS, et al. An integrated research–practice partnership for physical activity promo- tion in a statewide program. Transl J Am Coll Sports Med. 2017;2:57-67. doi:10.1249/TJX.0000000000000035 22. Bohlmeijer E, ten Klooster PM, Fledderus M, Veehof M, Baer R. Psychometric properties of the five facet mindfulness ques- tionnaire in depressed adults and development of a short form. Assessment. 2011;18:308-320. doi:10.1177/1073191111408231 32. Centers for Disease Control and Prevention. Exercise or phys- ical activity. https://www.cdc.gov/nchs/fastats/exercise.htm. Published February 26, 2019. Accessed April 9, 2019. 33. Birdee GS, Sohl SJ, Wallston K. Development and psycho- metric properties of the Yoga Self-Efficacy Scale (YSES). BMC Complement Altern Med. 2016;16:3. doi:10.1186/ s12906-015-0981-0 23. Baer RA, Smith GT, Hopkins J, Krietemeyer J, Toney L. Using self-report assessment methods to explore facets of mindfulness. Assessment. 2006;13:27-45. 24. Neuendorf KA. The Content Analysis Guidebook. Thousand Oaks, CA: Sage; 2002. 34. Brems C, Colgan D, Freeman H, et al. References Elements of yogic practice: perceptions of students in healthcare programs. Int J Yoga. 2016;9:121-129. 25. Denzin NK, Lincoln YS. Handbook of Qualitative Research. Thousand Oaks, CA: Sage; 1994. 35. Birdee GS, Ayala SG, Wallston KA. Cross-sectional analysis of health-related quality of life and elements of yoga prac- tice. BMC Complement Altern Med. 2017;17:83. doi:10.1186/ s12906-017-1599-1 26. Corey SM, Epel E, Schembri M, et al. Effect of restorative yoga vs stretching on diurnal cortisol dynamics and psy- chosocial outcomes in individuals with the metabolic syn- drome: the PRYSMS randomized controlled trial. Psychoneu- roendocrinology. 2014;49:260-271. doi:10.1016/j.psyneuen .2014.07.012 36. Estabrooks PA, Harden SM, Almeida FA, et al. Using inte- grated research-practice partnerships to move evidence-based principles into practice. Exerc Sport Sci Rev. 2019;47:176- 187. doi:10.1249/JES.0000000000000194 27. Burke SM, Carron AV, Eys MA, Ntoumanis N, Estabrooks PA. Group versus individual approach? A meta-analysis of
https://openalex.org/W3124986798	https://publikationen.bibliothek.kit.edu/1000129539/102185531	English	null	Efficient Communication Protection of Many-Core Systems against Active Attackers	Electronics	2,021	cc-by	22,556	Article Efﬁcient Communication Protection of Many-Core Systems against Active Attackers † Sadia Moriam 1,,‡, Elke Franz 2,‡, Paul Walther 2,‡, Akash Kumar 3 , Thorsten Strufe 4,5 and Gerhard Fettweis 1 Elke Franz 2,‡, Paul Walther 2,‡, Akash Kumar 3 , Thorsten Strufe 4,5 and Gerhard Fettweis 1 Sadia Moriam 1,,‡, Elke Franz 2,‡, Paul Walther 2,‡, Akash Kumar 3 , Thorsten Strufe 4,5 an 1 Vodafone Chair Mobile Communication Systems, Technische Universität Dresden, 01069 Dresden, Germany gerhard.fettweis@tu-dresden.de g 2 Chair of Privacy and Data Security, Technische Universität Dresden, 01069 Dresden, Germany; elke.franz@tu-dresden.de (E.F.); paul.walther@tu-dresden.de (P.W.) 3 Chair of Processor Design, Technische Universität Dresden, 01069 Dresden, Germany; akash.kumar@tu-dresden.de 4 Chair of IT Security, Karlsruhe Institute of Technology (KIT), 76131 Karlsruhe, Germany; strufe@kit.edu 5 Centre for Tactile Internet Technische Universität Dresden 01069 Dresden Germany y gy y 5 Centre for Tactile Internet, Technische Universität Dresden, 01069 Dresden, Germany * Correspondence: sadia_moriam@yahoo.com p y † This paper is an extended version of our paper published in the Proceedings of ACM Great Lakes S i VLSI (GLSVLSI) M i t l Chi IL USA 2018 † This paper is an extended version of our paper published in the Proceedings of ACM Great Lakes Symposium on VLSI (GLSVLSI), Moriam et al., Chicago, IL, USA, 2018. Symposium on VLSI (GLSVLSI), Moriam et al., Chicago, IL, USA, 2018. ‡ These authors contributed equally to this work. Abstract: Many-core system-on-chips, together with their established communication infrastructures, Networks-on-Chip (NoC), are growing in complexity, which encourages the integration of third- party components to simplify and accelerate production processes. However, this also adversely exposes the surface for attacks through the injection of hardware Trojans. This work addresses active attacks on NoCs and focuses on the integrity and availability of transmitted data. In particular, we consider the modiﬁcation and/or dropping of data during transmission as active attacks that might be performed by malicious routers. To mitigate the impact of such active attacks, we propose two lightweight solutions that respect the performance constraints of NoCs. Assuming the presence of symmetric keys, these approaches combine lightweight authentication codes for integrity protection with network coding for increased efﬁciency and robustness. The proposed solutions prevent undetected modiﬁcations and signiﬁcantly increase availability through a reliable detection of attacks. The efﬁciency of these solutions is investigated in different scenarios using cycle-accurate simulations and the area overhead is analyzed relative to state-of-the-art many-core system. Citation: Moriam, S.; Franz, E.; Walther, P.; Kumar, A.; Strufe, T.; Fettweis, G. Efﬁcient Communication Protection of Many-Core Systems against Active Attackers. Electronics 2021, 10, 238. https://doi.org/ 10.3390/electronics10030238 Received: 21 December 2020 Accepted: 17 January 2021 Published: 21 January 2021 Keywords: Networks-on-Chip; integrity; availability; network coding; performance Publisher’s Note: MDPI stays neu- tral with regard to jurisdictional clai- ms in published maps and institutio- nal afﬁliations. Article Efﬁcient Communication Protection of Many-Core Systems against Active Attackers † The results demonstrate that one authentication scheme with network coding protects the integrity of data to a low residual error of 1.36% at 0.2 attack probability with an area overhead of 2.68%. For faster and more ﬂexible evaluation, an analytical approach is developed which is validated against the cycle-accurate simulations. The analytical approach is more than 1000× faster while having a maximum estimation error of 5%. Moreover, the analytical model provides a deeper insight into the system’s behavior. For example, it reveals which factors inﬂuence the performance parameters. Citation: Moriam, S.; Franz, E.; Walther, P.; Kumar, A.; Strufe, T.; Fettweis, G. Efﬁcient Communication Protection of Many-Core Systems against Active Attackers. Electronics 2021, 10, 238. https://doi.org/ 10.3390/electronics10030238 electronics 1. Introduction The shift from single core to multi-processor systems-on-chip (MPSoCs) [1] has facil- itated a massive increase in performance while keeping the power consumption within limits. Since MPSoCs can consist of thousands of cores, it is of utmost importance to provide a scalable and efﬁcient communication medium for them. Here, classical bus-based systems are increasingly being replaced by systems based on packet-switched Networks-on-Chip (NoC) as a solution to the interconnection problem [2–4]. 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/). The growing complexity of the systems implies a higher susceptibility to errors [5]. This increased complexity is also reﬂected in the respective supply chain: own implemen- tation of such design processes as well as owning the respective factories is associated with https://www.mdpi.com/journal/electronics Electronics 2021, 10, 238. https://doi.org/10.3390/electronics10030238 Electronics 2021, 10, 238 2 of 31 immense costs [6]. Hence, due to complexity and costs, MPSoC design and production pro- cesses increasingly rely on the integration of third-party components, potentially also from untrusted sources [7]. Even tools used for design and development may be compromised, and hence pose a tangible threat to MPSoCs [8]. Such threats can be realized by inserting hardware Trojans (HT), as reported, e.g., in [7,9–12]. When designing NoCs it thus becomes increasingly important to consider not only the resilience against errors but the systems security as well. Although different methods for HT detection have been proposed, the general presence of an HT can never be excluded [6]. Hence, system designs have to be Trojan-tolerant, i.e., the system itself works even if an HT is present [6]. j y NoCs are an attractive target for attackers due to their basic functionality: By design, the MPSoCs’ entire data exchange passes through the NoC, hence, an attacker has the maximum possibilities of intervention at this point [13]. Purely passive HTs which act as eavesdroppers and aim to exﬁltrate conﬁdential data over local or remote channels can be countered by using end-to-end encryption of data. Active attackers, on the other hand, who additionally modify or discard data, are much harder to deal with. Even if the modiﬁcations or losses of ﬂow control units (ﬂits) are detected, the respective retransmissions signiﬁcantly increase the network load and thereby reduce the performance or even basic functionality. 1. Introduction y p y In this work, we present protocols, which ensure integrity and increase availability in NoCs, even in the presence of HTs in routers [14]. For this, we combine the features of network coding [15] with cryptographic authentication schemes. The authentication schemes relying on lightweight message authentication codes ensure that ﬂit modiﬁcations are detected and handled accordingly. Since message authentication codes are symmetric cryptographic primitives, they require a shared secret between the communicating nodes— this key exchange is out of the scope of this work, but could be realized by pre-sharing keys during an initiation phase. Additionally, network coding measures create robustness against the dropping of ﬂits by HTs and also against the discarding of modiﬁed ﬂits, which in combination effectively reduces retransmissions. We evaluate our proposed schemes with cycle-accurate simulations under realistic trafﬁc assumptions as well as with an analytical model. The results demonstrate that our schemes can ensure a secure data transmission in the presence of active attackers by reducing the respective error probability by up to 85.6% at a very reasonable overhead. Finally, we analyze to area overhead inﬂicted by the newly proposed schemes and show that only a 2.68% area increase is needed in comparison to a state-of-the-art MPSoC. In summary, we make the following contributions in this paper: (1) we propose protocols providing integrity protection and availability enhancement for NoC communications, (2) we subsequently evaluate the performance of these protocols extensive simulations (2) we subsequently evaluate the performance of these protocols extensive simulations, (3) we develop an analytical model for faster and more ﬂexible evaluation, q y p p (3) we develop an analytical model for faster and more ﬂexible evaluation, p y (4) ﬁnally, we analyze our solutions in terms of additional chip area required. The remaining work is organized as described below: The state of the art regarding NoC security and HTs is laid out in Section 2. Section 3 describes our system model and the respective assumptions. In Section 4, we propose our solution for integrity protection. We then present our analytical model for further analysis of the proposed security solutions in Section 5. Here, we also detail the results of this model accompanied by the respective simulation results and a discussion of the inﬂicted area overhead. Finally, Section 6 summarizes the paper and describes possible further research questions. 2. Related Work and State of the Art One of the main attack angles against MPSoCs is the embedding of HTs into them, e.g., [6,13,16,17]. HTs are covert autonomous functional units directly incorporated by attackers into the hardware of the attacked system. Due to the physical hardware in- tegration, HTs have direct access to all processed data at the lowest level. This allows them to arbitrarily read, possibly exﬁltrate, modify or discard data [6]. Their ability to execute arbitrary attacks, especially in the domain of internal communications, demands Electronics 2021, 10, 238 3 of 31 effective countermeasures [13]. A presentation of signiﬁcant work in this area is given in the following. Ancajas et al. [10] were the ﬁrst to present the threat potential of HTs in NoCs and presented initial solutions on how to diminish such threats. These solutions include measures such as scrambling the data at the lowest layer, certifying individual transmitted packets, and a process migration scheme, which serves to obfuscate the respective target application [10]. These measures are primarily intended to prevent data interception by a malicious NoC by complicating the tracing of logical data streams. Additionally, these schemes intend to prevent the initial triggering of an HT. A similar strategy is followed by the solution of Setumadhaven et al. [7], where the triggering events of HTs are also inhibited. Despite the effectiveness of these measures, they involve considerable computational effort and thus signiﬁcantly degrade the performance of the system. g y g p y Another approach to minimizing the risks posed by HTs is to detect and remove them at runtime. Frey and Yu [18] proposed a system based on a ﬁnite state machine, which represents the possible execution paths for the single MPSoC components. If a component deviates from the set of valid execution paths, it is assumed to be compromised and accordingly, this component is disabled. Since the ﬁnite state machine must represent the respective executions in real time and store the individual states of all monitored components, this system comes with a non-negligible computational and memory overhead. The authors proposed a more lightweight solution for the detection of HTs that alter ﬂits and their attached meta-data, e.g., routing information [18]. Since this system operates within the NoCs routers, an additional overhead for each router and in turn a reduced NoC performance is implied. A fundamentally different approach is to prevent HTs from being embedded in the system up front [7,19]. 3.1. System Model and Attacker Model Our system model assumes a 2D mesh network consisting of N × N nodes as under- lying NoC topology. An example of this topology is visualized in Figure 1. Each network node is composed of a processing element (PE), an interface to the NoC (NoC interface, NI), and a router. The routers will process the ﬂits in ﬁrst-in-ﬁrst-out (FIFO) order. The respective routing decision is determined via XY routing. Hence, the order of the single ﬂits traversing the NoC is not altered by this transmission. Finally, we assume a ﬂit size of approximately 150 bits. PE NI PE NI PE NI PE NI PE NI PE NI PE NI PE NI PE NI (0,0) (0,1) (0,2) (1,0) (1,1) (1,2) (2,0) (2,1) (2,2) Router Figure 1. Example Networks-on-Chip (NoC) topology (2D mesh) with attached processing elements (PE) and network interfaces (NI). Figure 1. Example Networks-on-Chip (NoC) topology (2D mesh) with attached processing elements (PE) and network interfaces (NI). In this paper, we use a spatial uniform trafﬁc distribution with a constant injection rate per module. To simplify matters, we consider the transmission of single ﬂits, i.e., each ﬂit contains the necessary header ﬁelds such as source and target address. The actual ﬂit injection rate into the network is directly inﬂuenced by the communication scheme used—for example, schemes based on network coding will inject redundant ﬂits to increase robustness and, thereby, alter the actual injection rate. To account for this changed injection rate, the creation rate of ﬂits in the PE is adapted accordingly. Therefore, the actual ﬂit injection rate is kept equal for all schemes at ≈0.2 ﬂits/module/cycle. Furthermore, we assume that the NIs are equipped with retransmission buffers of sufﬁcient size to avert ﬂit loss (see Section 5). The data interface between PE and NI will be 64 bit wide, i.e., a single data chunk will be of size 64 bit. To prepare this data for transmission through the NoC, the NI will add the following meta-data to it: The payload itself is extended by a 24-bit ﬂit identiﬁer FID (corresponding to network coded transmission, Section 3.2), which enables a unique identiﬁcation in case of retransmission.Hence, the payload is composed of a mode ﬁeld (4 bit) specifying the ﬂit type, an address ﬁeld (32 bit) for memory accesses, the 24-bit FID, and 64 bits of data. 3. Background and Assumptions g p 3.1. System Model and Attacker Model 3.1. System Model and Attacker Model 2. Related Work and State of the Art For this purpose, different systems have been proposed that verify the design procedure of outsourced production processes by performing static backdoor analysis during the complete design and production phases. Alternatively, the entire design process is changed to security orientated development procedures. However, HTs are an attack vector that cannot be mitigated completely [6]. Therefore, other strategies aim to diminish the exposed risk by securing the communication during attacks. Following the classiﬁcation in [6], such solutions are Trojan-tolerant, as they provide the functionality even in the presence of an active HT. Notable works include the solution by Boraten and Kodi [12], who propose the use of algebraic manipulation detection codes for the identiﬁcation of ﬂit modiﬁcations. The authors claim a minimal performance impact of 1% compared to NoCs. No evaluation with respect to security is presented. Alternatively, Kapoor et al. proposed to protect the communication using authenticated encryption [20]. However, the chosen cryptographic primitive, AES-128 in GCM mode, is heavy-weight regarding efﬁciency and area overhead and, therefore, it is an infeasible solution for NoCs [9]. For a comprehensive presentation of the state-of-the-art regarding HTs and the protec- tion against them, we would like to refer the reader to surveys like those by Rajesh et al. [13], which details the threats and attack vectors exposed through HTs, Xiao et al. [6], presenting a classiﬁcation for HTs itself as well as for the respective countermeasures and, ﬁnally, Shakya et al. [8], describing HT deployment strategies and their detectability. In summary, of the classic security objectives, conﬁdentiality, integrity, and availability, the presented works usually consider only the ﬁrst two. If all three are considered, the efﬁciency and chip area are neglected instead. In this work, we consider both integrity and availability, taking both under the special circumstances of high latency requirements and limited area. Additionally, our solution is completely independent of the actual application and thus can be used widely. 4 of 31 Electronics 2021, 10, 238 3.1. System Model and Attacker Model Furthermore, a header is added, which contains the information to route the ﬂit through the NoC. This header is composed of a single bit indicating burst transmission (burst mode will be considered in future work) as well as x and y coordinates of source and target nodes. To support a 2D mesh of up to 16 × 16 modules, we assume 4 bits for each coordinate. Hence, in total, the single ﬂits exiting the NI will be 141 bits long. g g g In accordance with [21], we assume the PE and NI to be trustworthy, whereas the routers in the NoC are considered to be corrupted. This assumption is rooted in the respective functionality Electronics 2021, 10, 238 5 of 31 of these elements. The PE and NI typically contain the business logic of the MPSoC and are therefore developed in house in a controlled environment. The NoC’s routers, on the other hand, realize deterministic functionality, which makes them suitable candidates for outsourcing and thus vulnerable to attack. Since we assume the NIs to be trustworthy, the additional functionality proposed by our protocols is placed within those components as shown in Sections 4.2 and 4.3. p p y p p p This work is focused on active attacks executed by the NoC routers, i.e., they modify a traversing flit with a certain modification probability pm or drop it with a drop probability pd. At- tackers are computationally restricted so that they are not able to break cryptographic measures. In general, active attacks cannot be prevented but only detected. A simple approach would be to stop the system whenever an attack was detected. However, then an attacker can disturb the availability of the system with a single modiﬁcation or drop. In contrast, we aim at a robust system that allows transmitting data even in the presence of an active attacker. We assume that an attacker tries to keep undetected what implies that he will not manipulate all transmissions. Hence, when the receiver recognizes modiﬁcations or losses, an automatic repeat request (ARQ) will be issued in order to trigger the retransmission of the affected ﬂits. ARQs have a similar structure than data ﬂits; the data ﬁeld can be used to specify details regarding the retransmission. The solution design for an appropriate protection scheme must account for the special requirements of MPSoCs and NoCs. 3.1. System Model and Attacker Model Speciﬁcally, this means that the high speed of the NoC must be maintained and the use of the limited available chip area and in turn the energy used must be minimized. Thus, the proposed security concept must neither cause signiﬁcant performance losses nor use a comparatively large amount of chip area. 3.2. Network Coded Transmission Electronics 2021, 10, 238 6 of 31 Hence, a generation in the case of network coded transmission always consists of two ﬂits ⃗f1, ⃗f2: ⃗f1 ⃗f2 ! = β1,1 β1,2 x1,1 x1,2 · · · x1,n β2,1 β2,2 x2,1 x2,2 · · · x2,n = 1 0 x1,1 x1,2 · · · x1,n 0 1 x2,1 x2,2 · · · x2,n (2) (2) Only the sending node computes linear combinations so that the forwarders are not burdened with computational overhead. It selects at random the encoding coef- ﬁcients αi,j ∈Fq, i = 1, 2, . . . , C; j = 1, 2 for the computation of the linear combinations ⃗ck, k = 1, 2, . . . , C. For example, the three linear combinations in the case of G2C3 are computed according to the following equation:   ⃗c1 ⃗c2 ⃗c3  =   α1,1 α1,2 α2,1 α2,2 α3,1 α3,2  × ⃗f1 ⃗f2 ! (3) (3) Given the generation size of G = 2, two linear independent combinations ⃗ci,⃗cj are sufﬁcient for decoding the resulting matrix of these combinations by multiplying it with the inverse of the 2 × 2 matrix A of their corresponding encoding coefﬁcients: Given the generation size of G = 2, two linear independent combinations ⃗ci,⃗cj are sufﬁcient for decoding the resulting matrix of these combinations by multiplying it with the inverse of the 2 × 2 matrix A of their corresponding encoding coefﬁcients: ⃗f1 ⃗f2 ! = A−1 × ⃗ci ⃗cj = αi,1 αi,2 αj,1 αj,2 −1 × ⃗ci ⃗cj (4) (4) Invertibility of A ensures that the matrix of linear combinations can be decoded. This invertibility depends both on the size of A (given by G) and on the size of the ﬁnite ﬁeld q [24]. The probability pinv(G, q) that A can be inverted is given by [25] Invertibility of A ensures that the matrix of linear combinations can be decoded. This invertibility depends both on the size of A (given by G) and on the size of the ﬁnite ﬁeld q [24]. The probability pinv(G, q) that A can be inverted is given by [25] pinv(G, q) = G ∏ i=1 1 −1 qi . 3.2. Network Coded Transmission For the transmission of flits, we consider the use of network coding to increase robustness against loss of flits. We also implement uncoded transmission as a baseline communication scenario to be able to evaluate the benefits of network coding. In a network coded transmission, linear combinations of data to be sent is computed. The network coded transmission applied in this work follows the approach of Practical Network Coding [22]. In that approach, data to be transmitted is divided into blocks ⃗xi = (xi,1, xi,2, . . . , xi,n) ∈Fn q with q = 2m. Each block is enlarged by a so-called global encoding vector (GEV) (βi,1, βi,2, ..., βi,G) ∈FG q , βi,j̸=i = 0, βi,j=i = 1. The enlarged blocks ⃗x′ i = (βi,1, βi,2, ..., βi,G, xi,1, xi,2, . . . , xi,n) are arranged in matrices (generations). G blocks constitute the rows of one generation of size G. For each generation, the sender randomly selects encoding coefficients αi,j ∈Fq, i = 1, 2, . . . , C; j = 1, 2, . . . , G and computes C ≥G linear combinations⃗ci: ⃗ci = G ∑ j=1 αi,j ·⃗x′ j (1) (1) The computations are done in the underlying finite field Fq. The elements of the GEV re- flect the linear combinations applied to the original blocks what enables the receiver to decode. The computations are done in the underlying finite field Fq. The elements of the GEV re- flect the linear combinations applied to the original blocks what enables the receiver to decode. Whenever the receiver got at least G linear independent combinations of one generation, he can decode by solving a system of linear equations. Since the receiver needs to know to which generation the combinations belong, they are tagged with a generation identifier GID. For data transmission with a NoC, network coding is applied to ﬂits. To meet the demand for low latencies, we use a generation size of G = 2 that achieves average latencies comparable to an uncoded transmission [23]. We employed three communication scenarios during our evaluations: UC: uncoded transmission, UC: uncoded transmission, G2C3: network coded transmission with G = 2, C = 3, G2C3: network coded transmission with G = 2, C = 3, G2C4: network coded transmission with G = 2, C = 4. 4.1. Possible Approaches Losses of ﬂits will be detected using timers (details are given in the following sections). The common measure to achieve integrity is authentication through a symmetric or asym- metric cryptographic system. By both approaches, an authentication tag is computed that is used to verify the integrity of the message: a digital signature in the case of asymmetric authentication or a message authentication code in the case of symmetric authentication. Digital signatures are not suited for authentication of ﬂits since they are too long to be included within a ﬂit. Furthermore, their computation requires a high computational effort. Under the consideration of these drawbacks, we decided to use symmetric authentication schemes for the computation of the tags. The necessary key exchange is out of the scope of this paper. One possibility is to exchange keys during an initialization phase of the system. Losses of ﬂits will be detected using timers (details are given in the following sections). The common measure to achieve integrity is authentication through a symmetric or asym- metric cryptographic system. By both approaches, an authentication tag is computed that is used to verify the integrity of the message: a digital signature in the case of asymmetric authentication or a message authentication code in the case of symmetric authentication. Digital signatures are not suited for authentication of ﬂits since they are too long to be included within a ﬂit. Furthermore, their computation requires a high computational effort. An adequate solution for communication within a NoC has to fulﬁll the speciﬁc de- mands on delay and area overhead. The Advanced Encryption Standard (AES) is not suited because it requires 1032 cycles per block encryption [26]. Therefore, a lightweight method that is optimized for implementation in hardware should be chosen as the cryptographic scheme. We selected mCrypton [27] since that algorithm provides a low delay of only 13 cycles per block and requires an area of 2681 gate equivalents only [26]. The authentication tag prevents undetected modiﬁcation of data to be transmitted since the computation of this tag requires the knowledge of the secret key. In addition to the data ﬁeld, it is also necessary to protect the metadata ﬁelds since modiﬁcations of these ﬁelds can also harm the system. If the receiver detects a modiﬁcation, he discards the affected ﬂit. In the case of network coding, he might still be able to decode due to the included redundancy. 3.2. Network Coded Transmission (5) (5) According to Table I in [24], the ﬁeld GF (24) already achieves an inverting probability of 0.93384 for a matrix of size 2 × 2 as considered here. Besides, only the sender selects encoding coefﬁcients so that the invertibility of the resulting matrix can be easily checked. Hence, we assume a symbol size of 4 bits for encoding coefﬁcients and data symbols. y g y The network coded ﬂits that are injected into the network comprise nearly the same ﬁelds as the uncoded ﬂits (Section 3.1). There is only one extra ﬁeld of 8 bits for the GEV (2 symbols with 4 bits each) and the GID replaces the FID. Figure 2 depicts the structure of a network coded ﬂit. We selected a size of 24 bits for the GID to prevent replay attacks, i.e., an injection of formerly intercepted combinations into a transmission. Such an injection would prevent the successful decoding of the corresponding generation since the injected ﬂit is not a valid combination for that generation. Details regarding the prevention of replay attacks are given in the next section. 7 of 31 Electronics 2021, 10, 238 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 bit bit 149 bits 17 bits 132 bits Source X B Source Y Target X Target Y Mode Address Address (cont.) GID data GID (cont.) GEV Header Payload Figure 2. Structure of a network coded ﬂit with 64-bit data ﬁeld; uncoded ﬂits would have nearly the same structure but without the global encoding vector (GEV) and a ﬂit identiﬁer (FID) instead of the generation identiﬁer (GID). Figure 2. Structure of a network coded ﬂit with 64-bit data ﬁeld; uncoded ﬂits would have nearly the same structure but without the global encoding vector (GEV) and a ﬂit identiﬁer (FID) instead of the generation identiﬁer (GID). Figure 2. Structure of a network coded ﬂit with 64-bit data ﬁeld; uncoded ﬂits would have nearly the same structure but without the global encoding vector (GEV) and a ﬂit identiﬁer (FID) instead of the generation identiﬁer (GID). 4.2. S1: Send Data and Tag in Two Separate Flits In this protocol, the symmetric block cipher mCrypton is directly used to compute the tag. Therefore, CBC-MAC is employed that bases on cipher block chaining (CBC) mode [28]: Using an initialization vector of zero, the input blocks are encrypted employing CBC, and the last ciphertext block serves as tag. CBC-MAC has security deﬁciencies for messages of arbitrary length [29], but in the proposed protocol, the number of input blocks that need to be authenticated is constant. The block size of the underlying cipher determines the size of the tag. The block size of the chosen cipher mCrypton is 64 bits, hence, it equals the size of the data ﬁeld. The tag is put into the data ﬁeld of an additionally generated ﬂit, the so-called tag ﬂit. The mode ﬁeld indicates the ﬂit type. All other ﬁelds of data and tag ﬂit are the same. Uncoded transmission (UC): In the case of UC, the sender computes for each original ﬂit delivered by the PE a tag ﬂit. If the computation of the tag is ﬁnished, data ﬂit and tag ﬂit are put into the transmission buffer and sent consecutively. Additionally, a copy of both ﬂits is stored in the retransmission buffer so that the tag does not need to be computed again in case of retransmission. When the receiver gets a data ﬂit, the computation of the tag can immediately start. If the computed tag equals the received tag, veriﬁcation was successful and the data ﬂit can be delivered to the PE. Otherwise, an ARQ for both data ﬂit and tag ﬂit has to be issued since it is not possible to decide which of them was modiﬁed. The arrival of a flit also triggers the start of a timer at the receiver to allow for the recognition of losses. If there is a time-out, the receiver issues an ARQ. If the receiver gets first a tag flit, it can directly issue an ARQ since the order of flits is not changed during transmission. Network Coded transmission (NC): When G original ﬂits arrived at the NI of the sender, the computation of the C linear combinations starts (Figure 3). Afterward, the sender computes a tag for each of these combinations, puts the resulting 2 · C ﬂits into the transmission buffer, and sends them consecutively. Similar to UC, copies of all ﬂits are stored in the retransmission buffer. We investigated two different possibilities for authentication: We investigated two different possibilities for authentication: • Solution 1 (S1): send data and tag in two separate ﬂits • Solution 2 (S2): include data and tag in one ﬂit 4.1. Possible Approaches Otherwise or in case of uncoded transmission, the receiver issues an ARQ to initiate the retransmission of the modiﬁed ﬂits. In the case of network coded transmission, tags are computed for the linear combina- tions. Hence, the receiver can check the validity of combinations after arrival and use only valid ones for decoding. Network coding implies in general the need to use homomorphic authentication schemes. However, since we assume that only the sender computes linear combinations, this is not necessary. To avoid computational overhead for intermediate nodes, the tags are used for an end-to-end authentication, i.e., only the receiver veriﬁes the validity of ﬂits received. y An attacker is not able to undetectably modify ﬂits but he could try to disturb trans- mission by injecting a valid ﬂit sent by the same sender (replay attack). The FID or GID is an increasing number so that the receiver can recognize the injection of an already received ﬁt. Hence, the FID or GID must not repeat as long as the same key for the computation of the tag is used to prevent a replay attack. A length of 24 bits allows 224 different values for the identiﬁer. Given the payload of 64 bits per ﬂit, a sender can send 128 MiB data using uncoded transmission to the same receiver before the corresponding key needs to be changed. In the case of network coded transmission, the sender can transmit 256 MiB since the two ﬂits of a generation get the same GID. The amount of data that can be exchanged using the same key can be increased by enlarging the FID or GID. 8 of 31 Electronics 2021, 10, 238 We investigated two different possibilities for authentication: 4.2. S1: Send Data and Tag in Two Separate Flits When the receiver gets a data ﬂit (i.e., a linear combination), it starts the computation of the tag. After successful veriﬁcation of at least G combined ﬂits, decoding starts. Finally, the decoded ﬂits are delivered to the PE. Due to the redundancy, decoding may still be possible even if modiﬁcations are detected. For example, if 2 of 4 received combinations failed veriﬁcation, decoding is possible if the remaining 2 combinations are unmodiﬁed. If there are not enough valid combinations, the receiver issues an ARQ for both data and tag ﬂit. g For the recognition of losses, timers are used as described for UC. However, an ARQ is only necessary if not enough valid combinations arrived at the receiver. Input for the computation of the tag is the whole data ﬂit of 141 bits (UC) or 149 bits (NC). Given the block length of 64 bits, there are three input blocks for mCrypton, the last one padded with zeros. Hence, the computation of the tag implies a delay of 3 × 13 = 39 cycles for both sender and receiver. Since the injection rate of ﬂits is much higher, it is necessary to consider a reasonable number of crypto modules for each NI in the NoC so that authentication of different ﬂits can be performed in parallel (Section 5.5). Although we consider the transmission of single ﬂits, the protocols imply that more than one ﬂit will be injected into the network. The ﬂits that are needed for a successful transmission form a transmission unit. In the uncoded case, a transmission unit comprises the data ﬂit and tag ﬂit, in the network coded case, it comprises the C linear combinations and the corresponding tag ﬂits. 9 of 31 Electronics 2021, 10, 238 encode generation authenticate combinations store copies in RTB loss detected issue ARQ pass to PE tag =tag’ store for veriﬁcation retrieve ﬂit(s) from RTB decode generation compute tag’ tag? ARQ? yes no no no yes yes Network Interface From PE To router From router To PE sending receiving Figure 3. Steps performed by the network interface in case of S1/Network Coded (NC) transmission. authenticate combinations encode generation store copies in RTB issue ARQ no Figure 3. Steps performed by the network interface in case of S1/Network Coded (NC) transmission. 4.3. S2: Include Data and Tag in One Flit The authentication bits are stored in the second halves of Electronics 2021, 10, 238 10 of 31 10 of 31 the corresponding ﬂits. The resulting ﬂits are sent consecutively while a copy of each of them is stored in the retransmission buffer. the corresponding ﬂits. The resulting ﬂits are sent consecutively while a copy of each of them is stored in the retransmission buffer. After the arrival of a ﬂit, the receiver starts to compute the pseudo-random key, computes the tag bits, and veriﬁes the integrity of the ﬂit by a comparison of received and computed tag bits. Since one data block is divided into two ﬂits, both ﬂits are necessary for successful transmission. If veriﬁcation of one or both ﬂits failed, the receiver issues an ARQ. However, in contrast to S1, if only one ﬂit was affected, the retransmission of that single ﬂit is sufﬁcient. Equivalently to solution S1, timers are employed for the recognition of losses. Network Coded transmission: Again, the 64-bit data ﬁeld is split into two halves that are distributed to two ﬂits. These two ﬂits establish a generation of size G = 2 so that the sender can immediately compute the C linear combinations (Figure 4). split ﬂit encode generation authenticate combinations store copies in RTB loss detected issue ARQ merge ﬂits tag =tag’ store for veriﬁcation retrieve ﬂit(s) from RTB decode generation compute tag’ ARQ? yes no yes Network Interface From PE To router From router To PE sending receiving Figure 4. Steps performed by the network interface in case of S2/NC. no Figure 4. Steps performed by the network interface in case of S2/NC. Since the GEV also belongs to the metadata, the size of the input for the block cipher is 85 bits which also results in two input blocks. The subsequent processing is equivalent to the uncoded case: For each of the two ﬂits, the input blocks are encrypted, the tag bits are computed and stored in the second halves of the data ﬁeld and the ﬂits are sent. As in the uncoded case, the receiver starts to compute the pseudo-random key im- mediately after receiving a ﬂit. For successful decoding, G of the C linear combinations need to arrive successfully at the receiver. If this is the case, the receiver decodes, generates the original 64-bit data block and delivers it to the PE. 4.3. S2: Include Data and Tag in One Flit Otherwise, it issues an ARQ for the retransmission of a single ﬂit. For both uncoded and network coded transmission, there are two input blocks for the block cipher what requires 2 × 13 = 26 cycles for the selected algorithm. For S2, a transmission unit contains the two ﬂits in case of the uncoded transmission and all C linear combinations in case of network coded transmission. 4.3. S2: Include Data and Tag in One Flit The advantage of this approach is that the receiver only needs one ﬂit for the veriﬁca- tion of integrity. However, if we do not want to increase the ﬂit size, the tag size is a problem. Adding a tag of 64 bits would imply an increase of the ﬂit size by 45.4% for UC and by 42.9% for NC. Nevertheless, the block size should not be smaller for security reasons. y Therefore, we decided to use an authentication code [30] as an alternative. As shown in Table 1, this authentication code requires two randomly selected key bits for the au- thentication of one message bit. For each message and tag bit, two possible keys remain. Hence, when an attacker wants to modify an intercepted ﬂit, he can only guess with a probability of 0.5 the correct key bits for every single bit he wants to change. Authentication of x message bits requires a stream of 2x key bits. Thus, 64 key bits can be used for the authentication of 32 message bits. The resulting 32 tag bits can be put together with the 32 data bits in the data ﬁeld of one ﬂit. Table 1. Example for the authentication of one bit. key bits 00 01 10 11 message bit 0 0 0 1 1 1 0 1 0 1 Table 1. Example for the authentication of one bit. Veriﬁcation requires the generation of the identical stream of key bits at the receiver side, even if previous ﬂits are lost. Encryption of all ﬁelds but the ﬁeld allows pseudo- randomly generating the necessary key bits. Uncoded transmission: The sender splits each 64-bit data block into two blocks containing only 32 bits. These blocks are distributed to two ﬂits where they are stored as the ﬁrst half of the data ﬁeld. The other ﬁelds of these ﬂits are equal despite the ﬂit ID that indicates which ﬂits belong together. For UC, there are 77 bits of metadata that need to be encrypted to generate the 64-bit key for authentication. Hence, there are two input blocks for encryption, the second one padded with zeros. The block cipher is used in CBC mode, and the second ciphertext block serves as the key. The actual computation of the tag bits can be done by a simple look-up in Table 1 with negligible effort. 4.4. Security Analysis: Integrity and Availability in Case of Losses and Modiﬁcations The goals of the proposed protocols are to ensure integrity and to increase availability in NoCs in presence of an active attacker who modiﬁes or drops ﬂits. Integrity means that data is correct or that it is unnoticeably not the case. In other words, there must be no undetectable modiﬁcations of transmitted data. In the proposed protocols, protection of integrity bases on the use of tags. Given that the cryptographic primitives used are secure, an attacker is not able to compute a valid tag for modiﬁed data without knowledge Electronics 2021, 10, 238 11 of 31 11 of 31 of the symmetric key shared between sender and receiver. Hence, any modiﬁcation will be detected by the receiver. The attacker may also change address information so that the receiver uses the wrong symmetric key, but of course, this will also result in a failed veriﬁcation of the tag. Hence, integrity will be guaranteed—modiﬁed ﬂits are discarded so that there is no chance for an attacker to inﬂuence subsequent processing by injecting falsiﬁed data. of the symmetric key shared between sender and receiver. Hence, any modiﬁcation will be detected by the receiver. The attacker may also change address information so that the receiver uses the wrong symmetric key, but of course, this will also result in a failed veriﬁcation of the tag. Hence, integrity will be guaranteed—modiﬁed ﬂits are discarded so that there is no chance for an attacker to inﬂuence subsequent processing by injecting falsiﬁed data. The second goal is to increase availability. Of course, availability includes integrity, thus, this protection goal is inﬂuenced by both modiﬁcations and losses. Three requirements need to be fulﬁlled in order to ensure that ﬂits sent are available at the receiver: (R1) The receiver must be able to detect losses and modiﬁcations, (R2) the receiver must be able to issue useful ARQs (an ARQ is useful if it triggers the retransmission of the needed ﬂit(s)), and (R3) the retransmission must be successful. Of course, it is not possible to enforce availability if there is an attacker that is able to massively disturb the system so that at least one of these requirements cannot be met. However, this is a general limitation and not speciﬁc for the protocols proposed in this paper. Under the assumption that the attacker tries to hide his activities, there will be a limited number of losses and modiﬁcations. 4.4. Security Analysis: Integrity and Availability in Case of Losses and Modiﬁcations In this case, the suggested protocols still allow transmiting data. g Availability in case of losses: Losses will be recognized by means of timers (R1). The receiver always starts a timer after the arrival of a ﬂit, in case of a timeout, an ARQ is issued. Availability in case of losses: Losses will be recognized by means of timers (R1). The receiver always starts a timer after the arrival of a ﬂit, in case of a timeout, an ARQ is issued. A special case is the loss of the ﬁrst ﬂit of a transmission unit, but this will be detected as well: A special case is the loss of the ﬁrst ﬂit of a transmission unit, but this will be detected as well: • S1 UC: The receiver will immediately recognize this loss. Since the order of ﬂits is not changed during transmission, the arrival of a tag ﬂit indicates the loss of the corresponding data ﬂit. p g • S1 NC: The same applies here, the arrival of a tag ﬂit indicates the loss of the corre- sponding linear combination. If both data and tag ﬂit are dropped, decoding may still be possible due to the included redundancy, otherwise, an ARQ will be issued. p y • S2 UC: The ﬂit identiﬁer of the second ﬂit indicates the loss of the ﬁrst ﬂit. • S2 NC: Detection of loss is not possible in this case; but equivalently to S1 NC, either the redundancy is sufﬁcient or an ARQ is triggered by the timer. • S2 NC: Detection of loss is not possible in this case; but equivalently to S1 NC, either the redundancy is sufﬁcient or an ARQ is triggered by the timer. Hence, losses will be detected given that at least one ﬂit of a transmission unit arrives at the receiver. If there is only loss, the receiver is always able to specify the ﬂit to be retransmitted, i.e., to send a useful ARQ. In S2 NC, the receiver cannot directly specify the lost ﬂit; however, the ﬂits already received. A limitation of the second requirement (R2) is given by the need to limit the number of possible ARQs in order to restrict the increase of the network load. However, the number of ARQs per transmission unit is a system parameter that can be set. 4.4. Security Analysis: Integrity and Availability in Case of Losses and Modiﬁcations The fulﬁllment of the third requirement (R3) depends on the size of the retransmis- sion buffer (a system parameter that needs to be set accordingly) and the success of the retransmission itself. The selection of another path is suggested to increase the chance to use a path without corrupted routers. Availability in case of modiﬁcations: As stated above, any modiﬁcation will be de- tected and the affected ﬂit(s) will be discarded. Hence, availability requires that the receiver is able to issue a useful ARQ if a modiﬁcation was detected (R2) and that the necessary ﬂit(s) can be successfully retransmitted (R3). The former condition requires a closer exami- nation of modiﬁcations of the ﬁelds contained in a ﬂit (Figure 2). We will not discuss the modiﬁcation of the burst bit since it is not used here. Hence, it is always set to zero. A modiﬁcation of the data ﬁeld that contains the data or the tag is the simplest case since the required ﬂits can be correctly speciﬁed in the ARQ. The same applies to the address ﬁeld. Modiﬁcations of the remaining ﬁelds need to be further considered; thereby, we assume that at least one ﬂit of a transmission unit is correctly transmitted: • Source address: If the modiﬁed source address is not valid for the given topology, there is no possibility to issue an useful ARQ. However, the receiver will recognize • Source address: If the modiﬁed source address is not valid for the given topology, there is no possibility to issue an useful ARQ. However, the receiver will recognize Electronics 2021, 10, 238 12 of 31 12 of 31 the loss of that ﬂit as described above. If the modiﬁed source address is valid, the ARQ will be sent to the wrong sender, a retransmission is not possible. However, the arrival of a further ﬂit at the correct recipient will imply the recognition of loss so that an ARQ is issued. • Target address: The incorrect receiver will issue an ARQ; the sender cannot ﬁnd the requested ﬂit(s) in its retransmission buffer and will discard the ARQ. However, detection of loss by the correct recipient will issue a useful ARQ. • Target address: The incorrect receiver will issue an ARQ; the sender cannot ﬁnd the requested ﬂit(s) in its retransmission buffer and will discard the ARQ. However, detection of loss by the correct recipient will issue a useful ARQ. 4.4. Security Analysis: Integrity and Availability in Case of Losses and Modiﬁcations • Mode: For S1, the mode can be changed from data to tag and vice versa. This will imply a detection of loss and the retransmission of the modiﬁed ﬂit. A change of data or tag mode to ARQ will imply that the receiver tries to select the requested ﬂit from its retransmission buffer what is of course not possible. However, if the other ﬂit, either data or tag, is not modiﬁed, a useful ARQ will be triggered due to the recognition of a loss. A change of mode ARQ to data or tag prevents that the intended retransmission is successful; the issued ARQ is not useful. For S2 there are only two modes data or ARQ The implications of changes are similar For S2, there are only two modes, data or ARQ. The implications of changes are similar to S1. For S2, there are only two modes, data or ARQ. The implications of changes are similar to S1. • FID/GID: For S1, veriﬁcation cannot be completed since two corresponding ﬂits are necessary for all communication schemes. A change of the FID or GID implies that the modiﬁed ﬂit seems to belong to another transmission unit, hence, the receiver recognizes two lost ﬂits and will issue two ARQs. Examples given, if the FID of the data ﬂit in case of S1/UC is modiﬁed, the receiver will issue one ARQ using the modiﬁed FID for the tag belonging to the modiﬁed data ﬂit and another ARQ using the correct FID for the data ﬂit belonging to the correct tag ﬂit. Only the latter is useful but this is sufﬁcient for a successful transmission. • FID/GID: For S1, veriﬁcation cannot be completed since two corresponding ﬂits are necessary for all communication schemes. A change of the FID or GID implies that the modiﬁed ﬂit seems to belong to another transmission unit, hence, the receiver recognizes two lost ﬂits and will issue two ARQs. Examples given, if the FID of the data ﬂit in case of S1/UC is modiﬁed, the receiver will issue one ARQ using the modiﬁed FID for the tag belonging to the modiﬁed data ﬂit and another ARQ using the correct FID for the data ﬂit belonging to the correct tag ﬂit. Only the latter is useful but this is sufﬁcient for a successful transmission. In case of S2, computation of the tag can start immediately. 4.4. Security Analysis: Integrity and Availability in Case of Losses and Modiﬁcations The redundancy included in NC is sufﬁcient for decoding if only one ﬂit was modiﬁed. Nevertheless, it is y In case of S2, computation of the tag will immediately start. The redundancy included in NC is sufﬁcient for decoding if only one ﬂit was modiﬁed. Nevertheless, it is possible to issue a useful ARQ by including the GEVs of the unmodiﬁed ﬂits. Given a limited number of losses or modiﬁcations, data can still be successfully transmitted. If there is only one modiﬁcation or loss per transmission unit, availability can be guaranteed (a change of the mode ﬁeld from ARQ to data or tag will not appear in this case). In some cases, unnecessary ARQs are issued what increases the network load, but transmission can be completed due to the recognition of losses. To support the selection of ﬂits from the retransmission buffer, ARQs specify required ﬂits as well as successfully received ﬂits. 4.4. Security Analysis: Integrity and Availability in Case of Losses and Modiﬁcations Basically, the change of the FID or GID also implies that the receiver treats the modiﬁed ﬂit as part of another transmission unit and recognizes loss for both received ﬂits. For the correct ﬂit, the receiver will detect loss and issue a useful ARQ. For the modiﬁed ﬂit, both loss of the corresponding (but not existing) ﬂit and modiﬁcation are detected. It depends on the timer for the recognition of loss as well as on the time needed for the computation of the tag what problem is detected ﬁrst. In both cases, the issued ARQ is not useful but also not necessary for a successful transmission. y GEV: In case of S1, there are always two ﬂits with the same GEV. Hence, a change of the GEV implies the detection of loss in two cases; veriﬁcation cannot be completed since the receiver does not have two corresponding ﬂits. The second ﬂit of the pair is the tag ﬂit. Since the order of ﬂits is not changed, the arrival of the tag ﬂit with a GEV that was not contained in the data ﬂit received before indicates the loss of a tag ﬂit and a data ﬂit. To enable the sender to select the correct ﬂit for retransmission, we consider to include in the ARQ the GEVs and modes of already received ﬂits. In case of S2, computation of the tag will immediately start. The redundancy included in NC is sufﬁcient for decoding if only one ﬂit was modiﬁed. Nevertheless, it is possible to issue a useful ARQ by including the GEVs of the unmodiﬁed ﬂits. y GEV: In case of S1, there are always two ﬂits with the same GEV. Hence, a change of the GEV implies the detection of loss in two cases; veriﬁcation cannot be completed since the receiver does not have two corresponding ﬂits. The second ﬂit of the pair is the tag ﬂit. Since the order of ﬂits is not changed, the arrival of the tag ﬂit with a GEV that was not contained in the data ﬂit received before indicates the loss of a tag ﬂit and a data ﬂit. To enable the sender to select the correct ﬂit for retransmission, we consider to include in the ARQ the GEVs and modes of already received ﬂits. In case of S2, computation of the tag will immediately start. 5.1. Parameters and Performance Metrics 5.1. Parameters and Performance Metrics The performance of our proposed authenticated schemes was evaluated by simula- tion in a cycle-accurate NoC simulator as well as using an analytical model. The speciﬁc Electronics 2021, 10, 238 13 of 31 13 of 31 simulation parameters and their corresponding values used in the investigations are sum- marized in Table 2. In all scenarios, the NoC was injected with a total average ﬂit injection rate of 0.2 ﬂits/module/cycle for all schemes. Due to the coding and communication schemes, the actual ﬂit injection rate is different from the ﬂit generation rate at the PE so that the latter must be adjusted to ensure an equal injection rate of 0.2 ﬂits/module/cycle. In UC, two ﬂits are generated from the original ﬂit (e.g., due to tag generation in S1 or due to data ﬂit halving in S2), so that the ﬂit generation rate at the PE was corrected to 0.1 ﬂits/module/cycle. In the NC schemes, to account for the redundant combinations as well as the tag ﬂit generation, the ﬂit generation rates at the PE were similarly adapted to 0.067 and 0.05 for G2C3 and G2C4 respectively. Table 2. Simulation parameters and their respective values. Topology 2D mesh of size 8 × 8 Routing Deterministic, dimension-ordered XY Arbitration Round-robin Injection rate (ﬂits/module/cycle) λ = 0.2 Communication models S1/{UC, G2C3, G2C4}, S2/{UC, G2C3, G2C4} Malicious routers 8 (at random locations in the 8 × 8 NoC) Modiﬁcation probability for a malicious router pm Drop probability for a malicious router pd Attack probability pa = wdpd + wmpm pa = 0.01 · i, i = 0, 1, . . . , 20 Loss detection timer 8 cycles Simulation run time 50,000 cycles Table 2. Simulation parameters and their respective values. In the evaluations, a limitation of the maximum number of retransmission and ARQs to 1 per logical transmission unit was applied to avoid very high loads that would saturate the network. This limitation also keeps the error control system simple but effective. It was assumed that ARQs can be dropped but not modified and hence ARQs are not authenticated. This assumption is reasonable since a modified ARQ would cause the retransmission of the flits unrelated to the desired logical transmission unit and therefore would have the same effect as if the ARQ had been lost. 5.1. Parameters and Performance Metrics The following performance metrics were considered: The following performance metrics were considered: Acceptance rate (A): also denoted as the network load, this is given by the total number of ﬂits (including redundant ﬂits due to NC, tag ﬂits, ARQs and retransmissions) injected per node in a clock cycle. Information rate (I):the ratio of actual data ﬂits transmitted to the total ﬂits transmitted, which includes the redundant ﬂits, tag ﬂits, ARQs, and retransmissions. Residual error probability (ϵ):The proportion of transmitted data ﬂits that failed to reach the destination, due to dropping and modiﬁcations, under the assumed limitation of ARQs and retransmission. Finally, the metric Latency was evaluated throughout the simulations and within the analytical model, which described the average path latency with respect to the varying pm and pd. However, there are some problems with latency results. Due to the rising drop and modiﬁcation probabilities, there will be more ﬂits lost over the course of the experiments, which ﬁnally will lead to the loss of complete transmission units. Since this loss cannot be detected, the overall load in the network will become smaller, which in turn results in lower path latencies, giving the false impression that the performance regarding latency would be better with increasing pm/pd. This effect was veriﬁed by the simulations as well Electronics 2021, 10, 238 14 of 31 14 of 31 as the analytical model. Due to this diminished information value, we do not consider the latency throughout the upcoming discussion. Nevertheless, the latency results are presented for the base and analytical model to show the mentioned effects. 5.2. Simulation Scenarios 5.2. Simulation Scenarios Simulations were conducted in different overall scenarios. In a single scenario, a key component of the simulation is changed, to inspect the respective impact on the solutions and the key metrics. The following scenarios are considered: Base The base scenario, which uses the simulation parameters as described in Table 2 and the weights wd = wm = 0.5 for the sum pa. No drops In this scenario, the attacker does not perform any dropping attacks but solely modiﬁcation attacks. This means the weighted sum of pa is calculated with weights wd = 0 and wm = 1, hence pa = pm. Therefore, the impact of modiﬁcations can be further evaluated. No modiﬁcations The same as above but reversed: the rogue routers will not perform modiﬁcations, but solely dropping attacks, i.e., wd = 1, wm = 0 and thereby pa = pd. No modiﬁcations The same as above but reversed: the rogue routers will not perform modiﬁcations, but solely dropping attacks, i.e., wd = 1, wm = 0 and thereby pa = pd. More attackers To further investigate the effect of different numbers of attacking nodes, the number of rogue routers will be varied. In this case, the number of attackers will be increased from formerly 8 to here 16. More attackers To further investigate the effect of different numbers of attacking nodes, the number of rogue routers will be varied. In this case, the number of attackers will be increased from formerly 8 to here 16. Fewer attackers In this scenario, the number of attacking routers will be decreased to 4. Fewer attackers In this scenario, the number of attacking routers will be decreased to All parameters not explicitly mentioned will be kept constant in the single scenarios. All parameters not explicitly mentioned will be kept constant in the single scenarios. 5.3. Simulation Results The proposed authentication schemes were evaluated by cycle-accurate simulations in a C++ simulation framework [31] extended to include the authentication schemes and NC, including delays implied by network coding and cryptographic primitives. The different performance parameters were evaluated in response to a varying attack probability pa. To remove the effect of the locations of the malicious routers within the NoC, the random position of the malicious routers were varied over 1000 iterations and then the obtained results were averaged. The results of the base simulation scenario are depicted in Figure 5. The effect on the acceptance rate is shown in Figure 5a, in which the curves rise from a starting value of 0.2 at 0.0 attack probability as the ARQs and retransmissions increase to tackle the drops and modiﬁcations as the attack probability increases to 0.2. The effect is more severe in the UC scheme in which the lack of redundancy results in ARQ and retransmission for each ﬂit loss or modiﬁcation. In comparison, NC due to the inherent redundancies requires retransmissions only when the generation could not be completed at the receiver. G2C4 performs better than G2C3 since it has greater redundancy than G2C3. S1 has a lower performance in comparison to S2 since S1 requires greater retransmissions. This is because due to the separate transmission of the data ﬂit and the corresponding tag ﬂit, not only is the susceptibility to attack increased but also because when one of them is modiﬁed, both of them must be retransmitted since since the receiver cannot decide which was modiﬁed. Figure 5b depicts the information rate with respect to the attack probability. As the ARQs and retransmissions increase, the total rate of information transmitted is decreased and for the same reasons as mentioned before, S2 performs better than S1. 5.3. Simulation Results 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.205 0.21 0.215 0.22 0.225 0.23 0.235 0.24 0.245 0.25 Acceptance rate (flits/node/cycle) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC 06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 6 0.08 0.1 0.12 0 Flit attack probability 6 0.08 0.1 0.12 0 Flit attack probability (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.02 0.04 0.06 0.08 0.1 0.12 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability (b) Information Rate (b) Information Rate (a) Acceptance Rate (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 50 60 70 80 90 100 110 120 Latency (cycles) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 Flit attack probability (c) Residual error probability (c) Residual error probability Figure 5. Simulation results for the base scenario: 8 × 8 2D mesh with 8 attacking routers. The residual error probability (Figure 5c) starts from 0 and increases with increasing attack probability, as fewer logical transmission units are received at the destination. In general, S1 has higher error probabilities, because of its higher attack susceptibility. For the greatest attack probability of pa = 0.2, the lowest error probability of ≈0.014% is achieved by S2/G2C4. The redundancy of the NC schemes displays its advantage over the UC schemes; however, from the curve of S1/G2C3, it can be observed that the redundancy of G2C3 is insufﬁcient at higher error rates to counter the higher attack susceptibility so that the error performance of S1/G2C3 degrades rapidly and becomes worse than UC at attack probabilities pa = 0.1 and pa = 0.13 for S1 and S2 respectively. The further simulation scenarios in general supported these results. A complete set of ﬁgures depicting the results of these runs can be found in Appendix A. 5.3. Simulation Results 15 of 31 Electronics 2021, 10, 238 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.205 0.21 0.215 0.22 0.225 0.23 0.235 0.24 0.245 0.25 Acceptance rate (flits/node/cycle) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.02 0.04 0.06 0.08 0.1 0.12 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 50 60 70 80 90 100 110 120 Latency (cycles) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (d) Latency Figure 5. Simulation results for the base scenario: 8 × 8 2D mesh with 8 attacking routers. 5.3. Simulation Results The scenario no drops mainly affects the UC cases, whereas the NC solutions mostly deliver the same results as in the base scenario. The most obvious difference is an increased acceptance rate for the UC cases, which are ≈4% higher than in the base scenario. As acceptance and information rate are tightly coupled, the information rate for these cases drop more than in the base scenario. In the no modiﬁcations scenario the most prominent change is that both UC cases now follow the same slope for all metrics. The diagrams also show that the acceptance rate is in general a bit lower for all schemes. The biggest decrease can be observed for S1 UC. A closer look at the impact of drop and modiﬁcation delivers an explanation for these results. In case of a drop, only the missed ﬂit needs to be retransmitted. In contrast, a Electronics 2021, 10, 238 16 of 31 16 of 31 modiﬁcation always requires to retransmit data and tag, i.e., two ﬂits in case of S1. Conse- quently, a drop in S1 UC has the same effect as a drop in S2 UC since there is no redundancy. The differences between the NC schemes stem from the different redundancy (differences between G2C3 and G2C4) and from possible cases that may cause a retransmission (differ- ences between S1 and S2). The analytical model provides a deeper insight into the reasons for these differences. The two scenarios varying the number of attacking routers, more attackers and fewer attackers, have straight forward effects on the metrics: Reducing the number of attackers from 8 to 4 more than halves the residual error probability while the overall shape of the curves is pertained. For pa of 0.2 the highest residual error probabilities (S1/UC and S1/G2C3) are reduced from over 0.1 to ≈0.04 (0.045 resp.). The worst solution in the base scenario (S1/UC) added 0.05 to the acceptance rate—with 4 attackers only 0.025 are added, while again pertaining the overall slope. Finally, the same holds for the information rate, where the attackers negative effect is approximately halved. For more attackers, the effects are comparable: the residual error probability is ap- proximately doubled for all solutions. 5.3. Simulation Results The acceptance rate is not affected in the same linear way: for pa = 0.2, the added impact for S1/UC is not increased by 100% to 0.1 but solely by 50% to ≈0.07 since we limited the number of ARQs in the simulation runs. Similar results are obtained for the information rate, where the impact is increased by a factor of 1.5. Overall, the extend evaluation with more diverse scenarios validates the results from the base scenario: the network coded approaches are signiﬁcantly more robust than the uncoded transmission. Among the coded solutions, S2 outperformed S1, especially in terms of residual error probability and acceptance rate. 5.4.1. S1 Authentication Scheme In S1 authentication scheme, ﬂits are always transmitted in pairs, the ﬁrst ﬂit con- taining the actual data and the second containing the authentication tag for the data ﬂit. Both of these must be received for authentication to occur. When a ﬂit loss is detected, that ﬂit must be retransmitted. The expression d′ b,ad′ a,bm′ a,b gives the probability that (in the case of a ﬂit drop from a to b), the ARQ arrived successfully (with probability d′ b,a) and the retransmitted ﬂit was not dropped or modiﬁed (d′ a,bm′ b,a). Thus, the probability that the ARQ or the retransmission was unsuccessful, denoted as Ra,b, is given by 1 −d′ b,ad′ a,bm′ a,b. If authentication fails, an ARQ is issued requesting the retransmission of both data and tag ﬂit as it is impossible to determine which was modiﬁed. In this case, the probability that the ARQ or the retransmission was unsuccessful, denoted as Ta,b must take into account that 2 are retransmitted and is thus given by 1 −d′ b,ad′ a,b 2m′ a,b 2. 5.4. Analytical Model An analytical model gives us a deeper insight into the system behavior so that we can understand how the performance is affected by the different factors. Furthermore, with an analytic model it possible to investigate with greater ﬂexibility different scenarios such as other topologies or various NoC sizes. In contrast to the extremely time-consuming cycle-accurate simulations, the model computes results signiﬁcantly faster (more than ×1000 faster for the 8 × 8 NoC), allowing investigating the performance of large NoCs that would be impossible to investigate through cycle-accurate simulations. In the following, we develop analytical expressions of Residual error probability (ϵ), Acceptance rate (A), Information rate (I) and Latency (ℓ) for S1 and S2. These will be then applied to compute the performance results for the 8 × 8 NoC and compared to that obtained from cycle-accurate simulations. Furthermore, as an application of the analytical model, we use it to determine the system performance of a very large NoC consisting of over a 1000 nodes when using S1 and S2 authentication schemes. The different parameter symbols used in the analytic model are summarized in Table 3. Certain symbols (such as the drop and modiﬁcation probabilities, pd and pm) were already introduced in Table 2 and is not repeated here. p The total ﬂit drop or the modiﬁcation probability between two modules are two quantities which are extensively used in the expressions of the performance metrics. This metric depends on the number of attacking routers, Na,b encountered along the XY route between two modules a and b: da,b = 1 −(1 −pd)Na,b (6) ma,b = 1 −(1 −pm)Na,b (7) (6) (7) 17 of 31 Electronics 2021, 10, 238 Table 3. Analytical model parameter symbols. Table 3. Analytical model parameter symbols. M Number of modules in the NoC λa,b Flit injection rate from module a to module b λ′ a,b Total ﬂit injection rate from a to b including ARQs and retransmissions ℓa,b Latency from module a to module b ℓ′ a,b Latency from a to b with retransmission Na,b Total number of attacking routers in the XY route from a to b da,b Total ﬂit drop probability from a to b d′ a,b = 1 −da,b Probability of no drop from a to b ma,b Total ﬂit modiﬁcation probability from a to b m′ a,b = 1 −ma,b Probability of no modiﬁcation from a to b ϵ Average residual error probability UC Transmission S1 Residual Error Probability: When one ﬂit (from a transmission unit of two ﬂits) is dropped ((2 1)da,bd′ a,b), an ARQ is issued to request a retransmission. However, if the ARQ or retransmission fails (with probability Ra,b), then error occurs. Error also occurs if both ﬂits are received but one or both are modiﬁed and the ARQ/retransmission fails (d′ a,b 2(1 −m′ a,b 2)Ta,b). The limitation of 1 ARQ per transmission unit further increases the error probability, e.g., when one ﬂit of a pair was dropped but the received ﬂit was modiﬁed ((2 1)da,bd′ a,bma,b). Here, an ARQ was already issued for the dropped ﬂit and since another ARQ cannot be issued for the modiﬁed ﬂit, error occurs. If both ﬂits are dropped (da,b 2), the receiver is not aware of the loss and does not issue an ARQ, resulting in error. UC Transmission S1 • error-free: latency = ℓa,b (NI injection and ejection delays + router traversal delays) + 2ℓtag (Authentication tag computation time at the sender and at the receiver). g p • with retransmission of a lost ﬂit: the loss of a ﬂit is detected by a timer tracking the inter-arrival delays of ﬂits and an ARQ is issued. If the ARQ and retransmis- sion is successful ,the retransmission reaches the receiver after a round trip delay (RTD) of 2ℓa,b plus some buffering delays (RTDa,b = 2ℓa,b + δ), after which the authentication occurs. • with retransmission of a modiﬁed ﬂit: the retransmitted data and tag ﬂit reach the destination after the RTD, if the ARQ and retransmission was successful (with prob- ability 1 −T). Here, there is an additional ℓtag cycles compared to the ﬂit drop case, since retransmitted data ﬂit must be veriﬁed again at the receiver. The total latency, ℓ′ b is given by: with retransmission of a modiﬁed ﬂit: the retransmitted data and tag ﬂit reach the destination after the RTD, if the ARQ and retransmission was successful (with prob- ability 1 −T). Here, there is an additional ℓtag cycles compared to the ﬂit drop case, since retransmitted data ﬂit must be veriﬁed again at the receiver. The total latency, ℓ′ a,b is given by: ℓ′ a,b = d′ a,b 2m′ a,b 2 · ℓa,b + 2ℓtag + 2 1 da,bd′ a,bm′ a,b(1 −Ra,b) · ℓa,b + RTDa,b + 2ℓtag + d′ a,b 2 1 −m′ a,b 2 (1 −Ta,b) · ℓa,b + RTDa,b + 2ℓtag + ℓtag . (14) The average latency is computed by averaging ℓ′ over all sender-receiver pairs. UC Transmission S1 By combining all these error scenarios and averaging over all source-destination pairs of the NoC (since we assumed uniform communication), we obtain the average residual error probability: ϵ = 1 M(M −1) M ∑ a=1 M ∑ b=1 b̸=a da,b 2 + 2 1 da,bd′ a,b(m′ a,bRa,b + ma,b) + d′ a,b 2(1−m′ a,b 2)Ta,b (8) (8) Acceptance Rate: The total ﬂit injection rate λ′ a,b at any module consists of the regular ﬂit injection, λa,b and the issued ARQs (λarq_a,b) and retransmissions (λretr_a,b): λ′ a,b = λa,b + λarq_a,b + λretr_a,b (9) λ′ a,b = λa,b + λarq_a,b + λretr_a,b (9) Electronics 2021, 10, 238 18 of 31 18 of 31 One ARQ is allowed per pair of ﬂits (so that rate of λarq is half in comparison to λ) and is issued by a module, a for a dropped ﬂit ((2 1)db,ad′ b,a) or for a ﬂit pair when authentication fails (d′ b,a 2(1 −m′ b,a 2)): λarq_a,b = λb,a 2 2 1 db,ad′ b,a + d′ b,a 2(1 −m′ b,a 2) (10) (10) In the above discussed cases, if the ARQ from a module b successfully arrives at a module a (with probability d′ b,a), the missing ﬂit (or ﬂits in the case of authentication failure) is retransmitted: λretr_a,b = λa,bd′ b,a 2 2 1 da,bd′ a,b + 2d′ a,b 2(1 −m′ a,b 2) (11) (11) The acceptance rate is computed by averaging the total ﬂit injection rate λ′ over all modules: 1 M M A = 1 M · M ∑ a=1 M ∑ b=1 b̸=a λ′ a,b (12) (12) Information Rate: The information rate is deﬁned as the proportion of data ﬂits to all transmitted ﬂits (data, ARQ, retransmission and tag ﬂits) and is therefore computed by the ratio of λa,b 2 (half of λa,b are tag ﬂits) to λ′ a,b using Equations (9)–(11): I = 1 2 ∑M a=1 ∑M b=1 b̸=a λa,b ∑M a=1 ∑M b=1 b̸=a λ′ a,b (13) (13) Latency: The latency is computed by considering those ﬂits which reached the des- tination successfully, within the limit of 1 retransmission. There are three possible cases for these: • error-free: latency = ℓa,b (NI injection and ejection delays + router traversal delays) + 2ℓtag (Authentication tag computation time at the sender and at the receiver). NC Transmission S1 Residual Error Probability: Here, C pairs of data and tag ﬂits are transmitted at the sender. Depending on how many unmodiﬁed ﬂits are received at the destination, different error cases can arise, as summarized in Table 4. 19 of 31 Electronics 2021, 10, 238 Table 4. Error cases for NC transmission with S1 authentication scheme. Number of Flits Received, n Possible Flit Combinations The Cases, Where Error Occurs n < 2G −1 ∑2G−2 n=0 (2C n ) Always, as 1 ARQ is insufﬁcient to complete G pairs. n = 2G −1 G −1 pairs + 1 f lit, ( 2G−1 G−1 pairs):( C G−1)(2C−2(G−1) 1 ) If any received ﬂit is modiﬁed or issued ARQ/retransmission fails, with probability R. < G −1 pairs, ( 2G−1 < G−1 pairs):( 2C 2G−1) −( C G−1)(2C−2(G−1) 1 ) Always, as 1 ARQ is insufﬁcient to complete G pairs. n = 2G ∽2C with k number of pairs k ≥G pairs, ( n ≥G pairs):∑n/2 k=G (C k)( C−k n−2k)(2 1) n−2k If less than G unmodiﬁed pairs are received, even with ARQ and retransmission k = G −1 pairs, ( n G−1 pairs):( C G−1)( C−(G−1) n−2(G−1))(2 1) n−2(G−1) If any of the received ﬂits are modiﬁed or issued ARQ/retransmission fails, with probability T. k < G −1 pairs, ( n < G−1 pairs):(2C n ) −( n ≥G pairs)− ( n G−1 pairs) Always, as 1 ARQ is insufﬁcient to complete G pairs. Table 4. Error cases for NC transmission with S1 authentication scheme. Table 4. Error cases for NC transmission with S1 authentication scheme. Table 4. Error cases for NC transmission with S1 authentication scheme. The Cases, Where Error Occurs n = 2G ∽2C with k number of pairs By combining these error cases, we obtain the residual error rate, ϵa,b for a transmission from module a to module b. • 0 < n ≤2G −1. Here, 1 ARQ is issued for a missing ﬂit. NC Transmission S1 We obtain the average residual error probability, ϵ after averaging ϵa,b over all sender-receiver pairs: ϵa,b = ϵa,b_<2G−1 + ϵa,b_2G−1 + ϵa,b_≥2G−1 (15) ϵa,b_n<2G−1 = 2G−2 ∑ n=0 2C n d′ a,b nda,b 2C−n (16) (15) (16) ϵa,b_n=2G−1 = d′ a,b 2G−1da,b 2C−(2G−1) 2G −1 G −1 pairs m′ a,b 2G−1Ra,b + 1 −m′ a,b 2G−1 + 2G −1 < G −1 pairs (17) −1 = d′ a,b 2G 1da,b 2C (2G 1) 2G −1 G −1 pairs m′ a,b 2G−1Ra,b + 1 −m′ a,b 2G−1 + 2G −1 < G −1 pairs (17) (17) ϵa,b_n≥2G−1 = 2C ∑ n=2G d′ a,b nda,b 2C−n n k ≥G pairs ( k G −1 m′ a,b 2(G−1) 1 −m′ a,b 2k−(G−1) Ta,b + G−2 ∑ t=0 k t m′ a,b 2t 1 −m′ a,b 2k−t ) + n G −1 pairs m′ a,b 2G−1Ra,b + 1 −m′ a,b 2G−1 + n < G −1 pairs (18) (18) Acceptance Rate: By considering the error cases in Table 4, we can accordingly deduce the issue of ARQs depending on the number of ﬂits received, n : • 0 < n ≤2G −1. Here, 1 ARQ is issued for a missing ﬂit. • 0 < n ≤2G −1. Here, 1 ARQ is issued for a missing ﬂit. Electronics 2021, 10, 238 20 of 31 20 of 31 The rate of ARQ (λarq) is 1 2C of λ as 1 ARQ or retransmission is allowed per generation: λarq_a,b = λb,a 2C " 2G−1 ∑ n=1 2C n d′ b,a ndb,a 2C−n + 2C ∑ n=2G d′ b,a ndb,a 2C−n· ( n k ≥G pairs G−1 ∑ t=0 k t m′ b,a 2t 1 −m′ b,a 2k−t + n < G pairs )# (19) (19) If the ARQ reaches the target without being dropped, the retransmission of the requested ﬂit (or ﬂits in case of modiﬁcation) is done. NC Transmission S1 Similar to ARQs, the rate of retrans- mission (λretr) is 1 2C of λ: λretr_a,b = λa,b 2C d′ b,a " 2G−1 ∑ n=1 2C n d′ a,b nda,b 2C−n + 2C ∑ n=2G d′ a,b nda,b 2C−n· ( 2 · n k ≥G pairs G−1 ∑ t=0 k t m′ a,b 2t 1 −m′ a,b 2k−t + n < G pairs )# (20) (20) Putting Equations (19) and (20), in Equations (9) and (12), we obtain the average acceptance rate. Putting Equations (19) and (20), in Equations (9) and (12), we obtain the average acceptance rate. Information Rate: The average information rate can be determined from ratios of λa,b to λ′ a,b, with factor 1 2 · G C to account for coding as well as for the tag ﬂits: I = 1 2 · G C ∑M a=1 ∑M b=1 b̸=a λa,b ∑M a=1 ∑M b=1 b̸=a λ′ a,b (21) (21) Latency: The regular path latency of NC transmission, ℓNCa,b includes some additional delays in comparison to the UC case such as waiting for G ﬂits at the sender and at the receiver for encoding and decoding. Thus, the latency components are: • error free case (G or more pairs of unmodiﬁed ﬂits are received): ℓNCa,b + 2ℓtag. ℓ′ NCa,b = ℓNCa,b + 2ℓtag · 2C ∑ n=2G ( d′ a,b nda,b 2C−n n k ≥G pairs k ∑ t=G k t m′ a,b 2t(1 −m′ a,b 2)k−t ) + ℓNCa,b + RTDa,b + 2ℓtag · 2C ∑ n=2G−1 d′ a,b nda,b 2C−n n G −1 pairs m′ a,b 2(G−1)+1(1 −Ra,b) + ℓNCa,b + RTDa,b + 2ℓtag + ℓtag · 2C ∑ n=2G d′ a,b nda,b 2C−n n k ≥G pairs · k G −1 m′ a,b 2(G−1)(1 −m′ a,b 2) k−(G−1)(1 −Ta,b) (22) (22) Relation between S1/NC and S1/UC equations: Relation between S1/NC and S1/UC equations: Relation between S1/NC and S1/UC equations: When considered, it is apparent that UC S1 is the G1C1 version of NC transmission since the transmission unit consists of only 1 data ﬂit (and its corresponding tag ﬂit). In NC, C pairs are transmitted and G pairs are needed for decoding and veriﬁcation at the receiver, which is valid also for UC with G1C1. This relation is apparent in the equations, e.g., when putting G = 1, C = 1 Equations (15)–(18) (residual error probability S1/NC), we ﬁnd that we arrive exactly at residual error probability of S1/UC (Equation (8)). Similarly, we obtain the equations of UC acceptance rate, information rate and latency by putting G = 1, C = 1 in the respective equations of NC. 21 of 31 21 of 31 Electronics 2021, 10, 238 5.4.2. S2 Authentication Scheme 5.4.2. S2 Authentication Scheme In S2 authentication scheme, the data is split over two ﬂits occupying half of the data ﬁeld. The remaining half of the data ﬁeld is used for transmitting the tag for the data. Thus, each ﬂit can be authenticated individually without depending on another ﬂit, in contrast to S1 authentication scheme. However, both ﬂits must be received (unmodiﬁed) in order to retrieve the original data. If a ﬂit is lost or modiﬁed, an ARQ is issued to request the retransmission of this ﬂit only. Similar to S1, the number of ARQs and retransmissions is limited to 1 per transmission unit. The expression Ra,b = 1 −d′ b,ad′ a,bm′ a,b gives the proba- bility that in case of a ﬂit loss or modiﬁcation, the ARQ is dropped or the retransmission is either dropped or modiﬁed. UC Transmission S2 Residual Error Probability: Error occurs if both ﬂits are dropped (da,b 2) or if one ﬂit is received (2 1)da,bd′ a,b but the ARQ/retransmission for the missing ﬂit fails (with probability R). However, if the received ﬂit was modiﬁed, then no further ARQs can be issued and so error occurs, regardless whether the ARQ/retransmission was successful or not. If both ﬂits are received, but one is modiﬁed d′ a,b 2(2 1)m′ a,bma,b then error occurs if the ARQ/retransmission fails. If both received ﬂits are modiﬁed, error occurs as only 1 ARQ/retransmission can be issued. ϵ = 1 M(M −1) M ∑ a=1 M ∑ b=1 b̸=a da,b 2 + 2 1 da,bd′ a,b m′ a,bRa,b + ma,b + d′ a,b 2 2 1 m′ a,bma,bRa,b + ma,b 2 (23) (23) Acceptance Rate and Information Rate: An ARQ is issued whenever either of two ﬂits are missing or if both ﬂits were received but one or both of them are modiﬁed: λarq_a,b = λb,a 2 2 1 db,ad′ b,a + d′ b,a 2(1 −m′ b,a 2) (24) (24) If the ARQ reaches the target successfully i.e., without being dropped, a retransmission of the requested ﬂit is done. Thus, the rate of retransmissions is equal to the rate of ARQs, provided the ARQ is not dropped: λretr_a,b = d′ b,aλarq_b,a (25) (25) Using Equations (24) and (25) in Equations (9) and (12), we can obtain the acceptance rate. Similarly, we can use Equation (13) to determine the information rate. The factor 1 2 is also necessary here for determining the information rate to account for the splitting of a data over 2 ﬂits. Latency: At the sender, ℓtag cycles are required to compute the authentication tags in parallel for the 2 data parts. At the receiver, after individual authentication, the two halves of the data are combined and forwarded to the module. The latency components are: • error free case: latency = ℓa,b + 2ℓtag cycles. • error free case: latency = ℓa,b + 2ℓtag cycles. • with retransmission of a lost ﬂit: if the received ﬂit is not modiﬁed, the latency is increased by the round trip delay, RTD, provided the ARQ/retransmission are not dropped or modiﬁed. NC Transmission S2 In S2 NC transmission scheme, after splitting the data over 2 ﬂits (considered to be a generation, i.e., G = 2), these are linearly combined into C ﬂits. At the receiver, after authentication, G ﬂits are decoded to retrieve the original data. To compensate for lost or modiﬁed ﬂits, 1 ARQ/retransmission is allowed per generation . p g Residual Error Probability: Error occurs according to the number o Residual Error Probability: Error occurs according to the number of flits received, n • n < G −1, as 1 ARQ is insufﬁcient to complete the generation. G 1 • n < G −1, as 1 ARQ is insufﬁcient to complete the generation. • n = G −1 and these are all unmodiﬁed, but the ARQ/retransmission fails (m′ a,b G−1Ra,b) or if one or more ﬂits are modiﬁed (1 −m′ a,b G−1) because then more than 1 ARQ would be needed. k • n = G −1 and these are all unmodiﬁed, but the ARQ/retransmission fails (m′ a,b G−1Ra,b) or if one or more ﬂits are modiﬁed (1 −m′ a,b G−1) because then more than 1 ARQ would be needed. k • n ≥G but too many flits are modified, ∑G−2 k=0 (n k)m′ a,b kma,bn−k so that 1 ARQ is insufficient to complete the generation or there are G −1 unmodified flits but ARQ/retransmission fails (( n G−1)m′ a,b G−1ma,bn−(G−1)Ra,b). • n ≥G but too many flits are modified, ∑G−2 k=0 (n k)m′ a,b kma,bn−k so that 1 ARQ is insufficient to complete the generation or there are G −1 unmodified flits but ARQ/retransmission fails (( n G−1)m′ a,b G−1ma,bn−(G−1)Ra,b). A i ll d i i bt i th id l b bilit , Averaging over all sender-receiver pairs, we obtain the average residual error probability: ϵ = 1 M(M −1) M ∑ a=1 M ∑ b=1 b̸=a " G−2 ∑ n=0 C n d′ a,b nda,b C−n + C G −1 d′ a,b G−1da,b C−(G−1) m′ a,b G−1Ra,b + 1 −m′ a,b G−1 + C ∑ n=G C n d′ a,b nda,b C−n ( n G −1 m′ a,b G−1ma,b n−(G−1)Ra,b + G−2 ∑ k=0 n k m′ a,b kma,b n−k )# (27) (27) Acceptance Rate and Information Rate: An ARQ is issued requesting the retransmis- sion of a dropped ﬂit when less than G ﬂits were received. UC Transmission S2 • with retransmission of a lost ﬂit: if the received ﬂit is not modiﬁed, the latency is increased by the round trip delay, RTD, provided the ARQ/retransmission are not dropped or modiﬁed. • with retransmission of modiﬁed ﬂit: if the ARQ/retransmission is successful, the latency is increased by RTDa,b along with another ℓtag cycles for the authentication of the retransmitted ﬂit. Electronics 2021, 10, 238 22 of 31 ℓ′ a,b = d′ a,b 2m′ a,b 2 · ℓa,b + 2ℓtag + 2 1 da,bd′ a,bm′ a,b(1 −Ra,b) · ℓa,b + RTDa,b + 2ℓtag + d′ a,b 2 · 2 1 ma,bm′ a,b(1 −Ra,b) · ℓa,b + RTDa,b + 2ℓtag + ℓtag (26) (26) NC Transmission S2 When G or more ﬂits were received but less than G are found to unmodiﬁed, then an ARQ requesting the retransmis- sion of a modiﬁed ﬂit is issued. When the ARQ successfully reaches the destination, the retransmission of the requested ﬂit is done: λarq_a,b = λb,a C h ∑G−1 n=1 (C n)d′ b,a ndb,a C−n + ∑C n=G (C n)d′ b,a ndb,a C−n ∑G−1 k=0 (n k)m′ b,a kmb,an−ki (28) λretr_a,b = d′ b,aλarq_b,a (29) G λarq_a,b = λb,a C h ∑G−1 n=1 (C n)d′ b,a ndb,a C−n + ∑C n=G (C n)d′ b,a ndb,a C−n ∑G−1 k=0 (n k)m′ b,a kmb,an−ki (28) (28) λretr_a,b = d′ b,aλarq_b,a (29) (29) The factor G/C 2 needs to be included for the computation of the information rate to account for the data splitting as well as the encoding: The factor G/C 2 needs to be included for the computation of the information rate to account for the data splitting as well as the encoding: INC = G/C 2 ∑M a=1 ∑M b=1 b̸=a λa,b ∑M a=1 ∑M b=1 b̸=a λ′ a,b (30) (30) Latency: In the error free case, i.e., when G or more unmodiﬁed ﬂits were received, the total latency includes the usual path latency, ℓNCx,y as well as 2 × 26 cycles for the authentication tag computation. The latency is increased by the round-trip delay, RTD when G −1 unmodiﬁed ﬂits were received and the ARQ and retransmission is received successfully with probability 1 −R. Relation between S2/NC and S2/UC equations: Similar to S1, S2/UC appears to be the G2C2 version of S2/NC. However, it must be noted that there is no encoding here in contrast to NC where it is possible to encode 2 ﬂits to generate 2 combinations. Putting G = 2, C = 2 in Equation (27) (residual error probability S2/NC), we arrive exactly at residual error probability of S2/UC (Equation (23)). Similarly, we can also obtain the equations for UC acceptance rate, information rate and latency by putting G = 2, C = 2 in the respective equations of S2/NC. NC Transmission S2 When G or more ﬂits were received but only G −1 Electronics 2021, 10, 238 23 of 31 23 of 31 ﬂits were found to be unmodiﬁed, then the latency is increased by RTD as well as another 26 cycles for the authentication of the retransmitted ﬂit: ﬂits were found to be unmodiﬁed, then the latency is increased by RTD as well as another 26 cycles for the authentication of the retransmitted ﬂit: ℓ′ NCa,b = ℓNCa,b + 2ℓtag · ( C ∑ n=G C n d′ a,b nda,b C−n n ∑ t=G n t m′ a,b tma,b n−t !) + ℓNCa,b + RTDa,b + 2ℓtag · C G −1 d′ a,b G−1da,b C−(G−1)m′ a,b G−1ma,b C−(G−1)(1 −Ra,b) + ℓNCa,b + RTDa,b + 2ℓtag + ℓtag · ( C ∑ n=G C n d′ a,b nda,b C−n n G −1 m′ a,b G−1ma,b n−(G−1)(1 −Ra,b) ) (31) (31) Relation between S2/NC and S2/UC equations: 5.4.3. Results and Discussion The performance of the authentication schemes for the 8 × 8 NoC was evaluated with the analytical model. The results were averaged over 1000 different locations of attacking routers and it was found that these results matched closely with those obtained from the simulations. To evaluate the effectiveness of the analytical model, the maximum difference between the performance parameter results from the cycle-accurate and the analytical model simulations was determined. From the summary depicted in Table 5, it is evident that the analytical model matches very closely the cycle-accurate simulator, with a maximum relative error of 5%. The exception to this are the latency calculations, which in the simulations is affected by congestion in the NoC which, however, is not covered by the analytical model. As a result, the results obtained by simulation are greater than those by the analytical model, particularly for S1 G2C3 in which the relatively higher error rates result in many ARQs and retransmissions leading to greater congestion and delay. Due to its greater speed of calculation and accuracy, we use the analytical model next to compute the results for a large NoC of 1024 modules. Application of the model to 32 × 32 NoC: Application of the model to 32 × 32 NoC: The 32 × 32 NoC was investigated with an identical ratio of attacking routers as in the 8 × 8 NoC, so that there are 128 attacking routers in this scenario. The total ﬂit attack probability is similarly varied from 0 to 0.2 in steps of 0.01. The results were computed over 5000 different locations of attacking routers and the average results for the residual error probability, acceptance rate and information rate are displayed in Figure 6. p y p p y g To understand the performance results for the 32 × 32 NoC, we need to ﬁrst consider the average path length of such a NoC. With uniform random trafﬁc pattern, the average path length of the 32 × 32 NoC is 22.33 hops, whereas for the 8 × 8 NoC it is 6.33 hops. The probability of encountering an attacking router with average path lengths of 6 hops and 22 hops are: (32) Pr6hops(≥1 attacking routers) = 1 −56 64 × 55 63 × 54 62 × 53 61 × 52 60 × 51 59 = 0.5669 Pr22hops(≥1 attacking routers) = 1 −896 1024 × 895 1023 × 896 1022 × 877 899 × 876 898 × 875 897 = 0.9487 24 of 31 Electronics 2021, 10, 238 Table 5. Maximum relative error between the analytical model and simulation. Table 5. Maximum relative error between the analytical model and simulation. Table 5. Maximum relative error between the analytical model and simulation. Application of the model to 32 × 32 NoC: S1 S2 UC G2C3 G2C4 UC G2C3 G2C4 Base: Error probability 1% 3% 2% 1% 4% 4% pa = 0.5pd + 0.5pm Information rate <1% <1% <1% <1% <1% <1% Acceptance rate <1% <1% <1% <1% <1% <1% Latency 2% 7.5% 5% 3% 1% 1% No drops: Error probability 1% 3% 2% 1% 2% 5% pa = pm Information rate <1% <1% <1% <1% <1% <1% Acceptance rate <1% <1% <1% <1% <1% <1% No modiﬁcations: Error probability 1% 5% 1% 1% 3% 6% pa = pd Information rate <1% <1% <1% <1% <1% <1% Acceptance rate <1% <1% <1% <1% <1% <1% 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.1 0.2 0.3 0.4 0.5 0.6 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Residual error probability 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Information Rate Figure 6. Analytical model results for 32 × 32 2D mesh with 128 attacking routers. Since there is a signiﬁcantly higher probability of passing through an attacking router in the 32 × 32 NoC, we can expect signiﬁcantly higher rates of ﬂit drops and modiﬁcations. As a result, residual error rates (Figure 6a) are considerably higher in comparison to the 8 × 8 NoC. Similar to the performance for 8 × 8 NoC, in the 32 × 32 NoC S2 behaves better than S1: S2/G2C4 has less than half the residual error probability of S1/G2C4 at 0.2 ﬂit attack probability. For G2C3, S1 has a 87% higher residual error probability than S2 at 0.2 attack probability. Application of the model to 32 × 32 NoC: For UC case, the performance of S1 and S2 are similar, 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.1 0.2 0.3 0.4 0.5 0.6 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC 6 0.08 0.1 0.12 0 Flit attack probability 6 0.08 0.1 0.12 0 Flit attack probability (a) Residual error probability b) Acceptance Rate ) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Information Rate Figure 6. Analytical model results for 32 × 32 2D mesh with 128 attacking routers. Figure 6. Analytical model results for 32 × 32 2D mesh with 128 attacking routers. Since there is a signiﬁcantly higher probability of passing through an attacking router in the 32 × 32 NoC, we can expect signiﬁcantly higher rates of ﬂit drops and modiﬁcations. As a result, residual error rates (Figure 6a) are considerably higher in comparison to the 8 × 8 NoC. Similar to the performance for 8 × 8 NoC, in the 32 × 32 NoC S2 behaves better than S1: S2/G2C4 has less than half the residual error probability of S1/G2C4 at 0.2 ﬂit attack probability. For G2C3, S1 has a 87% higher residual error probability than S2 at 0.2 attack probability. For UC case, the performance of S1 and S2 are similar, Since there is a signiﬁcantly higher probability of passing through an attacking router in the 32 × 32 NoC, we can expect signiﬁcantly higher rates of ﬂit drops and modiﬁcations. As a result, residual error rates (Figure 6a) are considerably higher in comparison to the 8 × 8 NoC. Similar to the performance for 8 × 8 NoC, in the 32 × 32 NoC S2 behaves better than S1: S2/G2C4 has less than half the residual error probability of S1/G2C4 at 0.2 ﬂit attack probability. For G2C3, S1 has a 87% higher residual error probability than S2 at 0.2 attack probability. Application of the model to 32 × 32 NoC: For UC case, the performance of S1 and S2 are similar, Electronics 2021, 10, 238 25 of 31 25 of 31 which is already expected from Equations (8) and (23). In S1, it is observed that the redundancy of NC is no longer sufﬁcient since at the high attack rates it becomes less likely to receive G pairs of unmodiﬁed ﬂits, resulting in a worse performance than UC. At 0.2 attack probability G2C3 has a 18.6% higher error rate and G2C4 has approximately the same error rate as UC. Another reason for the lower performance of NC compared to UC in S1 is the limitation of the ARQs and retransmissions to one per transmission unit, which is C pairs of ﬂits for NC and one pair for UC. The effect of this can be observed in Figure 6b where the high error rates result in an increase in the issue of ARQs and retransmissions, increasing the average rates of ﬂit injection, i.e., the ﬂit acceptance rate. In S1, UC has a 25.8% and 33.68% higher ﬂit acceptance rate than G2C3 and G2C4 respectively. Between S1 and S2, the latter starts with lower acceptance rates. However, in the S2 coded cases the ARQ and retransmission rates increase rapidly with increasing attack probability becoming close to S1 at 0.2 attack probability. This results in a rapidly decreasing information rate as can be observed in Figure 6c so that for the coded cases, the difference between S1 and S2 becomes less than 5% at 0.2 attack probability. However, since with similar acceptance rate S2 achieves a much lower residual error rate, we can select S2 as the superior authentication scheme. In the S2 scheme, we ﬁnd that G2C4 has 31% lower error rate but also 18.5% lower information rate than G2C3 at 0.2 attack probability. This means that to transmit the same amount of data, G2C4 requires approximately 20% more transmissions than G2C3. This may be a point to consider when choosing a transmission scheme for latency critical applications especially if the probability of attack is low. 5.5. Area Overhead Unit Area per NI Crypto modules 18 × 2681 = 48, 258 GEs LUTs (for network coding) 7080 GEs (S1 G2C3) or 16,992 GEs (S1 G2C4) Retransmission buffer (depth = 10) 19 × 10 = 190 bytes Table 6. Overview of area overhead. The matrix multiplications in the GF domain performed in network coding (Section 3.2) also increase the area. To simplify the multiplication process in the GF domain, we use look-up tables (LUTs) in which all the results of the multiplication over GF(24) are stored. The LUT was implemented in Verilog hardware description language and functionally veriﬁed. When synthesized in 65 nm CMOS technology, each LUT had an area of 118 GEs. For each ﬂit in S1, 18 symbols of 4 bits each (2 symbols for the GEV and 16 symbols for the 64 bits data) need to be encoded. Due to the generation size of G = 2, 2 encoding coefﬁcients are necessary for the computation of one linear combination. Hence in S1, 18 × 2 or 36 LUTs are required to produce one combination. In S2, fewer symbols need to be encoded since the data block is only 32 bits (Section 4.3). The multiplication of this ﬂit of size 10 symbols with the 2 encoding coefﬁcients requires thus 20 LUTs. The NC scheme G2C4 has the greatest number of combinations and for this case in S1, we need a total of 4 × 36 or 144 LUTs. These LUTs incur a total area overhead of 144 × 118 or 16, 992 GEs in each NI. Our considered state-of-art MPSoC ([33]) has 10 NIs so that the area overhead incurred is 10 × 16, 992 GEs, which is an increase of 0.7% for the MPSoC. The decoding of the generation at the receiver requires a multiplication with the inverse of the 2 × 2 encoding matrix. This can be achieved easily via the determinant method. As shown in Section 3.2, the decoding process requires fewer multiplications as the matrices are smaller, so that further LUTs are not required. q We know that depending on their size, buffers can occupy large area and also consume signiﬁcant power. Buffers are used in the NI to store transmitted ﬂits so that the tag does not need to be generated again for a retransmission. 5.5. Area Overhead As can expected, securing the NoC communication incurs some area overhead. How- ever, as we demonstrate in this section, the overhead is a reasonable one. Area overhead results due to authentication as well as due to network coding. Moreover, buffers are used at each sender NI to store a copy of transmitted ﬂits to allow for retransmission when needed. The main contributors to the area overhead of the proposed schemes are summarized in Table 6. The mCrypton modules [27] contribute to a signiﬁcant area increase since a certain number of them is required in the NI to generate the authentication tags for the information ﬂits, injected to and from the NoC. In S1 scheme, each tag generation requires 39 cycles whereas in S2 scheme 26 cycles are needed. Data ﬂits injected into the NoC and data ﬂits incoming from the NoC (also including retransmitted ﬂits) are served by a number of crypto modules working in parallel. The number of crypto modules must be sufﬁcient so that the rate of ﬂits queuing up is balanced by the service rate of the crypto modules. We assumed the total average ﬂit injection at the sender side of the NI is 0.2 ﬂits/cycle. As we assumed uniform random communication, there is also an equal ﬂit injection into receiver side of the NI. Flits at both the sender and receiver side must be authenticated. However, in S1 half of the injected ﬂits are tag ﬂits. Thus, the total incoming rate of ﬂits for the tag generation queue ( denoted as λq) is 0.2 ﬂits/node/cycle. Similarly, in S2, λq = 0.4 ﬂits/node/cycle. However, exact value of λq is affected by drops and by (successful) retransmissions of ﬂits incoming from NoC. The total ﬂit incoming rate (from NoC) consists of ﬂits which are not dropped, either originally transmitted ﬂits or retransmitted ﬂits. 5.5. Area Overhead Considering UC case (which has the highest number of retransmissions), we can evaluate λq at a module a using Equation (11) and Equation (25) for S1 and S2 respectively: λq S1 = 0.1 + M ∑ b=1 b̸=a ( λb,a 2 d′ b,a 2 + λb,ad′ a,b 2 db,ad′ b,a 2 + d′ b,a 4(1 −m′ b,a 2) ) λq S2 = 0.2 + M ∑ b=1 b̸=a n λb,ad′ b,a + λretr_b,ad′ b,a o Electronics 2021, 10, 238 26 of 31 26 of 31 By averaging over all modules, we obtain the value of λq for S1 and S2 at different drop and modiﬁcation probabilities. With no drop probability i.e., pa = pm, the maximum value of λq is obtained: 0.2185 ﬂits/node/cycle for S1 and 0.4214 ﬂits/node/cycle for S2. q y y Using Erlang’s C formula [32], we next estimate the number of crypto modules needed so that probability that an incoming ﬂit ﬁnds all crypto modules busy and must be queued is less than 0.05. The service rate of the crypto modules for S1 and S2 are 1/39 ﬂits/cycle and 1/26 ﬂits/cycle respectively. Since S2 has the higher λq, we use this value in our estimation to ﬁnd that with 18 crypto modules, a service level greater than 96% is achieved. Even a slight reduction to 15 crypto modules decreases the service level to 81%. The area of each mCrypton unit as given in [27] is 2681 gate equivalents (GEs) so that for 18 mCryptons, the total area overhead is 18 × 2681 GEs. To determine the actual overhead of these cryptomodules, we compare it to the total area of a state-of-the-art MPSoC. We thus consider the MPSoC Tomahawk 4 [33] with total area 24.43 MGEs comprised of a hexagonal NoC connecting 6 processing modules in addition to a global memory. A total 10 NIs are present whose communication should be protected. Assuming 18 crypto modules per NI, this means an area overhead of 10 × 18 × 2681 GEs over 24.43 MGEs or only ≈1.98%. Table 6. Overview of area overhead. Unit Area per NI Crypto modules 18 × 2681 = 48, 258 GEs LUTs (for network coding) 7080 GEs (S1 G2C3) or 16,992 GEs (S1 G2C4) Retransmission buffer (depth = 10) 19 × 10 = 190 bytes Table 6. Overview of area overhead. 6. Summary and Outlook In this work, we proposed and thoroughly evaluated efﬁcient authentication schemes to protect NoC communication against active attacks. By combining the usage of MACs for authentication and network coding for performance and resilience, we devised secure, highly robust, and efﬁcient solutions. The evaluation of these new schemes is twofold: ﬁrst of all, we performed extensive simulations with an cycle accurate NoC simulator covering different system parameters and attacker scenarios. Additionally, we developed an analytical model which describes the main performance metrics in a formalized manner. Thereby, we were able to examine the performance of our proposed schemes in additional scenarios, which are unfeasible to simulate. Furthermore, the analytical models provides insight into the system behaviour. Finally, the impact of our solution regarding chip area was analyzed. Our evaluation showed, that the proposed solutions realize a robust protection scheme for NoC communication. This robustness is primarily rooted in the network coding: In the base scenario, the best solution S2/G2C4 reduces the residual error probability by up to ≈85.9% compared to uncoded UC solution. The acceptance rate reduction of ≈15.7% also reﬂects a more robust transmission as fewer ﬂits were transferred. Additionally, this means that the overall network load is reduced with the proposed coded schemes. In the best scenario the residual error probability is reduced by ≈90.06% and the acceptance rate by ≈22.13%. y The addition of authentication and robustness implies additional costs: First, despite using an efﬁcient lightweight block cipher, mCrpyton, as cryptographic primitive, the computation and veriﬁcation of the MAC each takes up 39/26 cycles for S1/S2, respectively. In contrast, the network coding implementation via LUTs has negligible impact on the latency. Second, the addition of network coding decreases the information rate, i.e., for successful transmission more ﬂits are sent per data ﬂit. For the most robust solution G2C4 this means 4 ﬂits instead of 2 in the uncoded case need to be send. The developed analytical model overall conﬁrmed the simulation results obtained in the different scenarios. Additionally, it showed that for a 32 × 32 NoC with 128 attacking routers matching results could be obtained: Although successful transmission becomes signiﬁcantly harder in this extreme setting, the coded solution S2/G2C4 provided the best results, with residual error rates reduced by ≈54.76% compared to the uncoded case. Overall, the solution S2/G2C4 consistently provided the most robust efﬁcient protec- tion over all scenarios and parameters. 5.5. Area Overhead To determine how large buffers are needed, let us consider for how long ﬂits should be stored in these buffers. This duration should be long enough so that when the ARQ arrives, the ﬂit is still present in the buffer. In our investigated scenarios, the ﬂits of a generation are monitored at the receiver using a timer that is restarted whenever a ﬂit of the generation arrives. A ﬂit is assumed to be missing and an ARQ is issued if no new ﬂits arrive within 8 cycles after the last ﬂit. The ARQ reaches the target after a total round trip delay plus the 8 cycles. In the 8 × 8 NoC, the highest distance between 2 nodes is 15 hops, considering XY routing. In our NoC, each hop required 2 cycles so that the ARQ reaches the target node after a total delay of 2 × 30 + 8 Electronics 2021, 10, 238 27 of 31 27 of 31 or 68 cycles. With a retransmission buffer of size Nb ﬂits deep and an injection rate of 0.2 ﬂits/node/cycle, the original ﬂit is stored in the buffer for 5 · Nb cycles on average before it is overwritten. For the original ﬂit to be found when the ARQ arrives, the ﬂit must be in the buffer for at least 68 cycles, i.e., 5 · Nb ≥68. Thus, with Nb = 68 5 or ∼14, the original ﬂit can be found. A ﬂit modiﬁcation is detected after 26 or 39 cycles after receiving the ﬂit, after which an ARQ is issued. Thus, the ARQ will reach the sender 2 × 30 + 39 or 2 × 30 + 26 cycles respectively after the original ﬂit was transmitted. Thus, for the original ﬂit (or ﬂits in S1) still to be present in the buffer, the ﬂit must be present for at least 99 or 86 cycles. Thus, 5 · Nb ≥99 or 86, i.e., Nb ∼20 or 18 ﬂits deep. This demonstrates that the buffer required to store transmitted ﬂits is very small. Our reference MPSoC has a very small network size where the nodes have a maximum distance of 3 hops so that even smaller sized buffers are necessary. The effect of this buffer on the total area can be considered insigniﬁcant. Thus, in total the area overhead is (1.98 + 0.7)% or 2.68%. 6. Summary and Outlook With this approach, we provide efﬁcient detection of active attacks. Moreover, the redundancy provided by network coding is very effective against dropping and modiﬁcation of ﬂits. This enhanced robustness and additional security can be achieved with a minimal area overhead of ≈2.68% in comparison to the total area of a state-of-the-art MPSoC. Electronics 2021, 10, 238 28 of 31 28 of 31 Among many future research topics addressing the protection of NoC communication is the investigation of efﬁcient key management for symmetric block ciphers. Current protection schemes do explicitly exclude key management and assume pairwise symmetric keys. The distribution and management of these secrets pose a demanding issue, since there is no out-of-band medium available. Therefore, the untrusted medium and endpoints need to be used to securely distribute symmetric keys. y y y Another possible topic for future work could be the investigation of the proposed schemes using different cryptographic primitives. A promising contender can be PRINCE [34] as proposed in [35]. Its main advantage is the performance of 1 cycle per block, which would significantly reduce the latency of the current solution. Furthermore, the number of required cryptographic modules and their respective queues could be reduced. Although a PRINCE module has a greater area compared to mCrypton, the reduced total number of these required could result in a comparably similar or even lower area overhead. Another intended way forward is to further analyze the trade-offs and system parame- ters of the communication scheme: The application of more sophisticated routing schemes, e.g., Valiant [36] or ROMM [37], and multipath routing could lead to enhanced robustness against attackers. In fact, the analytical model is ﬂexible and already applicable to multi- path routing. Furthermore, the impact of the retransmission solution will be analyzed—the retransmission limit could be altered or removed to allow for additional retransmission and the usage of ARQs could be combined with an ACK-based solution. Finally, the application of burst mode for message transmission in the NoC and its implications for the proposed secure protocols will be investigated. Author Contributions: Conceptualization, S.M., E.F. and P.W.; methodology, S.M., E.F. and P.W.; software, S.M., E.F. and P.W.; validation, S.M., E.F. and P.W.; formal analysis, S.M.; investigation, S.M., E.F. and P.W.; resources, A.K. and T.S.; writing—original draft preparation, S.M., E.F. and P.W.; writing—review and editing, S.M., E.F., P.W. and A.K.; visualization, S.M., E.F. and P.W.; supervision, A.K.; funding acquisition, G.F. 6. Summary and Outlook All authors have read and agreed to the published version of the manuscript. Funding: This research was funded in part by the German Research Foundation (DFG) in the Collaborative Research Center 912 “Highly Adaptive Energy-Efﬁcient Computing” (HAEC). Data Availability Statement: Data sharing not applicable. Data Availability Statement: Data sharing not applicable. Acknowledgments: The authors gratefully acknowledge the GWK support for funding this project by providing computing time through the Center for Information Services and HPC (ZIH) at TU Dresden on the HPC-DA. Conﬂicts 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 interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Abbreviations The following abbreviations are used in this manuscript: ARQ automatic repreat request FID ﬂit identiﬁer GE gate equivalent GEV global encoding vector GID generation identiﬁer HT hardware Trojan LUT look-up table MPSoCs multi-processor systems-on-chip NC network coding NI network interface NoC Networks-on-Chip Electronics 2021, 10, 238 29 of 31 RTD round trip delay S1/S2 solution 1/solution 2 UC uncoded Appendix A. Additional Simulation Results Appendix A.1. Fixed Probabilities 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.19 0.2 0.21 0.22 0.23 0.24 0.25 0.26 0.27 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 Residual error probability (c) Residual error probability Figure A1. Simulation results for 8 × 8 2D mesh in scenario no drop, i.e., pa = pm. Appendix A. Additional Simulation Results Appendix A. Additional Simulation Results Appendix A.1. Fixed Probabilities Appendix A.1. Fixed Probabilities Appendix A.1. Fixed Probabilities 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.19 0.2 0.21 0.22 0.23 0.24 0.25 0.26 0.27 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 Residual error probability (c) Residual error probability (c) Residual error probability (c) Residual error probability (a) Acceptance Rate Figure A1. Simulation results for 8 × 8 2D mesh in scenario no drop, i.e., pa = pm. Abbreviations Different Number of Attackers 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.195 0.2 0.205 0.21 0.215 0.22 0.225 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability Figure A3. Simulation results for 8 × 8 2D mesh with 4 attacking routers. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.195 0.2 0.205 0.21 0.215 0.22 0.225 0.23 0.235 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate (c) Residual error probability (a) Acceptance Rate (b) Information Rate Figure A2. Simulation results for 8 × 8 2D mesh in scenario no modiﬁcation, i.e., pa = pd. Appendix A.2. Different Number of Attackers Appendix A.2. Different Number of Attackers Appendix A.2. Different Number of Attackers Appendix A.2. Different Number of Attackers 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.195 0.2 0.205 0.21 0.215 0.22 0.225 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability Figure A3. Simulation results for 8 × 8 2D mesh with 4 attacking routers. Abbreviations 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.195 0.2 0.205 0.21 0.215 0.22 0.225 0.23 0.235 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.02 0.04 0.06 0.08 0.1 0.12 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability Figure A2. Simulation results for 8 × 8 2D mesh in scenario no modiﬁcation, i.e., pa = pd. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.02 0.04 0.06 0.08 0.1 0.12 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.195 0.2 0.205 0.21 0.215 0.22 0.225 0.23 0.235 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.02 0.04 0.06 0.08 0.1 0.12 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability Figure A2. Simulation results for 8 × 8 2D mesh in scenario no modiﬁcation, i.e., pa = pd. Appendix A.2. Abbreviations 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.195 0.2 0.205 0.21 0.215 0.22 0.225 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate (c) Residual error probability (c) Residual error probability (b) Information rate (a) Acceptance Rate Figure A3. Simulation results for 8 × 8 2D mesh with 4 attacking routers. 30 of 31 30 of 31 Electronics 2021, 10, 238 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.19 0.2 0.21 0.22 0.23 0.24 0.25 0.26 0.27 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.05 0.1 0.15 0.2 0.25 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability Figure A4. Simulation results for 8 × 8 2D mesh with 16 attacking routers. Abbreviations 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Information rate S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (b) Information Rate 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0 0.05 0.1 0.15 0.2 0.25 Residual error probability S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (c) Residual error probability 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Flit attack probability 0.19 0.2 0.21 0.22 0.23 0.24 0.25 0.26 0.27 Acceptance rate (flits/cycle/node) S1 G2C4 S1 G2C3 S1 UC S2 G2C4 S2 G2C3 S2 UC (a) Acceptance Rate (c) Residual error probability (b) Information Rate Figure A4. Simulation results for 8 × 8 2D mesh with 16 attacking routers. References 1. Borkar, S. Thousand Core Chips: A Technology Perspective. In Proceedings of the 44th Annual Design Automation Conference, San Diego, CA, USA, 4–8 June 2007; pp. 746–749. 2. Benini, L.; De Micheli, G. Networks on chips: A new SoC paradigm. Computer 2002, 35, 70–78. [CrossRef] 3. Dally, W.; Towles, B. Route packets, not wires: On-chip interconnection networks. In Proceedings of the 38th Annual Design Automation Conference, Las Vegas, NV, USA, 22 June 2001; pp. 684–689. p p g p 3. Dally, W.; Towles, B. Route packets, not wires: On-chip interconnection networks. In Proceedings of the 38th Annual Design Automation Conference, Las Vegas, NV, USA, 22 June 2001; pp. 684–689. g pp 4. Jantsch, A.; Tenhunen, H. Will Networks on Chip Close the Productivity Gap? In Networks on Chip; Jantsch, A., Tenhunen, H., Eds.; Springer: Berlin/Heidelberg, Germany, 2004; pp. 3–18. g pp 4. Jantsch, A.; Tenhunen, H. Will Networks on Chip Close the Productivity Gap? In Networks on Chip; Jantsch, A., Tenhunen, H., Eds.; Springer: Berlin/Heidelberg, Germany, 2004; pp. 3–18. 5. Radetzki, M.; Feng, C.; Zhao, X.; Jantsch, A. Methods of Fault Tolerance in Networks-on-Chip. ACM Comput. Surv. 2013, 46, 1–38. [CrossRef] ao, K.; Forte, D.; Jin, Y.; Karri, R.; Bhunia, S.; Tehranipoor, M. Hardware trojans: Lessons learned after one d CM Trans Des Autom Electron Syst (TODAES) 2016 22 1 23 [CrossRef] 6. Xiao, K.; Forte, D.; Jin, Y.; Karri, R.; Bhunia, S.; Tehranipoor, M. Hardware trojans: Lessons learned afte 6. Xiao, K.; Forte, D.; Jin, Y.; Karri, R.; Bhunia, S.; Tehranipoor, M. Hardware trojans: Lessons learned after one decade of research. ACM Trans. Des. Autom. Electron. Syst. (TODAES) 2016, 22, 1–23. [CrossRef] h dh k h d f d , ; , ; J , ; , ; , ; p , j ACM Trans. Des. Autom. Electron. Syst. (TODAES) 2016, 22, 1–23. [CrossRef] S.; Waksman, A.; Suozzo, M.; Huang, Y.; Eum, J. Trustworthy Hardware from Untrusted Components. Comm 0–71. [CrossRef] 7. Sethumadhavan, S.; Waksman, A.; Suozzo, M.; Huang, Y.; Eum, J. Trustworthy Hardware from Untr ACM 2015, 58, 60–71. [CrossRef] 8. Shakya, B.; He, T.; Salmani, H.; Forte, D.; Bhunia, S.; Tehranipoor, M. Benchmarking of hardware trojans and maliciously affected circuits. J. Hardw. Syst. Secur. 2017, 1, 85–102. [CrossRef] y 9. Evain, S.; Diguet, J.P. From NoC Security Analysis to Design Solutions. References In Proceedings of the IEEE Workshop on Signal Processing Systems Design and Implementation, Athens, Greece, 2–4 November 2005. y , S.; Diguet, J.P. From NoC Security Analysis to Design Solutions. In Proceedings of the IEEE Workshop on Sig ms Design and Implementation, Athens, Greece, 2–4 November 2005. y 9. Evain, S.; Diguet, J.P. From NoC Security Analysis to Design Solutions. In Proceedings of the IEEE Works 9. Evain, S.; Diguet, J.P. From NoC Security Analysis to Design Solutions. In Proceedings of th Systems Design and Implementation, Athens, Greece, 2–4 November 2005. , S.; Diguet, J.P. From NoC Security Analysis to Design Solutions. In Proceedings of the IEEE Workshop on Sig ms Design and Implementation, Athens, Greece, 2–4 November 2005. g y y g g Systems Design and Implementation, Athens, Greece, 2–4 November 2005. 10. Ancajas, D.M.; Chakraborty, K.; Roy, S. Fort-NoCs: Mitigating the Threat of a Compromised NoC. In Proceedings of the 51st Annual Design Automation Conference, San Francisco, CA, USA, 1–5 June 2014; pp. 158:1–158:6. 10. Ancajas, D.M.; Chakraborty, K.; Roy, S. Fort-NoCs: Mitigating the Threat of a Compromised NoC Annual Design Automation Conference San Francisco CA USA 1–5 June 2014; pp 158:1–158:6 10. Ancajas, D.M.; Chakraborty, K.; Roy, S. Fort-NoCs: Mitigating the Threat of a Compromised NoC Annual Design Automation Conference, San Francisco, CA, USA, 1–5 June 2014; pp. 158:1–158:6. g pp 11. Frey, J.; Yu, Q. Exploiting State Obfuscation to Detect Hardware Trojans in NoC Network Interfaces. In Proceedings of the 2015 IEEE 58th International Midwest Symposium on Circuits and Systems (MWSCAS), Fort Collins, CO, USA, 2–5 August 2015; pp. 1–4. [CrossRef] pp [ ] 12. Boraten, T.; Kodi, A.K. Packet Security with Path Sensitization for NoCs. In Proceedings of the 2016 Design, Automation & Test in Europe Conference & Exhibition (DATE), Dresden, Germany, 14–18 March 2016; pp. 1136–1139. 13. Rajesh, J.; Chakraborty, K.; Roy, S. Hardware Trojan Attacks in SoC and NoC. In The Hardware Trojan War; Springer: Berlin/Heidelberg, Germany, 2018; pp. 55–74. 14. Moriam, S.; Franz, E.; Walthers, P.; Kumar, A.; Strufe, T.; Fettweis, G.P. Protecting Communication in Ma Active Attackers. In Proceedings of the ACM Great Lakes Symposium on VLSI (GLSVLSI), Chicago, I 14. Moriam, S.; Franz, E.; Walthers, P.; Kumar, A.; Strufe, T.; Fettweis, G.P. Protecting Communication in Many-Core Systems against Active Attackers. In Proceedings of the ACM Great Lakes Symposium on VLSI (GLSVLSI), Chicago, IL, USA, 23–25 May 2018. 15. References Ahlswede, R.; Cai, N.; Li, S.Y.R.; Yeung, R.W. Network information ﬂow. IEEE Trans. Inf. Theory 2000, 46, 1204–1216. [CrossRef] 16. Alkabani, Y.; Koushanfar, F. Extended abstract: Designer’s hardware Trojan horse. In Proceedings of the IEEE International Workshop on Hardwareoriented Security and Trust Anaheim CA USA 9 June 2008; pp 82–83 [CrossRef] 14. Moriam, S.; Franz, E.; Walthers, P.; Kumar, A.; Strufe, T.; Fettweis, G.P. Protecting Communication in Many-Core Systems against Active Attackers. In Proceedings of the ACM Great Lakes Symposium on VLSI (GLSVLSI), Chicago, IL, USA, 23–25 May 2018. 15. Ahlswede, R.; Cai, N.; Li, S.Y.R.; Yeung, R.W. Network information ﬂow. IEEE Trans. Inf. Theory 2000, 46, 1204–1216. [CrossRef] g y p g 15. Ahlswede, R.; Cai, N.; Li, S.Y.R.; Yeung, R.W. Network information ﬂow. IEEE Trans. Inf. Theory 2000, 15. Ahlswede, R.; Cai, N.; Li, S.Y.R.; Yeung, R.W. Network information ﬂow. IEEE Trans. Inf. Theory 2000, 46, 1204–1216. [CrossRef] 16. Alkabani, Y.; Koushanfar, F. Extended abstract: Designer’s hardware Trojan horse. In Proceedings of the IEEE International Workshop on Hardwareoriented Security and Trust, Anaheim, CA, USA, 9 June 2008; pp. 82–83. [CrossRef] 16. Alkabani, Y.; Koushanfar, F. Extended abstract: Designer’s hardware Trojan horse. In Proceedings Workshop on Hardwareoriented Security and Trust, Anaheim, CA, USA, 9 June 2008; pp. 82–83. [Cros 16. Alkabani, Y.; Koushanfar, F. Extended abstract: Designer’s hardware Trojan horse. In Proceedings of the IEEE International Workshop on Hardwareoriented Security and Trust, Anaheim, CA, USA, 9 June 2008; pp. 82–83. [CrossRef] , ; , g j g Workshop on Hardwareoriented Security and Trust, Anaheim, CA, USA, 9 June 2008; pp. 82–83. [CrossRef] 17. Jin, Y.; Kupp, N.; Makris, Y. Experiences in Hardware Trojan design and implementation. In Proceedings of the 2009 IEEE International Workshop on Hardware-Oriented Security and Trust, Francisco, CA, USA, 27–27 July 2009; pp. 50–57. [CrossRef] 18 F J Y Q A h d d t k hi d i i ti h d T j iti ti th d I t VLSI J 2016 17. Jin, Y.; Kupp, N.; Makris, Y. Experiences in Hardware Trojan design and implementation. In Proceedings of the 2009 IEEE International Workshop on Hardware-Oriented Security and Trust, Francisco, CA, USA, 27–27 July 2009; pp. 50–57. [CrossRef] 18. Frey, J.; Yu, Q. A hardened network-on-chip design using runtime hardware Trojan mitigation methods. Integr. VLSI J. 2016, 56, 15–31. [CrossRef] 19. Kocher, P.; Lee, R.; McGraw, G.; Raghunathan, A. References Security as a New Dimension in Embedded System Design. In Proceedings of the Proceedings of the 41st annual Design Automation Conference, San Diego, CA, USA, 7–11 June 2004; pp. 753–760. 19. Kocher, P.; Lee, R.; McGraw, G.; Raghunathan, A. Security as a New Dimension in Embedded System Design. In Proceedings of the Proceedings of the 41st annual Design Automation Conference, San Diego, CA, USA, 7–11 June 2004; pp. 753–760. 20. Kapoor, H.K.; Rao, G.B.; Arshi, S.; Trivedi, G. A Security Framework for NoC Using Authenticated Encryption and Session Keys. Circuits Syst. Signal Process. 2013, 32, 2605–2622. [CrossRef] the Proceedings of the 41st annual Design Automation Conference, San Diego, CA, USA, 7 11 June 2004; pp. 753 760. 20. Kapoor, H.K.; Rao, G.B.; Arshi, S.; Trivedi, G. A Security Framework for NoC Using Authenticated Encryption and Session Keys. Circuits Syst. Signal Process. 2013, 32, 2605–2622. [CrossRef] 20. Kapoor, H.K.; Rao, G.B.; Arshi, S.; Trivedi, G. A Security Fra Circuits Syst. Signal Process. 2013, 32, 2605–2622. [CrossRef] 21. Sepúlveda, J.; Zankl, A.; Flórez, D.; Sigl, G. Towards Protected MPSoC Communication for Information Protection against a Malicious NoC. In Procedia Computer Science, Proceedings of the International Conference on Computational Science, ICCS 2017, Zurich, Switzerland, 12–14 June 2017; Elsevier B.V.: Amsterdam, The Netherlands, 2017; Volume 108, pp. 1103–1112. 31 of 31 31 of 31 Electronics 2021, 10, 238 22. Chou, P.A.; Wu, Y.; Jain, K. Practical Network Coding. In Proceedings of the Annual Allerton Conference on Communication Control and Computing, Monticello, IL, USA, 1–3 October 2003. p g 23. Moriam, S.; Yan, Y.; Fischer, E.; Franz, E.; Fettweis, G.P. Resilient and Efﬁcient Communication in Many-Core Systems using Network Coding. In Proceedings of the 2015 IEEE 34th International Performance Computing and Communications Conference (IPCCC), Nanjing, China, 14–16 December 2015. ( ) j g 24. Pfennig, S.; Franz, E. Security Aspects of Conﬁdential Network Coding. In Proceedings of the 2017 IEEE International Conference on Communications (ICC), Paris, France, 21–25 May 2017. ( ) y 25. Charlap, S.L.; Rees, H.D.; Robbins, D.P. The asymptotic probability that a random biased matrix is invertible. Discret. Math. 1990, 82, 153–163. [CrossRef] 26. Eisenbarth, T.; Kumar, S.; Paar, C.; Poschmann, A.; Uhsadel, L. A Survey of Lightweight-Cryptography Implementations. IEEE Des. Test Comput. 2007, 24, 522–533. [CrossRef] 27. Lim, C.H.; Korishko, T. mCrypton—A Lightweight Block Cipher for Security of Low-Cost RFID Tags and Sensors. References In International Workshop on Information Security Applications; Springer: Berlin/Heidelberg, Germany, 2005. 28. NIST Special Publication 800-38A. Recommendation for Block Cipher Modes of Operation—Methods and Techniques; Technical Report; National Institute of Standards and Technology: Gaithersburg, MD, USA, 2001 28. NIST Special Publication 800-38A. Recommendation for Block Cipher Modes of Operation—Meth National Institute of Standards and Technology: Gaithersburg, MD, USA, 2001 gy g A.J.; Vanstone, S.A.; Oorschot, P.C.V. Handbook of Applied Cryptography; CRC Press: Boca Raton, FL, USA, 1 29. Menezes, A.J.; Vanstone, S.A.; Oorschot, P.C.V. Handbook of Applied Cryptography; CRC Press: Boca R Simmons, G.J. (Ed.) Contemporary Cryptography—The Science of Information Integrity; IEEE Press: New York Ed.) Contemporary Cryptography—The Science of Information Integrity; IEEE Press: New York, NY, USA, 1991. 31. Winter, M.; Fettweis, G.P. A Network-on-Chip Channel Allocator for Run-Time Task Scheduling in Multi-Processor System-on- Chips. In Proceedings of the 2008 11th EUROMICRO Conference on Digital System Design Architectures, Methods and Tools, Parma, Italy, 3–5 September 2008; pp. 133–140. 32. Kleinrock, L. Queueing Systems—1 : Theory; Wiley: New York, NY, USA, 1975. g y y y 33. Haas, S.; Seifert, T.; Nöthen, B.; Scholze, S.; Höppner, S.; Dixius, A.; Adeva, E.P.; Augustin, T.; Pauls, F.; Moriam, S.; et al. A heterogeneous SDR MPSoC in 28 nm CMOS for low-latency wireless applications. In Proceedings of the 54th Annual Design Automation Conference, Austin, TX, USA, 18–22 June 2017. 34. Borghoff, J.; Canteaut, A.; Güneysu, T.; Kavun, E.B.; Knudsen, L.R.; Le, G.; Paar, C.; Rechberger, C.; Rombouts, P. PRINCE—A Low-Latency Block Cipher for Pervasive Computing Applications (Full Version); Technical Report 529; Springer: New York, NY, USA, 2012. 35. Harttung, J.; Franz, E.; Moriam, S.; Walther, P. Lightweight Authenticated Encryption for Network-on-C Proceedings of the 2019 on Great Lakes Symposium on VLSI, Tysons Corner, VA, USA, 9–11 May 2019 36. Valiant, L.G.; Brebner, G.J. Universal schemes for parallel communication. In Proceedings of the Thirteenth Annual ACM Symposium on Theory of Computing, Milwaukee, WI, USA, 11–13 May 1981; pp. 263–277. 37. Nesson, T.; Johnsson, S.L. ROMM routing on mesh and torus networks. In Proceedings of the Seventh Annual ACM Symposium on Parallel Algorithms and Architectures, Santa Barbara, CA, USA, 16–18 July 1995; pp. 275–287.
https://openalex.org/W2065461770	https://escholarship.org/content/qt9xh3c50t/qt9xh3c50t.pdf?t=o4g07y	English	null	Search for first generation scalar leptoquarks in pp collisions at <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" altimg="si1.gif" overflow="scroll"><mml:msqrt><mml:mi>s</mml:mi></mml:msqrt><mml:mo>=</mml:mo><mml:mn>7</mml:mn><mml:mtext> TeV</mml:mtext></mml:math> with the ATLAS detector	Physics letters. B	2,012	cc-by	21,099	UC Santa Cruz UC Santa Cruz Previously Published Works UC Santa Cruz Previously Published Works Title Search for first generation scalar leptoquarks in pp collisions at s=7 TeV with the ATLAS detector Copyright Information This work is made available under the terms of a Creative Commons Attribution License, availalbe at https://creativecommons.org/licenses/by/4.0/ Copyright Information This work is made available under the terms of a Creative Commons Attribution License, availalbe at https://creativecommons.org/licenses/by/4.0/ UC Santa Cruz UC Santa Cruz Previously Published Works Title Search for first generation scalar leptoquarks in pp collisions at s=7 TeV with the ATLAS detector Permalink https://escholarship.org/uc/item/9xh3c50t Journal Physics Letters B, 709(3) ISSN 0370-2693 Authors Collaboration, ATLAS Aad, G Abbott, B et al. Publication Date 2012-03-01 DOI 10.1016/j.physletb.2012.02.004 Copyright Information This work is made available under the terms of a Creative Commons Attribution License, availalbe at https://creativecommons.org/licenses/by/4.0/ Peer reviewed UC Santa Cruz UC Santa Cruz Previously Published Works Title Search for first generation scalar leptoquarks in pp collisions at s=7 TeV with the ATLAS detector Permalink https://escholarship.org/uc/item/9xh3c50t Journal Physics Letters B, 709(3) ISSN 0370-2693 Authors Collaboration, ATLAS Aad, G Abbott, B et al. Publication Date 2012-03-01 DOI 10.1016/j.physletb.2012.02.004 Copyright Information This work is made available under the terms of a Creative Commons Attribution License, availalbe at https://creativecommons.org/licenses/by/4.0/ Peer reviewed UC Santa Cruz UC Santa Cruz Previously Publis Title Search for first generation scalar leptoquarks detector Permalink https://escholarship.org/uc/item/9xh3c50t Journal Physics Letters B, 709(3) ISSN 0370-2693 Authors Collaboration, ATLAS Aad, G Abbott, B et al. Publication Date 2012-03-01 DOI 10.1016/j.physletb.2012.02.004 Copyright Information This work is made available under the terms availalbe at https://creativecommons.org/lice Peer reviewed UC Santa Cruz UC Santa Cruz Previously Pub Title Search for first generation scalar leptoquar detector Permalink https://escholarship.org/uc/item/9xh3c50t Journal Physics Letters B, 709(3) ISSN 0370-2693 Authors Collaboration, ATLAS Aad, G Abbott, B et al. Publication Date 2012-03-01 DOI 10.1016/j.physletb.2012.02.004 Copyright Information This work is made available under the term availalbe at https://creativecommons.org/li Peer reviewed 1. Introduction ready exist from searches of LQ produced in pairs at the LHC [4,5], Tevatron [6] and LEP [7]. Limits on single LQ production come from HERA [8] and other experiments [9]. Similarities between leptons and quarks in the Standard Model (SM) suggest that they might be a part of some symmetry at en- ergy scales above the electroweak symmetry breaking scale. In this type of symmetry, transitions between leptons and quarks, medi- ated by a new type of gauge boson, a leptoquark (LQ), may occur. LQs are putative color-triplet bosons with spin 0 or 1, and frac- tional electric charge [1]. They are predicted in many extensions of the SM, such as Grand Uniﬁcation models, and possess both quark and lepton quantum numbers. The Yukawa coupling λLQ −l−q of a leptoquark to a lepton and a quark, and the branching ra- tio (β) to a charged lepton, are model dependent. In pp collisions, if λLQ −l−q is of the order of the electroweak coupling strength, leptoquarks are predominantly produced in pairs via the strong interaction. At the LHC, the pair production cross section is dom- inated by gluon fusion for LQ masses mLQ ≲1 TeV, whereas at higher masses it is dominated by quark–antiquark annihilation. Un- der these assumptions, the production rate for scalar LQs depends only on the known QCD coupling constant and the unknown LQ mass, and has been calculated at up to next-to-leading order. It is usually assumed that leptoquarks only couple to one generation of SM isospin multiplet to accommodate experimental constraints on ﬂavor-changing neutral currents, and lepton and baryon num- ber violation [2]. Consequently, they are classiﬁed as ﬁrst, second, or third generation according to the fermion generation to which they couple [3]. Lower mass limits on the ﬁrst generation LQs al- In this Letter we present updated results on a search for the pair production of ﬁrst generation scalar leptoquarks in pp colli- sions at √ s = 7 TeV. The search is performed with a dataset corre- sponding to an integrated luminosity of 1.030 ± 0.035 fb−1 [10] of data collected by the ATLAS detector at the LHC from March 2011 to July 2011. We search for leptoquarks in two different ﬁnal states. 1 ATLAS uses a right-handed coordinate system with its origin at the nominal interaction point in the center of the detector and z axis coinciding with the axis of the beam pipe. The x axis points from the interaction point to the center of the LHC ring, and the y axis points upward. Cylindrical coordinates (r,φ) are used in the transverse plane, φ being the azimuthal angle around the beam pipe. The pseudorapidity is deﬁned in terms of the polar angle θ as η = −ln tan(θ/2). a r t i c l e i n f o We report a search for ﬁrst generation scalar leptoquarks using 1.03 fb−1 of proton–proton collisions data produced by the Large Hadron Collider at √ s = 7 TeV and recorded by the ATLAS experiment. Leptoquarks are sought via their decay into an electron or neutrino and a quark, producing events with two oppositely charged electrons and at least two jets, or events with an electron, missing transverse momentum and at least two jets. Control data samples are used to validate background predictions from Monte Carlo simulation. In the signal region, the observed event yields are consistent with the background expectations. We exclude at 95% conﬁdence level the production of ﬁrst generation scalar leptoquark with masses mLQ < 660 (607) GeV when assuming the branching fraction of a leptoquark to a charged lepton is equal to 1.0 (0.5). Article history: Received 20 December 2011 Received in revised form 1 February 2012 Accepted 2 February 2012 Available online 7 February 2012 Editor: H. Weerts © 2012 CERN. Published by Elsevier B.V. Open access under CC BY-NC-ND license. 1. Introduction In the ﬁrst one both LQs decay into an electron and a quark, while in the second ﬁnal state one of the LQs decays into an electron and a quark and the other LQ decays into an electron–neutrino and a quark. These result in two different experimental signatures. One such signature is the production of two electrons and two jets and the other one comprises one electron, two jets, and missing trans- verse momentum (the magnitude of which is denoted as Emiss T ). The results from the two ﬁnal states are combined and presented in the mLQ versus β plane, where β is the branching ratio for a single LQ to decay into a charged lepton and a quark. ✩© CERN for the beneﬁt of the ATLAS Collaboration. ⋆E-mail address: atlas.publications@cern.ch. the LH in the pseudo 0370-2693 © 2012 CERN. Published by Elsevier B.V. doi:10.1016/j.physletb.2012.02.004 Open access under CC BY-NC-ND license. Powered by the California Digital Library University of California eScholarship.org Physics Letters B 709 (2012) 158–176 Search for ﬁrst generation scalar leptoquarks in pp collisions at √ s = 7 TeV with the ATLAS detector ✩ .ATLAS Collaboration ⋆ 0370-2693 © 2012 CERN. Published by Elsevier B.V. doi:10.1016/j.physletb.2012.02.004 Open access under CC BY-NC-ND license. 2. The ATLAS detector The ATLAS detector [11] is a general-purpose particle detector with cylindrical geometry,1 which consists of several subdetectors ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 159 of radius R = ( η)2 + ( φ)2 = 0.2 centered on the electron track, excluding the electron contribution, and corrected for the energy from event pile-up and the electron energy leakage inside the cone. surrounding the interaction point, and providing nearly 4π cov- erage in solid angle. The location of the interaction point and momenta of charged particles are determined by the multi-layer silicon pixel and strip detectors covering \|η\| < 2.5 in pseudora- pidity η, and a transition radiation tracker extending to \|η\| < 2.0, which are inside a superconducting solenoid producing a ﬁeld of 2 T. The tracking system is surrounded by a high-granularity liquid-argon (LAr) sampling electromagnetic calorimeter with cov- erage up to \|η\| < 3.2. An iron-scintillator tile hadronic calorimeter provides coverage in the range \|η\| < 1.7. In the end-cap and for- ward regions LAr calorimeters provide both electromagnetic and hadronic measurements and cover the region 1.5 < \|η\| < 4.9. The muon spectrometer, consisting of precision tracking detectors and superconducting toroids, is located outside the calorimeters. Jets are deﬁned as localized energy deposits in the calorimeter and are reconstructed using the anti-kt algorithm [23] with a dis- tance parameter of 0.4 and by performing a four-vector sum over calorimeter clusters. Reconstructed jets are corrected for the non- compensating calorimeter response, upstream material and other effects by using pT- and η-dependent correction factors derived from MC and validated with test-beam and collision data [24]. We further require that jets satisfy ET > 30 GeV, \|η\| < 2.8 and are sep- arated from electrons passing the above selection within R > 0.4. Selected jets must also pass quality requirements to reject jets arising from electronic noise bursts, cosmic rays and beam back- ground, originating mainly from beam-gas events and beam-halo events [25]. We perform the search in the data sample selected by a three- level trigger requiring at least one high transverse energy (ET) electron. The trigger is fully eﬃcient for electrons with ET > 30 GeV, as measured in an inclusive Z →ee control sample [12]. The presence of neutrinos is inferred from the missing trans- verse momentum ⃗pmiss T (and its magnitude Emiss T ) [26]. 3. Simulated samples Samples of Monte Carlo (MC) events are used to devise selec- tion criteria and validate background predictions. Background and signal samples are processed through the full ATLAS detector simu- lation based on GEANT4 [13], followed by the same reconstruction algorithms as used for collision data. The effects from in-time and out-of-time proton–proton collisions are included in the MC simu- lation. In the simulated samples, an event weight is applied to the average number of additional proton–proton collisions occurring in the same bunch crossing (event pile-up), to ensure that the num- ber of interactions per bunch crossing, amounting to an average of 6, is well modeled. Corrections are made to the simulated samples to ensure a good description of the energy resolution and the trigger and reconstruction eﬃciencies. These are determined in control data samples and applied to both simulated background and signal samples. These corrections change the total expected yields by less than 2%. 5. Event selection We deﬁne event selections to create samples with high signal and background acceptance. Events are selected to be consistent with the LQ LQ →eeq¯q/eνq¯q decays. In the eejj topology we re- quire two electrons and at least two jets as deﬁned in Section 4 and an invariant mass of the electron pair mee > 40 GeV. In the eν jj topology, one electron, at least two jets and Emiss T > 30 GeV are required, together with a requirement on the transverse mass of the electron and the ⃗pmiss T , mT = 2⃗pe T⃗pmiss T (1 −cos( φ)) > 40 GeV, where φ is the angle between the electron pT and ⃗pmiss T . In addition, we require that φ(jet, ⃗pmiss T ) > 4.5 × (1 − Emiss T /45 GeV) in the eν jj channel for events with Emiss T < 45 GeV to reduce residual contamination from MJ events. Events with ad- ditional identiﬁed electrons as deﬁned in Section 4 or muons with pT > 30 GeV and \|η\| < 2.4 are rejected. The dominant backgrounds to the leptoquark signal include W and Z boson production in association with one or more jets, sin- gle and pair production of top quarks, QCD multi-jet (MJ) and diboson processes. The ALPGEN [14] generator is used for the sim- ulation of the W , Z boson production in association with n par- tons. This program is interfaced to HERWIG [15] and JIMMY [16] to model parton showers and multiple parton interactions, respec- tively. The MLM [14] jet-parton matching scheme is used to form inclusive W /Z + jets MC samples. MC@NLO [17] is used to es- timate single and pair production of top quarks. Diboson events are generated using HERWIG, and scaled to next-to-leading (NLO) cross section predictions [17,18]. Signal LQ samples are produced with PYTHIA [19] and nor- malized with NLO cross sections determined from Ref. [20] using CTEQ6.6 [21] parton distribution functions. After all the selection criteria are applied the signal acceptance is of 70% for a LQ signal of mLQ = 600 GeV for both channels, but the sample is still dominated by background events. 2. The ATLAS detector ⃗pmiss T is deﬁned as the negative vector sum of the transverse momenta of reconstructed electrons, muons and jets, as well as calorime- ter clusters not associated to reconstructed objects. 6. Background determination The 600 GeV LQ signal is also shown for β = 1.0. The solid line (band) in the lower plots shows the Gaussian statistical (statistical + systematic) e of the difference between data and the prediction. Fig. 1. Data and SM background comparisons of the input LLR variables for the eejj channel. (a) Invariant mass of the two electrons in the event; (b) Average LQ mass resulting from the best (electron, jet) combinations in each event, and (c) ST. The stacked distributions show the various background contributions, and data are indicated by the points with error bars. The 600 GeV LQ signal is also shown for β = 1.0. The solid line (band) in the lower plots shows the Gaussian statistical (statistical + systematic) signiﬁcance of the difference between data and the prediction. leading jets and the transverse energy of the two electrons in the eejj channel. In the ST deﬁnition in the eν jj channel, the second electron ET is substituted by the Emiss T . but failing the nominal electron identiﬁcation criteria described in Section 4. The MJ enhanced samples are corrected to remove the residual contamination from real electrons. In the eejj channel, the ﬁts are applied to the sample selected following the criteria of Sec- tion 5, as well as to control regions (i) and (ii), and the W + jets background is estimated together with the MJ background. In the eν jj channel, the ﬁts are applied to the selected sample as well as to control regions (iii)–(v). In the eejj topology we deﬁne two control regions (i) Z + jets: formed by events with at least two jets and in which the two electrons are required to have an invariant mass within a Z mass window 81 < mee < 101 GeV, and (ii) t¯t: events with at least two jets and exactly one electron and one muon [27], deﬁned as in Sec- tion 4. 6. Background determination This search is based on selecting events with a high ET electron, two high pT jets, and an additional electron or large Emiss T . Electron candidates are reconstructed as energy deposits in the electromag- netic calorimeter. Electrons are required to have a shower proﬁle consistent with that expected for this particle, and to have a track pointing to the energy deposit in the calorimeter. The pattern of the energy deposits on the ﬁrst layer of the EM calorimeter is used to reject hadrons, while contamination from photon conversions is reduced by requiring a hit in the ﬁrst layer of the pixel detec- tor [22]. In addition to these criteria, we require electrons to have a transverse energy ET > 30 GeV and fall within a well instru- mented region of the detector. Further rejection against hadrons is achieved by requiring the electron candidates to be isolated from additional energy deposits in the calorimeter by requiring that E0.2 T /ET < 0.1, where E0.2 T is the transverse energy in a cone The MJ background estimate is derived directly from data, whereas MC samples are used to predict the other backgrounds. We verify the shape of the V + jets (V = W ±, Z) and top quark background prediction using control regions, which are deﬁned to enhance either the V + jets or the top quark production contri- bution, while keeping a negligible LQ signal contamination. These control regions are also used to derive the ﬁnal normalization of the V + jets and top quark backgrounds. The V + jets and top quark control regions are deﬁned by ap- plying additional selection criteria on mee and mT to the selected sample. The remaining signal contamination is reduced by apply- ing an upper threshold to the summed transverse momentum in the event, ST, deﬁned as the scalar sum of the pT of the two ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 160 ATLAS Collaboration / Physics Letters B 709 (2012) 158 176 a and SM background comparisons of the input LLR variables for the eejj channel. (a) Invariant mass of the two electrons in the event; (b) Average LQ mass rom the best (electron, jet) combinations in each event, and (c) ST. The stacked distributions show the various background contributions, and data are indicated by with error bars. 6. Background determination In the eν jj topology we deﬁne three control regions (iii) W + 2 jets: events with exactly two jets, an electron and Emiss T such that the transverse mass of the electron and the Emiss T is in the region of the W Jacobian peak, 40 < mT < 120 GeV, and an ST < 225 GeV requirement to limit the presence of signal events, (iv) W + 3 jets: as in (iii) but with three or more jets, and (v) t¯t: events with at least 4 jets, where the thresholds on the ﬁrst and second jets are raised to 50 GeV and 40 GeV, respectively. We observe 5615 data events in the eejj channel and 76 855 data events in the eν jj channel, with SM expectations of 5600 ± 1000 and 74 000±11 000, respectively. For mLQ = 600 GeV, we ex- pect 7.5±0.5 signal events in the eejj channel and 4.5±0.2 signal events in the eν jj channel. The aforementioned uncertainties fully account for (the dominant) systematic and statistical uncertainties. 8. Systematic uncertainties summed background and signal distributions respectively, and x j is the value of that variable for the j-th event in a given sample. Sep- arate LS distributions are created for several signal mass points, allowing mass-dependent optimization. Using the aforementioned quantities, a likelihood ratio is deﬁned as LLR = log(LS/LB) and is used as the ﬁnal variable to determine whether or not there is a LQ signal present in our data. Systematic uncertainties affect both background normalizations and shapes of the input distributions into the LLR. We consider systematic uncertainties from a variety of sources. These are de- scribed as follows. The jet energy scale (JES) and resolution (JER) uncertainties are considered independently, and applied by varying the JES (JER) within its uncertainty of 4% to 6.5% (14%) depending on the jet pT and η [28,29] for all simulated events. These variations are also propagated to the Emiss T in the eν jj channel. The resulting uncer- tainties for the mLQ = 600 GeV signal and background are 5% (8%) and 11% for the eejj (eν jj) ﬁnal state. The following discriminating variables, selected to give the best separation between signal and background, are used. For the eejj channel, we use mee, ST = Ee1 T + Ee2 T + pjet1 T + pjet2 T and the av- erage invariant LQ mass ¯mLQ . For the eν jj topology, we use mT(e, Emiss T ), ST, the transverse LQ mass mLQ T (jet, Emiss T ) and the invariant LQ mass mLQ (e, jet). To obtain the LQ masses, we calcu- late the invariant mass of the electron-jet system and the trans- verse mass of the Emiss T -jet system. Since the LQs are produced in pairs, there are two possible mass combinations for the electron- jet and Emiss T -jet pairs, and the combination giving the smallest mass difference is used. In the eejj channel, two possible electron- jet combinations arise from this procedure, and we take their average ¯mLQ for the analysis. The discriminating variables are shown in Figs. 1 and 2 for the eejj and the eν jj channels, re- spectively. Systematic uncertainties on the electron energy scale (1.6%) and resolution (0.6%), and on the electron trigger, reconstruction and identiﬁcation eﬃciencies are derived by varying the selection criteria deﬁning the Drell–Yan control sample used for the various measurements [12]. 7. Likelihood analysis To estimate the MJ background, we perform ﬁts to the mee dis- tribution in the eejj channel, and to the Emiss T distribution in the eν jj channel. In these ﬁts, the relative fraction of the MJ back- ground is a free parameter. Templates for the MJ background distri- butions are derived from MJ enhanced samples, which are formed using electron candidates passing relaxed selection requirements We use a likelihood ratio method to separate signal and SM background. The likelihoods are constructed separately for back- ground (LB) and signal (LS) hypotheses from a set of discriminat- ing variables as follows: LB ≡ bi(x j), LS ≡ si(x j), where bi, si are the probabilities of the i-th input variable from the normalized ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 161 a and SM background comparisons of the input LLR variables for the eν jj channel. (a) Transverse mass of the electron and the Emiss T in the event, (b) ST, (c) LQ (d) LQ transverse masses. The stacked distributions show the various background contributions, and data are indicated by the points with error bars. The 600 GeV s also shown for β = 0.5. The solid line (band) in the lower plots shows the Gaussian statistical (statistical + systematic) signiﬁcance of the difference between he prediction. Fig. 2. Data and SM background comparisons of the input LLR variables for the eν jj channel. (a) Transverse mass of the electron and the Emiss T in the event, (b) ST, (c) LQ mass, and (d) LQ transverse masses. The stacked distributions show the various background contributions, and data are indicated by the points with error bars. The 600 GeV LQ signal is also shown for β = 0.5. The solid line (band) in the lower plots shows the Gaussian statistical (statistical + systematic) signiﬁcance of the difference between data and the prediction. 8. Systematic uncertainties In addition, a 1% uncertainty is included to account for the eﬃciency of the isolation requirement. They lead to total signal and background yield uncertainties of 8% and 5% (3.5%), respectively, for the eejj (eν jj) channel and for a signal of mass mLQ = 600 GeV. ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 162 162 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 Fig. 3. LLR distributions for the eejj (a) and for the eν jj (b) ﬁnal states. The data are indicated with the points and the ﬁlled histograms show the SM background. The MJ background is estimated from data, while the other background contribu- tions are obtained from simulated samples as described in the text. The LQ signal corresponding to a LQ mass of 600 GeV is indicated by a solid line, and is normal- ized assuming β = 1.0 (0.5) in the eejj (eν jj) channel. The lowest bin corresponds to background events regions of the phase space for which no signal events are expected. The solid line (band) in the lower plots shows the Gaussian statistical (statistical + systematic) signiﬁcance of the difference between data and the predic- tion. The systematic uncertainty for the production model of V + jets is taken to be the largest difference between the nominal data- driven prediction using ALPGEN and that obtained by using SHERPA [30], giving an uncertainty of 1.5% and 3% for the eejj and the eν jj channels, respectively. The systematic uncertainty for the t¯t production model is eval- uated by comparing the yields between events generated with MC@NLO and those generated with various alternate samples. These include samples generated with POWHEG [31], a different top mass (170 GeV and 175 GeV instead of the nominal value equal to 172.5 GeV), and a different amount of initial and ﬁnal state-radiation (ISR/FSR). The result is an uncertainty in the t¯t yield of 10% and 15% for the single electron and dielectron analyses, re- spectively. Systematic uncertainties are determined for the MJ backgrounds by comparing results from alternative normalizations to those from the methods described earlier. The largest variation is taken, re- Table 1 The predicted and observed yields in a signal enhanced region deﬁned by requiring LLR > 0 for both channels. Background predictions are scaled as described in Sec- tion 6. Table 1 The predicted and observed yields in a signal enhanced region deﬁned by requiring LLR > 0 for both channels. Background predictions are scaled as described in Sec- tion 6. The eejj (eν jj) channel signal yields are computed assuming β = 1.0 (0.5). Statistical and systematic uncertainties added in quadrature are shown. The predicted and observed yields in a signal enhanced region deﬁned by requiring LLR > 0 for both channels. Background predictions are scaled as described in Sec- tion 6. The eejj (eν jj) channel signal yields are computed assuming β = 1.0 (0.5). Statistical and systematic uncertainties added in quadrature are shown. Fig. 3. LLR distributions for the eejj (a) and for the eν jj (b) ﬁnal states. The data are indicated with the points and the ﬁlled histograms show the SM background. The MJ background is estimated from data, while the other background contribu- tions are obtained from simulated samples as described in the text. The LQ signal corresponding to a LQ mass of 600 GeV is indicated by a solid line, and is normal- ized assuming β = 1.0 (0.5) in the eejj (eν jj) channel. The lowest bin corresponds to background events regions of the phase space for which no signal events are expected. The solid line (band) in the lower plots shows the Gaussian statistical (statistical + systematic) signiﬁcance of the difference between data and the predic- tion. h i i f h d i d l f j Statistical and systematic uncertainties added in quadrature are shown. Source eejj Channel eν jj Channel 400 GeV 600 GeV 400 GeV 600 GeV W + jets – – 1500 ± 670 670 ± 210 Z + jets 98 ± 53 26 ± 14 45 ± 41 18 ± 19 t¯t 15 ± 9 4.6 ± 2.2 430 ± 180 150 ± 38 Single t 1.4 ± 0.9 0.7 ± 0.4 53 ± 19 23 ± 4 Dibosons 1.5 ± 0.8 0.7 ± 0.3 25 ± 11 11 ± 2 MJ 9.2 ± 4.5 2.3 ± 1.5 170 ± 35 75 ± 15 Total 120 ± 55 34 ± 14 2200 ± 690 950 ± 220 Data 82 22 2207 900 LQ 120 ± 8 7.5 ± 0.5 69 ± 4 4.5 ± 0.2 Fig. 4. 8. Systematic uncertainties The eejj (eν jj) channel signal yields are computed assuming β = 1.0 (0.5). Statistical and systematic uncertainties added in quadrature are shown. Source eejj Channel eν jj Channel 400 GeV 600 GeV 400 GeV 600 GeV W + jets – – 1500 ± 670 670 ± 210 Z + jets 98 ± 53 26 ± 14 45 ± 41 18 ± 19 t¯t 15 ± 9 4.6 ± 2.2 430 ± 180 150 ± 38 Single t 1.4 ± 0.9 0.7 ± 0.4 53 ± 19 23 ± 4 Dibosons 1.5 ± 0.8 0.7 ± 0.3 25 ± 11 11 ± 2 MJ 9.2 ± 4.5 2.3 ± 1.5 170 ± 35 75 ± 15 Total 120 ± 55 34 ± 14 2200 ± 690 950 ± 220 Data 82 22 2207 900 LQ 120 ± 8 7.5 ± 0.5 69 ± 4 4.5 ± 0.2 Fig. 4. 95% CL upper limit on the pair production cross section times branching ra- tio of the ﬁrst generation leptoquarks for the eejj channel at β = 1.0 (a) and for the eν jj channel at β = 0.5 (b). The solid lines indicate the individual observed limits, while the expected limits are indicated by the dashed lines. The theory pre- diction is indicated by the dotted line, which includes the systematic uncertainties due to the choice of the PDF and due to the renormalization and factorization scales. The dark green (light yellow) solid band contains 68% (95%) of possible out- comes from pseudo-experiments in which the yield is Poisson-ﬂuctuated around the background-only expectation. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this Letter.) sulting in an uncertainty of 20% and 28% in the MJ normalization for the eν jj and the eejj channels, respectively. An uncertainty of 3.7% [10] on the integrated luminosity is applied to both dibo- son and single top background yields, as well as to expected signal yields. Finally, further uncertainties on the simulated background con- tributions originate from ﬁnite statistics in the MC samples used. Table 1 95% CL upper limit on the pair production cross section times branching ra- tio of the ﬁrst generation leptoquarks for the eejj channel at β = 1.0 (a) and for the eν jj channel at β = 0.5 (b). The solid lines indicate the individual observed limits, while the expected limits are indicated by the dashed lines. The theory pre- diction is indicated by the dotted line, which includes the systematic uncertainties due to the choice of the PDF and due to the renormalization and factorization scales. The dark green (light yellow) solid band contains 68% (95%) of possible out- comes from pseudo-experiments in which the yield is Poisson-ﬂuctuated around the background-only expectation. (For interpretation of the references to color in thi ﬁ l d th d i f d t th b i f thi L tt ) Source eejj Channel eν jj Channel 400 GeV 600 GeV 400 GeV 600 GeV W + jets – – 1500 ± 670 670 ± 210 Z + jets 98 ± 53 26 ± 14 45 ± 41 18 ± 19 t¯t 15 ± 9 4.6 ± 2.2 430 ± 180 150 ± 38 Single t 1.4 ± 0.9 0.7 ± 0.4 53 ± 19 23 ± 4 Dibosons 1.5 ± 0.8 0.7 ± 0.3 25 ± 11 11 ± 2 MJ 9.2 ± 4.5 2.3 ± 1.5 170 ± 35 75 ± 15 Total 120 ± 55 34 ± 14 2200 ± 690 950 ± 220 Data 82 22 2207 900 LQ 120 ± 8 7.5 ± 0.5 69 ± 4 4.5 ± 0.2 Fig. 3. LLR distributions for the eejj (a) and for the eν jj (b) ﬁnal states. The data are indicated with the points and the ﬁlled histograms show the SM background. The MJ background is estimated from data, while the other background contribu- tions are obtained from simulated samples as described in the text. The LQ signal corresponding to a LQ mass of 600 GeV is indicated by a solid line, and is normal- ized assuming β = 1.0 (0.5) in the eejj (eν jj) channel. The lowest bin corresponds to background events regions of the phase space for which no signal events are expected. The solid line (band) in the lower plots shows the Gaussian statistical (statistical + systematic) signiﬁcance of the difference between data and the predic- tion. Table 1 The systematic uncertainty for the production model of V + jets is taken to be the largest difference between the nominal data- driven prediction using ALPGEN and that obtained by using SHERPA [30], giving an uncertainty of 1.5% and 3% for the eejj and the eν jj channels, respectively. Fig. 4. 95% CL upper limit on the pair production cross section times branching ra- tio of the ﬁrst generation leptoquarks for the eejj channel at β = 1.0 (a) and for the eν jj channel at β = 0.5 (b). The solid lines indicate the individual observed limits, while the expected limits are indicated by the dashed lines. The theory pre- diction is indicated by the dotted line, which includes the systematic uncertainties due to the choice of the PDF and due to the renormalization and factorization scales. The dark green (light yellow) solid band contains 68% (95%) of possible out- comes from pseudo-experiments in which the yield is Poisson-ﬂuctuated around the background-only expectation. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this Letter.) The systematic uncertainty for the t¯t production model is eval- uated by comparing the yields between events generated with MC@NLO and those generated with various alternate samples. These include samples generated with POWHEG [31], a different top mass (170 GeV and 175 GeV instead of the nominal value equal to 172.5 GeV), and a different amount of initial and ﬁnal state-radiation (ISR/FSR). The result is an uncertainty in the t¯t yield of 10% and 15% for the single electron and dielectron analyses, re- spectively. sulting in an uncertainty of 20% and 28% in the MJ normalization for the eν jj and the eejj channels, respectively. An uncertainty of 3.7% [10] on the integrated luminosity is applied to both dibo- son and single top background yields, as well as to expected signal yields. Systematic uncertainties are determined for the MJ backgrounds by comparing results from alternative normalizations to those from the methods described earlier. The largest variation is taken, re- Finally, further uncertainties on the simulated background con- tributions originate from ﬁnite statistics in the MC samples used. ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 163 Fig. 5. 95% CL exclusion region resulting from the combination of the two channels shown in the β versus leptoquark mass plane. Acknowledgements We thank CERN for the very successful operation of the LHC, as well as the support staff from our institutions without whom ATLAS could not be operated eﬃciently. We acknowledge the support of ANPCyT, Argentina; YerPhI, Ar- menia; ARC, Australia; BMWF, Austria; ANAS, Azerbaijan; SSTC, Belarus; CNPq and FAPESP, Brazil; NSERC, NRC and CFI, Canada; CERN; CONICYT, Chile; CAS, MOST and NSFC, China; COLCIENCIAS, Colombia; MSMT CR, MPO CR and VSC CR, Czech Republic; DNRF, DNSRC and Lundbeck Foundation, Denmark; ARTEMIS, European Union; IN2P3-CNRS, CEA-DSM/IRFU, France; GNAS, Georgia; BMBF, DFG, HGF, MPG and AvH Foundation, Germany; GSRT, Greece; ISF, MINERVA, GIF, DIP and Benoziyo Center, Israel; INFN, Italy; MEXT and JSPS, Japan; CNRST, Morocco; FOM and NWO, Netherlands; RCN, Norway; MNiSW, Poland; GRICES and FCT, Portugal; MERYS (MECTS), Romania; MES of Russia and ROSATOM, Russian Federa- tion; JINR; MSTD, Serbia; MSSR, Slovakia; ARRS and MVZT, Slove- nia; DST/NRF, South Africa; MICINN, Spain; SRC and Wallenberg Foundation, Sweden; SER, SNSF and Cantons of Bern and Geneva, Switzerland; NSC, Taiwan; TAEK, Turkey; STFC, the Royal Soci- ety and Leverhulme Trust, United Kingdom; DOE and NSF, United States of America. Fig. 5. 95% CL exclusion region resulting from the combination of the two channels shown in the β versus leptoquark mass plane. The shaded area indicates the D0 exclusion limit [6], while the thick dotted line indicates the CMS exclusion [4]. The dotted and dotted-dashed lines indicate the individual limits for the eejj and the eν jj, respectively. The combined observed limit is indicated by the solid black line. The combined expected limit is indicated by the dashed line, together with the solid band containing 68% of possible outcomes from pseudo-experiments in which the yield is Poisson-ﬂuctuated around the background-only expectation. These range from 2%–9%, depending on the LQ mass under consid- eration. Additional signal uncertainties considered arise from the choice of the PDF, which results in an uncertainty on the signal ac- ceptance of 1%–8% for LQ masses between 300 GeV and 700 GeV, and from ISR/FSR effects, resulting in an uncertainty of 2% for both channels. The crucial computing support from all WLCG partners is ac- knowledged gratefully, in particular from CERN and the ATLAS Tier-1 facilities at TRIUMF (Canada), NDGF (Denmark, Norway, Sweden), CC-IN2P3 (France), KIT/GridKA (Germany), INFN-CNAF (Italy), NL-T1 (Netherlands), PIC (Spain), ASGC (Taiwan), RAL (UK) and BNL (USA) and in the Tier-2 facilities worldwide. Table 1 The shaded area indicates the D0 exclusion limit [6], while the thick dotted line indicates the CMS exclusion [4]. The dotted and dotted-dashed lines indicate the individual limits for the eejj and the eν jj, respectively. The combined observed limit is indicated by the solid black line. The combined expected limit is indicated by the dashed line, together with the solid band containing 68% of possible outcomes from pseudo-experiments in which the yield is Poisson-ﬂuctuated around the background-only expectation. m > 607 (587) GeV when assuming its branching fraction to a charged lepton to be equal to 1.0 and 0.5, respectively. These are the most stringent limits to date arising from direct searches for leptoquarks. References [1] B. Schrempp, F. Schrempp, Phys. Lett. B 153 (1985) 101; J. Pati, A. Salam, Phys. Rev. D 10 (1974) 275; G. Senjanovic, A. Sokorac, Z. Phys. C 20 (1983) 255; P. Frampton, B.-H. Lee, Phys. Rev. Lett. 64 (1990) 619; P. Frampton, T. Kephart, Phys. Rev. D 42 (1990) 3892; E. Witten, Nucl. Phys. B 258 (1985) 75; K. Lane, Nucl. Phys. (1993). Lectures given at Theoretical Advanced Study Insti- K. Lane, Nucl. Phys. (1993). Lectures given at Theoretical Advanced Study Insti- tute (TAS193) in: Elem. Part. Physics, 1233; arXiv:hep-ph/9401324. tute (TAS193) in: Elem. Part. Physics, 1233; arXiv:hep-ph/9401324. ] S. Kovalenko, I. Schmidt, Phys. Lett. B 562 (2003) 104. [2] S. Kovalenko, I. Schmidt, Phys. Lett. B 562 (2003) 104. [3] W. Buchmüller, R. Rückl, D. Wyler, Phys. Lett. B 191 (1987) 442; [3] W. Buchmüller, R. Rückl, D. Wyler, Phys. Lett. B 191 (1987) 442; W. Buchmüller, R. Rückl, D. Wyler, Phys. Lett. B 448 (1999) 320, [3] W. Buchmüller, R. Rückl, D. Wyler, Phys. Lett. B 191 (1987) 442; W Buchmüller R Rückl D Wyler Phys Lett B 448 (1999) 320 W. Buchmüller, R. Rückl, D. Wyler, Phys. Lett. B 448 (1999) 320, Erratum 9. Results The LLR distributions for data, backgrounds and a LQ signal as- suming mLQ = 600 GeV are shown in Fig. 3 for both channels. The observed and predicted event yields requiring LLR > 0 for the ma- jor background sources, as well as the expected signal, are shown in Table 1. We do not observe any excess of events at high LLR values where signal is expected, indicating no evidence of scalar LQ pair production. Given the absence of signal we determine 95% CL upper limits on the LQ pair production cross sections using a modiﬁed frequentist C Ls method based on a Poisson log-likelihood ratio statistical test [32,33]. Systematic and statistical uncertain- ties are treated as nuisance parameters with a Gaussian probability density function, and the full LLR distribution is considered. The effect of the various systematic uncertainties on the shape of the LLR distribution are included on the calculation by integrating over a Gaussian distribution with standard deviation equal to the frac- tional change in the yield between the systematically adjusted dis- tribution and the nominal case for each individual uncertainty in each bin. The 95% CL upper bounds on the cross section for LQ pair production as a function of mass are shown in Fig. 4 for both the eejj and the eν jj channels for β = 1.0 and β = 0.5, respectively. The obtained cross section limits are combined, and reinterpreted as limits in the β vs. mLQ plane as shown in Fig. 5. Open access This article is published Open Access at sciencedirect.com. It is distributed under the terms of the Creative Commons Attribu- tion License 3.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are credited. S. Frixione, P. Nason, B.R. Webber, JHEP 0308 (2 S. Frixione, P. Nason, B.R. Webber, JHEP 0308 (2003) 007. [18] J.M. Campbell, R.K. Ellis, Phys. Rev. D 60 (1999) 113006. [18] J.M. Campbell, R.K. Ellis, Phys. Rev. D 60 (1999) 113006. [29] ATLAS Collaboration, ATLAS-CONF-2010-054, http://cdsweb.cern.ch/record/ 1281311. [29] ATLAS Collaboration, ATLAS-CONF-2010-054, http://cdsweb.cern.ch/record/ 1281311. [19] T. Sjöstrand, S. Mrenna, P. Skands, JHEP 0605 (2006) 026. [20] M. Kramer, T. Plehn, M. Spira, P.M. Zerwas, Phys. Rev. D 7 [20] M. Kramer, T. Plehn, M. Spira, P.M. Zerwas, Phys. Rev. D 71 (2005) 057503. [30] T. Gleisberg, et al., JHEP 0902 (2009) 007; [21] D. Stump, et al., JHEP 0310 (2003) 046. [21] D. Stump, et al., JHEP 0310 (2003) 046. 30] T. Gleisberg, et al., JHEP 0902 (2009) 007; S. Schumann, F. Krauss, JHEP 0803 (2008) 038; S. Schumann, F. Krauss, JHEP 0803 (2008) 038; [22] ATLAS Collaboration, arXiv:1110.3174. . Hoeche, F. Krauss, S. Schumann, F. Siegert, JHEP 0905 (2009) 053. [23] M. Cacciari, G.P. Salam, G. Soyez, JHEP 0804 (2008) 063; S. Hoeche, F. Krauss, S. Schumann, F. Siegert, JHEP 0905 [23] M. Cacciari, G.P. Salam, G. Soyez, JHEP 0804 (2008) 0 M. Cacciari, G.P. Salam, Phys. Lett. B 641 (2006) 57. [31] S. Frixione, P. Nason, C. Oleari, JHEP 0711 (2007) 070. M. Cacciari, G.P. Salam, Phys. Lett. B 641 (2006) 57. [32] W. Fisher, FERMILAB-TM-2386-E. [24] ATLAS Collaboration, CERN-PH-EP-2011-191, http://cdsweb.cern.ch/record 1399505. [33] T. Junk, Nucl. Instrum. Meth. A 434 (1999) 435. 10. Conclusions [4] CMS Collaboration, Phys. Lett. B 703 (2011) 246. [5] ATLAS Collaboration, Phys. Rev. D 83 (2011) 112006. [6] D0 Collaboration, Phys. Rev. D 84 RC (2011) 071104. [6] D0 Collaboration, Phys. Rev. D 84 RC (2011) 071104. We report on a search for pair production of ﬁrst generation scalar leptoquarks at ATLAS using a data sample corresponding to an integrated luminosity of 1.03 fb−1. No excess over SM back- ground expectations is observed in the data in the signal enhanced region, and 95% CL upper bounds on the production cross section are thus determined. These are translated into lower observed (ex- pected) limits on leptoquark masses of m > 660 (650) GeV and [7] OPAL Collaboration, Eur. Phys. J. C 31 (2003) 281 [8] H1 Collaboration, Phys. Lett. B 704 (2011) 388, doi:10.1016/j.physletb.2011.09. 017. [9] K. Nakamura, et al., Particle Data Group, J. Phys. G 37 (2010) 075021. [10] ATLAS Collaboration, ATLAS-CONF-2011-116, http://cdsweb.cern.ch/record/ 1376384; [10] ATLAS Collaboration, ATLAS-CONF-2011-116, http://cdsweb.cern.ch/record/ 1376384; ATLAS Collaboration, Eur. Phys. J. C 71 (2011) 1630. [10] ATLAS Collaboration, ATLAS-CONF-2011-116, htt 1376384; ATLAS Collaboration, Eur. Phys. J. C 71 (2011) 1630. ATLAS Collaboration, Eur. Phys. J. C 71 (2011) 1630. [11] ATLAS Collaboration, JINST 3 (2008) S08003. [11] ATLAS Collaboration, JINST 3 (2008) S08003. ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 164 [25] ATLAS Collaboration, ATLAS-CONF-2010-038, http://cdsweb.cern.ch/record/ 1277678; ATLAS Collaboration, arXiv:1112.6426, Eur. Phys. J. C, submitted for publication. [12] ATLAS Collaboration, arXiv:1110.3174, Eur. Phys. J. C, submitted for publication. [13] S. Agostinelli, et al., Nucl. Instrum. Meth. A 506 (2003) 2 [13] S. Agostinelli, et al., Nucl. Instrum. Meth. A 506 (2003) 250; G. Aad, Eur. Phys. J. C 70 (2010) 823. [ ] g G. Aad, Eur. Phys. J. C 70 (2010) 823. ATLAS Collaboration, arXiv:1112.6426, Eur. Phys. J. C, submitted for publication. [26] ATLAS Collaboration, Eur. Phys. J. C 72 (2012) 1844, doi:10.1140/epjc/ s10052-011-1844-6. [26] ATLAS Collaboration, Eur. Phys. J. C 72 (2012) 1844, doi:10.1140/epjc/ s10052-011-1844-6. [14] M. Mangano, et al., JHEP 0307 (2003) 001. [15] G. Corcella, et al., JHEP 0101 (2001) 010, ar [15] G. Corcella, et al., JHEP 0101 (2001) 010, arXiv:hep-ph/0011363. [27] ATLAS Collaboration, ATLAS-CONF-2011-063, http://cdsweb.cern.ch/record/ 1345743. [16] J. Butterworth, J. Forshaw, M. Seymour, Z. Phys. C 72 (1996) 637. [16] J. Butterworth, J. Forshaw, M. Seymour, Z. Phys [17] S. Frixione, B.R. Webber, JHEP 0206 (2002) 029; [17] S. Frixione, B.R. Webber, JHEP 0206 (2002) 029; S. ATLAS Collaboration Abdallah 11, A.A. Abdelalim 49, A. Abdesselam 118, O. Abdinov 10, B. Abi 112, M. Abolins 88, O.S. AbouZeid 158, H. Abramowicz 153, H. Abreu 115, E. Acerbi 89a,89b, B.S. Acharya 164a,164b, G. Aad 48, B. Abbott 111, J. Abdallah 11, A.A. Abdelalim 49, A. Abdesselam 118, O. Abdinov 10, B. Abi 112, M. Abolins 88, O.S. AbouZeid 158, H. Abramowicz 153, H. Abreu 115, E. Acerbi 89a,89b, B.S. Acharya 164a,164b, M. Abolins 88, O.S. AbouZeid 158, H. Abramowicz 153, H. Abreu 115, E. Acerbi 89a,89b, B.S. Acharya 164a,164b, L. Adamczyk 37, D.L. Adams 24, T.N. Addy 56, J. Adelman 175, M. Aderholz 99, S. Adomeit 98, P. Adragna 75, T. Adye 129, S. Aefsky 22, J.A. Aguilar-Saavedra 124b,a, M. Aharrouche 81, S.P. Ahlen 21, F. Ahles 48, A. Ahmad 148, M. Ahsan 40, G. Aielli 133a,133b, T. Akdogan 18a, T.P.A. Åkesson 79, G. Akimoto 155, A V Aki 94 A Aki 67 M S Al 1 M A Al 76 J Alb 169 S Alb d 55 M Al k 29 , , , , , p , M. Alhroob 20, M. Aliev 15, G. Alimonti 89a, J. Alison 120, M. Aliyev 10, P.P. Allport 73, S.E. Allwood-Spiers 53, J. Almond 82, A. Aloisio 102a,102b, R. Alon 171, A. Alonso 79, B. Alvarez Gonzalez 88, M.G. Alviggi 102a,102b, K. Amako 66, P. Amaral 29, C. Amelung 22, V.V. Ammosov 128, A. Amorim 124a,b, G. Amorós 167, N. Amram 153, C. Anastopoulos 29, L.S. Ancu 16, N. Andari 115, T. Andeen 34, C.F. Anders 20, G. Anders 58a, K.J. Anderson 30, A. Andreazza 89a,89b, V. Andrei 58a, M.-L. Andrieux 55, X.S. Anduaga 70, A. Angerami 34, F. Anghinolﬁ29, A. Anisenkov 107, N. Anjos 124a, A. Annovi 47, A. Antonaki 8, M. Antonelli 47, A. Antonov 96, J. Antos 144b, F. Anulli 132a, S. Aoun 83, L. Aperio Bella 4, R. Apolle 118,c, G. Arabidze 88, I. Aracena 143, Y. Arai 66, A.T.H. Arce 44, J.P. Archambault 28, J.-F. Arguin 14, E. Arik 18a,∗, M. Arik 18a, A.J. Armbruster 87, O. Arnaez 81, A. Artamonov 95, G. Artoni 132a,132b, D. Arutinov 20, S. Asai 155, R. Asfandiyarov 172, S. Ask 27, B. Åsman 146a,146b, L. Asquith 5, K. Assamagan 24, A. Astbury 169, A. Astvatsatourov 52, B. Aubert 4, E. Auge 115, K. Augsten 127, M. Aurousseau 145a, G. Avolio 163, R. Avramidou 9, D. Axen 168, C. Ay 54, G. ATLAS Collaboration G. Aad 48, B. Abbott 111, J. Abdallah 11, A.A. Abdelalim 49, A. Abdesselam 118, O. Abdinov 10, B. Abi 112, M. Abolins 88, O.S. AbouZeid 158, H. Abramowicz 153, H. Abreu 115, E. Acerbi 89a,89b, B.S. Acharya 164a,164b, L. Adamczyk 37, D.L. Adams 24, T.N. Addy 56, J. Adelman 175, M. Aderholz 99, S. Adomeit 98, P. Adragna 75, T. Adye 129, S. Aefsky 22, J.A. Aguilar-Saavedra 124b,a, M. Aharrouche 81, S.P. Ahlen 21, F. Ahles 48, A. Ahmad 148, M. Ahsan 40, G. Aielli 133a,133b, T. Akdogan 18a, T.P.A. Åkesson 79, G. Akimoto 155, A.V. Akimov 94, A. Akiyama 67, M.S. Alam 1, M.A. Alam 76, J. Albert 169, S. Albrand 55, M. Aleksa 29, I.N. Aleksandrov 65, F. Alessandria 89a, C. Alexa 25a, G. Alexander 153, G. Alexandre 49, T. Alexopoulos 9, M. Alhroob 20, M. Aliev 15, G. Alimonti 89a, J. Alison 120, M. Aliyev 10, P.P. Allport 73, S.E. Allwood-Spiers 53, J. Almond 82, A. Aloisio 102a,102b, R. Alon 171, A. Alonso 79, B. Alvarez Gonzalez 88, M.G. Alviggi 102a,102b, K. Amako 66, P. Amaral 29, C. Amelung 22, V.V. Ammosov 128, A. Amorim 124a,b, G. Amorós 167, N. Amram 153, C. Anastopoulos 29, L.S. Ancu 16, N. Andari 115, T. Andeen 34, C.F. Anders 20, G. Anders 58a, K.J. Anderson 30, A. Andreazza 89a,89b, V. Andrei 58a, M.-L. Andrieux 55, X.S. Anduaga 70, A. Angerami 34, F. Anghinolﬁ29, A. Anisenkov 107, N. Anjos 124a, A. Annovi 47, A. Antonaki 8, M. Antonelli 47, A. Antonov 96, J. Antos 144b, F. Anulli 132a, S. Aoun 83, L. Aperio Bella 4, R. Apolle 118,c, G. Arabidze 88, I. Aracena 143, Y. Arai 66, A.T.H. Arce 44, J.P. Archambault 28, J.-F. Arguin 14, E. Arik 18a,∗, M. Arik 18a, A.J. Armbruster 87, O. Arnaez 81, A. Artamonov 95, G. Artoni 132a,132b, D. Arutinov 20, S. Asai 155, R. Asfandiyarov 172, S. Ask 27, B. Åsman 146a,146b, L. Asquith 5, K. Assamagan 24, A. Astbury 169, A. Astvatsatourov 52, B. Aubert 4, E. Auge 115, K. Augsten 127, M. Aurousseau 145a, G. Avolio 163, R. Avramidou 9, D. Axen 168, C. Ay 54, G. Azuelos 93,d, Y. Azuma 155, M.A. Baak 29, G. Baccaglioni 89a, C. Bacci 134a,134b, A.M. Bach 14, H Bachacou 136 K Bachas 29 G Bachy 29 M Backes 49 M Backhaus 20 E Badescu 25a G. Aad 48, B. Abbott 111, J. 10. Conclusions Frixione, P. Nason, B.R. Webber, JHEP 0308 (20 [28] ATLAS Collaboration, ATLAS-CONF-2011-032, http://cdsweb.cern.ch/record/ 1337782. ATLAS Collaboration Bobb 107 79 44 118 41 J , J , , , , N. Bousson 83, A. Boveia 30, J. Boyd 29, I.R. Boyko 65, N.I. Bozhko 128, I. Bozovic-Jelisavcic 12b, J. Bracinik 17, 29 134 57 7 118 54 156 , , g , , , , B. Brau 84, J.E. Brau 114, H.M. Braun 174, B. Brelier 158, J. Bremer 29, R. Brenner 166, S. Bressler 171, 11 3 2 20 88 1 1 3 J F. Broggi 89a, C. Bromberg 88, J. Bronner 99, G. Brooijmans 34, W.K. Brooks 31b, G. Brown 82, H. Brown 7, 38 144b 48 61 19a 19a M. Bruschi 19a, T. Buanes 13, Q. Buat 55, F. Bucci 49, J. Buchanan 118, N.J. Buchanan 2, P. Buchholz 141, L. Bugge 117, O. Bulekov 96, M. Bunse 42, T. Buran 117, H. Burckhart 29, S. Burdin 73, T. Burgess 13, 129 33 3 166 29 143 21 C.M. Buttar 53, J.M. Butterworth 77, W. Buttinger 27, S. Cabrera Urbán 167, D. Caforio 19a,19b, O. Cakir 3a, 14 78 98 106 23 132 132b 33 E. Castaneda-Miranda 172, V. Castillo Gimenez 167, N.F. Castro 124a, G. Cataldi 72a, F. Cataneo 29, A. Catinaccio 29, J.R. Catmore 71, A. Cattai 29, G. Cattani 133a,133b, S. Caughron 88, D. Cauz 164a,164c, P. Cavalleri 78, D. Cavalli 89a, M. Cavalli-Sforza 11, V. Cavasinni 122a,122b, F. Ceradini 134a,134b, D. Chakraborty 106, K. Chan 2, B. Chapleau 85, J.D. Chapman 27, J.W. Chapman 87, E. Chareyre 78, D G Charlton 17 V Cha da 82 C A Cha e Barajas 29 S Cheatham 85 S Chekano 5 S V Chekulae 159a V. Chernyatin 24, E. Cheu 6, S.L. Cheung 158, L. Chevalier 136, G. Chiefari 102a,102b, L. Chikovani 51a, J.T. Childers 58a, A. Chilingarov 71, G. Chiodini 72a, M.V. Chizhov 65, G. Choudalakis 30, S. Chouridou 137, I.A. Christidi 77, A. Christov 48, D. Chromek-Burckhart 29, M.L. Chu 151, J. Chudoba 125, G. Ciapetti 132a,132 K. Ciba 37, A.K. Ciftci 3a, R. Ciftci 3a, D. Cinca 33, V. Cindro 74, M.D. Ciobotaru 163, C. Ciocca 19a, A. Ciocio 1 M. Cirilli 87, M. Ciubancan 25a, A. Clark 49, P.J. Clark 45, W. Cleland 123, J.C. Clemens 83, B. Clement 55, C. Clement 146a,146b, R.W. Clifft 129, Y. Coadou 83, M. Cobal 164a,164c, A. Coccaro 172, J. Cochran 64, P. Coe 1 J.G. Cogan 143, J. Coggeshall 165, E. ATLAS Collaboration Azuelos 93,d, Y. Azuma 155, M.A. Baak 29, G. Baccaglioni 89a, C. Bacci 134a,134b, A.M. Bach 14, H Bachacou 136 K Bachas 29 G Bachy 29 M Backes 49 M Backhaus 20 E Badescu 25a y P. Bagnaia 132a,132b, S. Bahinipati 2, Y. Bai 32a, D.C. Bailey 158, T. Bain 158, J.T. Baines 129, O.K. Baker 175 M D Baker 24 S Baker 77 E Banas 38 P Banerjee 93 Sw Banerjee 172 D Banﬁ29 A Bangert 150 , , , , , , T. Barber 48, E.L. Barberio 86, D. Barberis 50a,50b, M. Barbero 20, D.Y. Bardin 65, T. Barillari 99, V. Bartsch , R.L. Bates , L. Batkova , J.R. Batley , A. Battaglia , M. Battistin , F. Bauer , H.S. Bawa 143,e, S. Beale 98, B. Beare 158, T. Beau 78, P.H. Beauchemin 161, R. Beccherle 50a, P. Bechtle 20, , , g , , J , , , V.A. Bednyakov 65, C.P. Bee 83, M. Begel 24, S. Behar Harpaz 152, P.K. Behera 63, M. Beimforde 99, . e o o , . e o , O. e obo odo a , . e ots y , O. e t a e o , S. e , O. Benary 153, D. Benchekroun 135a, C. Benchouk 83, M. Bendel 81, N. Benekos 165, Y. Benhammou 153, E. Benhar Noccioli 49, J.A. Benitez Garcia 159b, D.P. Benjamin 44, M. Benoit 115, J.R. Bensinger 22, K. Benslama 130, S. Bentvelsen 105, D. Berge 29, E. Bergeaas Kuutmann 41, N. Berger 4, F. Berghaus 169, E. Berglund 105, J. Beringer 14, P. Bernat 77, R. Bernhard 48, C. Bernius 24, T. Berry 76, C. Bertella 83, A. Bertin 19a,19b, F. Bertinelli 29, F. Bertolucci 122a,122b, M.I. Besana 89a,89b, N. Besson 136, S. Bethke 99, W. Bhimji 45, R.M. Bianchi 29, M. Bianco 72a,72b, O. Biebel 98, S.P. Bieniek 77, K. Bierwagen 54, J. Biesiada 14, M. Biglietti 134a,134b, H. Bilokon 47, M. Bindi 19a,19b, S. Binet 115, A. Bingul 18c, C. Bini 132a,132b, W. Bhimji 45, R.M. Bianchi 29, M. Bianco 72a,72b, O. Biebel 98, S.P. Bieniek 77, K. Bierwagen 54, J. Biesiada 14, M. Biglietti 134a,134b, H. Bilokon 47, M. Bindi 19a,19b, S. Binet 115, A. Bingul 18c, C. Bini 132a,132b, ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 165 C. Blocker 22, J. Blocki 38, A. Blondel 49, W. Blum 81, U. Blumenschein 54, G.J. ATLAS Collaboration Espinal Curull 11, B. Esposito 47, F. Etienne 83, A.I. Etienvre 136, E. Etzion 153, D. Evangelakou 54, H. Evans 61, L. Fabbri 19a,19b, C. Fabre 29, R.M. Fakhrutdinov 128, S. Falciano 132a, Y. Fang 172, M. Fanti 89a,89b, A. Farbin 7, A. Farilla 134a, J. Farley 148, T. Farooque 158, S.M. Farrington 118, P. Farthouat 29, P. Fassnacht 29, D. Fassouliotis 8, B. Fatholahzadeh 158, A. Favareto 89a,89b, L. Fayard 115, S. Fazio 36a,36b, R. Febbraro 33, P. Federic 144a, O.L. Fedin 121, W. Fedorko 88, M. Fehling-Kaschek 48, L. Feligioni 83, C. Feng 32d, E.J. Feng 30, A.B. Fenyuk 128, J. Ferencei 144b, J. Ferland 93, W. Fernando 109, S. Ferrag 53, J. Ferrando 53, V. Ferrara 41, A. Ferrari 166, P. Ferrari 105, R. Ferrari 119a, A. Ferrer 167, M.L. Ferrer 47, D. Ferrere 49, C. Ferretti 87, A. Ferretto Parodi 50a,50b, M. Fiascaris 30, F. Fiedler 81, A. Filipˇciˇc 74, A. Filippas 9, F. Filthaut 104, M. Fincke-Keeler 169, M.C.N. Fiolhais 124a,g, L. Fiorini 167, A. Firan 39, P. Fischer 20, M.J. Fisher 109, M. Flechl 48, I. Fleck 141, J. Fleckner 81, P. Fleischmann 173, S. Fleischmann 174, T. Flick 174, L.R. Flores Castillo 172, M.J. Flowerdew 99, M. Fokitis 9, T. Fonseca Martin 16, D.A. Forbush 138, A. Formica 136, A. Forti 82, D. Fortin 159a, J.M. Foster 82, D. Fournier 115, A. Foussat 29, A.J. Fowler 44, K. Fowler 137, H. Fox 71, P. Francavilla 122a,122b, S. Franchino 119a,119b, D. Francis 29, T. Frank 171, M. Franklin 57, S. Franz 29, M. Fraternali 119a,119b, S. Fratina 120, S.T. French 27, F. Friedrich 43, R. Froeschl 29, D. Froidevaux 29, J.A. Frost 27, C. Fukunaga 156, E. Fullana Torregrosa 29, J. Fuster 167, C. Gabaldon 29, O. Gabizon 171, T. Gadfort 24, S. Gadomski 49, G. Gagliardi 50a,50b, P. Gagnon 61, C. Galea 98, E.J. Gallas 118, V. Gallo 16, B.J. Gallop 129, P. Gallus 125, K.K. Gan 109, Y.S. Gao 143,e, V.A. Gapienko 128, A. Gaponenko 14, F. Garberson 175, M. Garcia-Sciveres 14, C. García 167, J.E. García Navarro 167, R.W. Gardner 30, N. Garelli 29, H. Garitaonandia 105, V. Garonne 29, J. Garvey 17, C. Gatti 47, G. Gaudio 119a, O. Gaumer 49, B. Gaur 141, L. Gauthier 136 I.L. Gavrilenko 94 C. Gay 168 G. Gaycken 20 J.-C. Gayde 29 E.N. Gazis 9 P. ATLAS Collaboration Cogneras 177, J. Colas 4, A.P. Colijn 105, C. Collard 115, N.J. Collins 17, C. Collins-Tooth 53, J. Collot 55, G. Colon 84, P. Conde Muiño 124a, E. Coniavitis 118, M.C. Conidi 11, M. Consonni 104, V. Consorti 48, S. Constantinescu 25a, C. Conta 119a,119b, F. Conventi 102a,h, J. Cook 29, M. Cooke 14, B.D. Cooper 77, A.M. Cooper-Sarkar 118, K. Copic 14, T. Cornelissen 174, M. Corradi 19a, F. Corriveau 85,i, A. Cortes-Gonzalez 165, G. Cortiana 99, G. Costa 89a, M.J. Costa 167, D. Costanzo 139, T. Costin 30, D. Côté 29, L. Courneyea 169, G. Cowan 76, C. Cowden 27, B.E. Cox 82, K. Cranmer 108, F. Crescioli 122a,122b, M. Cristinziani 20, G. Crosetti 36a,36b, R. Crupi 72a,72b, S. Crépé-Renaudin 55, C.-M. Cuciuc 25a, C. Cuenca Almenar 175, T. Cuhadar Donszelmann 139, M. Curatolo 47, C.J. Curtis 17, C. Cuthbert 150, P. Cwetanski 61, H. Czirr 141, P. Czodrowski 43, Z. Czyczula 175, S. D’Auria 53, M. D’Onofrio 73, A. D’Orazio 132a,132b, P.V.M. Da Silva 23a, C. Da Via 82, W. Dabrowski 37, T. Dai 87, C. Dallapiccola 84, M. Dam 35, M. Dameri 50a,50b, D.S. Damiani 137, H.O. Danielsson 29, D. Dannheim 99, V. Dao 49, G. Darbo 50a, G.L. Darlea 25b, C. Daum 105, W. Davey 20, T. Davidek 126, N. Davidson 86, 166 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 R. Davidson 71, E. Davies 118,c, M. Davies 93, A.R. Davison 77, Y. Davygora 58a, E. Dawe 142, I. Dawson 139, J.W. Dawson 5,∗, R.K. Daya 22, K. De 7, R. de Asmundis 102a, S. De Castro 19a,19b, P.E. De Castro Faria Salgado 24, S. De Cecco 78, J. de Graat 98, N. De Groot 104, P. de Jong 105, C. De La Taille 115, H. De la Torre 80, B. De Lotto 164a,164c, L. de Mora 71, L. De Nooij 105, D. De Pedis 132a, A. De Salvo 132a, U. De Sanctis 164a,164c, A. De Santo 149, J.B. De Vivie De Regie 115, S. Dean 77, W.J. Dearnaley 71, R. Debbe 24, C. Debenedetti 45, D.V. Dedovich 65, J. Degenhardt 120, M. Dehchar 118, C. Del Papa 164a,164c, J. Del Peso 80, T. Del Prete 122a,122b, T. Delemontex 55, M. Deliyergiyev 74, A. Dell’Acqua 29, L. Dell’Asta 21, M. Della Pietra 102a,h, D. della Volpe 102a,102b, M. Delmastro 4, N. Delruelle 29, P.A. Delsart 55, C. Deluca 148, S. Demers 175, M. ATLAS Collaboration Demichev 65, B. Demirkoz 11,j, J. Deng 163, S.P. Denisov 128, D. Derendarz 38, J.E. Derkaoui 135d, F. Derue 78, P. Dervan 73, K. Desch 20, E. Devetak 148, P.O. Deviveiros 105, A. Dewhurst 129, B. DeWilde 148, S. Dhaliwal 158, R. Dhullipudi 24,k, A. Di Ciaccio 133a,133b, L. Di Ciaccio 4, A. Di Girolamo 29, B. Di Girolamo 29, S. Di Luise 134a,134b, A. Di Mattia 172, B. Di Micco 29, R. Di Nardo 47, A. Di Simone 133a,133b, R. Di Sipio 19a,19b, M.A. Diaz 31a, F. Diblen 18c, E.B. Diehl 87, J. Dietrich 41, T.A. Dietzsch 58a, S. Diglio 86, K. Dindar Yagci 39, J. Dingfelder 20, C. Dionisi 132a,132b, P. Dita 25a, S. Dita 25a, F. Dittus 29, F. Djama 83, T. Djobava 51b, M.A.B. do Vale 23c, A. Do Valle Wemans 124a, T.K.O. Doan 4, M. Dobbs 85, R. Dobinson 29,∗, D. Dobos 29, E. Dobson 29,l, M. Dobson 163, J. Dodd 34, C. Doglioni 49, T. Doherty 53, Y. Doi 66,∗, J. Dolejsi 126, I. Dolenc 74, Z. Dolezal 126, B.A. Dolgoshein 96,∗, T. Dohmae 155, M. Donadelli 23d, M. Donega 120, J. Donini 33, J. Dopke 29, A. Doria 102a, A. Dos Anjos 172, M. Dosil 11, A. Dotti 122a,122b, M.T. Dova 70, J.D. Dowell 17, A.D. Doxiadis 105, A.T. Doyle 53, Z. Drasal 126, J. Drees 174, N. Dressnandt 120, H. Drevermann 29, C. Driouichi 35, M. Dris 9, J. Dubbert 99, S. Dube 14, E. Duchovni 171, G. Duckeck 98, A. Dudarev 29, F. Dudziak 64, M. Dührssen 29, I.P. Duerdoth 82, L. Duﬂot 115, M.-A. Dufour 85, M. Dunford 29, H. Duran Yildiz 3b, R. Duxﬁeld 139, M. Dwuznik 37, F. Dydak 29, M. Düren 52, W.L. Ebenstein 44, J. Ebke 98, S. Eckweiler 81, K. Edmonds 81, C.A. Edwards 76, N.C. Edwards 53, W. Ehrenfeld 41, T. Ehrich 99, T. Eifert 143, G. Eigen 13, K. Einsweiler 14, E. Eisenhandler 75, T. Ekelof 166, M. El Kacimi 135c, M. Ellert 166, S. Elles 4, F. Ellinghaus 81, K. Ellis 75, N. Ellis 29, J. Elmsheuser 98, M. Elsing 29, D. Emeliyanov 129, R. Engelmann 148, A. Engl 98, B. Epp 62, A. Eppig 87, J. Erdmann 54, A. Ereditato 16, D. Eriksson 146a, J. Ernst 1, M. Ernst 24, J. Ernwein 136, D. Errede 165, S. Errede 165, E. Ertel 81, M. Escalier 115, C. Escobar 123, X. ATLAS Collaboration Ge 32d ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 166 R. Davidson 71, E. Davies 118,c, M. Davies 93, A.R. Davison 77, Y. Davygora 58a, E. Dawe 142, I. Dawson 139, 5 22 7 102 19 19b R. Debbe , C. Debenedetti , D.V. Dedovich , J. Degenhardt , M. Dehchar , C. Del Papa , , J. Del Peso 80, T. Del Prete 122a,122b, T. Delemontex 55, M. Deliyergiyev 74, A. Dell’Acqua 29, L. Dell’Asta 21, 102 h 102 102b 4 29 55 148 , p , , , , S. Demers 175, M. Demichev 65, B. Demirkoz 11,j, J. Deng 163, S.P. Denisov 128, D. Derendarz 38, 135d 78 73 20 148 105 J , , , , , , A. Dewhurst 129, B. DeWilde 148, S. Dhaliwal 158, R. Dhullipudi 24,k, A. Di Ciaccio 133a,133b, L. Di Ciaccio 4, A Di Gi l 29 B Di Gi l 29 S Di L i 134a 134b A Di M tti 172 B Di Mi 29 R Di N d 47 , , , , , A. Di Simone 133a,133b, R. Di Sipio 19a,19b, M.A. Diaz 31a, F. Diblen 18c, E.B. Diehl 87, J. Dietrich 41, A i h 58a S i li 86 i d i 39 i f ld 20 C i i i 132a 132b i 25a S i ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 167 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 1 V. Gilewsky 91, D. Gillberg 28, A.R. Gillman 129, D.M. Gingrich 2,d, J. Ginzburg 153, N. Giokaris 8, M.P. Giordani 164c, R. Giordano 102a,102b, F.M. Giorgi 15, P. Giovannini 99, P.F. Giraud 136, D. Giugni 89a, M. Giunta 93, P. Giusti 19a, B.K. Gjelsten 117, L.K. Gladilin 97, C. Glasman 80, J. Glatzer 48, A. Glazov 41, G.L. Glonti 65, J.R. Goddard 75, J. Godfrey 142, J. Godlewski 29, M. Goebel 41, T. Göpfert 43, C. Goeringer 81, C. Gössling 42, T. Göttfert 99, S. Goldfarb 87, T. Golling 175, S.N. Golovnia 128, A. Gomes 124a,b, L.S. Gomez Fajardo 41, R. Gonçalo 76, J. Goncalves Pinto Firmino Da Costa 41, L. Gonella 20, A. Gonidec 29, S. Gonzalez 172, S. González de la Hoz 167, G. Gonzalez Parra 11, M.L. Gonzalez Silva 26, S. Gonzalez-Sevilla 49, J.J. Goodson 148, L. Goossens 29, P.A. Gorbounov 95, H.A. Gordon 24, I. Gorelov 103, G. Gorﬁne 174, B. ATLAS Collaboration Gorini 29, E. Gorini 72a,72b, A. Gorišek 74, E. Gornicki 38, S.A. Gorokhov 128, V.N. Goryachev 128, B. Gosdzik 41, M. Gosselink 105, M.I. Gostkin 65, I. Gough Eschrich 163, M. Gouighri 135 D. Goujdami 135c, M.P. Goulette 49, A.G. Goussiou 138, C. Goy 4, S. Gozpinar 22, I. Grabowska-Bold 37, P. Grafström 29, K.-J. Grahn 41, F. Grancagnolo 72a, S. Grancagnolo 15, V. Grassi 148, V. Gratchev 121, N. Grau 34, H.M. Gray 29, J.A. Gray 148, E. Graziani 134a, O.G. Grebenyuk 121, T. Greenshaw 73, Z.D. Greenwood 24,k, K. Gregersen 35, I.M. Gregor 41, P. Grenier 143, J. Griﬃths 138, N. Grigalashvili 65, A.A. Grillo 137, S. Grinstein 11, Y.V. Grishkevich 97, J.-F. Grivaz 115, M. Groh 99, E. Gross 171, J. Grosse-Knetter 54, J. Groth-Jensen 171, K. Grybel 141, V.J. Guarino 5, D. Guest 175, C. Guicheney 33, A. Guida 72a,72b, S. Guindon 54, H. Guler 85,m, J. Gunther 125, B. Guo 158, J. Guo 34, A. Gupta 30, Y. Gusakov 65, V.N. Gushchin 128, A. Gutierrez 93, P. Gutierrez 111, N. Guttman 153, O. Gutzwiller 172, C. Guyot 136, C. Gwenlan 118, C.B. Gwilliam 73, A. Haas 143, S. Haas 29, C. Haber 14, R. Hackenburg 24, H.K. Hadavand 39, D.R. Hadley 17, P. Haefner 99, F. Hahn 29, S. Haider 29, Z. Hajduk 38, H. Hakobyan 176, J. Haller 54, K. Hamacher 174, P. Hamal 113, M. Hamer 54, A. Hamilton 145b, S. Hamilton 161, H. Han 32a, L. Han 32b, K. Hanagaki 116, K. Hanawa 160, M. Hance 14, C. Handel 81, P. Hanke 58a, J.R. Hansen 35, J.B. Hansen 35, J.D. Hansen 35, P.H. Hansen 35, P. Hansson 143, K. Hara 160, G.A. Hare 137, T. Harenberg 174, S. Harkusha 90, D. Harper 87, R.D. Harrington 45, O.M. Harris 138, K. Harrison 17, J. Hartert 48, F. Hartjes 105 T. Haruyama 66, A. Harvey 56, S. Hasegawa 101, Y. Hasegawa 140, S. Hassani 136, M. Hatch 29, D. Hauff 99, S. Haug 16, M. Hauschild 29, R. Hauser 88, M. Havranek 20, B.M. Hawes 118, C.M. Hawkes 17, R.J. Hawkings 29, A.D. Hawkins 79, D. Hawkins 163, T. Hayakawa 67, T. Hayashi 160, D. Hayden 76, H.S. Hayward 73, S.J. Haywood 129, E. Hazen 21, M. He 32d, S.J. Head 17, V. Hedberg 79, L. Heelan 7, S. Heim 88, B. Heinemann 14, S. Heisterkamp 35, L. Helary 4, C. Heller 98, M. ATLAS Collaboration Heller 29, S. Hellman 146a,146b D. Hellmich 20, C. Helsens 11, R.C.W. Henderson 71, M. Henke 58a, A. Henrichs 54, A.M. Henriques Correia 29, S. Henrot-Versille 115, F. Henry-Couannier 83, C. Hensel 54, T. Henß 174, C.M. Hernandez 7, Y. Hernández Jiménez 167, R. Herrberg 15, A.D. Hershenhorn 152, G. Herten 48, R. Hertenberger 98, L. Hervas 29, N.P. Hessey 105, E. Higón-Rodriguez 167, D. Hill 5,∗, J.C. Hill 27, N. Hill 5, K.H. Hiller 41, S. Hillert 20, S.J. Hillier 17, I. Hinchliffe 14, E. Hines 120, M. Hirose 116, F. Hirsch 42, D. Hirschbuehl 174, J. Hobbs 148, N. Hod 153, M.C. Hodgkinson 139, P. Hodgson 139, A. Hoecker 29, M.R. Hoeferkamp 103, J. Hoffman 39, D. Hoffmann 83, M. Hohlfeld 81, M. Holder 141, S.O. Holmgren 146a, T. Holy 127, J.L. Holzbauer 88, Y. Homma 67, T.M. Hong 120, L. Hooft van Huysduynen 108, T. Horazdovsky 127, C. Horn 143, S. Horner 48, J.-Y. Hostachy 55, S. Hou 151, M.A. Houlden 73, A. Hoummada 135a, J. Howarth 82, D.F. Howell 118, I. Hristova 15, J. Hrivnac 115, I. Hruska 125, T. Hryn’ova 4 P.J. Hsu 81, S.-C. Hsu 14, G.S. Huang 111, Z. Hubacek 127, F. Hubaut 83, F. Huegging 20, A. Huettmann 41, T.B. Huffman 118, E.W. Hughes 34, G. Hughes 71, R.E. Hughes-Jones 82, M. Huhtinen 29, P. Hurst 57, M. Hurwitz 14, U. Husemann 41, N. Huseynov 65,n, J. Huston 88, J. Huth 57, G. Iacobucci 49, G. Iakovidis 9, M. Ibbotson 82, I. Ibragimov 141, R. Ichimiya 67, L. Iconomidou-Fayard 115, J. Idarraga 115, P. Iengo 102a,102b O. Igonkina 105, Y. Ikegami 66, M. Ikeno 66, Y. Ilchenko 39, D. Iliadis 154, N. Ilic 158, D. Imbault 78, M. Imori 155, T. Ince 20, J. Inigo-Golﬁn 29, P. Ioannou 8, M. Iodice 134a, V. Ippolito 132a,132b, A. Irles Quiles 167, C. Isaksson 166, A. Ishikawa 67, M. Ishino 68, R. Ishmukhametov 39, C. Issever 118, S. Istin 18a, A.V. Ivashin 128, W. Iwanski 38, H. Iwasaki 66, J.M. Izen 40, V. Izzo 102a, B. Jackson 120, J.N. Jackson 73, P. Jackson 143, M.R. Jaekel 29, V. Jain 61, K. Jakobs 48, S. Jakobsen 35, J. Jakubek 127, D.K. Jana 111, E. Jankowski 158, E. Jansen 77, H. Jansen 29, A. Jantsch 99, M. Janus 20, G. Jarlskog 79, L. Jeanty 57, K. Jelen 37, I. Jen-La Plante 30, P. Jenni 29, A. ATLAS Collaboration Jeremie 4, P. Jež 35, S. Jézéquel 4, M.K. Jha 19a, 172 81 148 32b 41 32b 32a 157 g , , g , J g , , Giordano 102a,102b, F.M. Giorgi 15, P. Giovannini 99, P.F. Giraud 136, D. Giugni 89a, g g usti 19a, B.K. Gjelsten 117, L.K. Gladilin 97, C. Glasman 80, J. Glatzer 48, A. Glazov 41, J J B. Gorini 29, E. Gorini 72a,72b, A. Gorišek 74, E. Gornicki 38, S.A. Gorokhov 128, J g g N. Grau 34, H.M. Gray 29, J.A. Gray 148, E. Graziani 134a, O.G. Grebenyuk 121, T. Greenshaw 73, 24 k 35 41 143 138 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 168 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 169 G. Luijckx , D. Lumb , L. Luminari , E. Lund , B. Lund Jensen , B. Lundberg , J. Lundberg 146a,146b, J. Lundquist 35, M. Lungwitz 81, G. Lutz 99, D. Lynn 24, J. Lys 14, E. Lytken 79, H. Ma 24 172 i 93 47 hi l 99 ˇ k 74 h d i 124a , J , , , , J g , R. Mackeprang 35, R.J. Madaras 14, W.F. Mader 43, R. Maenner 58c, T. Maeno 24, P. Mättig 174, S. Mättig 41, 29 54 153 48 17 132 132b J , , , , g , Magradze 54, Y. Mahalalel 153, K. Mahboubi 48, G. Mahout 17, C. Maiani 132a,132b g g C. Maidantchik 23a, A. Maio 124a,b, S. Majewski 24, Y. Makida 66, N. Makovec 115, P. Mal 136, B. Malaescu 29, 38 38 121 55 63 5 143 , , j , , , Pa. Malecki 38, P. Malecki 38, V.P. Maleev 121, F. Malek 55, U. Mallik 63, D. Malon 5, S. Maltezos 9, V. Malyshev 107, S. Malyukov 29, R. Mameghani 98, J. Mamuzic 12b, S. Maltezos 9, V. Malyshev 107, S. Malyukov 29, R. Mameghani 98, J. Mamuzic 12b, A. Manabe 66, 89 74 15 124 88 , y , y , g , J , L. Mandelli 89a, I. Mandi´c 74, R. Mandrysch 15, J. Maneira 124a, P.S. Mangeard 88, L. Manhaes de Andrade Filho 23a, I.D. Manjavidze 65, A. Mann 54, P.M. Manning 137, 8 136 99 29 29 j g A. Manousakis-Katsikakis 8, B. Mansoulie 136, A. Manz 99, A. Mapelli 29, L. Mapell p p J.F. Marchand 28, F. ATLAS Collaboration Marchese 133a,133b, G. Marchiori 78, M. Marcisovsky 125, A. Marin 21,∗, C.P. Marino 1 q J B. Martin 29, B. Martin 88, F.F. Martin 120, J.P. Martin 93, Ph. Martin 55, T.A. Martin 17, V.J. Martin 45, A.C. Martyniuk 169, M. Marx 82, F. Marzano 132a, A. Marzin 111, L. Masetti 81, T. Mashimo 155, y , , , , , , R. Mashinistov 94, J. Masik 82, A.L. Maslennikov 107, I. Massa 19a,19b, G. Massaro 105, N. Massol 4, P. Mastrandrea 132a,132b, A. Mastroberardino 36a,36b, T. Masubuchi 155, M. Mathes 20, H. Matsumoto 155, H. Matsunaga 155, T. Matsushita 67, C. Mattravers 118,c, J.M. Maugain 29, J. Maurer 83, S.J. Maxﬁeld 73, H. Matsunaga , T. Matsushita , C. Mattravers , J.M. Maugain , J. Maurer , S.J. Maxﬁeld , D.A. Maximov 107, E.N. May 5, A. Mayne 139, R. Mazini 151, M. Mazur 20, M. Mazzanti 89a, E. Mazzoni 122a,122b, S.P. Mc Kee 87, A. McCarn 165, R.L. McCarthy 148, T.G. McCarthy 28, N.A. McCubbin 129, K.W. McFarlane 56, J.A. Mcfayden 139, H. McGlone 53, G. Mchedlidze 51b, R.A. McLaren 29, T. Mclaughlan 17, S.J. McMahon 129, R.A. McPherson 169,i, A. Meade 84, J. Mechnich 105, M. Mechtel 174, M. Medinnis 41, R. Meera-Lebbai 111, T. Meguro 116, R. Mehdiyev 93, S. Mehlhase 35, A. Mehta 73, K. Meier 58a, g A. Mengarelli 19a,19b, S. Menke 99, C. Menot 29, E. Meoni 11, K.M. Mercurio 57, P. Mermod 49, ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 170 170 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 A. Neusiedl 81, R.M. Neves 108, P. Nevski 24, P.R. Newman 17, V. Nguyen Thi Hong 136, R.B. Nickerson 118, R. Nicolaidou 136, L. Nicolas 139, B. Nicquevert 29, F. Niedercorn 115, J. Nielsen 137, T. Niinikoski 29, N. Nikiforou 34, A. Nikiforov 15, V. Nikolaenko 128, K. Nikolaev 65, I. Nikolic-Audit 78, K. Nikolics 49, K. Nikolopoulos 24, H. Nilsen 48, P. Nilsson 7, Y. Ninomiya 155, A. Nisati 132a, T. Nishiyama 67, R. Nisius 99, L. Nodulman 5, M. Nomachi 116, I. Nomidis 154, M. Nordberg 29, B. Nordkvist 146a,146b, P.R. Norton 129, J. Novakova 126, M. Nozaki 66, L. Nozka 113, I.M. Nugent 159a, A.-E. Nuncio-Quiroz 20, G. Nunes Hanninger 86, T. Nunnemann 98, E. Nurse 77, T. Nyman 29, B.J. O’Brien 45, S.W. O’Neale 17,∗, D.C. O’Neil 142, V. O’Shea 53, L.B. Oakes 98, F.G. ATLAS Collaboration Oakham 28,d, H. Oberlack 99, J. Ocariz 78, A. Ochi 67, S. Oda 155, S. Odaka 66, J. Odier 83, H. Ogren 61, A. Oh 82, S.H. Oh 44, C.C. Ohm 146a,146b, T. Ohshima 101, H. Ohshita 140, T. Ohsugi 59, S. Okada 67, H. Okawa 163, Y. Okumura 101, T. Okuyama 155, A. Olariu 25a, M. Olcese 50a, A.G. Olchevski 65, M. Oliveira 124a,g, D. Oliveira Damazio 24, E. Oliver Garcia 167, D. Olivito 120, A. Olszewski 38, J. Olszowska 38, C. Omachi 67, A. Onofre 124a,v, P.U.E. Onyisi 30, C.J. Oram 159a, M.J. Oreglia 30, Y. Oren 153, D. Orestano 134a,134b, C. Oropeza Barrera 53, R.S. Orr 158, B. Osculati 50a,50b, R. Ospanov 120, C. Osuna 11, G. Otero y Garzon 26, J.P. Ottersbach 105, M. Ouchrif 135d, E.A. Ouellette 169, F. Ould-Saada 117, A. Ouraou 136, Q. Ouyang 32a, A. Ovcharova 14, M. Owen 82, S. Owen 139, V.E. Ozcan 18a, N. Ozturk 7, A. Pacheco Pages 11, C. Padilla Aranda 11, S. Pagan Griso 14, E. Paganis 139, F. Paige 24, P. Pais 84, K. Pajchel 117, G. Palacino 159b, C.P. Paleari 6, S. Palestini 29, D. Pallin 33, A. Palma 124a, J.D. Palmer 17, Y.B. Pan 172, E. Panagiotopoulou 9, B. Panes 31a, N. Panikashvili 87, S. Panitkin 24, D. Pantea 25a, M. Panuskova 125, V. Paolone 123, A. Papadelis 146a, Th.D. Papadopoulou 9, A. Paramonov 5, W. Park 24,w, M.A. Parker 27, F. Parodi 50a,50b, J.A. Parsons 34, U. Parzefall 48, E. Pasqualucci 132a, S. Passaggio 50a, A. Passeri 134a, F. Pastore 134a,134b, Fr. Pastore 76, G. Pásztor 49,x, S. Pataraia 174, N. Patel 150, J.R. Pater 82, S. Patricelli 102a,102b, T. Pauly 29, M. Pecsy 144a, M.I. Pedraza Morales 172, S.V. Peleganchuk 107, H. Peng 32b, R. Pengo 29, A. Penson 34, J. Penwell 61, M. Perantoni 23a, K. Perez 34,y, T. Perez Cavalcanti 41, E. Perez Codina 11, M.T. Pérez García-Estañ 167, V. Perez Reale 34, L. Perini 89a,89b, H. Pernegger 29, R. Perrino 72a, P. Perrodo 4, S. Persembe 3a, V.D. Peshekhonov 65, B.A. Petersen 29, J. Petersen 29, T.C. Petersen 35, E. Petit 4, A. Petridis 154, C. Petridou 154, E. Petrolo 132a, F. Petrucci 134a,134b, D. Petschull 41, M. Petteni 142, R. Pezoa 31b, A. Phan 86, P.W. Phillips 129, G. Piacquadio 29, E. Piccaro 75, M. Piccinini 19a,19b, S.M. Piec 41, R. Piegaia 26, D.T. Pignotti 109, J.E. ATLAS Collaboration Pilcher 30, A.D. Pilkington 82, J. Pina 124a,b, M. Pinamonti 164a,164c, A. Pinder 118, J.L. Pinfold 2, J. Ping 32c, B. Pinto 124a,b, O. Pirotte 29, C. Pizio 89a,89b, R. Placakyte 41, M. Plamondon 169, M.-A. Pleier 24, A.V. Pleskach 128, A. Poblaguev 24, S. Poddar 58a, F. Podlyski 33, L. Poggioli 115, T. Poghosyan 20, M. Pohl 49, F. Polci 55, G. Polesello 119a, A. Policicchio 138, A. Polini 19a, J. Poll 75, V. Polychronakos 24, D.M. Pomarede 136, D. Pomeroy 22, K. Pommès 29, L. Pontecorvo 132a, B.G. Pope 88, G.A. Popeneciu 25a, D.S. Popovic 12a, A. Poppleton 29, X. Portell Bueso 29, C. Posch 21, G.E. Pospelov 99, S. Pospisil 127, I.N. Potrap 99, C.J. Potter 149, C.T. Potter 114, G. Poulard 29, J. Poveda 172, R. Prabhu 77, P. Pralavorio 83, A. Pranko 14, S. Prasad 57, R. Pravahan 7, S. Prell 64, K. Pretzl 16, L. Pribyl 29, D. Price 61, J. Price 73, L.E. Price 5, M.J. Price 29, D. Prieur 123, M. Primavera 72a, K. Prokoﬁev 108, F. Prokoshin 31b, S. Protopopescu 24, J. Proudfoot 5, X. Prudent 43, M. Przybycien 37, H. Przysiezniak 4, S. Psoroulas 20, E. Ptacek 114, E. Pueschel 84, J. Purdham 87, M. Purohit 24,w, P. Puzo 115, Y. Pylypchenko 63, J. Qian 87, Z. Qian 83, Z. Qin 41, A. Quadt 54, D.R. Quarrie 14, W.B. Quayle 172, F. Quinonez 31a, M. Raas 104, V. Radescu 58b, B. Radics 20, P. Radloff 114, T. Rador 18a, F. Ragusa 89a,89b, G. Rahal 177, A.M. Rahimi 109, D. Rahm 24, S. Rajagopalan 24, M. Rammensee 48, M. Rammes 141, A.S. Randle-Conde 39, K. Randrianarivony 28, P.N. Ratoff 71, F. Rauscher 98, M. Raymond 29, A.L. Read 117, D.M. Rebuzzi 119a,119b, A. Redelbach 173, G. Redlinger 24, R. Reece 120, K. Reeves 40, A. Reichold 105, E. Reinherz-Aronis 153, A. Reinsch 114, I. Reisinger 42, D. Reljic 12a, C. Rembser 29, Z.L. Ren 151, A. Renaud 115, P. Renkel 39, M. Rescigno 132a, S. Resconi 89a, B. Resende 136, P. Reznicek 98, R. Rezvani 158, A. Richards 77, R. Richter 99, E. Richter-Was 4,z, M. Ridel 78, M. Rijpstra 105, M. Rijssenbeek 148, A. Rimoldi 119a,119b, L. Rinaldi 19a, R.R. Rios 39, I. Riu 11, G. Rivoltella 89a,89b, F. Rizatdinova 112, E. Rizvi 75, S.H. Robertson 85,i, A. Robichaud-Veronneau 118, D. Robinson 27, J.E.M. Robinson 77, M. Robinson 114, A. ATLAS Collaboration Robson 53, J.G. Rocha de Lima 106, C. Roda 122a,122b, D. Roda Dos Santos 29, D. Rodriguez 162, A. Roe 54, S. Roe 29, O. Røhne 117, V. Rojo 1, S. Rolli 161, A. Romaniouk 96, M. Romano 19a,19b, V.M. Romanov 65, G. Romeo 26, E. Romero Adam 167, L. Roos 78, E. Ros 167, S. Rosati 132a,132b, K. Rosbach 49, A. Rose 149, M. Rose 76, A. Neusiedl 81, R.M. Neves 108, P. Nevski 24, P.R. Newman 17, V. Nguyen Thi Hong 136, R.B. Nickerson 118, R. Nicolaidou 136, L. Nicolas 139, B. Nicquevert 29, F. Niedercorn 115, J. Nielsen 137, T. Niinikoski 29, N. Nikiforou 34, A. Nikiforov 15, V. Nikolaenko 128, K. Nikolaev 65, I. Nikolic-Audit 78, K. Nikolics 49, K. Nikolopoulos 24, H. Nilsen 48, P. Nilsson 7, Y. Ninomiya 155, A. Nisati 132a, T. Nishiyama 67, R. Nisius 99, L. Nodulman 5, M. Nomachi 116, I. Nomidis 154, M. Nordberg 29, B. Nordkvist 146a,146b, P.R. Norton 129, J Novakova 126 M Nozaki 66 L Nozka 113 I M Nugent 159a A E Nuncio Quiroz 20 J y C.J. Oram 159a, M.J. Oreglia 30, Y. Oren 153, D. Orestano 134a,134b, C. Oropeza Barrera 53, R.S. Orr 158, 50 50b 120 11 26 105 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 171 G.A. Rosenbaum 158, E.I. Rosenberg 64, P.L. Rosendahl 13, O. Rosenthal 141, L. Rosselet 49, V. Rossetti 11, E. Rossi 132a,132b, L.P. Rossi 50a, M. Rotaru 25a, I. Roth 171, J. Rothberg 138, D. Rousseau 115, C.R. Royon 136, A. Rozanov 83, Y. Rozen 152, X. Ruan 115,aa, I. Rubinskiy 41, B. Ruckert 98, N. Ruckstuhl 105, V.I. Rud 97, C. Rudolph 43, F. Rühr 6, F. Ruggieri 134a,134b, A. Ruiz-Martinez 64, V. Rumiantsev 91,∗, L. Rumyantsev 65, K. Runge 48, Z. Rurikova 48, N.A. Rusakovich 65, D.R. Rust 61, J.P. Rutherfoord 6, C. Ruwiedel 14, P. Ruzicka 125, Y.F. Ryabov 121, V. Ryadovikov 128, P. Ryan 88, M. Rybar 126, G. Rybkin 115, N.C. Ryder 118, S. Rzaeva 10, A.F. Saavedra 150, I. Sadeh 153, H.F.-W. Sadrozinski 137, R. Sadykov 65, F. Safai Tehrani 132a,132b, H. Sakamoto 155, G. Salamanna 75, A. Salamon 133a, M. Saleem 111, D. Salihagic 99, A. Salnikov 143, J. Salt 167, B.M. Salvachua Ferrando 5, D. Salvatore 36a,36b, F. Salvatore 149, A. Salvucci 104, A. Salzburger 29, D. Sampsonidis 154, B.H. Samset 117, A. Sanchez 102a,102b, H. ATLAS Collaboration Sandaker 13, H.G. Sander 81, M.P. Sanders 98, M. Sandhoff 174, T. Sandoval 27, C. Sandoval 162, R. Sandstroem 99, S. Sandvoss 174, D.P.C. Sankey 129, A. Sansoni 47, C. Santamarina Rios 85, C. Santoni 33, R. Santonico 133a,133b, H. Santos 124a, J.G. Saraiva 124a, T. Sarangi 172, E. Sarkisyan-Grinbaum 7, F. Sarri 122a,122b, G. Sartisohn 174, O. Sasaki 66, T. Sasaki 66, N. Sasao 68, I. Satsounkevitch 90, G. Sauvage 4, E. Sauvan 4, J.B. Sauvan 115, P. Savard 158,d, V. Savinov 123, D.O. Savu 29, L. Sawyer 24,k, D.H. Saxon 53, L.P. Says 33, C. Sbarra 19a, A. Sbrizzi 19a,19b, O. Scallon 93, D.A. Scannicchio 163, M. Scarcella 150, J. Schaarschmidt 115, P. Schacht 99, U. Schäfer 81, S. Schaepe 20, S. Schaetzel 58b, A.C. Schaffer 115, D. Schaile 98, R.D. Schamberger 148, V. Scharf 58a, V.A. Schegelsky 121, D. Scheirich 87, M. Schernau 163, M.I. Scherzer 34, C. Schiavi 50a,50b, J. Schieck 98, M. Schioppa 36a,36b, S. Schlenker 29, J.L. Schlereth 5, E. Schmidt 48, K. Schmieden 20, C. Schmitt 81, S. Schmitt 58b, M. Schmitz 20, A. Schöning 58b, M. Schott 29, D. Schouten 159a, J. Schovancova 125, M. Schram 85, C. Schroeder 81, N. Schroer 58c, S. Schuh 29, G. Schuler 29, J. Schultes 174, H.-C. Schultz-Coulon 58a, H. Schulz 15, J.W. Schumacher 20, M. Schumacher 48, B.A. Schumm 137, Ph. Schune 136, C. Schwanenberger 82, A. Schwartzman 143, Ph. Schwemling 78, R. Schwienhorst 88, R. Schwierz 43, J. Schwindling 136, T. Schwindt 20, M. Schwoerer 4, W.G. Scott 129, J. Searcy 114, G. Sedov 41, E. Sedykh 121, E. Segura 11, S.C. Seidel 103, A. Seiden 137, F. Seifert 43, J.M. Seixas 23a, G. Sekhniaidze 102a, K.E. Selbach 45, D.M. Seliverstov 121, B. Sellden 146a, G. Sellers 73, M. Seman 144b, N. Semprini-Cesari 19a,19b, C. Serfon 98, L. Serin 115, R. Seuster 99, H. Severini 111, M.E. Sevior 86, A. Sfyrla 29, E. Shabalina 54, M. Shamim 114, L.Y. Shan 32a, J.T. Shank 21, Q.T. Shao 86, M. Shapiro 14, P.B. Shatalov 95, L. Shaver 6, K. Shaw 164a,164c, D. Sherman 175, P. Sherwood 77, A. Shibata 108, H. Shichi 101, S. Shimizu 29, M. Shimojima 100, T. Shin 56, M. Shiyakova 65, A. Shmeleva 94, M.J. Shochet 30, D. Short 118, S. Shrestha 64, M.A. Shupe 6, P. Sicho 125, A. Sidoti 132a,132b, F. Siegert 48, Dj. Sijacki 12a, O. Silbert 171, J. ATLAS Collaboration Silva 124a,b, Y. Silver 153, D. Silverstein 143, S.B. Silverstein 146a, V. Simak 127, O. Simard 136, Lj. Simic 12a, S. Simion 115, B. Simmons 77, M. Simonyan 35, P. Sinervo 158, N.B. Sinev 114, V. Sipica 141, G. Siragusa 173, A. Sircar 24, A.N. Sisakyan 65, S.Yu. Sivoklokov 97, J. Sjölin 146a,146b, T.B. Sjursen 13, L.A. Skinnari 14, H.P. Skottowe 57, K. Skovpen 107, P. Skubic 111, N. Skvorodnev 22, M. Slater 17, T. Slavicek 127, K. Sliwa 161, J. Sloper 29, V. Smakhtin 171, S.Yu. Smirnov 96, L.N. Smirnova 97, O. Smirnova 79, B.C. Smith 57, K.M. Smith 53, M. Smizanska 71, 127 94 82 111 105 24 124 S. Schlenker 29, J.L. Schlereth 5, E. Schmidt 48, K. Schmieden 20, C. Schmitt 81, S. Schmitt 58b, M. Schmitz 2 58b 29 159 125 85 81 M. Simonyan 35, P. Sinervo 158, N.B. Sinev 114, V. Sipica 141, G. Siragusa 173, A. Sircar 24, A.N. Sisakyan 65, S.Yu. Sivoklokov 97, J. Sjölin 146a,146b, T.B. Sjursen 13, L.A. Skinnari 14, H.P. Skottowe 57, K. Skovpen 107, P. Skubic 111, N. Skvorodnev 22, M. Slater 17, T. Slavicek 127, K. Sliwa 161, J. Sloper 29, V. Smakhtin 171, S.Yu. Smirnov 96, L.N. Smirnova 97, O. Smirnova 79, B.C. Smith 57, K.M. Smith 53, M. Smizanska 71, K. Smolek 127, A.A. Snesarev 94, S.W. Snow 82, J. Snow 111, J. Snuverink 105, S. Snyder 24, M. Soares 124a, R. Sobie 169,i, J. Sodomka 127, A. Soffer 153, C.A. Solans 167, M. Solar 127, J. Solc 127, E. Soldatov 96, U. Soldevila 167, E. Solfaroli Camillocci 132a,132b, A.A. Solodkov 128, O.V. Solovyanov 128, N. Soni 2, V. Sopko 127, B. Sopko 127, M. Sosebee 7, R. Soualah 164a,164c, A. Soukharev 107, S. Spagnolo 72a,72b, F. Spanò 76, R. Spighi 19a, G. Spigo 29, F. Spila 132a,132b, R. Spiwoks 29, M. Spousta 126, T. Spreitzer 158, B. Spurlock 7, R.D. St. Denis 53, T. Stahl 141, J. Stahlman 120, R. Stamen 58a, E. Stanecka 38, R.W. Stanek 5, C. Stanescu 134a, S. Stapnes 117, E.A. Starchenko 128, J. Stark 55, P. Staroba 125, P. Starovoitov 91, A. Staude 98, P. Stavina 144a, G. Stavropoulos 14, G. Steele 53, P. Steinbach 43, P. Steinberg 24, I. Stekl 127, B. Stelzer 142, H.J. Stelzer 88, O. Stelzer-Chilton 159a, H. Stenzel 52, S. Stern 99, K. Stevenson 75, G.A. Stewart 29, J.A. Stillings 20, M.C. ATLAS Collaboration Stockton 85, K. Stoerig 48, G. Stoicea 25a, S. Stonjek 99, P. Strachota 126, A.R. Stradling 7, A. Straessner 43, J. Strandberg 147, S. Strandberg 146a,146b, A. Strandlie 117, M. Strang 109, E. Strauss 143, M. Strauss 111, P. Strizenec 144b, R. Ströhmer 173, D.M. Strom 114, J.A. Strong 76,∗, R. Stroynowski 39, J. Strube 129, B. Stugu 13, I. Stumer 24,∗, J. Stupak 148, P. Sturm 174, N.A. Styles 41, D.A. Soh 151,r, D. Su 143, HS. Subramania 2, A. Succurro 11, Y. Sugaya 116, T. Sugimoto 101, C. Suhr 106, K. Suita 67, M. Suk 126, V.V. Sulin 94, S. Sultansoy 3d, T. Sumida 68, X. Sun 55, J.E. Sundermann 48, K. Suruliz 139, S. Sushkov 11, G. Susinno 36a,36b, M.R. Sutton 149, Y. Suzuki 66, Y. Suzuki 67, M. Svatos 125, Yu.M. Sviridov 128, ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 172 S. Swedish 168, I. Sykora 144a, T. Sykora 126, B. Szeless 29, J. Sánchez 167, D. Ta 105, K. Tackmann 41, A. Taffard 163, R. Taﬁrout 159a, N. Taiblum 153, Y. Takahashi 101, H. Takai 24, R. Takashima 69, H. Takeda 67, T. Takeshita 140, Y. Takubo 66, M. Talby 83, A. Talyshev 107, M.C. Tamsett 24, J. Tanaka 155, R. Tanaka 115, S. Tanaka 131, S. Tanaka 66, Y. Tanaka 100, A.J. Tanasijczuk 142, K. Tani 67, N. Tannoury 83, G.P. Tappern 29, S. Tapprogge 81, D. Tardif 158, S. Tarem 152, F. Tarrade 28, G.F. Tartarelli 89a, P. Tas 126, M. Tasevsky 125, E. Tassi 36a,36b, M. Tatarkhanov 14, Y. Tayalati 135d, C. Taylor 77, F.E. Taylor 92, G.N. Taylor 86, W. Taylor 159b, M. Teinturier 115, M. Teixeira Dias Castanheira 75, P. Teixeira-Dias 76, K.K. Temming 48, H. Ten Kate 29, P.K. Teng 151, S. Terada 66, K. Terashi 155, J. Terron 80, M. Testa 47, R.J. Teuscher 158,i, J. Thadome 174, J. Therhaag 20, T. Theveneaux-Pelzer 78, M. Thioye 175, S. Thoma 48, J.P. Thomas 17, E.N. Thompson 34, P.D. Thompson 17, P.D. Thompson 158, A.S. Thompson 53, E. Thomson 120, M. Thomson 27, R.P. Thun 87, F. Tian 34, M.J. Tibbetts 14, T. Tic 125, V.O. Tikhomirov 94, Y.A. Tikhonov 107, S. Timoshenko 96, P. Tipton 175, F.J. Tique Aires Viegas 29, S. Tisserant 83, J. Tobias 48, B. Toczek 37, T. Todorov 4, S. Todorova-Nova 161, B. Toggerson 163, J. Tojo 66, S. Tokár 144a, K. ATLAS Collaboration Tokunaga 67, K. Tokushuku 66, K. Tollefson 88, M. Tomoto 101, L. Tompkins 30, K. Toms 103, G. Tong 32a, A. Tonoyan 13, C. Topfel 16, N.D. Topilin 65, I. Torchiani 29, E. Torrence 114, H. Torres 78, E. Torró Pastor 167, J. Toth 83,x, F. Touchard 83, D.R. Tovey 139, T. Trefzger 173, L. Tremblet 29, A. Tricoli 29, I.M. Trigger 159a, S. Trincaz-Duvoid 78, T.N. Trinh 78, M.F. Tripiana 70, W. Trischuk 158, A. Trivedi 24,w, B. Trocmé 55, C. Troncon 89a, M. Trottier-McDonald 142, M. Trzebinski 38, A. Trzupek 38, C. Tsarouchas 29, J.C.-L. Tseng 118, M. Tsiakiris 105, P.V. Tsiareshka 90, D. Tsionou 4,ab, G. Tsipolitis 9, V. Tsiskaridze 48, E.G. Tskhadadze 51a, I.I. Tsukerman 95, V. Tsulaia 14, J.-W. Tsung 20, S. Tsuno 66, D. Tsybychev 148, A. Tua 139, A. Tudorache 25a, V. Tudorache 25a, J.M. Tuggle 30, M. Turala 38, D. Turecek 127, I. Turk Cakir 3e, E. Turlay 105, R. Turra 89a,89b, P.M. Tuts 34, A. Tykhonov 74, M. Tylmad 146a,146b, M. Tyndel 129, G. Tzanakos 8, K. Uchida 20, I. Ueda 155, R. Ueno 28, M. Ugland 13, M. Uhlenbrock 20, M. Uhrmacher 54, F. Ukegawa 160, G. Unal 29, D.G. Underwood 5, A. Undrus 24, G. Unel 163, Y. Unno 66, D. Urbaniec 34, G. Usai 7, L. Vacavant 83, V. Vacek 127, B. Vachon 85, S. Vahsen 14, J Valenta 125 P Valente 132a S Valentinetti 19a,19b S Valkar 126 E Valladolid Gallego 167 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 173 W.C. Wong 40, G. Wooden 87, B.K. Wosiek 38, J. Wotschack 29, M.J. Woudstra 84, K.W. Wozniak 38, K. Wraight 53, C. Wright 53, M. Wright 53, B. Wrona 73, S.L. Wu 172, X. Wu 49, Y. Wu 32b,ae, E. Wulf 34, R. Wunstorf 42, B.M. Wynne 45, S. Xella 35, M. Xiao 136, S. Xie 48, Y. Xie 32a, C. Xu 32b,af , D. Xu 139, G. Xu 32a, B. Yabsley 150, S. Yacoob 145b, M. Yamada 66, H. Yamaguchi 155, A. Yamamoto 66, K. Yamamoto 64, S. Yamamoto 155, T. Yamamura 155, T. Yamanaka 155, J. Yamaoka 44, T. Yamazaki 155, Y. Yamazaki 67, Z. Yan 21, H. Yang 87, U.K. Yang 82, Y. Yang 61, Y. Yang 32a, Z. Yang 146a,146b, S. Yanush 91, Y. Yasu 66, G.V. Ybeles Smit 130, J. Ye 39, S. Ye 24, M. Yilmaz 3c, R. ATLAS Collaboration Zeller 175, M. Zeman 125, A. Zemla 38, C. Zendler 20, O. Zenin 128, T. Ženiš 144a, Z. Zenonos 122a,122b, S. Zenz 14, D. Zerwas 115, G. Zevi della Porta 57, Z. Zhan 32d, D. Zhang 32b,ad, H. Zhang 88, J. Zhang 5, X. Zhang 32d, Z. Zhang 115, L. Zhao 108, T. Zhao 138, Z. Zhao 32b, A. Zhemchugov 65, S. Zheng 32a, J. Zhong 118, B. Zhou 87, N. Zhou 163, Y. Zhou 151, C.G. Zhu 32d, H. Zhu 41, J. Zhu 87, Y. Zhu 32b, X. Zhuang 98, V. Zhuravlov 99, D. Zieminska 61, R. Zimmermann 20, S. Zimmermann 20, S. Zimmermann 48, M. Ziolkowski 141, R. Zitoun 4, L. Živkovi´c 34, V.V. Zmouchko 128,∗, G. Zobernig 172, A. Zoccoli 19a,19b, Y. Zolnierowski 4, A. Zsenei 29, M. zur Nedden 15, V. Zutshi 106, L. ATLAS Collaboration Zwalinski 29 1 University at Albany, Albany, NY, United States y y y 2 Department of Physics, University of Alberta, Edmonton AB, Canada Department of Physics, University of Alberta, Edmonton AB, Canada 3 (a)Department of Physics, Ankara University, Ankara; (b)Department of Physics, Dumlupinar University, Kutahya; (c)Department of Physics, Gazi University, Ankara; (d)Division of Physics, TOBB University of Economics and Technology, Ankara; (e)Turkish Atomic Energy Authority, Ankara, Turkey 3 (a)Department of Physics, Ankara University, Ankara; (b)Department of Physics, Dumlupinar University, Ku TOBB University of Economics and Technology, Ankara; (e)Turkish Atomic Energy Authority, Ankara, Turkey 4 3 (a)Department of Physics, Ankara University, Ankara; (b)Department of Physics, Dumlupinar University, Kutahya; (c)Department of Physics, Gazi University, Ankara; (d)Division of Physics, TOBB University of Economics and Technology, Ankara; (e)Turkish Atomic Energy Authority, Ankara, Turkey 4 4 LAPP, CNRS/IN2P3 and Université de Savoie, Annecy-le-Vieux, France , / , y , 5 High Energy Physics Division, Argonne National Laboratory, Argonne, IL, United States 6 5 High Energy Physics Division, Argonne National Laboratory, Argonne, IL, United States 5 High Energy Physics Division, Argonne National Laboratory, Argonne, IL, Unit 6 6 Department of Physics, University of Arizona, Tucson, AZ, United States 7 Department of Physics, The University of Texas at Arlington, Arlington, TX, United States 8 8 Physics Department, University of Athens, Athens, Greece y p y f 9 Physics Department, National Technical University of Athens, Zografou, Greece y p y f 9 Physics Department, National Technical University of Athens, Zografou, Greece 9 Physics Department, National Technical University of Athe 10 Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan f y j y f j 11 Institut de Física d’Altes Energies and Departament de Física de la Universitat Autònoma de Barcelona and ICREA, Barcelona, Spain b 11 Institut de Física d’Altes Energies and Departament de Física de la Universitat Autònoma de Barcelo 11 Institut de Física d’Altes Energies and Departament de Física de la Un g p 12 (a)Institute of Physics, University of Belgrade, Belgrade; (b)Vinca Institute of Nuclear Sciences, Belgrade, Serbia 12 (a)Institute of Physics, University of Belgrade, Belgrade; (b)Vinca Institute of Nuclear Sciences, Belgrade, Ser 12 (a)Institute of Physics, University of Belgrade, Belgrade; (b)Vinca Insti 13 Department for Physics and Technology, University of Bergen, Bergen p f y gy y f g g y 14 Physics Division, Lawrence Berkeley National Laboratory and University of California, Berkeley, CA, United States 14 Physics Division, Lawrence Berkeley National Laboratory and University of California, Berkeley, CA, 14 Physics Division, Lawrence Berkeley National Laboratory and Un 15 Department of Physics, Humboldt University, Berlin, Germany 15 Department of Physics, Humboldt University, Berlin, Germany p f y y y 16 Albert Einstein Center for Fundamental Physics and Laboratory 17 School of Physics and Astronomy, University of Birmingham, Birmingham, United Kingdom 8 ( ) (b) 17 School of Physics and Astronomy, University of Birmingham, Birmingham, United Kingdom School of Physics and Astronomy, University of Birmingham, Birmingham, United Kingdom 18 (a)Department of Physics, Bogazici University, Istanbul; (b)Division of Physics, Dogus University, Istanbul; (c)Department of Physics Engineering, Gaziantep University, Gaziantep; (d)Department of Physics, Istanbul Technical University, Istanbul, Turkey 19 (a)INFN S i di B l (b)Di i di Fi i U i i à di B l B l I l f y y y f g g g 18 (a)Department of Physics, Bogazici University, Istanbul; (b)Division of Physics, Dogus University, Istanbul; (c)Department of Physics Engineering, Gaziantep University, Gaziantep; (d)Department of Physics, Istanbul Technical University, Istanbul, Turkey f y y y f g g g 18 (a)Department of Physics, Bogazici University, Istanbul; (b)Division of Physics, Dogus University, Istanbul; (c)Department of Physics Engineering, Gaziantep University, Gaziantep; (d) f h b l h l b l k 18 (a)Department of Physics, Bogazici University, Istanbul; (b)Division of Physics, Dogus Univ 18 (a)Department of Physics, Bogazici University, Istanbul; (b)Division of Physics, Dogus Univers 18 (a)Department of Physics, Bogazici University, Istanbul; (b)Division of Physics, Dogus University, Istanbul; (c)Department of Physics En d (d)Department of Physics, Istanbul Technical University, Istanbul, Turkey 19 (a)INFN Sezione di Bologna; (b)Dipartimento di Fisica, Università di Bologna, Bologna, Italy 20 19 (a)INFN Sezione di Bologna; (b)Dipartimento di Fisica, Università 19 (a)INFN Sezione di Bologna; (b)Dipartimento di Fisica, Università di Bologna, Bologna, Italy 20 Physikalisches Institut, University of Bonn, Bonn, Germany 21 Department of Physics, Boston University, Boston, MA, United States p f y y 22 Department of Physics, Brandeis University, Waltham, MA, United States 22 Department of Physics, Brandeis University, Waltham, MA, United States p f y y 23 (a)Universidade Federal do Rio De Janeiro COPPE/EE/IF, Rio de Janeiro; (b)Federal University of Juiz de Fora (UFJF), Juiz de Fora; (c)Federal University of Sao Joao del Rei (UFSJ), Sao Joao del Rei; (d)Instituto de Fisica, Universidade de Sao Paulo, Sao Paulo, Brazil 23 (a)Universidade Federal do Rio De Janeiro COPPE/EE/IF, Rio de Janeiro; (b)Federal University of Juiz de Fora (UFJF), Juiz de Fora; (c)Federal University of Sao Joao del Rei (UFSJ), Sao Joao (d) 23 (a)Universidade Federal do Rio De Janeiro COPPE/EE/IF, Rio de Janeiro; (b)Federal University of Juiz de Fora (UFJF), Juiz de Fora; (d) J / / J del Rei; (d)Instituto de Fisica, Universidade de Sao Paulo, Sao Paulo, Brazil 24 Physics Department, Brookhaven National Laboratory, Upton, NY, United States b 24 Physics Department, Brookhaven National Laboratory, Upton, NY, United States aboratory, Upton, NY, United States Engineering, Bucharest; (b)University Politehnica Bucharest, Bucharest; (c)West University in Timisoara, Timisoara, Romania nos Aires Buenos Aires Argentina Physics Department, Brookhaven National Laboratory, Upton, NY, United States 25 (a)National Institute of Physics and Nuclear Engineering, Bucharest; (b)University Politehnica Bucharest, Bucharest; (c)West University in Timisoara, Timisoara, Romania 26 D t t d Fí i U i id d d B Ai B Ai A ti 25 (a)National Institute of Physics and Nuclear Engineering, Bucharest; (b)University Politehni 26 25 (a)National Institute of Physics and Nuclear Engineering, Bucharest; (b)University Politehnica Bucharest, Bucharest; (c)West University in Timisoara, Timisoara, Romania 26 26 Departamento de Física, Universidad de Buenos Aires, Buenos Aires, Argentina 26 Departamento de Física, Universidad de Buenos Aires, Buenos Aires, Argentina 27 Cavendish Laboratory, University of Cambridge, Cambridge, United Kingdom 27 Cavendish Laboratory, University of Cambridge, Cambridge, United Kingdom 28 Department of Physics, Carleton University, Ottawa ON, Canada 28 Department of Physics, Carleton University, Ottawa ON, Canada 29 CERN, Geneva, Switzerland ATLAS Collaboration Yoosoofmiya 123, K. Yorita 170, R. Yoshida 5, C. Young 143, S. Youssef 21, D. Yu 24, J. Yu 7, J. Yu 112, L. Yuan 32a,ag, A. Yurkewicz 106, B. Zabinski 38, V.G. Zaets 128, R. Zaidan 63, A.M. Zaitsev 128, Z. Zajacova 29, L. Zanello 132a,132b, P. Zarzhitsky 39, A. Zaytsev 107, C. Zeitnitz 174, M. Zeller 175, M. Zeman 125, A. Zemla 38, C. Zendler 20, O. Zenin 128, T. Ženiš 144a, Z. Zenonos 122a,122b, S. Zenz 14, D. Zerwas 115, G. Zevi della Porta 57, Z. Zhan 32d, D. Zhang 32b,ad, H. Zhang 88, J. Zhang 5, X. Zhang 32d, Z. Zhang 115, L. Zhao 108, T. Zhao 138, Z. Zhao 32b, A. Zhemchugov 65, S. Zheng 32a, J. Zhong 118, B. Zhou 87, N. Zhou 163, Y. Zhou 151, C.G. Zhu 32d, H. Zhu 41, J. Zhu 87, Y. Zhu 32b, X. Zhuang 98, V. Zhuravlov 99, D. Zieminska 61, R. Zimmermann 20, S. Zimmermann 20, S. Zimmermann 48, M. Ziolkowski 141, R. Zitoun 4, L. Živkovi´c 34, V.V. Zmouchko 128,∗, G. Zobernig 172, A. Zoccoli 19a,19b, Y. Zolnierowski 4, A. Zsenei 29, M. zur Nedden 15, V. Zutshi 106, L. Zwalinski 29 W.C. Wong 40, G. Wooden 87, B.K. Wosiek 38, J. Wotschack 29, M.J. Woudstra 84, K.W. Wozniak 38, K. Wraight 53, C. Wright 53, M. Wright 53, B. Wrona 73, S.L. Wu 172, X. Wu 49, Y. Wu 32b,ae, E. Wulf 34, R. Wunstorf 42, B.M. Wynne 45, S. Xella 35, M. Xiao 136, S. Xie 48, Y. Xie 32a, C. Xu 32b,af , D. Xu 139, G. Xu 32a, B. Yabsley 150, S. Yacoob 145b, M. Yamada 66, H. Yamaguchi 155, A. Yamamoto 66, K. Yamamoto 64, S. Yamamoto 155, T. Yamamura 155, T. Yamanaka 155, J. Yamaoka 44, T. Yamazaki 155, Y. Yamazaki 67, Z. Yan 21, H. Yang 87, U.K. Yang 82, Y. Yang 61, Y. Yang 32a, Z. Yang 146a,146b, S. Yanush 91, Y. Yasu 66, G.V. Ybeles Smit 130, J. Ye 39, S. Ye 24, M. Yilmaz 3c, R. Yoosoofmiya 123, K. Yorita 170, R. Yoshida 5, C. Young 143, S. Youssef 21, D. Yu 24, J. Yu 7, J. Yu 112, L. Yuan 32a,ag, A. Yurkewicz 106, B. Zabinski 38, V.G. Zaets 128, R. Zaidan 63, A.M. Zaitsev 128, Z. Zajacova 29, L. Zanello 132a,132b, P. Zarzhitsky 39, A. Zaytsev 107, C. Zeitnitz 174, M. , , rsity, Ankara; (b)Department of Physics, Dumlupinar University, Kutahya; (c)Department of Physics, Gazi University, Ankara; (d)Division of Physi logy, Ankara; (e)Turkish Atomic Energy Authority, Ankara, Turkey 29 CERN, Geneva, Switzerland Stepanov Institute of Physics, National Academy of Sciences of Belarus, Minsk, Belarus 1 National Scientiﬁc and Educational Centre for Particle and High Energy 92 Department of Physics, Massachusetts Institute of Technology, Cambridge, MA, United States 92 Department of Physics, Massachusetts Institute of Technology, Cambridge, MA, United States 93 Group of Particle Physics, University of Montreal, Montreal QC, Canada 4 P.N. Lebedev Institute of Physics, Academy of Sciences, Moscow, Russia 95 Institute for Theoretical and Experimental Physics (ITEP), Moscow, Russia 96 Moscow Engineering and Physics Institute (MEPhI), Moscow, Russia 97 Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State U 98 Fakultät für Physik, Ludwig-Maximilians-Universität München, München, Germany 98 Fakultät für Physik, Ludwig-Maximilians-Universität München, München, Germany 99 Max-Planck-Institut für Physik (Werner-Heisenberg-Institut), Mün 99 Max-Planck-Institut für Physik (Werner-Heisenberg-Institu 100 Nagasaki Institute of Applied Science, Nagasaki, Japan 101 Graduate School of Science, Nagoya University, Nagoya, Japan 102 (a)INFN Sezione di Napoli; (b)Dipartimento di Scienze Fisiche, Un 102 (a)INFN Sezione di Napoli; (b)Dipartimento di Scienze Fisiche, Università di Napoli, Napoli, Italy p p p p y 103 Department of Physics and Astronomy, University of New Mexico, Albuquerque, NM, United States 103 Department of Physics and Astronomy, University of New Mexico, Albuquerque, NM, United States 03 Department of Physics and Astronomy, University of New Mexico, Al p f y y y f q q 104 Institute for Mathematics, Astrophysics and Particle Physics, Radboud University Nijmegen/Nikhef, Nijmegen, Netherlands 104 Institute for Mathematics, Astrophysics and Particle Physics, Radboud University Nijmegen/Nikhef, Nij 04 Institute for Mathematics, Astrophysics and Particle Physics, Radboud 105 Nikhef National Institute for Subatomic Physics and University of Amsterdam, Amsterdam, Netherlands 105 Nikhef National Institute for Subatomic Physics and University of Amsterdam, Amsterdam, Netherlands Nikhef National Institute for Subatomic Physics and University of Amste 106 Department of Physics, Northern Illinois University, DeKalb, IL, United States 106 Department of Physics, Northern Illinois University, DeKalb, IL, United States 107 Budker Institute of Nuclear Physics (BINP), Novosibirsk, Russia 107 Budker Institute of Nuclear Physics (BINP), Novosibirsk, Russia 108 Department of Physics, New York University, New York, NY, Unit 108 Department of Physics, New York University, New York, NY, United States 109 Ohio State University, Columbus, OH, United States 110 Faculty of Science, Okayama University, Okayama, Japan y f y y y J p 111 Homer L. Dodge Department of Physics and Astronomy, University of Oklahoma, Norman, OK 111 Homer L. 29 CERN, Geneva, Switzerland Andronikashvili Institute of Physics, Georgian Academy of Sciences, Tbilisi; (b)High Energy P I Physikalisches Institut, Justus-Liebig-Universität Giessen, Giessen, Germ y J g y 53 SUPA – School of Physics and Astronomy, University of Glasgow, Glasgow, United Kingdom y J g y 53 SUPA – School of Physics and Astronomy, University of Glasgow, Glasgow, United Kingdom 3 SUPA – School of Physics and Astronomy, University of Glasgow, Glasgo 54 II Physikalisches Institut, Georg-August-Universität, Göttingen, Germany 54 II Physikalisches Institut, Georg-August-Universität, Göttingen, Germany 5 Laboratoire de Physique Subatomique et de Cosmologie, Université Jose 56 Department of Physics, Hampton University, Hampton, VA, United States 56 Department of Physics, Hampton University, Hampton, VA, United States 57 Laboratory for Particle Physics and Cosmology, Harvard University, Cambridge, MA, United States 57 Laboratory for Particle Physics and Cosmology, Harvard University, Cambridge, MA, United States 7 Laboratory for Particle Physics and Cosmology, Harvard University, Cam 8 (a)Kirchhoff-Institut für Physik, Ruprecht-Karls-Universität Heidelberg technische Informatik, Ruprecht-Karls-Universität Heidelberg, Mannheim, Germany technische Informatik, Ruprecht-Karls-Universität Heidelberg, M technische Informatik, Ruprecht-Karls-Universität Heidelberg, Mannheim, Germany 59 Faculty of Science, Hiroshima University, Hiroshima, Japan 60 Faculty of Applied Information Science, Hiroshima Institute 61 Department of Physics, Indiana University, Bloomington, IN, United States 62 Institut für Astro- und Teilchenphysik, Leopold-Franzens-Universität, Innsbruck, Aust 62 Institut für Astro- und Teilchenphysik, Leopold-F 63 University of Iowa, Iowa City, IA, United States 64 Department of Physics and Astronomy, Iowa State University, Ames, IA, United States 64 Department of Physics and Astronomy, Iowa State University, Ames, IA, United States 64 Department of Physics and Astronomy, Iowa Sta 65 Joint Institute for Nuclear Research, JINR Dubna, Dubna, Russia 65 Joint Institute for Nuclear Research, JINR Dubna, Dubna, Russia J f J 66 KEK, High Energy Accelerator Research Organization, Tsukuba, Japan 66 KEK, High Energy Accelerator Research Organization, Tsukuba, Japan g gy g 67 Graduate School of Science, Kobe University, Kobe, Japan 67 Graduate School of Science, Kobe University, Kobe, Jap 68 Faculty of Science, Kyoto University, Kyoto, Japan 68 Faculty of Science, Kyoto University, Kyoto, Japan y f y y y J p 69 Kyoto University of Education, Kyoto, Japan 69 Kyoto University of Education, Kyoto, Japan y y f y J p 70 Instituto de Física La Plata, Universidad Nacional de La Plata and CONICET, La Plata, Argentina 70 Instituto de Física La Plata, Universidad Nacional de La Plata and CONICET, La Plata, Argentina 71 Physics Department, Lancaster University, Lancaster, Unite 72 (a)INFN Sezione di Lecce; (b)Dipartimento di Fisica, Università del Salento, Lecce, Italy 72 (a)INFN Sezione di Lecce; (b)Dipartimento di Fisica, Università del Salento, Lec 73 Oliver Lodge Laboratory, University of Liverpool, Liverpool, United Kingdom 73 Oliver Lodge Laboratory, University of Liverpool, Liverpool, United Kingdom 74 Department of Physics, Jožef Stefan Institute and University of Ljub 75 School of Physics and Astronomy, Queen Mary University of London, London, United Kingdom 75 School of Physics and Astronomy, Queen Mary University of London, London, United Kingdom 76 Department of Physics, Royal Holloway University of London, Surrey, United Kingdom 77 Department of Physics and Astronomy, University College London, London, United Kingdom 78 Laboratoire de Physique Nucléaire et de Hautes Energies, UPMC and Université Paris-Diderot 79 Fysiska institutionen, Lunds universitet, Lund, Sweden 80 Departamento de Fisica Teorica C-15, Universidad Aut 81 Institut für Physik, Universität Mainz, Mainz, German 82 School of Physics and Astronomy, University of Manchester, Manche f y y y f 83 CPPM, Aix-Marseille Université and CNRS/IN2P3, Marseille, France 83 CPPM, Aix-Marseille Université and CNRS/IN2P3, Marseille, France / 84 Department of Physics, University of Massachusetts, Amherst, MA, United States 84 Department of Physics, University of Massachusetts, Amherst, MA, U 85 Department of Physics, McGill University, Montreal QC, Canada 86 School of Physics, University of Melbourne, Victoria, Australia 86 School of Physics, University of Melbourne, Victoria, Australia 87 Department of Physics, The University of Michigan, Ann Arbor, MI, United States 89 (a)INFN Sezione di Milano; (b)Dipartimento di Fisica, Università di Milano, Milano, Italy 89 (a)INFN Sezione di Milano; (b)Dipartimento di Fisica, Università di Milano, Milano, Italy 90 B.I. 29 CERN, Geneva, Switzerland 30 Enrico Fermi Institute, University of Chicago, Chicago, IL, United States 30 Enrico Fermi Institute, University of Chicago, Chicago, IL, United States 31 (a)Departamento de Fisica, Pontiﬁcia Universidad Católica de Chile, Santiago; (b)Departamento de 32 ( ) (b) 31 (a)Departamento de Fisica, Pontiﬁcia Universidad Católica de Chile, Santiago; (b)Departamento de Física, Universidad Técnica Federico Santa María, Valparaíso, Chile 32 (a)Institute of High Energy Physics, Chinese Academy of Sciences, Beijing; (b)Department of Modern Physics, University of Science and Technology of China, Anhui; (c)Department of 31 (a)Departamento de Fisica, Pontiﬁcia Universidad Católica de Chile, Santiago; (b)Departamento de Física, Universidad Técnica Federico Santa María, Valparaíso, Chile 32 (a)Institute of High Energy Physics Chinese Academy of Sciences Beijing; (b)Department of Modern Physics University of Science and Technology of China Anhui; (c)Department of p ﬁ g p 32 (a)Institute of High Energy Physics, Chinese Academy of Sciences, Beijing; (b)Department of Modern Physics, University of Science Physics, Nanjing University, Jiangsu; (d)High Energy Physics Group, Shandong University, Shandong, China Physics, Nanjing University, Jiangsu; (d)High Energy Physics Group, Shandong University, Shandong, Chi 33 Laboratoire de Physique Corpusculaire, Clermont Université and Université Blaise Pascal 33 Laboratoire de Physique Corpusculaire, Clermont Université and Université Blaise Pascal and CNRS/IN2P3, Aubiere Cedex, 33 Laboratoire de Physique Corpusculaire, Clermont Université and Univ 33 Laboratoire de Physique Corpusculaire, Clermont Université and Université Blaise Pascal and CNRS/IN2P3, Aubiere Cedex, France 34 Nevis Laboratory, Columbia University, Irvington, NY, United States 34 Nevis Laboratory, Columbia University, Irvington, NY, United States 35 Niels Bohr Institute, University of Copenhagen, Kobenhavn, Denmark 35 Niels Bohr Institute, University of Copenhagen, Kobenhavn, Denma 36 (a)INFN Gruppo Collegato di Cosenza; (b)Dipartimento di Fisica, Università della Cal 36 (a)INFN Gruppo Collegato di Cosenza; (b)Dipartimento di Fisica, Univ 36 (a)INFN Gruppo Collegato di Cosenza; (b)Dipartimento di Fisica, Università della Calabria, Arcavata di Rende, Italy 37 37 Faculty of Physics and Applied Computer Science, AGH-University of Science and Technology, Krakow 37 Faculty of Physics and Applied Computer Science, AGH-University of Science and Technology, Krakow, Poland 38 38 The Henryk Niewodniczanski Institute of Nuclear Physics, Polish Academy of Sciences, Krakow, Poland 39 39 Physics Department, Southern Methodist University, Dallas, TX, Unite 39 Physics Department, Southern Methodist University, Dallas, TX, United States 40 Physics Department, University of Texas at Dallas, Richardson, TX, United States 40 Physics Department, University of Texas at Dallas, Richardson, TX, United States 41 DESY, Hamburg and Zeuthen, Germany 42 Institut für Experimentelle Physik IV, Technische Universität Dortmund, Dortmund, Germany 42 Institut für Experimentelle Physik IV, Technische Universität Dortmund, Dortmund, Germany f p y 43 Institut für Kern- und Teilchenphysik, Technical University Dresden, Dresden, Germany 43 Institut für Kern- und Teilchenphysik, Technical University Dresden, Dresden, Germany 43 Institut für Kern- und Teilchenphysik, Technical University Dresden 44 Department of Physics, Duke University, Durham, NC, United States Department of Physics, Duke University, Durham, NC, United States 45 SUPA – School of Physics and Astronomy, University of Edinburgh, Edinburgh, United Kingdom 45 SUPA – School of Physics and Astronomy, University of Edinburgh, Edinburgh, United Kingdom 46 45 SUPA – School of Physics and Astronomy, University of Edinburgh, E 46 Fachhochschule Wiener Neustadt, Johannes Gutenbergstrasse 3 2700 Wiener Neustadt, Austria 46 Fachhochschule Wiener Neustadt, Johannes Gutenbergstrasse 3 2700 Wiener Neustadt, Austria 46 Fachhochschule Wiener Neustadt, Johannes Gutenbe 47 INFN Laboratori Nazionali di Frascati, Frascati, Italy 48 Fakultät für Mathematik und Physik, Albert-Ludwigs-Universität, Freiburg i.Br., Germany 48 Fakultät für Mathematik und Physik, Albert-Ludwigs-Universität, Freiburg i.Br., Germany 48 Fakultät für Mathematik und Physik, Albert-Ludwigs-Universität, 49 Section de Physique, Université de Genève, Geneva, Switzerland 49 Section de Physique, Université de Genève, Geneva, Switzerland y q 50 (a)INFN Sezione di Genova; (b)Dipartimento di Fisica, Università d ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 174 ili Institute of Physics, Georgian Academy of Sciences, Tbilisi; (b)High Energy Physics Institute, Tbilisi State University, Tbilisi, Georgia i J Li bi U i i ä Gi Gi G 51 (a)E. 29 CERN, Geneva, Switzerland Andronikashvili Institute of Physics, Georgian Academy of Sciences, Tbilisi; (b)High Energy Physics Institute, Tbilisi State Universi 51 (a)E. 29 CERN, Geneva, Switzerland Dodge Department of Physics and Astronomy, University of Oklahoma, Norman, OK, United State 111 Homer L. Dodge Department of Physics and Astronomy, Univers f Physics and Astronomy, University of Oklahoma, Norman, OK, United S 112 Department of Physics, Oklahoma State University, Stillwater, OK, United States 112 Department of Physics, Oklahoma State University, Stillwater, OK, United States 113 Palacký University, RCPTM, Olomouc, Czech Republic 114 Center for High Energy Physics, University of Oregon, Eugene, OR, U 114 Center for High Energy Physics, University of Oregon, Eugene, OR, United States 115 LAL, Univ. Paris-Sud and CNRS/IN2P3, Orsay, France 116 Graduate School of Science, Osaka University, Osaka, Japan 117 Department of Physics, University of Oslo, Oslo, Norway 118 Department of Physics, Oxford University, Oxford, United Kingdom 118 Department of Physics, Oxford University, Oxford, United Kingdom 119 (a)INFN Sezione di Pavia; (b)Dipartimento di Fisica Nucleare e Teori p 120 Department of Physics, University of Pennsylvania, Philadelphia, PA, United States 120 Department of Physics, University of Pennsylvania, Philadelphia, PA, United States 120 Department of Physics, University of Pennsylvania, Philadelph 121 Petersburg Nuclear Physics Institute, Gatchina, Russia 21 Petersburg Nuclear Physics Institute, Gatchina, Russia 122 (a)INFN Sezione di Pisa; (b)Dipartimento di Fisica E. Fermi, Universi 123 Department of Physics and Astronomy, University of Pittsburgh, Pittsburgh, PA, United States ( ) (b 123 Department of Physics and Astronomy, University of Pittsburgh, Pittsburgh, PA, United State 124 (a)Laboratorio de Instrumentacao e Fisica Experimental de Particulas – LIP, Lisboa, Portugal; (b)Departamento de Fisica Teo d i 124 (a)Laboratorio de Instrumentacao e Fisica Experimental de Particulas – LIP, Lisboa, Portugal; (b)Departamento de Fisica Teorica y del Cosm Granada, Spain 125 Institute of Physics, Academy of Sciences of the Czech Republic, Praha, Czech Republic 125 Institute of Physics, Academy of Sciences of the Czech Republic, Praha, Czech Republic f y y f f p p 126 Faculty of Mathematics and Physics, Charles University in Prague, Praha, Czech Republic 126 Faculty of Mathematics and Physics, Charles University in Prague, Praha, Czech Republic 127 Czech Technical University in Prague, Praha, Czech Republic 175 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 2390, Marrakech 40000; Faculté des Sciences, Université Mohamed Premier and LPTPM, Oujda; (e)Faculté des Sciences, Université Mohammed V, Rabat, Morocco Oujda; (e)Faculté des Sciences, Université Mohammed V, Rabat, Morocc LPTPM, Oujda; (e)Faculté des Sciences, Université Mohammed V, Rabat, Morocco j ; , , , /IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Saclay (Commissariat a l’Energie Atomique), Gif-sur-Yvette, France 6 DSM/IRFU (Institut de Recherches sur les Lois Fondamentales de l’Unive 137 Santa Cruz Institute for Particle Physics, University of California Santa Cruz, Santa Cruz, CA, United States 138 7 Santa Cruz Institute for Particle Physics, University of California Santa 137 Santa Cruz Institute for Particle Physics, University of California Santa Cruz, Santa Cruz, CA, United States 138 Department of Physics, University of Washington, Seattle, WA, United States 138 Department of Physics, University of Washington, Seattle, WA, United States 138 Department of Physics, University of Washington, Seattle, WA, Unite p f y y f g 139 Department of Physics and Astronomy, University of Sheﬃeld, Sheﬃeld, United Kingdom 139 Department of Physics and Astronomy, University of Sheﬃeld, Sheﬃeld, United Kingdo 139 Department of Physics and Astronomy, University of Sheﬃel 140 Department of Physics, Shinshu University, Nagano, Japan 140 Department of Physics, Shinshu University, Nagano, Japan 141 Fachbereich Physik, Universität Siegen, Siegen, Germany y g g y 142 Department of Physics, Simon Fraser University, Burnaby BC, Canada 142 Department of Physics, Simon Fraser University, Burnaby BC, Canada 143 SLAC National Accelerator Laboratory, Stanford, CA, United States 143 SLAC National Accelerator Laboratory, Stanford, CA, United States y f 144 (a)Faculty of Mathematics, Physics & Informatics, Comenius University, Bratislava; (b)Department of Subnuclear Physics, Institute of Experimental Physics of the Slovak Academy of Sciences Kosice Slovak Republic 144 (a)Faculty of Mathematics, Physics & Informatics, Comenius Univers 144 (a)Faculty of Mathematics, Physics & Informatics, Comenius University, Bratislava; (b)Department of Subnuclear Physics, Institute of Experi Sciences, Kosice, Slovak Republic 144 (a)Faculty of Mathematics, Physics & Informatics, Comenius University, Bratislava; (b)Department of Subnuclear Physics, Institute of Experimental Physics of the Slovak Academy of Sciences, Kosice, Slovak Republic Sciences, Kosice, Slovak Republic Sciences, Kosice, Slovak Republic p t of Physics, University of Johannesburg, Johannesburg; (b)School of Physics, University of the Witwatersrand, Johannesburg, South Africa t of Physics Stockholm University; (b)The Oskar Klein Centre Stockholm Sweden p 145 (a)Department of Physics, University of Johannesburg, Johannesburg; (b)School of Physics, University of the Witwatersrand, Johannesburg, S 146 ( ) (b) p 145 (a)Department of Physics, University of Johannesburg, Johannesburg; (b)School of Physics, University of the Witwatersrand, Johannesburg, South Africa 146 (a)Department of Physics, Stockholm University; (b)The Oskar Klein Centre, Stockholm, Sweden 145 (a)Department of Physics, University of Johannesburg, Johannesburg; (b)School of Physics, University of 145 (a)Department of Physics, University of Johannesburg, Johannesburg; (b)School of Physics, University of the Witwatersrand, Johannesburg, South Africa 146 (a)Department of Physics, Stockholm University; (b)The Oskar Klein Centre, Stockholm, Sweden p f y , y f J g, J g; f y , y f , J g, f 146 (a)Department of Physics, Stockholm University; (b)The Oskar Klein Centre, Stockholm, Sweden 6 (a)Department of Physics, Stockholm University; (b)The Oskar Klein Cen 147 Physics Department, Royal Institute of Technology, Stockholm, Sweden 147 Physics Department, Royal Institute of Technology, Stockholm, Sweden 48 148 Department of Physics and Astronomy, Stony Brook University, Stony Brook, NY, United States 148 Department of Physics and Astronomy, Stony Brook University, Stony Brook, NY, United State 148 Department of Physics and Astronomy, Stony Brook University, Ston p f y y y y y 149 Department of Physics and Astronomy, University of Sussex, Brighton, United K 149 Department of Physics and Astronomy, University of Sussex 149 Department of Physics and Astronomy, University of Sussex, Brighton, United Kingdo 150 School of Physics, University of Sydney, Sydney, Australia 151 Institute of Physics, Academia Sinica, Taipei, Taiwan 151 Institute of Physics, Academia Sinica, Taipei, Taiwan 152 Department of Physics, Technion: Israel Inst. of Technology, Haifa, Isr 152 Department of Physics, Technion: Israel Inst. ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 128 State Research Center Institute for High Energy Physics, Protvino, Russia 128 State Research Center Institute for High Energy Physics, Protvino, Russia 129 Particle Physics Department, Rutherford Appleton Laboratory, D 130 Physics Department, University of Regina, Regina SK, Canada 130 Physics Department, University of Regina, Regina SK, Canada 131 Ritsumeikan University, Kusatsu, Shiga, Japan y g J p 132 (a)INFN Sezione di Roma I; (b)Dipartimento di Fisica, Universit e di Roma I; (b)Dipartimento di Fisica, Università La Sapienza, Roma, Ital (b) p p y 133 (a)INFN Sezione di Roma Tor Vergata; (b)Dipartimento di Fisica, Università di Roma Tor Vergata, Roma, Italy 134 ( ) (b) 133 (a)INFN Sezione di Roma Tor Vergata; (b)Dipartimento di Fisica, Università di Roma Tor Verga e di Roma Tor Vergata; (b)Dipartimento di Fisica, Università di Roma Tor 134 (a)INFN Sezione di Roma Tre; (b)Dipartimento di Fisica, Università Roma Tre, Roma, Italy 134 (a)INFN Sezione di Roma Tre; (b)Dipartimento di Fisica, Università Roma Tre, Roma, Italy ; p , , , y 135 (a)Faculté des Sciences Ain Chock, Réseau Universitaire de Physique des Hautes Energies – Université Hassan II, Casablanca; (b)Centre National de l’Energie des Sciences Techniques Nucleaires, Rabat; (c)Université Cadi Ayyad, Faculté des sciences Semlalia Département de Physique, B.P. 2390, Marrakech 40000; (d)Faculté des Sciences, Université Mohamed Premier and LPTPM, Oujda; (e)Faculté des Sciences, Université Mohammed V, Rabat, Morocco 5 (a)Faculté des Sciences Ain Chock, Réseau Universitaire de Physique des Nucleaires, Rabat; (c)Université Cadi Ayyad, Faculté des sciences Semlalia Département de Physique, B.P. 2390, Marrakech 40000; (d)Faculté de LPTPM Oujda; (e)Faculté des Sciences Université Mohammed V Rabat Morocco Nucleaires, Rabat; (c)Université Cadi Ayyad, Faculté des sciences Semlalia Département de Physique, B.P. 2390 ucleaires, Rabat; (c)Université Cadi Ayyad, Faculté des sciences Semlalia D Nucleaires, Rabat; Université Cadi Ayyad, Faculté des sciences Semlalia Département de Physique, B.P. a Also at Laboratorio de Instrumentacao e Fisica Experimental de Particulas – LIP, Lisboa, Portugal. a Also at Laboratorio de Instrumentacao e Fisica Experimental de Particulas – LIP, Lisboa, Portugal. b Also at Faculdade de Ciencias and CFNUL, Universidade de Lisboa, Lisboa, Portugal. c Also at Particle Physics Department, Rutherford Appleton Laborat c Also at Particle Physics Department, Rutherford Appleton Laboratory, Didcot, United d Also at TRIUMF, Vancouver BC, Canada d Also at TRIUMF, Vancouver BC, Canada. e Also at Department of Physics, California State University, Fresno, CA, United States. e Also at Department of Physics, California State University, Fres f Also at Fermilab, Batavia, IL, United States. f Also at Fermilab, Batavia, IL, United States. g Also at Department of Physics, University of Coimbra, Coimbra, Portugal. g Also at Department of Physics, University of Coimbra, Coimbra, Portugal. g Also at Department of Physics, University of Coimbra, C h Also at Università di Napoli Parthenope, Napoli, Italy. h Also at Università di Napoli Parthenope, Napoli, Italy. i Also at Institute of Particle Physics (IPP), Canada. i Also at Institute of Particle Physics (IPP), Canada. j Also at Department of Physics, Middle East Technical Universi k Also at Louisiana Tech University, Ruston, LA, United States. k Also at Louisiana Tech University, Ruston, LA, United States. l Also at Department of Physics and Astronomy, University College London, London, United Kingdom. l Also at Department of Physics and Astronomy, University Col l Also at Department of Physics and Astronomy, University Co roup of Particle Physics, University of Montreal, Montreal QC, Cana n Also at Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan. n Also at Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan. o Also at Institut für Experimentalphysik, Universität Hamburg, Hamburg, Germany. o Also at Institut für Experimentalphysik, Universität Hamburg, Hamburg, Germa o Also at Institut für Experimentalphysik, Universität Hamburg, H p Also at Manhattan College, New York, NY, United States. q Also at CPPM, Aix-Marseille Université and CNRS/IN2P3, Marseille, France. q Also at CPPM, Aix-Marseille Université and CNRS/IN2P3, Marseille, France. r Also at School of Physics and Engineering, Sun Yat-sen University, Guanzhou, China. r Also at School of Physics and Engineering, Sun Yat-sen University, Guanzhou, China t Also at High Energy Physics Group, Shandong University, Shandong, China. t Also at High Energy Physics Group, Shandong University, Shandong, China. u Also at Section de Physique, Université de Genève, Geneva, Switzerland. u Also at Section de Physique, Université de Genève, Geneva, Switzerland. Sciences, Kosice, Slovak Republic of Technology, Haifa, Israel p f y f gy f 153 Raymond and Beverly Sackler School of Physics and Astronomy, Tel Aviv University, Tel Aviv, Israel 154 3 Raymond and Beverly Sackler School of Physics and Astronomy, Tel Avi 153 Raymond and Beverly Sackler School of Physics and Astronomy, Tel Aviv University, Tel Aviv, Israel 4 Department of Physics, Aristotle University of Thessaloniki, Thessalonik p f y , y f , , 155 International Center for Elementary Particle Physics and Department of Physics, The University of Tokyo, Tokyo, Japan p f y y f 155 International Center for Elementary Particle Physics and Department of Physics, The University p f y y f 155 International Center for Elementary Particle Physics and Department of Physics, T 5 International Center for Elementary Particle Physics and Department o 156 Graduate School of Science and Technology, Tokyo Metropolitan University, Tokyo, Japan 156 Graduate School of Science and Technology, Tokyo Metropolitan University, Tokyo, Japan 157 Department of Physics, Tokyo Institute of Technology, Tokyo, Japan 158 Department of Physics, University of Toronto, Toronto ON, Canada 158 Department of Physics, University of Toronto, Toronto ON, Canada p f y y f 159 (a)TRIUMF, Vancouver BC; (b)Department of Physics and Astronomy, York University, Toronto ON, Canada 160 159 (a)TRIUMF, Vancouver BC; (b)Department of Physics and Astronomy, York University, Toronto ON, Canada 159 (a)TRIUMF, Vancouver BC; (b)Department of Physics and Astronomy 0 Institute of Pure and Applied Sciences, University of Tsukuba, 1-1-1 Ten 161 Science and Technology Center, Tufts University, Medford, MA, United States 161 Science and Technology Center, Tufts University, Medford, MA, United States 162 Centro de Investigaciones, Universidad Antonio Narino, Bogota, Colombia 162 Centro de Investigaciones, Universidad Antonio Narino, Bogota, Colombia 3 Department of Physics and Astronomy, University of California Irvine, I p f y y, y f f , , , 164 (a)INFN Gruppo Collegato di Udine; (b)ICTP, Trieste; (c)Dipartimento di Chimica, Fisica e Ambiente, Università di Udine, Udine, Italy 6 164 (a)INFN Gruppo Collegato di Udine; (b)ICTP, Trieste; (c)Dipartimento di Chimica, Fisica e Ambiente, Un a)INFN Gruppo Collegato di Udine; (b)ICTP, Trieste; (c)Dipartimento di C 4 (a)INFN Gruppo Collegato di Udine; (b)ICTP, Trieste; (c)Dipartimento di 164 (a)INFN Gruppo Collegato di Udine; (b)ICTP, Trieste; (c)Dipartimento di Chimica, Fisica e Ambi 165 Department of Physics, University of Illinois, Urbana, IL, United States 165 Department of Physics, University of Illinois, Urbana, IL, United States 6 Department of Physics and Astronomy, University of Uppsala, Uppsala, y, y f pp , pp , ar (IFIC) and Departamento de Física Atómica, Molecular y Nuclear and Departamento de Ingenierá Electrónica and Instituto de Microelectrónic y of Valencia and CSIC Valencia Spain 167 Instituto de Física Corpuscular (IFIC) and Departamento de Física Atómica, Molecular y Nuclear and Departamento de Ingenierá Electr 7 Instituto de Física Corpuscular (IFIC) and Departamento de Física Ató 167 Instituto de Física Corpuscular (IFIC) and Departamento de Física Atómica, Molec rcelona (IMB-CNM), University of Valencia and CSIC, Valencia, Spain Barcelona (IMB-CNM), University of Valencia and CSIC, Valencia, Spai 8 Department of Physics, University of British Columbia, Vancouver BC, C 168 Department of Physics, University of British Columbia, Vancouver 169 Department of Physics and Astronomy, University of Victoria, Victoria BC, Canada 169 Department of Physics and Astronomy, University of Victoria, Victoria BC, Canada 170 Waseda University, Tokyo, Japan y y J p 171 Department of Particle Physics, The Weizmann Institute of Science, Rehovot, Israel y y J p 171 Department of Particle Physics, The Weizmann Institute of Science, Rehovot, Israe 172 Department of Physics, University of Wisconsin, Madison, WI, United States 172 Department of Physics, University of Wisconsin, Madison, WI, United States 173 Fakultät für Physik und Astronomie, Julius-Maximilians-Universität, Würzburg, Ger 174 174 Fachbereich C Physik, Bergische Universität Wuppertal, Wuppertal, Germany 174 Fachbereich C Physik, Bergische Universität Wuppertal, Wuppertal, Germany 175 Department of Physics, Yale University, New Haven, CT, United States 175 Department of Physics, Yale University, New Haven, CT, United States 176 Yerevan Physics Institute, Yerevan, Armenia 176 Yerevan Physics Institute, Yerevan, Armenia y 177 Domaine scientiﬁque de la Doua, Centre de Calcul CNRS/IN2P3, Villeurbanne Cedex, France a Also at Laboratorio de Instrumentacao e Fisica Experimental de Particulas – LIP, Lisboa, Portugal. a Also at Laboratorio de Instrumentacao e Fisica Experimental de Particulas – LIP, Lisboa, Portugal. so at Departamento de Fisica, Universidade de Minho, Braga, Portug g Also at Laboratoire de Physique Nucléaire et de Hautes Energies, UPMC and Université Paris-Diderot and CNRS/IN2P3, Paris, France ∗Deceased ae Also at Department of Physics, The University of Michigan, Ann Arbor, MI, United States. af Also at DSM/IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Saclay (Commissariat a l’Energie Atomique), Gif-sur-Yvette, France. ag Also at Laboratoire de Physique Nucléaire et de Hautes Energies, UPMC and Université Paris-Diderot and CNRS/IN2P3, Paris, France. ∗Deceased so at DSM/IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Saclay (Commissariat a l’Energie Atomique), G so at Laboratoire de Physique Nucléaire et de Hautes Energies, UPMC and Université Paris-Diderot and CNRS/IN2P3, Paris, France. d f Also at DSM/IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Saclay (Commissariat a l’Energie Atomique g Al L b i d Ph i N lé i d H E i UPMC d U i i é P i Did d CNRS/IN2P3 P i F ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 ATLAS Collaboration / Physics Letters B 709 (2012) 158–176 176 w Also at Department of Physics and Astronomy, University of South Carolina, Columbia, SC, United States. w Also at Department of Physics and Astronomy, University of South Carolina, Columbia, SC, United States lso at KFKI Research Institute for Particle and Nuclear Physics, Bud y Also at California Institute of Technology, Pasadena, CA, United States. y Also at California Institute of Technology, Pasadena, CA, United States. z Also at Institute of Physics, Jagiellonian University, Krakow, Poland. Also at Institute of High Energy Physics, Chinese Academy of Scie b Also at Department of Physics and Astronomy, University of Sheﬃeld, Sheﬃeld, United Kingdom. ab Also at Department of Physics and Astronomy, University of Sheﬃeld, Sheﬃeld, United Kingdom. ac Also at Department of Physics, Oxford University, Oxford, United Kingdom. ac Also at Department of Physics, Oxford University, Oxford, United Kingdom. ad Also at Institute of Physics, Academia Sinica, Taipei, Taiwan. e Also at Department of Physics, The University of Michigan, Ann Arbor, MI, United States. ae Also at Department of Physics, The University of Michigan, Ann Arbor, MI, United States ae Also at Department of Physics, The University of Michigan, An af Also at DSM/IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Sacla
https://openalex.org/W3043194582	https://europepmc.org/articles/pmc7366473?pdf=render	English	null	Effect of oxygenation modalities among patients with postoperative respiratory failure: a pairwise and network meta-analysis of randomized controlled trials	Journal of intensive care	2,020	cc-by	8,551	RESEARCH Open Access Zayed et al. Journal of Intensive Care (2020) 8:51 https://doi.org/10.1186/s40560-020-00468-x Zayed et al. Journal of Intensive Care (2020) 8:51 https://doi.org/10.1186/s40560-020-00468-x Effect of oxygenation modalities among patients with postoperative respiratory failure: a pairwise and network meta- analysis of randomized controlled trials Yazan Zayed1* , Babikir Kheiri2, Mahmoud Barbarawi1, Laith Rashdan1, Inderdeep Gakhal1, Esra’a Ismail3, Josiane Kerbage4, Fatima Rizk5, Saadia Shafi1, Areeg Bala1, Shima Sidahmed1, Ghassan Bachuwa1 and Elfateh Seedahmed6 Abstract Background: Postoperative respiratory failure is associated with increased perioperative complications. Our aim is to compare outcomes between non-invasive ventilation (NIV), high-flow nasal cannula (HFNC), and standard oxygen in patients at high-risk for or with established postoperative respiratory failure. Methods: Electronic databases including PubMed, Embase, and the Cochrane Library were reviewed from inception to September 2019. We included only randomized controlled trials (RCTs) that compared NIV, HFNC, and standard oxygen in patients at high risk for or with established postoperative respiratory failure. We performed a Bayesian network meta-analysis to calculate the odds ratio (OR) and Bayesian 95% credible intervals (CrIs). Results: Nine RCTs representing 1865 patients were included (the mean age was 61.6 ± 10.2 and 64.4% were males). In comparison with standard oxygen, NIV was associated with a significant reduction in intubation rate (OR 0.23; 95% Cr.I. 0.10–0.46), mortality (OR 0.45; 95% Cr.I. 0.27–0.71), and intensive care unit (ICU)-acquired infections (OR 0.43, 95% Cr.I. 0.25–0.70). Compared to standard oxygen, HFNC was associated with a significant reduction in intubation rate (OR 0.28, 95% Cr.I. 0.08–0.76) and ICU-acquired infections (OR 0.41; 95% Cr.I. 0.20–0.80), but not mortality (OR 0.58; 95% Cr.I. 0.26–1.22). There were no significant differences between HFNC and NIV regarding different outcomes. In a subgroup analysis, we observed a mortality benefit with NIV over standard oxygen in patients undergoing cardiothoracic surgeries but not in abdominal surgeries. Furthermore, in comparison with standard oxygen, NIV and HFNC were associated with lower intubation rates following cardiothoracic surgeries while only NIV reduced the intubation rates following abdominal surgeries. Conclusions: Among patients with post-operative respiratory failure, HFNC and NIV were associated with significantly reduced rates of intubation and ICU-acquired infections compared with standard oxygen. Moreover, NIV was associated with reduced mortality in comparison with standard oxygen. ostoperative respiratory failure, High-flow nasal cannula, Non-invasive ventilation, Standard oxygen, Keywords: Postoperative respiratory failure, High-flow nasal cannula, Non-invasive ventilation, Standard oxygen, Meta-analysis * Correspondence: yzayed1@hurleymc.com; yz.alzayed@yahoo.com 1Department of Internal Medicine, Hurley Medical Center/Michigan State University, One Hurley Plaza, Suite 212, Flint, MI 48503, USA Full list of author information is available at the end of the article Study design and study selection Our study is a meta-analysis and systematic review per- formed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRIS MA-P) 2015 Statement [16]. Two reviewers (M.B., I.G) independently and separately performed a literature search utilizing electronic databases including PubMed, Cochrane Library, and Embase from inception to Sep- tember 2019 without language restrictions. Articles were first screened by titles and abstracts before exclusion. © 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 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 11 Page 2 of 11 Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care (2020) 8:51 Inclusion criteria and study selection y Only RCTs were eligible for inclusion in our analysis. We included studies that compared different oxygen- ation strategies in patients who developed or were deemed at high risk for developing post-operative hyp- oxemic respiratory failure. Patients at risk were defined to have intermediate to high risk for development of post-operative pulmonary complications according to ei- ther Assess Respiratory Risk in Surgical Patients in Cata- lonia (ARISCAT) score of ≥26 points [4]. Patients who failed spontaneous breathing trial and those who passed spontaneous breathing trials but had risk factors for failed extubation such as cardiac dysfunction, obesity (BMI > 30), or failure of previous extubation were also considered high risk. Post-operative hypoxemic respira- tory failure was defined as the development of tachypnea with a respiratory rate of ≥25 respirations per minute, intense work of breathing with the use of accessory mus- cles, hypoxemia (oxygen saturation ≤92% or partial ar- terial oxygen pressure to fraction of inspired oxygen ≤ 300) in the immediate post-operative period or within 7 days post-operatively. We excluded studies that investi- gated prophylactic use of NIV and HFNC as a routine therapy in the post-operative period. ( ) [ ] In nonsurgical patients, oxygenation modalities for hypoxemic respiratory failure are varied. Non-invasive ventilation (NIV) has shown promising results for redu- cing intubation rates among patients with cardiogenic pulmonary edema and chronic obstructive pulmonary disease exacerbations [10–12]. High-flow nasal cannula (HFNC) is a new oxygenation strategy that delivers oxy- gen at high concentrations and a high flow rate and has been increasingly utilized due to its ease of application, tolerance, and potential clinical benefits [13, 14]. Never- theless, the current European Respiratory Society/ American Thoracic Society (ERS/ATS) guidelines have conditional recommendations regarding the use of NIV in postoperative respiratory failure [15]. In addition, few randomized controlled trials (RCTs) have been con- ducted to evaluate the efficacy of HFNC vs NIV in post- operative patients. Therefore, we conducted a meta- analysis to compare the efficacy and safety of HFNC, NIV, and standard oxygen therapy in the treatment of patients who developed or were considered high risk for post-operative respiratory failure. Data were extracted into a predesigned table inde- pendently and separately by two reviewers (L.R and S.S.). Any discrepancies were solved by consensus with a third reviewer (Y.Z.). Quality assessment Cochrane Collaboration’s tool for assessing risk of bias in randomized controlled trials was used for quality as- sessment for the included RCTs [17]. Each of the in- cluded RCTs was assessed for random sequence generation, allocation concealment, blindness of partici- pants and health-care personnel, blindness of outcome assessment, incomplete outcome data, selective report- ing, and other biases if any were present. Methodology Study design and study selection Introduction Full texts of eligible articles were reviewed for final in- clusion or exclusion. Mesh terms used were as follows: “postoperative respiratory failure”, “respiratory failure”, “postoperative”, “hypoxemic”, “hypoxic”, “non-invasive ventilation”, “NIV”, “high-flow nasal cannula”, “HFNC”, “high-flow nasal therapy”, “HFNT”, “high-flow nasal oxy- gen”, “HFNO”, “oxygen”, “facemask”, and “ventilation”. References of relevant articles were also reviewed for possible inclusion. A third reviewer (YZ) resolved any discrepancies. Full texts of eligible articles were reviewed for final in- clusion or exclusion. Mesh terms used were as follows: “postoperative respiratory failure”, “respiratory failure”, “postoperative”, “hypoxemic”, “hypoxic”, “non-invasive ventilation”, “NIV”, “high-flow nasal cannula”, “HFNC”, “high-flow nasal therapy”, “HFNT”, “high-flow nasal oxy- gen”, “HFNO”, “oxygen”, “facemask”, and “ventilation”. References of relevant articles were also reviewed for possible inclusion. A third reviewer (YZ) resolved any discrepancies. Postoperative respiratory failure is associated with in- creased perioperative complications such as reintuba- tion, invasive mechanical ventilation, and healthcare- associated infections, which can lead to increases in mortality, intensive care unit (ICU) and hospital length of stay, delays in hospital discharges, and higher health- care resource utilization [1–4]. Several post-operative pulmonary complications may result in post-operative hypoxemic respiratory failure, including pneumonia, atelectasis, bronchospasm, pneumothorax, and pleural effusion. The incidence of these complications is variable and ranges between 5 and 40% according to the type of surgery, as well as other risk factors including anesthetic technique, dur- ation of surgery, and severity of illness [5–9]. Cardiac surgery has the highest rate of post-operative respiratory complications (up to 40%), followed by thoracic surgery (30%), while abdominal and vascular surgeries have a low incidence of post-operative pulmonary complica- tions (6–7%) [5–7]. Outcomes Our main outcome was the intubation rate following surgery. Secondary outcomes included mortality at the Zayed et al. Journal of Intensive Care (2020) 8:51 Page 3 of 11 Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care longest follow-up period provided by each study and ICU-acquired infections. In an exploratory analysis, we performed a meta- regression analysis to explain any significant heterogen- eity (> 25%) for NIV vs standard oxygen therapy direct meta-analysis. Moderators included study-level covari- ates: age, gender, body mass index, Simplified Acute Physiology Score (SAPS) II, respiratory rate, PaO2/FiO2 ratio, and partial arterial pressure of carbon dioxide (PaCO2). All data were analyzed using RevMan v5.3 Windows, Comprehensive Meta-Analysis software v3, NetMetaXL v1.6.1, and WinBUGS v1.4.3. Statistical analysis y An informative prior Bayesian framework for the net- work meta-analysis was performed using the Markov Chain Monte Carlo simulation to derive the posterior distribution of the parameter estimates. We used a beta distribution of (0, 2) for binominal likelihood. We used the Brooks-Gelman-Rubin method to assess for conver- gence. A consistency model which contains treatment as a fixed effect and trial as a random effect was used. Re- sults were reported as odds ratios (ORs) and Bayesian 95% credible intervals (Cr.Is). Inconsistency was assessed using the deviance residuals and deviance information criteria statistics. Sensitivity analysis was performed by including only trials that included patients who had de- veloped respiratory failure. Furthermore, subgroup ana- lysis according to the type of surgery (cardiothoracic or abdominal) was performed. In addition, to show the val- idity of our results, we performed a direct pairwise meta-analysis for comparisons that have three or more studies comparing directly the two interventions. Discussion In this first network meta-analysis comparing various oxygenation strategies in patients at risk for hypoxemic respiratory failure or established respiratory failure within 7 days of surgery, we have found that NIV and HFNC were associated with a significant reduction in in- tubation rates and ICU-acquired infections when com- pared to standard oxygen therapy. However, when compared to standard oxygen therapy, only NIV was found to have a mortality benefit in this patient popula- tion. We found HFNC and NIV to have no significant differences in the primary or secondary outcomes. Fur- thermore, in a subgroup analysis, patients undergoing cardiothoracic surgery had a significantly lower rate of intubation when treated with HFNC or NIV in compari- son with standard oxygen therapy, but mortality was sig- nificantly lower in patients treated with NIV in comparison with standard oxygen therapy. Additionally, in patients with abdominal surgeries, only NIV was asso- ciated with a significant reduction in intubation rates compared to standard oxygen, but there was no signifi- cant difference in mortality between competing interventions. g Sensitivity analysis was performed by including only pa- tients who developed acute hypoxemic respiratory failure (but not patients at increased risk), which showed similar results. In a subgroup analysis for patients undergoing car- diothoracic surgery, both NIV and HFNC were associated with a similar reduction in intubation rates compared with standard oxygen therapy (NIV vs standard oxygen (OR 0.08; 95% Cr.I. 0.03–0.19) and HFN vs standard oxygen (OR 0.08; 95% Cr.I. 0.03–0.21)) (Fig. 3a). However, in pa- tients undergoing abdominal surgery, NIV (but not HFNC) was associated with significantly reduced intub- ation rates compared with standard therapy (NIV vs standard oxygen (OR 0.51; 95% Cr.I. 0.26–0.87)) (Fig. 3b). In an exploratory meta-regression analysis, we found that higher PaCO2 was associated with lower risk for in- tubation when NIV was compared to standard oxygen therapy (P < 0.05) (Supplementary Figure 3). Summary of the included studies After review of 1369 articles, 9 studies were included in the final analysis representing 1865 patients [18–26]. Figure 1 illustrates the search process. The mean age was 61.6 ± 10.2, and 64.4% were males. Four RCTs in- cluded patients undergoing cardiac and/or lung surgeries [19, 20, 24, 26], 3 RCTs involved patients undergoing ab- dominal surgeries [21–23], and 2 RCTs included patients following organ transplantation [18, 25]. Two trials in- cluded patients considered at high risk of post-operative pulmonary complications and respiratory failure [23, 26] and one trial included patients at risk for respiratory Fig. 1 Flow chart of literature search and study selection Fig. 1 Flow chart of literature search and study selection Page 4 of 11 Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care (2020) 8:51 0.45; 95% Cr.I. 0.27–0.71). Additionally, there was no significant difference between NIV and HFNC (OR 0.78; 95% Cr.I. 0.41–1.50) or HFNC and standard oxygen (OR 0.58; 95% Cr.I. 0.26–1.22) as shown in Fig. 4. failure or patients with established respiratory failure [20] while six trials included patients who developed re- spiratory failure in the immediate post-operative period or up to 7 days postoperatively [18, 19, 21, 22, 24, 25]. Two trials compared HFNC vs NIV [18, 20], five trials compared NIV vs standard oxygen therapy [19, 21, 22, 24, 25], and 2 trials compared HFNC vs standard oxygen therapy [23, 26]. Table 1 explains the characteristics of the included trials, and Supplementary Figure 1 illus- trates the network geometry. NIV was the most com- monly used treatment (41.2% of patients), HFNC was used in 31.6% of cases, and 27.2% of patients were treated with standard oxygen therapy. Table 2 explains the baseline and demographic characteristics of included patients. In a subgroup analysis based on the type of surgery (cardiothoracic or abdominal), mortality benefit of NIV was limited to those undergoing cardiothoracic surgery compared with standard oxygen therapy (OR 0.31; 95% Cr.I. 0.13–0.70), unlike those undergoing abdominal sur- geries (OR 0.56; 95% Cr.I. 0.27–1.08) (Fig. 5). Direct pairwise meta-analysis results We have performed direct pairwise meta-analysis com- paring NIV versus standard oxygen which showed con- sistent results of the network meta-analysis (Supplementary Figure 4). However, we did not perform the direct meta-analysis for HFNC vs NIV or HFNC vs standard oxygen because studies that compared directly between these interventions were one or two studies. Supplementary Figures 5 and 6 show the results of these individual studies for different outcomes. ICU-acquired infections HFNC and NIV were associated with a decreased risk for ICU-acquired infections in comparison with standard oxygen therapy (OR 0.41; 95% Cr.I. 0.20–0.80) and (OR 0.43; 95% Cr.I. 0.25–0.70), respectively. No significant difference was found between HFNC and NIV (Fig. 6). Included studies were noted to have inevitable per- formance bias as blinding of participants and personnel was difficult given the nature of the intervention. De- tailed quality assessment was not performed for one study as we only found the abstract with no full article explaining the methods. Supplementary Figure 2 shows the risk of bias in each included RCT based on the au- thors’ judgment. Rate of intubation NIV and HFNC were associated with significant reduc- tions in intubation rates when compared to standard oxygen therapy (OR 0.23; 95% Cr.I. 0.10–0.46) and (OR 0.28; 95% Cr.I. 0.08–0.76), respectively. However, there was no significant difference between HFNC and NIV with regard to the intubation rates (OR 0.82; 95% Cr.I. 0.30–2.50), Fig. 2. Mortality Patients older than 19 who had undergone laparoscopic or non- laparoscopic elective or nonelective ab- dominal surgery under general anesthesia that were diagnosed with ARF within 7 days of surgical procedure defined as persistence of more than 30 min of hypoxemia Patients who had undergone cardiothoracic surgery who developed ARF (failure of SBT or successful SBT but failed extubation) or were deemed at risk for respiratory failure post- extubation due to preexisting risk factors HFNC: initial rate of 50 L/min with initial FiO2 50% adjusted to maintain SpO2 92% or more BiPAP: full facemask connected to ventilator with adjustments made to PEEP and FiO2 to maintain SpO2 of 92% or more 3 days NPPV: BiPAP via facemask. FiO2 adjusted to maintain SpO2 of around 92% SO: standard medical care and oxygen therapy as needed Length of hospital stay CPAP: treated with FiO2 of 0.5 plus CPAP of 7.5. After 6 h, patients underwent 1-h screening test breathing O2 through a venture mask at an FiO2 of 0.3. Patients returned to assigned treatment if PaO2/FiO2 ratio was 300 or less, and treatment was interrupted if the ratio was higher than 300 SO: 8 to 10 L/min oxygen. Length of hospital stay Patients who after cardiac surgery developed ARF after initial extubation who were hemodynamically stable with no evidence of bleeding Post-op elective abdominal surgery under GA if surgery required laparotomy and time of viscera exposure longer than 90 min. Patients were extubated after surgery, and if they developed a PaO2/FiO2 of 300 less, they were included in study. Patients with AHRI following lung resection if they met at least three of the following criteria: dyspnea at rest, active contraction of accessory respiratory muscles, PaO2/FiO2 less than 200, chest radiographic abnormalities NPPV: cushion bridge nasal mask with BiPAP. PS was increased to achieve exhaled TV of 8–10 mL/kg and RR of less than 25 breaths/min. FiO2 was adjusted to obtain SpO2 above 90% SO: O2 supplementation to achieve SaO2 above 90% 120 days NIV: ventilator connected to full-face mask with titration of PS to obtain ex- haled TV of 8 to 10 mL/kg, RR less than 25/min. PEEP increased gradually and up to 10 cm H2O until FiO2 require- ment was 0.6 or less. Settings were ad- justed based on continuous oximetry and measurements of ABG. Patients with AHRI following lung resection if they met at least three of the following criteria: dyspnea at rest, active contraction of accessory respiratory muscles, PaO2/FiO2 less than 200, chest radiographic abnormalities HFNC high-flow nasal cannula, SO standard oxygen, NIV non-invasive ventilation, PPC postoperative pulmonary complications, ARISCAT assess respiratory risk in surgical patients in Catalonia, FiO2 fraction of inspired oxygen, SpO2 peripheral capillary oxygen saturation, L liters, min minute, ABG arterial blood gas, EPAP expiratory positive airway pressure, IPAP inspiratory positive airway pressure, ARF acute respiratory failure, ICU intensive care unit, PEEP positive end-expiratory pressure, BiPAP bilevel positive airway pressure, SBT spontaneous breathing trail, CPAP continuous positive airway pressure, PaO2 partial pressure of oxygen, AHRI acute hypoxemic respiratory insufficiency, mm millimeter, Hg mercury, TV tidal volume, PS pressure support Mortality Hypoxemia occurs frequently in the post-operative period and can lead to acute respiratory failure. Several NIV was associated with a significant reduction of mor- tality in comparison with standard oxygen therapy (OR Page 5 of 11 Zayed et al. Journal of Intensive Care (2020) 8:51 Table 1 Characteristics of the included studies Study (author, year) Study groups Study design Inclusion criteria S c Yu, 2017 HFNC 56, SO 54 Multicenter, prospective, randomize, interventional trial Patients who underwent thoracoscopic lobectomy because of lung tumor and were at intermediate to high risk for PPC as determined by an ARISCAT score ≥26. Patients were immunocompetent, not pregnant, between 18 and 80 years old H 6 t m S o b S Futier, 2016 HFNC 108, SO 112 Multicenter, randomized controlled trial Adult patients scheduled for planned or unplanned abdominal, or abdominal and thoracic surgery with and anticipated duration of 2 h or more and an ARISCAT score ≥26 H m S n a Gupta, 2016 HFNC 10, NIV 10 Pilot study, single-center, randomized controlled trial Postoperative hypoxemia in post-liver transplant patients H m N c Jaber, 2016 SO 145, NIV 148 Multicenter, randomized, parallel-group clinical trial Patients older than 19 who had undergone laparoscopic or non- laparoscopic elective or nonelective ab- dominal surgery under general anesthesia that were diagnosed with ARF within 7 days of surgical procedure defined as persistence of more than 30 min of hypoxemia S 1 9 N N a l Stephan, 2015 HFNC 414, NIV 416 Multicenter, randomized, noninferiority trial Patients who had undergone cardiothoracic surgery who developed ARF (failure of SBT or successful SBT but failed extubation) or were deemed at risk for respiratory failure post- extubation due to preexisting risk factors H F 9 B v P 9 Zhu, 2013 NIV 48, SO 47 Single-center, prospective, randomized control study Patients who after cardiac surgery developed ARF after initial extubation who were hemodynamically stable with no evidence of bleeding N a 9 S t Squadrone, 2005 NIV 105, SO 104 Multicenter, randomized, controlled, unblinded study Post-op elective abdominal surgery under GA if surgery required laparotomy and time of viscera exposure longer than 90 min. Patients were extubated after surgery, and if they developed a PaO2/FiO2 of 300 less, they were included in study. NA Mortality C C u O o t l t S Auriant, 2001 NIV 24, SO 24 Prospective, randomized controlled trial Patients with AHRI following lung resection if they met at least three of the following criteria: dyspnea at rest, active contraction of accessory respiratory muscles, PaO2/FiO2 less than 200, chest radiographic abnormalities N B e l a S S Antonelli, 2000 NIV 20, SO 20 Single center, prospective, randomized study Recipients of solid organ transplants with acute hypoxemic respiratory failure. Criteria included acute respiratory distress, respiratory rate greater than 35/min, ratio of PaO2/ FiO2 of less than 200, active contraction of accessory muscles or paradoxical abdominal motion N m h 2 u m j a S w a HFNC high-flow nasal cannula, SO standard oxygen, NIV non-invasive ventilation, PPC postoperative pulmonary co surgical patients in Catalonia, FiO2 fraction of inspired oxygen, SpO2 peripheral capillary oxygen saturation, L liter Settings of experimental group and control group intervention Follow-up period HFNC: received at a flow rate if 35 to 60 L/min and FiO2 was titrated from 45 to 100% to maintain a SpO2 of 95% or more SO: received oxygen via nasal prongs or facemasks with FiO2 titrated between 45 and 100% to maintain SpO2 of 95% or more 72 h following extubation HFNC: flow rate of 50 to 60 L/min to maintain an SpO2 of 95% or more SO: O2 delivered continuously using nasal prongs or facemasks to maintain an SpO2 of 95% or more 7 days post-op HFNC: initiated at a flow rate of 60 L/ min and titrated according to ABG NIV: set EPAP of 5 cm and IPAP at 10 cm and titrated according to ABG 48 h post- op. SO: supplemental O2 at a rate of up to 15 L/min to maintain SpO2 of at least 94% NIV: facemask connected to an ICU or NIV dedicated ventilator titrating PEEP and FiO2 to maintain an SpO2 of at least 94% 90 days post-op. Antonelli, 2000 NIV 20, SO 20 Single center, prospective, randomized study Mortality Standard oxygen: Venturi mask started with FiO2 of 40% and titrated to achieve a level of SpO2 90% Recipients of solid organ transplants with acute hypoxemic respiratory failure. Criteria included acute respiratory distress, respiratory rate greater than 35/min, ratio of PaO2/ FiO2 of less than 200, active contraction of accessory muscles or paradoxical abdominal motion HFNC high-flow nasal cannula, SO standard oxygen, NIV non-invasive ventilation, PPC postoperative pulmonary complications, ARISCAT assess respiratory risk in surgical patients in Catalonia, FiO2 fraction of inspired oxygen, SpO2 peripheral capillary oxygen saturation, L liters, min minute, ABG arterial blood gas, EPAP expiratory positive airway pressure, IPAP inspiratory positive airway pressure, ARF acute respiratory failure, ICU intensive care unit, PEEP positive end-expiratory pressure, BiPAP bilevel positive airway pressure, SBT spontaneous breathing trail, CPAP continuous positive airway pressure, PaO2 partial pressure of oxygen, AHRI acute hypoxemic respiratory insufficiency, mm millimeter, Hg mercury, TV tidal volume, PS pressure support Zayed et al. Mortality Although mortality rates were lower in the NIV group (14 vs 21%), the difference did not reach a statis- tical significance in their study [22]. Similarly, patients undergoing cardiothoracic surgery and treated with NIV for postoperative respiratory failure had lower rates of intubation and mortality when compared to patients treated with standard oxygen therapy [19, 24]. factors play a role in the development of post-operative respiratory failure, including diaphragmatic dysfunction, retained secretions, and atelectasis and alveolar collapse which promote bacterial growth and infections [27–29]. Non-invasive ventilation (NIV) improves oxygenation by recruiting collapsed alveoli and increasing tidal volume participating in gas exchange without hemodynamic ad- verse events [1, 30]. However, previous studies and meta-analyses had not shown a significant reduction in intubation rates with prophylactic use of NIV after sur- gery, despite the reduction in the incidence of post- operative pulmonary complications [31–33]. factors play a role in the development of post-operative respiratory failure, including diaphragmatic dysfunction, retained secretions, and atelectasis and alveolar collapse which promote bacterial growth and infections [27–29]. Non-invasive ventilation (NIV) improves oxygenation by recruiting collapsed alveoli and increasing tidal volume participating in gas exchange without hemodynamic ad- verse events [1, 30]. However, previous studies and meta-analyses had not shown a significant reduction in intubation rates with prophylactic use of NIV after sur- gery, despite the reduction in the incidence of post- operative pulmonary complications [31–33]. yg y HFNC is a new oxygenation strategy that has been used more frequently in patients with respiratory failure. It is found to be more comfortable than NIV and can deliver concentrated oxygen reaching 100% with a high flow rate up to 60 mL/min [34, 35]. Furthermore, it can provide positive end-expiratory pressure up to 2– 3mmHG [34, 35]. The use of HFNC has shown benefi- cial effects in patients who developed post-extubation respiratory failure or when used during intubation to prevent hypoxemia when compared to standard oxygen [36–38]. Additionally, Frat et al. found lower mortality rates with HFNC in comparison with NIV and conven- tional oxygen in patients with non-hypercapnic hypox- emic respiratory failure. However, other trials did not find differences between HFNC and standard oxygen therapy [39–42]. The use of HFNC in the post-operative Currently, NIV is recommended in the treatment of patients with post-operative respiratory failure according to the ERS/ATS guidelines [15]. Mortality Journal of Intensive Care (2020) 8:51 Page 6 of 11 Table 2 Baseline demographic and clinical characteristics of included patients Study name Study groups Total number Age Male (%) BMI SAPS II score Respiratory rate PaO2/FiO2 ratio PaCO2 Yu 2017 HFNC 56 56.31 ± 7.03 54 26.32 ± 4.73 NA 18.43 ± 3.45 350 ± 33.87 41.73 ± 6.33 SO 54 55.82 ± 7.92 52 25.19 ± 5.02 NA 17.98 ± 3.87 341 ± 40.65 43.52 ± 4.93 Jaber 2016 NIV 148 62.5 ± 14.5 78.4 27.2 ± 5.9 33.6 ± 12.8 28.2 + 7.7 201 ± 69 39 ± 7 SO 145 64.4 ± 13.1 74.5 27.1 ± 6.2 33.4 ± 11.7 28.8 + 7.3 188 ± 71 37 ± 7 Gupta 2016 HFNC 10 NA NA NA NA NA NA NA NIV 10 NA NA NA NA NA NA NA Futier 2016 HFNC 108 62 ± 12 56 25 ± 4 NA NA NA NA SO 112 61 ± 13 57 25 ± 4 NA NA NA NA Stephan 2015 NIV 416 63.9 (62.6– 65.2) 67 28.2 (27.6– 28.7) 28.8 (27.7– 30.0) 23.2 (22.6– 24.0) 203 (195–212) 39.1 (38.4– 39.8) HFNC 414 63.8 (62.5– 65.2) 66 28.3 (27.8– 28.8) 29.0 (27.8– 30.1) 22.8 (22.1– 23.5) 196 (187–204) 38.7 (38.1– 39.4) Zhu 2013 NIV 48 62 ± 10.3 66 25.3 ± 4.6 NA 28.3 ± 8.6 NA 38.9 ± 12.2 SO 47 61 ± 12.2 57 24.4 ± 3.5 NA 25.4 ± 6.7 NA 38.3 ± 11.3 Squadrone 2005 NIV 105 66 ± 9 68 26.5 ± 4.7 27 ± 7 NA 247 ± 33 39 ± 7 SO 104 65 ± 10 62 26.3 ± 4.5 28 ± 8 255 ± 31 39 ± 5 Aurian 2001 NIV 24 58.9 ± 10 NA NA 16.9 ± 5.4 26.25 ± 13.2 124 ± 50.2 63.9 ± 20.5 SO 24 63 ± 9 NA NA 16.8 ± 4.4 29.5 ± 6.9 111 ± 54.3 43.4 ± 9.3 Antonelle 2000 NIV 20 45 ± 19 65 NA NA 38 ± 3 NA 42 ± 10 SO 20 44 ± 10 60 NA NA 37 ± 1 NA 38 ± 10 Data are provided percent (%), mean ± SD, or median (interquartile range) HFNC high-flow nasal cannula, SO standard oxygen, NIV non-invasive ventilation, BMI body mass index, SAPS simplified acute physiology score, PaO2/FiO2 partial pressure of arterial oxygen to fraction of inspired oxygen ratio, PaCO2 partial pressure of arterial carbon dioxide, NA not available Table 2 Baseline demographic and clinical characteristics of included patients Data are provided percent (%), mean ± SD, or median (interquartile range) HFNC high-flow nasal cannula, SO standard oxygen, NIV non-invasive ventilation, BMI body mass index, SAPS simplified acute physiology score, PaO2/FiO2 partial pressure of arterial oxygen to fraction of inspired oxygen ratio, PaCO2 partial pressure of arterial carbon dioxide, NA not available days on mechanical ventilator, and significantly lower rates of healthcare-associated infections, including pneu- monia. Mortality Our results indicate that intubation rates and mortality are significantly lower in patient populations who are at an increased risk or have developed postoperative respiratory failure treated with NIV in comparison with standard oxygen. In our sub- group analysis, mortality benefit was only noted in pa- tients undergoing cardiothoracic surgeries but not in abdominal surgeries. In an RCT examining NIV vs standard oxygen therapy in patients with respiratory fail- ure after abdominal surgeries, Jaber and Antonelli found that NIV was associated with lower intubation rates, less Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care Page 7 of 11 Fig. 2 Forest plot for the rate of intubation between competing interventions Fig. 2 Forest plot for the rate of intubation between competing interventions results were also found in both subgroups (cardiothor- acic surgeries and abdominal surgeries). period was investigated by several RCTs. In a large RCT involving more than 800 patients after cardiac surgery, the use of HFNC and NIV in the treatment of high-risk patients or those who had developed post-operative re- spiratory failure was similar between both interventions with similar intubation rates, mortality, and rates of hospital-acquired infections [20]. In our analysis, there was no difference between HFNC and NIV in intubation rates, mortality, and ICU-acquired infections. Similar In addition, HFNC was associated with lower intub- ation rates in patients following cardiothoracic surgeries but not following abdominal surgeries when compared to standard oxygen therapy. This could be explained by the fact that in thoracic surgery, HFNC could minimize lung decruitment post-extubation by providing some level of continuous positive airway pressure through Fig. 3 Forest plot for the rate of intubation between competing interventions following cardiothoracic surgery (a) and abdominal (b) surgery Fig. 5 Forest plot for mortality between competing interventions following cardiothoracic surgery (a) and abdominal surgery (b) Fig. 3 Forest plot for the rate of intubation between competing interventions following cardiothoracic surgery (a) and abdominal (b) surgery Fig. 3 Forest plot for the rate of intubation between competing interventions following cardiothoracic surgery ( Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care Page 8 of 11 Fig. 4 Forest plot for mortality between competing interventions Fig. 4 Forest plot for mortality between competing interventions Although there was no significant difference between HFNC and NIV with regard to rates of intubation, mor- tality, and ICU-associated infections, when each of these two strategies was compared to standard oxygen, NIV was associated with a survival benefit especially in pa- tients who had cardiothoracic surgery. Additionally, there was a trend toward lower mortality in abdominal surgeries, but HFNC had no mortality benefit in the high-flow ventilation, though this positive pressure can be variable due to the leak around the nasal cannula and nonguaranteed closed mouth of the patients [26]. Furthermore, we found lower rates of infections with HFNC and NIV when compared to standard oxygen are attributed to lower intubation rates in both interven- tions, which avoids the need for mechanical ventilation and decreases catheter-associated infections. Fig. 5 Forest plot for mortality between competing interventions following cardiothoracic surgery (a) and abdominal surgery (b) Limitations Our analysis has several limitations. First, we were un- able to perform analysis based on various risk factors, duration of surgery, severity scores, and different surgi- cal types as we lack patients’ level data. Second, blinding of intervention and personnel was impossible given the nature of intervention. Third, there were few sample size and limited events, and therefore, larger trials and long- term outcomes are needed. Fourth, we used informative prior module for our analysis which could affect the re- sults given the small number of included trials. Fifth, there was a significant time gap between included stud- ies through which there was a significant development in the ICU management, preoperative and postoperative evaluation and care, supportive management, and cri- teria for admission to the ICU. Supplementary figure 3: Regression of PaCO2 on intubation rate between non-invasive ventilation and standard oxygen. Higher PaCO2 was associated with a lower risk for intubation with NIV use (P < 0.05). Supplementary figure 4:. Direct meta-analysis results between NIV ver- sus standard oxygen showing forest plots for intubation rate, mortality, and ICU acquired infections. Supplementary figure 5:. Results of individual studies comparing between HFNC and standard oxygen for different outcomes. Supplementary figure 5:. Results of individual studies comparing between HFNC and standard oxygen for different outcomes. Supplementary figure 6:. Results of individual studies comparing between NIV and HFNC for different outcomes. Supplementary figure 6:. Results of individual studies comparing between NIV and HFNC for different outcomes. Supplementary figure 6:. Results of individual studies comparing between NIV and HFNC for different outcomes. Supplementary information Supplementary information Supplementary information accompanies this p 1186/s40560-020-00468-x. Supplementary information Supplementary information accompanies this paper at https://doi.org/10. 1186/s40560-020-00468-x. Nevertheless, due to the low events, further RCTs are needed to compare between both interventions in differ- ent types of surgeries to determine its effect on various long-term clinical outcomes and quality of life and also to examine whether certain patients’ risk factors could affect the beneficial effects of these interventions to- wards reduction of intubation rates and mortality. y Supplementary information accompanies this paper at https://doi.org/10. 1186/s40560-020-00468-x. Supplementary informatio 1186/s40560-020-00468-x. Supplementary figure 1:. Network geometry. Number of participants in each group represented with node size and the edge widths are proportional to the number of studies between different interventions. A = High-flow nasal cannula; B = Non-invasive ventilation; C = standard oxygen therapy. Supplementary figure 1:. Network geometry. Number of participants in each group represented with node size and the edge widths are proportional to the number of studies between different interventions. A = High-flow nasal cannula; B = Non-invasive ventilation; C = standard oxygen therapy. Supplementary figure 2:. Risk of bias assessment based on authors’ judgment for each of the included RCTs. Blank items indicate unclear risk of bias. Supplementary figure 2:. Risk of bias assessment based on authors’ judgment for each of the included RCTs. Blank items indicate unclear risk of bias. ig. 5 Forest plot for mortality between competing interventions following cardiothoracic surgery (a) and abdominal surgery Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care Page 9 of 11 Fig. 6 Forest plot for the incidence of ICU-acquired infections between competing interventions ig. 6 Forest plot for the incidence of ICU-acquired infections between competing interventions oxygen therapy. In addition, HFNC was associated with reduced rates of intubation and ICU-acquired infections but not mortality in comparison with standard oxygen. There was no significant difference between HFNC and NIV on the various studied clinical outcomes. total patient population and both subgroups. Whether a lower number of patients included in the comparison between HFNC and standard oxygen or certain other factors could have contributed to the inability to detect a mortality benefit despite a significant reduction of in- tubation rates is needed to be addressed in further larger and well-controlled trials. Abbreviations NIV N i i NIV: Non-invasive ventilation; HFNC: High-flow nasal cannula; ICU: Intensive care unit; RCT: Randomized controlled trial; OR: Odds ratio; Cr.I.: Credible interval Ethics approval and consent to participate NA Ethics approval and consent to participate NA Competing interests h h h 14. Parke R, McGuinness S, Dixon R, Jull A. Open-label, phase II study of routine high-flow nasal oxygen therapy in cardiac surgical patients. Br J Anaesth [Internet]. 2013;111:925–31 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/23921199. The authors have no competing interests to declare. Consent for publication Th h i h J The authors give the Journal of Intensive Care the consent for publishing this manuscript. There are no personal information that required consent. 13. Frat J-P, Coudroy R, Thille AW. Non-invasive ventilation or high-flow oxygen therapy: when to choose one over the other? Theatr Res Int. 2018; Available from: http://www.ncbi.nlm.nih.gov/pubmed/30406954. Conclusion Y.Z.: study design, literature search, data analysis, data extraction, drafting the manuscript, final approval of the manuscript. B.K.: study design, data analysis, drafting the manuscript, final approval of the manuscript. M.B.: literature search, drafting the manuscript, final approval of the manuscript. L.R.: data extraction, drafting the manuscript, final approval of the manuscript. I.G.: literature search, drafting the manuscript, final approval of the manuscript. E.I.: study design, data analysis, literature search, Among patients who are at risk for developing post- operative respiratory failure, or have developed post- operative respiratory failure, the use of NIV was associ- ated with reduced rates of intubation, mortality, and ICU-acquired infections in comparison with standard Page 10 of 11 Page 10 of 11 Page 10 of 11 Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care (2020) 8:51 8. Futier E, Constantin J-M, Paugam-Burtz C, Pascal J, Eurin M, Neuschwander A, et al. A trial of intraoperative low-tidal-volume ventilation in abdominal surgery. N Engl J Med [Internet]. 2013;369:428–37 Available from: http:// www.ncbi.nlm.nih.gov/pubmed/23902482. drafting the manuscript, final approval of the manuscript. J.K.: literature search, drafting the manuscript, final approval of the manuscript. F.R.: drafting the manuscript, final approval of the manuscript. S.S.: data extraction, drafting the manuscript, final approval of the manuscript. A.B.: data analysis, drafting the manuscript, final approval of the manuscript. S.Se.: data extraction, drafting the manuscript, final approval of the manuscript. G.B.: drafting the manuscript, final approval of the manuscript. E.S.: study design, drafting the manuscript, final approval of the manuscript. The authors read and approved the final manuscript. 9. Canet J, Gallart L. Postoperative respiratory failure: pathogenesis, prediction, and prevention. Curr Opin Crit Care [Internet]. 2014;20:56–62 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24240985. 10. Brochard L, Mancebo J, Wysocki M, Lofaso F, Conti G, Rauss A, et al. Noninvasive ventilation for acute exacerbations of chronic obstructive pulmonary disease. N Engl J Med [Internet]. 1995;333:817–22 Available from: http://www.ncbi.nlm.nih.gov/pubmed/7651472. Funding There was no funding for this work. References 18. Gupta S, Govil D, Srinivasan S, Patel SJ, Gupta A, Tomar DS, et al. High flow nasal cannula (HFNC) as an alternative to noninvasive ventilation (NIV) in acute respiratory failure (ARF) in immunosuppressed patients-an Indian post liver transplant experience. Intensive Care Med Exp Conf 29th Annu Congr Eur Soc intensive care Med ESICM 2016 Italy [Internet]. 2016;4 Available from: https://www.cochranelibrary.com/central/doi/10.1002/central/CN- 01407482/full. 1. Chiumello D, Chevallard G, Gregoretti C. Non-invasive ventilation in postoperative patients: a systematic review. Intensive Care Med [Internet]. 2011;37:918–29 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/21424246. 2. Pearse RM, Moreno RP, Bauer P, Pelosi P, Metnitz P, Spies C, et al. Mortality after surgery in Europe: a 7 day cohort study. Lancet (London, England) [Internet]. 2012;380:1059–65 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/22998715. 19. Zhu G, Wang D, Liu S, Jia M, Jia S. Efficacy and safety of noninvasive positive pressure ventilation in the treatment of acute respiratory failure after cardiac surgery. Chin Med J (Engl) [Internet]. 2013;126:4463–9 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24286408. 3. Serpa Neto A, Hemmes SNT, Barbas CSV, Beiderlinden M, Fernandez- Bustamante A, Futier E, et al. Incidence of mortality and morbidity related to postoperative lung injury in patients who have undergone abdominal or thoracic surgery: a systematic review and meta-analysis. Lancet Respir Med [Internet]. 2014;2:1007–15 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/25466352. 3. Serpa Neto A, Hemmes SNT, Barbas CSV, Beiderlinden M, Fernandez- Bustamante A, Futier E, et al. Incidence of mortality and morbidity related to postoperative lung injury in patients who have undergone abdominal or thoracic surgery: a systematic review and meta-analysis. Lancet Respir Med [Internet]. 2014;2:1007–15 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/25466352. 20. Stéphan F, Barrucand B, Petit P, Rézaiguia-Delclaux S, Médard A, Delannoy B, et al. High-flow nasal oxygen vs noninvasive positive airway pressure in hypoxemic patients after cardiothoracic surgery: a randomized clinical trial. JAMA [Internet]. 2015;313:2331–9 Available from: http://www.ncbi.nlm.nih. gov/pubmed/25980660. 4. Canet J, Gallart L, Gomar C, Paluzie G, Vallès J, Castillo J, et al. Prediction of postoperative pulmonary complications in a population-based surgical cohort. Anesthesiology [Internet]. 2010;113:1338–50 Available from: http:// www.ncbi.nlm.nih.gov/pubmed/21045639. 4. Canet J, Gallart L, Gomar C, Paluzie G, Vallès J, Castillo J, et al. Prediction of postoperative pulmonary complications in a population-based surgical cohort. Anesthesiology [Internet]. 2010;113:1338–50 Available from: http:// www.ncbi.nlm.nih.gov/pubmed/21045639. 21. Squadrone V, Coha M, Cerutti E, Schellino MM, Biolino P, Occella P, et al. Continuous positive airway pressure for treatment of postoperative hypoxemia: a randomized controlled trial. JAMA [Internet]. 2005;293:589–95 Available from: http://www.ncbi.nlm.nih.gov/pubmed/15687314. 5. Funding h 11. Keenan SP, Sinuff T, Cook DJ, Hill NS. Which patients with acute exacerbation of chronic obstructive pulmonary disease benefit from noninvasive positive-pressure ventilation? A systematic review of the literature. Ann Intern Med [Internet]. 2003;138:861–70 Available from: http:// www.ncbi.nlm.nih.gov/pubmed/12779296. There was no funding for this work. Availability of data and materials Data and materials are available and can be presented upon request. Availability of data and materials Data and materials are available and can be presented upon request. y Data and materials are available and can be presented upon request. 12. Lightowler JV, Wedzicha JA, Elliott MW, Ram FSF. Non-invasive positive pressure ventilation to treat respiratory failure resulting from exacerbations of chronic obstructive pulmonary disease: Cochrane systematic review and meta-analysis. BMJ [Internet]. 2003;326:185 Available from: http://www.ncbi. nlm.nih.gov/pubmed/12543832. Author details 1 1Department of Internal Medicine, Hurley Medical Center/Michigan State University, One Hurley Plaza, Suite 212, Flint, MI 48503, USA. 2Knight Cardiovascular Institute, Oregon Health & Science University, Portland, Oregon, USA. 3College of Human Medicine, Michigan State University, East Lansing, MI, USA. 4Department of Anesthesia, Lebanese University, Beirut, Lebanon. 5College of Osteopathic Medicine, Michigan State University, East Lansing, MI, USA. 6Department of Pulmonary and Critical Care, Hurley Medical Center/Michigan State University, Flint, MI, USA. 1Department of Internal Medicine, Hurley Medical Center/Michigan State University, One Hurley Plaza, Suite 212, Flint, MI 48503, USA. 2Knight Cardiovascular Institute, Oregon Health & Science University, Portland, Oregon, USA. 3College of Human Medicine, Michigan State University, East Lansing, MI, USA. 4Department of Anesthesia, Lebanese University, Beirut, 5 15. Rochwerg B, Brochard L, Elliott MW, Hess D, Hill NS, Nava S, et al. Official ERS/ATS clinical practice guidelines: noninvasive ventilation for acute respiratory failure. Eur Respir J [Internet], Available from. 2017;50 http:// www.ncbi.nlm.nih.gov/pubmed/28860265. 16. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev [Internet]. 2015;4:1 Available from: http://www.ncbi.nlm.nih.gov/pubmed/25554246. 17. Higgins JPT, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ [Internet]. 2011;343:d5928 Available from: http://www.ncbi.nlm. nih.gov/pubmed/22008217. Received: 3 April 2020 Accepted: 8 July 2020 References Mazo V, Sabaté S, Canet J, Gallart L, de Abreu MG, Belda J, et al. Prospective external validation of a predictive score for postoperative pulmonary complications. Anesthesiology [Internet]. 2014;121:219–31 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24901240. 5. Mazo V, Sabaté S, Canet J, Gallart L, de Abreu MG, Belda J, et al. Prospective external validation of a predictive score for postoperative pulmonary complications. Anesthesiology [Internet]. 2014;121:219–31 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24901240. 22. Jaber S, Lescot T, Futier E, Paugam-Burtz C, Seguin P, Ferrandiere M, et al. Effect of noninvasive ventilation on tracheal reintubation among patients with hypoxemic respiratory failure following abdominal surgery: a randomized clinical trial. JAMA [Internet]. 2016;315:1345–53 Available from: http://www.ncbi.nlm.nih.gov/pubmed/26975890. 6. Gallart L, Canet J. Post-operative pulmonary complications: Understanding definitions and risk assessment. Best Pract Res Clin Anaesthesiol [Internet]. 2015;29:315–30. Available from. http://www.ncbi.nlm.nih.gov/ pubmed/26643097. 7. Arozullah AM, Khuri SF, Henderson WG, Daley J. Participants in the National Veterans Affairs Surgical Quality Improvement Program. Development and validation of a multifactorial risk index for predicting postoperative pneumonia after major noncardiac surgery. Ann Intern Med [Internet]. 2001; 135:847–57 Available from: http://www.ncbi.nlm.nih.gov/pubmed/11712875. 6. Gallart L, Canet J. Post-operative pulmonary complications: Understanding definitions and risk assessment. Best Pract Res Clin Anaesthesiol [Internet]. 2015;29:315–30. Available from. http://www.ncbi.nlm.nih.gov/ pubmed/26643097. 23. Futier E, Paugam-Burtz C, Godet T, Khoy-Ear L, Rozencwajg S, Delay J-M, et al. Effect of early postextubation high-flow nasal cannula vs conventional oxygen therapy on hypoxaemia in patients after major abdominal surgery: a French multicentre randomised controlled trial (OPERA). Intensive Care Med [Internet]. 2016;42:1888–98 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/27771739. 7. Arozullah AM, Khuri SF, Henderson WG, Daley J. Participants in the National Veterans Affairs Surgical Quality Improvement Program. Development and validation of a multifactorial risk index for predicting postoperative pneumonia after major noncardiac surgery. Ann Intern Med [Internet]. 2001; 135:847–57 Available from: http://www.ncbi.nlm.nih.gov/pubmed/11712875. 7. Arozullah AM, Khuri SF, Henderson WG, Daley J. Participants in the National Veterans Affairs Surgical Quality Improvement Program. Development and validation of a multifactorial risk index for predicting postoperative pneumonia after major noncardiac surgery. Ann Intern Med [Internet]. 2001; 135:847–57 Available from: http://www.ncbi.nlm.nih.gov/pubmed/11712875. Page 11 of 11 Page 11 of 11 Zayed et al. Journal of Intensive Care (2020) 8:51 Zayed et al. Journal of Intensive Care (2020) 8:51 Engl J Med [Internet]. 2015;372:2185–96 Available from: http://www.ncbi. nlm.nih.gov/pubmed/25981908. Engl J Med [Internet]. 2015;372:2185–96 Available from: http://www.ncbi. nlm.nih.gov/pubmed/25981908. 24. Auriant I, Jallot A, Hervé P, Cerrina J, Le Roy LF, Fournier JL, et al. References Noninvasive ventilation reduces mortality in acute respiratory failure following lung resection. Am J Respir Crit Care Med [Internet]. 2001;164: 1231–5 Available from: http://www.ncbi.nlm.nih.gov/pubmed/11673215. 40. Lemiale V, Mokart D, Mayaux J, Lambert J, Rabbat A, Demoule A, et al. The effects of a 2-h trial of high-flow oxygen by nasal cannula versus Venturi mask in immunocompromised patients with hypoxemic acute respiratory failure: a multicenter randomized trial. Crit Care [Internet]. 2015;19:380 Available from: http://www.ncbi.nlm.nih.gov/pubmed/26521922. 25. Antonelli M, Conti G, Bufi M, Costa MG, Lappa A, Rocco M, et al. Noninvasive ventilation for treatment of acute respiratory failure in patients undergoing solid organ transplantation: a randomized trial. JAMA [Internet]. 2000;283:235–41 Available from: http://www.ncbi.nlm.nih.gov/pubmed/1 0634340. 41. Jones PG, Kamona S, Doran O, Sawtell F, Wilsher M. Randomized controlled trial of humidified high-flow nasal oxygen for acute respiratory distress in the emergency department: the HOT-ER study. Respir Care [Internet]. 2016; 61:291–9 Available from: http://www.ncbi.nlm.nih.gov/pubmed/26577199. 41. Jones PG, Kamona S, Doran O, Sawtell F, Wilsher M. Randomized controlled trial of humidified high-flow nasal oxygen for acute respiratory distress in the emergency department: the HOT-ER study. Respir Care [Internet]. 2016; 61:291–9 Available from: http://www.ncbi.nlm.nih.gov/pubmed/26577199. 26. Yu Y, Qian X, Liu C, Zhu C. Effect of high-flow nasal cannula versus conventional oxygen therapy for patients with thoracoscopic lobectomy after extubation. Can Respir J [Internet]. 2017;2017:7894631 Available from: http://www.ncbi.nlm.nih.gov/pubmed/28298878. 26. Yu Y, Qian X, Liu C, Zhu C. Effect of high-flow nasal cannula versus conventional oxygen therapy for patients with thoracoscopic lobectomy after extubation. Can Respir J [Internet]. 2017;2017:7894631 Available from: http://www.ncbi.nlm.nih.gov/pubmed/28298878. 42. Azoulay E, Lemiale V, Mokart D, Nseir S, Argaud L, Pène F, et al. Effect of high-flow nasal oxygen vs standard oxygen on 28-day mortality in immunocompromised patients with acute respiratory failure: the HIGH randomized clinical trial. JAMA [Internet]. 2018;320:2099–107 Available from: http://www.ncbi.nlm.nih.gov/pubmed/30357270. 42. Azoulay E, Lemiale V, Mokart D, Nseir S, Argaud L, Pène F, et al. Effect of high-flow nasal oxygen vs standard oxygen on 28-day mortality in immunocompromised patients with acute respiratory failure: the HIGH randomized clinical trial. JAMA [Internet]. 2018;320:2099–107 Available from: http://www.ncbi.nlm.nih.gov/pubmed/30357270. 27. Ahmed AA. Dysfunction of the diaphragm. N Engl J Med [Internet]. 2012; 366:2036–7 author reply 2037. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/22621640. 27. Ahmed AA. Dysfunction of the diaphragm. N Engl J Med [Internet]. 2012; 366:2036–7 author reply 2037. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/22621640. 28. Canet J, Mazo V. Postoperative pulmonary complications. Minerva Anestesiol [Internet]. 2010;76:138–43 Available from: http://www.ncbi.nlm. nih.gov/pubmed/20150855. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 29. van Kaam AH, Lachmann RA, Herting E, De Jaegere A, van Iwaarden F, Noorduyn LA, et al. Reducing atelectasis attenuates bacterial growth and translocation in experimental pneumonia. Am J Respir Crit Care Med [Internet]. 2004;169:1046–53 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/14977624. 30. Jaber S, Antonelli M. Preventive or curative postoperative noninvasive ventilation after thoracic surgery: still a grey zone? Intensive Care Med [Internet]. 2014;40:280–3 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/24445592. 31. Zarbock A, Mueller E, Netzer S, Gabriel A, Feindt P, Kindgen-Milles D. Prophylactic nasal continuous positive airway pressure following cardiac surgery protects from postoperative pulmonary complications: a prospective, randomized, controlled trial in 500 patients. Chest [Internet]. 2009;135:1252–9 Available from: http://www.ncbi.nlm.nih.gov/pubmed/1901 7864. 32. Al Jaaly E, Fiorentino F, Reeves BC, Ind PW, Angelini GD, Kemp S, et al. Effect of adding postoperative noninvasive ventilation to usual care to prevent pulmonary complications in patients undergoing coronary artery bypass grafting: a randomized controlled trial. J Thorac Cardiovasc Surg [Internet]. 2013;146:912–8 Available from: http://www.ncbi.nlm.nih.gov/pubmed/23582 830. 33. Ireland CJ, Chapman TM, Mathew SF, Herbison GP, Zacharias M. Continuous positive airway pressure (CPAP) during the postoperative period for prevention of postoperative morbidity and mortality following major abdominal surgery. Cochrane database Syst Rev [Internet]. 2014:CD008930 Available from: http://www.ncbi.nlm.nih.gov/pubmed/25081420. 34. Mauri T, Turrini C, Eronia N, Grasselli G, Volta CA, Bellani G, et al. Physiologic effects of high-flow nasal cannula in acute hypoxemic respiratory failure. Am J Respir Crit Care Med [Internet]. 2017;195:1207–15 Available from: http://www.ncbi.nlm.nih.gov/pubmed/27997805. 35. Corley A, Caruana LR, Barnett AG, Tronstad O, Fraser JF. Oxygen delivery through high-flow nasal cannulae increase end-expiratory lung volume and reduce respiratory rate in post-cardiac surgical patients. Br J Anaesth [Internet]. 2011;107:998–1004 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/21908497. 36. Maggiore SM, Idone FA, Vaschetto R, Festa R, Cataldo A, Antonicelli F, et al. Nasal high-flow versus Venturi mask oxygen therapy after extubation. Effects on oxygenation, comfort, and clinical outcome. Am J Respir Crit Care Med [Internet]. 2014;190:282–8 Available from: http://www.ncbi.nlm.nih.gov/ pubmed/25003980. 37. Rittayamai N, Tscheikuna J, Rujiwit P. High-flow nasal cannula versus conventional oxygen therapy after endotracheal extubation: a randomized crossover physiologic study. Respir Care [Internet]. 2014;59:485–90 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24046462. 37. Rittayamai N, Tscheikuna J, Rujiwit P. High-flow nasal cannula versus conventional oxygen therapy after endotracheal extubation: a randomized crossover physiologic study. Respir Care [Internet]. 2014;59:485–90 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24046462. 38. Publisher’s Note Hernández G, Vaquero C, González P, Subira C, Frutos-Vivar F, Rialp G, et al. Effect of postextubation high-flow nasal cannula vs conventional oxygen therapy on reintubation in low-risk patients: a randomized clinical trial. JAMA [Internet]. 2016;315:1354–61 Available from: http://www.ncbi.nlm.nih. gov/pubmed/26975498. 38. Hernández G, Vaquero C, González P, Subira C, Frutos-Vivar F, Rialp G, et al. Effect of postextubation high-flow nasal cannula vs conventional oxygen therapy on reintubation in low-risk patients: a randomized clinical trial. JAMA [Internet]. 2016;315:1354–61 Available from: http://www.ncbi.nlm.nih. gov/pubmed/26975498. 39. Frat J-P, Thille AW, Mercat A, Girault C, Ragot S, Perbet S, et al. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N 39. Frat J-P, Thille AW, Mercat A, Girault C, Ragot S, Perbet S, et al. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N
https://openalex.org/W2294667940	https://www.frontiersin.org/articles/10.3389/fmicb.2016.00193/pdf	English	null	RNA-seq Profiling Reveals Novel Target Genes of LexA in the Cyanobacterium Synechocystis sp. PCC 6803	Frontiers in microbiology	2,016	cc-by	11,852	Edited by: Takashi Osanai, Meiji University, Japan Reviewed by: Lei Chen, Tianjin University, China Rei Narikawa, Shizuoka University, Japan Correspondence: Yukako Hihara hihara@molbiol.saitama-u.ac.jp Edited by: Takashi Osanai, Meiji University, Japan Reviewed by: Lei Chen, Tianjin University, China Rei Narikawa, Shizuoka University, Japan hihara@molbiol.saitama-u.ac.jp Specialty section: This article was submitted to Microbiotechnology, Ecotoxicology and Bioremediation, a section of the journal Frontiers in Microbiology Received: 07 January 2016 Accepted: 04 February 2016 Published: 19 February 2016 Specialty section: This article was submitted to Microbiotechnology, Ecotoxicology and Bioremediation, a section of the journal Frontiers in Microbiology ORIGINAL RESEARCH published: 19 February 2016 doi: 10.3389/fmicb.2016.00193 RNA-seq Proﬁling Reveals Novel Target Genes of LexA in the Cyanobacterium Synechocystis sp. PCC 6803 Ayumi Kizawa 1, Akihito Kawahara 2, Yasushi Takimura 2, Yoshitaka Nishiyama 1 and Yukako Hihara 1, 3 1 Department of Biochemistry and Molecular Biology, Graduate School of Science and Engineering, Saitama University, Saitama, Japan, 2 Biological Science Laboratories, KAO Corporation, Wakayama, Japan, 3 Core Research of Evolutional Science and Technology, Japan Science and Technology Agency, Saitama, Japan LexA is a well-established transcriptional repressor of SOS genes induced by DNA damage in Escherichia coli and other bacterial species. However, LexA in the cyanobacterium Synechocystis sp. PCC 6803 has been suggested not to be involved in SOS response. In this study, we performed RNA-seq analysis of the wild-type strain and the lexA-disrupted mutant to obtain the comprehensive view of LexA-regulated genes in Synechocystis. Disruption of lexA positively or negatively affected expression of genes related to various cellular functions such as phototactic motility, accumulation of the major compatible solute glucosylglycerol and subunits of bidirectional hydrogenase, photosystem I, and phycobilisome complexes. We also observed increase in the expression level of genes related to iron and manganese uptake in the mutant at the later stage of cultivation. However, none of the genes related to DNA metabolism were affected by disruption of lexA. DNA gel mobility shift assay using the recombinant LexA protein suggested that LexA binds to the upstream region of pilA7, pilA9, ggpS, and slr1670 to directly regulate their expression, but changes in the expression level of photosystem I genes by disruption of lexA is likely a secondary effect. INTRODUCTION The LexA protein in Escherichia coli has been well-characterized as the key regulator of the SOS response induced by DNA damage (Butala et al., 2009). Under non-stress conditions, LexA binds to the promoter regions of more than 40 genes involved in the SOS response and represses their expression. When DNA is damaged, LexA undergoes autoproteolytic cleavage upon association with RecA protein activated through binding of single-stranded DNA fragments. As a consequence of auto-cleavage of the Ala84-Gly85 peptide bond carried out by Ser119 and Lys156, LexA loses DNA binding activity, thereby inducing the SOS response. Received: 07 January 2016 Accepted: 04 February 2016 Published: 19 February 2016 Received: 07 January 2016 Accepted: 04 February 2016 Keywords: cyanobacteria, LexA, RNA-seq, Synechocystis, transcriptome Citation: DNA microarray analysis revealed that expression of the pilA genes encoding the subunits of the type IV pilus- like structure was lowered in the mutant. Although regulation of various cellular processes has been suggested, we currently have still a fragmentary understanding of the function of LexA in S.6803. DNA microarray analysis has been the most popular methods of genome-wide transcriptome proﬁling. However, it has been supplanted by RNA-seq analysis in which isolated transcripts are converted into the complementary DNA (cDNA) followed by direct sequence in a massively parallel DNA sequencing-based approach. The advantages of RNA-seq over DNA microarray are its higher resolution and better dynamic range of detecting diﬀerential gene expression (Zhao et al., 2014). In order to obtain the comprehensive view of LexA-regulated genes in S.6803, here we performed RNA-seq analysis of the wild-type (WT) strain and the lexA-disrupted mutant. The results of RNA-seq analysis indicate that LexA in S.6803 regulates speciﬁc cellular functions such as phototactic motility accumulation of the major in several cyanobacterial species were suggested not to be involved in the typical E. coli-type SOS regulation. In Anabaena sp. PCC 7120, auto-cleavage of the Ala84-Gly85 bond of LexA does not occur at physiological pH even in the presence of activated RecA (Kumar et al., 2015). In the case of Synechocystis sp. PCC 6803 (S.6803), LexA lacks the conserved Ala-Gly auto- cleavage site and the serine of the Ser-Lys dyad required for auto- cleavage activity (Patterson-Fortin et al., 2006) and auto-cleavage of LexA in S.6803 has not been reported so far. DNA microarray analysis revealed that LexA depletion did not aﬀect the expression level of genes involved in DNA metabolism (Domain et al., 2004). (sll1626)-disrupted mutant (1lexA) was grown under the same conditions, except that 20 µg mL−1 kanamycin (Km) was added to the medium. Cell density was estimated by measuring OD730 using a spectrophotometer (model UV-160A, Shimadzu). RNA Gel Blot Analysis Isolation of total RNA by the hot phenol method and RNA gel blot analyses, using DIG RNA Labeling and Detection Kit (Roche), were performed as described previously (Muramatsu and Hihara, 2003). Template DNA fragments for in vitro transcription to generate RNA probes were prepared by PCR using the primers shown in Table S1. DNA microarray analysis has been the most popular methods of genome-wide transcriptome proﬁling. However, it has been supplanted by RNA-seq analysis in which isolated transcripts are converted into the complementary DNA (cDNA) followed by direct sequence in a massively parallel DNA sequencing-based approach. The advantages of RNA-seq over DNA microarray are its higher resolution and better dynamic range of detecting diﬀerential gene expression (Zhao et al., 2014). In order to obtain the comprehensive view of LexA-regulated genes in S.6803, here we performed RNA-seq analysis of the wild-type (WT) strain and the lexA-disrupted mutant. The results of RNA-seq analysis indicate that LexA in S.6803 regulates speciﬁc cellular functions such as phototactic motility, accumulation of the major compatible solute glucosylglycerol and subunits of bidirectional hydrogenase, and photosynthetic complexes, but not the SOS response. DNA gel mobility shift assay using the recombinant LexA protein suggested that LexA binds to the upstream region of pilA7, pilA9, ggpS, and slr1670 to directly regulate their expression. Generation of the lexA (sll1626)-Disrupted Mutant The coding region of lexA (612 bp, from nucleotide 1319330 to 1318719 according to numbering in CyanoBase) was disrupted by insertion of a kanamycin resistance (Kmr) cassette. The upstream and downstream fragments including the lexA coding sequence were ampliﬁed by PCR from the genomic DNA of the WT strain using the primer sets lexA-F and Km- lexA-R (for ampliﬁcation of 404 bp upstream fragment, from nucleotide 1319525 to 1319122) and Km-lexA-F and lexA-R (for ampliﬁcation of 394 bp downstream fragment, from nucleotide 1318996 to 1318603; Table S1). Kmr cassette was PCR ampliﬁed from the pRL161 plasmid using the primer set Km-F and Km- R (Table S1). The ampliﬁed lexA fragments and Kmr cassette were fused together by the fusion PCR method (Wang et al., 2002) using the primer set lexA-F and lexA-R. The WT strain was transformed with the fusion PCR product and transformants (1lexA mutant) were selected in the presence of Km. g The cellular processes regulated by LexA in S.6803 have been implied by studies reporting isolation of LexA as a binding factor to the promoter region of speciﬁc genes, such as the hoxEFUYH operon encoding bidirectional hydrogenase (Gutekunst et al., 2005; Oliveira and Lindblad, 2005), crhR encoding RNA helicase (Patterson-Fortin et al., 2006), and sbtA encoding sodium-dependent bicarbonate transporter (Lieman- Hurwitz et al., 2009). Domain et al. (2004) performed DNA microarray analysis of the LexA-depleted strain and found that most of genes aﬀected were previously reported to be regulated by the availability of inorganic carbon (Wang et al., 2004). Kamei et al. (2001) reported that the lexA-disrupted mutant of the motile strain of S.6803 (denoted PCC strain) showed non-motile phenotype. DNA microarray analysis revealed that expression of the pilA genes encoding the subunits of the type IV pilus- like structure was lowered in the mutant. Although regulation of various cellular processes has been suggested, we currently have still a fragmentary understanding of the function of LexA in S.6803. Immunoblot Analysis Total proteins were extracted from Synechocystis cells as described previously (Ishii and Hihara, 2008) and separated by 15% (w/v) SDS-PAGE, followed by electroblotting onto PVDF membranes (Immobilon-P; Millipore). Immunodetection was done using a rabbit polyclonal antibody raised against His-LexA recombinant protein. Goat anti-rabbit IgG conjugated to alkaline phosphatase was used as a secondary antibody. Determination of Pigment Contents In vivo absorption spectra of whole cells suspended in BG- 11 medium were measured at room temperature using a spectrophotometer (V-650 Spectrometer, JASCO) with ISV-722 integrating sphare. Chlorophyll and phycocyanin contents were calculated from the peak heights of absorption spectra using the equations described in Arnon et al. (1974). MATERIALS AND METHODS RNA-seq analysis was carried out using cultures at OD730 = 0.5 and OD730 = 1.0 with three biological replicates. WT and 1lexA were inoculated into new media at OD730 = 0.1 and incubated for 50 and 80 h, respectively, to be harvested at OD730 = 0.5. Similarly, WT and 1lexA were inoculated at OD730 = 0.1 and incubated for 70 and 120 h, respectively, to be harvested at OD730 = 1.0. Isolation of total RNA by the hot phenol method Citation: Kizawa A, Kawahara A, Takimura Y, Nishiyama Y and Hihara Y (2016) RNA-seq Proﬁling Reveals Novel Target Genes of LexA in the Cyanobacterium Synechocystis sp. PCC 6803. Front. Microbiol. 7:193. doi: 10.3389/fmicb.2016.00193 Kizawa A, Kawahara A, Takimura Y, Nishiyama Y and Hihara Y (2016) RNA-seq Proﬁling Reveals Novel Target Genes of LexA in the Cyanobacterium Synechocystis sp. PCC 6803. Front. Microbiol. 7:193. doi: 10.3389/fmicb.2016.00193 Genes encoding LexA homologs are highly conserved in bacterial genomes and LexA-dependent transcriptional regulation of genes involved in DNA repair has been reported in various bacterial species (Erill et al., 2007; Butala et al., 2009), indicating that the regulation of SOS regulon by LexA might be a universal adaptation strategy of bacteria to DNA damage. However, LexA homologs February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org RNA-seq of lexA in Synechocystis Kizawa et al. in several cyanobacterial species were suggested not to be involved in the typical E. coli-type SOS regulation. In Anabaena sp. PCC 7120, auto-cleavage of the Ala84-Gly85 bond of LexA does not occur at physiological pH even in the presence of activated RecA (Kumar et al., 2015). In the case of Synechocystis sp. PCC 6803 (S.6803), LexA lacks the conserved Ala-Gly auto- cleavage site and the serine of the Ser-Lys dyad required for auto- cleavage activity (Patterson-Fortin et al., 2006) and auto-cleavage of LexA in S.6803 has not been reported so far. DNA microarray analysis revealed that LexA depletion did not aﬀect the expression level of genes involved in DNA metabolism (Domain et al., 2004). The cellular processes regulated by LexA in S.6803 have been implied by studies reporting isolation of LexA as a binding factor to the promoter region of speciﬁc genes, such as the hoxEFUYH operon encoding bidirectional hydrogenase (Gutekunst et al., 2005; Oliveira and Lindblad, 2005), crhR encoding RNA helicase (Patterson-Fortin et al., 2006), and sbtA encoding sodium-dependent bicarbonate transporter (Lieman- Hurwitz et al., 2009). Domain et al. (2004) performed DNA microarray analysis of the LexA-depleted strain and found that most of genes aﬀected were previously reported to be regulated by the availability of inorganic carbon (Wang et al., 2004). Kamei et al. (2001) reported that the lexA-disrupted mutant of the motile strain of S.6803 (denoted PCC strain) showed non-motile phenotype. DNA Gel Mobility Shift Assay was performed as described previously (Muramatsu and Hihara, 2003). To eliminate genomic DNA from total RNA samples, each sample was added with DNase I (TaKaRa) and incubated at 37◦C for 3 h. Total RNA concentration was measured with Nanodrop 2000 (Thermo Fisher Scientiﬁc). The Ribo-Zero Magnetic Kit for Bacteria (Epicentre) was used to remove ribosomal RNA from each sample. Concentration and quality of mRNA samples were examined using an Agilent 2100 Bioanalyzer. TruSeq RNA Sample Prep Kit v2 (Illumina) was used for cDNA library construction, and the libraries were sequenced using the Illumina MiSeq system. 12 samples in total were analyzed using two cartridge of MiSeq Reagent Kit v3 (Illumina). y y Probes for DNA gel mobility shift assays were obtained by PCR ampliﬁcation with primers shown in Table S1 using genomic DNA as a template. The 3′ end of the DNA fragment for each probe was labeled with digoxigenin (DIG)-ddUTP by using the terminal transferase method according to the manufacturer’s instructions (DIG gel shift kit 2nd generation; Roche). Gel mobility shift assays were performed by using a DIG gel shift kit 2nd generation (Roche) according to the manufacturer’s instruction except that 1 mM DTT was added to the reaction mixture. A total of 64 million reads data was obtained from 12 samples. To quantify expression level of each gene, nucleotide sequences of obtained reads were mapped to the genomic sequence of GT-I strain of S.6803 (Kanesaki et al., 2012) (NC_017038.1; http://www.ncbi.nlm.nih.gov/nuccore/NC_017038) using CLC Genomics Workbench 7.5.1 software (Qiagen). Raw read counts were divided by length of the transcripts and total number of million mapped reads in each sample to obtain reads per kilobase per million (RPKM) values (Mortazavi et al., 2008). TCC package of R software (Sun et al., 2013) was used to detect the diﬀerentially expressed genes between WT and 1lexA. A false discovery rate of <0.01 was considered to be signiﬁcant. RNA-seq Transcriptome Analysis RNA seq Transcriptome Analysis To investigate the diﬀerence in gene expression proﬁle between WT and 1lexA, total RNA was isolated from cultures incubated under normal growth conditions and RNA-seq analysis was performed. Figure 2 shows MA plots of the gene expression data obtained from cultures at OD730 = 0.5 and OD730 = 1.0. There were 1011 genes diﬀerentially expressed between strains at OD730 = 0.5 as shown in magenta (Figure 2A). Among them, expression levels of 315 genes were more than two-fold higher and those of 28 genes were more than two-fold lower in 1lexA than in WT (Table S2). In the case of WT and 1lexA cells at OD730 = 1.0, there were 447 genes diﬀerentially expressed between strains (Figure 2B). Among them, expression levels of 360 genes were more than two-fold higher and those of 21 genes were more than two-fold lower in 1lexA than in WT (Table S3). p p g E. coli BL21(DE3) harboring the His-LexA expression construct was grown to an OD600 = 0.6 in 250 mL of 2 × yeast extract-tryptone (YT) medium containing 20 µg mL−1 Km at 37◦C and induced with 0.013% of isopropyl β-D-thiogalactoside for 3 h. The cells were pelleted by centrifugation at 5800 g for 2 min, resuspended in 50 mM sodium phosphate buﬀer, pH 7.4, containing 0.5 M NaCl and 60 mM imidazole, and disrupted by three rounds of sonication with Soniﬁer 450 (Branson) for 2 min with interval of 1 min on ice. After the removal of whole cells and insoluble material by centrifugation, the soluble protein fraction was ﬁltered through a 0.2 µm ﬁlter (DISMIC-25CS; ADVANTEC). His-LexA was puriﬁed by nickel-aﬃnity column chromatography using a HisTrap FF crude (GE Healthcare). The soluble protein fraction was applied to the column equilibrated with 20 mM phosphate buﬀer, pH 7.4, containing 0.5 M NaCl and 60 mM imidazole, washed with 20 mM phosphate buﬀer, pH 7.4, containing 0.5 M NaCl and 80 mM imidazole, and eluted with 20 mM phosphate buﬀer, pH 7.4, containing 0.5 M NaCl and 300 mM imidazole. Puriﬁed His-LexA was desalted by a HiTrap Desalting column (GE Healthcare). Protein composition was examined by 15% (w/v) SDS-PAGE followed by staining with Coomassie Brilliant Blue R-250. Table 1 shows the list of genes whose expression was aﬀected by disruption of lexA. Strains and Culture Conditions Strains and Culture Conditions A glucose-tolerant non-motile strain (GT strain) of Synechocystis sp. PCC 6803 was grown at 32◦C in BG-11 medium containing 20 mM HEPES-NaOH, pH 7.0, under continuous illumination at 20 µmol photons m−2 s−1 with bubbling of air. The lexA February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 2 RNA-seq of lexA in Synechocystis Kizawa et al. Overexpression and Puriﬁcation of His-lexA The coding region of the lexA gene was ampliﬁed by PCR using the primers lexA-NdeI-F and lexA-XhoI-R (Table S1), containing NdeI and XhoI sites at their 5′ end, respectively. The ampliﬁed lexA coding fragment was cloned into the pT7Blue T- vector (Novagen), digested with NdeI and XhoI and subcloned into the same restriction sites in pET28a vector (Novagen) to express the LexA protein with an N-terminal 6 × His-tag. Frontiers in Microbiology \| www.frontiersin.org Characterization of the lexA (sll1626)-Disrupted Mutant To reveal the function of LexA in GT strain of S.6803, we disrupted the lexA gene by inserting a Kmr cassette within the coding region (Figure 1A). Although a fully segregated mutant was not obtained (Figure 1B), RNA gel blot and immunoblot analyses revealed that both the lexA transcript (Figure 1C) and LexA protein (Figure 1D) levels were below the detection limit in the partially segregated mutant (1lexA) grown under normal growth conditions. Under the same conditions, 1lexA displayed several abnormal phenotypes. The doubling time of 1lexA was longer (31.4 h) than that of WT (19.5 h) at log phase, whereas the diﬀerence in growth rate between strains became smaller at stationary phase (Figure 1E). Amounts of chlorophyll and phycocyanin in 1lexA calculated from the peak heights of cellular absorption spectra were 93 and 80% of WT levels, respectively (Figure 1F). Microscopic observation revealed that cell size of 1lexA was heterogeneous and tended to be larger than that of WT (Figure 1G). RNA-seq Transcriptome Analysis The higher resolution and better dynamic range of RNA-seq analysis compared to DNA microarray analysis enabled listing of small ORFs such as ssl1577, ggpR (ssl3076), ssr1251, ssr1473 and ssr3589, and genes with low expression level (low RPKM value) that cannot be detected by previous DNA microarray analyses. Diﬀerentially expressed genes can Table 1 shows the list of genes whose expression was aﬀected by disruption of lexA. The higher resolution and better dynamic range of RNA-seq analysis compared to DNA microarray analysis enabled listing of small ORFs such as ssl1577, ggpR (ssl3076), ssr1251, ssr1473 and ssr3589, and genes with low expression level (low RPKM value) that cannot be detected by previous DNA microarray analyses. Diﬀerentially expressed genes can February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 3 RNA-seq of lexA in Synechocystis Kizawa et al. FIGURE 1 \| Generation and characterization of the 1lexA mutant. (A) Scheme of the construct for disruption of lexA. The kanamycin resistance (Kmr) cartridge was inserted into the coding region. Arrows indicate the primers used for PCR ampliﬁcation shown in (B). (B) PCR ampliﬁcation of the lexA gene using genomic DNA from WT and the 1lexA mutant as templates. (C) RNA gel blot analysis of the lexA transcripts detected by single-stranded RNA probe. 3 µg of total RNA were loaded per lane. Total RNA was stained with methylene blue to show the equal loading. (D) Immunoblot analysis of the LexA protein detected by anti-LexA antibody. 5 µg of total protein from cell lysate were loaded per lane. (E) Growth curves of WT (open circles) and the 1lexA mutant (closed circles) under normal growth conditions. (F) Amounts of photosynthetic pigments calculated from the peak heights of cellular absorption spectra. (G) Observation of cell morphology by differential interference contrast microscopy. FIGURE 1 \| Generation and characterization of the 1lexA mutant. (A) Scheme of the construct for disruption of lexA. The kanamycin resistance (Kmr) cartridge was inserted into the coding region. Arrows indicate the primers used for PCR ampliﬁcation shown in (B). (B) PCR ampliﬁcation of the lexA gene using genomic DNA from WT and the 1lexA mutant as templates. (C) RNA gel blot analysis of the lexA transcripts detected by single-stranded RNA probe. 3 µg of total RNA were loaded per lane. Total RNA was stained with methylene blue to show the equal loading. RNA-seq Transcriptome Analysis (D) Immunoblot analysis of the LexA protein detected by anti-LexA antibody. 5 µg of total protein from cell lysate were loaded per lane. (E) Growth curves of WT (open circles) and the 1lexA mutant (closed circles) under normal growth conditions. (F) Amounts of photosynthetic pigments calculated from the peak heights of cellular absorption spectra. (G) Observation of cell morphology by differential interference contrast microscopy. Furthermore, we observed that several genes other than pilA involved in motility were aﬀected by disruption of lexA. Expression of pixG-pixH-pixI-pixJ1-pixJ2-pixL (sll0038-0043) encoding regulatory factors involved in positive phototaxis increased in the mutant (Tables S2). It has been reported that motility is controlled by cAMP level in S.6803 and inactivation of cya1 encoding adenylate cyclase or sycrp1 encoding cAMP receptor protein results in loss of motility (Terauchi and Ohmori, 1999; Yoshimura et al., 2002a). Although expression of cya1 and sycrp1 itself was not so much aﬀected by disruption of lexA, decrease in expression levels of ﬁve genes, pilA9- pilA10-pilA11-slr2018 and cccS (slr1667), out of six genes reported to be decreased by disruption of sycrp1(Yoshimura et al., 2002b) was observed (Table 1). cccS is also considered to be related to motility, since its disruption resulted in loss of the thick pili (Yoshimura et al., 2010). We observed expression level of sycrp2 is lower in 1lexA (Table 1), although involvement of SYCRP2 in regulation of motility has not been reported. be categorized into several groups according to related cellular functions as mentioned below. Frontiers in Microbiology \| www.frontiersin.org Motility-Related Genes The motile strain of S.6803 exhibits phototactic motility dependent on the type IV-like thick pilus structure (Brahamsha and Bhaya, 2014). In S.6803 genome, there are multiple genes homologous to the pilA gene encoding the subunit of the type IV pilus-like structure. Among them, pilA1 was shown to be responsible for the thick pilus structure, motility, and transformation eﬃciency (Bhaya et al., 1999; Yoshihara et al., 2001), whereas functions of other pilA-like genes are unknown. We observed that their expression is positively or negatively aﬀected by disruption of the lexA gene. Expression of pilA7-pilA8 was largely enhanced whereas that of pilA9-pilA10-pilA11 and pilA1-pilA2 decreased. The observed decrease in expression level of pilA1 and pilA9-pilA10-pilA11 was consistent with the results of DNA microarray analysis of the 1lexA mutant in the motile PCC strain (Kamei et al., 2001). February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 4 RNA-seq of lexA in Synechocystis Kizawa et al. TABLE 1 \| Genes with altered expression in the lexA-disrupted mutant. Gene No. Frontiers in Microbiology \| www.frontiersin.org Motility-Related Genes Gene symbol Deﬁnition OD730 = 0.5 OD730 = 1.0 Average RPKM Ratio Average RPKM Ratio WT 1lexA 1lexA/WT WT 1lexA 1lexA/WT MOTILITY sll1694 pilA1 Pilin polypeptide PilA1 7672.37 3764.79 0.49 9711.78 3875.35 0.40 sll1695 pilA2 Pilin polypeptide PilA2 245.64 161.17 0.66 272.03 149.86 0.55 slr1930 pilA7 Type 4 pilin-like protein 110.72 1163.62 10.51 148.23 1231.28 8.31 slr1931 pilA8 Type 4 pilin-like protein 193.59 1186.22 6.13 226.51 1250.18 5.52 slr2015 pilA9 Type 4 pilin-like protein 68.30 16.96 0.25 55.96 19.50 0.35 slr2016 pilA10 Type 4 pilin-like protein 38.83 23.24 0.60 37.41 16.90 0.45 slr2017 pilA11 Type 4 pilin-like protein 72.06 31.47 0.44 73.43 35.82 0.49 slr2018 Unknown protein 84.75 48.63 0.57 97.49 56.12 0.58 sll1291 taxP2 Two-component response regulator PatA subfamily 185.08 63.22 0.34 203.50 86.70 0.43 slr1667 cccS Hypothetical protein (target gene of sycrp1) 25.66 13.57 0.53 45.84 18.36 0.40 GLUCOSYLGLYCEROL METABOLISM slr1670 Unknown protein 29.83 309.81 10.39 24.85 250.51 10.08 slr1672 glpK Glycerol kinase 52.30 299.46 5.73 43.84 227.47 5.19 slr1673 spoU Probable tRNA/rRNA methyltransferase 38.42 179.60 4.67 42.98 144.32 3.36 ssl3076 ggpR Unknown protein 2.03 16.39 8.07 0.00 0.87 N.D sll1566 ggpS Glucosylglycerolphosphate synthase 34.65 497.47 14.36 27.78 431.85 15.54 sll1085 glpD Glycerol-3-phosphate dehydrogenase 32.37 191.71 5.92 36.05 215.11 5.97 slr0529 ggtB Glucosylglycerol transport system substrate-binding protein 17.28 79.09 4.58 18.76 76.80 4.09 slr0530 ggtC Glucosylglycerol transport system permease protein 21.55 102.21 4.74 23.34 88.37 3.79 HYDROGENASE sll1220 hoxE Diaphorase subunit of the bidirectional hydrogenase 100.28 48.94 0.49 62.85 32.54 0.52 sll1221 hoxF Diaphorase subunit of the bidirectional hydrogenase 64.58 31.77 0.49 49.47 31.90 0.64 sll1223 hoxU Diaphorase subunit of the bidirectional hydrogenase 96.63 49.39 0.51 68.18 47.23 0.69 sll1224 hoxY Hydrogenase subunit of the bidirectional hydrogenase 70.59 35.93 0.51 37.57 28.59 0.76 ssl2420 Unknown protein 54.08 25.39 0.47 42.93 24.51 0.57 slr1675 hypA1 Putative hydrogenase expression/formation protein HypA1 31.16 266.71 8.56 33.07 195.50 5.91 PHOTOSYNTHESIS slr0737 psaD Photosystem I subunit II 9924.04 5201.83 0.52 6914.54 4585.06 0.66 slr1835 psaB P700 apoprotein subunit Ib 34540.96 22842.73 0.66 42976.65 25014.56 0.58 smr0004 psaI Photosystem I subunit VIII 3093.51 2356.63 0.76 290.74 157.92 0.54 ssl0563 psaC Photosystem I subunit VII 10241.74 5794.74 0.57 5317.26 3374.82 0.63 ssr0390 psaK1 Photosystem I subunit X 2883.40 1628.37 0.56 1635.89 1152.40 0.70 slr0012 rbcS Rubisco small subunit 3477.08 1936.92 0.56 4542.25 3204.02 0.71 slr0011 rbcX Possible Rubisco chaperonin 3913.65 2224.26 0.57 5157.07 3785.52 0.73 sll0247 isiA Iron-stress chlorophyll-binding protein 55.34 124.36 2.25 53.76 761.15 14.16 sll0248 isiB Flavodoxin 10.09 35.32 3.50 5.91 121.62 20.59 sll1577 cpcB Phycocyanin beta subunit 62726.44 35125.15 0.56 54762.14 27956.67 0.51 sll1578 cpcA Phycocyanin alpha subunit 79538.85 42145.92 0.53 69158.50 35179.81 0.51 sll1579 cpcC2 Phycobilisome rod linker polypeptide 13587.17 7921.46 0.58 12570.33 6318.76 0.50 sll1580 cpcC1 Phycobilisome rod linker polypeptide 14076.98 8090.37 0.57 12675.97 6193.81 0.49 ssl3093 cpcD Phycobilisome small rod linker polypeptide 4048.58 2772.20 0.68 3023.79 1747.70 0.58 sll1471 cpcG2 Phycobilisome rod-core linker polypeptide 847.25 349.04 0.41 638.78 289.60 0.45 ssl2542 hliA High light-inducible polypeptide HliA 22.74 133.52 5.87 26.31 167.07 6.35 ssr2595 hliB High light-inducible polypeptide HliB 71.98 314.95 4.38 40.31 186.15 4.62 (Continued) sll1291 taxP2 Two-component response regulator PatA subfamily 185.08 63.22 0.34 203.50 86.70 0.43 slr1667 cccS Hypothetical protein (target gene of sycrp1) 25.66 13.57 0.53 45.84 18.36 0.40 GLUCOSYLGLYCEROL METABOLISM slr1670 Unknown protein 29.83 309.81 10.39 24.85 250.51 10.08 slr1672 glpK Glycerol kinase 52.30 299.46 5.73 43.84 227.47 5.19 slr1673 spoU Probable tRNA/rRNA methyltransferase 38.42 179.60 4.67 42.98 144.32 3.36 ssl3076 ggpR Unknown protein 2.03 16.39 8.07 0.00 0.87 N.D sll1566 ggpS Glucosylglycerolphosphate synthase 34.65 497.47 14.36 27.78 431.85 15.54 sll1085 glpD Glycerol-3-phosphate dehydrogenase 32.37 191.71 5.92 36.05 215.11 5.97 slr0529 ggtB Glucosylglycerol transport system substrate-binding protein 17.28 79.09 4.58 18.76 76.80 4.09 slr0530 ggtC Glucosylglycerol transport system permease protein 21.55 102.21 4.74 23.34 88.37 3.79 HYDROGENASE sll1220 hoxE Diaphorase subunit of the bidirectional hydrogenase 100.28 48.94 0.49 62.85 32.54 0.52 sll1221 hoxF Diaphorase subunit of the bidirectional hydrogenase 64.58 31.77 0.49 49.47 31.90 0.64 sll1223 hoxU Diaphorase subunit of the bidirectional hydrogenase 96.63 49.39 0.51 68.18 47.23 0.69 sll1224 hoxY Hydrogenase subunit of the bidirectional hydrogenase 70.59 35.93 0.51 37.57 28.59 0.76 ssl2420 Unknown protein 54.08 25.39 0.47 42.93 24.51 0.57 slr1675 hypA1 Putative hydrogenase expression/formation protein HypA1 31.16 266.71 8.56 33.07 195.50 5.91 PHOTOSYNTHESIS slr0737 psaD Photosystem I subunit II 9924.04 5201.83 0.52 6914.54 4585.06 0.66 l 1835 B P700 t i b it Ib 34540 96 22842 73 0 66 42976 65 25014 56 0 58 GLUCOSYLGLYCEROL METABOLISM February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 5 RNA-seq of lexA in Synechocystis Kizawa et al. Motility-Related Genes TABLE 1 \| Continued Gene No. Frontiers in Microbiology \| www.frontiersin.org Motility-Related Genes Gene symbol Deﬁnition OD730 = 0.5 OD730 = 1.0 Average RPKM Ratio Average RPKM Ratio WT 1lexA 1lexA/WT WT 1lexA 1lexA/WT slr0506 por Light-dependent NADPH-protochlorophyllide oxidoreductase 247.75 186.82 0.75 318.96 181.53 0.57 slr0749 chlL Light-independent protochlorophyllide reductase iron protein subunit ChlL 489.76 109.67 0.22 98.28 41.69 0.42 slr0750 chlN Light-independent protochlorophyllide reductase subunit ChlN 149.90 49.34 0.33 240.63 81.30 0.34 CHAPERONES sll0430 htpG HtpG, heat shock protein 90 156.26 589.81 3.77 115.48 607.00 5.26 sll0909 dnaJ DnaJ, heat shock protein 40 24.21 216.45 8.94 22.63 304.56 13.46 sll1514 hspA 16.6 kDa small heat shock protein 100.08 1082.85 10.82 121.59 1307.29 10.75 REGULATORY FUNCTIONS sll0094 hik37 Two-component sensor histidine kinase 52.56 33.12 0.63 74.37 39.50 0.53 sll0775 Unknown protein 29.04 391.93 13.50 40.24 339.89 8.45 sll0776 spkD Serine/threonine kinase 16.14 248.58 15.40 27.99 235.49 8.41 sll0777 Putative carboxypeptidase 24.39 219.13 8.98 35.22 210.97 5.99 sll0778 ABC transporter, ATP-binding protein 14.17 72.87 5.14 17.63 71.67 4.07 sll0790 hik31 Two-component sensor histidine kinase 49.77 256.96 5.16 74.75 321.59 4.30 sll0797 nrsR, rppA Redox-responsive and/or Ni(II)-responsive regulator, two-component response regulator OmpR subfamily 12.59 7.00 0.56 6.80 10.22 1.50 sll1003 hik13 Two-component sensor histidine kinase 10.65 45.77 4.30 10.70 50.35 4.71 sll1626 lexA LexA repressor 1658.80 10652.12 6.42 1451.28 9173.74 6.32 sll1924 sycrp2 cAMP receptor protein sycrp1 homolog 23.44 13.60 0.58 11.10 13.47 1.21 slr0895 prqR Transcriptional regulator 15.85 67.22 4.24 19.15 62.39 3.26 slr1564 sigF Group 3 RNA polymerase sigma factor 297.21 217.21 0.73 330.72 209.54 0.63 slr1594 patA Two-component response regulator PatA subfamily 40.38 195.88 4.85 49.36 265.06 5.37 slr1760 rre8 Two-component response regulator 15.72 104.05 6.62 26.41 88.06 3.33 slr2098 hik21 Two-component hybrid sensor and regulator 26.67 129.37 4.85 38.41 152.72 3.98 TRANSPORT AND BINDING PROTEINS sll1404 exbB3 Biopolymer transport ExbB protein homolog 84.83 131.51 1.55 16.17 290.69 17.98 sll1405 exbD, sll1405 Biopolymer transport ExbD protein homolog 28.62 59.22 2.07 13.99 118.68 8.48 sll1406 fhuA Ferrichrome-iron receptor 29.10 57.61 1.98 25.09 132.90 5.30 sll1598 mntC Mn transporter MntC 9.96 28.07 2.82 9.36 112.24 11.99 sll1599 mntA Manganese transport system ATP-binding protein MntA 4.04 17.11 4.24 4.18 65.76 15.74 slr1295 futA1 Iron transport system substrate-binding protein 479.40 669.20 1.40 183.65 1131.71 6.16 slr0513 futA2 Iron transport system substrate-binding protein 425.93 832.10 1.95 366.04 2076.58 5.67 slr1488 Multidrug resistance family ABC transporter 15.16 46.45 3.06 17.28 130.69 7.56 OTHER CATEGORIES sll1358 mncA Oxalate decarboxylase, periplasmic protein 248.08 94.87 0.38 133.04 73.57 0.55 sll1688 thrC Threonine synthase 1010.37 704.75 0.70 1408.81 822.45 0.58 sll1709 gdh 3-ketoacyl-acyl carrier protein reductase 99.76 504.56 5.06 59.54 418.64 7.03 slr0518 abfB Similar to alpha-L-arabinofuranosidase B 40.86 27.97 0.68 45.82 26.80 0.58 slr0786 mapB Methionine aminopeptidase 6.73 20.91 3.11 6.85 35.69 5.21 slr1204 degP Protease 109.11 511.99 4.69 122.90 572.69 4.66 slr1764 capA Similar to tellurium resistance protein TerE 26.98 222.68 8.25 23.67 125.43 5.30 slr2097 glbN Cyanoglobin 128.40 1919.90 14.95 116.38 1589.74 13.66 slr2116 spsA Probable glycosyltransferase 22.84 11.76 0.51 16.60 14.06 0.85 ssr1720 tyrS Similar to tyrosyl tRNA synthetase 5.26 14.68 2.79 3.77 24.90 6.61 (Continued) February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 6 RNA-seq of lexA in Synechocystis Kizawa et al. Motility-Related Genes TABLE 1 \| Continued Gene No. Motility-Related Genes Gene symbol Deﬁnition OD730 = 0.5 OD730 = 1.0 Average RPKM Ratio Average RPKM Ratio WT 1lexA 1lexA/WT WT 1lexA 1lexA/WT UNKNOWN PROTEIN sll0249 Hypothetical protein 10.39 23.76 2.29 6.55 84.74 12.95 sll0327 Unknown protein 113.13 1929.23 17.05 157.08 1310.83 8.35 sll0328 Unknown protein 47.79 859.88 17.99 48.45 595.60 12.29 sll0443 Unknown protein 71.82 419.75 5.84 75.08 336.96 4.49 sll0444 Unknown protein 116.90 563.08 4.82 97.11 410.52 4.23 sll0445 Unknown protein 114.46 528.48 4.62 117.45 444.43 3.78 sll0448 Unknown protein 10.17 49.14 4.83 7.21 42.81 5.94 sll0543 Hypothetical protein 677.95 37.76 0.06 535.50 31.01 0.06 sll0783 Unknown protein 84.80 38.09 0.45 86.35 68.05 0.79 sll0846 Hypothetical protein 133.72 577.77 4.32 106.12 532.70 5.02 sll0910 Unknown protein 27.90 188.61 6.76 23.33 243.59 10.44 sll0911 Unknown protein 29.09 170.08 5.85 16.73 124.75 7.46 sll1009 Unknown protein 611.42 3545.01 5.80 926.18 3621.41 3.91 sll1119 Hypothetical protein 109.22 83.03 0.76 73.68 44.88 0.61 sll1236 Unknown protein 30.41 472.85 15.55 27.60 136.10 4.93 sll1239 Unknown protein 88.93 685.38 7.71 41.49 494.27 11.91 sll1240 Unknown protein 23.31 240.78 10.33 18.06 205.72 11.39 sll1241 Unknown protein 26.35 199.16 7.56 14.51 181.65 12.52 sll1247 Hypothetical protein 137.85 61.68 0.45 194.48 104.16 0.54 sll1359 Unknown protein 76.42 37.70 0.49 48.30 36.91 0.76 sll1396 Unknown protein 59.93 13.16 0.22 54.18 13.94 0.26 sll1472 Unknown protein 97.90 48.95 0.50 61.97 49.50 0.80 sll1483 Periplasmic protein 57.32 302.45 5.28 46.76 183.74 3.93 sll1549 Salt-enhanced periplasmic protein 232.67 121.52 0.52 18.94 215.48 11.37 sll1660 Hypothetical protein 45.49 351.04 7.72 48.25 355.40 7.37 sll1722 Hypothetical protein 18.55 130.15 7.01 13.41 45.98 3.43 sll1723 Probable glycosyltransferase 11.63 67.76 5.83 7.87 24.45 3.11 sll1851 Unknown protein 136.11 100.98 0.74 15.16 103.21 6.81 sll1913 Hypothetical protein 24.83 103.47 4.17 18.85 108.19 5.74 sll1921 Hypothetical protein 136.06 1118.71 8.22 152.96 1173.63 7.67 sll1956 Hypothetical protein 68.61 42.42 0.62 60.44 37.54 0.62 slr0105 Hypothetical protein 40.10 327.60 8.17 51.00 312.12 6.12 slr0106 Unknown protein 58.01 324.48 5.59 74.96 308.21 4.11 slr0179 Hypothetical protein 11.14 405.28 36.37 19.47 345.32 17.74 slr0196 Unknown protein 38.61 187.90 4.87 14.43 111.92 7.75 slr0317 Hypothetical protein 18.01 103.04 5.72 20.82 119.39 5.73 slr0362 Hypothetical protein 48.52 240.11 4.95 55.46 204.71 3.69 slr0364 Hypothetical protein 5.58 25.68 4.61 5.73 12.25 2.14 slr0393 Unknown protein 17.69 35.91 2.03 7.26 38.90 5.36 slr0442 Unknown protein 175.06 105.14 0.60 207.58 121.90 0.59 slr0572 Unknown protein 350.05 18.16 0.05 194.88 16.91 0.09 slr0573 Unknown protein 18.80 3.65 0.19 22.65 6.20 0.27 slr0581 Unknown protein 79.35 334.01 4.21 60.26 159.45 2.65 slr0617 Unknown protein 85.17 16.67 0.20 89.51 25.91 0.29 slr0709 Hypothetical protein 88.58 76.13 0.86 96.00 56.95 0.59 slr0868 Unknown protein 20 22 326 40 16 14 13 93 203 19 14 59 OD730 = 0.5 OD730 = 1.0 Average RPKM Ratio Average RPKM Ratio WT 1lexA 1lexA/WT WT 1lexA 1lexA/WT 10.39 23.76 2.29 6.55 84.74 12.95 113.13 1929.23 17.05 157.08 1310.83 8.35 47.79 859.88 17.99 48.45 595.60 12.29 71.82 419.75 5.84 75.08 336.96 4.49 116.90 563.08 4.82 97.11 410.52 4.23 114.46 528.48 4.62 117.45 444.43 3.78 10.17 49.14 4.83 7.21 42.81 5.94 677.95 37.76 0.06 535.50 31.01 0.06 84.80 38.09 0.45 86.35 68.05 0.79 133.72 577.77 4.32 106.12 532.70 5.02 27.90 188.61 6.76 23.33 243.59 10.44 29.09 170.08 5.85 16.73 124.75 7.46 611.42 3545.01 5.80 926.18 3621.41 3.91 109.22 83.03 0.76 73.68 44.88 0.61 30.41 472.85 15.55 27.60 136.10 4.93 88.93 685.38 7.71 41.49 494.27 11.91 23.31 240.78 10.33 18.06 205.72 11.39 26.35 199.16 7.56 14.51 181.65 12.52 137.85 61.68 0.45 194.48 104.16 0.54 76.42 37.70 0.49 48.30 36.91 0.76 59.93 13.16 0.22 54.18 13.94 0.26 97.90 48.95 0.50 61.97 49.50 0.80 57.32 302.45 5.28 46.76 183.74 3.93 232.67 121.52 0.52 18.94 215.48 11.37 45.49 351.04 7.72 48.25 355.40 7.37 18.55 130.15 7.01 13.41 45.98 3.43 11.63 67.76 5.83 7.87 24.45 3.11 136.11 100.98 0.74 15.16 103.21 6.81 24.83 103.47 4.17 18.85 108.19 5.74 136.06 1118.71 8.22 152.96 1173.63 7.67 68.61 42.42 0.62 60.44 37.54 0.62 40.10 327.60 8.17 51.00 312.12 6.12 58.01 324.48 5.59 74.96 308.21 4.11 11.14 405.28 36.37 19.47 345.32 17.74 38.61 187.90 4.87 14.43 111.92 7.75 18.01 103.04 5.72 20.82 119.39 5.73 48.52 240.11 4.95 55.46 204.71 3.69 5.58 25.68 4.61 5.73 12.25 2.14 17.69 35.91 2.03 7.26 38.90 5.36 175.06 105.14 0.60 207.58 121.90 0.59 350.05 18.16 0.05 194.88 16.91 0.09 18.80 3.65 0.19 22.65 6.20 0.27 79.35 334.01 4.21 60.26 159.45 2.65 85.17 16.67 0.20 89.51 25.91 0.29 88.58 76.13 0.86 96.00 56.95 0.59 20.22 326.40 16.14 13.93 203.19 14.59 (Continued) February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 7 RNA-seq of lexA in Synechocystis Kizawa et al. Motility-Related Genes TABLE 1 \| Continued Gene No. Gene symbol Deﬁnition OD730 = 0.5 OD730 = 1.0 Average RPKM Ratio Average RPKM Ratio WT 1lexA 1lexA/WT WT 1lexA 1lexA/WT slr0869 Hypothetical protein 23.29 185.20 7.95 24.87 165.85 6.67 slr0870 Hypothetical protein 31.36 196.83 6.28 14.82 110.90 7.48 slr0871 Unknown protein 12.74 102.10 8.01 5.49 63.27 11.53 slr1161 Hypothetical protein 306.59 134.80 0.44 251.76 84.91 0.34 slr1162 Unknown protein 131.90 66.56 0.50 104.40 63.19 0.61 slr1278 Hypothetical protein YCF62 32.54 27.87 0.86 78.79 41.89 0.53 slr1484 Unknown protein 48.58 129.12 2.66 29.37 273.35 9.31 slr1485 Salt-induced periplasmic protein 12.48 53.60 4.29 14.24 110.54 7.76 slr1704 Hypothetical protein 162.27 1747.93 10.77 179.40 617.74 3.44 slr1767 Hypothetical protein 39.67 197.11 4.97 19.06 96.15 5.04 slr1772 Probable hydrolase, periplasmic protein 49.33 225.97 4.58 49.08 245.26 5.00 slr1788 Unknown protein 33.57 388.00 11.56 65.92 359.18 5.45 slr1789 Unknown protein 16.30 152.34 9.34 29.57 152.66 5.16 slr1798 Unknown protein 155.94 109.76 0.70 186.54 116.12 0.62 slr1920 Unknown protein 69.46 571.56 8.23 58.59 561.87 9.59 slr2119 Unknown protein 60.74 16.37 0.27 41.99 14.00 0.33 ssl1046 Hypothetical protein 573.18 21.06 0.04 291.05 10.30 0.04 ssl1378 Hypothetical protein 69.57 33.34 0.48 194.48 104.16 0.54 ssl1577 Hypothetical protein 20.16 114.07 5.66 7.78 45.35 5.83 ssr0332 Hypothetical protein 218.38 154.04 0.71 120.96 73.51 0.61 ssr1155 Hypothetical protein 670.58 374.80 0.56 164.79 140.34 0.85 ssr1251 Hypothetical protein 52.57 15.34 0.29 6.04 2.22 0.37 ssr1473 Hypothetical protein 13.23 91.14 6.89 10.36 45.68 4.41 ssr2194 Unknown protein 14.79 614.16 41.52 8.50 181.59 21.37 ssr2615 Hypothetical protein 24.65 17.45 0.71 27.18 9.41 0.35 ssr2962 Hypothetical protein 63.09 276.40 4.38 41.60 191.31 4.60 ssr3570 Unknown protein 61.19 27.67 0.45 30.69 17.30 0.56 ssr3589 Hypothetical protein 19.15 115.47 6.03 9.11 61.37 6.73 Glucosylglycerol-Related Genes reporting that LexA acts as a transcriptional activator for the hox operon (Gutekunst et al., 2005). On the other hand, the expression level of the hypA1 gene involved in hydrogenase maturation increased in 1lexA. y g y In S.6803, glucosylglycerol (GG) is a major compatible solute to adapt to high-salt or high-osmotic pressure conditions (Klähn and Hagemann, 2011). A set of genes related to GG biosynthesis (ggp, glp) and uptake (ggt) are organized into several gene clusters such as ggtBCD (slr0529-0531), ggpS-glpD (sll1566-sll1085), ggpP-ggtA (slr0746-0747), and slr1670-glpK- spoU-slr1674-hypA1(slr1670-1675) in S.6803 genome (Mikkat and Hagemann, 2000; Klähn et al., 2010). Photosynthesis-Related Genes Photosynthesis Related Genes In S.6803, photosystem (PS) I complex is comprised of 11 subunits and genes encoding these subunits (psa) are dispersed throughout the genome (Kaneko et al., 1996). We found that the expression level of every PSI gene was lower in 1lexA than WT. The expression level of genes encoding subunits of phycobilisome (cpc, apc) was also lower in the mutant, whereas the expression level of genes encoding PSII subunits (psb) was not so much aﬀected by disruption of lexA. Expression levels of chlL-chlN encoding subunits of light-independent protochlorophyllide reductase and that of por encoding light- dependent protochlorophyllide reductase were lower in 1lexA. Both light-dependent and -independent enzymes catalyzing the last step of chlorophyll biosynthesis are likely to be under the control of LexA. On the other hand, expression level of hliA and Motility-Related Genes RNA-seq analysis revealed that expression levels of these gene clusters were signiﬁcantly higher in 1lexA than WT (Table 1 and Table S2). Klähn et al. (2010) reported that a small ORF, ggpR (ssl3076), exists overlapping with the transcription initiation site of ggpS and its promoter region. Expression of ggpR was also induced by disruption of lexA (Table 1). Effects of Disruption of the lexA Gene in S.6803 In this study, we created the gene-disrupted mutant of lexA in GT strain of S.6803 to obtain the comprehensive view of LexA regulon by RNA-seq analysis. Although Kamei et al. (2001) successfully obtained the fully-segregated lexA mutant from the motile PCC strain, in most cases the 1lexA mutant invariably retained the WT copy of the lexA gene (Domain et al., 2004; Gutekunst et al., 2005) and we also could not obtain fully-segregated mutant (Figure 1B). The heterogeneous appearance of the 1lexA mutant cells (Figure 1G) may be caused by diﬀerence in the extent of segregation. However, despite the existence of the WT copy of lexA, immunoblot analysis revealed that LexA protein level was below the detection limit in our mutant (Figure 1D). Hydogenase-Related Genes Expression level of the hoxE-hoxF-hoxU-hoxY-hoxH operon encoding subunits of bidirectional NiFe-hydrogenase was lower in 1lexA. This observation is consistent with the previous study February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 8 RNA-seq of lexA in Synechocystis Kizawa et al. FIGURE 2 \| MA plots of RNA-seq data obtained from WT and 1lexA cells at OD730 = 0.5 (A) and OD730 = 1.0 (B). The MA plot, a scatterplot of log2-fold-change (1lexA /WT) versus average expression in log2 scale for each gene, was produced using TCC package. Dots shown in magenta indicate differentially expressed genes with a false discovery rate < 0.01. FIGURE 2 \| MA plots of RNA-seq data obtained from WT and 1lexA cells at OD730 = 0.5 (A) and OD730 = 1.0 (B). The MA plot, a scatterplot of log2-fold-change (1lexA /WT) versus average expression in log2 scale for each gene, was produced using TCC package. Dots shown in magenta indicate differentially expressed genes with a false discovery rate < 0.01. hliB encoding high-light inducible proteins was higher in 1lexA than WT. transcribed operons, ggpS-glpD and slr1670-glpK-spoU-slr1674- hypA1, both of which are highly induced in 1lexA. His-LexA bound to the promoter fragment of each operon (for the ggpS operon from nucleotide 1949371 to 1949186 and for the slr1670 operon from nucleotide 1949332 to 1949534). It is notable that LexA-binding site for the the ggpS operon is within the coding region of ggpR (nucleotide 1949372 to 1949100). Our results suggest that LexA binds to at least two binding site located in the intergenic region of the ggpS and slr1670 operons to repress their expression. Next, we examined the binding of LexA to the upstream region of PSI genes by using light-responsive promoter fragments containing the HLR1 sequence recognized by the response regulator RpaB (Seino et al., 2009). Binding of His-LexA to the promoter region of PSI genes was not observed (Figure 3) or much weaker than that to the pilA7, pilA9, ggpS, and slr1670 promoters and not reproducible. This indicates that decrease in expression levels of PSI genes in 1lexA may be a secondary eﬀect. SOS-Response Related Genes Previous studies suggested that LexA in S.6803 is not involved in the SOS response. Neither lexA nor recA expression was induced upon UV-irradiation (Domain et al., 2004; Patterson-Fortin et al., 2006) and none of DNA metabolism-related genes was listed as genes induced or repressed by LexA depletion (Kamei et al., 2001; Domain et al., 2004). Similarly, induction or repression of DNA metabolism-related genes by disruption of lexA was not observed in our RNA-seq analysis. Genes Differentially Expressed in 1lexA at the Later Stage of Growth Several genes expressed under iron-limiting conditions such as exbB-exbD-fhuA operon involved in inorganic iron uptake (Jiang et al., 2015), futA1 and futA2 encoding subunits of iron transporter (Katoh et al., 2001), and isiA-isiB operon encoding iron-stress inducible proteins (Vinnemeier et al., 1998) were highly induced in 1lexA at OD730 = 1.0 but not in OD730 = 0.5. In the case of mntA and mntC encoding subunits of manganese transporter (Bartsevich and Pakrasi, 1995), their expression level was already higher in 1lexA at OD730 = 0.5 and showed further increase at OD730 = 1.0. Frontiers in Microbiology \| www.frontiersin.org DNA Gel Mobility Shift Assay DNA gel mobility shift assay was performed to examine whether LexA directly regulates expression of putative target genes listed by RNA-seq analysis (Figure 3). We observed induction of the pilA7-pilA8 operon and repression of the pilA9-pilA10-pilA11 operon in 1lexA (Table 1). Binding of His-LexA to the promoter regions of both operons (for the pilA7 operon from nucleotide 2222102 to 2222304 and for the pilA9 operon from nucleotide 755577 to 755778, according to numbering in CyanoBase) was observed, indicating that LexA directly activates or represses expression of these pilA operons. We also examined whether His-LexA binds to the upstream region of the two divergently To date, LexA in S.6803 has been reported to be involved in transcriptional regulation of genes related to various cellular functions. Our RNA-seq data are consistent with some of these Frontiers in Microbiology \| www.frontiersin.org February 2016 \| Volume 7 \| Article 193 9 RNA-seq of lexA in Synechocystis Kizawa et al. FIGURE 3 \| DNA gel mobility shift assay of the promoter segments of putative target genes with His-LexA. DIG-labeled promoter segments of pilA7, pilA9, ggpS, slr1670, and psaD were incubated for 25 min at room temperature with His-LexA added at indicated concentrations. ﬁve-fold and 50-fold excess amounts of the non-labeled promoter segments were added as a competitor. Samples were separated on a 6% polyacrylamide gel. FIGURE 3 \| DNA gel mobility shift assay of the promoter segments of putative target genes with His-LexA. DIG-labeled promoter segments of pilA7, pilA9, ggpS, slr1670, and psaD were incubated for 25 min at room temperature with His-LexA added at indicated concentrations. ﬁve-fold and 50-fold excess amounts of the non-labeled promoter segments were added as a competitor. Samples were separated on a 6% polyacrylamide gel. Furthermore, expression levels of several genes related to positive phototaxis and cAMP signaling were aﬀected. Although many non-motile mutants were so far isolated from the PCC strain, information on the mechanism of transcriptional regulation of motility-related genes is limited. Bhaya et al. (1999) reported decrease in expression level of pilA1 and pilA2 by disruption of the sigF gene encoding an alternative sigma factor. Yoshimura et al. (2002b) and Dienst et al. (2008) reported decrease in expression level of pilA9-pilA10-pilA11-slr2018 and cccS-cccP by disruption of sycrp1 encoding cAMP receptor protein and hfq encoding RNA chaperone homolog, respectively. Panichkin et al. DNA Gel Mobility Shift Assay (2006) reported decrease in expression level of pilA9-pilA10- pilA11-slr2018 and increase in that of pilA5-pilA6 and pilA1- pilA2 by disruption of spkA encoding Serine/threonine protein kinase. None of these reports showed the direct interaction of these regulatory factors with pilA genes and LexA in this study is the ﬁrst report of binding of transcriptional regulator to their upstream region (Figure 3). Involvement of SYCRP1 in transcriptional regulation of pilA genes through the direct regulation of LexA is not likely, since no SYCRP1 binding sequence has been detected in the upstream region of the lexA gene (Omagari et al., 2008; Xu and Su, 2009). Further examination of relationship between LexA and previously identiﬁed regulatory factors which aﬀect motility may be a key to understanding of signal transduction mechanism regulating phototactic motility. reports, e.g., positive regulation of the hox operon reported by Gutekunst et al. (2005) and positive regulation of the pilA genes reported by Kamei et al. (2001). However, we could not observe the large eﬀect of LexA depletion on carbon metabolism-related genes reported by Domain et al. (2004). Domain et al. isolated RNA for DNA microarray analysis from concentrated cultures incubated on plates for 2 h. The growth condition must be largely diﬀerent from our liquid culture, which may cause the diﬀerence in gene expression proﬁle. in vitro transcription/translation assay performed by Patterson-Fortin et al. (2006) showed that CrhR protein accumulation decreased in response to increasing LexA concentration. However, in our data, expression level of crhR was not aﬀected by disruption of lexA. RNA-seq data in this study suggested involvement of LexA in regulation of (1) phototactic motility, (2) accumulation of GG, (3) bidirectional hydrogenase, and (4) photosystem I and phycobilisome complexes. We also observed increase in expression level of genes related to iron and manganese uptake in 1lexA at OD730 = 1.0. LexA may be involved in stage speciﬁc repression of these genes, but it is also possible that these genes were upregulated as a consequence of iron and manganese limitation in the mutant culture during prolonged incubation. We will discuss regulation of cellular processes (1)–(4) by LexA in the following sections. Frontiers in Microbiology \| www.frontiersin.org Phototactic Motility In order to acclimate to high-salt or high-osmotic pressure conditions, S.6803 accumulates the compatible solute GG. Upon a salt shock, genes related to both GG biosynthesis (ggp, glp) and uptake (ggt) are induced (Kanesaki et al., 2002; Marin et al., 2004). GG is synthesized by a two-step reaction in S.6803. First, condensation of ADP-glucose and glycerol 3-phosphate is catalyzed by GG-phosphate synthase (GgpS) and then the intermediate is dephosphorylated by GG-phosphate phosphatase (GgpP) (Hagemann and Erdmann, 1994). Glycerol-3-phosphate Kamei et al. (2001) reported that disruption of the lexA gene in the motile PCC strain resulted in decrease in expression level of pilA genes and loss of thick pili and motility. Our RNA- seq analysis showed that expression levels of genes related to phototactic motility are largely aﬀected by disruption of lexA also in the non-motile strain. In addition to the decrease in expression level of pilA1 and pilA9-pilA10-pilA11 reported in Kamei et al. (2001), we observed signiﬁcant induction of pilA7-pilA8. February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 10 RNA-seq of lexA in Synechocystis Kizawa et al. dehydrogenase (GlpD) and glycerol kinase (GlpK) are involved in the metabolism of glycerol-3-phosphate, a precursor of GG. Uptake of GG from the environment is performed by ABC transporter consisting of an ATP-binding protein (GgtA), a substrate-binding protein (GgtB) and two integral membrane proteins (GgtC and GgtD) in S.6803 (Mikkat and Hagemann, 2000). All of these genes are induced by the disruption of lexA (Table 1 and Table S2). DNA gel mobility shift assay revealed that His-LexA protein binds to the upstream region of two divergently transcribed operons, ggpS-glpD and slr1670-glpK-spoU-slr1674- hypA1 (Figure 3). To date, sigma factors SigF (Marin et al., 2002) and SigB (Nikkinen et al., 2012), a small protein GgpR (Klähn et al., 2010) and a response regulator Slr1588 (Chen et al., 2014) were reported to be involved in transcriptional regulation of the ggpS-glpD operon. Our result suggests the existence of the additional regulatory mechanism, namely, repression of the divergent ggpS and slr1670 operons by LexA. Expression of ggpR may also be repressed by LexA, judging from the fact its expression was induced by the disruption of lexA (Table 1). Phototactic Motility Salt- stress inducible genes such as hliA, hliB, hspA, prqR, degP, sll1723, sll0846, sll1483, slr1704, sll1236, slr0581, and ssr2194, reported in the previous DNA microarray studies (Kanesaki et al., 2002; Marin et al., 2004), were also induced by the disruption of lexA (Table 1). There is possibility that LexA acts as a repressor for multiple salt-stress inducible genes as well as the ggpS and slr1670 operons. LexA in regulation of photosynthetic gene expression. However, clear and reproducible band shift was not observed when binding of His-LexA to the promoter regions of PSI genes was examined (Figure 3). It is possible that changes in expression levels of photosynthesis-related genes in 1lexA are not the consequence of loss of regulation by LexA but a secondary eﬀect. Search for LexA Binding Sites in the Target Promoters Our results of DNA gel mobility shift assay suggest that LexA binds to the upstream region of piA7, pilA9, ggpS and slr1670 to directly regulate their expression (Figure 3). To date, several nucleotide sequences for LexA binding site have been identiﬁed by DNA gel mobility shift assay, for example, 5′-TTTATTTGAACTATTTTT-3′, 5′-TTTTTCGTT GTCTAAATT-3′ (Oliveira and Lindblad, 2005), 5′-CTA-N9(AT- rich)-CTA-3′ (Patterson-Fortin and Owttrim, 2008), and 5′-AGT AACTAGTTCG-3′ (Gutekunst et al., 2005) in S.6803 and 5′- TAGTACTAATGTTCTA-3′ in A.7120. (Mazón et al., 2004). However, these LexA binding sequences could not be found in the promoter fragments to which His-LexA bound. Instead, we found that a 5′-TTTTG(A/T)TNAC-3′ sequence commonly exists in these promoter fragments (Figure S1). The sequence is located around the putative transcription start site in the case of the negatively-regulated target genes, ggpS, piA7, and slr1670, whereas it is located further upstream region in the case of the positively-regulated pilA9 gene. It has been reported that a certain global transcriptional regulator, such as NtcA and RpaB in S.6803, can act as both repressor and activator dependent on the location of the binding site (García-Domínguez et al., 2000; Seino et al., 2009). Binding of the transcriptional regulator causes repression when its binding site overlaps the RNA polymerase- binding site, whereas activating eﬀect is observed when the binding site is located further upstream. The location of 5′- TTTTG(A/T)TNAC-3′ sequence in four LexA-target promoters seems consistent with the scheme. Bidirectional Hydogenase Regulation of the hox operon by LexA has been extensively studied in S.6803 (Oliveira and Lindblad, 2009). LexA was shown to bind to two distinct regions of the hox promoter, −198 to −338 and −592 to −690, relative to the start codon of hoxE (Gutekunst et al., 2005; Oliveira and Lindblad, 2005) and work for positive regulation of hydrogenase activity (Gutekunst et al., 2005). Regulation of hydrogenase-related genes by LexA may be common among cyanobacterial species, judging from the reports on LexA homologs in Anabaena sp. PCC 7120 (Sjöholm et al., 2007) and Lyngbya majuscula CCAP 1446/4 (Ferreira et al., 2007). Frontiers in Microbiology \| www.frontiersin.org REFERENCES Appl. Microbiol. Biotechnol. 96, 183–196. doi: 10.1007/s00253-012- 4307-6 Arnon, D. I., McSwain, B. D., Tsujimoto, H. Y., and Wada, K. (1974). Photochemical activity and components of membrane preparations from blue- green algae. I. Coexistence of two photosystems in relation to chlorophyll a and removal of phycocyanin. Biochim. Biophys. Acta 357, 231–245. doi: 10.1016/0005-2728(74)90063-2 Dienst, D., Duhring, U., Mollenkopf, H.-J., Vogel, J., Golecki, J., Hess, W. R., et al. (2008). The cyanobacterial homologue of the RNA chaperone Hfq is essential for motility of Synechocystis sp. PCC 6803. Microbiology 154, 3134–3143. doi: 10.1099/mic.0.2008/020222-0 Domain, F., Houot, L., Chauvat, F., and Cassier-Chauvat, C. (2004). Function and regulation of the cyanobacterial genes lexA, recA and ruvB: LexA is critical to the survival of cells facing inorganic carbon starvation. Mol. Microbiol. 53, 65–80. doi: 10.1111/j.1365-2958.2004.04100.x Bartsevich, V. V., and Pakrasi, H. B. (1995). Molecular identiﬁcation of an ABC transporter complex for manganese: analysis of a cyanobacterial mutant strain impaired in the photosynthetic oxygen evolution process. EMBO J. 14, 1845–1853. 65–80. doi: 10.1111/j.1365-2958.2004.04100.x Erill, I., Campoy, S., and Barbé, J. (2007). Aeons of distress: an evolutionary perspective on the bacterial SOS response. FEMS Microbiol. Rev. 31, 637–656. doi: 10.1111/j.1574-6976.2007.00082.x Bhaya, D., Watanabe, N., Ogawa, T., and Grossman, A. R. (1999). The role of an alternative sigma factor in motility and pilus formation in the cyanobacterium Synechocystis sp. strain PCC6803. Proc. Natl. Acad. Sci. U.S.A. 96, 3188–3193. doi: 10.1073/pnas.96.6.3188 Ferreira, D., Leitão, E., Sjöholm, J., Oliveira, P., Lindblad, P., Moradas-Ferreira, P., et al. (2007). Transcription and regulation of the hydrogenase(s) accessory genes, hypFCDEAB, in the cyanobacterium Lyngbya majuscula CCAP 1446/4. Arch. Microbiol. 188, 609–617. doi: 10.1007/s00203-007-0281-2 Blot, N., Mella-Flores, D., Six, C., Le Corguille, G., Boutte, C., Peyrat, A., et al. (2011). Light history inﬂuences the response of the marine cyanobacterium Synechococcus sp. WH7803 to oxidative stress. Plant Physiol. 156, 1934–1954. doi: 10.1104/pp.111.174714 García-Domínguez, M., Reyes, J. C., and Florencio, F. J. (2000). NtcA represses transcription of gifA and gifB, genes that encode inhibitors of glutamine synthetase type I from Synechocystis sp. PCC 6803. Mol. Microbiol. 35, 1192–1201. doi: 10.1046/j.1365-2958.2000.01789.x Brahamsha, B., and Bhaya, D. (2014). “Motility in unicellular and ﬁlamentous Cyanobacteria,” in The Cell Biology of Cyanobacteria, eds E. Flores and A. Herrero (Dorset: Caister Academic Press), 233–262. Gutekunst, K., Phunpruch, S., Schwarz, C., Schuchardt, S., Schulz-Friedrich, R., and Appel, J. (2005). LexA regulates the bidirectional hydrogenase in the cyanobacterium Synechocystis sp. Photosystem I and Phycobilisome Then, what is the physiological meaning of the coordinated regulation of phototactic motility, GG accumulation, and hydogenase activity by LexA in S.6803? We searched for environmental conditions where LexA-target genes are coordinately regulated using CyanoEXpress gene expression database (http://cyanoexpress.sysbiolab.eu/) and found that salt stress causes induction of GG metabolism-related genes and repression of hox operon and pilA genes in WT (Shoumskaya et al., 2005; Dickson et al., 2012). The expression proﬁle is similar to that observed by disruption of lexA (Table 1), indicating the possibility that transcriptional regulation by LexA is temporarily inactivated under salt stress conditions. et al., 2004; Patterson-Fortin et al., 2006). Similarly, in Anabaena sp. PCC 7120, expression of lexA was not induced upon UV-B exposure or treatment with a DNA damaging agent mitomycin C (Kumar et al., 2015). In these freshwater species, LexA-independent protection mechanism for DNA damage may have evolved and LexA may have become devoted to regulating other cellular processes. Then, what is the physiological meaning of the coordinated regulation of phototactic motility, GG accumulation, and hydogenase activity by LexA in S.6803? We searched for environmental conditions where LexA-target genes are coordinately regulated using CyanoEXpress gene expression database (http://cyanoexpress.sysbiolab.eu/) and found that salt stress causes induction of GG metabolism-related genes and repression of hox operon and pilA genes in WT (Shoumskaya et al., 2005; Dickson et al., 2012). The expression proﬁle is similar to that observed by disruption of lexA (Table 1), indicating the possibility that transcriptional regulation by LexA is temporarily inactivated under salt stress conditions. use the transcriptional regulator LexA for better adaptation to changing environment. FUNDING This work was ﬁnancially supported by the Core Research of Evolutional Science & Technology (CREST) programs from the Japan Science and Technology Agency (JST). SUPPLEMENTARY MATERIAL Recently, it has been suggested that the SOS response in the marine Synechococcus is regulated by LexA like E. coli (Blot et al., 2011; Tetu et al., 2013). Cyanobacterial LexA genes can be clustered into three groups, Clade A containing Gloeobacter violaceus PCC 7421, Clade C containing marine picocyanobacteria and Clade B containing most remaining species (Li et al., 2010). There may exist high degree of variation of LexA regulons among species belonging to these three clades. By examination of what kind of cellular processes LexA regulates, we will be able to know decision of each species about how to The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.00193 Table S1 \| Oligonucleotides used in this study. Table S1 \| Oligonucleotides used in this study. Table S1 \| Oligonucleotides used in this study. Table S2 \| RNA-seq data of WT and 1lexA at OD730 = 0.5. Table S2 \| RNA-seq data of WT and 1lexA at OD730 = 0.5. Table S3 \| RNA-seq data of WT and 1lexA at OD730 = 1.0. Table S3 \| RNA-seq data of WT and 1lexA at OD730 = 1.0. Figure S1 \| Consensus sequences for LexA binding site in four LexA-target promoters identiﬁed with MEME. AUTHOR CONTRIBUTIONS The study was conceived by AYK and YH, with design input from AKK. Experiments were performed by AYK and AKK. Data analysis and interpretation was done by all authors. The manuscript was prepared by AYK and YH, and reviewed by all authors. p p y Data analysis and interpretation was done by all authors. The manuscript was prepared by AYK and YH, and reviewed by all authors. Photosystem I and Phycobilisome Photosystem I and Phycobilisome In the 1lexA mutant, chlorophyll and phycocyanin contents were lower than those in WT (Figure 1F). This may be caused by decreased expression level of genes encoding subunits of PSI (psa), subunits of phycobilisome (cpc, apc) and both light- dependent and -independent protochlorophyllide reductase (chlL, chlB, por). It is known that these photosynthesis-related genes show the quite similar response to the changing light environment (Muramatsu and Hihara, 2012). The response regulator RpaB regulates high-light response of photosynthesis- related genes by binding to their promoter regions under low-light conditions (Wilde and Hihara, 2016). PSI genes and hli genes are positively- and negatively-regulated target genes of RpaB, respectively (Kappell and van Waasbergen, 2007; Seki et al., 2007; Seino et al., 2009). Repression of PSI genes and induction of hli genes by disruption of LexA (Table 1) seem to suggest overlapping roles of RpaB and LexA Results of RNA-seq analysis (Table 1) together with DNA gel mobility shift assay (Figure 3) suggest LexA in S.6803 can positively or negatively regulate various cellular processes such as phototactic motility, GG accumulation and hydogenase activity. Regulation of such a wide range of cellular processes by LexA was reported in other bacterial species. For example, the lexA mutant of Clostridium diﬃcile showed pleiotrophic phenotypes such as ﬁlamentous structure due to inhibition of cell division, decreased sporulation, decrease in swimming motility and increased bioﬁlm formation (Walter et al., 2015). In this case, LexA acts as a regulator of DNA damage in addition to the above mentioned biological functions. In contrast, DNA microarray data from diﬀerent research groups (Kamei et al., 2001; Domain et al., 2004) and our RNA-seq data suggest LexA in S.6803 is not involved in regulation of SOS genes. In S.6803, expression of lexA and recA was not induced upon UV-irradiation (Domain February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 11 RNA-seq of lexA in Synechocystis Kizawa et al. et al., 2004; Patterson-Fortin et al., 2006). Similarly, in Anabaena sp. PCC 7120, expression of lexA was not induced upon UV-B exposure or treatment with a DNA damaging agent mitomycin C (Kumar et al., 2015). In these freshwater species, LexA-independent protection mechanism for DNA damage may have evolved and LexA may have become devoted to regulating other cellular processes. REFERENCES PCC 6803 as a transcription activator. Mol. Microbiol. 58, 810–823. doi: 10.1111/j.1365-2958.2005.04867.x Butala, M., Žgur-Bertok, D., and Busby, S. J. W. (2009). The bacterial LexA transcriptional repressor. Cell. Mol. Life Sci. 66, 82–93. doi: 10.1007/s00018- 008-8378-6 Chen, L., Wu, L., Zhu, Y., Song, Z., Wang, J., and Zhang, W. (2014). An orphan two-component response regulator Slr1588 involves salt tolerance by directly regulating synthesis of compatible solutes in photosynthetic Synechocystis sp. PCC 6803. Mol. BioSyst. 10, 1765–1774. doi: 10.1039/c4mb00095a Hagemann, M., and Erdmann, N. (1994). Activation and pathway of glucosylglycerol synthesis in the cyanobacterium Synechocystis sp. PCC 6803. Microbiology 140, 1427–1431. doi: 10.1099/00221287-140-6-1427 Ishii, A., and Hihara, Y. (2008). An AbrB-like transcriptional regulator, Sll0822, is essential for the activation of nitrogen-regulated genes in Synechocystis sp. PCC 6803. Plant Physiol. 148, 660–670. doi: 10.1104/pp.108.123505 Dickson, D. J., Luterra, M. D., and Ely, R. L. (2012). Transcriptomic responses of Synechocystis sp. PCC 6803 encapsulated in silica gel. February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 12 RNA-seq of lexA in Synechocystis Kizawa et al. Jiang, H.-B., Lou, W.-J., Ke, W.-T., Song, W.-Y., Price, N. M., and Qiu, B.-S. (2015). New insights into iron acquisition by cyanobacteria: an essential role for ExbB-ExbD complex in inorganic iron uptake. ISME J. 9, 297–309. doi: 10.1038/ismej.2014.123 in Synechocystis sp. PCC 6803. Planta 216, 446–453. doi: 10.1007/s00425-002- 0859-5 Muramatsu, M., and Hihara, Y. (2012). Acclimation to high-light conditions in cyanobacteria: from gene expression to physiological responses. J. Plant Res. 125, 11–39. doi: 10.1007/s10265-011-0454-6 Kamei, A., Hihara, Y., Yoshihara, S., Geng, X., Kanehisa, M., and Ikuechi, M. (2001). “Functional Analysis of lexA-like gene, sll1626 in Synechocystis sp. PCC 6803 using DNA microarray,” in PS2001 Proceedings of 12th International Congress on Photosynthesis (Melbourne, VIC: CSIRO Publishing), S41–013. Nikkinen, H.-L., Hakkila, K., Gunnelius, L., Huokko, T., Pollari, M., and Tyystjarvi, T. (2012). The SigB factor regulates multiple salt acclimation responses of the cyanobacterium Synechocystis sp. PCC 6803. Plant Physiol. 158, 514–523. doi: 10.1104/pp.111.190058 Kaneko, T., Sato, S., Kotani, H., Tanaka, A., Asamizu, E., Nakamura, Y., et al. (1996). Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions. DNA Res. 3, 109–136. doi: 10.1093/dnares/3.3.109 Oliveira, P., and Lindblad, P. (2005). LexA, a transcription regulator binding in the promoter region of the bidirectional hydrogenase in the cyanobacterium Synechocystis sp. PCC 6803. REFERENCES FEMS Microbiol. Lett. 251, 59–66. doi: 10.1016/j.femsle.2005.07.024 Oliveira, P., and Lindblad, P. (2009). Transcriptional regulation of the cyanobacterial bidirectional Hox-hydrogenase. Dalton Trans. 7, 9990–9996. doi: 10.1039/b908593a Kanesaki, Y., Shiwa, Y., Tajima, N., Suzuki, M., Watanabe, S., Sato, N., et al. (2012). Identiﬁcation of substrain-speciﬁc mutations by massively parallel whole- genome resequencing of Synechocystis sp. PCC 6803. DNA Res. 19, 67–79. doi: 10.1093/dnares/dsr042 Omagari, K., Yoshimura, H., Suzuki, T., Takano, M., Ohmori, M., and Sarai, A. (2008). Delta G-based prediction and experimental conﬁrmation of SYCRP1- binding sites on the Synechocystis genome. FEBS J. 275, 4786–4795. doi: 10.1111/j.1742-4658.2008.06618.x Kanesaki, Y., Suzuki, I., Allakhverdiev, S. I., Mikami, K., and Murata, N. (2002). Salt stress and hyperosmotic stress regulate the expression of diﬀerent sets of genes in Synechocystis sp. PCC 6803. Biochem. Biophys. Res. Commun. 290, 339–348. doi: 10.1006/bbrc.2001.6201 Panichkin, V. B., Arakawa-Kobayashi, S., Kanaseki, T., Suzuki, I., Los, D. A., Shestakov, S. V., et al. (2006). Serine/threonine protein kinase SpkA in Synechocystis sp. strain PCC 6803 is a regulator of expression of three putative pilA operons, formation of thick pili, and cell motility. J. Bacteriol. 188, 7696–7699. doi: 10.1128/JB.00838-06 Kappell, A. D., and van Waasbergen, L. G. (2007). The response regulator RpaB binds the high light regulatory 1 sequence upstream of the high-light-inducible hliB gene from the cyanobacterium Synechocystis PCC 6803. Arch. Microbiol. 187, 337–342. doi: 10.1007/s00203-007-0213-1 Patterson-Fortin, L. M., Colvin, K. R., and Owttrim, G. W. (2006). A LexA- related protein regulates redox-sensitive expression of the cyanobacterial RNA helicase, crhR. Nucleic Acids Res. 34, 3446–3454. doi: 10.1093/nar/ gkl426 Katoh, H., Hagino, N., and Grossman, A. R. (2001). Genes essential to iron transport in the cyanobacterium Synechocystis sp. strain PCC 6803. J. Bacteriol. 183, 2779–2784. doi: 10.1128/JB.183.9.2779-2784.2001 g Patterson-Fortin, L. M., and Owttrim, G. W. (2008). A Synechocystis LexA- orthologue binds direct repeats in target genes. FEBS Lett. 582, 2424–2430. doi: 10.1016/j.febslet.2008.06.009 Klähn, S., and Hagemann, M. (2011). Compatible solute biosynthesis in cyanobacteria. Environ. Microbiol. 13, 551–562. doi: 10.1111/j.1462- 2920.2010.02366.x Klähn, S., Höhne, A., Simon, E., and Hagemann, M. (2010). The gene ssl3076 encodes a protein mediating the salt-induced expression of ggpS for the biosynthesis of the compatible solute glucosylglycerol in Synechocystis sp. strain PCC 6803. J. Bacteriol. 192, 4403–4412. doi: 10.1128/JB.00481-10 Seino, Y., Takahashi, T., and Hihara, Y. (2009). REFERENCES The response regulator RpaB binds to the upstream element of photosystem I genes to work for positive regulation under low-light conditions in Synechocystis sp. strain PCC 6803. J. Bacteriol. 191, 1581–1586. doi: 10.1128/JB.01588-08 Seki, A., Hanaoka, M., Akimoto, Y., Masuda, S., Iwasaki, H., and Tanaka, K. (2007). Induction of a group 2 σ factor, RPOD3, by high light and the underlying mechanism in Synechococcus elongatus PCC 7942. J. Biol. Chem. 282, 36887–36894. doi: 10.1074/jbc.M707582200 Kumar, A., Kirti, A., and Rajaram, H. (2015). LexA protein of cyanobacterium Anabaena sp. strain PCC7120 exhibits in vitro pH-dependent and RecA- independent autoproteolytic activity. Int. J. Biochem. Cell Biol. 59, 84–93. doi: 10.1016/j.biocel.2014.12.003 Li, S., Xu, M., and Su, Z. (2010). Computational analysis of LexA regulons in cyanobacteria. BMC Genomics 11:527. doi: 10.1186/1471-2164-11-527 Shoumskaya, M. A., Paithoonrangsarid, K., Kanesaki, Y., Los, D. A., Zinchenko, V. V., Tanticharoen, M., et al. (2005). Identical Hik-Rre systems are involved in perception and transduction of salt signals and hyperosmotic signals but regulate the expression of individual genes to diﬀerent extents in Synechocystis. J. Biol. Chem. 280, 21531–21538. doi: 10.1074/jbc.M412174200 Lieman-Hurwitz, J., Haimovich, M., Shalev-Malul, G., Ishii, A., Hihara, Y., Gaathon, A., et al. (2009). A cyanobacterial AbrB-like protein aﬀects the apparent photosynthetic aﬃnity for CO2 by modulating low-CO2- induced gene expression. Environ. Microbiol. 11, 927–936. doi: 10.1111/j.1462- 2920.2008.01818.x J. Biol. Chem. 280, 21531–21538. doi: 10.1074/jbc.M412174200 Sjöholm, J., Oliveira, P., and Lindblad, P. (2007). Transcription and regulation of the bidirectional hydrogenase in the cyanobacterium Nostoc sp. strain PCC 7120. Appl. Environ. Microbiol. 73, 5435–5446. doi: 10.1128/AEM.00756-07 Marin, K., Huckauf, J., Fulda, S., and Hagemann, M. (2002). Salt-dependent expression of glucosylglycerol-phosphate synthase, involved in osmolyte synthesis in the cyanobacterium Synechocystis sp. strain PCC 6803. J. Bacteriol. 184, 2870–2877. doi: 10.1128/JB.184.11.2870-2877.2002 Sun, J., Nishiyama, T., Shimizu, K., and Kadota, K. (2013). TCC: an R package for comparing tag count data with robust normalization strategies. BMC Bioinformatics 14:219. doi: 10.1186/1471-2105-14-219 Marin, K., Kanesaki, Y., Los, D. A., Murata, N., Suzuki, I., and Hagemann, M. (2004). Gene expression proﬁling reﬂects physiological processes in salt acclimation of Synechocystis sp. strain PCC 6803. Plant Physiol. 136, 300–329. doi: 10.1104/pp.104.045047 Terauchi, K., and Ohmori, M. (1999). An adenylate cyclase, cya1, regulates cell motility in the cyanobacterium Synechocystis sp. PCC 6803. Plant Cell Physiol. 40, 248–251. doi: 10.1093/oxfordjournals.pcp.a029534 Tetu, S. G., Johnson, D. A., Varkey, D., Phillippy, K., Stuart, R. K., Dupont, C. L., et al. (2013). REFERENCES Impact of DNA damaging agents on genome-wide transcriptional proﬁles in two marine Synechococcus species. Front. Microbiol. 4:232. doi: 10.3389/fmicb.2013.00232 Mazón, G., Lucena, J. M., Campoy, S., Fernández de Henestrosa, A. R., Candau, P., and Barbé, J. (2004). LexA-binding sequences in Gram-positive and cyanobacteria are closely related. Mol. Genet. Genomics 271, 40–49. doi: 10.1007/s00438-003-0952-x Vinnemeier, J., Kunert, A., and Hagemann, M. (1998). Transcriptional analysis of the isiAB operon in salt-stressed cells of the cyanobacterium Synechocystis sp. PCC 6803. FEMS Microbiol. Lett. 169, 323–330. doi: 10.1111/j.1574- 6968.1998.tb13336.x Mikkat, S., and Hagemann, M. (2000). Molecular analysis of the ggtBCD gene cluster of Synechocystis sp. strain PCC6803 encoding subunits of an ABC transporter for osmoprotective compounds. Arch. Microbiol. 174, 273–282. doi: 10.1007/s002030000201 Walter, B. M., Cartman, S. T., Minton, N. P., Butala, M., and Rupnik, M. (2015). The SOS response master regulator LexA is associated with sporulation, motility and bioﬁlm formation in Clostridium diﬃcile. PLoS ONE 10:e0144763. doi: 10.1371/journal.pone.0144763 Mortazavi, A., Williams, B. A., McCue, K., Schaeﬀer, L., and Wold, B. (2008). Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat. Methods 5, 621–628. doi: 10.1038/nmeth.1226 Muramatsu, M., and Hihara, Y. (2003). Transcriptional regulation of genes encoding subunits of photosystem I during acclimation to high-light conditions Wang, H.-L., Postier, B. L., and Burnap, R. L. (2002). Polymerase chain reaction-based mutagenesis identify key transporters belonging to multigene February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 13 RNA-seq of lexA in Synechocystis Kizawa et al. families involved in Na+ and pH homeostasis of Synechocystis sp. PCC 6803. Mol. Microbiol. 44, 1493–1506. doi: 10.1046/j.1365-2958.2002. 02983.x Yoshimura, H., Yanagisawa, S., Kanehisa, M., and Ohmori, M. (2002b). Screening for the target gene of cyanobacterial cAMP receptor protein SYCRP1. Mol. Microbiol. 43, 843–853. doi: 10.1046/j.1365-2958.2002.02790.x Wang, H.-L., Postier, B. L., and Burnap, R. L. (2004). Alterations in global patterns of gene expression in Synechocystis sp. PCC 6803 in response to inorganic carbon limitation and the inactivation of ndhR, a LysR family regulator. J. Biol. Chem. 279, 5739–5751. doi: 10.1074/jbc.M311336200 Yoshimura, H., Yoshihara, S., Okamoto, S., Ikeuchi, M., and Ohmori, M. (2002a). A cAMP receptor protein, SYCRP1, is responsible for the cell motility of Synechocystis sp. PCC 6803. Plant Cell Physiol. 43, 460–463. doi: 10.1093/pcp/pcf050 y Chem. 279, 5739–5751. doi: 10.1074/jbc.M311336200 Zhao, S., Fung-Leung, W.-P., Bittner, A., Ngo, K., and Liu, X. (2014). Comparison of RNA-Seq and microarray in transcriptome proﬁling of activated T cells. Frontiers in Microbiology \| www.frontiersin.org REFERENCES PLoS ONE 9:e78644. doi: 10.1371/journal.pone.0078644 Wilde, A., and Hihara, Y. (2016). Transcriptional and posttranscriptional regulation of cyanobacterial photosynthesis. Biochim. Biophys. Acta 1857, 296–308. doi: 10.1016/j.bbabio.2015.11.002 Xu, M., and Su, Z. (2009). Computational prediction of cAMP receptor protein (CRP) binding sites in cyanobacterial genomes. BMC Genomics 10:23. doi: 10.1186/1471-2164-10-23 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. Yoshihara, S., Geng, X., Okamoto, S., Yura, K., Murata, T., Go, M., et al. (2001). Mutational analysis of genes involved in pilus structure, motility and transformation competency in the unicellular motile cyanobacterium Synechocystis sp. PCC 6803. Plant Cell Physiol. 42, 63–73. doi: 10.1093/pcp/pce007 Copyright © 2016 Kizawa, Kawahara, Takimura, Nishiyama and Hihara. 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. Yoshimura, H., Kaneko, Y., Ehira, S., Yoshihara, S., Ikeuchi, M., and Ohmori, M. (2010). CccS and CccP are involved in construction of cell surface components in the cyanobacterium Synechocystis sp. strain PCC 6803. Plant Cell Physiol. 51, 1163–1172. doi: 10.1093/pcp/pcq081 February 2016 \| Volume 7 \| Article 193 Frontiers in Microbiology \| www.frontiersin.org 14
https://openalex.org/W1481136666	https://repositorio.unesp.br/bitstream/11449/131421/1/PMC4494706.pdf	English	null	Differences in Transcriptional Activity of Human Papillomavirus Type 6 Molecular Variants in Recurrent Respiratory Papillomatosis	PloS one	2,015	cc-by	6,669	RESEARCH ARTICLE Differences in Transcriptional Activity of Human Papillomavirus Type 6 Molecular Variants in Recurrent Respiratory Papillomatosis Caroline Measso do Bonfim1, João Simão Sobrinho2, Rodrigo Lacerda Nogueira3, Daniel Salgado Kupper3, Fabiana Cardoso Pereira Valera3, Maurício Lacerda Nogueira4, Luisa Lina Villa2,5,6, Paula Rahal1, Laura Sichero2* 1 Laboratory of Genomic Studies, Universidade do Estado de São Paulo, UNESP, São José do Rio Preto, SP, Brazil, 2 Molecular Biology Laboratory, Center of Translational Oncology, Instituto do Câncer do Estado de São Paulo, ICESP, São Paulo, Brazil, 3 Department of Ophthalmology/Otorhinolaryngology and Head/ Neck Surgery, Discipline Otorhinolaryngology, Faculty of Medicine of Ribeirão Preto, Universidade de São Paulo, USP, São Paulo, Brazil, 4 Laboratory of Research in Virology, Faculty of Medicine of São José do Rio Preto, FAMERP, São José do Rio Preto, Brazil, 5 Department of Radiology and Oncology, School of Medicine, Universidade de São Paulo, USP, São Paulo, Brazil, 6 School of Medicine, Santa Casa de São Paulo and HPV Institute, São Paulo, Brazil OPEN ACCESS * lsichero@gmail.com Citation: Measso do Bonfim C, Simão Sobrinho J, Lacerda Nogueira R, Salgado Kupper D, Cardoso Pereira Valera F, Lacerda Nogueira M, et al. (2015) Differences in Transcriptional Activity of Human Papillomavirus Type 6 Molecular Variants in Recurrent Respiratory Papillomatosis. PLoS ONE 10(7): e0132325. doi:10.1371/journal.pone.0132325 * lsichero@gmail.com Introduction role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Recurrent Respiratory Papillomatosis (RRP) is characterized by the proliferation of multiple papillomas within the respiratory tract affecting especially the larynx [1,2]. RRP can affect per- sons of all ages; but a bimodal age distribution is often observed: disease peaks in children younger than 5 years and between the third and fourth decade of life. Thus, based on age distri- bution of affected individuals, RRP has been clinically divided into juvenile onset (JORRP) and adult onset RRP (AORRP) [3]. Human Papillomaviruses (HPV) types 6 and 11 are etiologically associated with RRP development [4]. Although RRP is considered a benign disease, high recurrence rates are associated to significant morbidity and is occasionally fatal [5]. Competing Interests: LLV is a consultant of Merck Sharp & Dohme for HPV vaccines. None of the other authors have conflicts of interest to report. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials. The prototype HPV-6b clone was originally isolated from a condyloma acuminatum speci- men [6]. Subsequently, HPV-6a and HPV-6vc non-prototypic genomes were identified by var- iations in restriction patterns, and gave rise to subtype classification [7]. Because the term subtype was redefined, HPV-6a, -6b, and -6vc were assigned to molecular variants since nucle- otide sequence among these isolates diverges less than 2% within the L1 gene [8]. Recently, the analysis of 190 complete viral genomes clustered HPV-6 isolates into two variants lineages named A and B with the latter enclosing 5 sublineages (differences in whole genome sequence of 0.5–1.0%) [9]. While the HPV-6b-reference isolate is one of the highly related members of lineage A, HPV-6vc and HPV-6a sort to B1 and B3 sublineages, respectively [10]. The viral long control region (LCR) comprises approximately 10% of the viral genome and encloses cis-regulatory elements for cellular and viral transcription factors (TFs) that modulate early gene expression and replication [11]. The repertoire of potential TFs binding sites within the LCR vary largely among distinct HPV types and variants [12,13] due to nucleotide diver- gences, and may impact binding affinity, transcriptional activity and ultimately the clinical out- come associated with HPV infections [14]. HPVs 16 and 18 intratype variability has been epidemiologically correlated to viral persis- tence and development of clinically relevant lesions [15–18]. Introduction Concerning the prevalence of low risk HPVs and the functional implications of viral heterogeneity, data is still scarce. In the pres- ent study, we thought to characterize the complete LCR of HPV-6 detected in a group of indi- viduals diagnosed with RRP and analyze nucleotide variability and age of disease development. Further, we investigated the impact of LCR nucleotide heterogeneity upon transcription. In sil- ico analysis revealed putative binding sites for GATA1, ELF1 and FOXA1 overlapping diver- gent nucleotide positions. We then analyzed the binding and influence of these TFs upon transcriptional activity of HPV-6 variants. Materials and Methods Clinical samples We analyzed 23 laryngeal papillomas biopsy specimens from 13 patients diagnosed with RRP from 2005 to 2010, and treated at the Laryngology Clinic of the School of Medicine, University of São Paulo, Ribeirão Preto, Brazil. Disease severity was accessed using the Derkay system at each surgical intervention [19]. This study was approved by the Institutional Ethics Research Committee of the Sao Paulo State University at Sao José do Rio Preto, São Paulo, and written informed consent was obtained from patients or parents of underage patients. Transcriptional Regulation of HPV-6 Variants Abstract A significant proportion of recurrent respiratory papillomatosis (RRP) is caused by human papillomavirus type 6 (HPV-6). The long control region (LCR) contains cis-elements for reg- ulation of transcription. Our aim was to characterize LCR HPV-6 variants in RRP cases, compare promoter activity of these isolates and search for cellular transcription factors (TFs) that could explain the differences observed. The complete LCR from 13 RRP was analyzed. Transcriptional activity of 5 variants was compared using luciferase assays. Dif- ferences in putative TFs binding sites among variants were revealed using the TRANSFAC database. Chromatin immunoprecipation (CHIP) and luciferase assays were used to evalu- ate TF binding and impact upon transcription, respectively. Juvenile-onset RRP cases har- bored exclusively HPV-6vc related variants, whereas among adult-onset cases HPV-6a variants were more prevalent. The HPV-6vc reference was more transcriptionally active than the HPV-6a reference. Active FOXA1, ELF1 and GATA1 binding sites overlap variable nucleotide positions among isolates and influenced LCR activity. Furthermore, our results support a crucial role for ELF1 on transcriptional downregulation. We identified TFs impli- cated in the regulation of HPV-6 early gene expression. Many of these factors are mutated in cancer or are putative cancer biomarkers, and must be further studied. Editor: Peter C. Angeletti, University of Nebraska- Lincoln, UNITED STATES Copyright: © 2015 Measso do Bonfim 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: All relevant data are within the paper. Novel HPV-6 sequences were submitted to GenBank (n° KF436496-KF436509). Funding: This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) (grant numbers 10/00029-0 to CMB and 08/57889-1 to LLV); Conselho Nacional de Desenvolvimento Científico e Tencnológico (CNPq) (grant number 573799/2008-3 to LLV); and Ludwig Institute for Cancer Research. The funders had no 1 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Validation Assays C33A cells were co-transfected in 96-well plates using Fugene (Promega) with 100ng of pGL3-LCR-Luc plasmids of different variants of HPV-6 together with increasing amounts of selected TFs expression plasmids (120, 240, 360ng), and 50ng of pCMV-β-Gal vector. Cells were harvested 48 hours after transfection by addition of 100μL of 1X Reporter Lysis Buffer (Promega). Relative luciferase activity was normalized to β-galactosidase measurements. Tran- scriptional activity was compared to basal activity of the same variant with no TFs superex- pressed. Averages and standard deviations are based on the mean of triplicates of at least twelve independent experiments. Cell culture C33A (ATCC HTB-31) and C33A cells permanently transfected with pGL3-HPV-6-LCR were grown in Dulbecco’s Modified Eagle medium supplemented with 10% fetal calf serum and antibiotics. Primary human foreskin keratinocytes (PHFK) (Clonetics), and PHFK infected with pLXSN-HPV-6b-E6/E7 or pLXSN-HPV-16-E6/E7 were maintained in keratinocyte serum free medium (KSFM), supplemented with 5ng/mL epidermal growth factor and 50μg/mL bovine pituitary extract (Invitrogen). Promoter activity assays The complete LCR from the different HPV-6 variants was cloned upstream the luciferase gene in the pGL3-Basic vector (Promega). Accurate construction of all plasmids was confirmed by sequencing. Approximately 24 hours before transfection, 4x106 C33A cells were plated in 10cm diameter dishes. Cells were co-transfected using Lipofectamine (Invitrogen) with 4μg of pGL3-LCR-Luc plasmids and 1μg of pCMV-βGal vector for internal control of transfection efficiency [23]. Extracts were obtained 48 hours after transfection by addition of 800μL of 1X Reporter Lysis Buffer (Promega). Luciferase activity was measured using the Promega Lucifer- ase Assay System kit (Promega) in a Victor Light Luminescence Counter (Perkin Elmer). The β-Galactosidase Enzyme Assay System (Promega) was used to determine β-Galactosidase activity at 420nm on a Benchmark plate reader (Bio-Rad). Relative luciferase activity measure- ments were normalized to β-Galactosidase activity and protein content measured using the Bradford assay (Bio-Rad). Averages were based on the mean of triplicates of three independent experiments. Sequence analysis DNA was extracted using the QIAamp DNA Micro kit (Qiagen Inc.). The complete HPV-6 LCR (nt 7292 to 101) was amplified in two independent reactions. Amplicons were cloned using the pCR XL-TOPO Vector (Invitrogen), and clones were purified using the GeneJET 2 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Transcriptional Regulation of HPV-6 Variants Plasmid Miniprep Kit (Fermentas Life Sciences). Sequencing reactions were performed in an ABI 3130XL sequencer (Applied Biosystems) using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). Variants were identified by alignment of sequences with the references of HPV-6b (GenBank n° X00203), HPV-6a (GenBank n° L41216) and HPV-6vc (GenBank n° AF092932) using CLUSTALW [20] enclosed in the BioEdit 7.0.9.0 package [21]. Sequences were submitted to GenBank (accession numbers: KF436496-KF436509). TFs bind- ing sites were predicted online (http://trap.molgen.mpg.de/cgi-bin/home.cgi) using the TRANSFAC database [22]. Statistical Analysis Fisher´s exact test was used to compare HPV-6 variant distribution between JORRP and AORRP cases. We used T-test to compare Derkay mean values with 95% confidence interval (CI) between HPV-6a and -6vc-related RRP cases. These analyses were conducted using the GraphPad Prism statistical package (version 5.01). Kruskal-Wallis and Tukey’s tests were 3 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Transcriptional Regulation of HPV-6 Variants employed to compare LCR activity among HPV-6 variants. For TFs validation assays, one-way analysis of variance and Tukey’s multiple comparison tests were performed using the SPSS sta- tistical package version 20.0. P< 0.05 was considered significant for all tests. Chromatin Immunoprecipitation (ChIP) ChIP assays were perfomed using the MAGnify Chromatin Immunoprecipitation System (Invitrogen). Briefly, C33A cells permanently transfected with pGL3-HPV-6-LCR from differ- ent variants were cross-linked with methanol-free formaldehyde (Polyscience) and further son- icated using a Sonic Dismembrator Model 100 (Fisher Scientific). Production of DNA fragments of 100-500bp was checked by gel electrophoresis in agarose 2%. Approximately 1x105 cells were incubated with 20μg of anti-GATA1 (Abcam), anti-ELF1 (Santa Cruz Biotech- nology), or anti-FOXA1 (Santa Cruz Biotechnology) overnight at 4°C with moderate agitation. Recovered DNA was subjected to PCR using specific primers surrounding putative TFs bind- ing sites. Western blotting Cells were washed with ice-cold phosphate-buffered saline (PBS), scraped, and centrifuged. Radio immunoprecipitation assay (RIPA) buffer (20mM Tris-HCl [pH 7.4], 150mM sodium chloride, 1mM ethylenediaminetetraacetic acid, 1mM ethylene glycol tetraacetic acid, 0.1% sodium dodecyl sulfate [SDS], 1% sodium deoxycholate, 1% Triton X-100, 1mM sodium ortho- vanadate) containing complete protease inhibitor cocktail (Roche) was used for protein extrac- tion. Eighty micrograms of each lysate were submitted to 12% SDS polyacrylamide gel electrophoresis, and further transferred to nylon membranes (Hybond). Incubations with anti- GATA1 (1:500), anti-ELF1 (1:200), anti-FOXA1 (1:500), or anti-tubulin (1:1.000) (Sigma) were conducted overnight in PBS-T (PBS, 0.05% Tween 20) at 4°C with 5% milk. We followed incubation with anti-HRP-conjugated antirabbit secondary antibody (1:5.000) (GE Healthcare) for 1 hour at room temperature. Proteins were visualized using the Amersham ECL Western Blotting Detection Reagents (GE Healthcare) in a ImageQuant LAS4.000 equipment (GE Healthcare), and quantified using the ImageQuant TL software (GE Healthcare). Transcriptional Regulation of HPV-6 Variants bility within the long control region (LCR) of human papillomavirus type 6 (HPV-6) molecular variants. Table 1. Nucleotide sequence variability within the long control region (LCR) of human papillomavirus type 6 (HPV-6) molecular variants. Number of samples HPV-6 LCRvariant 7320 7350 7520 7626 7631 7633 7681 7762 7875 7919 7978 16 2 6a-ref A G C T A A A C C A C G 1 6a-var1 A 1 6a-var2 G 6 6vc-ref T I1 T I2 G A C 1 6vc-var1 T I1 G T I2 G A C 1 6vc-var2 T I1 T I2 G A C T 1 6vc-var3 T I1 T I2 G G A C Genomic positions containing speciﬁc mutations are indicated vertically across the top. Genomic positions without mutations compared to the HPV-6a-ref sequence (), Insertions (I): I1 = TTATTGTATATCTTGTTACA; I2 = C nucleotide insertion. de sequence variability within the long control region (LCR) of human papillomavirus type 6 (HPV-6) molecu Table 1. Nucleotide sequence variability within the long control region (LCR) of human papillomavirus t Genomic positions containing speciﬁc mutations are indicated vertically across the top. Genomic positions without mutations compared to the HPV-6a-ref sequence (), Insertions (I): I1 = TTATTGTATATCTTGTTACA; I2 = C nucleotide insertion. Genomic positions containing speciﬁc mutations are indicated vertically across the top. Genomic positions without mutations compared to the HPV-6a-ref sequence (), Insertions (I): I1 = TTATTGTATATCTTGTTACA; I2 = C nucleotide insertion. tations are indicated vertically across the top. Genomic positions without mutations compared to the HPV-6a-ref TATATCTTGTTACA; I2 = C nucleotide insertion. Transcriptional activity of human pappilomavirus (HPV) type 6 variants Luciferase assays were used to access the impact of LCR nucleotide sequence heterogeneity on early viral transcription. Unfortunately, we were unable to construct a recombinant plasmid containing the HPV-6a-var2 LCR. We observed an 11.5 fold enhanced promoter activity for HPV-6vc-ref variant as compared to HPV-6a-ref. A transition at nucleotide position 16 detected solely in HPV-6a-var1 led to an increase in transcriptional activity similar to that of the HPV-6vc-ref, whereas the transversion at position 7630 inherent of the HPV-6vc-var1 abolished the increased activity observed for HPV-6-vc-ref. HPV-6vc-var2 and -6vc-var3 showed promoter activities similar to HPV-6vc-ref (Fig 1A) Putative cis-elements within HPV-6 LCR Search for putative TFs binding sites was restricted to nucleotide positions divergent among HPV-6a-ref, -6a-var1, -6vc-ref, -6vc-var1 because these isolates showed differences in tran- scription. This approach revealed that nucleotide heterogeneity of HPV-6 variants may impact on binding of different TFs including FOXA1, GATA1 and ELF1 (Fig 1B). Human papillomavirus (HPV) type 6 variant characterization and distribution We sequenced the LCR of 23 HPV-6 RRP specimens obtained from 13 patients. Individuals with more than one biopsy harbored identical HPV genomes. We detected 7 different HPV-6 variants (Table 1). Maximum genomic distance was 7 mutations between HPV-6a-ref and -6vc-ref (2 inser- tions and 5 substitutions). HPV-6a-var1, -6a-var2, -6vc-var1, -6vc-var2 and -6vc-var3 have unique substitutions at nucleotide positions 16 (G!A), 7320 (A!G), 7626 (T!G), 7978 (C!T) and 7681 (A!G), respectively. We detected HPV-6a-related sequences (HPV-6a-ref,- var1,-var2) in 32.5% (4/13) patients. Additionally, 67.5% (9/13) individuals contained HPV- 6vc-related sequences (HPV-6vc-ref,-var1,-var2,-var3), of which HPV-6vc-ref was the most prevalent. Age at RRP diagnosis ranged from newborn to 51 years old. All JORRP cases harbored HPV-6vc-related variants, whereas most AORRP specimens had HPV-6a-related variants (66.6%; 4/6) (p = 0.02) (Table 2). Average Derkay scores were similarly distributed in patients with different variants. 4 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 doi:10.1371/journal.pone.0132325.t002 FOXA1, ELF1 and GATA1 binding and influence upon HPV-6 LCR The substitution in nucleotide position 16 (G!A) detected in HPV-6a-var1 leads to the elimi- nation of putative binding sites for GATA1 and ELF1 as compared to the HPV-6a-ref (Fig 1B). We observed by transfecting increasing amounts of plasmids expressing GATA1 or ELF1 that both TFs significantly reduce HPV-6a-ref transcription, whereas have no significant impact Table 2. Clinical data of recurrent respiratory papillomatosis (RRP) patients harboring human papillo- mavirus type -6a and -6vc (HPV-6a and -6vc) related variants. HPV-6a-related HPV-6vc-related p-value JORRP n = 0 n = 7 P = 0.02a AORRP n = 4 n = 2 Mean Derkay score 6.4 (CI 95%: 4.8 to 7.9) 6.8 (CI 95%: 6.3 to 7.3) P = 0.46b Abbreviations: JORRP, juvenile onset recurrent respiratory papillomatosis; AORRP adult onset recurrent respiratory papillomatosis. a Fisher's exact test. b T t t Table 2. Clinical data of recurrent respiratory papillomatosis (RRP) patients harboring human papillo- mavirus type -6a and -6vc (HPV-6a and -6vc) related variants. HPV-6a-related HPV-6vc-related p-value JORRP n = 0 n = 7 P = 0.02a AORRP n = 4 n = 2 Mean Derkay score 6.4 (CI 95%: 4.8 to 7.9) 6.8 (CI 95%: 6.3 to 7.3) P = 0.46b Abbreviations: JORRP, juvenile onset recurrent respiratory papillomatosis; AORRP adult onset recurrent respiratory papillomatosis. a Fisher's exact test. b T-test. Table 2. Clinical data of recurrent respiratory papillomatosis (RRP) patients harboring human papillo- mavirus type -6a and -6vc (HPV-6a and -6vc) related variants. doi:10.1371/journal.pone.0132325.t002 5 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Transcriptional Regulation of HPV-6 Variants Fig 1. Molecular variants of HPV-6 differ in early promoter activity. (A) Transcriptional activity of human papillomavirus type (HPV-6) molecular variants isolated from recurrent respiratory papillomatosis (RRP) cases. Data are presented as mean ± SD relative values from 3 independent experiments. Transcriptional activity was normalized to that of HPV-6a-ref which was arbitrarily defined as the reference and set to value 1. Kruskal-Wallis and Tukey’s tests were used to compare LCR activity among the different HPV-6 molecular variants. Asterisks indicate isolates that showed statistically significant different transcriptional activities compared to the corresponding references (HPV-6a-ref or HPV-6vc-ref). (P<0.001). (B) Putative cellular transcription factors binding sites within the long control region (LCR) that differ among the different molecular variants of human papillomavirus (HPV) type 6 detected. Fig 1. Molecular variants of HPV-6 differ in early promoter activity. FOXA1, ELF1 and GATA1 binding and influence upon HPV-6 LCR (A) Transcriptional activity of human papillomavirus type (HPV-6) molecular variants isolated from recurrent respiratory papillomatosis (RRP) cases. Data are presented as mean ± SD relative values from 3 independent experiments. Transcriptional activity was normalized to that of HPV-6a-ref which was arbitrarily defined as the reference and set to value 1. Kruskal-Wallis and Tukey’s tests were used to compare LCR activity among the different HPV-6 molecular variants. Asterisks indicate isolates that showed statistically significant different transcriptional activities compared to the corresponding references (HPV-6a-ref or HPV-6vc-ref). (P<0.001). (B) Putative cellular transcription factors binding sites within the long control region (LCR) that differ among the different molecular variants of human papillomavirus (HPV) type 6 detected. doi:10.1371/journal.pone.0132325.g001 doi:10.1371/journal.pone.0132325.g001 upon HPV-6a-var1 LCR (Fig 2A). ChIP assays were performed to verify the functional activity of the cis-elements suggested in silico. PCR amplification of recovered DNA using primers sur- rounding nucleotide position 16 revealed the in vivo association of both ELF1 and GATA1 with the HPV-6a-ref LCR but not with the HPV-6a-var1 isolate suggesting a direct effect upon transcriptional activity (Fig 2B). Divergence in nucleotide position 7626 (T!G) between HPV-6vc-ref and HPV-6vc-var1 results in the replacement of a putative ELF1 for a FOXA1 binding site as suggested by compu- tational analysis (Fig 1B). ELF1 inhibited transcription activity of both isolates in a similar Fig 2. Binding and activity of ELF1 and GATA1 to HPV-6a-ref and HPV-6a-var1. Luciferase reporter assay for (A) ELF1 and (B) GATA1. (C) Amplification of LCR fragments following chromatin immunoprecipitation (ChIP) using primers surrounding nucleotide position 16 that differs among HPV-6a-ref and HPV-6a-var1 variants. Input-nonimmunoprecipated samples. doi:10.1371/journal.pone.0132325.g002 Fig 2. Binding and activity of ELF1 and GATA1 to HPV-6a-ref and HPV-6a-var1. Luciferase reporter assay for (A) ELF1 and (B) GATA1. (C) Amplification of LCR fragments following chromatin immunoprecipitation (ChIP) using primers surrounding nucleotide position 16 that differs among HPV-6a-ref and HPV-6a-var1 variants. Input-nonimmunoprecipated samples. doi:10.1371/journal.pone.0132325.g002 6 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Transcriptional Regulation of HPV-6 Variants Fig 3. Binding and activity of ELF1 and FOXA1 to HPV-6vc-ref and HPV-6vc-var1. Luciferase reporter assay for (A) ELF1 and (B) FOXA1. (C) Amplification of LCR fragments following chromatin immunoprecipitation (ChIP) using primers surrounding nucleotide position 7626 that differs among HPV-vc- ref and HPV-6vc-var1 variants. Input-nonimmunoprecipated samples. doi 10 1371/jo rnal pone 0132325 g003 Fig 3. Binding and activity of ELF1 and FOXA1 to HPV-6vc-ref and HPV-6vc-var1. Discussion RRP is considered a non-malignant disease; however, in more severe clinical cases lesions spread throughout the lower respiratory tract, affecting bronchi, trachea, esophagus and lung [24]. HPV-6 and -11 DNA are detected in more than 90% of the cases [4]. While HPV-6 is more prevalent in RRP individuals, HPV-11 infection has been linked to a more aggressive clinical outcome [25]. Besides, co-infection with different variants of HPV-6 and -11 is very rarely detected [25, 26] and it has been suggested that RRP is the consequence of persistent infection with the initial viral genomic variant [27]. HPV-6 variants distribution is not as geo- graphically restricted as has been reported for HPV-16 and -18 variants [9,28]. However, a pos- sible lesion-specific preference of some HPV-6 variants was suggested due to the association of HPV-6vc-related isolates with anogenital infections [9]. We evaluated the complete HPV-6 LCR sequence of 13 laryngeal biopsies from RRP patients, and describe for the first time 5 novel HPV-6 LCR variants: HPV-6a-var1, -6a-var2, -6vc-var1, -6vc-var2 and -6vc-var3. In this series, we observed uneven distribution of HPV-6 variants: HPV-6vc and HPV-6a-related variants were more prevalent in JORRP and AORRP cases, respectively. It has been reported that pregnant women with evident anogenital warts or recent HPV infection at the time of delivery are more prone of having children affected by JORRP. Thus, the predominance of HPV-6vc-related variants in JORRP in this series could reflect the predominance of these genotypes in anogenital warts and anal cancer samples, as previously described [9,29,30]. HPV-6vc-related variants were also frequently detected in laryngeal papillomas from Australia [29], Slovenia [31] and South Africa [32]. However, com- parison among studies is hindered by the lack of clinical information and restriction of analyses to pediatric cases in these reports. Even though some severe RRP cases are observed in adults, JORRP is a more aggressive condition being associated with increased risk of lesion develop- ment in the lower respiratory tract [33]. Furthermore, there is a relation of age and surgical procedures, and in the case of JORRP, due to the small airway of children, multiple surgeries are required to avoid airway obstruction [33]. Although the regularity and number of surgical interventions in an individual is considered when estimating Derkay values, average scores were not differently distributed in individuals harboring different molecular variants of HPV-6 in the series of cases we analyzed. FOXA1, ELF1 and GATA1 binding and influence upon HPV-6 LCR Luciferase reporter assay for (A) ELF1 and (B) FOXA1. (C) Amplification of LCR fragments following chromatin immunoprecipitation (ChIP) using primers surrounding nucleotide position 7626 that differs among HPV-vc- ref and HPV-6vc-var1 variants. Input-nonimmunoprecipated samples. doi:10.1371/journal.pone.0132325.g003 manner (Fig 2A). Although FOXA1 significantly transactivated HPV-6vc-ref, enhancement of transcription was much stronger for the HPV-6vc-var1 isolate. Interestingly, both FOXA1 and ELF1 were shown to bind solely to the HPV-6vc-var1 LCR; no binding of both protein to the HPV-6vc-ref was detected by ChIP (Fig 3B). Nucleotide differences in positions 7631/33 leads to a substitution of a putative CEBP for an ELF1 binding site in HPV-6vc-ref as compared to HPV-6a-ref (Fig 1B). Additionally, alteration in nucleotide position 7762 determines the replacement of a putative HNF4a for a GATA1 binding site in HPV-6vc-ref as compared to HPV-6a-ref. ELF1 significantly repressed both HPV-6a-ref and HPV-6vc-ref (Figs 2A and 3A). However, ELF1 binding to HPV-6vc-ref LCR was not observed. In this case, ELF1 is most probably bound to the the ELF1 binding site at nucleotide position 7626 within the HPV-6a-ref LCR. Furthermore, we observed that GATA1 represses transcription of both HPV-6a-ref and HPV-6vc-ref (Figs 2A and 4A), although repression was more pronounced for the HPV-6vc-ref. GATA1 was shown to bind the LCR of both variants using ChIP (Fig 4B). We observed that ELF1, GATA1 and FOXA1 are detected Fig 4. Binding and activity of ELF1, GATA1 and FOXA1 to HPV-6a-ref and HPV-6vc-ref. (A) Luciferase reporter assay for GATA1. Amplification of LCR fragments following chromatin immunoprecipitation (ChIP) using primers surrounding nucleotide position (B) 7631/33 and (C) 7762 and that differ between HPV-6a-ref and HPV-6vc-var1 variants. Input-nonimmunoprecipated samples. doi:10.1371/journal.pone.0132325.g004 Fig 4. Binding and activity of ELF1, GATA1 and FOXA1 to HPV-6a-ref and HPV-6vc-ref. (A) Luciferase reporter assay for GATA1. Amplification of LCR fragments following chromatin immunoprecipitation (ChIP) using primers surrounding nucleotide position (B) 7631/33 and (C) 7762 and that differ between HPV-6a-ref and HPV-6vc-var1 variants. Input-nonimmunoprecipated samples. doi:10.1371/journal.pone.0132325.g004 doi:10.1371/journal.pone.0132325.g004 7 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Transcriptional Regulation of HPV-6 Variants not only in PHFK, but also in keratinocytes transducing E6/E7 from both HPV-16 and HPV- 6b (S1 Fig). not only in PHFK, but also in keratinocytes transducing E6/E7 from both HPV-16 and HPV- 6b (S1 Fig). PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Discussion The lack of clinical correlates of disease requires the study of larger series. Nevertheless, it is of note that in addition to age, other host factors such as genetic and immunological profiles appear to contribute to the greater aggressiveness of RRP [34,35]. LCR activity is crucial for viral replication, transcription and host cell proliferation mediated by these viruses [11]. Some studies described the impact of high-risk HPV-16 and -18 LCR var- iability upon viral early transcriptional activity, which may support augmented E6 and E7 lev- els and finally confer enhanced oncogenic potential to specific variants [14,36,37]. LCR activity is crucial for viral replication, transcription and host cell proliferation mediated by these viruses [11]. Some studies described the impact of high-risk HPV-16 and -18 LCR var- iability upon viral early transcriptional activity, which may support augmented E6 and E7 lev- els and finally confer enhanced oncogenic potential to specific variants [14,36,37]. Furthermore, LCR heterogeneity inherent of HPV-16 variants has been shown to influence expression of viral E2 and E1 replicating proteins and affect viral replication efficiency [38]. Concerning low risk viral types, a duplication of the HPV-11 early viral promoter sequence was associated to a higher degree of disease severity [39]. Additionally, increased LCR activity of HPV-11 variants was shown to associate with clinically RRP aggressiveness (30 to 33 epi- sodes) [26]. Thus, one may hypothesize that infection with more replicative HPV variants could result in more efficient proliferation of papillomas and thus respond for the augmented frequency of surgical processes required to control disease and avoid respiratory hitch. Furthermore, LCR heterogeneity inherent of HPV-16 variants has been shown to influence expression of viral E2 and E1 replicating proteins and affect viral replication efficiency [38]. Concerning low risk viral types, a duplication of the HPV-11 early viral promoter sequence was associated to a higher degree of disease severity [39]. Additionally, increased LCR activity of HPV-11 variants was shown to associate with clinically RRP aggressiveness (30 to 33 epi- sodes) [26]. Thus, one may hypothesize that infection with more replicative HPV variants could result in more efficient proliferation of papillomas and thus respond for the augmented frequency of surgical processes required to control disease and avoid respiratory hitch. PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 8 / 12 Transcriptional Regulation of HPV-6 Variants Some differences were observed in transcriptional activity among the HPV-6 molecular var- iants analyzed. Discussion In silico analysis revealed that some nucleotide position in which substitutions were detected overlap putative TF binding sites. Further, cis-acting elements were created or abolished due to nucleotide heterogeneity. Even so, we may not discard the indirect influence of some cellular TFs upon HPV transcription. The substitution in nucleotide position 16 of HPV-6a-var1 isolate in the background of the HPV-6a-ref sequence eliminates putative GATA1 and ELF1 binding sites. GATA1 is one of the members of GATA TFs family and has been linked to cancer development associated to chromosome 21 trisomy [40]. Further, GATA1 is a master regulator of erythroid cell development [41]. ELF1 is mainly expressed in lymphoid cells and is essential for the regulation of hematopoiesis and angiogenesis during development [42]. Once CHIP assays revealed the in vivo association of both TFs exclusively to the HPV-6a-var1 LCR, we believe that GATA1 and ELF1 induced inhibition of HPV-6a-ref transcription is determinant for the diminished transcriptional activity of this isolate as com- pared to the HPV-6a-var1. HPV-6vc-var1 LCR was approximately 11 times less active than the HPV-6vc-ref. The sin- gle nucleotide alteration that differentiates these LCR leads to loss and creation of a putative ELF1 and FOXA1, respectively. FOXA1 is a pioneer factor which binding to promoters and enhancers sequences permits the access of other specific TFs to the chromatin [43]. FOXA1 has recently been shown by us to strongly enhance HPV-16 and -18 transcriptions [14]. Among the TFs tested by us, FOXA1 was the only protein that enhanced HPV-6 transcription. Even though FOXA1 was shown to activate much more HPV-6vc-var1 as compared to HPV- 6vc-ref, and ELF1 repressed both variants transcriptional activity similarly, the binding of ELF1 solely to the HPV-6vc-var1 may explain the reduced transcriptional activity observed for this isolate. Particular changes inherent of HPV-6vc-var2 and -6vc-var3 did not influence the LCR activity. Although computational analysis indicated a putative GATA1 binding site exclu- sively in HPV-6a-ref as compared to HPV-6vc-ref, we observed that this TF was able to inter- act in vivo and decrease transcriptional activity of both variants. For this reason, we hypothesize that ELF1 binding and repression of viral transcription is crucial for the lower transcriptional activity inherent of HPV-6a-ref. Altogether, we described five novel HPV-6 LCR variants. We report uneven distribution of HPV-6 variants in JORRP and AORRP; HPV-6vc-related variants were the only isolates detected among JORRP cases. Supporting Information S1 Fig. Transcription factors levels in primary human foreskin keratinocytes. (A) ELF1, (B) GATA1 and (C) ELF1 levels in primary human foreskin keratinocytes (PHFK), and PHFK infected with pLXSN-HPV-6b-E6/E7 or pLXSN-HPV-16-E6/E7. One representative experi- ment from two is shown. (TIFF) Discussion Overall, HPV-6vc-related variants were the more transcription- ally active, and we further identified cellular TFs capable of influencing transcriptional activity of the different HPV-6 isolates. Our results support a crucial role of ELF1 on transcriptional downregulation in the differences observed. Further studies might shed light on these findings. References 1. Derkay CS, Wiatrak B. Recurrent Respiratory Papillomatosis: A Review. Laryng 2008; 118(7):1236– 47. 2. Kashima H, Mounts P, Leventhal B, Hruban RH. Sites of predilection in recurrent respiratory papilloma- tosis. Ann Otol Rhinol Laryngol 1993 Aug; 102(8 Pt 1):580–3. PMID: 8394667 3. Larson DA, Derkay CS. Epidemiology of recurrent respiratory papillomatosis. APMIS 2010 Jun; 118 (6–7):450–4. doi: 10.1111/j.1600-0463.2010.02619.x PMID: 20553527 4. Donne AJ, Clarke R. Recurrent respiratory papillomatosis: an uncommon but potentially devastating effect of human papillomavirus in children. Int J STD AIDS 2010 Jun; 21(6):381–5. doi: 10.1258/ijsa. 2010.010073 PMID: 20606216 5. Bishai D, Kashima H, Shah K. The cost of juvenile-onset recurrent respiratory papillomatosis. Arch Oto- laryngol Head Neck Surg 2000 Aug; 126(8):935–9. PMID: 10922223 6. de Villiers EM, Gissmann L, zur Hausen H. Molecular cloning of viral DNA from human genital warts. J Virol 1981 Dec; 40(3): 932–935. PMID: 6275126 7. Rando RF, Groff DE, Chirikjian JG, Lancaster WD. Isolation and characterization of a novel human pap- illoma- virus type 6 dna from an invasive vulvar carcinoma. J Virol 1986 Jan; 57(1):353–6. PMID: 3001357 8. Heinzel PA, Chan SY, Ho L, O'Connor M, Balaram P, Campo MS, et al. Variation of human papillomavi- rus type 6 (HPV-6) and HPV-11 genomes sampled throughout the world. J Clin Microbiol 1995 Jul; 33 (7):1746–54. PMID: 7665641 9. Jelen MM, Chen Z, Kocjan BJ, Burt FJ, Chan PKS, Chouhy D, et al. Global genomic diversity of human papillomavirus 6 based on 724 isolates and 190 complete genome sequences. J Virol 2014 July; 88 (13):7307–7316. doi: 10.1128/JVI.00621-14 PMID: 24741079 10. Burk RD, Chen Z, Harari A, Smith BC, Kocjan BJ, Maver PJ, et al. Classification and nomenclature sys- tem for human Alphapapillomavirus variants: general features, nucleotide landmarks and assignment of HPV6 and HPV11 isolates to variant lineages. Acta Dermatovenerol Alp Pannonica Adriat 2011 Sep; 20(3):113–123. PMID: 22131111 11. Bernard HU. Gene expression of genital human papillomaviruses and considerations on potential anti- viral approaches. Antivir Ther 2002 Dec; 7(4):219–37. PMID: 12553476 12. García-Vallvé S, Iglesias-Rozas JR, Alonso Á, Bravo IG. Different papillomaviruses have different rep- ertoires of transcription factor binding sites: convergence and divergence in the upstream regulatory region. BMC Evol Biol 2006; 6:20. PMID: 16526953 13. Romanczuk H, Villa LL, Schlegel R, Howley PM. Author Contributions Conceived and designed the experiments: CMdB LLV PR LS. Performed the experiments: CMdB JSS LS. Analyzed the data: CMdB JSS RLN DSK FCPV MLN LS. Contributed reagents/ materials/analysis tools: CMdB JSS RLN DSK FCPV MLN LLV PR LS. Wrote the paper: CMdB LLV PR LS. Acknowledgments We thank E Boccardo and M Morale (University of São Paulo, Brazil) for kind gift of primary human foreskin keratinocytes (PHFK) infected with pLXSN-HPV-6b-E6/E7. We thank KR Kao (Memorial University of Newfoundland) for providing pCS2+-Elf-1 expression vector, PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 9 / 12 Transcriptional Regulation of HPV-6 Variants and S Ezoe (Osaka University Graduate School of Medicine) for providing pcDNA3-GATA1 recombinant plasmid. We thank José Antônio Cordeiro from the Universidade do Estado de São Paulo—UNESP, São Jose do Rio Preto, SP, Brazil for statistical analysis." PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 References PMID: 2495524 24. Donne AJ, Hampson L, Homer JJ, Hampson IN. The role of HPV type in Recurrent Respiratory Papillo- matosis. Int J Pediatr Otorhinolaryngol 2010 Jan; 74 (1):7–14. doi: 10.1016/j.ijporl.2009.09.004 PMID: 19800138 25. Komlos KF, Kocjan BJ, Kosorok P, Luzar B, Meglič L, Potočnik M, et al. Tumor-specific and gender- specific pre-vaccination distribution of human papillomavirus types 6 and 11 in anogenital warts and laryngeal papillomas: a study on 574 tissue specimens. J Med Virol 2012 Aug; 84 (8): 1233–41. doi: 10.1002/jmv.23318 PMID: 22711351 26. Gall T, Kis A, Feher E, Gergely L, Szarka K. Virological failure of intralesional cidofovir therapy in recur- rent respiratory papillomatosis is not associated with genetic or epigenetic changes of HPV11: com- plete genome comparison of sequential isolates. Antiviral Res 2011 Nov; 92(2): 356–358. doi: 10. 1016/j.antiviral.2011.09.007 PMID: 21945249 27. Kocjan BJ, Gale N, Hocevar Boltezar I, Seme K, Komloš KF, Hošnjak L, et al. Identical Human Papillo- mavirus (HPV) Genomic Variants Persist in Recurrent Respiratory Papillomatosis for up to 22 Years. J Infect Dis 2013; 207 (4): 583–587. doi: 10.1093/infdis/jis733 PMID: 23204170 28. de Matos RPA, Sichero L, Mansur IM, Bonfim CM, Bittar C, Nogueira RL, et al. Nucleotide and phyloge- netic analysis of human papillomavirus types 6 and 11 isolated from recurrent respiratory papillomato- sis in Brazil. Infect Genet Evol 2013; 16: 282–9. doi: 10.1016/j.meegid.2012.12.033 PMID: 23466889 29. Danielewski JA, Garland SM, McCloskey J, Hillman RJ, Tabrizi SN. Human Papillomavirus Type 6 and 11 Genetic Variants Found in 71 Oral and Anogenital Epithelial Samples from Australia. PLoS One 2013 May; 8 (5):638–92. 30. Obalek S, Jablonska S, Favre M, Walczak L, Orth G. Condylomata acuminata in children: Frequent association with papillomaviruses responsible for cutaneous warts. J Am Acad Dermatol 1990 Aug; 23 (2 Pt 1):205–13. PMID: 2170467 31. Kocjan BJ, Poljak M, Cimerman M, Gale N, Potocnik M, Bogovac Z, Seme K. Prevaccination genomic diversity of human papillomavirus genotype 6 (HPV 6). Virol 2009 Sep; 391 (2):274–283. 32. Combrinck CE, Seedat RY, Randall C, Roodt Y, Burt FJ. Novel HPV-6 variants of human papillomavi- rus causing recurrent respiratory papillomatosis in southern Africa. Epidemiol Infect 2012 Jun; 140 (6):1095–101. doi: 10.1017/S0950268811001580 PMID: 21875449 33. Buchinsky FJ, Donfack J, Derkay CS, Choi SS, Conley SF, Myer CM, et al. Age of child, more than HPV type, is associated with clinical course in recurrent respiratory papillomatosis. PLoS One 2008 May; 3(5):2263. 34. References The viral transcriptional regulatory region upstream of the E6 and E7 genes is a major determinant of the differential immortalization activities of human papil- lomavirus types 16 and 18. J Virol 1991 May; 65 (5):2739–44. PMID: 1850048 14. Sichero L, Sobrinho JS, Villa LL. Identification of novel cellular transcription factors that regulate early promoters of Human Papillomavirus Types 18 and 16. J Infect Dis 2012 Sep; 206(6):867–74. doi: 10. 1093/infdis/jis430 PMID: 22740717 15. Sichero L, Ferreira S, Trottier H, Duarte-Franco E, Ferenczy A, Franco EL, et al. High grade cervical lesions are caused preferentially by non-European variants of HPVs 16 and 18. Int J Cancer 2007 Apr; 120 (8):1763–8. PMID: 17230525 16. Lizano M, De la Cruz-Hernández E, Carrillo-García A, García-Carrancá A, Ponce de Leon-Rosales S, Dueñas-González A, et al. Distribution of HPV16 and 18 intratypic variants in normal cytology, intrae- pithelial lesions, and cervical cancer in a Mexican population. Gynecol Oncol 2006 Aug; 102 (2):230– 5. PMID: 16427686 10 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Transcriptional Regulation of HPV-6 Variants 17. Berumen J, Ordoñez RM, Lazcano E, Salmeron J, Galvan SC, Estrada RA, et al. Asian-American vari- ants of human papillomavirus 16 and risk for cervical cancer: a case-control study. J Natl Cancer Inst 2001; 93(17): 1325–1330. PMID: 11535707 18. Hildesheim A, Schiffman M, Bromley C, Wacholder S, Herrero R, Rodriguez AC, et al. Human papillo- mavirus type 16 variants and risk of cervical cancer. J Natl Cancer Inst 2001; 93(4): 315–318. PMID: 11181779 19. Derkay CS, Malis DJ, Zalzal G, Wiatrak BJ, Kashima HK, Coltrera MD. A staging system for assessing severity of disease and response to therapy in recurrent respiratory papillomatosis. Laryngoscope 1998 June; 108 (6):935–37. PMID: 9628513 20. Thompson DI, French RM. The use of computerized risk adjustment tools in clinical process improve- ment. Healthc Inf Manage 1994; 8 (4):45–52. PMID: 10141373 21. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Win- dows 95/98/NT. Nucleic Acids Symposium Series 1999; 41:95–98. 22. Thomas-Chollier M, Hufton A, Heinig M, O'Keeffe S, Masri NE, Roider HG, et al. Transcription factor binding predictions using TRAP for the analysis of ChIP-seq data and regulatory SNPs. Nature Proto- cols 2011 Nov; 6(12):1860–9. doi: 10.1038/nprot.2011.409 PMID: 22051799 23. McGregor GR, Caskey CT. Construction of plasmids that express E. coli β-galactosidase in mamma- lian cells. Nucleic Acids Res 1989; 17:2365–70. 1989 Mar; 17(6): 2365. PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 References Lucs AV, Wu R, Mullooly V, Abramson AL, Steinberg BM. Constitutive overexpression of the oncogene Rac1 in the airway of recurrent respiratory papillomatosis patients is a targetable host-susceptibility fac- tor. Mol Med 2012 Mar; 18:244–9. doi: 10.2119/molmed.2011.00447 PMID: 22113496 35. Gelder CM, Williams OM, Hart KW, Wall S, Williams G, Ingramset D, et al. HLA class II polymorphisms and susceptibility to recurrent respiratory papillomatosis. J Virol 2003 Feb; 77(3): 1927–1939. PMID: 12525627 36. Sichero L, Franco EL, Villa LL. Different P105 promotor activities among natural variants of human pap- illomavirus type 18. J Infect Dis 2005; 191 (5): 739–742. PMID: 15688288 11 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0132325 July 7, 2015 Transcriptional Regulation of HPV-6 Variants 37. Kämmer C, Warthorst U, Torrez-Martinez N, Wheeler CM, Pfister H, et al. Sequence analysis of the long control region of human papillomavirus type 16 variants and functional consequences for P97 pro- moter activity. J Gen Virol 2000 Aug; 81 (8):1975–1981. 38. Hubert WG. Variant upstream regulatory region sequences differentially regulate human papillomavirus type 16 DNA replication throughout the viral life cycle. J Virol 2005 May; 79(10): 5914–5922. PMID: 15857977 39. Yuan H, Myers S, Wang J, Dan Zhou, Woo JA, Kallakury B, et al. Use of reprogrammed cells to identify therapy for respiratory papillomatosis. N Engl J Med 2012 Sep; 367 (13):1220–122. PMID: 23013073 40. Rainis L, Bercovich D, Strehl S, Teigler-Schlegel A, Stark B, Trka J, et al. Mutations in exon 2 of GATA1 are early events in megakaryocytic malignancies associated with trisomy 21. Blood 2003 Aug; 102 (3): 981–986. PMID: 12649131 41. Katsumura KR, DeVilbiss AW, Pope NJ, Johnson KD, Bresnick EH. Transcriptional mechanisms underlying hemoglobin synthesis. Cold Spring Harb Perspect Med 2013 Sep; 3(9):a01541. 42. Gallant S, Gilkeson G. ETS transcription factors and regulation of immunity. Arch Immunol Ther Exp 2006 May-Jun; 54(3):149–6. 43. Friedman JR, Kaestner KH. The FoxA family of transcription factors in development and metabolism. Cell Mol Life Sci 2006 Oct; 63(19–20):2317–28. PMID: 16909212 12 / 12
https://openalex.org/W4281390499	https://www.frontiersin.org/articles/10.3389/fnut.2022.896552/pdf	English	null	Black Soldier Fly Larvae Meal in the Diet of Gilthead Sea Bream: Effect on Chemical and Microbiological Quality of Filets	Frontiers in nutrition	2,022	cc-by	14,802	Black Soldier Fly Larvae Meal in the Diet of Gilthead Sea Bream: Effect on Chemical and Microbiological Quality of Filets Marianna Oteri 1, Biagina Chiofalo 1, Giulia Maricchiolo 2, Giovanni Toscano 3, Luca Nalbone 1, Vittorio Lo Presti 1 and Ambra Rita Di Rosa 1* 1 Department of Veterinary Sciences, University of Messina, Messina, Italy, 2 Institute of Biological Resources and Marine Biotechnologies, National Research Council, Messina, Italy, 3 Department of Chemical, Biological, Pharmaceutical and Environmental Sciences, University of Messina, Messina, Italy The chemical and microbiological characteristics of ﬁlets of Spaurus aurata L. specimens fed with diets containing a Hermetia illucens meal (HIM) at the 25, 35, and 50%, as a partial replacement for ﬁsh meal (FM) were evaluated. The diets, formulated to satisfy the nutritional needs of ﬁsh, were isoenergetic (22 MJ/kg gross energy), isonitrogenous (43 g/100 g, a.f.), and isolipidic (19 g/100 g, a.f.). Seventy-two specimens were randomly killed after 186 days of growing trials. Then, the ﬁlets were analyzed for chemical proﬁle, fatty acids, amino acids, minerals, and microbial ﬂora. Data were subjected to statistical analysis. No signiﬁcant differences were observed in chemical composition. The sum of polyunsaturated fatty acids (PUFAs) showed a similar content in the ﬁlets; eicosapentaenoic acid was similar in the ﬁlets of HIM0, HIM35%, and HIM50%, whereas docosahexaenoic acid was higher in ﬁlets of the HIM0 group. n3/n6 PUFA ratio and the sum of EPA + DHA showed a high value (p < 0.001) in ﬁlets of the group fed with FM. No signiﬁcant difference was observed in thrombogenic index and hypocholesterolaemic/hypercholesterolaemic ratio in the groups; the atherogenic index showed a higher value (p = 0.001) in the HIM50% group. Indispensable amino acids showed some signiﬁcant (p < 0.0001) differences in the groups; arginine and phenylalanine content was higher in the ﬁlets of ﬁsh fed with FM; isoleucine and valine content was higher in the ﬁlets of HIM50%; leucine, lysine and methionine content was lower in the ﬁlets of HIM35%; histidine content was lower in the ﬁlets of HIM25%; tryptophan content was lower in ﬁlets of the HIM50% group. EAA/NEAA ratio showed highest value in the ﬁlets of the group that received FM. The presence of HIM in the three diets kept chromium, manganese, iron, copper, zinc, and nickel levels lower than those recommended by various authorities. Ca/P ratio showed a higher level (p < 0.0001) in the group fed with FM than those fed with diets containing HIM. The insect meal in the diets did not inﬂuence the microbiological proﬁle of ﬁsh. Black Soldier Fly Larvae Meal in the Diet of Gilthead Sea Bream: Effect on Chemical and Microbiological Quality of Filets Use of HIM as an unconventional feed ingredient in Sparus aurata diet looks promising, although the quality of ﬁlets may be affected. ORIGINAL RESEARCH published: 24 May 2022 doi: 10.3389/fnut.2022.896552 ORIGINAL RESEARCH published: 24 May 2022 doi: 10.3389/fnut.2022.896552 Edited by: Fatih Ozogul, Çukurova University, Turkey Edited by: Fatih Ozogul, Çukurova University, Turkey Reviewed by: Maryam El Bakali, Abdelmalek Essaadi University, Morocco Semra Çiçek, Atatürk University, Turkey Correspondence: Ambra Rita Di Rosa dirosaa@unime.it Reviewed by: Maryam El Bakali, Abdelmalek Essaadi University, Morocco Semra Çiçek, Atatürk University, Turkey Reviewed by: Maryam El Bakali, Abdelmalek Essaadi University, Morocco Semra Çiçek, Atatürk University, Turkey Correspondence: Ambra Rita Di Rosa dirosaa@unime.it Specialty section: This article was submitted to Nutrition and Sustainable Diets, a section of the journal Frontiers in Nutrition Specialty section: This article was submitted to Nutrition and Sustainable Diets, a section of the journal Frontiers in Nutrition Received: 15 March 2022 Accepted: 19 April 2022 Published: 24 May 2022 Keywords: nutritional quality, mineral proﬁle, microbiological quality, seafood, insect meal INTRODUCTION (27, 28). However, it may be possible to modify the nutritional composition of H. illucens through the feeding media of the insect (29–31), as observed for H. illucens-fed ﬁsh oﬀal and algae where signiﬁcant amounts of EPA and DHA were found (27, 32). The global demand and consumption of ﬁsh are increasing to meet the needs of the growing population at a faster rate than the demand for ﬁsh feed ingredients; this is leading to a rapid decline in ﬁsh meal (FM) availability and simultaneous rise in prices (1, 2). FM is the optimal animal protein source used in commercial ﬁsh feeds (3), with high bioavailability of nutrients and an adequate nutritional composition, which meets the requirements of essential amino acids and fatty acids of ﬁsh species (4, 5). This study is part of a much larger research project, “Feed Insects For Aquaculture” (FIFA), which aims to reveal the nutritional value of a protein-rich insect meal (IM) produced from H. illucens larvae and used as a partial substitute of ﬁsh meal in Sparus aurata feeding. In a previous study, the proximate, fatty acid, amino acid, and mineral compositions, microbiological proﬁle, and organoleptic characteristics of four diets formulated for Sparus aurata and containing 25, 35, and 50% of defatted HIM (33) as replacement for FM were characterized. Given the growing interest in HIM as an alternative protein source to replace FM in ﬁsh feeds, in another study on Sparus aurata fed with the diets described above, the growth performance and feed utilization eﬃciency were reported (not yet published), and the organoleptic properties of the ﬁlets were analyzed using a sensor- based instrument platform consisting of E-eye, E-nose with 18 MOS sensors, and a potentiometric E-tongue with 7 chemical sensors (34). This study mainly focussed on the chemical and microbiological characteristics of ﬁlets of Spaurus aurata L. fed with diets containing increasing levels of HIM as a partial replacement for FM. However, the use of FM is unsustainable from both environmental and economic points of view. The aquaculture industry’s most positive contribution is the search for alternative ingredients that are integrated into sustainable farming systems and provide high quality protein and lipids without negatively impacting ﬁsh health, performance, and disease resistance (6, 7), and without compromising the nutritional value of farmed ﬁsh for humans (8, 9). Citation: Oteri M, Chiofalo B, Maricchiolo G, Toscano G, Nalbone L, Lo Presti V and Di Rosa AR (2022) Black Soldier Fly Larvae Meal in the Diet of Gilthead Sea Bream: Effect on Chemical and Microbiological Quality of Filets. Front. Nutr. 9:896552. doi: 10.3389/fnut.2022.896552 May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org Sparus aurata Fed Insect Meal Oteri et al. INTRODUCTION The use of nonconventional ingredients such as insect materials with nutritional and nutraceutical potential for human and animal health has been proposed as a relevant sustainable element of the agri-food chain (10–12). Insect meal (IM) is considered an adequate protein and lipid source that can be used as a substitute for FM in aquaculture feed because of its protein, amino acid, and fatty acid proﬁles (13). However, the use of insects as feed is a relatively new practice on a commercial scale, and many questions remain to be tackled: (i) the risk of transfer of pathogens into the production system (14) so much so that EFSA identiﬁes the substrate used to feed insects as the key entry point for contamination (15); (ii) the optimal level of food replacement of FM for IM, which can vary considerably from 25 to 100% because of diﬀerent compositions of larvae, ﬁsh species, and diets (16). MATERIALS AND METHODS The experimental protocol was designed according to the Italian legislation (35) and guidelines of the current European Directive (36) on the protection of animals utilized for scientiﬁc purposes. The experimental protocol was authorized by the Italian Ministry of Health (Ministerial Authorization number 491/2019-PR released on 4 July 2019). Among diﬀerent insect species considered for possible use in aquaculture, Hermetia illucens is one of the most interesting because its sustainability is linked to its abilityto convert food waste materials or manure into high-quality insect nutrients (17). The proximate composition of the H. illucens meal (HIM) is highly variable; based on dry matter, protein and lipid contents were reported for de-fatted HIM of 47.2 and 11.8% (18) and 51.8 and 14.8% (19), respectively, and for full-fat HIM of 36.2 and 18%, respectively (20). Ash content ranged between 11 and 15% with high mineral concentrations characterized by high Ca/P ratio (21). Frontiers in Nutrition \| www.frontiersin.org Diet Formulation All the diets were developed to meet the nutritional requirements of Sparus aurata and beisoenergetic (about 22 MJ/kg gross energy), isonitrogenous (about 43 g/100, as fed), and isolipidic (about 19 g/100, as fed). A control formula (HIM0) containing ﬁsh meal (FM) as an exclusive protein source of animal origin was developed. For the other three diets (HIM25, HIM35, HIM50%), the defatted Hermetia illucens meal was added at 25, 35, and 50% (as fed basis) to the control formula, replacing FM, to create three formulations characterized by diﬀerent amounts of HIM (79, 110, and 157 g/kg). The other ingredients of the diets were adjusted to obtain iso-energetic formulas. Although studies on the use of black soldier ﬂy larvae meal in ﬁsh ﬁnishing started in 1987 (22), it became popular again, especially after 2017 when the European Commission allowed the use of proteins derived from 7 species of insects as alternative protein sources for aquafeed formulation (23). Previous studies have shown that replacing FM with IM in ﬁsh feeds results in changes in ﬁlet quality (24), without adversely aﬀecting ﬁsh growth (25, 26). In particular, one problem for both producers and consumers is the consequent decrease in alfa-linolenic acid (ALA; C18:3n3), eicosapentaenoic acid (EPA; C20:5n3), and docosahexaenoic acid (DHA; C22:6n3) levels in ﬁsh ﬁlets due to the inclusion of HIM in the feed The diets were prepared by SPAROS Lda (Olhao, Portugal), which was commissioned to prepare the extruded ﬁsh diets. The ingredients were weighed, mixed, and grounded, and the feeds were extruded in a single screw extruder using a die diameter (dd) of 4 mm; after extrusion, kibbles were dried and coated with oil. The ingredients and proximate composition of the diets (HIM0, HIM25, HIM35, and HIM50%) are reported in Table 1. The fatty acid, amino acid, and mineral compositions, and microbiological proﬁle were previously reported by Oteri et al. (33). May 2022 \| Volume 9 \| Article 896552 2 Sparus aurata Fed Insect Meal Oteri et al. TABLE 1 \| Ingredients and proximate composition of the experimental diets. Diet Formulation DIET HIM0 HIM25% HIM35% HIM50% Ingredients, % as fed Fish meal 25.00 18.75 16.25 12.50 Hermetiaillucens meal 0 7.90 11.00 15.70 Soy protein concentrate 5.00 5.00 5.00 5.00 Wheat gluten 5.00 5.00 5.00 5.00 Corn gluten 5.00 5.00 5.00 5.00 Soybean meal 48 15.00 15.00 15.00 15.00 Rapeseed meal 5.00 5.00 5.00 5.00 Wheat meal 17.45 15.17 14.21 12.88 Whole peas 4.00 4.00 4.00 4.00 Fish oil 5.00 5.00 5.00 5.00 Rapeseed oil 10.00 9.80 9.80 9.80 Vitamin and mineral premix 1.00 1.00 1.00 1.00 Vitamin C35 0.03 0.03 0.03 0.03 Vitamin E50 0.02 0.02 0.02 0.02 Antioxidant 0.30 0.30 0.30 0.30 Sodium propionate 0.10 0.10 0.10 0.10 MCP, monocalcium phosphate 1.50 2.20 2.50 2.80 L-Lysine 0.30 0.35 0.37 0.40 L-Tryptophan - 0.03 0.04 0.05 DL-Methionine 0.10 0.15 0.18 0.22 L-Taurine 0.20 0.20 0.20 0.20 Proximate analysis, g/100g as fed Dry matter 92.33 92.78 92.90 92.64 Crude protein 42.7 42.7 42.7 42.7 Crude fat 18.6 18.6 18.6 18.7 Crude ﬁber 2.3 2.2 2.2 2.1 Ash 9.3 9.3 9.4 9.3 NFE* 19.43 19.98 20.00 19.84 HIM0, ﬁsh meal; HIM25, HIM35, and HIM50%, Hermetia illucens meal at the 25, 35, and 50% substitution rates of ﬁsh meal, respectively. *Nitrogen-free extract, NFE (g/100g) = 100–(crude protein + crude fat + crude ﬁber + ash). TABLE 1 \| Ingredients and proximate composition of the experimental diets. using a professional multi-parametric probe (YSI Professional Plus Multi-Parameters Water Quality Meter probe; Xylem Inc., Yellow Springs, OH, United States). The ﬁsh were fed with the commercial diet and adapted for a further 7 days to the experimental conditions. Twice a day (at 9:00 and 16:00 h) and 6 days a week, or for over 180 days (18 February−24th August), the ﬁsh were fed with the experimental feeds (HIM0, HIM25, HIM35, and HIM50%), initially at 0.8% and with an increase of up to 1.5% of body weight depending on water temperature. Throughout the growth trial period, tank biomass was weighed in bulk every 20 days to update daily feed ration. The tanks were inspected daily for mortality, which was, throughout the duration of the experiment,.003%. At the end of the trial, all the ﬁsh were starved for 24 h and killed by overdose (500 mg/L) with an anesthetic (tricaine methanesulfonate solution, MS-222; Sigma-Aldrich, Italy); body weight (390 ± 49, on average) and length (28 ± 49 cm on average) were determined individually for all the ﬁsh. Diet Formulation A subsample of 72 specimens (n = 18 ﬁsh per diet and 6 ﬁsh/tank) was randomly slaughtered and transported, in dry ice to the Laboratories of the Department of Veterinary Sciences-Unit of Animal Production, University of Messina (Messina, Italy), where they were gutted, ﬁleted, deskinned, sampled in small aliquots, vacuum-packed, and freeze-dried prior to being subjected to scheduled analyses. Then, each aliquot of the ﬁlets was defrosted and homogenized with a common laboratory knife mill (Grindomix GM 200; Retsch GmbH, Haan, Germany) for the analyses described in section 2.3. A total of 36 specimens (n = 9 ﬁsh per diet and 3 ﬁsh/tank) were randomly slaughtered and transported in sterile plastic bags in dry ice to the Laboratories of the Department of Veterinary Sciences–Unit of Inspection of Food of Animal Origin, University of Messina (Messina, Italy) and processed within 3 h. Analyses of Chemical Composition of Fish Muscle The proximate composition of the ground ﬁlets from the four groups of ﬁsh (total number = 72; 18 ﬁsh per diet and 6 ﬁsh per replication) was determined following the AOAC (37) methods for moisture (method 950.46), crude protein (method 981.10), and ash (method 920.153). Feeding Trial and Facilities Frontiers in Nutrition \| www.frontiersin.org Feeding Trial and Facilities The three indices were calculated as follows: performic acid was performed for deamination prior to acid hydrolysis with an HCl solution (6M). For tryptophan analysis, acid hydrolysis was performed using 10 ml of a NaOH solution (4M) at 112◦C for 16 h; then, cooling and neutralization of each sample were performed with acetic acid (44). Amino acids were analyzed with a Trace 1310 chromatograph (Thermo Fisher Scientiﬁc, Milan, Italy) provided with a ﬂame ionization detector (FID) and a ZB-AAA Amino Acid column (10 m × 0.25 mm ID); oven temperature was programmed from 110 to 320◦C at 32 ◦C/min, with a ﬁnal isotherm of 320 ◦C (1 min). Injector and detector temperatures were 250 and 320◦C, respectively. Injection volume and split ratio were 2.5 µl and 1:15, respectively. Procedures for puriﬁcation, pre-column derivatization, and quali-quantitative analyses of each amino acid were performed using EZ:Faast Kit (Phenomenex, Torrance, CA, United States). performic acid was performed for deamination prior to acid hydrolysis with an HCl solution (6M). For tryptophan analysis, acid hydrolysis was performed using 10 ml of a NaOH solution (4M) at 112◦C for 16 h; then, cooling and neutralization of each sample were performed with acetic acid (44). Amino acids were analyzed with a Trace 1310 chromatograph (Thermo Fisher Scientiﬁc, Milan, Italy) provided with a ﬂame ionization detector (FID) and a ZB-AAA Amino Acid column (10 m × 0.25 mm ID); oven temperature was programmed from 110 to 320◦C at 32 ◦C/min, with a ﬁnal isotherm of 320 ◦C (1 min). Injector and detector temperatures were 250 and 320◦C, respectively. Injection volume and split ratio were 2.5 µl and 1:15, respectively. Procedures for puriﬁcation, pre-column derivatization, and quali-quantitative analyses of each amino acid were performed using EZ:Faast Kit (Phenomenex, Torrance, CA, United States). IA = [C12 : 0 + (4 × C14 : 0) + C16 : 0]/( X n6 − PUFA + X n3 −PUFA + X MUFA) (1) The mineral proﬁle of ﬁsh samples, deprived of bones and scales, was analyzed with a high-performance dispersing instrument. About 0.5 g of the sample ﬁlets were exactly weighed in a pre-cleaned PTFE vessel by acidic wash and then digested with 7 ml of 69% Nitric acid TraceSelect and 1 ml of H2O2 at 30% (OptimaTM for Ultra Trace Analysis), both purchased from Honeywell Fluka (Seelze, Germany). Feeding Trial and Facilities The closed vessels were introduced into a microwave digestion system (Ethos 1; Milestone, Bergamo, Italy) and treated with a warm-up program of 20 min at 1,000 W of microwave power. After the cooling time, the digested samples were diluted to a ﬁnal volume of 25 ml with ultrapure water (resistivity 18.2 MΩ/cm) obtained from a Milli-Q Integral 3 device (Merck Millipore, Merck KGaA, Darmstadt, Germany). Samples of Mussel Tissue (CE278k) and Bovine Muscle (BB184), both from ERM (European Reference Material, Geel, Belgium), were used to verify the accuracy of the analytical procedures described above. All the solutions were stored in high-density PE bottles cleaned with a 10% (v/v) solution of HNO3, and were sonicated and rinsed afterward with ultrapure water. For the analysis of minerals, an ICP-OES instrument, Avio200, equipped with a vertical DualView optical system coupled with an S10 autosampler was used. Lengths of the analytical lines (nm) utilized for the analyses are reported in Table 2; the Argon line at 420.069 nm was applied as an internal standard. Table 3 reports the operational conditions of the ICP- OES. Data acquisition and processing were performed using the PerkinElmer SyngistixTM for ICP software (Perkin Elmer, Waltham, MA, United States). IT = (C14 : 0 + C16 : 0 + C18 : 0)/[(.5 × X MUFA) + (.5× X n6 −PUFA) + (3 × X n3 −PUFA) + ( X n3 −−PUFA/ X n6 −PUFA)] (2) (2) H/H = (C18 : 1n9 + C18 : 2n6 + C20 : 4n6 + C18 : 3n3 + C20 : 5n3 + C22 : 5n3 + C22 : 6n3)/(C14 : 0 + C16 : 0) (3) (3) Moreover, the peroxidation index (PI), that expresses a measure of peroxidation susceptibility and peroxidative lipid damage for a particular phospholipid membrane, was calculated using the following formula reported by Luciano et al. (42): PI = (%dienoic×1) + (%trienoic × 2) + (4) (%tetraenoic × 3) + (%pentaenoic ×4) + (%hexaenoic ×5) (4) For amino acid proﬁle, aliquots (approximately.25 g) of wet ﬁlet muscles from the four ﬁsh groups (n = 72) were hydrolyzed in 10 ml of an HCl solution (6M) at 110◦C for 24 h. During acid hydrolysis, asparagine and glutamine were converted to aspartic and glutamic acids (43); therefore, they were calculated as the sum of aspartic acid plus asparagine and of the glutamic acid plus glutamine. Feeding Trial and Facilities (Omegawax 250; Supelco, Bellefonte, PA, United States) 25-µm ﬁlm, and maintained at 100◦C for 5 min, from 100 to 240◦C at 4 ◦C/min and a ﬁnal isotherm of 20 min at 240◦C. Injector and detector temperatures were set at 250◦C. Injection volume and split ratio were 0.5 µl and 1:50, respectively. The carrier gas, helium (He), was set at a ﬂow rate of 1 ml/min. Data acquisition and processing were performed using ChromeleonTM Software (Thermo Fisher Scientiﬁc, Milan, Italy). Fatty acids of the ﬁsh samples were identiﬁed by comparing the relative retention times of FAMEs with those of a standard mix solution (mix 37 FAMEs; Supelco, Inc., Bellefonte, PA, United States) under the same analytical conditions. FA concentrations were expressed as g/100 g, where 100 g was the total of all areas of the identiﬁed FAMEs. Nutritional indices that consider diﬀerent fatty acids according to their diﬀerent contributions to the promotion or prevention of cardiovascular disorders were calculated from the identiﬁed fatty acids. Atherogenic (AI) and thrombogenic (TI) indices were calculated using the Ulbricht and Southgate equations (40), while hypocholesterolemic/hypercholesterolemic ratio (H/H), was calculated using the equation proposed by Santos-Silva et al. (41). The three indices were calculated as follows: (Omegawax 250; Supelco, Bellefonte, PA, United States) 25-µm ﬁlm, and maintained at 100◦C for 5 min, from 100 to 240◦C at 4 ◦C/min and a ﬁnal isotherm of 20 min at 240◦C. Injector and detector temperatures were set at 250◦C. Injection volume and split ratio were 0.5 µl and 1:50, respectively. The carrier gas, helium (He), was set at a ﬂow rate of 1 ml/min. Data acquisition and processing were performed using ChromeleonTM Software (Thermo Fisher Scientiﬁc, Milan, Italy). Fatty acids of the ﬁsh samples were identiﬁed by comparing the relative retention times of FAMEs with those of a standard mix solution (mix 37 FAMEs; Supelco, Inc., Bellefonte, PA, United States) under the same analytical conditions. FA concentrations were expressed as g/100 g, where 100 g was the total of all areas of the identiﬁed FAMEs. Nutritional indices that consider diﬀerent fatty acids according to their diﬀerent contributions to the promotion or prevention of cardiovascular disorders were calculated from the identiﬁed fatty acids. Atherogenic (AI) and thrombogenic (TI) indices were calculated using the Ulbricht and Southgate equations (40), while hypocholesterolemic/hypercholesterolemic ratio (H/H), was calculated using the equation proposed by Santos-Silva et al. (41). Feeding Trial and Facilities The experimental study on Sparus aurata specimens was carried out at the IRBIM facility in the CNR headquarter in Messina (Italy). On 3 February 2020, 332 ﬁsh purchased by the Ittica Caldoli Company (Lesina, Foggia, Italy) were transported to the IRBIM-CNR facility and transferred to a large tank (4.5 m3) for about 1 week to acclimatize to the breeding conditions. During that time, ﬁsh were fed with a commercial diet (46% protein, 16% fat; 20.7% NFE, 2.3% crude ﬁber; Aller Blue Omega 3 mm; Aller Aqua Company, Christiansfeld, Denmark). After 1 week of acclimation, a total of 324 mixed-sex specimens were individually weighed (average initial weight: 143.65 ± 25.94 g) and randomly divided into 12 indoor ﬁberglass tanks of 1.4 m3 (27 ﬁsh/tank, 3 replicate tanks per diet, and total of 81 ﬁsh per diet), in an open circuit system, with intake and discharge of 12 L/min of water from and to the sea (12 complete tank renewals a day). Some water parameters (pH, O2, temperature) were monitored daily For determination of total lipid n, the aliquots (approximately 2 g) of wet ﬁlet muscles from the four ﬁsh groups (n = 72) were ground, and the oil was extracted using a chloroform/methanol (2:1, v/v) solution (38). Each chemical analysis was performed in triplicate. Then, total lipids were used to prepare fatty acid methyl esters (FAMEs) for the analysis of fatty acid (FA) proﬁle, according to the method of Christie (39). In particular, on each sample of total lipid, 2 ml of methanol:sulfuric acid (9:1, v/v) solution was added, and the mixture was heated at 100◦C for 1 h. The FAMEs were analyzed with a Trace 1310 chromatograph (Thermo Fisher Scientiﬁc, Milan, Italy) provided with a ﬂame ionization detector (FID). Separation of FAMEs was carried out using a 30 m ×.25 mm (length × internal diameter). fused silica capillary column May 2022 \| Volume 9 \| Article 896552 3 Sparus aurata Fed Insect Meal Oteri et al. TABLE 2 \| Analytical line length (nm) utilized for analysis. Element nm Element nm Cr 267.716 Se 196.026 Cu 327.393 Ni 231.064 Fe 238.204 Mn 257.610 K 766.490 Zn 213.857 P 213.617 Mg 285.592 Na 589.592 Ca 317.933 TABLE 2 \| Analytical line length (nm) utilized for analysis. Frontiers in Nutrition \| www.frontiersin.org Statistical Analysis For chemical composition of the ﬁlets, all the data were analyzed with the ANCOVA procedure of the XLSTAT statistical package (51). The diets (HIM0, HIM25, HIM35, and HIM50%) were used as a ﬁxed eﬀect and the ﬁnal body weight as the covariate. In this way, the possible eﬀects of diet and body weight have been separated. Separation of means was assessed by Tukey’s test, and diﬀerences were signiﬁcant if p < 0.05. To evaluate the inﬂuence of the diﬀerent dietary formulations on the microbiological proﬁle of the skin, intestine, and muscle of the ﬁsh, the microbial loads of each parameter between the diﬀerent groups were compared. The normal distribution of the raw data was tested by a D’Agostino-Pearson omnibus test, and a one-way analysis of variance (ANOVA) was conducted to evaluate any signiﬁcant diﬀerences between each group. A post hoc Tukey’s test was conducted for the multiple comparisons in the obtained ANOVA data. Critical signiﬁcance level (p) was set at 5% (0.05), and all the tests were performed two-sided. All the statistical analyses were carried out with the Graph Pad Prism 9 software (San Diego, CA, United States). Feeding Trial and Facilities For cysteine analysis (43), an oxidation reaction using Optical optimization of the ICP-OES was conducted with the procedure of the SyngistixTM software, and torch position May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 4 Sparus aurata Fed Insect Meal Oteri et al. TABLE 3 \| Operational conditions of the ICP-OES. Parameter Conditions Radiofrequency power (W) 1,500 Plasma gas ﬂow (L/min) 10.0 Auxiliary gas (L/min) 0.2 Nebulizer gas (L/min) 0.7 Sample uptake (mL/min) 1.0 TABLE 3 \| Operational conditions of the ICP-OES. on Iron Agar (Lyngby) (Oxoid Ltd., Basingstoke, Hampshire, England), incubated at 20 ± 1◦C for 72 h counting black colonies as sulﬁde producers and white colonies as sulﬁde non-producers. The LODs were 10 CFU/g for the count of Enterobacteriaceae, Aeromonas spp., Clostridium spp. and SSOs, and 100 CFU/g for the count of Pseudomonas spp. The other aliquots for each sample of skin, dorsal ﬂesh, and intestine were processed for the detection of Listeria monocytogenes (50) as follows: they were homogenized with Listeria Fraser Broth Half Concentration (Biolife, Milano, Italy), incubated at 30 ± 1◦C for 20 h, followed by a passage in Listeria Fraser Broth (Biolife, Milano, Italy) incubated at 37 ± 1◦C for 24 h and spread both on Agar Listeria according to Ottaviani and Agosti (Biolife, Milano, Italy) and Listeria Palcam Agar (Biolife, Milano, Italy), both incubated at 37 ± 1◦C for 24–48 h. was optimized before the analytical step using an Mn analytical line with a 1 mg/l Mn solution. The quantiﬁcation of each mineral was made with external calibration curves using a set of solutions of 0.05, 0.25, 0.5, and 1 ppm arranged from a Perkin Elmer multi-element analytical solution for ICP analysis. The calibration curves for all elements were established using a calibration blank and a reagent blank, and all were found to have correlation coeﬃcients (r2) ranging from 0.998 to 0.999. Detection limits (DLs) were determined by analyzing a matrix blank represented by reagents and quantities same as those used for sample preparation. Recoveries from the certiﬁed materials have reached acceptable values and higher than 85% for most of the elements and up to 95% for Zn and Cu. Analysis of Microbiological Proﬁle of Fish Muscle The analyses were carried out on samples of skin dissected from the tail to the opercula, on dorsal ﬂeshes, and on the intestine dissected from the pylorus to the anal opening. Sampling was performed with sterile scissors and forceps by collecting the skin ﬁrst from one side and the underlying ﬂesh portion, and then repeating the operation on the opposite side and ﬁnally sampling the intestine. Each sample of skin, dorsal ﬂeshes, and intestine was split into two aliquots and subjected to microbiological analysis. The aliquots for each sample of skin, dorsal ﬂesh, and intestine were homogenized with buﬀered peptone water (Biolife, Milano, Italy, at a ratio of 1:9 w/v and with a stomacher (400 Circulator; International PBI s.p.a., Milano, Italy) for 60 s at 230 rpm and tested for the following parameters: (i) enumeration of Enterobacteriaceae (45) on Violet Red Bile Glucose Agar (Biolife, Milano, Italy), incubated at 37 ± 1◦C for 24 h; (ii) enumeration of Clostridium spp. (46) on Tryptose Sulﬁte Cycloserine Agar (Biolife, Milano, Italy), incubated at 37 ± 1◦C for 24 h under anaerobic conditions; (iii) detection of Salmonella spp. (47) on Chromogenic Salmonella Agar (Biolife, Milano, Italy) and Xylose Lysine Deoxycholate Agar (Biolife, Milano, Italy), both incubated at 37 ± 1◦C for 24 h; (iv) detection and enumeration of Pseudomonas spp. on Pseudomonas Agar Base (HiMedia Laboratories, Mumbai, India), incubated at 25 ± 1 ◦C for 48 h; (v) detection and enumeration of Aeromonas spp. on GSP Agar (Pseudomonas Aeromonas Selective Agar Base) acc. to KIELWEIN (Merck, Darmstadt, Germany), incubated at 30 ± 1◦C for 48 h; (vi) detection and enumeration of Vibrio spp. (48) on TCBS (thiosulfate citrate bile sucrose agar;bioMerieux, Marcy l’Etoile, France), incubated at 37 ± 1◦C for 24 h; (vii) detection and enumeration of Speciﬁc Spoilage Organisms (SSOs) (49) Fish Growth Performance Fish Growth Performance In brief, at the end of the feeding trial, all the ﬁsh almost tripled their mean body weight, but there were no signiﬁcant diﬀerences (p > 0.05) between the dietary groups for any of the considered growth performance and feed utilization eﬃciency indices. Frontiers in Nutrition \| www.frontiersin.org Chemical Composition of Filets Table 4 reports the chemical composition of sea bream ﬁlet muscle. Moisture, crude protein, total lipids, and ash contents of the ﬁlets were not aﬀected by the dietary treatments. The fatty acid composition of the sea bream ﬁlet muscle is shown in Table 5. The saturated fatty acid, the lauric acid (C12:0) and the myristic acid (C14:0) showed signiﬁcantly higher values in the HIM50% group than those observed in the ﬁsh fed ﬁsh meal and lower inclusions of Hermetia illucens meal. The unsaturated fatty acids, myristoleic acid (C14:1), alpha- linolenic acid (ALA, C18:2n6), and arachidonic acid (ARA, C20:4n6) showed signiﬁcantly higher values in the HIM50% May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 5 Sparus aurata Fed Insect Meal Oteri et al. TABLE 4 \| Proximate composition (g/100 g of wet weight) of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. GROUP p-value SEM1 HIM0 HIM25% HIM35% HIM50% D2 BW3 Fish 18 18 18 18 Moisture 67.11 67.24 66.79 67.68 0.354 0.110 0.287 Crude Protein 20.41 19.99 20.07 19.82 0.242 0.824 0.183 Total Lipids 10.78 11.13 11.48 10.83 0.273 0.304 0.275 Ash 1.70 1.64 1.66 1.67 0.869 0.392 0.049 HIM0, ﬁsh meal group; HIM25, HIM35, and HIM50%, Hermetia illucens meal at 25, 35, and 50% substitution rates of ﬁsh meal groups, respectively. Fish: 18 per each diet, 6 ﬁsh per tank, and 3 replications per diet. 1Standard error of the mean. 2Diet. 3Body weight. TABLE 4 \| Proximate composition (g/100 g of wet weight) of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. mposition (g/100 g of wet weight) of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. amino acids in the ﬁlets, arginine and phenylalanine showed signiﬁcantly higher values in ﬁsh fed with ﬁsh meal than in ﬁsh fed with diﬀerent inclusion of the insect meal; isoleucine and valine showed signiﬁcantly higher values in ﬁsh fed with higher inclusion of the insect meal (HIM50%) than in ﬁsh fed with ﬁsh meal and with lower levels of inclusion of the insect meal; leucine, lysine and methionine showed signiﬁcantly lower levels in ﬁsh of the HIM35% group, histidine in ﬁsh of the HIM25% group, and tryptophan in ﬁsh of the HIM50% group than those in the other groups. Threonine showed similar values (p > 0.05) in the groups. Chemical Composition of Filets Among the dispensable amino acids, alanine showed a signiﬁcantly higher level in ﬁlets of the HIM50% group than those observed in the HIM25 and HIM35% groups; aspartate + asparagine showed a signiﬁcantly higher level in ﬁlets of the HIM50% group than that observed in the ﬁlets of ﬁsh fed with ﬁsh meal; glycine showed a signiﬁcantly higher level in ﬁlets of the HIM50% group than that of HIM35% group; cysteine and tyrosine showed signiﬁcantly higher levels in the ﬁlets of ﬁsh fed with ﬁsh meal than those of the ﬁsh fed with diﬀerent inclusion of the insect meal. Hydroxylysine, hydroxyproline, and serine showed similar values in the groups. The EAA/NEAA ratio in the ﬁlets showed a signiﬁcantly higher value in the ﬁsh fed with ﬁsh meal than in the ﬁsh fed with diﬀerent inclusion of the insect meal. Body weight did not signiﬁcantly (p > 0.05) aﬀect the indispensable or the dispensable amino acid proﬁle. group than those observed in the ﬁsh fed with ﬁsh meal and lower inclusions of the Hermetia illucens meal; oleic acid (C18:1n9) showed a signiﬁcantly higher level in the HIM25 and HIM35% groups than in the HIM50% group; eicosapentaenoic acid (EPA, C20:5n3) showed a similar content in the HIM0, HIM35 and HIM50% groups, which was signiﬁcantly higher than that in the HIM25% group; docosahexaenoic acid (DHA, C22:6n3) showed a signiﬁcantly higher content in the ﬁsh fed with ﬁsh meal than in the ﬁsh fed with the insect meal. The fatty acid classes of the ﬁlets are reported in Table 6. The sum of the saturated fatty acids (SFAs) showed a signiﬁcantly higher value in the HIM50% group than the observed value in the HIM0 group. The sum of the monounsaturated fatty acids (MUFAs) showed a signiﬁcantly lower value in the HIM50% group than the observed in the ﬁsh with fed ﬁsh meal and lower inclusions of the Hermetia illucens meal. The sum of the polyunsaturated fatty acids (PUFAs) and PUFAs of the n3-series showed a similar content in the HIM0, HIM35, and HIM50% groups, which was higher than that in the HIM25% group, whereas the fatty acids of the n6-series showed a signiﬁcantly higher value in the HIM50% group than the observed in the ﬁsh fed with ﬁsh meal and with lower levels of inclusion of the insect meal. Frontiers in Nutrition \| www.frontiersin.org Chemical Composition of Filets n3/n6 PUFA ratio (Table 6), as well as the sum of EPA+DHA, showed signiﬁcantly higher levels in the ﬁsh with fed with ﬁsh meal than those of the ﬁsh fed with diﬀerent inclusion of the insect meal. The indices of nutritional interest, i.e., the atherogenic (AI), thrombogenic (TI), and peroxidation (PI) indices and hypocholesterolemic/hypercholesterolemic (H/H) ratio are reported in Table 6. No signiﬁcant (p > 0.05) diﬀerence was observed for TI and H/H in the groups, whereas the AI showed a similar level in the HIM0, HIM25, and HIM35% groups but was signiﬁcantly lower and, therefore, better than that recorded in the HIM50% group. On the contrary, the PI showed a similar content in the HIM0, HIM35, and HIM50% groups but was higher than that in the HIM25% group. Body weight did not signiﬁcantly (p > 0.05) aﬀect the fatty acid classes, the ratio, or the nutritional indices. Table 8 shows the average mineral content values in the sea bream ﬁlet muscle: macrominerals (phosphorus, sodium, potassium, calcium, and magnesium), microminerals (zinc, iron, manganese, copper, and chromium), and trace minerals (nickel). Phosphorus, showed a signiﬁcantly higher value in the ﬁsh of the HIM25% group than that observed in the HIM0, HIM35, and HIM50% groups. Calcium showed a signiﬁcantly higher value in the ﬁlets of the HIM0 group than those observed in the HIM50% group, whereas the HIM25 and HIM35% groups showed intermediate values. Sodium, potassium, and magnesium showed similar values (p > 0.05) in the groups. Due to the antagonist interaction of the Ca and P, the concentration ratio between these macroelements was calculated (52). A signiﬁcantly higher level of the Ca/P ratio was observed in the HIM0 group than in the ﬁsh fed with diﬀerent inclusions of Hermetia illucens. Twenty amino acids, ten indispensable amino acids (EAA), and ten dispensable ones (NEAA), were identiﬁed and quantiﬁed in the sea bream ﬁlet muscle (Table 7). Among the indispensable May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 6 Sparus aurata Fed Insect Meal Oteri et al. TABLE 5 \| Fatty acid composition (g/100 g of fatty acid methyl esters)# of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. Chemical Composition of Filets GROUP p-value SEM1 HIM0 HIM25% HIM35% HIM50% D2 BW3 Fish 18 18 18 18 C12:0 0.05 d 0.23 c 0.40 b 0.58 a <0.0001 0.265 0.001 C14:0 2.57 b 2.59 b 2.63 ab 2.75 a 0.001 0.636 0.012 C14:1 0.06 b 0.06 b 0.06 b 0.07 a 0.001 0.320 <0.0001 C15:0 0.20 a 0.19 b 0.19 b 0.19 b 0.001 0.690 <0.0001 C16:0 15.29 15.31 15.12 15.41 0.626 0.944 0.257 C16:1 4.17 4.16 4.25 4.16 0.610 0.716 0.031 C17:0 0.17 a 0.16 b 0.16 b 0.17 a 0.003 0.610 <0.0001 C17:1 0.03 0.03 0.03 0.03 0.296 0.407 <0.0001 C18:0 3.41 3.45 3.40 3.38 0.898 0.113 0.054 C18:1n9 40.62 ab 41.11 a 40.78 a 40.02 b <0.0001 0.715 0.308 C18:1n7 3.11 a 3.15 a 3.06 ab 3.01 b 0.001 0.996 0.062 C18:2n6 12.83 c 12.78 c 13.17 b 13.52 a <0.0001 0.231 0.062 C18:3n6 0.17 0.16 0.18 0.15 0.058 0.031 0.001 C18:3n3 3.26 3.28 3.25 3.30 0.661 0.348 0.010 C20:0 0.25 0.26 0.25 0.26 0.352 0.770 <0.0001 C20:1n9 2.02 a 2.02 a 1.82 b 1.79 b <0.0001 0.339 0.006 C20:2n6 0.29 b 0.29 b 0.31 b 0.33 a <0.0001 0.460 0.001 C20:3n6 0.16 0.17 0.17 0.17 0.505 0.112 <0.0001 C20:4n3 0.39 a 0.35 b 0.37 ab 0.36 b 0.0003 0.682 0.001 C20:4n6 ARA 0.12 b 0.12 b 0.13 ab 0.14 a 0.0002 0.224 <0.0001 C20:5n3 EPA 3.41 a 3.20 b 3.33 a 3.32 a 0.0002 0.026 0.012 C22:0 0.14 0.14 0.14 0.15 0.675 0.823 <0.0001 C22:1n9 1.12 a 0.99 b 0.93 b 0.94 b 0.0005 0.738 0.012 C22:5n3 DPA 1.28 b 1.27 b 1.34 a 1.34 a 0.047 0.677 0.006 C22:6n3 DHA 4.80 a 4.48 b 4.46 b 4.41 b 0.008 0.588 0.077 HIM0 ﬁh l HIM25 HIM35 d HIM50% H ti ill l t 2 35 d 50% b tit ti t f ﬁh l ti l ty acid methyl esters)# of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. 3Body weight. colony of the researched microorganisms was detected. No Clostridium spp., Salmonella spp., Aeromonasspp., Vibrio spp., and L. monocytogenes were detected in the skin and intestine of the tested specimens. The Enterobacteriaceae loads detected in the intestine of ﬁsh of all the groups were <2 log cfu/g, while they were not detected in the skin of any ﬁsh. Chemical Composition of Filets SSO loads detected in the skin and intestine were <2 log cfu/g in the HIM0 and HIM25% groups but slightly higher than 2 Log cfu/g in the HIM35 and HIM50% groups although did not reach statistical signiﬁcance. Regarding SSOs in the samples of the HIM0% group, only white colonies were detected while, in the samples of the HIM25, HIM35, and HIM50% groups, the white colonies compared to the black colonies accounted for majority (> 98%). Pseudomonas spp. was detected in the skin and intestine of all the specimens tested (≤2 log cfu/g on average). Signiﬁcantly (p < 0.05) higher loads were observed for the Among the microminerals, manganese, copper, and chromium showed similar values (p > 0.05) in the groups. Zinc showed a signiﬁcantly lower value in ﬁsh of the HIM25 and HIM50% groups than in those of the HIM0 and HIM35% groups. Iron was signiﬁcantly higher (p = 0.011) in the ﬁlets of the HIM0 group than in those of HIM50%, whereas intermediate values in the HIM25% and HIM35% groups were observed. The only trace mineral identiﬁed was nickel, which showed a signiﬁcantly higher (p = 0.044) value in ﬁsh of the HIM25% group than in those of the HIM35% group; intermediate values in ﬁsh of the HIM0 and HIM50% groups were observed. Body weight did not signiﬁcantly (p > 0.05) aﬀect the mineral proﬁle or the Ca/P ratio. h meal group; HIM25, HIM35, and HIM50%, Hermetia illucens meal at 2, 35, and 50% substitution rates of ﬁsh meal groups, respectively. per diet, 6 ﬁsh per tank, and 3 replications per diet. Microbiological Proﬁle of Filets g The results of the microbial analysis are summarized in Table 9. In the dorsal ﬂesh of the ﬁsh from all the tested groups, no May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 7 Sparus aurata Fed Insect Meal Oteri et al. TABLE 6 \| Fatty acid classes and nutritional indices of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. GROUP p-value SEM1 HIM0 HIM25% HIM35% HIM50% D2 BW3 Fish 18 18 18 18 SFA 22.08 b 22.33 ab 22.28 ab 22.88 a 0.042 0.619 0.469 MUFA 51.14 a 51.51 a 50.93 a 50.00 b <0.0001 0.792 0.360 PUFA 26.72 a 26.10 b 26.72 a 27.05 a <0.0001 0.318 0.224 n3 13.15 a 12.57 b 12.75 ab 12.74 ab 0.003 0.512 0.134 n6 13.57 c 13.52 c 13.97 b 14.31 a <0.0001 0.380 0.071 n3/n6 0.97 a 0.93 b 0.91 bc 0.89 c <0.0001 0.995 0.001 EPA+DHA 8.21 a 7.68 b 7.79 b 7.73 b 0.001 0.747 0.100 AI 0.33 b 0.33 b 0.34 b 0.35 a 0.001 0.941 <0.0001 TI 0.29 0.30 0.30 0.30 0.182 0.501 <0.0001 H/H ratio 3.72 3.70 3.75 3.64 0.379 0.964 0.023 PI 64.62 a 61.95 b 63.17 ab 63.30 ab 0.004 0.651 2.910 HIM0, ﬁsh meal group; HIM25, HIM35, and HIM50%, Hermetia illucens meal at 25, 35, and 50% substitution rates of ﬁsh meal groups, respectively. Fish: 18 per diet, 6 ﬁsh per tank, and 3 replications per diet. SFA, sum of the saturated fatty acids; MUFA, sum of the monounsaturated fatty acids; PUFA, sum of the polyunsaturated fatty acids; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; AI, atherogenic Index; TI, thrombogenic Index; PI, peroxidation index; H/H, hypocholesterolaemic/hypercholesterolaemic ratio. Mean values with different letters in the same row are signiﬁcantly different at p < 0.05. 1Standard error of the mean. 2Diet. 3Body weight. 6 \| Fatty acid classes and nutritional indices of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. 2Diet. 3Body weight. content in the muscle, signiﬁcantly lower values were found in ﬁsh fed with insect meal than in those fed with the basal diet (100% ﬁsh meal), although the diets contained similar levels of DHA [Oteri et al. (33)] and were formulated to provide DHA above the estimated EFA requirements (55). The results are in agreement with the observations of Belforti et al. (8) and Pulido et al. Microbiological Proﬁle of Filets (56). EPA appeared reduced in muscle fatty acids (3.32%, on average) compared to dietary concentrations (4.75%, on average) (33). The results agree with Sealey et al.’s observations (54), but they are not in agreement with St-Hilaire et al. (26) and Ewald et al. (28) who observed a decrease in ALA and EPA in ﬁsh fed with the inclusion of HIM in the feed. It is assumed that marine ﬁsh have a deﬁcient capacity to bioconvert 18C precursors (C18: 2n6 and C18: 3n3) into LC- PUFAs with 20 or 22 carbon atoms such ARA, EPA, and DHA, thanks to the activity of 16, 15, and 14 fatty acid desaturases (57, 58). For this reason, marine ﬁsh require the presence of preformed LC-PUFAs in the diet. However, Seiliez et al. (59) demonstrated the presence and nutritional modulation of a 16 fatty acid desaturase in Sparus aurata. The same authors did not detect 15 and 14 fatty acid desaturases; the latter are responsible for the synthesis of DHA from DPA (C22: 5n3). Therefore, clear explanation and interpretation of the results obtained appear diﬃcult, as the metabolic pathway of long-chain polyunsaturated fatty acids (LC-PUFAs) in marine ﬁsh is still under debate. Pseudomonas spp. in the skin of the HIM50% group compared to the HIM25% group. Frontiers in Nutrition \| www.frontiersin.org h meal group; HIM25, HIM35, and HIM50%, Hermetia illucens meal at 25, 35, and 50% substitution rates of ﬁsh meal groups, respectively. per diet 6 ﬁsh per tank and 3 replications per diet DISCUSSION The diet containing insect meal did not aﬀect the chemical composition of sea bream (Sparus aurata) muscle. A slight decrease in protein content was observed in ﬁsh fed with insect meal, although this variation was not statistically signiﬁcant. The data are in accordance with studies on Atlantic salmons fed with a larva meal from Hermetia illucens (9) and sea breams fed with a larva meal from Tenebrio molitor (53). Among the fatty acids in ﬁsh ﬁlets, the most interesting results were regarding the signiﬁcant diﬀerences observed for some polyunsaturated fatty acids of nutritional interest. These observations are not consistent with the dogma in that diﬀerences in the fatty acid composition of muscle lipids reﬂect diﬀerences in dietary fatty acid contents (53, 54). In fact, despite the diﬀerent Hermetia illucens inclusions into the diet (33) not being able to modify the fatty acid content of the feeds, the fatty acids in the sea bream muscles showed a diﬀerent trend. Moreover, as observed by Sealey et al. (54), some muscle fatty acid concentrations were attenuated relative to dietary content. Among these, oleic acid (C18:1n9) ranged from 43 to 45% in the diets (33), but it only ranged from 40 to 41% in the muscles, and alfa-linolenic acid (C18:3n3) ranged from 4 to 5% in the diets (33) but was only about 3% in the muscles of all the groups. A higher concentration of DHA (C22:6n3) was detected in the muscle (4.54% on average) in comparison to its content in the diets (3.61% on average). Regarding the trend of DHA It must be mentioned tha tEPA and DHA are essential for the growth, development and health, and regulation of expression of several genes involved in lipid metabolism (60). ARA and EPA play a major role in eicosanoid production (61). The results showed that in the ﬁlets of all the experimental groups, the May 2022 \| Volume 9 \| Article 896552 8 Sparus aurata Fed Insect Meal Oteri et al. TABLE 7 \| Amino acid composition (g/100 g of wet weight) of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. DISCUSSION GROUP p-value SEM1 HIM0 HIM25% HIM35% HIM50% D2 BW3 Fish 18 18 18 18 Indispensable amino acids Arginine 1.77 a 1.11 b 1.05 c 1.16 b <0.001 0.295 0.014 Histidine 0.83 a 0.77 b 0.82 ab 0.86 a <0.001 0.701 0.013 Isoleucine 1.23 b 1.25 b 1.20 b 1.34 a <0.0001 0.404 0.014 Leucine 2.12 a 2.07 a 1.93 b 2.11 a <0.001 0.463 0.018 Lysine 3.66 a 3.65 a 3.44 b 3.55 ab 0.011 0.236 0.047 Methionine 0.66 a 0.59 b 0.56 c 0.66 a <0.001 0.884 0.008 Phenylalanine 1.56 a 1.16 c 1.34 b 1.04 d <0.001 0.626 0.011 Threonine 1.10 1.18 1.16 1.16 0.212 0.691 0.027 Valine 1.15 b 1.17 b 1.13 b 1.26 a <0.001 0.705 0.015 Tryptophan 0.02 b 0.03 a 0.03 a 0.02 c <0.001 0.855 0.011 Dispensable amino acids Hydroxylysine 0.14 0.12 0.13 0.12 0.630 0.276 0.009 Alanine 0.96 ab 0.95 b 0.96 b 1.00 a 0.013 0.187 0.011 Aspartate+Asparigine 1.39 b 1.41 ab 1.39 ab 1.48 a 0.032 0.841 0.024 Cysteine 0.06 a 0.03 b 0.04 b 0.01 c <0.001 0.766 0.003 Glycine 0.93 ab 0.91 ab 0.85 b 0.95 a 0.007 0.478 0.019 Glutamate+ Glutammine 0.77 0.74 0.76 0.75 0.823 0.598 0.021 Proline 0.73 a 0.72 ab 0.68 b 0.76 a 0.001 0.666 0.013 Hydroxyproline 0.33 0.33 0.34 0.34 0.456 0.193 0.008 Tyrosine 1.03 a 0.97 b 0.90 c 0.91 c <0.001 0.889 0.014 Serine 1.07 1.15 1.14 1.09 0.058 0.615 0.024 EAA/NEAA 1.91 a 1.77 b 1.76 b 1.78 b <0.001 0.206 0.018 HIM0 ﬁh l HIM25 HIM35 d HIM50% H ti ill l t 25 35 d 50% b tit ti t f ﬁh l ti l 3Body weight. with cholesterol-lowering properties in ﬁsh (62, 63), it binds with lipid micelles (cholesterol), inhibits their absorption, and increases the excretion of bile acid; thus, it interferes with the absorption of cholesterol (64). As observed by Iaconisi et al. (53), the atherogenic index showed the worst value in ﬁlets of ﬁsh fed with highest inclusion of HIM (HIM50% group), testifying a greater probability of fatty acids to aﬀect the incidence of cardiovascular diseases (40). However, the AI and TI values observed in ﬁlets appeared much lower and, therefore, better than those reported in terrestrial animal foods (8, 65). The peroxidation index considers the contribution that PUFAs make in inﬂuencing oxidative degradation. Frontiers in Nutrition \| www.frontiersin.org h meal group; HIM25„ HIM35, and HIM50%, Hermetia illucens meal at 25, 35 and 50% substitution rates of ﬁsh meal groups, respectively. di t 6 ﬁh t k d 3 li ti di t DISCUSSION GROUP p-value SEM1 HIM0 HIM25% HIM35% HIM50% D2 BW3 Fish 18 18 18 18 Macrominerals P - Phosphorus 2,356 c 12,016 a 9,459 b 8,880 b <0.0001 0.575 300.323 Na-Sodium 665 578 520 6340 0.254 0.652 53.05 K-Potassium 3286 3414 3503 3462 0.453 0.493 98.81 Ca-Calcium 943 a 625 ab 590 ab 484 b 0.019 0.484 96.01 Mg-Magnesium 358 344 367 356 0.709 0.526 13.83 Ca/P ratio 0.41 a 0.05 b 0.07 b 0.05 b <0.0001 0.088 0.03 Microminerals Zn-Zinc 24.96 a 17.92 b 22.36 a 16.44 b <0.0001 0.466 2.14 Fe-Iron 5.73 a 4.09 ab 4.47 ab 3.60 b 0.011 0.281 0.41 Mn-Manganese 0.41 0.50 0.49 0.43 0.635 0.080 0.06 Cu-Copper 0.31 0.35 0.52 0.35 0.110 0.693 0.06 Cr-Chromium 0.37 0.41 0.46 0.48 0.156 0.124 0.05 Se-Selenium 0.30 a 0.22 b 0.22 b 0.35 a <0.0001 0.377 0.02 Trace mineral Ni-Nickel 1.13 ab 1.17 a 0.75 b 1.10 ab 0.044 0.094 0.101 HIM0, ﬁsh meal group; HIM25, HIM35, and HIM50%, Hermetia illucens meal at 25, 35, and 50% substitution rates of ﬁsh meal groups, respectively. Fish: 18 per diet, 6 ﬁsh per tank, and 3 replications per diet. Mean values with different letters in the same row are signiﬁcantly different at p < 0.05. 1Standard error of the mean. TABLE 8 \| Mineral element proﬁle (mg/kg of wet weight) of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. TABLE 9 \| Microbial proﬁle (log cfu/g) of the skin, intestine, and muscle of the gilthead sea bream fed with the four exerimental diets. Items GROUP HIM0% HIM25% HIM35% HIM50% Skin Intestine Muscle Skin Intestine Muscle Skin Intestine Muscle Skin Intestine Muscle Enterobatteriaceae <1 1.76 <1 <1 1.83 <1 <1 1.83 <1 <1 1.83 <1 SSOs1 1.90 1.93 <1 1.86 1.96 <1 2.03 2.04 <1 2.04 2.04 <1 Pseudomonas spp. 1.66ab 1.67 <1 1.68b 1.82 <1 1.91ab 1.87 <1 1.95a 1.93 <1 Aeromonas spp. <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 Vibrio spp. <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 Clostridium spp. <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 L. monocytogenes Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Salmonella spp. Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent 1Speciﬁc spoilage organisms. DISCUSSION The control of this process, causing loss of nutritional value and formation of anti-nutritional molecules, can play a central role in maintaining muscle quality (66). Lastly, the highest level of lauric acid observed in ﬁsh ﬁlets of the HIM50% group, did not aﬀect (p > 0.05) ﬁsh growth performance (ﬁnal body weight: 390 ± 49 g; speciﬁc growth rate: 0.75 ± 0.02). This fatty acid is dominant in black soldier ﬂy larvae (28, 32) and is considered a bioactive compound for a possible role as an antimicrobial counteracting antibiotic resistance (67). ARA/EPA ratio did not change (0.04), and that the levels of DHA were more depressed than those of EPA, as indicated by the changes in the EPA/DHA ratio (from 0.71 to 0.75). Moreover, Pulido et al. (56), with the aim of evaluating to what extent replacing ﬁsh meal with insect meal could alter not only the fatty acid (FA) proﬁle of ﬁlets of Sparus aurata but also the FA distribution inside tryglicerides, observed that the inclusion of HIM reduced n3-PUFAs in sea bream ﬁlets but did not substantially change the presence of fatty acids important for human nutrition (e.g., EPA and DHA) in the sn-2 position of ﬁlet triglycerides, increasing the chances of being better assimilated and absorbed by potential consumers. The values recorded for some health lipid indices (TI: thrombogenic index and H/H: hypocholesterolemic/hypercholesterolemic ratio) appeared to be of interest, into account the contribution that each fatty acid has to inﬂuence the incidence of cardiovascular diseases (40). The similar values of these indices in the ﬁlets of all the groups suggest similar nutritional eﬀects of all the diets on the animals. This result could be due to chitin, the main component of the exoskeleton of insects. Chitin contains high levels of chitosan May 2022 \| Volume 9 \| Article 896552 9 Sparus aurata Fed Insect Meal Oteri et al. TABLE 8 \| Mineral element proﬁle (mg/kg of wet weight) of the ﬁlets of the gilthead sea bream fed with the four exerimental diets. 0, ﬁsh meal group; HIM25, HIM35, and HIM50%, Hermetia illucens meal at 25, 35, and 50% substitution rates of ﬁsh meal groups, respective 18 per diet, 6 ﬁsh per tank, and 3 replications per diet. ues with different letters in the same row are signiﬁcantly different at p < 0.05. d f h DISCUSSION (85), the content of metals was higher in the gills than in the muscles. Gills and liver are chosen as target organs for assessing metal accumulation. Therefore, the signiﬁcantly lower values of the Ca/P ratios in ﬁsh receiving insect meal should not cause a concern. In fact, the Ca e P content did not aﬀect anomalies in mineral homeostasis and bone mass (86). Heavy metals in the marine environment and ﬁsh contamination not only pose a threat to ﬁsh health, but by accumulating as they ﬂow down to the natural food chain, they also pose a risk to human health (87, 88). Therefore, it is necessary to determine their content in widely consumed ﬁsh species such as Sparus aurata. Heavy metals such as manganese, iron, cobalt and copper are necessary for ﬁsh metabolism (89) but are toxic at high concentrations (90), while cadmium, chromium, mercury, lead and nickel are toxic metals even if present in traces in both humans and animals (91) causing numerous damages to organs (92). Although chromium is a ubiquitous metal in the environment and trivalent chromium is essential for biolife, hexavalent chromium is said to be a toxic metal with mutagenic, carcinogenic, and harmful impacts on the biota. Researchers revealed that chromium aﬀects the physiological, behavioral, histological, biochemical, genetic, and immunological conditions of experimental organism (93). The chromium concentrations in the ﬁsh ﬁlets were found to be below the permitted level set by the European Union at 0.5 mg/kg wet weight (94). Manganese functions as a cofactor in several enzyme systems, including those involved in urea synthesis from ammonia, amino acid metabolism, fatty acid metabolism, and glucose oxidation (95). Manganese levels in ﬁsh ﬁlets were found to be below the permitted levels established by the FAO/WHO (96). Iron, essential in ﬁsh as a heme protein compound (e.g., hemoglobin myoglobin and cytochromes) or as a nonheme Minerals are divided into macroelements, whose needs by organisms are in large quantities, microelements, whose needs by organisms are in small quantities (79), and trace minerals typically required by organisms in such small quantities that a dietary supplement is not required (77). The functions of macrominerals include formation of skeletal structures and other hard tissues, electron transfer, regulation of acid: base equilibrium, production of membrane potentials, and osmoregulations (77). Among microelements, calcium and phosphorus are two of the major constituents of the inorganic portions of diets for ﬁsh. DISCUSSION Data reported are expressed as mean values of 9 samples analyzed (3 ﬁsh per tank and 3 replications per diet). Mean values with different letters in the same row are signiﬁcantly different at p < 0.05. 9 \| Microbial proﬁle (log cfu/g) of the skin, intestine, and muscle of the gilthead sea bream fed with the four exerimental diets. in the corresponding feeds (33). Quantitatively, their content in ﬁsh ﬁlets was similar in the groups, with the exception of the muscles of the HIM35% group, although the Lys and Leu content was higher in feeds containing insect meal as a partial substitute for ﬁsh meal. One possible explanation could be that, although all the diets were formulated to be isonitrogenous, the chitin content of the feeds containing insect meal may have reduced the Based on our knowledge, this study is the ﬁrst to test the eﬀects of Hermetia illucens meal dietary inclusion on the amino acid composition of sea bream ﬁlets, so few comparisons with the literature are possible. On the whole, the essential amino acid proﬁle of the ﬁsh ﬁlets reﬂected high protein quality. As observed by Iaconisi et al. (68), Lys and Leu were the most representative EAAs in ﬁsh ﬁlets and are the same EAAs contained at a high level May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 10 Sparus aurata Fed Insect Meal Oteri et al. levels of digestible proteins or, more speciﬁcally, available amino acids (69). In fact, chitin, an unbranched N-acetylglucosamine polymer, is indigestible for many ﬁsh species that are devoid of chitinase activity (70) or with limited activity (18, 71–73). This leads to impaired digestibility of other nutrients with consequent increase in bulk, reduced feces retention time, and reduced enzyme accessibility to substrates (74). Nonetheless, the growth performance of the ﬁsh (ﬁnal body weight: 390 ± 49 g; speciﬁc growth rate:.75 ± 0.02) was similar in the groups and was not aﬀected by the dietary incorporation of HIM. Furthermore, as reported by Belghit et al. (9), Hermetia illucens larvae have a well-balanced EAA proﬁle, with the exception of lysine, methionine, and tryptophan, close to that of the ﬁsh meal, considered as the protein with the best EAA proﬁle for ﬁsh (75). DISCUSSION In this trial, in relation to the levels of FM replacement by HIM, these amino acids were added to the diets to meet the needs of Sparus aurata as suggested by Magalhães et al. (76). As quantiﬁcation of EAA requirements is generally based on analysis of dose-response curves with weight gain used as response criterion (77), the similar results obtained for the in vivo performance of the ﬁsh of all groups (ﬁnal body weight: 390 ± 49 g; speciﬁc growth rate:.75 ± 0.02) demonstrate that all the diets meet the dietary amino acid requirement of Sparus aurata. NEAAs are not strictly necessary in the diet, because ﬁsh can synthesize them on their own; however, NEAAs can have beneﬁcial eﬀects on ﬁsh health and performance when present in the right concentration (78). In this study, although the diets containing the insect meal had a higher NEAA level (33) than the diet containing exclusively ﬁsh meal, the NEAA content in ﬁsh ﬁlets of all the groups was similar. This result is not in accordance with the observations of Belghit et al. (9) and could be due, as reported above, to the dietary content of chitin. to be of particular interest in relation to the results obtained in this study, where the phosphorus content was signiﬁcantly lower in the ﬁsh ﬁlets of the control group (HIM0 = 2.4 g/kg) fed with diet with highest phosphorus content (11.45 g/kg) and formulated exclusively with ﬁsh meal (33). As reported by Rodehutscord and Pfeﬀer (80), phosphorus concentrations in practical dietary formulations mainly based on ﬁsh meal considerably exceed the estimation requirements. Therefore, excess in dietary phosphorus and low amount of absorbed phosphorus by ﬁsh lead to a problem of environmental impact caused by surplus phosphorus discharge into the eﬄuents (81). However, muscle tissuess are not considered to be speciﬁc physiological sites for calcium and phosphorus (82). Phosphorus and calcium accumulate in largest amounts in bones. Borucka- JastrzeBska et al. (83) determined micro- and macroelement concentrations in diﬀerent tissues of ﬁsh, and they reported that calcium distribution followed the same pattern for all three analyzed species in decreasing order: gills > muscles > skin > liver > kidney > blood. Perkowska and Protasowicki (84) showed that high levels of heavy metals were in the liver, and that the lowest ones were in the muscles. Moreover, in ﬁsh species analyzed by Roméo et al. Frontiers in Nutrition \| www.frontiersin.org ETHICS STATEMENT The animal study was reviewed and approved by Italian Ministry of Health (Ministerial Authorization Number 491/2019-PR released on date 4th July 2019). Despite the observed diﬀerences in the lipid, protein, and mineral proﬁles of the ﬁlets, the organoleptic properties, in terms of color, volatile fraction, and taste, of the 4 groups of fresh ﬁlets resulted similar between the groups, suggesting that the use of HIM does not alter signiﬁcantly the organoleptic properties of Spaurus aurata ﬁlets (34). CONCLUSIONS Overall, considering that several studies indicated that the content of these microelements in ﬁsh is inﬂuenced by surrounding water (77, 101) and by food source (102), which is the major source of elements such as iron, zinc, manganese and copper, the data would seem to highlight that the inclusion of HIM into the three diets maintains lower heavy metal levels than those recommended by various authorities (FAO, WHO, and EU). This study indicates that the Hermetia illucens meal as a partial substitute for ﬁsh meal did not aﬀect the proximate composition of the ﬁsh ﬁlets but signiﬁcantly aﬀected the fatty acid and amino acid proﬁles. However, since no detrimental eﬀects on growth performance were found, the eﬀects on ﬁlet quality should be considered. Furthermore, the heavy metal content and microbiological quality of the ﬁsh ﬁlets underline the safety of the Hermetia illucens meal as animal feed. As the price of ﬁshmeal and ﬁsh oil increases, an economic analysis of incorporation of the Hermetia illucens meal into diets is needed to better assess its role as an aﬀordable and sustainable feed in aquaculture. DISCUSSION Quantitatively, calcium and phosphorus function primarily as structural components of hard tissues. Dietary deﬁciencies of most macrominerals such as calcium have been generally diﬃcult to produce with ﬁsh species because of the presence of these ions in water (77). On the contrary, concentrations of phosphorus in natural waters are generally very low (77). Deﬁciency of dietary phosphorus impairs intermediate metabolism and causes reduction in ﬁsh growth and feed conversion. Integrating phosphorus into ﬁsh diets is generally more critical, because its presence in water and use by ﬁsh are limited. However, the inﬂuence of excreted phosphorus on eutrophication of receiving waters has led to a signiﬁcant amount of research focused on phosphorus nutrition with the aim of minimizing phosphorus excretion (77). This appears May 2022 \| Volume 9 \| Article 896552 11 Sparus aurata Fed Insect Meal Oteri et al. FUNDING This study was supported by the Italian Ministry of Agricultural, Food and Forestry Policies, and by European Maritime and Fisheries Fund (PO FEAMP) 2014–2020 mis. 2.47 CUP J46C18000570006, project codex 03/INA/17, Title of the project FIFA-Feed Insects for Aquaculture; Scientiﬁc Responsible: Biagina Chiofalo. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. 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. AUTHOR CONTRIBUTIONS MO: formal analysis and writing (original draft preparation and review and editing). BC: conceptualization, methodology, investigation, data curation, writing (original draft preparation and review and editing), supervision, and funding acquisition. GM: investigation and resources. GT: formal analysis. LN and VL: formal analysis and writing (original draft preparation). ADR: methodology, software, data curation, and writing (review and editing). All authors contributed to the article and approved the submitted version. With regard to biological hazards, the EFSA opinion identiﬁes the substrate used to feed the insects as the key entrance point for contamination (15); therefore, pathogenic bacteria may be present in insects depending on the substrate used and rearing conditions. Mucosal tissues, including skin and gut, are in direct contact with the environment and, thus, are the ﬁrst contact points of microbes with their host representing a good control point for ﬁsh health and consumer safety (103). The reported results are conﬁrmed, as the insect meal incorporation into the diets did not signiﬁcantly inﬂuence the microbiological proﬁle of the ﬁsh. Pseudomonas spp. belongs to the group of the SSOs whose loads did not diﬀer between the various groups; therefore, although the loads of Pseudomonas spp. in the skin of the HIM50% group was signiﬁcantly higher than in HIM25%, the detected values do not represent a relevant risk to public health. The feed can impact the microbial quality of the ﬁsh both directly, if microbiologically poor, and indirectly if residing, due to inadequate breeding set up and management, leads to modiﬁcations of water parameters (104). Therefore, we could speculate that the very low loads observed in this study are related to the good microbial quality of the feeds used, characterized only by a few SSOs with loads below 1.85 log cfu/g, as previously reported by Oteri et al. (33), as well as to the good conditions of the experimental aquaculture facility. Further studies are desirable to evaluate their application in a real scenario. CONCLUSIONS showed values below the maximum levels set by the FAO/WHO (96). Copper is an important micromineral in ﬁsh metabolism and is important for hemoglobin synthesis in many enzymatic reactions (98), but high copper concentrations can cause liver and kidney damages (99). The copper concentration determined in the ﬁsh ﬁlets of this study was below the values set by the standard regulatory limits allowed in ﬁsh samples (96). Nickel is an environmental factor that occurs at a very low level and can cause serious lung health problems such as lung cancer, ﬁbrosis, emphysema, cancer, and kidney disease (100). In this study, nickel values were considerably lower than the permitted levels set by the FAO/WHO (96). Finally, zinc, involved in various metabolic pathways such as protein synthesis, growth, immunity, and energy metabolism in ﬁsh (79, 97) showed a signiﬁcantly lower level in the ﬁsh fed with an inclusion of 25 and 50% of HIM. Overall, considering that several studies indicated that the content of these microelements in ﬁsh is inﬂuenced by surrounding water (77, 101) and by food source (102), which is the major source of elements such as iron, zinc, manganese and copper, the data would seem to highlight that the inclusion of HIM into the three diets maintains lower heavy metal levels than those recommended by various authorities (FAO, WHO, and EU). showed values below the maximum levels set by the FAO/WHO (96). Copper is an important micromineral in ﬁsh metabolism and is important for hemoglobin synthesis in many enzymatic reactions (98), but high copper concentrations can cause liver and kidney damages (99). The copper concentration determined in the ﬁsh ﬁlets of this study was below the values set by the standard regulatory limits allowed in ﬁsh samples (96). Nickel is an environmental factor that occurs at a very low level and can cause serious lung health problems such as lung cancer, ﬁbrosis, emphysema, cancer, and kidney disease (100). In this study, nickel values were considerably lower than the permitted levels set by the FAO/WHO (96). Finally, zinc, involved in various metabolic pathways such as protein synthesis, growth, immunity, and energy metabolism in ﬁsh (79, 97) showed a signiﬁcantly lower level in the ﬁsh fed with an inclusion of 25 and 50% of HIM. Frontiers in Nutrition \| www.frontiersin.org REFERENCES doi: 10.2527/jas.2015-9201 29. Barragan-Fonseca KB, Dicke M. Loon JJAV. Nutritional value of the black soldier ﬂy (Hermetiaillucens L) and its suitability as animal feed - a review. J Insects as Food Feed. (2017) 3:105–20. doi: 10.3920/JIFF2016.0055 13. van Huis A. Potential of insects as food and feed in assuring food security. Annu Rev Èntomol. (2013) 58:563–83. doi: 10.1146/annurev-ento-120811-153704 30. Meneguz M, Schiavone A, Gai F, Dama A, Lussiana C, Renna M, et al. Eﬀect of rearing substrate on growth performance, waste reduction eﬃciency and chemical composition of black soldier ﬂy (Hermetiaillucens) larvae. J Sci Food Agric. (2018) 98:5776–84. doi: 10.1002/jsfa.9127 14. Payne CLR, Dobermann D, Forkes A, House J, Josephs J, McBride A, et al. Insects as food and feed: European perspectives on recent research and future priorities. J Insectsas Food Feed. (2016) 2:269–76. doi: 10.3920/JIFF2016.0011 31. Spranghers T, Ottoboni M, Klootwijk C, Ovyn A, Deboosere S, De Meulenaer B, et al. Nutritionalcomposition of black soldierﬂy (Hermetiaillucens) prepupaereared on diﬀerent organic wastesubstrates. J Sci Food Agric. (2017) 97:2594–600. doi: 10.1002/jsfa.8081 15. Finke MD, Rojo S, Roos N, van Huis A, Yen AL. The European Food Safety Authority scientiﬁc opinion on a risk proﬁle related to production and consumption of insects as food and feed. J Insects as Food Feed. (2015) 1:245–7. doi: 10.3920/JIFF2015.x006 32. Liland NS, Biancarosa I, Araujo P, Biemans D, Bruckner CG, Waagbø R, et al. Modulation of nutrient composition of black soldier ﬂy (Hermetiaillucens) larvae by feeding seaweed-enriched media. PLoS ONE. (2017) 12:e0183188. doi: 10.1371/journal.pone.0183188 16. Acar U, Giannetto A, Giannetto D, Kesbiç OS, Yilmaz S, Romano A, et al. Evaluation of an innovative and sustainable pre-commercial compound as replacement of ﬁsh meal in diets for rainbow trout during pre-fattening phase: eﬀects on growth performances, haematological parameters and ﬁllet quality traits. Animals. (2021) 11:3547. doi: 10.3390/ani11123547 33. Oteri M, Di Rosa AR, Lo Presti V, Giarratana F, Toscano G, Chiofalo B. Black Soldier Fly Larvae Meal as Alternative to Fish Meal for Aquaculture Feed. Sustainability. (2021) 13:5447. doi: 10.3390/su13105447 34. Di Rosa AR, Oteri M, Lo Presti V, Chiofalo B. Feed insects for aquaculture. Use of Hermetiaillucens L. meal for Sparus aurata L.: Emerging trends of advanced sensor-based instruments (E-eye, E-nose, E-tongue) for ﬁsh quality. In: Taylor, Francis, editor. ASPA 24th Congress Book of Abstract. Padua: Italy (2021). p. 49. 17. Boukid F, Riudavets J. del Arco L, Castellari M. REFERENCES chitin degradation in juvenile turbot (Psetta maxima). Aquaculture. (2012) 364–5:345–52. doi: 10.1016/j.aquaculture.2012.08.041 1. Parisi G, Tulli F, Fortina R, Marino R, Bani P, Dalle Zotte A, et al. Protein hunger of the feed sector: the alternatives oﬀered by the plant world. Ital J Anim Sci. (2020) 19:1204–25. doi: 10.1080/1828051X.2020.1827993 1. Parisi G, Tulli F, Fortina R, Marino R, Bani P, Dalle Zotte A, et al. Protein hunger of the feed sector: the alternatives oﬀered by the plant world. Ital J Anim Sci. (2020) 19:1204–25. doi: 10.1080/1828051X.2020.1827993 19. Cullere M, Tasoniero G, Giaccone V, Miotti-Scapin R, Claeys E, De Smet S, et al. Black soldierﬂyasdietaryprotein source for broilerquails: apparentdigestibility, excretamicrobial load, feed choice, performance, carcass and meat traits. Animal. (2016) 10:1923–30. doi: 10.1017/S1751731116001270 Anim Sci. (2020) 19:1204–25. doi: 10.1080/1828051X.2020.1 2. EUMOFA. Fishmeal and Fish Oil, Maritime Aﬀairs and Fisheries, Production and Trade Flows in the EU. Brussels: EUMOFA (2021). 20. Barroso FG, De Haro C, Sánchez-Muros M-J, Venegas E, Martínez-Sánchez A, Pérez-Bañón C. The potential of various insect species for use as food for ﬁsh. Aquaculture. (2014) 422–3:193–201. doi: 10.1016/j.aquaculture.2013.12.024 3. Tacon AGJ, Hasan MR, Metian M. Demand and Supply of Feed Ingredients for Farmed Fish and Crustaceans: Trends and Prospects. FAO Fisheries and Aquaculture Technical Paper No 564. Rome: FAO (2011). 4. Hardy RW. Utilization of plant proteins in ﬁsh diets: Eﬀects of global demand and supplies of ﬁshmeal. Aquac Res. (2010) 41:770–6. doi: 10.1111/j.1365-2109.2009.02349.x 21. Yu GH, Chen YH, Yu ZN, Cheng P. Research progress on the larvae and prepupae of black soldier ﬂy Hermetiaillucens used as animal feedstuﬀ. Chin Bull Entomol. (2009) 46:41–5. 5. Turchini GM, Torstensen BE, Ng WK. Fish oil replacement in ﬁnﬁsh nutrition. Rev Aquac. (2009) 1:10–57. doi: 10.1111/j.1753-5131.2008.01001.x 22. Bondari K. Sheppard DC. SoldierﬂyHermetiaillucens L, as feed for channelcatﬁsh, Ictaluruspunctatus (Raﬁnesque), and blue tilapia, Oreochromisaureus (Steindachner). Aquacult Fish Manage. (1987) 18:209–20. doi: 10.1111/j.1365-2109.1987.tb00141.x 6. Montero D, Izquierdo M. Welfare and health of ﬁsh fed vegetable oils as alternative lipid sources to ﬁsh oil. In: Turchini GM, Ng WK, Tocher DR, editors. Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Boca Raton, FL: CRC Press (2010). p. 439–85. 23. European Commission. Commission Regulation (EU) 2017/893 amending Annexes I and IV to Regulation (EC) No. 999/2001 of the European Parliament and of the Council and Annexes X, XIV and XV to Commission Regulation (EU) No 142/2011 as regards the provisions on processed animal protein. REFERENCES OﬀJ Eur Union. (2017) 138:92–116. 7. Torrecillas S, Román L, Rivero-Ramírez F, Caballero MJ, Pascual C, Robaina L, et al. Supplementation of arachidonic acid rich oil in Europeanseabassjuveniles (Dicentrarchuslabrax) diets: Eﬀects on leucocytes and plasma fatty acid proﬁles, selected immune parameters and circulatingprostaglandinslevels. Fish Shellﬁsh Immunol. (2017) 64:437–45. doi: 10.1016/j.fsi.2017.03.041 24. Grigorakis K. Compositional and organoleptic quality of farmed and wild gilthead sea bream (Sparus aurata) and sea bass (Dicentrarchuslabrax) and factors aﬀecting it: a review. Aquaculture. (2007) 272:55–75. doi: 10.1016/j.aquaculture.2007.04.062 8. Belforti M, Gai F, Lussiana C, Renna M, Malfatto V, Rotolo L, et al. The meaning of the sentence is not clear; please rewrite or conﬁrm that the sentence is correct. S, Schiavone A Tenebrio molitor meal in rainbow trout (Oncorhynchus mykiss) diets: eﬀects on animal performance, nutrient digestibility and chemical composition of ﬁllets Ital. J Anim Sci. (2015) 14:4170. doi: 10.4081/ijas.2015.4170 25. Lock ER, Arsiwalla T, Waagbø R. Insect larvae meal as an alternative source of nutrients in the diet of Atlantic salmon (Salmo salar) postsmolt. AquacultNutr. (2015) 22:1202–13. doi: 10.1111/anu.12343 26. St-Hilaire S, Cranﬁll K, McGuire MA, Mosley EE, Tomberlin JK, Newton L, et al. Fish oﬀal recycling by the black soldier ﬂy produces a foodstuﬀhigh in Omega-3 fatty acids. J World AquacSoc. (2007) 38:309–13. doi: 10.1111/j.1749-7345.2007.00101.x 9. Belghit I, Liland NS, Gjesdal P, Biancarosa I, Menchetti E, Li Y, et al. Black soldier ﬂy larvae meal can replace ﬁsh meal in diets of sea- water phase Atlantic salmon (Salmo salar). Aquaculture. (2019) 503:609– 19. doi: 10.1016/j.aquaculture.2018.12.032 27. St-Hilaire S, Sheppard C, Tomberlin J, Irving S, Newton L, McGuire M, et al. Fly prepupae as a feedstuﬀfor rainbow trout, oncorhynchus mykiss. J World AquacSoc. (2007) 38:59–67. doi: 10.1111/j.1749-7345.2006.00073.x 10. Makkar HP, Tran G, Heuzé V, Ankers P. State-of-the-art on use of insects as animal feed. Anim Feed Sci Technol. (2014) 197:1–33. doi: 10.1016/j.anifeedsci.2014.07.008 28. Ewald N, Vidakovic A, Langeland M, Kiessling A, Sampels S, Lalander C. Fatty acid composition of black soldier ﬂy larvae (Hermetiaillucens) - possibilities and limitations for modiﬁcation through diet. Waste Manag. (2020) 102:40–7. doi: 10.1016/j.wasman.2019.10.014 11. Veldkamp T, Bosch G. Insects: a protein-rich feed ingredient in pig and poultry diets. AnimFron. (2015) 5:45–50. doi: 10.2527/af.2015-0019 12. Bovera F, Loponte R, Marono S, Piccolo G, Parisi G, Iaconisi V, et al. Use of Tenebrio molitor larvaemealasprotein source in broilerdiet: eﬀect on growth performance, nutrientdigestibility, carcass and meat traits. J Anim Sci. (2016) 94:639–47. ACKNOWLEDGMENTS The authors would like to thank all the staﬀof the IRBIM facility of the Messina Headquarter of CNR (Messina, Italy) for the technical support in ﬁsh management. May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 12 Sparus aurata Fed Insect Meal Oteri et al. REFERENCES (1991) 338:985–92. doi: 10.1016/0140-6736(91)91846-M 60. Tocher DR. Omega-3 long-chain polyunsaturated fatty acids and aquaculture in perspective. Aquaculture. (2015) 449:94– 107. doi: 10.1016/j.aquaculture.2015.01.010 41. Santos-Silva J, Bessa RJB, Santos-Silva F. Eﬀect of genotype, feeding system and slaughter weight on the quality of light lambs. II Fatty acid composition of meat. Livest Prod Sci. (2002) 77:187–92. doi: 10.1016/S0301-6226(02)00059-3 61. World Health Organization. Diet, Nutrition, and the Prevention of Chronic Diseases: Report of a Joint WHO/FAO Expert Consultation. Geneva: World Health Organization (2003). 42. Luciano G, Pauselli M, Servili M, Mourvaki E, Serra A, Monahan FJ, et al. Dietary olive cake reduces the oxidation of lipids, including cholesterol, in lamb meat enriched in polyunsaturated fatty acids. Meat Sci. (2013) 93:703–14. doi: 10.1016/j.meatsci.2012.11.033 62. Chen Y, Zhu X, Yang Y, Han D, Jin J, Xie S. Eﬀect of dietary chitosan on growth performance, haematology, immune response, intestine morphology, intestine microbiota and disease resistance in gibel carp (Carassius auratus gibelio). Aquacult Nutr. (2014) 20:532–46. doi: 10.1111/anu.12106 43. Must˘a¸tea G, Ungureanu EL, Iorga E. Protein acidic hydrolysis for amino acids analysis in food - progress over time: a short review. J Hyg Eng Des. (2019) 26:81–7. 63. Shiau SY Yu YP. Dietary supplementation of chitin and chitosan depresses growth in tilapia, Oreochromis niloticus×O. Aureaus Aquaculture. (1999) 179:439–46. doi: 10.1016/S0044-8486(99)00177-5 44. Dogan G, Osman EO. Determination of amino acid and fatty acid composition of goldband goatﬁsh [Upeneusmoluccensis (Bleeker, 1855)] ﬁshing from the Gulf of Antalya (Turkey). Int Aquat Res. (2017) 9:313– 27. doi: 10.1007/s40071-017-0179-9 64. Khoushab F, Yamabhai M. Chitin research revisited. Mar Drugs. (2010) 8:1988–2012. doi: 10.3390/md8071988 65. Palmegiano GB, Costanzo MT, Daprà F, Gai F, Galletta MG, Maricchiolo G, et al. Rice protein concentrate mealaspotentialdietaryingredient in practicaldiets for blackspotseabream (Pagellusbogaraveo). AnimPhysiolAnimNutr. (2007) 91:235– 9. doi: 10.1111/j.1439-0396.2007.00697.x 45. ISO 21528-2:2017. Microbiology of the Food Chain-Horizontal Method for the Detection and Enumeration of Enterobacteriaceae - Part 2: Colony-Count Technique. Geneva: International Organization for Standardization (2017). 46. ISO 15213:2003. Microbiology of Food and Animal Feeding Stuﬀs- Horizontal Method for the Enumeration of Sulﬁte-Reducing Bacteria Growing under Anaerobic Conditions. Geneva: International Organization for Standardization (2003). 66. Secci G, Parisi G. From farm to fork: lipid oxidation in ﬁsh products. A review Ital J Anim Sci. (2016) 15:124–36. doi: 10.1080/1828051X.2015.1128687 67. Weththasinghe P, Lagos L, Cortés M, Hansen JØ, Øverland M. REFERENCES Dietary Inclusion of Black Soldier Fly (HermetiaIllucens) Larvae Meal and Paste Improved Gut Health but Had Minor Eﬀects on Skin Mucus Proteome and Immune Response in Atlantic Salmon (Salmo Salar). Front Immunol. (2021) 12:599530. doi: 10.3389/ﬁmmu.2021.599530 47. ISO 6579-1:2017. Microbiology of the Food Chain-Horizontal Method for the Detection, Enumeration and Serotyping of Salmonella - Part 1: Detection of Salmonella spp. Geneva: International Organization for Standardization (2017). 48. ISO 21872-1:2017. Microbiology of the Food Chain-Horizontal Method for the determination of Vibrio spp. - Part 1: Detection of potentially enteropathogenic Vibrio parahaemolyticus, Vibrio cholerae and Vibrio vulniﬁcus. Geneva: International Organization for Standardization (2017). 68. Iaconisi V, Secci G, Sabatino G, Piccolo G, Gasco L, Papini AM. Eﬀect of mealworm (Tenebrio molitor L) larvaemeal on amino acid composition of giltheadseabream (Sparus aurata L) and rainbowtrout (Oncorhynchusmykiss W) ﬁllets. Aquaculture. (2019) 513:734403. doi: 10.1016/j.aquaculture.2019.734403 49. Gram L, Trolle G, Huss HH. Detection of speciﬁc spoilage bacteria from ﬁsh stored at low (0◦C) and high (20◦C) temperatures. Int J Food Microbiol. (1987) 4:65–72. doi: 10.1016/0168-1605(87)90060-2 69. Sheppard DC, Newton GL, Thompson SA, Savage S. A value added manure management system using the black soldier ﬂy. Bioresour Technol. (1994) 50:275–9. doi: 10.1016/0960-8524(94)90102-3 50. ISO 11290-1:2017. Microbiology of the Food Chain-Horizontal Method for the Detection and Enumeration of Listeria Monocytogenes and of Listeria spp. - Part 1: Detection Method. Geneva: International Organization for Standardization (2017). 70. Rust MB. Nutritional physiology. In: Halver JE, Hardy RW, editors. Fish Nutrition. New York, NY: Academic Press (2002). p. 368–446. 71. Abro R, Sundell K, Sandblom E, Sundh H, Brännäs E, Kiessling A, et al. Evaluation of chitinolytic activities and membrane integrity in guttissues of Arcticcharr (Salvelinusalpinus) fedﬁshmeal and zygomycetebiomass. Comp Biochem Physiol B BiochemMol Biol. (2014) 175:1–8. doi: 10.1016/j.cbpb.2014.06.003 51. Addinsoft. XLSTAT Statistical and Data Analysis Solution. New York, NY: Addinsoft (2022). 52. Food and Agriculture Organization. Food and Agriculture Organization of the United Nations Minerals. Rome: Food and Agriculture Organization (2016). 72. Krogdahl A, Hemre G, Mommsen T. Carbohydrates in ﬁsh nutrition: digestion and absorption in postlarval stages. AquacultNutr. (2005) 11:103–22. doi: 10.1111/j.1365-2095.2004.00327.x 53. Iaconisi V, Marono S, Parisi G, Gasco L, Genovese L, Maricchiolo G, et al. Dietaryinclusion of Tenebrio molitor larvaemeal: eﬀects on growth performance and ﬁnalqualitytreats of blackspotseabream (Pagellusbogaraveo). Aquaculture. (2017) 476:49–58. doi: 10.1016/j.aquaculture.2017.04.007 73. Lindsay GJH, Walton MJ, Adron JW, Fletcher TC, Cho CY, Cowey CB. REFERENCES Impact of diets including agro-industrial by-products on the fatty acid and sterol Proﬁles of larvae biomass from ephestiakuehniella, tenebrio molitor and hermetiaillucens. Insects. (2021) 12:672. doi: 10.3390/insects12080672 18. Kroeckel S, Harjes A-GE, Roth I, Katz H, Wuertz S, Susenbeth A, et al. When a turbot catches a ﬂy: evaluation of a pre-pupae-meal of the Black Soldier Fly (Hermetiaillucens) as ﬁshmeal substitute - Growth performance and 35. European Directive. Legislative Decree No. 26 implementing European Directive 2010/63/EU of the European Parliament and of the Council of 22 May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 13 Sparus aurata Fed Insect Meal Oteri et al. (Sparus aurata) feddietscontainingdiﬀerentlipid sources. Aquaculture. (2003) 225:325–40. doi: 10.1016/S0044-8486(03)00299-0 September 2010 on the protection of animals used for scientiﬁc purposes. Oﬀ J Ital Leg. (2014) 61:1–36. September 2010 on the protection of animals used for scientiﬁc purposes. Oﬀ J Ital Leg. (2014) 61:1–36. g 36. European Directive. European Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientiﬁc purposes. OﬀJ Eur. (2010) 276:33–79. 56. Pulido L, Secci G, Maricchiolo G, Gasco L, Gai F, Serra A, et al. Eﬀect of dietary black soldier ﬂy larvae meal on fatty acid composition of lipids and sn-2 position of triglycerides of marketable size gilthead sea bream ﬁllets. Aquaculture. (2022) 546:737351–8. doi: 10.1016/j.aquaculture.2021.737351 37. Oﬃcial methods of analysis of AOAC international. In: 21st Association of Oﬃcial Analytical Chemists. Gaithersburg, MD: Oﬃcial methods of analysis of AOAC international (2019). 57. HendersonRJ, Tocher DR. The lipid composition and biochemistry of freshwater ﬁsh. Prog Lipid Res. (1987) 26:281– 347. doi: 10.1016/0163-7827(87)90002-6 38. Folch J, Lees M, Stanley GHS A. Simple method for the isolation and puriﬁcation of total lipides from animal tissues. J BiolChem. (1957) 226:497– 509. doi: 10.1016/S0021-9258(18)64849-5 58. Sargent JR, Tocher DR, Bell JG. The lipids. In: Halver J, Hardy E, editors. Fish Nutrition. 3rd ed. San Diego, CA: Academic (2002). p. 181–257. 39. Christie WW. Preparation of ester derivatives of fatty acids for chromatographic analysis. In: Christie WW, editor. Advances in Lipid Methodology-Two. Dundee: Oily Press (1993). p. 69-111. 59. SeiliezI, PanseratS, CorrazeG, Kaushik S, BergotP. Cloning and nutritional regulation of a 16-desaturase-like enzyme in the marine teleost gilthead seabream (Sparus aurata). Comp Biochem Physiol B Biochem. (2003) 135:449–60. doi: 10.1016/S1096-4959(03)00111-8 40. Ulbricht TL, Southgate DA. Coronary heart disease: seven dietary factors. Lancet. REFERENCES Sivaperumal P, Sankar TV, Viswanathan Nair PG. Heavy metal concentrations in ﬁsh, shellﬁsh and ﬁsh products from internal markets of India vis-a-vis international standards. Food Chem. (2007) 102:612–20. doi: 10.1016/j.foodchem.2006.05.041 80. Rodehutscord M, Pfeﬀer E. Requirement for phosphorus in rainbow trout (Oncorhynchus mykiss) growing from 50 to 200 g. Water Sci Technol. (1995) 31:137–41. doi: 10.2166/wst.1995.0370 81. Pimentel-Rodriguez AM, Oliva-Teles A. Phosphorus requirements of gilthead sea bream (Sparus aurata L.) juveniles. Aquac Res. (2001) 32:157–61. doi: 10.1046/j.1355-557x.2001.00013.x 99. Ikem A, Egiebor NO. Assessment of trace elements in canned ﬁshes (mackerel, tuna, salmon, sardines and herrings) marketed in Georgia and Alabama (United States of America). J Food Comp Anal. (2005) 18:771–87. doi: 10.1016/j.jfca.2004.11.002 82. Al-Yousuf MH, El-Shahawi MS, Al-Ghais SM. Trace metals in liver, skin and muscle of Lethrinuslentjan ﬁsh species in relation to body length and sex. Sci Total Environ. (2000) 256:87–94. doi: 10.1016/S0048-9697(99)00363-0 100. Forti E, Salovaara S, Cetin Y, Bulgheroni A, Tessadri R, Jennings P, et al. In vitro evaluation of the toxicity induced by nickel soluble and particulate forms in human airway epithelial cells. Toxicol In Vitro. (2011) 25:454– 61. doi: 10.1016/j.tiv.2010.11.013 83. Borucka-Jastrzebska E, Kawczuga D, Rajkowska M, Protasowicki M. Levels of micro-elements (Cu, Zn, Fe) and macroelements (Mg, Ca) in freshwater ﬁsh. J Elementol. (2009) 14:437–47. doi: 10.5601/jelem.2009.14.3.02 101. Oliva-Teles A. Nutrition and health of aquaculture ﬁsh. J Fish Dis. (2012) 35:83–108. doi: 10.1111/j.1365-2761.2011.01333.x 84. Perkowska A, Protasowicki M. Cadmium and lead in ﬁshes in selected elements of Swidwie Lake ecosystem. Acta IchthyolPiscat. (2000) 30:71–84. doi: 10.3750/AIP2000.30.2.07 102. Tang Q, Wang C, Xie C, Jin J, Huang Y. Dietary available phosphorus aﬀected growth performance, body composition, and hepatic antioxidant property of juvenile yellow catﬁsh (Pelteobagrusfulvidraco). Sci World J. (2012) 2012:987570. doi: 10.1100/2012/987570 85. Roméo M, Siau Y, Sidoumou Z, Gnassia-Barelli M. Heavy metal distribution in diﬀerent ﬁsh species from the Mauritania coast. Sci Total Environ. (1999) 232:169–75. doi: 10.1016/S0048-9697(99)00099-6 103. de Bruijn I, Liu Y, Wiegertjes GF, Raaijmakers JM. Exploring ﬁsh microbial communities to mitigate emerging diseases in aquaculture. FEMS Microbiol Ecol. (2018) 94:ﬁx161. doi: 10.1093/femsec/ﬁx161 86. Kumar V, Makkar HPS, Becker K. Dietary inclusion of detoxiﬁed Jatropha curcas kernel meal, eﬀects on growth performance and metabolic eﬃciency in common carp, Cyprinus carpio (Linnaeus). Fish Physiol Biochem. (2010) 36:1159–70. doi: 10.1007/s10695-010-9394-7 104. Rurangwa E, Verdegem MC. Microorganisms in recirculating aquaculture systems and their management. RevAquac. (2015) 7:117–30. doi: 10.1111/raq.12057 87. REFERENCES The growth of rainbow trout (Salmo gairdneri) given diets containing chitin and its relationship to chitinolytic enzymes and chitin digestibility. Aquaculture. (1984) 37:315–34. doi: 10.1016/0044-8486(84)90297-7 54. Sealey WM, Gaylord TG, Barrows FT, Tomberlin JK, McGuire MA, Ross C. Sensoryanalysis of rainbowtrout, Oncorhynchusmykiss, fedenriched black soldierﬂyprepupae, Hermetiaillucens. J World AquacultSoc. (2011) 42:34– 45. doi: 10.1111/j.1749-7345.2010.00441.x 74. Zhang Y, Zhou Z, Liu Y, Cao Y, He S, Huo F, et al. High-yield production of a chitinase from Aeromonas veronii B565 as a potential feed supplement for warm-water aquaculture. ApplMicrobiolBiotechnol. (2014) 98:1651–62. doi: 10.1007/s00253-013-5023-6 55. Caballero MJ, Izquierdo MS, Kjørsvik E, Montero D, Socorro J, Fernández AJ, et al. Morphologicalaspects of intestinalcells from giltheadseabream May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 14 Sparus aurata Fed Insect Meal Oteri et al. 75. Oliva-Teles A, Enes P, Peres H. Replacing ﬁshmeal and ﬁsh oil in industrial aquafeed. In: Davis A, editor. Feed and Feeding Practices in Aquaculture. Waltham, MA: Woodhead Publishing limited (2015). p. 203–33. 93. Bakshi A, Panigrahi AK. A comprehensive review on chromium induced alterations in fresh water ﬁshes. Toxicol Rep. (2018) 5:440–7. doi: 10.1016/j.toxrep.2018.03.007 94. European Commission. Commission Regulation (EC) No. 629/2008 of 2 July 2008 Amending Regulation (EC) No. 1881/2006 Setting Maximum Levels for Certain Contaminants in Foodstuﬀs. Brussels: European Commission (2008). p. L173/6–9. 76. Magalhães R, Sánchez-López A, Leal RS, Martínez-Llorens S, Oliva-Teles A, Peres H. Black soldier ﬂy (Hermetiaillucens) pre-pupae meal as a ﬁsh meal replacement in diets for European seabass (Dicentrarchuslabrax). Aquaculture. (2017) 476:79–85. doi: 10.1016/j.aquaculture.2017.04.021 95. Lall SP. The minerals. In: Halver JE, Hardy RW, editors. Fish Nutrition. London: Academic Press (2002). p. 259–308. 77. National Research Council. Minerals. In: Nutrient Requirements of Fish and Shrimp. Washington, CO: The National Academies Press (2011). p. 163–85. 96. FAO/WHO. Evaluation of Certain Food Additives and the Contaminants Mercury, Lead and Cadmium. Geneva: WHO Technical Report, Series No. 505 (1989). 78. Wu G, Wu Z, Dai Z, Yang Y, Wang W, Liu C, et al. Dietary requirements of “nutritionally non-essential amino acids” by animals and humans. Amino Acids. (2013) 44:1107–13. doi: 10.1007/s00726-012-1444-2 f Mercury, Lead and Cadmium. Geneva: WHO Technical Report, Series No. 505 (1989). 79. Food and Agriculture Organization of the United Nations. Aquaculture Feed and Fertilizer Resources Information System. Rome: Food and Agriculture Organization (2017). 97. Watanabe T, Kiron V, Satoh S. Trace minerals in ﬁshnutrition. Aquaculture. (1997) 151:185–207. doi: 10.1016/S0044-8486(96)01503-7 98. Frontiers in Nutrition \| www.frontiersin.org REFERENCES Yi Y, Yang Z, Zhang S. Ecological risk assessment of heavy metals in sediment and human health risk assessment of heavy metals in ﬁshes in the middle and lower reaches of the Yangtze River basin. Environ Pollut. (2011) 159:2575–85. doi: 10.1016/j.envpol.2011.06.011 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. 88. Ip CCM Li XD, Zhang G, Wong CSC, Zhang WL. Heavy metal and Pb isotopic compositions of aquatic organisms in the Pearl River Estuary, South China. EnvironPollut. (2005) 138:494–504. doi: 10.1016/j.envpol.2005.04.016 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. 89. Tuzen M, Soylak M. Determination of trace metals in canned ﬁsh marketed in Turkey. Food Chem. (2007) 101:1378–82. doi: 10.1016/j.foodchem.2006.03.044 90. Gulec AK, Yildrim NC, Danabas D, Yildirim N. Some Haematological and BiochemicalParameters in Common Carp (Cyprinus carpio L., 1758) in Munzur River, Tunceli, Turkey. Asian J Chem. (2011) 23:910–2. Available online at: https://hdl.handle.net/20.500.12406/428 Copyright © 2022 Oteri, Chiofalo, Maricchiolo, Toscano, Nalbone, Lo Presti and Di Rosa. 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. 91. Gu YG, Lin Q, Huang HH, Wang LG, Ning JJ, Du FY. Heavy metals in ﬁsh tissues/stomach contents in four marine wild commercially valuable ﬁsh species from the western continental shelf of South China Sea. Marine Poll Bull. (2017) 114:1125–9. doi: 10.1016/j.marpolbul.2016.10.040 92. Aslam S, Yousafzai AM. Chromium toxicity in ﬁsh: a review article. J Entomol Zool Stud. (2017) 5:1483–8. May 2022 \| Volume 9 \| Article 896552 Frontiers in Nutrition \| www.frontiersin.org 15
https://openalex.org/W4322626696	https://link.springer.com/content/pdf/10.1007/s40574-023-00349-9.pdf	English	null	Prime numbers in typical continued fraction expansions	Bollettino della Unione matematica italiana	2,023	cc-by	9,887	Bollettino dell’Unione Matematica Italiana (2023) 16:259–274 https://doi.org/10.1007/s40574-023-00349-9 Bollettino dell’Unione Matematica Italiana (2023) 16:259–274 https://doi.org/10.1007/s40574-023-00349-9 Mathematics Subject Classiﬁcation Primary 11K50 · 28D05 · 37A25 · 37C30 · 37A50 Mathematics Subject Classiﬁcation Primary 11K50 · 28D05 · 37A25 · 37C30 · 37A50 Mathematics Subject Classiﬁcation Primary 11K50 · 28D05 · 37A25 · 37C30 · 37A50 Tanja I. Schindler1 · Roland Zweimüller1 Received: 28 September 2022 / Accepted: 6 February 2023 / Published online: 28 February 2023 © The Author(s) 2023 1 Fakultät für Mathematik, Universität Wien, Oskar-Morgenstern-Platz 1, 1090 Wien, Austria Abstract We study, from the viewpoint of metrical number theory and (inﬁnite) ergodic theory, the probabilistic laws governing the occurrence of prime numbers as digits in continued fraction expansions of real numbers. Keywords Continued fractions · Prime numbers · Stochastic limit theorems 1 Introduction While many analogous versions of the following statements have directly been proven for the continued fraction digits, today, it is possible to deduce them or the version for the prime digits from more general theorems. The paper is structured as follows. In Sect.2 we give results about the pointwise behaviour of the prime digits in the continued fraction expansion and in Sect.3 about their distributional behaviour. Then we start with the proof sections, i.e. in Sect.4 we state general results about the Gauss map and in Sects.5 and 6 we give the proofs of the results from Sects.2 and 3 respectively. 1 Introduction Ever since Gauss [10] declared his interest in the intriguing statistical properties of sequences of digits an(x), n ≥1, in the continued fraction (CF) expansion of real numbers x ∈I := (0, 1], x = [a1(x), a2(x), . . .] = 1 a1(x) + 1 a2(x) + 1 a3(x) + · · · (and, in particular, mentioned that this led to questions he could not answer), the metrical theory of continued fractions has attracted the attention of many mathematicians. In the present paper we will be interested in the prime digits of x, i.e. those an(x) which happen to belong to the set P of prime numbers. To single them out, we deﬁne, for x ∈I and n ≥1, a′ n(x) := 1P(an(x)) · an(x) = an(x) if an(x) ∈P, 0 otherwise. (There is hardly any danger of misinterpreting this phonetically perfect symbol as a deriva- tive.) The purpose of this note is to point out that it is in fact possible—with the aid of the (There is hardly any danger of misinterpreting this phonetically perfect symbol as a deriva- tive.) The purpose of this note is to point out that it is in fact possible—with the aid of the 1 Fakultät für Mathematik, Universität Wien, Oskar-Morgenstern-Platz 1, 1090 Wien, Austria 123 260 T. I. Schindler, R. Zweimüller prime number theorem and recent work in (inﬁnite) ergodic theory and in the probability the- ory of dynamical systems—to derive a lot of information about the occurrences and values of prime digits in CF-expansions of (Lebesgue) typical numbers. Besides stating the theorems themselves it is also our aim to show some newer more general results in ergodic theory in action. While many analogous versions of the following statements have directly been proven for the continued fraction digits, today, it is possible to deduce them or the version for the prime digits from more general theorems. prime number theorem and recent work in (inﬁnite) ergodic theory and in the probability the- ory of dynamical systems—to derive a lot of information about the occurrences and values of prime digits in CF-expansions of (Lebesgue) typical numbers. Besides stating the theorems themselves it is also our aim to show some newer more general results in ergodic theory in action. 2 Main results—pointwise matters Remark 2.1 The exponent 0.475 in c) comes from estimates for the error term in the prime number theorem and might be improved by future research. Example 2.1 A straightforward calculation shows that ple 2.1 A straightforward calculation shows that λ({a′ n > n logγ 2 n i.o.}) = 1 if γ ≤1, 0 otherwise, and this remains true if a′ n is replaced by M′ n. We thus ﬁnd that and this remains true if a′ n is replaced by M′ n. We thus ﬁnd that lim n→∞ log a′ n −log n log3 n = lim n→∞ log M′ n −log n log3 n = 1 a.e. As a consequence of Theorem 2.1 b), observing that the series n≥1 1/(bn log bn) con- verges iff n≥1 1/(ρ bn log(ρ bn)) converges for every ρ ∈(0, ∞), we get Corollary 2.1 If (bn)n≥1 is non-decreasing, then Corollary 2.1 If (bn)n≥1 is non-decreasing, then lim n→∞ a′ n bn = lim n→∞ M′ n bn = ∞a.e. if n≥1 1 bn log bn = ∞, 0 a.e. otherwise. (2.3) (2.3) In particular, lim n→∞ a′ n n log2 n = lim n→∞ M′ n n log2 n = ∞a.e. (2.4) (2.4) A convenient condition for the criterion above is provided by Lemma 2.1 Let (bn)n≥1 be a sequence in (1, ∞) for which bn/n increases. Then A convenient condition for the criterion above is provided by Lemma 2.1 Let (bn)n≥1 be a sequence in (1, ∞) for which bn/n increases. Then lim n→∞ n log2 n bn > 0 implies n≥1 1 bn log bn = ∞. As in the case of the full digit sequence (an)n≥1, the peculiar properties of (a′ n)n≥1 are due to the fact that these functions are not integrable. A general fact for non-integrable non-negative stationary sequences is the non-existence of a non-trivial strong law of large numbers, made precise in a), c) and d) of the next result, where c) is in the spirit of [22]. However, it is sometimes possible to recover a meaningful limit by trimming, i.e. by removing maximal terms. In the case of (an)n≥1, this was ﬁrst pointed out in [8]. Assertion b) below gives the proper version for the (a′ n)n≥1. 2 Main results—pointwise matters We ﬁrst consider questions about the pointwise behaviour of the sequence (a′ n)n≥1 on I. Throughout, λ denotes Lebesgue measure on (the Borel σ-ﬁeld BI of) I, and almost every- where (a.e.) is meant w.r.t. λ. For the sake of completeness, we also include a few easy basic facts, e.g. that for a.e. x ∈I, the proportion of those k ∈{1, . . . , n} for which ak(x) is prime converges: Proposition 2.1 (Asymptotic frequency of prime digits)We have lim n→∞ 1 n n k=1 1P ◦ak = 1 log 2 log p∈P 1 + 1 p(p + 2) a.e. The results to follow can best be understood (and proved) by regarding (a′ n) as a stationary sequence with respect to the Gauss measure (cf. §4 below). The ﬁrst statement of the next theorem is parallel to the classical Borel-Bernstein theorem (cf. [5, 7]) the third and fourth statements are in accordance with [16]. As usual, i.o. is short for "inﬁnitely often", i.e. for "inﬁnitely many indices". We denote the iterated logarithms by log1 := log and logm+1 := log ◦logm, m ≥1. Furthermore, we deﬁne the maximal entry M′ n := max1≤k≤n a′ k, n ≥1. Theorem 2.1 (Pointwise growth of prime digits and maxima) a) Assume that (bn)n≥1 is a sequence in (1, ∞). Then λ({a′ n > bn i.o.}) = 1 if n≥1 1 bn log bn = ∞, 0 otherwise. (2.1) (2.1) b) Moreover, if (bn)n≥1 is non-decreasing, then b) Moreover, if (bn)n≥1 is non-decreasing, then b) Moreover, if (bn)n≥1 is non-decreasing, then λ({a′ n > bn i.o.}) = λ({M′ n > bn i.o.}). (2.2) (2.2) c) Let (cn)n≥1 and (dn)n≥1 be sequences in (1, ∞) with dn →∞and cn ≤d0.475 n for large n. Then λ({dn ≤a′ n ≤dn(1 + 1/cn) i.o.}) = 1 if n≥1 1 cndn log(dn) = ∞. 0 otherwise. d) Let (dn)n≥1 be a sequence of primes, then d) Let (dn)n≥1 be a sequence of primes, then λ({a′ n = dn i.o.}) = 1 if n≥1 1 d2n = ∞, 0 otherwise. 123 Prime numbers in typical... 261 Remark 2.1 The exponent 0.475 in c) comes from estimates for the error term in the prime number theorem and might be improved by future research. 2 Main results—pointwise matters Theorem 2.2 (Strong laws of large numbers) a) The prime digits satisfy Theorem 2.2 (Strong laws of large numbers) a) The prime digits satisfy Theorem 2.2 (Strong laws of large numbers) a) The prime digits satisfy lim n→∞ 1 n n k=1 a′ k = ∞a.e. lim n→∞ 1 n n k=1 a′ k = ∞a.e. b) Subtracting M′ n, we obtain a trimmed strong law, b) Subtracting M′ n, we obtain a trimmed strong law, lim n→∞ log 2 n log2 n n k=1 a′ k −M′ n = 1 a.e. (2.5) (2.5) c) For sequences (bn)n≥1 in (1, ∞) satisfying bn/n ↗∞as n →∞, c) For sequences (bn)n≥1 in (1, ∞) satisfying bn/n ↗∞as n →∞, lim n→∞ 1 bn n k=1 a′ k = ∞a.e. iff n≥1 1 bn log bn = ∞, (2.6) (2.6) 123 262 T. I. Schindler, R. Zweimüller erwise lim n→∞ 1 bn n k=1 a′ k = 0 a.e. (2.7) (2.7) d) But, deﬁning n( j) := e j log2 j, j ≥1, and d′ n := n( j) log2 n( j)/ log 2 for n ∈(n( j − 1), n( j)] gives a normalizing sequence for which d) But, deﬁning n( j) := e j log2 j, j ≥1, and d′ n := n( j) log2 n( j)/ log 2 for n ∈(n( j − 1), n( j)] gives a normalizing sequence for which d) But, deﬁning n( j) := e j log2 j, j ≥1, and d′ n := n( j) log2 n( j)/ log 2 for n ∈(n( j − 1), n( j)] gives a normalizing sequence for which lim n→∞ 1 d′n n k=1 a′ k = 1 a.e. (2.8) (2.8) The trimmed law from b) shows that the bad pointwise behaviour described in c) is due to a few exceptionally large individual terms a′ n which, necessarily, have to be of the order of the preceding partial sum n−1 k=1 a′ k. In fact, almost surely, the partial sum will inﬁnitely often be of strictly smaller order than the following term, see statement a) below. We can also ask whether, or to what extent, the terms from the thinner sequence (a′ n)n≥1 come close to the partial sums (n−1 k=0 ak)n≥1 of the unrestricted one. The answer is given by the dichotomy rule in statement b) of the next result. 2 Main results—pointwise matters for ρ > 1 while lim n→∞ a′ n logρ a′ n n−1 k=1 ak = 0 a.e. for ρ ≤1. lim n→∞ a′ n log a′ n n−1 k=1 ak = ∞a.e. for ρ > 1 while lim n→∞ a′ n logρ a′ n n−1 k=1 ak = 0 a.e. for ρ ≤1. b) In case g(t) := t log t/ logγ 2 t, γ ∈R, we ﬁnd for γ ≤1 b) In case g(t) := t log t/ logγ 2 t, γ ∈R, we ﬁnd for γ ≤1 lim n→∞ a′ n log a′ n/ log2 a′ n n−1 k=1 ak = ∞a.e. while, for γ > 1, while, for γ > 1, lim n→∞ a′ n log a′ n/ logγ 2 a′ n n−1 k=1 ak = 0 a.e. On the other hand, if we look at primes to some power γ we obtain—as a counterpart to Theorem 2.2 b)—the following result: Theorem 2.4 a) For γ < 1 there exists Kγ > 0 such that Theorem 2.4 a) For γ < 1 there exists Kγ > 0 such that Theorem 2.4 a) For γ < 1 there exists Kγ > 0 such that lim n→∞ n k=1 a′ k γ n = Kγ < ∞a.e. lim n→∞ n k=1 a′ k γ n = Kγ < ∞a.e. b) Let σ := σ(n,x) ∈Sn be a pointwise permutation, i.e. σ : I × {1, . . . , n} →{1, . . . , n}, such that a′ σ(1) ≥. . . ≥a′ σ(n) and Sk n := n j=k+1 a′ σ( j). If γ > 1, then for all (bn) ∈NN fulﬁlling bn = o(n1−ϵ) for some ϵ > 0 and b) Let σ := σ(n,x) ∈Sn be a pointwise permutation, i.e. σ : I × {1, . . . , n} →{1, . . . , n}, such that a′ σ(1) ≥. . . ≥a′ σ(n) and Sk n := n j=k+1 a′ σ( j). If γ > 1, then for all (bn) ∈NN fulﬁlling bn = o(n1−ϵ) for some ϵ > 0 and lim n→∞ bn log log n = ∞ (2.11) we have b lim n→∞ bn log log n = ∞ (2.11) lim n→∞ bn log log n = ∞ (2.11) we have lim n→∞ Sbn n dn = 1 a.e. 2 Main results—pointwise matters We shall tacitly interpret real sequences (gn)n≥0 as functions on R+ via t −→g[t], and write g(t) ∼h(t) as t →∞if g(t)/h(t) →1. Moreover, g(t) ≍h(t) means 0 < limt→∞g(t)/h(t) ≤limt→∞g(t)/h(t) < ∞. Theorem 2.3 (Relative size of digits and partial sums) a) We have Theorem 2.3 (Relative size of digits and partial sums) a) We have a) We have lim n→∞ a′ n n−1 k=1 a′ k = ∞a.e. (2.9) (2.9) Generally, for functions g : [0, ∞) →(3, ∞) fulﬁlling g(η(t)) ≍g(t) if η(t) ∼t as t →∞, we have lim n→∞ g(a′ n) n−1 k=1 a′ k = ∞a.e. iff ∞ c g(y) log2 g(y) dy y2 log y = ∞, (2.10) (2.10) while otherwise while otherwise lim n→∞ g(a′ n) n−1 k=1 a′ k = 0 a.e. b) In contrast, comparing to the unrestricted digit sum n−1 k=1 ak, one has lim n→∞ a′ n n−1 k=1 ak = 0 a.e. lim n→∞ a′ n n−1 k=1 ak = 0 a.e. lim n→∞ a′ n n−1 k=1 ak = 0 a.e. Generally, for functions g : [0, ∞) →(3, ∞) fulﬁlling g(η(t)) ≍g(t) if η(t) ∼t as t →∞, we have Generally, for functions g : [0, ∞) →(3, ∞) fulﬁlling g(η(t)) ≍g(t) if η(t) ∼t as t →∞, we have lim n→∞ g(a′ n) n−1 k=1 ak = ∞a.e. iff ∞ c g(y) log g(y) dy y2 log y = ∞, while otherwise lim n→∞ g(a′ n) n−1 k=0 ak = 0 a.e. 123 Prime numbers in typical... 263 c) Turning to a comparison of partial sums, we ﬁnd that lim n→∞ n k=1 a′ k n k=1 ak = 0 a.e. lim n→∞ n k=1 a′ k n k=1 ak = 0 a.e. Remark 2.2 A broad class of functions which satisfy g(η(t)) ≍g(t) if η(t) ∼t as t →∞, are the regularly varying functions. Recall that a measurable function g : (L, ∞) →(0, ∞) is regularly varying of index ρ ∈R at inﬁnity, written g ∈Rρ, if g(ct)/g(t) →cρ as t →∞ for all c > 0 (see Chapter 1 of [6] for more information). Whether or not the integrals diverge can easily be checked for many speciﬁc g’s: Example 2.2 a) Taking g(t) := t logρ t, ρ ∈R, part b) gives lim n→∞ a′ n log a′ n n−1 k=1 ak = ∞a.e. 3 Main results—distributional matters The second set of results we present focuses on the distributions of (various functions of) the digits a′ n. If (M, d) is a separable metric space with Borel σ-ﬁeld BM, a sequence (νn)n≥1 of probability measures on (M, BM) converges weakly to the probability measure ν on (M, BM), written νn ⇒ν, if the integrals of bounded continuous function ψ : M →R converge, i.e. ψ dνn −→ ψ dν as n →∞. If Rn : I →M, n ≥1, Borel measurable functions and ν a Borel probability on M (or R another random element of M, not necessarily deﬁned on I, with distribution ν) then (Rn)n≥1 converges in distribution to ν (or to R) under the probability measure P on BI , if the distributions P ◦R−1 n of the Rn w.r.t. P converge weakly to ν. Explicitly specifying the underlying measure, we denote this by Rn P ⇒ν or Rn P ⇒R. For sequences (Rn) deﬁned on an ergodic dynamical system, it is often the case that a distributional limit theorem Rn P ⇒R automatically carries over to a large collection of other probability measures: strong distributional convergence, written Rn L(λ) ⇒ν or Rn L(λ) ⇒R, means that Rn P ⇒R for all probability measures P ≪λ, see [24]. means that Rn P ⇒R for all probability measures P ≪λ, see [24]. We start by giving a counterpart to Theorem 2.4 for weak convergence, where b) is in the spirit of [19]. Theorem 3.1 a) For γ < 1 there exists Kγ > 0 such that Theorem 3.1 a) For γ < 1 there exists Kγ > 0 such that Theorem 3.1 a) For γ < 1 there exists Kγ > 0 such that n k=1 a′ k γ L(λ) K Theorem 3.1 a) For γ < 1 there exists Kγ > 0 such that n k=1 a′ k γ n L(λ) ⇒Kγ . n k=1 a′ k γ n L(λ) ⇒Kγ . b) For the case γ = 1 we have b) For the case γ = 1 we have n k=1 a′ k n log2 n L(λ) ⇒log 2. c) If γ > 1 we have c) If γ > 1 we have Sbn n dn L(λ) ⇒1, where Sbn n is deﬁned as in Theorem 2.4, (dn) is given as in (2.13) and limn→∞bn = ∞and bn = o(n1−ϵ). 2 Main results—pointwise matters (2.12) where dn ∼ 1 (γ −1)(log 2)γ · nγ b1−γ n (log n)γ . (2.13) (2.11) we have lim n→∞ Sbn n dn = 1 a.e. (2.12) where dn ∼ 1 (γ −1)(log 2)γ · nγ b1−γ n (log n)γ . (2.13) lim n→∞ Sbn n dn = 1 a.e. (2.12) where dn ∼ 1 (γ −1)(log 2)γ · nγ b1−γ n (log n)γ . (2.13) (2.13) Remark 2.3 It is not proven that a trimming rate slower than the one given in (2.11) is possible. However, by [11] one can deduce that for i.i.d. random variables with the same distribution function and bn ≍log log n a strong law of large numbers as in (2.12) is no longer possible. Remark 2.3 It is not proven that a trimming rate slower than the one given in (2.11) is possible. However, by [11] one can deduce that for i.i.d. random variables with the same distribution function and bn ≍log log n a strong law of large numbers as in (2.12) is no longer possible. However, if we only ask for convergence in probability, the picture looks much simpler and we refer the reader to Theorem 3.1 in the next section. 264 T. I. Schindler, R. Zweimüller 3 Main results—distributional matters Remark 3.1 Indeed by [17] the stronger result of convergence in mean follows for c). It is not proven that for the situation in c) convergence in probability can not hold for a lightly trimmed sum, i.e. a sum from which only a ﬁnite number of large entries, being independent of n is removed. However, it follows from [1] that n k=1(a′ k)γ normed by the right norming sequence converges to a non-degenerate Mittag-Lefﬂer distribution if γ > 1. On the other hand, by [18] it follows that light trimming does not have any inﬂuence on distributional convergence if the random variables considered are i.i.d. As we have seen in the previous section, the maximum M′ n has a large inﬂuence then the whole system, in the following we will give its distributional convergence. We let denote a positive random variable with Pr[ ≤y] = e−1/y, y > 0 and get the following counterpart to [21]. 123 Prime numbers in typical... 265 Theorem 3.2 (Distributional convergence of M′ n )The maximum M′ n of the prime digits con- verges in distribution, Theorem 3.2 (Distributional convergence of M′ n )The maximum M′ n of the prime digits con- verges in distribution, log 2 log n n · M′ n L(λ) ⇒ as n →∞. (3.1) (3.1) A related classical topic, introduced by Doeblin [9], is the Poissonian nature of occurrences of very large CF-digits. For l ≥1 let ϕl = ϕl,1 := inf{k ≥1 : ak ≥l}, the ﬁrst position in the CF-expansion at which a digit ≥l shows up, and ϕl,i+1 := inf{k ≥1 : aϕl,i+k ≥l} the distance between the ith and (i + 1)st occurrence. Deﬁning l : I →[0, ∞]N as l := (ϕl,1, ϕl,2, . . .) and letting Exp denote an i.i.d. sequence of normalized exponentially distributed random variables, we can express this classical result by stating that 1 log 2 1 l · l λ ⇒Exp as l →∞. 1 log 2 1 l · l λ ⇒Exp as l →∞. Turning to prime digits, we shall consider the corresponding quantities ϕ′ l,i with ϕ′ l,0 := 0 and ϕ′ l,i+1 := inf{k ≥1 : a′ ϕ′ l i+k ≥l}, i ≥0, and the processes ′ l := (ϕ′ l,1, ϕ′ l,2, . . 3 Main results—distributional matters .) Turning to prime digits, we shall consider the corresponding quantities ϕ′ l,i with ϕ′ l,0 := 0 and ϕ′ l,i+1 := inf{k ≥1 : a′ ϕ′ l i+k ≥l}, i ≥0, and the processes ′ l := (ϕ′ l,1, ϕ′ l,2, . . .) , of distances between consecutive occurrences of prime digits of size at least l. In fact, we also provide reﬁned versions of the limit theorem which show that, asymptotically, both the relative size compared to l of such a large prime digit a′ ϕ′ l,i and its residue class for a given modulus m, are stochastically independent of the positions ϕ′ l,i at which they occur. (These statements are parallel to Propositions 10.1 and 10.2 of [25]. A (q1, . . . , qd)-Bernoulli sequence is an iid sequence of random variables which can assume d different values with respective probabilities q1, . . . , qd.) Theorem 3.3 (Poisson limits for large prime CF-digits) The sequences ′ l of positions at which large prime digits occur satisfy the following. a) Their distances converge to an i.i.d. sequence of exponential variables, 1 log 2 1 l logl · ′ l L(λ) ⇒Exp as l →∞. (3.2) (3.2) b) Take any ϑ ∈(0, 1), let ψ′ l,i be the indicator function of {a′ ϕ′ l,i ≥l/ϑ} and set ′ l := (ψ′ l,1, ψ′ l,2, . . .), which identiﬁes those prime digits ≥l which are in fact ≥l/ϑ. Then b) Take any ϑ ∈(0, 1), let ψ′ l,i be the indicator function of {a′ ϕ′ l,i ≥l/ϑ} and set ′ l := (ψ′ l,1, ψ′ l,2, . . .), which identiﬁes those prime digits ≥l which are in fact ≥l/ϑ. Then 1 log 2 1 l logl · ′ l, ′ l L(λ) ⇒( Exp, ′) as l →∞, (3.3) (3.3) where ( Exp, ′) is an independent pair with ′ a (1 −ϑ, ϑ)-Bernoulli sequence. c) Fix an integer m ≥2. For l > m deﬁne υ′ l,i : I →{ j ∈{1, . . . , m} : j relatively prime to m} by υ′ l,i(x) := j if a′ ϕ′ l,i (x) ≡j mod m, so that ϒ′ l := (υ′ l,1, υ′ l,2, . . .) identiﬁes the residue classes mod m of the prime digits a′ ϕ′ l,i . 3 Main results—distributional matters Then, for Sn:= n k=1 1Ak the following central limit theorem holds: Sn − Sn dμG L(λ) ⇒N as n →∞ Then, for Sn:= n k=1 1Ak the following central limit theorem holds: Sn − Sn dμG Sn − Sn dμG 2 L(λ) ⇒N as n →∞. 3 Main results—distributional matters Then where ( Exp, ′) is an independent pair with ′ a (1 −ϑ, ϑ)-Bernoulli sequence. c) Fix an integer m ≥2. For l > m deﬁne υ′ l,i : I →{ j ∈{1, . . . , m} : j relatively prime to m} by υ′ l,i(x) := j if a′ ϕ′ l,i (x) ≡j mod m, so that ϒ′ l := (υ′ l,1, υ′ l,2, . . .) identiﬁes the residue classes mod m of the prime digits a′ ϕ′ l,i . Then 1 log 2 1 l logl · ′ l, ϒ′ l L(λ) ⇒( Exp, ϒ′) as l →∞, (3.4) (3.4) where ( Exp, ϒ′) is an independent pair with ϒ′ a 1 φ(m), . . . , 1 φ(m) -Bernoulli sequence. (Here φ(m) denotes the Euler totient.) We ﬁnally look at the distribution of a function which counts how many a′ n fall into particular sets An giving a limit theorem in the spirit of [16, 20]. We let N denote a positive random variable with Pr[N ≤y] = y 0 e−t2/2 dt/ √ 2π, y > 0. T. I. Schindler, R. Zweimüller 266 Theorem 3.4 (A CLT for counting primes in CF) Suppose that either Theorem 3.4 (A CLT for counting primes in CF) Suppose that either (A) An:= a′ n ≥bn with (bn) ∈RN and n:bn>1 1/bn log bn = ∞, (A) An:= a′ n ≥bn with (bn) ∈RN and n:bn>1 1/bn log bn = ∞, n (B) An:= a′ n = dn with (dn) a sequence of primes and n∈N 1/d2 n = ∞, n B) An:= a′ n = dn with (dn) a sequence of primes and n∈N 1/d2 n = ∞, (C) An:= dn ≤a′ n ≤dn · 1 + 1 cn with (dn) a sequence of natural numbers tending to inﬁnity, (cn) a sequence of positive numbers with cn ≤d0.475 n and ∞ n=1 1/(cndn log(dn)) = ∞. (C) An:= dn ≤a′ n ≤dn · 1 + 1 cn with (dn) a sequence of natural numbers tending to inﬁnity, (cn) a sequence of positive numbers with cn ≤d0.475 n and ∞ n=1 1/(cndn log(dn)) = ∞. 4 The Gauss map and the prime digit function The results announced above express properties of certain stochastic processes derived from the exceptionally well understood dynamical system generated by the ergodic continued fraction map (or Gauss map) S : (0, 1] →[0, 1], Sx := 1 x − 1 x = 1 x −k for x ∈ 1 k + 1, 1 k =: Ik, k ≥1 which, since [10], is known to preserve the probability density hG(x) := 1 log 2 1 1 + x , x ∈I. The invariant Gauss measure μG on BI deﬁned by the latter, μG(B) := B hG(x) dx, is exact (and hence ergodic). As hardly any textbook on ergodic theory fails to point out, iteration of S reveals the continued fraction digits of any x ∈I, in that x = [a1(x), a2(x), . . .] with an(x) = a ◦Sn−1(x), n ≥1, where a : I →N is the digit function corresponding to the partition ξ := {Ik : k ≥1}, i.e. a(x) := ⌊1/x⌋= k for x ∈Ik. The stationary sequence (a ◦Sn)n≥0 on the probability space (I, BI , μG) thus obtained exhibits interesting properties since a has inﬁnite expectation, I a dμG = k≥1 k μG(Ik) = ∞, as μG(Ik) = log( (k+1)2 k(k+2))/ log 2 ∼1/(log 2 · k2) for k →∞. As in classical probability theory, the tail behaviour of the distribution, given by μG ({a ≥K}) = 1 log 2 · log K + 1 K ∼ 1 log 2 · 1 K as K →∞ (which entails L(N) := I (a ∧N) dμG = N K=1 μG ({a ≥K}) ∼log N/ log 2 as N → ∞), is the key to ﬁne asymptotic results. However, the study of the CF digit sequence goes beyond standard results, since the random variables a ◦Sn are not independent. Yet, it is well known that they still satisfy a strong form of asymptotic independence or mixing in the following sense: Given any measure preserving transformation T on a probability space (, B, P), and a countable measurable partition γ (mod P), the ψ-mixing coefﬁcients of γ are deﬁned as ψγ (n) := sup k≥1 P(V ∩W) P(V )P(W) −1 : V ∈σ(k−1 j=0 T −jγ ), P(V ) > 0, W ∈T −(n+k−1)B, P(W) > 0 , n ≥1. 123 123 Prime numbers in typical... 4 The Gauss map and the prime digit function 267 The partition γ is said to be continued-fraction (CF-) mixing for the probability preserving system (, B, P, T ) if it is generating, and if ψγ (1) < ∞as well as ψγ (n) →0 for n →∞. (Note that (ψγ (n))n≥1 is non-increasing.) Of course, the nomenclature is due to the fact that (4.1) ξ is CF-mixing for (I, BI , μG, S). (4.1) Actually, this system is exponentially CF-mixing, in that there are constants C > 0 and ρ ∈(0, 1) such that ψξ(n) ≤C ρn for n ≥1 ψξ(n) ≤C ρn for n ≥1 ψξ(n) ≤C ρn for n ≥1 (which is related to Gauss’ famous question mentioned in the introduction, see e.g. [13] or [23]). (which is related to Gauss’ famous question mentioned in the introduction, see e.g. [13] or [23]). We are going to study occurrences of prime digits by considering the restricted digit function a′ := (1P ◦a) · a : I →{0} ∪P. As in the case of a, this function, as a random variable on (I, BI , μG), still has inﬁnite expectation. Indeed, the prime number theorem (PNT) enables us to quickly determine the all-important tail asymptotics for the distribution of a′. The following lemma is the key to our analysis of the prime digit sequence. Lemma 4.1 (Tailbehaviourandtruncatedexpectationofa′)Thedistributionofa′ (withrespect to the Gauss measure) satisﬁes μG a′ ≥K ∼ 1 log 2 · 1 K log K as K →∞. (4.2) μG a′ ≥K ∼ 1 log 2 · 1 K log K as K →∞. (4.2) In particular, a′ is not integrable, I a′ dμG = ∞. Moreover, L′(N) := I (a′ ∧N) dμG ∼log2 N log 2 as N →∞. (4.3) (4.2) In particular, a′ is not integrable, I a′ dμG = ∞. Moreover, log N In particular, a′ is not integrable, I a′ dμG = ∞. Moreover, L′(N) := I (a′ ∧N) dμG ∼log2 N log 2 as N →∞. (4.3) (4.3) so that a′(N) := N/L′(N) ∼log 2 · N/ log2 N is asymptotically inverse to b′(N) := (N log2 N)/ log 2. Proof First, the PNT is easily seen (cf. [12], Theorem 1.8.8) to imply that Proof First, the PNT is easily seen (cf. [12], Theorem 1.8.8) to imply that pn ∼n log n as n →∞, (4.4) (4.4) where pn denotes the nth prime number. 4 The Gauss map and the prime digit function Therefore, where pn denotes the nth prime number. Therefore, where pn denotes the nth prime number. Therefore, n≥N 1 p2n ∼ n≥N 1 n2(log n)2 ∼ 1 N(log N)2 as N →∞. Letting N(K) denote the least n with pn ≥K, we have, as K →∞, Letting N(K) denote the least n with pn ≥K, we have, as K →∞, μG a′ ≥K = p≥K,p∈P μG(Ip) ∼ 1 log 2 p≥K,p∈P 1 p2 = 1 log 2 n≥N(K) 1 p2n μG a′ ≥K = p≥K,p∈P μG(Ip) ∼ 1 log 2 p≥K,p∈P 1 p2 = 1 log 2 n≥N(K) 1 p2n and, by PNT, N(K) ∼K/ log K. Combining these observations yields (4.2). The second statement is an easy consequence thereof, since and, by PNT, N(K) ∼K/ log K. Combining these observations yields (4.2). The second statement is an easy consequence thereof, since L′(N) = N K=1 μG a′ ≥K ∼ 1 log 2 N K=2 1 K log K ∼ 1 log 2 N 2 dx x log x as N →∞. ⊓⊔ Straightforward calculation veriﬁes the assertions about a′ and b′. 123 268 T. I. Schindler, R. Zweimüller Remark 4.1 Several of the results allow for analogues in which prime digits are replaced by digits belonging to other subsets M of the integers for which πM(n) := #M ∩{1, . . . , n} is regularly varying with m∈M 1 m = ∞, like, for example, the set of integers which are the product of exactly k prime numbers, see Theorem 3.5.11 of [14]. (M. Thaler, personal communication.) 5 Proofs of the results on a.e. convergence We are now ready for the proofs of our pointwise convergence results. We can always work, without further mention, with the invariant measure μG, since it has the same null-sets as λ. Proof of Proposition 2.1 This, of course, is just the ergodic theorem, 1 n n k=1 1P ◦ak = 1 n n−1 k=0 1P ◦Sk −→μG(P) = p∈P μG(Ip) a.e. as n →∞. ⊓⊔ In the following we will repeatedly appeal to the following version of Rényi’s Borel- Cantelli Lemma (BCL) (as in Lemma 1 of [3]): Lemma 5.1 (Rényi’s Borel-Cantelli Lemma)Assume that (En)n≥1 is a sequence of events in the probability space (, B, P) for which there is some r ∈(0, ∞) such that P(E j ∩Ek) P(E j) P(Ek) ≤r whenever j, k ≥1, j ̸= k. Then P({En inﬁnitely often}) > 0 iff n≥1 P(En) = ∞. Then P({En inﬁnitely often}) > 0 iff n≥1 P(En) = ∞. Then P({En inﬁnitely often}) > 0 iff n≥1 P(En) = ∞. This lemma enables us to prove Theorem 2.1. This lemma enables us to prove Theorem 2.1. Proof of Theorem 2.1 a) Note that {a′ j > c} = S−( j−1){a′ > c} with {a′ > c} measurable w.r.t. ξ. As a consequence of the CF-mixing property (4.1), we see that Rényi’s BCL applies to show that μG({a′ n > bn i.o.}) > 0 iff n≥1 μG({a′ n > bn}) = ∞. (5.1) (5.1) By S-invariance of μG and Lemma 4.1, we have μG({a′ n ≥bn}) = μG({a′ ≥bn}) ∼ 1/(bn log bn), so that divergence of the right-hand series in (5.1) is equivalent to that of n≥1(bn log bn)−1. Finally, again because of {a′ j > c} = S−( j−1){a′ > c}, the set {a′ n > bn i.o.} is easily seen to belong to the tail-σ-ﬁeld n≥0 S−nBI of S. The system (I, BI , μG, S) being exact, the latter is trivial mod μG. Hence μG({a′ n > bn i.o.}) > 0 implies μG({a′ n > bn i.o.}) = 1. b) Statement (2.2) is seen by an easy routine argument, as in the proof of Proposition 3.1.8 of [13]. b) Statement (2.2) is seen by an easy routine argument, as in the proof of Proposition 3.1.8 of [13]. [ ] c) Without loss of generality we ﬁrst assume that cn ≤0.5, for all n. 5 Proofs of the results on a.e. convergence If this doesn’t hold, we can easily switch to a subsequence in which this holds and consider the subsequences c) Without loss of generality we ﬁrst assume that cn ≤0.5, for all n. If this doesn’t hold, we can easily switch to a subsequence in which this holds and consider the subsequences 123 269 Prime numbers in typical... separately. By the prime number theorem we have separately. By the prime number theorem we have separately. By the prime number theorem we have separately. By the prime number theorem we have λ(a′ 1 ∈[dn, dn(1 + 1/cn)]) = λ(a′ 1 ≥dn) −λ(a′ 1 ≥dn(1 + 1/cn)) ≍ 1 dn log dn − 1 dn(1 + 1/cn) log(dn(1 + 1/cn)) ∼ 1 cndn log dn . Next, we assume that cn > 0.5. We note that #P ∩(dn, dn(1 + 1/cn)] · λ(a1 = dn) ≤λ(a′ 1 ∈[dn, dn(1 + 1/cn)]) ≤#P ∩(dn, dn(1 + 1/cn)] · λ(a1 = dn(1 + 1/cn)). (5.2) Next, we assume that cn > 0.5. We note that Next, we assume that cn > 0.5. We note that #P ∩(dn, dn(1 + 1/cn)] · λ(a1 = dn) ≤λ(a′ 1 ∈[dn, dn(1 + 1/cn)]) ≤#P ∩(dn, dn(1 + 1/cn)] · λ(a1 = dn(1 + 1/cn)). (5.2) (5.2) Furthermore, λ(a1 = dn) ≍1 d2n ≍λ(a1 = dn(1 + 1/cn)). (5.3) (5.3) On the other hand, we have by [4], p. 562 that there exists K > 0 such that #P ∩(d d (1 + 1/c )] ∼π(d (1 + 1/c )) π(d ) ≤K dn On the other hand, we have by [4], p. 562 that there exists K > 0 such that On the other hand, we have by [4], p. 562 that there exists K > 0 such that #P ∩(dn, dn(1 + 1/cn)] ∼π(dn(1 + 1/cn)) −π(dn) ≤K · dn c log d . n the other hand, we have by [4], p. 562 that there exists K > 0 such that #P ∩(dn, dn(1 + 1/cn)] ∼π(dn(1 + 1/cn)) −π(dn) ≤K · dn cn log dn . #P ∩(dn, dn(1 + 1/cn)] ∼π(dn(1 + 1/cn)) −π(dn) ≤K · dn cn log dn . Combining this with (5.2) and (5.3) yields the statement of c). d) This follows immediately from [16, Theorem 6a]. Combining this with (5.2) and (5.3) yields the statement of c). d) This follows immediately from [16, Theorem 6a]. 5 Proofs of the results on a.e. convergence ⊓⊔ ⊓⊔ Proof of Lemma 2.1 By assumption there is some ε ∈(0, 1) such that the set M := {n ≥1 : (n log2 n)/bn ≥ε} is inﬁnite. Deﬁne c(x) := exp(√log x) and f (x) := x log x log2 x for x > 1, and note that c(x) < x for x > e. Suppose that n ∈M, n ≥4, and c(n) < k ≤n. Since k ≤n, we have bk = (bk/k)k ≤(bn/n)k ≤(1/ε)k log2 n, and thus bk log bk ≤(1/ε)k log2 n(log(1/ε) + log k + log3 n) = f (k) ε log2 n log2 k −log(ε) log k + 1 + log3 n log k . On the other hand, c(n) < k implies log k > √log n and hence log2 k > (1/2) log2 n, log k > 1, and log k > log3 n. On the other hand, c(n) < k implies log k > √log n and hence log2 k > (1/2) log2 n, log k > 1, and log k > log3 n. On the other hand, c(n) < k implies log k > √log n and hence On the other hand, c(n) < k implies log k > √log n and hence log2 k > (1/2) log2 n, log k > 1, and log k > log3 n. log2 k > (1/2) log2 n, log k > 1, and log k > log3 n. Using these estimates we see that Using these estimates we see that Using these estimates we see that bk log bk ≤f (k) C(ε) with C(ε) := ε 2 (−log(ε) + 2) > 0. Taking into account that log3 x is a primitive of 1/ f (x) we get Taking into account that log3 x is a primitive of 1/ f (x) we get k>c(n) 1 bk log bk ≥C(ε) c(n)<k≤n 1 f (k) ≥C(ε) n c(n) dx f (x) − 1 f ([c(n)]) = C(ε) log 2 − 1 f ([c(n)]) . Since this estimate holds for inﬁnitely many n, we see that Since this estimate holds for inﬁnitely many n, we see that lim n→∞ k>n 1 bk log bk ≥C(ε) log 2, 123 123 270 T. I. Schindler, R. Zweimüller proving that k≥1 1 bk log bk diverges. proving that k≥1 1 bk log bk diverges. 5 Proofs of the results on a.e. convergence proving that k≥1 1 bk log bk diverges. Proof of Theorem 2.2 a) Since I a′ dμG = ∞by Lemma 4.1, this is immediate from the ergodic theorem. b) We apply Theorem 1.1 of [2] to (I, BI , μG, S) and a′, observing that (in the notation of that paper), Na′ = 1 since J1 = n≥1 n2 log n log2 n −1 < ∞. Further- more, by using the estimate of Lemma 4.1 and setting a′(N) := N/L′(N) ∼log 2·N/ log2 N we get that its asymptotic inverse can be written as b′(N) := (N log2 N)/ log 2 which by the statement of the paper coincides with the norming sequence. c) Using Theorem 2.1 a), we ﬁrst note that n≥1 1/(bn log bn) = ∞implies limn→∞b−1 n n k=1 a′ k = ∞a.e. since a′ n ≤n k=1 a′ k. For the converse, assume that n≥1 1/(bn log bn) < ∞, which by Lemma 2.1 entails (n log2 n)/bn →0. In view of Theorem 2.1 a), our assumption implies that M′ n/bn →0 a.e. Together with statement b) above, these observations prove (2.7), because 1 bn n k=1 a′ k = M′ n bn + n log2 n log 2 · bn · log 2 n log2 n n k=1 a′ k −M′ n . d) Note ﬁrst that letting c′ n := d′ n/ log2(10 j) for n ∈(n( j −1), n( j)], provides us with a non-decreasing sequence satisfying n≥1 1/(c′ n log c′ n) < ∞(use generous estimates). By Theorem 2.1 therefore λ({M′ n > c′ n i.o.}) = 0. Since c′ n = o(d′ n), we see that for every ε > 0, {M′ n > ε d′ n i.o.} ⊆{M′ n > c′ n i.o.}. Combining these observations shows that M′ n d′n −→0 a.e. (5.4) (5.4) Together with (2.5) and n log2 n/(log 2 · d′ n) ≤1, this proves, via Together with (2.5) and n log2 n/(log 2 · d′ n) ≤1, this proves, via 1 d′n n k=1 a′ k = n log2 n log 2 · d′n · log 2 n log2 n n k=1 a′ k −M′ n + M′ n d′n , (5.5) (5.5) that lim n→∞ 1 d′n n k=1 a′ k ≤1 a.e. 5 Proofs of the results on a.e. convergence lim n→∞ 1 d′n n k=1 a′ k ≤1 a.e. Specializing (5.5), and using (2.5) and (5.4) again, we ﬁnd that Specializing (5.5), and using (2.5) and (5.4) again, we ﬁnd that Specializing (5.5), and using (2.5) and (5.4) again, we ﬁnd that 1 d′ n( j) n( j) k=1 a′ k = log 2 n( j) log2 n( j) ⎛ ⎝ n( j) k=1 a′ k −M′ n( j) ⎞ ⎠+ M′ n( j) d′ n( j) −→1 a.e. as j →∞, and our claim (2.8) follows. ⊓⊔ as j →∞, and our claim (2.8) follows. ⊓⊔ ⊓⊔ as j →∞, and our claim (2.8) follows. as j →∞, and our claim (2.8) follows. ⊓⊔ Proof of Theorem 2.3 a) Apply Theorem 4 of [3] to the system (I, BI , μG, S) with CF-mixing partition γ := ξ. Statement (2.9) is immediate if we take (a′, a′) as our pair (ϕ, ψ) of γ - measurable functions, cf. Remark 3 in [3]. Turning to the general version (2.10), we consider ϕ := g ◦a′ and ψ := a′. According to the result cited, Proof of Theorem 2.3 a) Apply Theorem 4 of [3] to the system (I, BI , μG, S) with CF-mixing partition γ := ξ. Statement (2.9) is immediate if we take (a′, a′) as our pair (ϕ, ψ) of γ - measurable functions, cf. Remark 3 in [3]. Turning to the general version (2.10), we consider ϕ := g ◦a′ and ψ := a′. According to the result cited, lim n→∞ g(a′ n) n−1 k=1 a′ k = ∞a.e. iff I a′ ◦g ◦a′ dμG = ∞, (with a′ from Lemma 4.1), while otherwise limn→∞g(a′ n)/(n−1 k=1 a′ k) = 0 a.e. The present assertion merely reformulates the divergence condition above: We see (using (4.4) and the (with a′ from Lemma 4.1), while otherwise limn→∞g(a′ n)/(n−1 k=1 a′ k) = 0 a.e. The present assertion merely reformulates the divergence condition above: We see (using (4.4) and the 123 271 Prime numbers in typical... 5 Proofs of the results on a.e. convergence Using then [15, Theorem 1.7 & erratum] we obtain that (2.12) holds for (bn) fulﬁlling bn = o(n) and limn→∞bn log2 n = ∞and for (dn) fulﬁlling dn ∼ 1/γ 1 −1/γ nγ b1−γ n L−γ # n bn γ , where ℓ# denotes the de Bruijn conjugate of a slowly varying function ℓ, see e.g. [6] for a pre- cise deﬁnition. In our case L−γ # (n) = ((log n)γ (log 2)γ /γ γ )# ∼γ γ / ((log n)γ (log 2)γ ). Hence, where ℓ# denotes the de Bruijn conjugate of a slowly varying function ℓ, see e.g. [6] for a pre- cise deﬁnition. In our case L−γ # (n) = ((log n)γ (log 2)γ /γ γ )# ∼γ γ / ((log n)γ (log 2)γ ). Hence, dn ∼ γ γ (γ −1)(log 2)γ nγ b1−γ n 1 (log (n/bn)γ )γ ∼ 1 (γ −1)(log 2)γ · nγ b1−γ n (log n)γ , where the last assymptotic follows from the assumption bn = o(n1−ϵ). ⊓⊔ 5 Proofs of the results on a.e. convergence regularity properties on g) that (for some constant c > 0) regularity properties on g) that (for some constant c > 0) regularity properties on g) that (for some constant c > 0) I a ◦g ◦a′ dμG ≍ n≥1 g(pn) log2 g(pn) μG(Ipn) ≍ n≥1 g(n log n) log2 g(n log n) 1 (n log n)2 ≍ ∞ c g(x log x) log2 g(x log x) dx (x log x)2 ≍ ∞ c g(y) log2 g(y) dy y2 log y . b) Same argument as in a), this time with ϕ := g ◦a′ and ψ := a, and replacing a′ above by a(t) := t/L(t) ∼log 2 · t/ log t as t →∞. c) We have b) Same argument as in a), this time with ϕ := g ◦a′ and ψ := a, and replacing a′ above by a(t) := t/L(t) ∼log 2 · t/ log t as t →∞. c) We have lim n→∞ n k=1 a′ k n k=1 ak ≤lim n→∞ n k=1 a′ k −M′ n n k=1 ak −Mn + lim n→∞ M′ n n k=1 ak −Mn ≤lim n→∞ n log2 n log 2 n log n log 2 + lim n→∞ n log2 n log 2 n log n log 2 = 0 which follows by b) of Theorem 2.2 together with the Diamond-Vaaler trimmed law, log 2 n k=1 ak −Mn /(n log n) →1 a.e. and ﬁnally by a) of Corollary 2.1. ⊓⊔ ⊓⊔ Proof of Theorem 2.4 a) We have that (a′)γ dλ < ∞and the statement follows by the ergodic theorem. b) We may apply [15, Theorem 1.7 & erratum]. That Property C is fulﬁlled with the bounded variation norm ∥·∥BV is a standard result. For Property D, we notice that ∥a·1{a≤ℓ}∥BV ≤2ℓ and ∥1{a≤ℓ}∥BV ≤2 implying that this property is fulﬁlled. In order to calculate the norming sequence (dn) we notice that In order to calculate the norming sequence (dn) we notice that μG a′ γ > n = μG a′ > n1/γ ∼ 1 log 2 n1/γ log n1/γ = γ log 2 n1/γ log n = L(n) n1/γ , μG a′ γ > n = μG a′ > n1/γ where L(n) = γ/(log 2 log n) is a slowly varying function. where L(n) = γ/(log 2 log n) is a slowly varying function. 6 Proofs of the results on distributional convergence , ∞} is given by ϕA(x) := inf{n ≥1 : Snx ∈A}, which is ﬁnite a.e. on I. Deﬁne SAx := SϕA(x)x for a.e. x ∈I, which gives the ﬁrst entrance map S : I →A Letting A′ : {a′ ≥l} l ≥1 we see that ϕ′ ϕ and more generally ϕ′ l,i = ϕA′ l ◦Si−1 A′ l for i ≥1. It is clear that A′ l is ξ-measurable, and according to Lemma 4.1, μG A′ l ∼(log 2 · l logl)−1 as l →∞. Therefore, Theorem 10.2.a) of [25] immediately implies statement a). b) This is a straightforward consequence of Theorem 10.2.b) in [25], because {a′ ≥l/ϑ} = A′ ⌊l/ϑ⌋is ξ -measurable and (4.2) entails μG a′ ≥l/ϑ ∼ϑ μG a′ ≥l as l →∞. c) Let P( j) := {p ∈P : p ≡j (mod m)}, then Dirichlet’s PNT for primes in residue classes (e.g. Theorem 4.4.4 of [14]) asserts that for each j relatively prime to m, # (P( j) ∩{2, . . . , N}) ∼ 1 φ(m) N log N as N →∞. Via an easy argument parallel to the proof of (4.2), this shows that Via an easy argument parallel to the proof of (4.2), this shows that μG A′ l ∩ a′ ≡j (mod m) ∼ 1 φ(m) log 2 · 1 l logl as l →∞, and hence μG A′ l( j) ∼μG A′ l /φ(m) with A′ l( j) := A′ l ∩ a′ ≡j (mod m) a ξ- measurable set. Another direct application of Theorem 10.2.b) in [25] then completes the proof of our theorem. ⊓⊔ and hence μG A′ l( j) ∼μG A′ l /φ(m) with A′ l( j) := A′ l ∩ a′ ≡j (mod m) a ξ- measurable set. Another direct application of Theorem 10.2.b) in [25] then completes the proof of our theorem. ⊓⊔ The result thus established essentially contains (3.1). The result thus established essentially contains (3.1). Proof of Theorem 3.2 Theorem 3.3 a) contains the statement that μG A′ l ϕ′ l,1 converges to a standard exponential law. Using the natural duality {M′ n < l} = {ϕ′ l,1 ≥n} this is easily seen to imply (3.1). ⊓⊔ ⊓⊔ Proof of Theorem 3.4 The result follows directly by [16, Theorem 3] by considering the sets An = {an ∈P ∩n}. 6 Proofs of the results on distributional convergence We are now ready for the proofs of our distributional convergence results. 123 272 T. I. Schindler, R. Zweimüller Proof of Theorem 3.1 In all cases we only need to check convergence in law w.r.t. μG. a) This follows directly from Theorem 2.4. Proof of Theorem 3.1 In all cases we only need to check convergence in law w.r.t. μG a) This follows directly from Theorem 2.4. b) This statement follows directly by [1]. We use the expression for L′(N) from (4.3) which is a slowly varying function. Since for bn = n log2 n/ log 2 we have nL(bn) ∼bn the statement follows. b) This statement follows directly by [1]. We use the expression for L′(N) from (4.3) which is a slowly varying function. Since for bn = n log2 n/ log 2 we have nL(bn) ∼bn the statement follows. c) This follows from [17, Theorem 1.8]. The conditions on the system and the asymptotic of the norming sequence (dn) we have already considered in the proof of Theorem 2.4. ⊓⊔ c) This follows from [17, Theorem 1.8]. The conditions on the system and the asymptotic of the norming sequence (dn) we have already considered in the proof of Theorem 2.4. ⊓⊔ Proof of Theorem 3.3 In each of the three statements it sufﬁces to prove distributional con- vergence under the invariant measure μG (see Propositions 3.1 and 5.1 in [25]). g μG ( p [ ]) a) For A ∈BI with λ(A) > 0, the (ﬁrst) hitting time function of A under the Gauss map S, ϕA : I →N := {1, 2, . . . , ∞} is given by ϕA(x) := inf{n ≥1 : Snx ∈A}, which is ﬁnite a.e. on I. Deﬁne SAx := SϕA(x)x for a.e. x ∈I, which gives the ﬁrst entrance map SA : I →A. Letting A′ l := {a′ ≥l}, l ≥1, we see that ϕ′ l = ϕA′ l and, more generally, ϕ′ l,i = ϕA′ l ◦Si−1 A′ l for i ≥1. It is clear that A′ l is ξ-measurable, and according to Lemma 4.1, μG A′ l ∼(log 2 · l logl)−1 as l →∞. Therefore, Theorem 10.2.a) of [25] immediately implies statement a). a) For A ∈BI with λ(A) > 0, the (ﬁrst) hitting time function of A under the Gauss map S, ϕA : I →N := {1, 2, . . . 6 Proofs of the results on distributional convergence The only thing to check is that ∞ n=1 λ(An) · λ(Ac n) = ∞. (6.1) (6.1) We note that ∞ n=1 λ(An) · λ(Ac n) ≥∞ n=1 λ(An) and thus for (A) (6.1) follows from the proof of Theorem 2.1 a). (B) corresponds to [16, Theorem 5A] and for (C) (6.1) follows from the proof of Theorem 2.1 c). ⊓⊔ ⊓⊔ 123 273 Prime numbers in typical... Acknowledgements The authors are indebted to M. Thaler for valuable comments and suggestions on an earlier version and to T. Trudgian for useful discussions regarding the error term of the prime number theorem. This research was supported by the Austrian Science Fund FWF: P 33943-N Funding Open access funding provided by Austrian Science Fund (FWF). Declarations Conﬂict of interest On behalf of all authors, the corresponding author states that there is no conﬂict of interest. Conﬂict of interest On behalf of all authors, the corresponding author states that there is no conﬂict of interest. Conﬂict of interest On behalf of all authors, the corresponding author states that there is no conﬂict of interest. 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/. References 1. Aaronson, J.: Random f -expansions. Ann. Probab. 14, 1037–1057 (1986). https://doi.org/10.1214/aop/ 1176992457 2. Aaronson, J., Nakada, H.: Trimmed sums for non-negative, mixing stationary processes. Stoch. Proc. Appl. 104, 173–192 (2003). https://doi.org/10.1017/S0143385704001051 pp p g 3. Aaronson, J., Thaler, M., Zweimüller, R.: Occupation times of sets of inﬁnite measure for ergodic transfor- mations. Ergod. Theory Dynam. Sys. 25, 959–976 (2005). https://doi.org/10.1017/S0143385704001051 3. Aaronson, J., Thaler, M., Zweimüller, R.: Occupation times of sets of inﬁnite measure for ergodic transfor- mations. Ergod. Theory Dynam. Sys. 25, 959–976 (2005). https://doi.org/10.1017/S0143385704001051 4. Baker, R.C., Harman, G., Pintz, J.: The difference between consecutive primes. II. Proc. Lond. Math. Soc. 83, 532–562 (2001). https://doi.org/10.1112/plms/83.3.532 4. Baker, R.C., Harman, G., Pintz, J.: The difference between consecutive primes. II. Proc. Lond. Math. Soc. 83, 532–562 (2001). https://doi.org/10.1112/plms/83.3.532 p g p 5. Bernstein, F.: Über eine Anwendung der Mengenlehre auf ein aus der Theorie der säkularen Störungen herrührendes Problem. Math. Ann. 71, 417–439 (1911). https://doi.org/10.1007/BF01456856 5. Bernstein, F.: Über eine Anwendung der Mengenlehre auf ein aus der Theorie der säkularen Störungen herrührendes Problem. Math. Ann. 71, 417–439 (1911). https://doi.org/10.1007/BF01456856 6. Bingham, N.H., Goldie, C.M., Teugels, J.L.: Regular variation. Cambridge University Press, Cambri (1989) Borel, E.: Les probabilités dénomerables et leurs applications arithmétiques. Rend. Circ. Mat. Palermo 27, 247–271 (1909). https://doi.org/10.1007/BF03019651 p g 8. Diamond, H.G., Vaaler, J.D.: Estimates for partial sums of continued fraction partial quotients. Paciﬁc J. Math. 122, 73–82 (1986) 8. Diamond, H.G., Vaaler, J.D.: Estimates for partial sums of continued fraction partial quotients. Paciﬁc J. Math. 122, 73–82 (1986) 9. Doeblin, W.: Remarques sur la théorie mé trique des fractiones continues. Comp. Math. 7, 353–371 (1940) 10. Gauss, C.F.: Letter to Laplace. Göttingen, January 30th 1812 Doeblin, W.: Remarques sur la théorie mé trique des fractiones continues. Comp. Math. 7, 353–371 (1940) 10. Gauss, C.F.: Letter to Laplace. Göttingen, January 30th 1812 10. Gauss, C.F.: Letter to Laplace. Göttingen, January 30th 1812 11. Haeusler, E.: A nonstandard law of the iterated logarithm for trimmed sums. Ann. Probab. 21, 831–860 (1993) 12. Hardy, G.H., Wright, E.M.: An introduction to the theory of numbers. Oxford University Press, London (1989) 13. Iosifescu, M., Kraaikamp, C.: Metrical theory of continued fractions. Kluwer (2002). https://doi.org/ 1007/978-94-015-9940-5 14. Jameson, G.J.O.: The prime number theorem. Cambridge University Press, Cambridge (2003) 15. Kesseböhmer, M., Schindler, T.I.: Strong laws of large numbers for intermediately trimmed Birkhoff sums of observables with inﬁnite mean. Stoch. Proc. Appl. 19. Khinchin, A.. Ya..: Metrische Kettenbruchprobleme. Comp. Math. 1, 361–382 (1935) Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. References 129, 4163–4207 (2019). https://doi.org/10. 1016/j.spa.2018.11.015 Corrigendum in Stoch. Proc. Appl., 130 (2020), 7019. https://doi.org/10.1016/j. spa.2020.07.014 p 16. Kesseböhmer, M., Schindler, T.I.: Limit theorems for counting large continued fraction digits. Lith. M J. 60, 189–207 (2020). https://doi.org/10.1007/s10986-020-09479-5 17. Kesseböhmer, M., Schindler, T.I.: Mean convergence for intermediately trimmed Birkhoff sums of observ- ables with regularly varying tails. Nonlinearity 33, 5543–5566 (2020). https://doi.org/10.1088/1361- 6544/ab9585 18. Kesten, H.: Convergence in distribution of lightly trimmed and untrimmed sums are equivalent. Math. Proc. Cambridge Philos. Soc. 113, 615–638 (1993). https://doi.org/10.1017/S0305004100076246 123 12 274 T. I. Schindler, R. Zweimüller 20. Philipp, W.: Some metrical theorems in number theory II. Duke Math. J. 38, 447–458 (1970). https://doi. org/10.1215/S0012-7094-70-03756-7 21. Philipp, W.: A conjecture of Erdös on continued fractions. Acta Arithmetica 28, 379–386 (1976). https:// doi.org/10.4064/aa-28-4-379-386 g 22. Philipp, W.: Limit theorems for sums of partial quotients of continued fractions. Monatsh. Math 1 195–206 (1988). https://doi.org/10.1007/BF01636928 p g 23. Zweimüller, R.: Kuzmin, coupling, cones, and exponential mixing. Forum Math. 16, 447–457 (200 https://doi.org/10.1515/form.2004.021 24. Zweimüller, R.: Mixing limit theorems for ergodic transformations. J. Theor. Probab. 20, 1059–1071 (2007). https://doi.org/10.1007/s10959-007-0085-y 24. Zweimüller, R.: Mixing limit theorems for ergodic transformations. J. Theor. Probab. 20, 1059–1071 (2007). https://doi.org/10.1007/s10959-007-0085-y ( ) p g y 25. Zweimüller, R.: Hitting times and positions in rare events. Ann. H. Lebesgue. 5, 1361–1415 (2022). https://doi.org/10.5802/ahl.154 p g y 25. Zweimüller, R.: Hitting times and positions in rare events. Ann. H. Lebesgue. 5, 1361–1415 (2022). https://doi.org/10.5802/ahl.154 25. Zweimüller, R.: Hitting times an https://doi.org/10.5802/ahl.154 https://doi.org/10.5802/ahl.154 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. 123
https://openalex.org/W2915000666	https://discovery.ucl.ac.uk/10069286/7/Chudasama_c9ob00339h.pdf	English	null	Recent advances in metal-free aerobic C–H activation	Organic & biomolecular chemistry	2,019	cc-by	5,748	Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence Cite this: Org. Biomol. Chem., 2019, 17, 2865 Cite this: Org. Biomol. Chem., 2019, 17, 2865 André Shamsabadi and Vijay Chudasama * Herein we describe recent developments in selective, metal-free, dioxygen-induced C–H activation. This method of activating C–H bonds is an attractive alternative to traditional methodologies as it uses dioxy- gen, an inherently sustainable and widely accessible oxidant, in place of expensive or toxic metals and/or hazardous peroxides. Reactions developed on the basis of using aerobic C–H activation are also discussed. Received 14th January 2019, Accepted 13th February 2019 DOI: 10.1039/c9ob00339h rsc.li/obc C–H bond activation waste minimisation when compared to classic cross-coupling reactions (e.g. Negishi coupling, Suzuki coupling, etc.). Over the past few decades, great strides have been made in the development of elegant C–H bond activation methodologies which has seen the process turn from being an aspirational idea to a commonplace reaction that is now heavily studied2 and used in both academia and industry.3,4 Modern C–H func- tionalisation reactions are carried out through (i) C–H bond homolysis via use of radical initiators followed by radical func- tionalisation, (ii) C–H insertion through use of carbenes or nitrenes, or (iii) metallic C–H activation via a C–metal inter- mediate by way of σ-bond metathesis, concerted metalation deprotonation, and oxidative addition. Unfortunately, each protocol commonly utilises harsh reagents or additives to acti- vate what is otherwise an unreactive C–H bond. Owing to the Direct C–H bond activation has long been credited as the ‘holy grail’ of organic chemistry.1 The concept of discriminatively activating what is typically an inert C–H bond is a fundamen- tally powerful tool in synthesis for a number of reasons: (i) C–H bonds are ubiquitous in organic molecules, (ii) C–H bond functionalisation typically reduces the amount of transform- ations in multi-step syntheses which can consist of long and laborious protection–deprotection sequences, (iii) it estab- lishes an ideal transformation regarding atom economy and Vijay Chudasama Dr Vijay Chudasama obtained his MSci degree and PhD from University College London in 2008 and 2011, respectively. Following post-doctoral studies under the supervision of Prof. Stephen Caddick, Vijay obtained a Ramsay Memorial Fellowship. During this time, he was made Technical Director of a biotech- nology spin-out (ThioLogics™). In April 2015, he was awarded a lectureship (Reader from October 2017) at UCL for him to focus on the research areas of aerobic C–H activation and various aspects of Chemical Biology. Vijay’s research has recently been highlighted by Forbes, Scientific American, CNN News, Nature Chemistry and the Royal Society of Chemistry. Vijay Chudasama André Shamsabadi André Shamsabadi was born in London, UK. He received his undergraduate education at University College London (UCL) during which he was awarded the Jackson-Lewis Scholarship (2015), C. K. Ingold Prize (2015), Franz Sondheimer Prize (2016) and was a Dean’s List commen- dee. In 2016, he was awarded a UCL Graduate School Research Scholarship, which allowed him to continue in a postgraduate position at UCL under the gui- dance of Dr Vijay Chudasama. Department of Chemistry, University College London, 20 Gordon Street, London WC1H 0AJ, UK. E-mail: v.chudasama@ucl.ac.uk C–H bond activation His research interests are based in radical organic synthesis and aerobic C–H activation. André Shamsabadi André Shamsabadi André Shamsabadi André Shamsabadi was born in London, UK. He received his undergraduate education at University College London (UCL) during which he was awarded the Jackson-Lewis Scholarship (2015), C. K. Ingold Prize (2015), Franz Sondheimer Prize (2016) and was a Dean’s List commen- dee. In 2016, he was awarded a UCL Graduate School Research Scholarship, which allowed him to continue in a postgraduate position at UCL under the gui- dance of Dr Vijay Chudasama. His research interests are based in radical organic synthesis and aerobic C–H activation. André Shamsabadi Vijay Chudasama André Shamsabadi This journal is © The Royal Society of Chemistry 2019 Org. Biomol. Chem., 2019, 17, 2865–2872 \| 2865 Organic & Biomolecular Chemistry View Article Online View Article Online Review Organic & Biomolecular Chemistry Organic & Biomolecular Chemistry Organic & Biomolecular Chemistry ever-increasing popularity and widespread usage of C–H acti- vation processes, there is significant pressure to improve the sustainability of C–H reactions from an environmental stand- point: radical initiators are generally toxic or shock sensitive, nitrenes and carbenes are high-energy materials that require intricate and heavily-monitored reaction conditions and organometallic C–H bond activation often requires the use of precious metals and stoichiometric amounts of oxidants (i.e. MnO2). Indeed, it has been highlighted recently by the ACS Green Chemistry Institute® that C–H activation processes uti- lising green oxidants whilst giving predictable site-selectivities is one of the top three research areas that would benefit from improvements in process ‘greenness’.5 To achieve this, che- mists have created increasingly elegant and boundary-pushing ideas for organometallic C–H bond activation including cataly- sis by Earth-abundant metals (e.g. iron,6 copper7) or forgoing the use of metals altogether.8 In the goal for eﬀectively using greener and milder oxidants, recent advances have been achieved in using arguably the most sustainable oxidant: dioxygen. Dioxygen is abundant; indeed, molecular dioxygen makes up 20.94% of the Earth’s atmosphere, and one of its most fascinating and under-exploited uses is its ability to initiate auto-oxidation (alternatively referred to as autoxida- tion) processes. Whilst process chemists have previously uti- lised an essentially free reagent (air) in the homogenous cata- lytic oxidation of organic and inorganic compounds, the use of dioxygen has very rarely been used as the focal point for C–H bond activation. Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Modern dioxygen-induced C–H bond functionalisation reac- tions have utilised aldehydes, ethers, benzylamines and glycine derivatives as precursors to aerobic activation, and reactions developed on this basis are described within. This journal is © The Royal Society of Chemistry 2019 C–H bond activation but intriguing discussion regarding auto-oxidation is the observation that molecular dioxygen can directly interact with certain C–H bonds to generate carbon radicals without the apparent use of any additional initiator species. This allows for the possibility of C–H functionalisation through reaction of the transiently-formed radical species. This review highlights recent examples of ground-breaking research in which dioxygen is used to activate C–H bonds as a means to achieve green and sustainable C–H bond functionali- sation (Scheme 2). Historically, aerobic C–H bond activation has been scarcely used in synthesis outside of simple C–H oxi- dation reactions,9 but its true potential in synthesis is starting to be uncovered with an increasing number of research groups utilising dioxygen as a radical initiator. It should be high- lighted that diﬀerent research groups have had diﬀering logis- tic protocols of introducing dioxygen in their procedures: use of a pressurised environment of O2 through use of an auto- clave (>1 atm), use of a dioxygen atmosphere via balloon (1 atm), bubbling compressed air into a reaction mixture under an inert environment, or by simple exposure of the reac- tion to the dioxygen in air (referred to as atmospheric dioxy- gen). We believe that, where possible, the strive should be made to use atmospheric dioxygen as the source of dioxygen as it represents the greenest and a freely accessible form of the oxidant to be utilised in synthesis. Aldehydes The auto-oxidation process describes the auto-catalytic oxi- dation of organic or inorganic systems. For organic molecules, it is usually used to describe the propagation reactions in a chain reaction cycle in which carbon radicals couple with dioxygen to form peroxy radical species, which in turn can then abstract H-atoms to form a peroxide product and regener- ate further carbon radicals (Scheme 1). However, in these radical chain reactions, use of a radical initiator or UV light is commonly employed to initially activate a C–H bond. A rare Aldehydes are amongst the most desirable substrates to achieve C–H functionalisation owing to ubiquity of the acyl group in molecular architecture. Radical C–H functionalisation is particularly useful for this purpose due to the susceptibility to hydrogen abstraction of aldehydic C–H bonds. Aldehydes are one of the most classical and studied examples of a reagent class that undergoes auto-oxidation.10,11 The mecha- nism for aldehyde auto-oxidation is identical to that shown above (Scheme 1) except that the peracyl product can undergo further reaction with another molecule of aldehyde to form an intermediate that breaks down into two molecules of car- boxylic acid as the final product.11 The immediate species Scheme 1 General dioxygen-initiated autoxidation pathway. Scheme 2 Model example of aerobic C–H bond activation via dioxy- gen-induced radical initiation through trapping of initial radical adduct species. Scheme 2 Model example of aerobic C–H bond activation via dioxy- gen-induced radical initiation through trapping of initial radical adduct species. Scheme 2 Model example of aerobic C–H bond activation via dioxy- gen-induced radical initiation through trapping of initial radical adduct species. Scheme 1 General dioxygen-initiated autoxidation pathway. 2866 \| Org. Biomol. Chem., 2019, 17, 2865–2872 View Article Online Organic & Biomolecular Chemistry Review Scheme 5 Hydroacylation of unsaturated esters 6–8 via aerobic C–H activation. formed through aldehyde auto-oxidation is the acyl radical which is nucleophilic in nature and can eﬃciently participate in radical addition to electron-deficient unsaturated bonds. Since 2009, hydroacylation of CvC and NvN bonds utilis- ing acyl radicals formed through the exposure of aldehydes 1 to atmospheric dioxygen has allowed the green and eﬃcient formation of various unsymmetrical ketones and acyl hydra- zide moieties in good to excellent yields.12 Caddick and co- workers reported the reaction of aldehydes with several vinyl sulfonates 2 under atmospheric conditions to aﬀord unsym- metrical ketones 3 in respectable yields (ca. Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Scheme 5 Hydroacylation of unsaturated esters 6–8 via aerobic C–H activation. dation) and increasing conversion of alkenes 6–8 to the desired ketone products 9–11. Chudasama et al. then extended the scope of the developed protocol to include vinyl phosphonates 12, which were suc- cessfully utilised as acyl radical acceptors (Scheme 6). A reac- tion temperature of 60 °C was again used to achieve eﬃcient conversion of the alkene.16 The addition of a dioxane radical to a vinyl phosphonate was also observed when this protocol was used, providing supporting evidence of radicals being formed in the reaction mixture. The Chudasama and Caddick groups then further utilised aerobically-generated acyl radicals to achieve hydroacylation of various vinyl sulfones 4. Here, aliphatic aldehydes were found to perform best in the reaction with vinyl sulfones to achieve consistent yields (ca. 60%) of the desired ketone product 5 (Scheme 4) at 21 °C using water as the reaction solvent. Chudasama et al. then expanded the scope of this hydroacy- lation methodology to include various unsaturated esters 6–8 where aliphatic aldehydes were once again found to be the best performers, aﬀording yields of 9–11 of up to 89% (Scheme 5). This work paved the way for expanding the scope of radical acceptors that were applicable for use in this meth- odology.15 It should be noted that eﬃcient reaction conversion was achieved through raising the reaction temperature to 60 °C. It was postulated that the increase in temperature resulted in a lowering of dissolved oxygen in the reaction mixture, combating over-oxidation of the acyl radical (auto-oxi- In 2011, using a similar aerobic activation protocol, acyl hydrazides 15 were synthesised in a highly eﬃcient manner when using azodicarboxylates 14 as radical acceptors (Scheme 7). Acyl hydrazides are extremely useful synthetic intermediates which have found use in the formation of many medicinally desirable entities (amides, indazoles, hydroxamic acids, etc.).17–22 Both diethyl azodicarboxylate (DEAD) and di- isopropyl azodicarboxylate (DIAD) were shown to be compatible with both aliphatic and aromatic aldehydes towards hydroacy- lation via aerobic C–H bond activation. The eﬃciency of the reaction was clearly demonstrated with excellent yields being Scheme 6 Hydroacylation of vinyl phosphonates 12 via aerobic C–H activation. Scheme 3 Hydroacylation of vinyl sulfonates 2 via aerobic C–H activation. Scheme 3 Hydroacylation of vinyl sulfonates 2 via aerobic C–H activation. Scheme 6 Hydroacylation of vinyl phosphonates 12 via aerobic C–H activation. Aldehydes 70%) and all was reported at room temperature (Scheme 3).13 Optimal yield was achieved when utilising 5 equivalents of aldehyde, which is very competitive with regards to existing hydroacylation meth- odologies. The reaction was completely inhibited when con- ducted in the presence of radical inhibitor BHT (2,6-di-tert- butyl-4-methylphenol), providing evidence of a radical mecha- nism. Whilst the initial account reported 1,4-dioxane as an optimal solvent, further research from the group found that the use of water as a solvent, the ideal “green” solvent, pro- duced similar if not better yields of the ketones.14 Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Scheme 9 Dioxygen-mediated radical dicarbofunctionalisation of alkene 20 with aldehyde 1. R’ can be a variety of functional groups (e.g. alkyl, amide, nitro, halo, etc.). Guin and co-workers have since expanded their method- ology to generate alkyl radicals 17 to achieve dicarbofunctiona- lisation of alkenes 18 (Scheme 9).25 As before, the methodology required a relatively high equivalence of aldehyde 1 (6–12 equivalents). Moderate yields where observed when secondary or tertiary aldehydes were utilised (ca. 60%) with poor yields only observed when a primary aldehyde was used (25%). When the reaction was repeated under an atmosphere of argon, only a trace amount of desired product 24 was observed, highlight- ing the importance of dioxygen for the reaction. Furthermore, no product formation was observed when the reaction was carried out in the presence of radical inhibitor TEMPO ((2,2,6,6-tetramethylpiperidin-1-yl)oxyl), with the alkyl-TEMPO product being detected by GCMS and HRMS, thereby corrobor- ating the radical mechanism for the reaction. It is important to note that whilst the optimised methodology utilises a dioxy- gen atmosphere, it is stated that the reaction can be carried out in air albeit with a slightly lower yield of desired product.† Scheme 9 Dioxygen-mediated radical dicarbofunctionalisation of alkene 20 with aldehyde 1. R’ can be a variety of functional groups (e.g. alkyl, amide, nitro, halo, etc.). Scheme 10 Oxidative radical 1,2-alkylarylation of α,β-unsaturated amides 25. Scheme 11 Aerobically-generated hypervalent iodine(III) species 31. Scheme 10 Oxidative radical 1,2-alkylarylation of α,β-unsaturated amides 25. Scheme 10 Oxidative radical 1,2-alkylarylation of α,β-unsaturated amides 25. Guin and co-workers have also used their procedure to describe the oxidative 1,2-alkylarylation of α,β-unsaturated amides 25 for the synthesis of biologically-active oxindoles 28 (Scheme 10).26 A large excess of aldehyde had to be utilised (6–12 equivalents), which is perhaps unexpected as in the pro- posed radical chain process an acyl radical is generated as reaction byproduct. As before, complete radical inhibition was observed when conducted in the presence of TEMPO. Nonetheless, good to excellent yields of product 28 (44–81% yield) were achieved when utilising secondary or tertiary alde- hydes, with the only poor yield (15%) observed when using butyraldehyde. Scheme 11 Aerobically-generated hypervalent iodine(III) species 31. Scheme 11 Aerobically-generated hypervalent iodine(III) species 31. †Precise yield not stated. Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Scheme 6 Hydroacylation of vinyl phosphonates 12 via aerobic C–H activation. Scheme 4 Hydroacylation of vinyl sulfones 4 via aerobic C–H activation. Scheme 4 Hydroacylation of vinyl sulfones 4 via aerobic C–H activation. Scheme 7 Hydroacylation of azodicarboxylates 14 via aerobic C–H activation. Scheme 4 Hydroacylation of vinyl sulfones 4 via aerobic C–H activation. Scheme 7 Hydroacylation of azodicarboxylates 14 via aerobic C–H activation. Org. Biomol. Chem., 2019, 17, 2865–2872 \| 2867 This journal is © The Royal Society of Chemistry 2019 View Article Online Review Review Organic & Biomolecular Chemistry attained even when utilising aldehydes as the limiting reagent.23 Chudasama et al. also showed that α-chiral alde- hydes could undergo aerobic hydroacylation with azodicarbox- ylates and vinyl sulfonates with complete retention of optical purity. These transformations represented the first examples of hydroacylation achieved using enantio-enriched aldehydes with retention of enantiomeric excess. Scheme 8 Dioxygen-mediated decarbonylative C–H alkylation of pro- tonated heteroaromatic bases 18 with aldehydes 1. Scheme 8 Dioxygen-mediated decarbonylative C–H alkylation of pro- tonated heteroaromatic bases 18 with aldehydes 1. In 2015, Guin and co-workers conducted reactions under a dioxygen environment through use of a dioxygen balloon as an alternative method to achieve aldehydic C–H bond activation. The group utilised a high reaction temperature (115 °C) as means of accessing alkyl radicals 17 through facile decarboxyl- ation of initially formed secondary and tertiary acyl radicals 16. In eﬀect, the group achieved eﬃcient ortho-alkylation on a plethora of acidified heteroaromatic bases 18 (Scheme 8).24 Regrettably, a large excess of aldehyde had to be utilised for optimal yield (6–20 equivalents). It is important to note that a moderate yield† of desired product 19 was observed when the reaction mixture was exposed to air as opposed to use of an oxygen balloon. This journal is © The Royal Society of Chemistry 2019 Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Scheme 12 Aerobically-generated hypervalent iodine(III) dicarboxylate species 33. was never carried out when utilising air as the dioxygen source, it was reported that the reaction procedure for the oxi- dative cleavage of 1,2-diols was almost as eﬀective when the reaction mixture was exposed to air (ca. 90% compared to quantitative yield when utilising a balloon). More recently, Troisi et al. extended the available reaction partners that can be coupled to THF following aerobic C–H bond activation.33 It was showed that allyl chloride 40 can be utilised as a chlorinating agent. Doing so formed an allyl radical capable of abstracting an ethereal H-atom, propagating the radical chain reaction. Although not explicitly stated, it was suspected that the resulting α-chloro product 41 had the same decomposition eﬀect displayed by the α-bromo species 36 described by Parsons and co-workers in which release of halide resulted in oxocarbenium species 37, which can then be trapped using alcohols 38. Utilising this method, Troisi et al. showed the tetrahydrofuranylation of simple (phenol, benzylic) and complex (cholesterol) alcohols (Scheme 14). Further studies confirmed the importance of dioxygen in their protocol by observing no product formation when the reaction was conducted under a nitrogen atmosphere and using deaer- ated THF. Evidence supporting a radical mechanism was then provided by showing complete reaction inhibition when Powers and co-workers carried out a similar experiment for the oxidation of iodobenzene 30 to iodobenzene diacetate 33 (Scheme 12).28 Their initial results highlighted the auto-oxi- dation of diﬀerent aldehydes as a means to oxidise iodoben- zene at 23 °C. Benzaldehyde, isobutyraldehyde and butyralde- hyde all proved ineﬀective with minimal amount of product formed (0%, 2% and 6% respectively). However, use of acet- aldehyde gave a somewhat capricious formation of desired product 33 with 42–91% yield observed over 5 repeats. The group speculated that the variability of oxidation eﬃciency was due to inconsistent initiation of radical auto-oxidation chem- istry. Although the work published showed the possibility of the reaction being able to proceed through a dioxygen- mediated C–H activation process, the group eventually settled on using a cobalt-based initiator (1 mol%) and utilised the in situ formation of the iodine(III) diacetate species to oxyge- nate a wide range of substrates (alkenes, β-keto esters, etc.). Scheme 13 Aerobically-initiated protocol for the protection of alcohols 38 as 2-tetrahydrofuranyl ethers 39 utilising bromotrichloromethane. This journal is © The Royal Society of Chemistry 2019 Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. which perisobutyric acid 29 was the predominant species present after 1 h.27 Further reaction upon addition of iodoben- zene 30 proceeded eﬃciently, granting a 75% yield of hyperva- lent iodine(III) species 31 after 5 h (Scheme 11). Miyamoto and co-workers then used this procedure of generating hypervalent iodine(III) intermediates to eﬀectively carry out oxidative clea- vage of 1,2-diols and Hofmann rearrangement of carboxa- mides. Whilst the isolation of hypervalent iodide(III) species 31 In 2017, Miyamoto and co-workers initially reported the facile auto-oxidation of isobutyraldehyde 1a under a dioxygen atmosphere (use of balloon) in 1,2-dichloroethane at 40 °C in This journal is © The Royal Society of Chemistry 2019 2868 \| Org. Biomol. Chem., 2019, 17, 2865–2872 View Article Online View Article Online Organic & Biomolecular Chemistry Organic & Biomolecular Chemistry Review Scheme 12 Aerobically-generated hypervalent iodine(III) dicarboxylate species 33. An early example of ethereal aerobic C–H bond activation was achieved in 2000 by Parsons and co-workers who used bro- motrichloromethane as a means of brominating the α-position on THF 34 (Scheme 13).31 The subsequent entity 36 was found to be unstable with regards to elimination and thereby resulted in oxocarbenium cation 37 formation. The group then used this cation as a means of tetrahydrofuranyl-protecting alcohol groups, a very desirable protecting group that can be selectively cleaved even in the presence of THP-ethers.32 It should be noted that owing to the use of THF as the reaction solvent, the substrate is used in a great excess (ca. 185 equivalents). It does not appear that eﬀorts were made to utilise the ether in a smaller stoichiometry coupled with possibly a greener, non- nucleophilic solvent (i.e. esters). Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Scheme 17 Synthesis of benzimidazoles 53 through aerobic C–H bond activation of benzylamines 44. activation, they were able to be eﬀectively utilised (45–77% yield) in the cross-coupling with benzylamines when used in excess (1.8 equivalents). Whilst the exact mechanistic pathway was not discerned, it was clear dioxygen played a role in the reaction via aerobic C–H bond activation. Following this, Troisi then briefly showed the possibility of the tetrahydrofuranyl radical 35 participating in radical addition to imines and an alkyne 42 (Scheme 15). Troisi postu- lated that the reaction was likely to follow a similar mecha- nism. Although substrate scope was scarcely explored, and THF was again used as the reaction solvent, it was shown that the ethereal radical was able to add to CvN bonds and CuC bonds, both of which were unexplored with respect to aerobi- cally-induced acyl radical addition. Nguyen et al. then reported a similar reaction that forgoes the need for a solvent altogether.35 Nguyen reported that simply heating benzylamines under an oxygen environment was suﬃcient for achieving oxidative coupling of amines to imines in good to excellent yields (62–83%) in 24 h. Nguyen reported that whilst eﬀective benzylamine auto-oxidation was observed utilising a flask exposed to air, evaporation of volatile benzylamines rendered the method inconvenient. Nonetheless, the reaction was proven to be eﬀective, and Nguyen et al. used this process of generating imines to con- struct many medicinally relevant nitrogen heterocycles (i.e. benzimidazoles, Scheme 17). Benzylamines Benzylamines represent a class of molecules that have gath- ered recent attention due to observed imine formation in reac- tions that are heated in the presence of air. The Fu and Nguyen groups have both exploited this and have shown that aerobically-induced C–H bond activation of benzylamines 44 resulted in oxidative coupling to form imines 50 (Scheme 16).34,35 Fu and co-workers showed that by utilising a dioxygen balloon, metal-free aerobic oxidative coupling of amines can be achieved whilst utilising water as the reaction solvent, with good to excellent yields (51–85%) achieved in 12–64 h. Furthermore, there was no product formation observed when the reaction was conducted under a nitrogen atmosphere, highlighting the significance of dioxygen for the reaction. The group also declared an unchanging yield when the reaction was conducted in a Teflon-reactor, excluding the possibility of light- or glass-catalysed processes. Although ali- phatic amines have low susceptibility to aerobic C–H bond This journal is © The Royal Society of Chemistry 2019 Ethers Owing to their common usage as reaction solvents, ethereal auto-oxidation is the most studied and understood auto-oxi- dation process,29,30 however it should be noted that it is uncommonly utilised as a means for aerobic C–H bond acti- vation. This is likely due to final oxidation product of ethers being relatively explosive peroxide species, meaning that his- torically, measures have been taken to prevent the activation of ethereal C–H bonds through dioxygen interaction (through use of an inert atmosphere or intentional doping with radical inhibitors). Nonetheless, a few research groups have shown the eﬀective utilisation of (cyclic) ethers as precursors for aerobic C–H bond activation. All ethereal C–H bond activations have been carried out whilst utilising the ether substrate as the reaction solvent, which represents a great excess of the ether substrate. Owing to the observed sensitivity of ethers to auto- oxidation, all methods have utilised transient dioxygen in the atmosphere and forgo the need for a dioxygen balloon. Scheme 13 Aerobically-initiated protocol for the protection of alcohols 38 as 2-tetrahydrofuranyl ethers 39 utilising bromotrichloromethane. Scheme 14 Aerobic protocol for the protection of alcohols 38 as 2-tetrahydrofuranyl ethers 39 utilising allyl chloride. Scheme 14 Aerobic protocol for the protection of alcohols 38 as 2-tetrahydrofuranyl ethers 39 utilising allyl chloride. Scheme 14 Aerobic protocol for the protection of alcohols 38 as 2-tetrahydrofuranyl ethers 39 utilising allyl chloride. This journal is © The Royal Society of Chemistry 2019 Org. Biomol. Chem., 2019, 17, 2865–2872 \| 2869 Scheme 15 Ethereal radical addition to phenylacetylene 42 to aﬀord alkene 43. Review View Article Online View Article Online Review Review Organic & Biomolecular Chemistry Organic & Biomolecular Chemistry Scheme 17 Synthesis of benzimidazoles 53 through aerobic C–H bond activation of benzylamines 44. Scheme 15 Ethereal radical addition to phenylacetylene 42 to aﬀord alkene 43. Scheme 15 Ethereal radical addition to phenylacetylene 42 to aﬀord alkene 43. carried out in the presence of TEMPO. Interestingly, Troisi et al. then provided further evidence by observing no reaction when utilising commercially stabilised THF, and then rapid subsequent reaction upon addition of CrCl2 oxidant to the reaction mixture, which consumed the antioxidant stabiliser and allowed uninhibited aerobic C–H bond activation. References 1 B. A. Arndtsen, R. G. Bergman, T. A. Mobley and T. H. Peterson, Acc. Chem. Res., 1995, 28, 154–162. 2 H. M. L. Davies and D. Morton, ACS Cent. Sci., 2017, 3, 936–943. 3 K. Godula and D. Sames, Science, 2006, 312, 67–72. 3 K. Godula and D. Sames, Science, 2006, 312, 67–72. 4 E. J. E. Caro-Diaz, M. Urbano, D. J. Buzard and R. M. Jones, Bioorg. Med. Chem. Lett., 2016, 26, 5378–5383. For the coupling of the glycine imine to styrene moieties 59, an open flask with access to air was utilised (Scheme 19). However, yields were comparatively low (28–53%) owing to an alternative pathway providing formation of self-oxidation side product 61 (7–40% yield). Finally, the use of electron-rich arenes allowed for eﬃcient nucleophilic addition to the proto- nated imine intermediate at room temperature and utilising atmospheric dioxygen. Moderate to good yields were observed (34–76%) when utilising a stoichiometric amount of arene to glycine ester. 5 M. C. Bryan, P. J. Dunn, D. Entwistle, F. Gallou, S. G. Koenig, J. D. Hayler, M. R. Hickey, S. Hughes, M. E. Kopach, G. Moine, P. Richardson, F. Roschangar, A. Steven and F. J. Weiberth, Green Chem., 2018, 20, 5082– 5103. 6 R. Shang, L. Ilies and E. Nakamura, Chem. Rev., 2017, 117, 9086–9139. 7 X. X. Guo, D. W. Gu, Z. Wu and W. Zhang, Chem. Rev., 2015, 115, 1622–1651. To provide further proof of the mechanism for benzylamine auto-oxidation, the group showed that when 1 mol% of MnO2 was added at the end of a reaction, gas evolution was observed. This suggested the formation of H2O2 as an auto-oxidation by-product since MnO2 reacts with H2O2 to form dioxygen and water. 8 M. A. Légaré, M. A. Courtemanche, É. Rochette and F. G. Fontaine, Science, 2015, 349, 513–516. 9 E. Roduner, W. Kaim, B. Sarkar, V. B. Urlacher, J. Pleiss, R. Gläser, W. D. Einicke, G. A. Sprenger, U. Beifuß, E. Klemm, C. Liebner, H. Hieronymus, S. F. Hsu, B. Plietker and S. Laschat, ChemCatChem, 2013, 5, 82– 112. 10 M. Conte, H. Miyamura, S. Kobayashi and V. Chechik, Chem. Commun., 2010, 46, 145–147. There are no conflicts to declare. For the reaction with indoles, a dioxygen balloon was uti- lised at 40 °C. Reaction times varied (4–60 h) with average to high yields reported (30–83%). Analogous reactions conducted under an argon atmosphere reported a significant drop in yield, thus indicating the significance of dioxygen for the reac- tion. Furthermore, no desired product was observed upon the addition of TEMPO, thereby providing further evidence of a radical mechanism. Open Access Article. Published on 13 February 2019. Downloaded on 9/11/2019 7:00:06 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Scheme 19 Utilising aerobic generation of 58 to facilitate Povarov/ aromatisation tandem reactions with styrenes 59. This journal is © The Royal Society of Chemistry 2019 Conﬂicts of interest nucleophilic attack on the imine intermediate formed via aerobic activation; styrenes, indoles and arenes were used as nucleophilic partners for C–C bond formation.36 There are no conflicts to declare. Glycine derivatives Pioneering research by Huo and co-workers has shown the functionalisation of the α-carbon in glycine derivatives 54.36,37 The modification of glycine is particularly attractive for medic- inal chemistry owing to the possibility of simple diversification of unnatural amino acids. Since 2014, Huo and co-workers have exploited the auto-oxidation of glycine derivatives 54. The mechanism for glycine auto-oxidation is similar to that dis- played by benzylamines in which an imine intermediate 58 is formed through the release of H2O2 (Scheme 18). The group have developed elegant methodologies for C–C formation via Scheme 16 Aerobic oxidative coupling of benzylamines 44. Scheme 18 Aerobic C–H bond activation of glycine derivatives 54 Scheme 16 Aerobic oxidative coupling of benzylamines 44. Scheme 18 Aerobic C–H bond activation of glycine derivatives 54. Scheme 16 Aerobic oxidative coupling of benzylamines 44. Scheme 18 Aerobic C–H bond activation of glycine derivatives 54. 2870 \| Org. Biomol. Chem., 2019, 17, 2865–2872 This journal is © The Royal Society of Chemistry 2019 View Article Online Organic & Biomolecular Chemistry Organic & Biomolecular Chemistry Review Scheme 19 Utilising aerobic generation of 58 to facilitate Povarov/ aromatisation tandem reactions with styrenes 59. g y hydes, ethers, benzylamines and glycine derivatives) it is hoped that these breakthroughs will lead to further eﬀorts in exploring novel motifs that are amenable to this methodology. For example, by looking more closely at other molecules/func- tionalities that “degrade” aerobically and/or appraising various parameters (e.g. dissolved oxygen concentration of solvent, stir- ring rate, surface area to volume ratio, etc.) that promote the aerobic oxidation of alternative C–H bearing functional groups. Conclusions 11 L. Vanoye, A. Favre-Réguillon, A. Aloui, R. Philippe and C. de Bellefon, RSC Adv., 2013, 3, 18931–18937. Aerobic C–H bond activation in which dioxygen interacts and activates a C–H bond is a fundamentally green progress with regards to atom economy and waste minimisation when com- pared with typical C–H activation reagents. Utilising dioxygen as the activating species ensures the use of an inherently green oxidant species (especially through the use of atmospheric dioxygen) which is of timely interest as scientists strive for more environmentally friendly protocols. In this review we highlight the realisation of new synthetic protocols that utilise aerobic C–H bond activation as a means of initiating radical reactions. Whilst these new reactions use relatively few func- tional groups as aerobic C–H activation precursors (e.g. alde- 12 V. Chudasama, A. R. Akhbar, K. A. Bahou, R. J. Fitzmaurice and S. Caddick, Org. Biomol. Chem., 2013, 11, 7301– 7317. 13 R. J. Fitzmaurice, J. M. Ahern and S. Caddick, Org. Biomol. Chem., 2009, 7, 235–237. 14 V. Chudasama, R. J. Fitzmaurice, J. M. Ahern and S. Caddick, Chem. Commun., 2010, 46, 133–135. 15 V. Chudasama, R. J. Fitzmaurice and S. Caddick, Nat. Chem., 2010, 2, 592–596. 16 V. Chudasama, J. M. Ahern, R. J. Fitzmaurice and S. Caddick, Tetrahedron Lett., 2011, 52, 1067–1069. Org. Biomol. Chem., 2019, 17, 2865–2872 \| 2871 View Article Online Review Organic & Biomolecular Chemistry Review Organic & Biomolecular Chemistry 17 A. Shamsabadi, J. Ren and V. Chudasama, RSC Adv., 2017, 7, 27608–27611. 28 A. Maity, S. M. Hyun and D. C. Powers, Nat. Chem., 2018, 10, 200–204. 29 S. Di Tommaso, P. Rotureau, O. Crescenzi and C. Adamo, Phys. Chem. Chem. Phys., 2011, 13, 14636–14645. 18 A. Maruani, M. T. W. Lee, G. Watkins, A. R. Akhbar, H. Baggs, A. Shamsabadi, D. A. Richards and V. Chudasama, RSC Adv., 2016, 6, 3372–3376. 30 S. Di Tommaso, P. Rotureau, B. Sirjean, R. Fournet, W. Benaissa, P. Gruez and C. Adamo, Process Saf. Prog., 2014, 33, 64–69. 19 A. Shamsabadi and V. Chudasama, Org. Biomol. Chem., 2017, 15, 17–33. 20 A. Shamsabadi and V. Chudasama, Chem. Commun., 2018, 54, 11180–11183. 31 J. M. Barks, B. C. Gilbert, A. F. Parsons and B. Upeandran, Tetrahedron Lett., 2000, 41, 6249–6252. 21 A. R. Akhbar, V. Chudasama, R. J. Fitzmaurice, L. Powell and S. Caddick, Chem. Commun., 2014, 50, 743–746. 32 C. G. Kruse, N. L. J. M. 2872 \| Org. Biomol. Chem., 2019, 17, 2865–2872 This journal is © The Royal Society of Chemistry 2019 Conclusions Broekhof and A. van der Gen, Tetrahedron Lett., 1976, 17, 1725–1728. 33 L. Troisi, C. Granito, L. Ronzini, F. Rosato and V. Videtta, Tetrahedron Lett., 2010, 51, 5980–5983. 22 G. N. Papadopoulos and C. G. Kokotos, Chem. – Eur. J., 2016, 22, 6964–6967. 34 L. Liu, S. Zhang, X. Fu and C. H. Yan, Chem. Commun., 2011, 47, 10148–10150. 23 V. Chudasama, J. M. Ahern, D. V. Dhokia, R. J. Fitzmaurice and S. Caddick, Chem. Commun., 2011, 47, 3269–3271. 35 T. B. Nguyen, L. Ermolenko and A. Al-Mourabit, Green Chem., 2013, 15, 2713–2717. 25 P. Biswas and J. Guin, J. Org. Chem., 2018, 83, 5629–5638. and J. Guin, J. Org. Chem., 2018, 83, 5629–5638. 36 C. Huo, Y. Yuan, M. Wu, X. Jia, X. Wang, F. Chen and J. Tang, Angew. Chem., Int. Ed., 2014, 53, 13544–13547. 26 P. Biswas, S. Paul and J. Guin, Synlett, 2017, 28, 1244–1249. 27 K. Miyamoto, J. Yamashita, S. Narita, Y. Sakai, K. Hirano, T. Saito, C. Wang, M. Ochiai and M. Uchiyama, Chem. Commun., 2017, 53, 9781–9784. 37 Y. Wei, J. Wang, Y. Wang, X. Yao, C. Yang and C. Huo, Org. Biomol. Chem., 2018, 16, 4985–4989. 2872 \| Org. Biomol. Chem., 2019, 17, 2865–2872
https://openalex.org/W3137830125	https://www.researchsquare.com/article/rs-186071/v1.pdf?c=1631871360000	English	null	Nonlinear dynamic response of a Negative Stiffness-Shape Memory Alloy isolation system	Research Square (Research Square)	2,021	cc-by	15,520	Nonlinear dynamic response of a Negative Stiffness-Shape Memory Alloy isolation system vatore (  a.salvatore@uniroma1.it ) degli Studi di Roma La Sapienza https://orcid.org/0000-0001-9718-9871 Nonlinear dynamic response of a Negative Stiffness-Shape Memory Alloy isolation system Andrea Salvatore (  a.salvatore@uniroma1.it ) Universita degli Studi di Roma La Sapienza https://orcid.org/0000-0001-9718-9871 Biagio Carboni Universita degli Studi di Roma La Sapienza Walter Lacarbonara Sapienza University of Rome: Universita degli Studi di Roma La Sapienza Research Article Keywords: High-static-low-dynamic stiffness isolator, Quasi-zero-stiffness isolation, Negative stiffness, Pseudo-elastic hysteresis Posted Date: March 1st, 2021 DOI: https://doi.org/10.21203/rs.3.rs-186071/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 Nonlinear Dynamics on July 18th, 2021. See the published version at https://doi.org/10.1007/s11071-021-06666-y. Nonlinear dynamic response of a Negative Stiffness-Shape Memory Alloy isolation system Andrea Salvatore (  a.salvatore@uniroma1.it ) Universita degli Studi di Roma La Sapienza https://orcid.org/0000-0001-9718-9871 Biagio Carboni Universita degli Studi di Roma La Sapienza Walter Lacarbonara Sapienza University of Rome: Universita degli Studi di Roma La Sapienza Research Article Keywords: High-static-low-dynamic stiffness isolator, Quasi-zero-stiffness isolation, Negative stiffness, Pseudo-elastic hysteresis Posted Date: March 1st, 2021 DOI: https://doi.org/10.21203/rs.3.rs-186071/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 Nonlinear Dynamics on July 18th, 2021. See the published version at https://doi.org/10.1007/s11071-021-06666-y. Version of Record: A version of this preprint was published at Nonlinear Dynamics on July 18th, 2021. See the published version at https://doi.org/10.1007/s11071-021-06666-y. Research Article License:   This work is licensed under a Creative Commons Attribution 4.0 International License. R d F ll Li 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 Nonlinear Dynamics on July 18th, 2021. See the published version at https://doi.org/10.1007/s11071-021-06666-y. Nonlinear Dynamics manuscript No. (will be inserted by the editor) Nonlinear Dynamics manuscript No. (will be inserted by the editor) Nonlinear dynamic response of a Negative Stiffness-Shape Memory Alloy isolation system Andrea Salvatore · Biagio Carboni · Walter Lacarbonara Received: date / Accepted: date Abstract The negative stiffness exhibited by bi-stable mechanisms together with tunable hysteresis in the context of vibration isolation devices can enhance the dynamic resilience of a structure. The effects of negative stiffness and shape memory alloy (SMA) damping in base-isolated structures are here explored by carrying out an extensive study of the non- linear dynamic response via pathfollowing, bifurcation analysis, and time integration. The frequency-response curves of the isolated structure, with and without the negative stiffness contribution, are numerically obtained for different excitation amplitudes to construct the acceleration and displacement transmissibility curves. The advantages of negative stiffness, damping augmentation and reduced accelerations and displacements transmissibility, as well as the existence of rich bifurcation scenarios giving rise to quasi-periodicity and synchro- nization, are extensively illustrated. Keywords High-static-low-dynamic stiffness isolator · Quasi-zero-stiffness isolation · Negative stiffness · Pseudo-elastic hysteresis A. Salvatore Sapienza University of Rome E-mail: a.salvatore@uniroma1.it 1 Introduction One of the main beneﬁts is the possibility of reaching levels of deformability of the iso- lation layer beyond the limit represented by the deformability of the material constituting the isolation devices, obtaining quasi zero stiffness and ultra-low frequency isolation, hence substantial reduction in acceleration transmissibility otherwise not achievable through exist- ing devices. Secondly, it is possible to leverage on the possibility of introducing high levels of hysteretic damping and amplifying the existing damping without loss of performance due to an increase in initial stiffness, given the fact that this increase is cancelled out by the ap- propriately tailored negative stiffness. For a careful investigation of the nonlinear stationary response of a SDOF representing the isolated structure controlled with a negative stiffness- SMA (NS-SMA) damping mechanism, both the nonlinear response of the bearing devices and the super-elastic response of SMA wires are modelled employing suitable hysteretic constitutive laws. 1 Introduction The stiffness of SMA wires overcomes the negative stiffness exhibited by the prestressed spring until phase transformation occurs. For larger displacements, the stiffness of SMA wires vanishes and the overall stiffness becomes quasi zero. The transmissibility of a SDOF with the proposed response was analytically evaluated using a piece-wise linear constitutive law for the SMA response and a linear elastic law for the negative stiffness contribution. In the present work, the dynamic response of a classical seismic isolation system made up of elastomeric bearings working in parallel with a neg- ative stiffness mechanism and SMA wires is explored. The SMA wires are used in order to realize the initial gap and, at the same time, provide hysteretic damping to the system while the negative stiffness contribution is used to achieve a zero dynamic stiffness away from the equilibrium position. The beneﬁts associated with a negative stiffness mechanism together with hysteretic damping for an improved seismic isolation system are multifaceted. One of the main beneﬁts is the possibility of reaching levels of deformability of the iso- lation layer beyond the limit represented by the deformability of the material constituting the isolation devices, obtaining quasi zero stiffness and ultra-low frequency isolation, hence substantial reduction in acceleration transmissibility otherwise not achievable through exist- ing devices. Secondly, it is possible to leverage on the possibility of introducing high levels of hysteretic damping and amplifying the existing damping without loss of performance due to an increase in initial stiffness, given the fact that this increase is cancelled out by the ap- propriately tailored negative stiffness. For a careful investigation of the nonlinear stationary response of a SDOF representing the isolated structure controlled with a negative stiffness- SMA (NS-SMA) damping mechanism, both the nonlinear response of the bearing devices and the super-elastic response of SMA wires are modelled employing suitable hysteretic constitutive laws. in order to achieve an initial gap in the ensuing restoring force of the device. Other ways to achieve negative stiffness responses are the use of reverse bending sliding surfaces or magneto-rheological dampers with linear voltage decay [3 -9]. 1 Introduction There are several develop- ments of this concept applied to seismic isolation that show very promising results in terms of reduction of both accelerations and displacements transmissibilities in [10 -32], where a linear vertical isolation system is improved by introducing negative stiffness correction in order to achieve a high preload stiffness and a Quasi-Zero-Stiffness (QZS) in the equilibrium position. The nonlinear isolator response is usually described by a Dufﬁng oscillator with a vanishing linear stiffness. Donmez et al. [33] studied the dynamic response of a dry-friction QZS isolator, showing that the hysteretic damping ensures a better performance than viscous damping in the out-of-resonance frequency range. In all of these works negative stiffness correction is used to achieve a zero stiffness in the equilibrium position, but as known, in a typical civil application dealing with seismic horizontal isolation, this is unwanted because of the need of a wind restraint. By delaying the negative stiffness contribution through an initial gap, the wind restraint is preserved, and transmissibility reduction can be achieved. Liu, Zhung et al. [34] proposed a novel isolation system composed by Shape Memory Alloy (SMA) wires and prestressed spring. The stiffness of SMA wires overcomes the negative stiffness exhibited by the prestressed spring until phase transformation occurs. For larger displacements, the stiffness of SMA wires vanishes and the overall stiffness becomes quasi zero. The transmissibility of a SDOF with the proposed response was analytically evaluated using a piece-wise linear constitutive law for the SMA response and a linear elastic law for the negative stiffness contribution. In the present work, the dynamic response of a classical seismic isolation system made up of elastomeric bearings working in parallel with a neg- ative stiffness mechanism and SMA wires is explored. The SMA wires are used in order to realize the initial gap and, at the same time, provide hysteretic damping to the system while the negative stiffness contribution is used to achieve a zero dynamic stiffness away from the equilibrium position. The beneﬁts associated with a negative stiffness mechanism together with hysteretic damping for an improved seismic isolation system are multifaceted. 1 Introduction Passive vibration isolation is one of the most popular and effective vibration control tech- niques. It is well known that for a SDOF oscillator, under harmonic forcing, there is a deam- pliﬁcation of the incoming accelerations for frequencies higher than √ 2f0, where f0 is the natural frequency of the oscillator. Therefore, the lower the natural frequency (i.e., the stiff- ness), the higher the isolated frequency bandwidth. On the other hand, an excessively low stiffness induces large static displacements. Hence, the need to obtain high static, low dy- namic stiffness isolation systems. A very promising technique to achieve this goal is the exploitation of negative stiffness devices in parallel with classical vibration isolation de- vices. Pasala, Nagarajaiah et al. [1,2] proposed and carried out analytical studies on an adaptive stiffness system with a negative stiffness component to simulate early yielding that can re- duce seismic demands, whereby a bilinear spring is used in parallel with negative stiffness 2 A. Salvatore et al in order to achieve an initial gap in the ensuing restoring force of the device. Other ways to achieve negative stiffness responses are the use of reverse bending sliding surfaces or magneto-rheological dampers with linear voltage decay [3 -9]. There are several develop- ments of this concept applied to seismic isolation that show very promising results in terms of reduction of both accelerations and displacements transmissibilities in [10 -32], where a linear vertical isolation system is improved by introducing negative stiffness correction in order to achieve a high preload stiffness and a Quasi-Zero-Stiffness (QZS) in the equilibrium position. The nonlinear isolator response is usually described by a Dufﬁng oscillator with a vanishing linear stiffness. Donmez et al. [33] studied the dynamic response of a dry-friction QZS isolator, showing that the hysteretic damping ensures a better performance than viscous damping in the out-of-resonance frequency range. In all of these works negative stiffness correction is used to achieve a zero stiffness in the equilibrium position, but as known, in a typical civil application dealing with seismic horizontal isolation, this is unwanted because of the need of a wind restraint. By delaying the negative stiffness contribution through an initial gap, the wind restraint is preserved, and transmissibility reduction can be achieved. Liu, Zhung et al. [34] proposed a novel isolation system composed by Shape Memory Alloy (SMA) wires and prestressed spring. 2 Negative stiffness-SMA damper for seismic isolation In previous works, a bilinear spring is used in parallel with the negative stiffness in order to achieve an initial gap in the ensuing restoring force of the device. This gap allows to maintain for low amplitudes the virgin isolation stiffness and this is useful to realize a wind restraint. In this work, a super-elastic spring (Shape Memory Alloy wire) is used instead of the bilinear spring in order to realize the initial gap and, at the same time, deliver hysteretic 3 Nonlinear dynamic response of a NS-SMA isolation system damping to the system. The total restoring force f of the proposed isolation system is thus the summation of the force fi provided by response of the traditional seismic elastomeric isolators, the force fs of SMA wires, and that provided by the negative stiffness mechanism denoted by fn; that is, (1) f = fi +( fn + fs) = fi + fns (1) where fns = fn + fs is the force of the proposed rheological device. For displacements below the gap, the stiffness of the wires is equal to the negative stiffness, hence, the overall response coincides with that of the elastomeric element. For displacements greater than the gap dis- placement, corresponding to the SMA wires yielding displacement, the negative stiffness strongly reduces the total force (base shear force) and stiffness. For larger displacements, the cubic term tends to overcome the negative stiffness contribution, thus making the overall response follow the baseline (elastomeric) backbone response (see Fig. 1). Fig. 1 Force-displacement cycles associated with (left) damper force fns (black line) ensuing from fs (red line) plus fn (blue line) and with (right) the overall system response with and without damper (gray and black lines, respectively). Fig. 1 Force-displacement cycles associated with (left) damper force fns (black line) ensuing from fs (red line) plus fn (blue line) and with (right) the overall system response with and without damper (gray and black lines, respectively). In the subsequent sections, 3 types of isolated SDOF systems are studied and compared. 2.1.1 Elastomeric isolators Elastomeric isolation systems are usually described by the Bouc-Wen model of hysteresis [35, 36] together with a linear viscous damping. The model is the direct summation of a viscous damping force, an elastic force and a hysteretic force. That is, fi = c˙x+αKix+(1−α)Kiz, (2) (2) fi = c˙x+αKix+(1−α)Kiz, where the hysteretic force z is governed by where the hysteretic force z is governed by ˙z = ˙x[1−(γ +βsign(z˙x)]\|z\|n, (3) (3) where c is the viscous damping coefﬁcient, α is the ratio between the post-elastic and the ini- tial stiffness Ki, γ and β control the shape of the hysteresis loops, n regulates the smoothness of transition between initial elastic and post-elastic stiffness. The upper and lower bounds of z are given by zm = np (1−α)Ki/(γ +β). In the present study, γ + β is restricted to be positive in order to have a softening behaviour and n is set to 1 (Fig. 3). Isolation system parameters Ki Initial stiffness 0.2 0.01 ! 0.09 n 1 c 0.1 Fig. 3 Force-displacement cycles of isolation devices (left) and model parameters (right). Isolation system parameters Ki Initial stiffness 0.2 0.01 ! 0.09 n 1 c 0.1 Fig. 3 Force-displacement cycles of isolation devices (left) and model parameters (right). 2 Negative stiffness-SMA damper for seismic isolation We conveniently identify them as follows: the baseline elastomeric isolation system (EIS) is re- ferred to as S1 and the associated restoring force is f = fi, S2 denotes the baseline isolation system together with the SMA damper alone and the associated restoring force is f = fi + fs and, ﬁnally, S3 indicates the proposed isolation system with the NS-SMA damper in parallel with the elastomeric bearings (see Fig. 2). In the subsequent sections, 3 types of isolated SDOF systems are studied and compared. We conveniently identify them as follows: the baseline elastomeric isolation system (EIS) is re- ferred to as S1 and the associated restoring force is f = fi, S2 denotes the baseline isolation system together with the SMA damper alone and the associated restoring force is f = fi + fs and, ﬁnally, S3 indicates the proposed isolation system with the NS-SMA damper in parallel with the elastomeric bearings (see Fig. 2). Fig. 2 Schematic representation of the 3 different isolated systems referred to as: S1, S2 and S3. Fig. 2 Schematic representation of the 3 different isolated systems referred to as: S1, S2 and S3. A. Salvatore et al 4 2.1 Rheological models 2.1 Rheological models 2.1.1 Elastomeric isolators 2.1.1 Elastomeric isolators 2.1.2 Negative stiffness mechanism In most passive bi-stable mechanisms, the negative stiffness is produced by geometric non- linearities that cause negative stiffness only in a given displacement range and positive stiff- ness outside. In this work the positive force exhibited past the displacement xf , where the force vanishes, is cancelled by means of a step function. The NS force reads fn = (−Knx+K3x3)(1+sign(xf −\|x\|)) 2 (4) (4) where Kn is the negative linear stiffness, K3 is the positive cubic stiffness, and x f is the dis- placement corresponding to a vanishing force and is equal to xf = p Kn/K3. Another char- acteristic displacement is the displacement where the maximum negative force is reached given by xn = p Kn/3K3 (see Fig.4). Nonlinear dynamic response of a NS-SMA isolation system 5 Negative stiffness parameters Kn Negative stiffness xf Limit displacement K3 Kn/xf2 Fig. 4 Force-displacement cycles provided by the negative stiffness force (left) and model parameters (right). Fig. 4 Force-displacement cycles provided by the negative stiffness force (left) and model parameters (right). 2.1.3 SMA wires The response of SMA wires is modelled according to the phenomenological super-elastic hysteretic model proposed by Charalampakis [37] given in rate form: ˙fs = (1−s)Ks[˙x−\|˙x\|sign( fs −βs)( \| fs−βs\| Y )ns]+sKm ˙x, (5) βs = Ksαs[x−fs Ks + fttanh(asx)[ 1+sign(−x˙x) 2 ]], (6) s = tanh[cs(\|x\|−xm)]+1 2 , (7) (5) (6) (7) (7) where Ks is the initial stiffness during the austenitic phase,Y is the yielding force, αs controls the post-elastic stiffness, ns regulates the smoothness of transition from the initial elastic to the post-elastic phase, ft controls the twinning hysteresis and super-elasticity, as governs the pinching around the origin during cyclic loading, Km is the stiffness during the fully martensitic phase, xm is the displacement at which the transition from the post-elastic to the fully martensitic phase occurs, while cs control the smoothness of this transition (see Fig. 5). where Ks is the initial stiffness during the austenitic phase,Y is the yielding force, αs controls the post-elastic stiffness, ns regulates the smoothness of transition from the initial elastic to the post-elastic phase, ft controls the twinning hysteresis and super-elasticity, as governs the pinching around the origin during cyclic loading, Km is the stiffness during the fully martensitic phase, xm is the displacement at which the transition from the post-elastic to the fully martensitic phase occurs, while cs control the smoothness of this transition (see Fig. 5). SMA wires parameters Y Yielding force xg Gap displ. Ks Y/xg αs 0.01 ft 6000 ns 3 as 0.06 Km 0.5 Ks cs 0.02 xm Martensitic displ. Fig. 5 Force-displacement cycles of the SMA wires (left) and model parameters (right). SMA wires parameters Y Yielding force xg Gap displ. Ks Y/xg αs 0.01 ft 6000 ns 3 as 0.06 Km 0.5 Ks cs 0.02 xm Martensitic displ. Fig. 5 Force-displacement cycles of the SMA wires (left) and model parameters (right). 6 A. Salvatore et al Because of the large variability and dependence of parameters ft and as on the remaining parameters, two new parameters with a more straightforward physical interpretation, ys and ˜as, are introduced. In terms of these new parameters, ft and as are expressed as: ft = (2Y −ysY)/(αsKs), (8) as = tanh−1( ˜asKs)/(Y −ysY). (9) (8) (9) (9) The advantage of introducing these new parameters is associated with the possibility of ob- taining the same cycles by changing a parameter of interest, useful for parametric analyses. 2.1.3 SMA wires The advantage of introducing these new parameters is associated with the possibility of ob- taining the same cycles by changing a parameter of interest, useful for parametric analyses. The ﬁrst parameter indicates the difference between the loading and unloading forces ex- pressed as percentage of Y, the second allows has been reformulated in order to obtain, with the same coefﬁcient, the same residual displacement, and therefore the same hysteresis loop shape, for different values of the other parameters (see Fig. 6). In particular, ˜as indicates the value assumed by tanh( ˜asx) for a displacement x at which the continuation of the unloading branch intersects the elastic loading branch. Fig. 6 Variation of the force-displacement cycles of SMA wires with the nondimensional parameters ys (left) and ˜as (right). Fig. 6 Variation of the force-displacement cycles of SMA wires with the nondimensional parameters ys (left) and ˜as (right). 2.1.4 Design Parameters The main goal is to investigate the effects of the NS-SMA device in parallel with a traditional elastomeric device as a hysteretic isolation system. Hence, the stiffness Ki of the isolation system, the gap displacement xg = Fw/Ki , (with Fw the maximum expected wind load) and the maximum allowed displacement xu are assumed as input parameters (see Fig. 7). In this study as reasonable value of xg is selected xg = 0.05xu. On the other hand, Kn, Y and ys are the design parameters. All the remaining parameters are either set on ﬁxed values (c = 0.1, α = 0.2, β = 0.09, γ = 0.01, n = 1, αs = 0.01, ns = 3, ˜as = 0.6, cs = 0.02) or determined accordingly xf = xm = 0.7xu, K3 = Kn/x2 f , Ks = Y/xg, ft = (2Y −ysY)/(αsKs), as = tanh−1( ˜asKs)/(Y −ysY), Km = 0.5Ks. Nonlinear dynamic response of a NS-SMA isolation system 7 Input parameters Ki Initial stiffness xg Gap displacement xu Maximum displacement Design parameters Kn Negative stiffness Y Yielding force ys Hysteretic ratio 0.7 Fig. 7 Overall force-displacement cycles provided by S3 (left) and system parameters (right) 0.7 Input parameters Ki Initial stiffness xg Gap displacement xu Maximum displacement Design parameters Kn Negative stiffness Y Yielding force ys Hysteretic ratio Fig. 7 Overall force-displacement cycles provided by S3 (left) and system parameters (right) 2.1.5 Nondimensional equation of motion 2.1.5 Nondimensional equation of motion The equation of motion of a SDOF mass m subject to the hysteretic restoring force given by Eq. (1) reads: m¨x+c˙x+αKix+(1−α)Kiz+(−Knx+K3x3)(1+sign(x f −\|x\|)) 2 + fs = Psin(Ωgt). (10) By choosing a characteristic displacement x0 = xu and a characteristic stiffness Ki (i.e., the initial stiffness of the isolation system, equal to the initial stiffness of the Bouc-Wen model), the following nondimensional variables are introduced ˜x = x/x0, ˜t = ωt, where ω = p Ki/m while the characteristic force is N0 = Kix0. By dividing the equation of motion by N0, the following nondimensional equation of motion is obtained: ¨˜x+ζ ˙˜x+α ˜x+(1−α)˜z+(−˜Kn ˜x+ ˜K3 ˜x3)(1+sign(˜xf −\|˜x\|)) 2 + ˜fs = ˜Psin( ˜Ωg˜t), (11) (11) with ˜P = P/N0, ˜Ωg = Ωg/ω. Consequently, the nondimensional elastomeric isolators force becomes f ˜fi = fi N0 = ζ ˙˜x+α ˜x+(1−α)˜z, (12) (12) where ζ = cω/Ki. 2.1.4 Design Parameters Equation (3) can be rewritten in nondimensional form as ˙˜z = ˙z x0ω = ˙˜x[1−( ˜γ + ˜βsign(˜z˙˜x))\|˜z\|n], (13) (13) where ˜z = z x0 , ˜γ = γxn 0, ˜β = βxn 0.The nondimensional form of equation (4) becomes ˜fn = fn N0 = (−˜Kn ˜x+ ˜K3 ˜x3)(1+sign(˜x f −\|˜x\|)) 2 , (14) (14) with ˜Kn = Kn Ki , ˜K3 = K3x2 0 Ki , ˜x f = xf x0 . Finally, also Eqs. (5), (6) and (7) are rendered nondimen- sional as follows: ˙˜fs = (1−˜s) ˜Ks[˙˜x−\|˙˜x\|sign( ˜fs −˜βs)( \| ˜fs−˜βs\| ˜Y )ns]+ ˜s ˜Km ˙˜x, (15) ˜βs = ˜Ksαs[˜x− ˜fs ˜Ks + ˜fttanh( ˜as ˜x)[ 1+sign(−˜x˙˜x) 2 ]], (16) ˜s = tanh[˜cs(\|˜x\|−˜xm)]+1 2 , (17) (17) A. Salvatore et al 8 after introducing the following nondimensional parameters: ˜Ks = Ks Ki , ˜Km = Km Ki , ˜Y = Y N0 , ˜ft = ft x0 , ˜as = asx0, ˜cs = asx0, ˜xm = xm x0 . after introducing the following nondimensional parameters: ˜Ks = Ks Ki , ˜Km = Km Ki , ˜Y = Y N0 , ˜ft = ft x0 , ˜as = asx0, ˜cs = asx0, ˜xm = xm x0 . 3 Equivalent stiffness and damping A useful way to characterize the dynamic behavior of the nonlinear hysteretic oscillator is to make use of the equivalent linearization. An equivalent linear stiffness and an equivalent damping ratio are deﬁned for each displacement amplitude (U) as: Keq = R U −U Ktdx 2U ,ξ = Ed 4πEk (18) (18) where Kt denotes the tangent stiffness at the current displacement x, Ed denotes the to- tal dissipated energy and Ek is the stored energy for a displacement amplitude U. To start with, we analyse the effects of adding SMA wires alone in parallel to the elastomeric iso- lation system. In order to clarify the effects of the parameters, four different cases for the rheological system S2 are reported in Fig. 8: (Y = Zm, ˜Ks = α,ys = 0.2)(red dotted line), (Y = Zm, ˜Ks = α,ys = 0.5)(red dashed line), (Y = Zm, ˜Ks = α,ys = 0.8)(red solid line), (Y = 1.6Zm, ˜Ks = 1.6α,ys = 0.8)(red dashed dotted line). where Kt denotes the tangent stiffness at the current displacement x, Ed denotes the to- tal dissipated energy and Ek is the stored energy for a displacement amplitude U. To start with, we analyse the effects of adding SMA wires alone in parallel to the elastomeric iso- lation system. In order to clarify the effects of the parameters, four different cases for the rheological system S2 are reported in Fig. 8: (Y = Zm, ˜Ks = α,ys = 0.2)(red dotted line), (Y = Zm, ˜Ks = α,ys = 0.5)(red dashed line), (Y = Zm, ˜Ks = α,ys = 0.8)(red solid line), (Y = 1.6Zm, ˜Ks = 1.6α,ys = 0.8)(red dashed dotted line). Fig. 8 Equivalent nondimensional stiffness vs. nondimensional displacement amplitude (left) and equivalent damping vs. nondimensional amplitude (right) for S1 (black line), S2 with Y = Zm,ys = (0.2,0.5,0.8) (dotted, dashed and solid red line, respectively) and with Y = 1.6Zm,ys = 0.8 (dashed-dotted red line). The sub-ﬁgures show the hysteresis loops of device (a) and system (b) for the parameters described above. Fig. 8 Equivalent nondimensional stiffness vs. nondimensional displacement amplitude (left) and equivalent damping vs. nondimensional amplitude (right) for S1 (black line), S2 with Y = Zm,ys = (0.2,0.5,0.8) (dotted, dashed and solid red line, respectively) and with Y = 1.6Zm,ys = 0.8 (dashed-dotted red line). The sub-ﬁgures show the hysteresis loops of device (a) and system (b) for the parameters described above. 3 Equivalent stiffness and damping The trends of equivalent stiffness and damping in Fig. 8 show that for low ratios of super- elastic hysteresis (see dotted red curve) the only effect is an increase of stiffness that causes a reduction of equivalent damping in the low amplitude range. For greater ratios of super- elastic hysteresis (ys), a small increase of damping is achieved for moderate and large dis- placements, whereas, for low amplitudes, a decrease of damping persists. Increasing the yielding force (Y) entails an increase of initial stiffness and hence damping reduction at low amplitudes is more signiﬁcant. Moreover, a damping increment for large displacements is affected. Therefore, the introduction of SMA dampers within the isolation system, like any other hysteretic damper, regardless of the rate of hysteresis, causes a decrease in seismic 9 Nonlinear dynamic response of a NS-SMA isolation system 9 isolation performance for low intensity earthquakes but with high frequency return (because of an increase of stiffness and a subsequent decrease of damping), and a slight improvement of performance for medium and high intensity earthquakes. By considering next the case (Y = Zm,ys = 0.8), Fig. 9 shows that the introduction of negative stiffness drastically am- pliﬁes the equivalent damping. In fact, while the maximum increase of damping achieved with SMA wires alone was of ≃2%, with ˜Kn = (0.5,1,1.2)α, the achieved increments are ≃(5,25,40)%, respectively. As already seen for the case with only SMA wires, also with negative stiffness the increase of hysteretic ratio, ys, entails an increase of equivalent damp- ing, while the increase of yielding force Y gives rise to an increase of initial stiffness and thus a drop of equivalent damping over a broad range of displacements. = = Fig. 9 Equivalent nondimensional stiffness vs. nondimensional displacement amplitude (left) and equivalent damping vs. nondimensional amplitude (right) for S1 (black line), S2 with Y = Zm,ys = 0.8 (red line), S3 with ˜Kn = (0.5,1,1.2)α (magenta, violet, blue lines, respectively); the blue dashed-dotted lines represents the case ˜Kn = 1.2α,Y = 1.6Zm,ys = 0.8, while the subﬁgures show the hysteresis loops for the assigned parameters. = = Fig. 9 Equivalent nondimensional stiffness vs. nondimensional displacement amplitude (left) and equivalent damping vs. 3 Equivalent stiffness and damping nondimensional amplitude (right) for S1 (black line), S2 with Y = Zm,ys = 0.8 (red line), S3 with ˜Kn = (0.5,1,1.2)α (magenta, violet, blue lines, respectively); the blue dashed-dotted lines represents the case ˜Kn = 1.2α,Y = 1.6Zm,ys = 0.8, while the subﬁgures show the hysteresis loops for the assigned parameters. -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 0 0.25 0.5 0.75 1 1.25 Δξ [%] =0.05 /α Y=1.6Zm Y=Zm -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 0 0.25 0.5 0.75 1 1.25 Δξ [%] =0.05 /α -10 0 10 20 30 40 50 60 0 0.25 0.5 0.75 1 1.25 Δξ [%] = 0.2 /α Y=1.6Zm Y=Zm Fig. 10 Damping ampliﬁcation ∆ξ vs. negative stiffness coefﬁcient for a nondimensional displacement of 0.05 (left) and of 0.2 (right) for the S3 system with Y = Zm,ys = (0.2,0.5,0.8) (black dotted, dashed and solid lines, respectively) and with Y = 1.6Zm,ys = 0.8 represented by dashed-dotted gray lines. -10 0 10 20 30 40 50 60 0 0.25 0.5 0.75 1 1.25 Δξ [%] = 0.2 /α Fig. 10 Damping ampliﬁcation ∆ξ vs. negative stiffness coefﬁcient for a nondimensional displacement of 0.05 (left) and of 0.2 (right) for the S3 system with Y = Zm,ys = (0.2,0.5,0.8) (black dotted, dashed and solid lines, respectively) and with Y = 1.6Zm,ys = 0.8 represented by dashed-dotted gray lines. 10 A. Salvatore et al In Fig. 10 the damping ampliﬁcation due to negative stiffness is shown for two different displacement levels equal to (0.05, 0.2) X0. For the system with the SMA damper without negative stiffness correction, a decrease of damping is observed for low amplitudes while for the second amplitude a slight increase is observed. With the introduction of negative stiffness, a strong damping ampliﬁcation is achieved for both amplitudes. In Fig. 10 the damping ampliﬁcation due to negative stiffness is shown for two different displacement levels equal to (0.05, 0.2) X0. For the system with the SMA damper without negative stiffness correction, a decrease of damping is observed for low amplitudes while for the second amplitude a slight increase is observed. With the introduction of negative stiffness, a strong damping ampliﬁcation is achieved for both amplitudes. 3 Equivalent stiffness and damping By numerically integrating the median line of the hysteresis cycles, the potential energy proﬁle is obtained for the NS-SMA device alone and for the combined system S3 (see Fig. 11). The potential energy proﬁle of the device shows its bi-stable/tri-stable nature. In fact, for ˜Kn = (0.2,0.5)α, the proﬁle has three minima giving rise to three stable equilibria and two unstable equilibria. On the other hand, when the negative stiffness is equal or greater than the SMA elastic stiffness, ˜Kn = (1,1.2)α, the two unstable equilibria coalesce into one unstable equilibrium point at the origin instead of the original stable equilibrium point and the device becomes bi-stable. The addition of the damper to the isolation system gives rise to an erosion of the global potential energy proﬁle bringing it closer to the damped energy proﬁle, thereby increasing the ratio Ed/Ek, hence, the equivalent damping. Fig. 11 Potential energy proﬁles of the NS-SMA damper (left) and of the combined system S3 (right) for ˜Kn = (0,0.2,0.5,1,1.2)α, denoted by red, fuchsia, magenta, violet and blue lines, respectively. Fig. 11 Potential energy proﬁles of the NS-SMA damper (left) and of the combined system S3 (right) for ˜Kn = (0,0.2,0.5,1,1.2)α, denoted by red, fuchsia, magenta, violet and blue lines, respectively. In Fig. 12, it can be seen that for ˜Kn = 1.2α the response shows a negative stiffness range and the damped energy proﬁle, along the associated branch, is greater than the potential proﬁle, thus the system becomes overdamped. For values of ˜Kn > 1.2α, the median line of hysteresis loops intersects the abscissa line (i.e., the force vanishes). Consequently, 2 lateral minima in the energy proﬁle are born, hence the system becomes globally tri-stable. In order to obtain a self-recentering system, the existence of 2 additional equilibrium positions is un- desirable. This is why the negative stiffness ˜Kn needs to be limited. In the light of the above some useful hints can be obtained on the optimum negative stiffness; indeed, a negative stiffness between α < ˜Kn < 1.2α shows a consistent reduction of stiffness and ampliﬁcation of damping without loss of self-recentering capacity. The limit value of negative stiffness past which the median line of the hysteresis loops intersects the abscissa axis and the system becomes tri-stable can be analytically obtained as ˜Kn MS→TS = −α −αs ˜Ks −3 3 s (−1+αs)2 ˜K3 ˜K2s ˜x2g 4 . 3 Equivalent stiffness and damping (19) (19) Nonlinear dynamic response of a NS-SMA isolation system 11 Fig. 12 (left) Potential energy proﬁles of system S3 with ˜Kn = (1,1.2,1.3)α represented by magenta, blue, and violet lines, respectively and damped energy proﬁle denoted by gray dashed lines. The subﬁgure shows the total force displacement cycles together with the average force in dashed-dotted lines. (right) Analytical bounding curves between mono-stability and tri-stability in the parameters space of the negative stiffness mechanism ( ˜Kn, ˜K3) for the S3 system with ys = 0.8 and Y = (1,1.6)Zm represented by black solid and black dashed-dotted lines, respectively. Fig. 12 (left) Potential energy proﬁles of system S3 with ˜Kn = (1,1.2,1.3)α represented by magenta, blue, and violet lines, respectively and damped energy proﬁle denoted by gray dashed lines. The subﬁgure shows the total force displacement cycles together with the average force in dashed-dotted lines. (right) Analytical bounding curves between mono-stability and tri-stability in the parameters space of the negative stiffness mechanism ( ˜Kn, ˜K3) for the S3 system with ys = 0.8 and Y = (1,1.6)Zm represented by black solid and black dashed-dotted lines, respectively. 4 Nonlinear dynamic response scenario The FRCs of the described nonlinear hysteretic oscillators endowed with the rheological devices S1, S2, S3 are numerically obtained for several excitation levels employing a con- tinuation procedure based on the Poincar`e map. The Poincar`e map and the associated mon- odromy matrix are computed via the fourth order Runge-Kutta integration scheme and the ﬁnite difference method, respectively. Monitoring the eigenvalues of the monodromy matrix (i.e., Floquet multipliers) allows to ascertain the stability and bifurcations along the path of periodic solutions [38, 39]. In the next subsections, a full parametric analysis is carried out to investigate the sensitivity of the frequency-response with respect to the main design parameters. Linear and nonlinear negative stiffness. In Fig. 13, the displacement and acceleration FRCs are reported for the baseline isolation system (S1 system denoted by black lines), for the isolation system with the SMA damper (S2 system with ˜Ks = α, ˜xg = 0.05,ys = 0.2, denoted by red lines) and for the isolation system with the same SMA damper plus the neg- ative stiffness (S3 system with ˜Kn = (0.5,1,1.2)α) for two different excitation amplitudes, ˜Ag = (0.01,0.015). As expected, the addition of SMA wires induces an increase of stiffness, larger at low amplitudes, resulting in a shift of the curves to the right, and then an increase of accelerations and a decrease of displacements. With the introduction of negative stiffness, we observe a reverse shift of the curves to the left and then a decrease of accelerations and an increase of displacements. It is interesting to note that despite the stiffness being much lower than the baseline system, the maximum displacement is always smaller, at most equal, to that of the original system, thanks to the beneﬁcial effect of the augmented damping. Be- sides the negative stiffness coefﬁcient, also the nonlinear stiffness coefﬁcient ˜K3 plays an important role on the nonlinear dynamic response. In Fig. 14, the displacement and acceler- ation FRCs are reported for the baseline isolation system (black line) and for the S3 system endowed with different nonlinear stiffness coefﬁcients. A. Salvatore et al 12 (a) (b) (c) (d) Fig. 13 Frequency-response curves (FRCs) in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration of 0.01 ((a) and (b)) and 0.015 ((c) and (d)). 4 Nonlinear dynamic response scenario The response of S1 is denoted by black lines, the response of S2 (when Y = Zm,ys = 0.2) by red lines while the response of S3 by magenta ( ˜Kn = 0.5α), violet ( ˜Kn = α) and blue lines ( ˜Kn = 1.2α), respectively. Fig. 14 Frequency-response curves (FRCs) in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration set to 0.015. The response of S1 is represented by black lines and that of S3 (when Y = Zm,ys = 0.2, ˜Kn = α and ˜K3 = (1,2,4,8,10)10−6 ˜Kn) are denoted by solid violet lines with increasing thickness for increasing ˜K3. (a) (b) (c) (d) Fig. 13 Frequency-response curves (FRCs) in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration of 0.01 ((a) and (b)) and 0.015 ((c) and (d)). The response of S1 is denoted by black lines, the response of S2 (when Y = Zm,ys = 0.2) by red lines while the response of S3 by magenta ( ˜Kn = 0.5α), violet ( ˜Kn = α) and blue lines ( ˜Kn = 1.2α), respectively. (a) (b) (c) (d) Fig. 13 Frequency-response curves (FRCs) in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration of 0.01 ((a) and (b)) and 0.015 ((c) and (d)). The response of S1 is denoted by black lines, the response of S2 (when Y = Zm,ys = 0.2) by red lines while the response of S3 by magenta ( ˜Kn = 0.5α), violet ( ˜Kn = α) and blue lines ( ˜Kn = 1.2α), respectively. (a) (c) (d) Fig. 13 Frequency-response curves (FRCs) in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration of 0.01 ((a) and (b)) and 0.015 ((c) and (d)). The response of S1 is denoted by black lines, the response of S2 (when Y = Zm,ys = 0.2) by red lines while the response of S3 by magenta ( ˜Kn = 0.5α), violet ( ˜Kn = α) and blue lines ( ˜Kn = 1.2α), respectively. Fig. 14 Frequency-response curves (FRCs) in terms of nondimensional displacement (left) and acceleration Fig. 14 Frequency-response curves (FRCs) in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration set to 0.015. 4 Nonlinear dynamic response scenario The response of S1 is represented by black lines and that of S3 (when Y = Zm,ys = 0.2, ˜Kn = α and ˜K3 = (1,2,4,8,10)10−6 ˜Kn) are denoted by solid violet lines with increasing thickness for increasing ˜K3. 13 Nonlinear dynamic response of a NS-SMA isolation system Note that an increase of ˜K3, associated with a smaller working displacement ˜xf , entails a stronger hardening nonlinearity that leads to a reduction of the peak displacement and to an increase of the peak acceleration. SMA mechanical characteristics. In Fig. 15 the FRCs of the baseline isolation system (black lines) are compared with those of the S2 system (red lines) and with those of the S3 system (violet lines) for different levels of hysteresis ratio ys and for two excitation ampli- tudes, ˜Ag = (0.01,0.015). The acceleration of the S2-isolated system shows, for low exci- tations and regardless of the hysteresis ratio, an increase of acceleration compared to the baseline S1 system. On the contrary, the S3-isolated system exhibits for both excitation am- plitudes a strong reduction in accelerations while for medium and high hysteresis ratios of the SMA damper, it also undergoes a strong reduction in displacements. By introduc- ing more SMA wires (with Y = 1.6Zm), an additional reduction of displacement amplitude can be achieved but at the expense of a stiffness increase and, accordingly, of accelerations transmissibility. (a) (b) (c) (d) Fig. 15 Frequency-response curves in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration set to 0.01 ((a) and (b)) and 0.015 ((c) and (d)). The response of the S1-isolated system is described by black lines, those of S2 (when Y = Zm,ys = (0.2,0.5,0.8)) by red lines with increasing thickness for increasing ys, and those of S3 (when ˜Kn = α,Y = Zm,ys = (0.2,0.5,0.8)) by violet lines with increasing thickness for increasing ys and blue lines ( ˜Kn = α,Y = 1.6Zm,ys = 0.8), respectively. (a) (b) (c) (d) Fig. 15 Frequency-response curves in terms of nondimensional displacement (left) and acceleration (right) for a ground acceleration set to 0.01 ((a) and (b)) and 0.015 ((c) and (d)). 4 Nonlinear dynamic response scenario The response of the S1-isolated system is described by black lines, those of S2 (when Y = Zm,ys = (0.2,0.5,0.8)) by red lines with increasing thickness for increasing ys, and those of S3 (when ˜Kn = α,Y = Zm,ys = (0.2,0.5,0.8)) by violet lines with increasing thickness for increasing ys and blue lines ( ˜Kn = α,Y = 1.6Zm,ys = 0.8), respectively. 14 A. Salvatore et al 5 Displacement and acceleration transmissibilities The parametric study has disclosed a meaningful sensitivity of the frequency-response with respect to the system parameters. Henceforth, the evolution of the response for increasing base accelerations is discussed. The FRCs of the S3-isolated system, are computed for dif- ferent excitation amplitudes (see Fig. 16). The strong softening-hardening nonlinearity of the system is reﬂected by the trend of the system backbone curves. Moreover, other interest- ing phenomena are observed such as the emergence of detached resonance curves. For the case with large negative stiffness ( ˜Kn = 1.2α) there occurs a disappearance of the peak in the response within the displacement range in which the system turns out to be overdamped (0.2 < ˜x < 0.4). Moreover, for the strongest base acceleration in the low frequency range, there exists a bandwidth in which no stationary solutions could be obtained, a circumstance that suggested the existence of quasi-periodic/nonperiodic responses. Fig. 16 FRCs in terms of nondimensional displacement (left) and acceleration (right) for the S3 system with ˜Kn = α,Y = Zm,ys = 0.2 (top) and with ˜Kn = 1.2α,Y = Zm,ys = 0.2 (bottom) when the base accelerations are set to (0.8,1,1.08,1.2,1.28,1.3,1.32,1.36,1.4,1.48,1.6,1.8,2)10−2. Fig. 16 FRCs in terms of nondimensional displacement (left) and acceleration (right) for the S3 system with ˜Kn = α,Y = Zm,ys = 0.2 (top) and with ˜Kn = 1.2α,Y = Zm,ys = 0.2 (bottom) when the base accelerations are set to (0.8,1,1.08,1.2,1.28,1.3,1.32,1.36,1.4,1.48,1.6,1.8,2)10−2. For the same excitation amplitudes, the FRCs are computed for the baseline isolation sys- tem, for the S2-isolated system with various hysteresis ratios and yielding force levels, and for the S3-isolated system with various negative stiffness and hysteresis levels. The ratios Nonlinear dynamic response of a NS-SMA isolation system 15 between the peak response of the S2 or S3-isolated system and the response of the baseline system are reported in Fig. 17. (b) Xmax/XmaxIS (a) (d) Xmax/XmaxIS (c) Amax/AmaxIS Amax/AmaxIS Fig. 17 Ratio between the peak response of the controlled system and that of the baseline system in terms of nondimensional displacements (left) and accelerations (right). The responses of S2 with Y = Zm and ys = (0.2,0.5,0.8) are represented by red dotted, red dashed and red solid lines, respectively, while the response of S2 with Y = 1.6Zm and ys = 0.8 is described by bordeaux solid lines. 5 Displacement and acceleration transmissibilities The responses of S3 with ˜Kn = α (top), 1.2α (bottom), Y = Zm,ys = (0.2,0.5,0.8) are described by blue dotted, blue dashed and blue solid lines, respectively, while the response of S3 with ˜Kn = α (top), 1.2α (bottom),Y = 1.6Zm,ys = 0.8 is indicated by violet solid lines. Xmax/XmaxIS (a) (b) Xmax/XmaxIS (c) A /A IS (d) max max Fig. 17 Ratio between the peak response of the controlled system and that of the baseline system in terms of nondimensional displacements (left) and accelerations (right). The responses of S2 with Y = Zm and ys = (0.2,0.5,0.8) are represented by red dotted, red dashed and red solid lines, respectively, while the response of S2 with Y = 1.6Zm and ys = 0.8 is described by bordeaux solid lines. The responses of S3 with ˜Kn = α (top), 1.2α (bottom), Y = Zm,ys = (0.2,0.5,0.8) are described by blue dotted, blue dashed and blue solid lines, respectively, while the response of S3 with ˜Kn = α (top), 1.2α (bottom),Y = 1.6Zm,ys = 0.8 is indicated by violet solid lines. As expected, the insertion of the SMA damper alone causes an increase of accelerations response for weak excitations. The increase is about 30% when Y = Zm and 60% when Y = 1.6Zm, respectively. Hence, it depends mainly on the initial stiffness of the hysteretic damping and mildly on the damping ratio. On the other hand, the damping ratio strongly affects the response for moderate to strong base accelerations. It turns out that reductions of 20%, 40%, 60% are obtained when ys = (0.2,0.5,0.8), respectively. By introducing the negative stiffness in parallel with the SMA damper in the S3 system, the ampliﬁcation of accelerations for weak excitations is totally cancelled and a mild reduction can be observed. In addition, a further 20% reduction of accelerations compared with the S2-isolated sys- tem is obtained for moderate and strong base excitations, achieving an overall acceleration reduction of 40%, 60%, and 80% for ys = (0.2,0.5,0.8), respectively. The trend in accel- eration reduction coincides with the trend depicting stiffness reduction. Therefore, it shows a bell shape with a peak corresponding to the displacement where the stiffness reduction is maximum. Given the presence of a cubic stiffness in the NS mechanism, for large displace- As expected, the insertion of the SMA damper alone causes an increase of accelerations response for weak excitations. 5 Displacement and acceleration transmissibilities The increase is about 30% when Y = Zm and 60% when Y = 1.6Zm, respectively. Hence, it depends mainly on the initial stiffness of the hysteretic damping and mildly on the damping ratio. On the other hand, the damping ratio strongly affects the response for moderate to strong base accelerations. It turns out that reductions of 20%, 40%, 60% are obtained when ys = (0.2,0.5,0.8), respectively. By introducing the negative stiffness in parallel with the SMA damper in the S3 system, the ampliﬁcation of accelerations for weak excitations is totally cancelled and a mild reduction can be observed. In addition, a further 20% reduction of accelerations compared with the S2-isolated sys- tem is obtained for moderate and strong base excitations, achieving an overall acceleration reduction of 40%, 60%, and 80% for ys = (0.2,0.5,0.8), respectively. The trend in accel- eration reduction coincides with the trend depicting stiffness reduction. Therefore, it shows a bell shape with a peak corresponding to the displacement where the stiffness reduction is maximum. Given the presence of a cubic stiffness in the NS mechanism, for large displace- 16 A. Salvatore et al ments, the equivalent stiffness, hence the maximum acceleration, reaches again the stiffness of the baseline system. The effect of increasing the damping ratio is an expansion of the bell, both in terms of width and height. If the initial stiffness of the SMA damper overcomes the negative stiffness, as is the case with ˜Kn = α and Y = 1.6Zm (violet line in Fig. 17b), a slight increase of peak acceleration for low excitations occurs, together with an expansion of the bell width and of the effective displacement range. By balancing the increase of SMA damper’s initial stiffness with an equivalent increase of negative stiffness as is the case with ˜Kn = 1.2α and Y = 1.6Zm, (violet line in Fig. 17d), the increase of peak acceleration for low excitation amplitudes is cancelled again. For moderate and strong excitations, the increase of negative stiffness from α to 1.2α causes a slight increase of acceleration reduction of about 10%. Regarding the displacements, it is worth highlighting a substantial coincidence between the reduction offered by the SMA system and by the NS-SMA system with ˜Kn = α for weak base excitations. 5 Displacement and acceleration transmissibilities The trend in reduction is for both systems initially quasilinear up to a maximum value corresponding to the displacement that yields the maximum damp- ing. The maximum displacement reduction is quite similar for both systems except for low damping ratios (ys = 0.2) where the stiffness reduction is not balanced by a robust damping augmentation and it is of about 20% for the S3 system and of about 40% for the S2 sys- tem. For the remaining damping ratios, the displacement peak reduction for both systems is about 55% and 65% with ys = (0.2,0.5), respectively. Past the maximum, the trend of peak displacement reduction of the S3-isolated system deviates from the trend of the S2 system, showing a smaller reduction. The maximum deviation between the two responses occurs where the stiffness reduction is maximum. Fig. 18 FRCs in terms of nondimensional displacements for the base accelerations set to (0.8,1,1.08,1.2,1.28,1.3,1.32,1.36,1.4,1.48,1.6,1.8,2)10−2 for the S3 system with ˜Kn = α,Y = Zm,ys = (0.5,0.8) (parts (a) and (b)) and ˜Kn = α,Y = 1.6Zm,ys = 0.8 (part (c)), and with ˜Kn = 1.2α,Y = Zm,ys = (0.5,0.8) (parts (d) and (e)) and ˜Kn = α,Y = 1.6Zm,ys = 0.8 (part (f)). Fig. 18 FRCs in terms of nondimensional displacements for the base accelerations set to (0.8,1,1.08,1.2,1.28,1.3,1.32,1.36,1.4,1.48,1.6,1.8,2)10−2 for the S3 system with ˜Kn = α,Y = Zm,ys = (0.5,0.8) (parts (a) and (b)) and ˜Kn = α,Y = 1.6Zm,ys = 0.8 (part (c)), and with ˜Kn = 1.2α,Y = Zm,ys = (0.5,0.8) (parts (d) and (e)) and ˜Kn = α,Y = 1.6Zm,ys = 0.8 (part (f)). 17 Nonlinear dynamic response of a NS-SMA isolation system The increase of yielding force (Y), hence of the initial SMA stiffness, from Y = Zm to Y = 1.6Zm moves the curves to the right, showing a smaller reduction for weak excitations and a larger reduction for moderate and strong base excitations. For the case with ˜Kn = 1.2α, the response reduction is smaller than that achieved with ˜Kn = α, while for a certain range of base excitations, there exists an increase in response. This range of base excitations cor- responds to the FRCs where the displacement peak is due to the superharmonic resonance, as one can note in Fig. 18. Next we address the force transmissibility in terms of frequency bandwidth where effective isolation is attained. 5 Displacement and acceleration transmissibilities We consider force transmissibility as the ratio between absolute acceler- ation and base acceleration. As known, a response is considered effectively controlled when the transmissibility is lower than 1. For the nondimensional base acceleration of 0.02, the acceleration peak reduction for the S3-isolated system is minimum (i.e., 40%) and is equal to that of the S2-isolated system (see Fig. 17b and Fig. 17d). By analysing the force transmis- sibility under the same base accelerations, further useful considerations can be drawn about the isolation performance of the proposed system. While Fig. 17 shows only the reduction in acceleration peak, Fig. 19 shows the bandwidth of the isolated frequencies. 1stfis Kn= SMA Kn=1.2 1stfis 1stfis Fig. 19 FRCs in terms of force transmissibility for a nondimensional base acceleration equal to 0.02. The response of the S1-isolated system is described by the black solid line, while the responses of the S2 system (when ys = 0.8,Y = (1,1.6)Zm) are denoted by the red solid and red dashed lines, respectively. The responses of the S3-isolated system (when ˜Kn = α,Y = (1,1.6)Zm) are described by the violet solid and dashed lines and those for the S3 system (when ˜Kn = 1.2α,Y = (1,1.6)Zm) are described by the blue solid and blue dashed lines, respectively. In all curves ys = 0.8. 1stfis Kn= SMA Kn=1.2 1stfis 1stfis Fig. 19 FRCs in terms of force transmissibility for a nondimensional base acceleration equal to 0.02. The response of the S1-isolated system is described by the black solid line, while the responses of the S2 system (when ys = 0.8,Y = (1,1.6)Zm) are denoted by the red solid and red dashed lines, respectively. The responses of the S3-isolated system (when ˜Kn = α,Y = (1,1.6)Zm) are described by the violet solid and dashed lines and those for the S3 system (when ˜Kn = 1.2α,Y = (1,1.6)Zm) are described by the blue solid and blue dashed lines, respectively. In all curves ys = 0.8. The introduction of the SMA damper alone increases the value of the ﬁrst isolated fre- quency, thus reducing the bandwidth of isolated frequencies of 69% and 87% for Y = Zm and Y = 1.6Zm, respectively. 5 Displacement and acceleration transmissibilities On the other hand, the negative stiffness mechanism in parallel with the SMA damper, in addition to a reduction of the peak response, yields an increase of bandwidth of isolated frequencies, reducing the value of the ﬁrst isolated frequency of 25% and 44% with ˜Kn = α and ˜Kn = 1.2α, respectively. It is also possible to observe that an increase of the SMA yielding force (Y) entailing an increase of initial SMA stiffness not accompanied by an increase of negative stiffness, yields a further reduction in peak response but at the expense of a reduction of isolated frequency bandwidth. 18 A. Salvatore et al ̃ ̃ Δf ̃ = 0, = 0, = 0 ̃ = 0, = , = 0.8 ̃ = 0, = 1.6 , = 0.8 ̃ = , = , = 0.8 ̃ = , = 1.6 , = 0.8 ̃ = 1.2 , = , = 0.8 ̃ = 1.2 , = 1.6 , = 0.8 Fig. 20 Force transmissibility performance for the three kinds of isolated systems, S1, S2 and S3. Fig. 20 Force transmissibility performance for the three kinds of isolated systems, S1, S2 and S3. 6 Primary, superharmonic and detached resonances De- spite the very low frequency range where the isolas appear, the outer isolas are detrimental for isolation purposes since they can give rise to unwanted dynamic ampliﬁcation. On the contrary, the inner islands can be used to reduce the response near the main frequency. The factors that determine whether the mass will move along the detached solution curve or along the main branch are the initial conditions or the perturbations causing jumps between the coexisting attractors. merge with each other and with the 1:3 superharmonic resonance branch (see Fig. 21g). At a higher excitation amplitude, the primary resonance branch and the outer superisola merges and give rise to an inner isola for slightly higher excitation amplitudes (see Fig. 21h). De- spite the very low frequency range where the isolas appear, the outer isolas are detrimental for isolation purposes since they can give rise to unwanted dynamic ampliﬁcation. On the contrary, the inner islands can be used to reduce the response near the main frequency. The factors that determine whether the mass will move along the detached solution curve or along the main branch are the initial conditions or the perturbations causing jumps between the coexisting attractors. In order to obtain basins of attraction of the system near the isolas (see Fig. 22), the equa- tions of motion are numerically integrated for a harmonic base excitation over 1,000 periods over a grid of initial conditions. The initial conditions, in terms of displacement and veloc- ity, that lead to different attractors are denoted by different colors. In particular, the initial conditions that lead to the low-amplitude solution are represented in red, while in blue those that lead to the high-amplitude solution. It is possible to note that the system with ˜Kn = α shows much thinner basins of attraction for both the outer and inner isola than the system with ˜Kn = 1.2α, denoted by a major blue portion and a major red portion for the outer and inner isolas, respectively. Fig. 22 Basins of attraction for the system with ˜Kn = α,Y = Zm,ys = 0.2 (top) for a base acceleration ˜Ag = 0.0128 and frequency ˜Ω2 = 0.022, corresponding to the outer isola (left), and for a base acceleration ˜Ag = 0.0132 and frequency ˜Ω2 = 0.05, corresponding to the inner isola (right). 6 Primary, superharmonic and detached resonances The severe softening nonlinearity associated with the softening hysteresis induces an inter- action between the primary and the superharmonic resonances causing the emergence of detached resonance curves, a phenomenon that is well documented in the literature [40, 41, 42]. However, a new phenomenology is here documented. In Fig. 21 the evolution of the isolas for the system with two levels of negative stiffness is shown. The qualitative pattern in both cases consists in the birth of an outer isola in the neighborhood of the superharmonic resonance frequency which, for increasing base acceleration levels, coalesces ﬁrst with the superharmonic resonance branch and thereafter, upon further increase of excitation ampli- tude, with the main resonance branch, giving rise to an inner island. Fig. 21 Evolution of isolas topology for the S3 system with ˜Kn = α,Y = Zm,ys = 0.2 (top) for a nondimen- sional ground acceleration equal to 0.0128 (a), 0.01288 (b), 0.0130 (c), 0.0132 (d) and for the S3 system with ˜Kn = 1.2α,Y = Zm,ys = 0.2 (bottom) for a nondimensional ground acceleration equal to 0.01072 (e), 0.01088 (f), 0.01112 (g) and 0.0112 (h). Fig. 21 Evolution of isolas topology for the S3 system with ˜Kn = α,Y = Zm,ys = 0.2 (top) for a nondimen- sional ground acceleration equal to 0.0128 (a), 0.01288 (b), 0.0130 (c), 0.0132 (d) and for the S3 system with ˜Kn = 1.2α,Y = Zm,ys = 0.2 (bottom) for a nondimensional ground acceleration equal to 0.01072 (e), 0.01088 (f), 0.01112 (g) and 0.0112 (h). It is possible to observe that for the S3 system with ˜Kn = α, there exists only a small outer is- land, while for the case with ˜Kn = 1.2α, different outer islands coexist. Indeed, for a ground acceleration of 0.01, two outer detached resonance isolas are visible in the proximity of the superharmonic resonances of order 1:3 and 1:5 (see Fig. 21e). For a higher amplitude, the two distinct isolas merge and a new isola is formed near the superharmonic resonance of order 1:7 (see Fig. 21f). Upon further increasing the excitation amplitude, all previous isolas Nonlinear dynamic response of a NS-SMA isolation system 19 merge with each other and with the 1:3 superharmonic resonance branch (see Fig. 21g). At a higher excitation amplitude, the primary resonance branch and the outer superisola merges and give rise to an inner isola for slightly higher excitation amplitudes (see Fig. 21h). 6 Primary, superharmonic and detached resonances 24, it is possible to see that, for the solution belonging to the main branch, the peak of main harmonic corresponds to the zero of all other harmonics having thus a relative phase of π/2n with n representing the order of the harmonic component. By focusing on the third harmonic, a relative phase of π/6 means that the peak of the main harmonic coincides with the zero value with negative derivative of superharmonic and this is equivalent to the out-of-phase condition or deampliﬁcation condition. On the other hand, the solution along the detached curve shows that the peak of the main harmonic coincides with the minimum of the third harmonic, thus giving rise to a π/2 relative phase. It can be said that, for the solutions belonging to the detached curve, the higher harmonics are phased with the main harmonic, producing a less distorted and larger motion. Zm,ys = 0.2, the force-displacement cycles, the phase portraits, the time histories and the FFTs of the response to a harmonic base excitation with ˜Ag = 0.01072 and ˜Ω2 = 0.022 are shown in Fig. 23. The response belonging to the main solution branch is richer due to the presence of more superharmonic components. In fact, while in the ﬁrst case, third, ﬁfth and seventh harmonics exist, in the case of the isola only the third harmonic is manifested. From the FFT it is also possible to see that, for solutions belonging to the isola, the amplitude of the main harmonic is larger than that of the response, sum of all harmonics, while the solution along the main curve shows the fundamental harmonic with the same amplitude of the response. This suggests for the two solutions a different relative phase between the main harmonic and higher harmonics. Through a harmonic decomposition of the two different responses shown in Fig. 24, it is possible to see that, for the solution belonging to the main branch, the peak of main harmonic corresponds to the zero of all other harmonics having thus a relative phase of π/2n with n representing the order of the harmonic component. By focusing on the third harmonic, a relative phase of π/6 means that the peak of the main harmonic coincides with the zero value with negative derivative of superharmonic and this is equivalent to the out-of-phase condition or deampliﬁcation condition. 6 Primary, superharmonic and detached resonances (bottom) Basins of attraction for the system with ˜Kn = 1.2α,Y = Zm,ys = 0.2 for a base acceleration of ˜Ag = 0.01072 and ˜Ω2 = 0.022, corresponding to the outer isola (left), and for a base acceleration of ˜Ag = 0.0112 and ˜Ω2 = 0.05, correspond- ing to the inner isola (right). In red the initial conditions that lead to the low-amplitude solution, while in blue those that lead to the high-amplitude solution. Fig. 22 Basins of attraction for the system with ˜Kn = α,Y = Zm,ys = 0.2 (top) for a base acceleration ˜Ag = 0.0128 and frequency ˜Ω2 = 0.022, corresponding to the outer isola (left), and for a base acceleration ˜Ag = 0.0132 and frequency ˜Ω2 = 0.05, corresponding to the inner isola (right). (bottom) Basins of attraction for the system with ˜Kn = 1.2α,Y = Zm,ys = 0.2 for a base acceleration of ˜Ag = 0.01072 and ˜Ω2 = 0.022, corresponding to the outer isola (left), and for a base acceleration of ˜Ag = 0.0112 and ˜Ω2 = 0.05, correspond- ing to the inner isola (right). In red the initial conditions that lead to the low-amplitude solution, while in blue those that lead to the high-amplitude solution. For a more thorough characterization of the two coexisting attractors, one along the main curve and the other along the detached curve, exhibited by the system with ˜Kn = 1.2α,Y = 20 A. Salvatore et al Zm,ys = 0.2, the force-displacement cycles, the phase portraits, the time histories and the FFTs of the response to a harmonic base excitation with ˜Ag = 0.01072 and ˜Ω2 = 0.022 are shown in Fig. 23. The response belonging to the main solution branch is richer due to the presence of more superharmonic components. In fact, while in the ﬁrst case, third, ﬁfth and seventh harmonics exist, in the case of the isola only the third harmonic is manifested. From the FFT it is also possible to see that, for solutions belonging to the isola, the amplitude of the main harmonic is larger than that of the response, sum of all harmonics, while the solution along the main curve shows the fundamental harmonic with the same amplitude of the response. This suggests for the two solutions a different relative phase between the main harmonic and higher harmonics. Through a harmonic decomposition of the two different responses shown in Fig. 6 Primary, superharmonic and detached resonances On the other hand, the solution along the detached curve shows that the peak of the main harmonic coincides with the minimum of the third harmonic, thus giving rise to a π/2 relative phase. It can be said that, for the solutions belonging to the detached curve, the higher harmonics are phased with the main harmonic, producing a less distorted and larger motion. Fig. 23 Force-displacement cycles (a), phase portraits (b), time histories (c) and FFTs (d) of the S3 system (with ˜Kn = 1.2α,Y = Zm,ys = 0.2) for ˜Ag = 0.01072 and ˜Ω2 = 0.022. In red the response to a zero initial condition along to the main solutions branch, while in blue the solution for the initial conditions ˜x = 0.2, ˜v = 0, giving rising to the isola solution curve. Fig. 23 Force-displacement cycles (a), phase portraits (b), time histories (c) and FFTs (d) of the S3 system (with ˜Kn = 1.2α,Y = Zm,ys = 0.2) for ˜Ag = 0.01072 and ˜Ω2 = 0.022. In red the response to a zero initial condition along to the main solutions branch, while in blue the solution for the initial conditions ˜x = 0.2, ˜v = 0, giving rising to the isola solution curve. Nonlinear dynamic response of a NS-SMA isolation system 21 Fig. 24 Harmonic decomposition of the superharmonic response along the main resonance branch (left) and detached resonance (right). The ﬁrst, third, ﬁfth and seventh harmonics and the total response are represented by red, blue, violet, magenta and black lines, respectively. Fig. 24 Harmonic decomposition of the superharmonic response along the main resonance branch (left) and detached resonance (right). The ﬁrst, third, ﬁfth and seventh harmonics and the total response are represented by red, blue, violet, magenta and black lines, respectively. Figure 25 shows the amplitudes and the phases of the overall response, of the main har- monic and of the ﬁrst superharmonic and the relative phase between the main harmonic and the ﬁrst superharmonic in the frequency domain. Note that the solutions belonging to the de- tached resonance curve show a phase equal to π/2, condition shared only with the solution of the main resonance frequency. Fig. 25 Nondimensional displacement (left) and phase angle (right) vs. nondimensional frequency for the S3 system with ˜Kn = 1.2α,Y = Zm,ys = 0.2 and ˜Ag = 0.01072. 6 Primary, superharmonic and detached resonances Black lines show the amplitude and phase angle of the total response, red lines and blue lines represent the amplitude and phase angle of the main harmonic and of the ﬁrst superharmonic of order 1:3, respectively. The phase angle between the main harmonic and the ﬁrst superharmonic is reported in gray lines. Fig. 25 Nondimensional displacement (left) and phase angle (right) vs. nondimensional frequency for the S3 system with ˜Kn = 1.2α,Y = Zm,ys = 0.2 and ˜Ag = 0.01072. Black lines show the amplitude and phase angle of the total response, red lines and blue lines represent the amplitude and phase angle of the main harmonic and of the ﬁrst superharmonic of order 1:3, respectively. The phase angle between the main harmonic and the ﬁrst superharmonic is reported in gray lines. 6.1 Bifurcation scenarios and quasi-periodicity As mentioned above, for low frequencies and high base accelerations, the periodic response of the S3 system with ˜Kn = 1.2α undergoes a loss of stability. By restricting our analysis to 22 A. Salvatore et al the frequency range reported in Fig. 26, a rich sequence of bifurcations is found. Moving from low to high frequencies, the ﬁrst encountered bifurcation is a Neimarck-Sacker or sec- ondary Hopf bifurcation (A), signalled by a pair of Floquet multipliers crossing the unit cir- cle away from the real axis (red circles). The solution emerging out of the Neimarck-Sacker bifurcation is a quasi-periodic solution. However, past the bifurcation the continuation of the unstable periodic solution cannot be successfully achieved. On the other hand, between C and D a stable branch of periodic solutions is found, which loses its stability at C due to a symmetry-breaking bifurcation. The two branches of mirror nonsimmetric periodic at- tractors lose their stability at B due to a period doubling bifurcation, circumstance signalled by the fact that one of the Floquet multipliers crosses the unit circle through -1. In D the solution experiences a fold bifurcation, whereby one of the Floquet multipliers crosses the unit circle along the positive real axis. Finally, in E a new Neimarck-Sacker bifurcation is manifested and afterwards path following of the stationary solutions breaks down. Fig. 26 FRCs for the S3 system with ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148 (left) and imaginary parts vs. real parts of Floquet multipliers (right). Fig. 26 FRCs for the S3 system with ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148 (left) and imaginary parts vs. real parts of Floquet multipliers (right). To investigate more in depth the scenario between the Neimarck-Sacker bifurcation in A and the period-doubling in B, bifurcation diagrams were constructed by direct numerical integration of the equations of motion (see Fig. 27). The time step was ﬁxed by dividing the excitation period into 4096 equally spaced points. The integration was carried out over 2,000 cycles of the excitation by considering as initial conditions those obtained at the pre- vious excitation frequency and the last 64 points of the Poincar´e map were recorded. 6.1 Bifurcation scenarios and quasi-periodicity It is possible to note that in A the solution becomes quasiperiodic, through the mentioned sec- ondary Hopf bifurcation, while from B to the left the solution becomes quasi-periodic by means of a more complex sequence of bifurcations. Figure 28 shows a symmetry breaking bifurcation at the frequency denoted by 2, singled out by the birth of an even superharmonic component, after the limit cycle exhibited at 1. This symmetry breaking paves the way to two distinct branches of nonsymmetric mirror solutions, whose orbits include the limit cycle and towards this they expand. In this frequency range the solution turns out to jump from one nonsymmetric branch to the other for any small frequency variation. Each of these branches undergoes a cascade of successive period-doubling bifurcation induced by the birth of a sub- harmonic component of order 2:1 at 3 and of subharmonics of order 4:1 and 2:1 at 4. Nonlinear dynamic response of a NS-SMA isolation system 23 Fig. 27 Bifurcation diagram for the system with ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148. Fig. 27 Bifurcation diagram for the system with ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148. This cascade of period-doubling bifurcations, giving rise to a nidiﬁcation of subharmonics and superharmonics, may lead to a choatic attractor. When the orbits of the nonsymmet- ric solutions touch the orbit of the limit cycle there is a reverse symmetry breaking and the solution regains symmetry, signalled by the disappearance of the even superharmonic component (6). In the frequency range between 2.04 and 2.09, the ratio between the modu- lation frequency and the carrier frequency locks into a rational number, due to the so-called frequency-locking phenomenon with three-period motions, supported by the 3:1 subhar- monic (7). Past this window, a new symmetry breaking is experienced by the 3-T solution, yielding two distinct branches of 3-T solutions in which, in addition to the 3:1 subharmonic, there appears an even superharmonic of order 1:2 (8). The symmetry breaking forces the solution to transition towards a quasiperiodic symmetric solution regime (9). Finally, the amplitude of the phase plane portion covered by the trajectory progressively gets reduced until 10, where a reverse secondary Hopf bifurcation makes the solution stable. (1) (2) (b) (a) (c) (d) Fig. 6.1.1 Bifurcation scenarios for the tri-stable conﬁguration For ˜Kn = 1.2α, the bifurcation scenario engages only ultra-low frequencies and a small displacements range. On the other hand, when ˜Kn > 1.2α, the bifurcation scenario involves much larger ranges of frequencies and displacements. As shown, for ˜Kn > 1.2α the system is tri-stable and the presence of the two lateral attractors breaks the symmetry of the response. Because of the existence of symmetry breaking and period-doubling bifurcations, the re- sponse is quasi-periodic for most of the frequencies within the considered range. In order to obtain the FRCs of the tri-stable conﬁguration (i.e., ˜Kn = 1.4α), the equations of motion are numerically integrated for a harmonic base excitation over 1,000 periods and the maximum amplitudes exhibited in the last 50 cycles, together with the Poincar´e sections, are recorded for each frequency within the range (see Fig. 31). Depending on the initial conditions, the mass can vibrate around the origin, the left or the right equilibrium positions. Regardless of the equilibrium around which the mass vibrates, the adjacent attractor breaks the symmetry of the response and leads to quasi-periodicity. Further, when the mass vibrates around one of the lateral equilibria, due to the stronger effects of the cubic stiffness, the acceleration transmissibility is higher compared with that exhibited by the mass vibrating around the ori- gin. For limited frequency intervals, the phase-locking phenomenon is observable together with a reduction of transmissibility with respect to the adjacent quasi-periodic response. By analysing more in depth the response in Fig. 31, we can note that for low frequencies (0 < ˜Ω2 < 0.022), the system has sufﬁcient energy to complete symmetric periodic cycles. For higher frequencies (0.022 < ˜Ω2 < 0.042), because of a folding bifurcation, two dif- ferent responses are exhibited by the system depending on the initial conditions. The high amplitude response, such as the response of the previous frequency range, exhibits stable cycles. On the other hand, the low amplitude solution, with a lower associated energy, suf- fers the attraction of the lateral equilibria and shows a quasi-periodic behavior. Increasing the frequency up to ˜Ω2 = 0.042, a downward jump occurs together with the birth of two mirrored asymmetric solutions, each of which experiences cascades of period-doubling bi- furcations. In the frequency range between 0.118 < ˜Ω2 < 0.18, the existing solutions regain For ˜Kn = 1.2α, the bifurcation scenario engages only ultra-low frequencies and a small displacements range. 6.1 Bifurcation scenarios and quasi-periodicity 28 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.2α,Y = Zm,ys = 0.5, the nondimensional ground acceleration is set to ˜Ag = 0.0148 for the frequencies corresponding to 1, 2. (d) (1) (a) (b) (c) (2) Fig. 28 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.2α,Y = Zm,ys = 0.5, the nondimensional ground acceleration is set to ˜Ag = 0.0148 for the frequencies corresponding to 1, 2. 24 A. Salvatore et al (6) (7) (8) (4) (5) (3) (b) (a) (c) (d) Fig. 29 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148 for the frequencies referred to as 3, 4, 5, 6, 7, 8 in Fig. 27. (4) (3) (a) (d) (b) (c) (4) (5) (6) (7) (8) Fig. 29 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148 for the frequencies referred to as 3, 4, 5, 6, 7, 8 in Fig. 27. Nonlinear dynamic response of a NS-SMA isolation system 25 (9) (10) (a) (9) (10) (b) (a) (c) (d) Fig. 30 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148 for the frequencies referred to as 9 and 10 in Fig. 27. (10) Fig. 30 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.2α,Y = Zm,ys = 0.5 for a nondimensional ground acceleration equal to ˜Ag = 0.0148 for the frequencies referred to as 9 and 10 in Fig. 27. 6.1.1 Bifurcation scenarios for the tri-stable conﬁguration On the other hand, when ˜Kn > 1.2α, the bifurcation scenario involves much larger ranges of frequencies and displacements. As shown, for ˜Kn > 1.2α the system is tri-stable and the presence of the two lateral attractors breaks the symmetry of the response. Because of the existence of symmetry breaking and period-doubling bifurcations, the re- sponse is quasi-periodic for most of the frequencies within the considered range. In order to obtain the FRCs of the tri-stable conﬁguration (i.e., ˜Kn = 1.4α), the equations of motion are numerically integrated for a harmonic base excitation over 1,000 periods and the maximum amplitudes exhibited in the last 50 cycles, together with the Poincar´e sections, are recorded for each frequency within the range (see Fig. 31). Depending on the initial conditions, the mass can vibrate around the origin, the left or the right equilibrium positions. Regardless of the equilibrium around which the mass vibrates, the adjacent attractor breaks the symmetry of the response and leads to quasi-periodicity. Further, when the mass vibrates around one of the lateral equilibria, due to the stronger effects of the cubic stiffness, the acceleration transmissibility is higher compared with that exhibited by the mass vibrating around the ori- gin. For limited frequency intervals, the phase-locking phenomenon is observable together with a reduction of transmissibility with respect to the adjacent quasi-periodic response. By analysing more in depth the response in Fig. 31, we can note that for low frequencies (0 < ˜Ω2 < 0.022), the system has sufﬁcient energy to complete symmetric periodic cycles. For higher frequencies (0.022 < ˜Ω2 < 0.042), because of a folding bifurcation, two dif- ferent responses are exhibited by the system depending on the initial conditions. The high amplitude response, such as the response of the previous frequency range, exhibits stable cycles. On the other hand, the low amplitude solution, with a lower associated energy, suf- fers the attraction of the lateral equilibria and shows a quasi-periodic behavior. Increasing the frequency up to ˜Ω2 = 0.042, a downward jump occurs together with the birth of two mirrored asymmetric solutions, each of which experiences cascades of period-doubling bi- furcations. In the frequency range between 0.118 < ˜Ω2 < 0.18, the existing solutions regain 26 A. Salvatore et al periodicity and two new mirrored asymmetric and quasi-periodic solutions appear along the lateral right and left equilibrium. 6.1.1 Bifurcation scenarios for the tri-stable conﬁguration Finally, when ˜Ω2 = 0.165, the ﬁrst two mirrored solutions coalesce into one symmetric periodic solution while the two lateral asymmetric solutions regain stability. In this latter frequency range, an ampliﬁcation of both accelerations and displacements is observable. Fig. 31 FRCs in terms of nondimensional displacements (a) and accelerations (b) and bifurcation diagram (c) for the S3 system with ˜Kn = 1.4α,Y = Zm,ys = 0.2 for a ground acceleration equal to ˜Ag = 0.0142. Blue lines represent the responses obtained for the forward frequency sweep while red lines denote those obtained in reverse sweep. Magenta and cyan lines indicate the responses of the system with initial conditions ˜x0 = 0.5 and of ˜x0 = −0.5, respectively. Finally, for comparative purposes, the responses of the mono-stable system with ˜Kn = 1.2α are represented by gray lines. Fig. 31 FRCs in terms of nondimensional displacements (a) and accelerations (b) and bifurcation diagram (c) for the S3 system with ˜Kn = 1.4α,Y = Zm,ys = 0.2 for a ground acceleration equal to ˜Ag = 0.0142. Blue lines represent the responses obtained for the forward frequency sweep while red lines denote those obtained in reverse sweep. Magenta and cyan lines indicate the responses of the system with initial conditions ˜x0 = 0.5 and of ˜x0 = −0.5, respectively. Finally, for comparative purposes, the responses of the mono-stable system with ˜Kn = 1.2α are represented by gray lines. In Fig. 32 the FFTs, hysteretic loops, phase portraits and Poincar`e maps associated with the frequencies highlighted in Fig. 31 are reported. By focusing on the last section (6) the coexistence of four different types of response for the same excitation frequency is noted. The basins of attraction associated to this frequency are numerically obtained and reported in Fig. 33. Nonlinear dynamic response of a NS-SMA isolation system 27 Fig. 32 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.4α,Y = Zm,ys = 0.2 and the nondimensional ground acceleration is set to ˜Ag = 0.0142 for the frequencies referred to as 0, 1, 2, 3, 4, 5, 6 in Fig. 31. Fig. 6.1.1 Bifurcation scenarios for the tri-stable conﬁguration 32 FFTs of the response (a), force-displacement cycles (b), phase portraits (c) and Poincar`e map (d) of the system when ˜Kn = 1.4α,Y = Zm,ys = 0.2 and the nondimensional ground acceleration is set to ˜Ag = 0.0142 for the frequencies referred to as 0, 1, 2, 3, 4, 5, 6 in Fig. 31. 28 A. Salvatore et al The coexistence of four different attractors, namely the left stable (LS) and unstable (LU) and the right stable (RS) and unstable (RU) equilibria, results in a high sensitivity of the dynamic response to the initial conditions, as manifested by the richness of the basins. From the progressive zooms of the basins of attraction (see Fig. 33c and Fig. 33d) it is notable that the strict adherence to the attractors succession order (LS, LU, RU, RS) leads to a fractal-like pattern. Fig. 33 (a) Basins of attraction for the system with ˜Kn = 1.4α,Y = Zm,ys = 0.2 for a base acceleration ˜Ag = 0.0142 and frequency ˜Ω2 = 0.125. Parts (b), (c) and (d) are zoomed-in regions bounded by the dashed rectangles. Magenta and red dots denote the initial conditions that lead to the left stable (LS) and unstable (LU) attractors, respectively, while, cyan and blue dots represent the initial conditions that lead to the right stable (RS) and unstable (RU) attractors, respectively. Fig. 33 (a) Basins of attraction for the system with ˜Kn = 1.4α,Y = Zm,ys = 0.2 for a base acceleration ˜Ag = 0.0142 and frequency ˜Ω2 = 0.125. Parts (b), (c) and (d) are zoomed-in regions bounded by the dashed rectangles. Magenta and red dots denote the initial conditions that lead to the left stable (LS) and unstable (LU) attractors, respectively, while, cyan and blue dots represent the initial conditions that lead to the right stable (RS) and unstable (RU) attractors, respectively. 7 Conclusions A novel vibration isolation system featuring a negative stiffness mechanism and shape mem- ory alloy nonlinear damping arranged in parallel with classical elastomeric hysteretic iso- lation devices is investigated parametrically for different levels of negative stiffness, SMA hysteretic damping ratio and yielding force. y p g y g The introduction of SMA alone, without negative stiffness, leads to a detrimental increase of the initial stiffness and, hence, to an increase of accelerations for low base excitations. On the Nonlinear dynamic response of a NS-SMA isolation system 29 other hand, by accurately tuning the negative stiffness with the SMA damping, a remarkable reduction of displacement and acceleration amplitudes can be achieved, while preserving a self-recentering capability without incurring an increase of acceleration transmissibility for low base excitations. To ensure an effective acceleration transmissibility reduction and, at the same time, the mono-stability and self-recentering capability of the isolated system, the optimum negative stiffness coefﬁcient ˜Kn must be bounded within the range α < ˜Kn < 1.2α, where α is the ratio between the post-elastic and the initial device stiffness. Regarding the SMA rheological element, the initial stiffness must be equal to the negative stiffness in order to keep the isolation system stiffness unaltered at the origin and, at the same time, develop a sufﬁciently high hysteretic damping. By considering ˜Kn = α, the optimum SMA element yielding force was found to be ˜Y = α ˜xg, where ˜xg is the ratio between the gap displacement and the maximum allowable displacement. Moreover, a high hysteresis ratio ys was shown to entail a better isolation performance. The study of the nonlinear dynamic response and its bifurcations has unfolded very rich bifurcation scenarios with detached resonances and unusual interactions between the pri- mary resonance and superharmonic resonances, or between superharmonic resonances of various orders, featuring multiplicity of coexisiting attractors, secondary Hopf bifurcations leading to quasi-periodicity, synchronization, symmetry breaking and period-doubling cas- cades. The detached resonance curves and bifurcations were numerically explored for high levels of negative stiffness and their nonlinear impact on isolation performance was dis- cussed. In particular, when ˜Kn = α the peak of the detached resonance curve was found to be lower than the peak of the main resonance curve, thus not affecting the isolation per- formance. Declarations Acknowledgments This work was partially supported by the PRIN Grant n. 2017L7X3CS −002 (3D PRINT- ING: A BRIDGE TO THE FUTURE. Computational methods, innovative applications, ex- perimental validations of new materials and technologies) and by the 2019 Sapienza Grant n. RG11916B8160BCCC (Vibration mitigation via advanced engineered devices and mate- rials) which are gratefully acknowledged. 7 Conclusions Instead, for a higher negative stiffness value (i.e., ˜Kn = 1.2α), the peak of the detached resonance curve was found to be larger than the primary resonance peak, hence, notably affecting the isolation performance. Moreover, the interesting quasi-periodicity of the response of the tri-stable conﬁguration obtained for higher levels of negative stiffness (i.e., ˜Kn = 1.4α), and the subsequent dynamic ampliﬁcation, were illustrated extensively. The obtained results pave the way towards a streamlined design process which aims to op- timize the isolation performance of the proposed negative stiffness-SMA device in terms of transmissibility and dynamic stability. Authors contributions All authors contributed to the study conception and design. Material preparation, data col- lection and analysis were performed by Andrea Salvatore. The ﬁrst draft of the manuscript was written by Andrea Salvatore and all authors commented on previous versions of the manuscript. All authors read and approved the ﬁnal manuscript. References 1. Pasala DTR, Sarlis AA, Nagarajaiah S, et al, Negative stiffness device for seismic response control of multistory buildings, Structures Congress, 2012 y g , g , 2. Taylor D, Nagarajaiah S, Reinhorn AM, et al, Adaptive negative stiffness: A new structural modiﬁcation approach for seismic protection, Journal of Structural Engineering (139), 2013 3. Attary N, Symans M, Nagarajaiah S, Development of a rotation-based negative stiffness device for seismic protection of structures, Journal of Vibration and Control (23), 853-867, 2017 rotection of structures, Journal of Vibration and Control (23), 853- 4. Iemura H, Application of pseudo-negative stiffness control to the benchmark bridge, Journal of Structural Control (10-3), 187-203, 2003 5. Iemura H, Kouchiyama O, Toyooka A, Shimoda I, Development of the Friction-Based Passive Negative Stiffness Damper and Its Veriﬁcation Tests Using Shaking Table, World Conference on Earthquake Engi- neering, 2008 g 6. Iemura H, Pradono MH, Advances in the development of pseudo-negative-stiffness dampers for seismic response control, Structural Control and Health Monitoring (16), 784-799, 2010 7. Pradono MH, Iemura H, Passively controlled MR damper in the benchmark structural control problem for seismically excited highway bridge, Structural Control and Health Monitoring (16), 626-638, 2010 8. Tan X, Wang B, Chen S, et al, A novel cylindrical negative stiffness structure for shock isolation, Com- posite Structures (214), 397-405, 2019 9. Winterﬂood J, Blair DG, Slagmolen B, High performance vibration isolation using springs in Euler col- umn buckling mode, Physics Letters A (300), 122-130, 2002 g y 10. Platus DL. Negative stiffness mechanism vibration isolation systems, Proceedings Volume 1619, Vibra- tion Control in Microelectronics, Optics, and Metrology, 1992 10. Platus DL. Negative stiffness mechanism vibration isolation systems, tion Control in Microelectronics, Optics, and Metrology, 1992 10. Platus DL. Negative stiffness mechanism vibration isolation tion Control in Microelectronics, Optics, and Metrology, 1992 p gy 11. Alhan C, Gavin HP, Aldemir U, Optimal control: Basis for performance comparison of passive and semiactive isolation systems, Journal of Engineering Mechanics (132), 2006 11. Alhan C, Gavin HP, Aldemir U, Optimal control: Basis f emiactive isolation systems, Journal of Engineering Mechanics (1 12. Carrella A, Brennan MJ, Waters TP, Static analysis of a passive vibration isolator with quasi-zero- stiffness characteristic, Journal of Sound and Vibration (301), 678-689, 2007 13. Conﬂict of interest The authors have no conﬂicts of interest to declare that are relevant to the content of this article. 30 A. Salvatore et al References Kovacic I, Brennan MJ, Waters TP, A study of a nonlinear vibration isolator with a quasi-zero stiffness characteristic, Journal of Sound and Vibration (315), 700-711, 2008 characteristic, Journal of Sound and Vibration (315), 700-711, 200 14. Zhou N, Liu K, A tunable high-static-low-dynamic stiffness vibration isolator, Journal of Sound and Vibration (329), 1254-1273, 2010 15. Le TD, Ahn KK, A vibration isolation system in low frequency excitation region using negative stiffness structure for vehicle seat, Journal of Sound and Vibration (330), 6311-6335, 2011 16. Carrella A, Brennan MJ, Waters TP, Lopes V, Force and displacement transmissibility of a nonlinear isolator with high-static-low-dynamic-stiffness, International Journal of Mechanical Sciences (55), 22-29, 2012 17. Yang J, Xiong YP, Xing JT, Dynamics and power ﬂow behaviour of a nonlinear vibration isolation system with a negative stiffness mechanism, Journal of Sound and Vibration (332), 167-183, 2013 18. Huang X, Liu X, Sun J, et al, Vibration isolation characteristics of a nonlinear isolator using euler buckled beam as negative stiffness corrector: A theoretical and experimental study, Journal of Sound and Vibration (333), 1132-1148, 2014 19. Hao Z, Cao Q, The isolation characteristics of an archetypal dynamical model with stable-quasi-zero- stiffness, Journal of Sound and Vibration (340), 61-79, 2015 20. Zheng Y, Zhang X, Luo Y, et al, Design and experiment of a high-static-low-dynamic stiffness isolator using a negative stiffness magnetic spring, Journal of Sound and Vibration (360), 31-52, 2016 21. Lu ZQ, Brennan M, Ding H, Chen LQ, High-static-low-dynamic-stiffness vibration isolation enhanced by damping nonlinearity, Science China Technological Sciences (62), 1103-1110, 2018 inearity, Science China Technological Sciences (62), 1103-1110, 2 22. Cimellaro GP, Domaneschi M, Warn G, Three-Dimensional Base Isolation Using Vertical Negative Stiff- ness Devices, Journal of Earthquake Engineering (24), 2004-2032, 2018 23. Wang X, Liu H, Chen Y, Gao P, Beneﬁcial stiffness design of a high-static-low-dynamic-stiffness vi- bration isolator based on static and dynamic analysis, International Journal of Mechanical Sciences (142), 235-244, 2018 Nonlinear dynamic response of a NS-SMA isolation system 31 24. Zhou Y, Chen P, Mosqueda G, Analytical and Numerical Investigation of Quasi-Zero Stiffness Vertical Isolation System, Journal of Engineering Mechanics (145), 2019 25. Sun M, Song G, Li Y, Huang Z, Effect of negative stiffness mechanism in a vibration isolator with asymmetric and high-static-low-dynamic stiffness, Mechanical Systems and Signal Processing (124), 388- 407, 2019 26. References Shi X, Zhu S, A comparative study of vibration isolation performance using negative stiffness and inerter dampers, Journal of the Franklin Institute (356), 7922-7946, 2019 27. Yan B, Ma H, Jian B, Wang K, Wu C, Nonlinear dynamics analysis of a bi-state nonlinear vibration isolator with symmetric permanent magnets, Nonlinear Dynamics (97), 2499-2519, 2019 y p g y 28. Bian J, Jing X, Analysis and design of a novel and compact X-structured vibration isolation mount (X- Mount) with wider quasi-zero-stiffness range, Nonlinear Dynamics (101), 2195-2222, 2020 29. Gao X, Teng HD, Dynamics and isolation properties for a pneumatic near-zero frequency vibration isolator with nonlinear stiffness and damping, Nonlinear Dynamics (102), 2205-2227, 2020 30. Liu C, Yu K, Accurate modeling and analysis of a typical nonlinear vibration isolator with quasi-zero stiffness, Nonlinear Dynamics (100), 2141-2165, 2020 31. Sonfack Bouna H, Nana Nbendjo BR, Woafo P, Isolation performance of a quasi-zero stiffness isolator in vibration isolation of a multi-span continuous beam bridge under pier base vibrating excitation, Nonlinear Dynamics (100), 1125-1141, 2020 y 32. Wang K, Zhou J, Chang Y, Ouyang H, Xu D, A nonlinear ultra-low-frequency vibration isolator with dual quasi-zero-stiffness mechanism, Nonlinear Dynamics (101), 755-773, 2020 33. Donmez A, Cigeroglu E, Ozgen OG, An improved quasi-zero stiffness vibration isolation system utiliz- ing dry friction damping, Nonlinear Dynamics (101), 107-121, 2020 ing dry friction damping, Nonlinear Dynamics (101), 107-121, 2020 34 Li M Zh P Li H N l S lf C i N i S iff D B d C bi i f Sh ing dry friction damping, Nonlinear Dynamics (101), 107-121, 20 34. Liu M, Zhou P, Li H, Novel Self-Centering Negative Stiffness Damper Based on Combination of Shape M All d P d S i J l f A E i i (31) 2018 34. Liu M, Zhou P, Li H, Novel Self-Centering Negative Stiffness Damper Based on Combination of Shape Memory Alloy and Prepressed Springs, Journal of Aerospace Engineering (31), 2018 Liu M, Zhou P, Li H, Novel Self-Centering Negative Stiffness Da oy and Prepressed Springs, Journal of Aerospace Engineering (31) 35. Bouc R, Forced vibration of mechanical systems with hysteresis, Materials Science 1967 36. Wen Y, Method for random vibration of hysteretic Systems, Journal of the Engineering Mechanics Divi- sion (102), 249-263, 1976 37. Charalampakis AE, Tsiatas GC, A Simple Rate-Independent Uniaxial Shape Memory Alloy (SMA) Model, Frontiers in Built Environment (4), 2018 38. References Carboni B, Lacarbonara W, Dynamic response of nonlinear oscillators with hysteresis, Proceedings of the ASME 2015 IDETC-CIE conference (6), 2015 39. Lacarbonara W, Vestroni F, and Capecchi D, Poincar`e map-based continuation of periodic orbits in dy- namic discontinuous and hysteretic systems, 17th Biennial ASME Conference on Mechanical Vibration and Noise, 1999 40. Habib G, Cirillo GI, Kerschenc G, Uncovering detached resonance curves in single-degree-of-freedom systems, Procedia Engineering (199), 649-656, 2017 y g g 41. Gatti G, Brennan MJ, Kovacic I, On the interaction of the responses at the resonance frequencies of a nonlinear two degrees-of-freedom system, Physica D: Nonlinear Phenomena (239), 591-599, 2016 42. Formica G, Lacarbonara W, Asymptotic dynamic modeling and response of hysteretic nanostructured beams, Nonlinear Dynamics (99), 227-248, 2020
https://openalex.org/W2612639955	https://hal.science/hal-01538152/file/Perspective24_Goebel_FR.pdf	French	null	Insect pest control in agriculture. Changing scale: from field to landscape	Perspective	2,013	cc-by	3,051	To cite this version: François-Régis Goebel. Lutte contre les insectes bioagresseurs en agriculture. Changer d’échelle : de la parcelle au paysage. Perspective, 2013, N° 24, 4 p. 10.18167/agritrop/00029. hal-01538152 Distributed under a Creative Commons Attribution 4.0 International License 24 Changer d’échelle : de la parcelle au paysage François-Régis Goebel Depuis une vingtaine d’années, la pression des insectes bioagresseurs sur l’agriculture augmente. Cette pression croissante s’explique par l’extension des monocultures et par l’intensification des pratiques culturales, qui modifient les paysages et réduisent la biodiversité. Elle est accentuée par le changement climatique, qui favorise les migrations des insectes tropicaux vers les zones tempérées et modifie la biologie des insectes. Lutter contre cette pression croissante, tout en réduisant ou en arrêtant l’usage des pesticides, suppose d’agir non plus seulement à l’échelle de la parcelle mais aussi à celle du paysage. Ce changement d’échelle permet de tirer parti de la biodiversité pour réguler les bioagresseurs, et aussi de coordonner les pratiques des acteurs, comme le montre la lutte contre les bioagresseurs de la canne à sucre et du cotonnier. Il suppose toutefois de s’appuyer sur une connaissance fine des interactions entre d’une part les populations de bioagresseurs et leurs auxiliaires, et d’autre part les composantes du paysage, la biodiversité et les activités humaines, ce qui ouvre de nouveaux champs de recherche transdisciplinaire. 24 Lutte contre les insectes bioagresseurs en agriculture Changer d’échelle : de la parcelle au paysage > 24 Lutte contre les insectes bioagresseurs en agriculture > HAL Id: hal-01538152 https://hal.science/hal-01538152v1 Submitted on 19 Jun 2017 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 4.0 International License Une pression croissante En Afrique par exemple, où la culture du cotonnier fait depuis longtemps l’objet de traitements insecticides, des chenilles dépré- datrices de la capsule sont devenues résistantes aux pyréthrinoïdes, et des insectes piqueurs- suceurs (pucerons, aleurodes) aux organophos- phorés. Aux États-Unis, en Inde et en Chine, ces chenilles sont aussi devenues résistantes aux toxines produites par des variétés génétiquement modifiées pour les combattre, et des ravageurs considérés comme secondaires sont devenus préoccupants. L’impact des intrants est d’autant plus fort que les préconisations ne sont pas toujours respectées. Par exemple, le surdosage de fertilisation azotée, fréquent pour augmenter les rendements des cultures de riz, maïs, canne à sucre…, accroît les infestations. pour faciliter la coupe et livrer aux usines des cannes débarrassées des pailles ; or il détruit des auxiliaires utiles et n’élimine pas les lépidoptères foreurs, qui se développent à l’intérieur des tiges. De même, le brûlage des tiges et des rameaux du cotonnier, encore fréquent en Afrique, ne détruit pas les bioagresseurs, qui se réfugient dans les débris végétaux tombés au sol. À ces pratiques néfastes s’ajoute le changement climatique, qui favorise les migrations des zones tropicales vers les zones tempérées, ce qui accroît la pression des bioagresseurs dans les pays de destination. Ainsi, le nombre d’insectes tropicaux ayant migré vers l’Europe ne cesse de croître – 1 500 espèces y sont arrivées depuis une ving- taine d’années – : Paysandisia archon, papillon originaire d’Argentine, qui dévaste les palmiers du sud de la France ; Cacyreus marshalli, papillon originaire d’Afrique du Sud, qui ravage les géra- niums ; frelons et coccinelles asiatiques, invasifs, qui supplantent les espèces européennes… De plus, le changement climatique modifie les cycles biologiques des insectes dans leur pays d’origine, ce qui provoque des pullulations plus fréquentes et plus virulentes. Comment, dans ce contexte, lutter contre les insectes bioagresseurs ? Pendant longtemps, l’utilisation du tout-insecticide à l’échelle de la parcelle a été privilégiée. Si les insecticides sont maintenant remis en question en raison des effets nocifs sur la santé humaine et sur l’envi- ronnement, la parcelle ou l’exploitation demeure l’échelle à laquelle sont conduites la plupart des actions. Or, dans la perspective d’une lutte agroécologique, cette échelle n’est plus suffisante. En effet, les insectes sont mobiles : ils naissent dans un habitat, puis en colonisent d’autres ; leur habitat ne se limite donc pas à la parcelle ou à l’exploitation, mais s’étend hors de l’espace agricole. > L’échelle du paysage permet de tirer parti de la biodiversité et de coordonner les acteurs. > Les monocultures et l’intensification des pratiques culturales augmentent la pression des bioagresseurs. Une pression croissante agricoles pour l’alimentation et l’énergie. Ces pratiques modifient ou uniformisent les paysages, réduisent la biodiversité et nuisent à certains services écosystémiques, comme la pollinisation par les abeilles. Des insectes bioagresseurs et leurs ennemis naturels – les auxiliaires (préda- teurs, parasitoïdes…) – voient leurs biotopes détruits et migrent vers d’autres biotopes, voire colonisent de nouveaux territoires. Les équilibres biologiques sont perturbés. Avec Perspective, le Cirad propose un espace d’expression de nouvelles pistes de réflexion et d’action, fondées sur des travaux de recherche et sur l’expertise, sans pour autant présenter une position institutionnelle. D D epuis une vingtaine d’années, la pres- sion des insectes bioagresseurs sur l’agriculture augmente. Cette pression croissante s’explique notamment par le défri- chement et le déboisement de vastes étendues pour y installer des monocultures industrielles et ainsi répondre à la demande en produits La pression croissante tient aussi à l’intensi- fication des pratiques culturales. Les grandes entreprises, aidées parfois par des institutions de recherche et développement, préconisent des paquets techniques : application systématique de pesticides (insecticides, herbicides, fongi- cides…) ; utilisation d’engrais et de variétés améliorées, génétiquement modifiées ou non. Ces préconisations altèrent le fonctionnement des écosystèmes. Des résistances aux pesticides apparaissent, et certains bioagresseurs ne sont plus maîtrisés. En Afrique par exemple, où la culture du cotonnier fait depuis longtemps l’objet de traitements insecticides, des chenilles dépré- datrices de la capsule sont devenues résistantes aux pyréthrinoïdes, et des insectes piqueurs- suceurs (pucerons, aleurodes) aux organophos- phorés. Aux États-Unis, en Inde et en Chine, ces chenilles sont aussi devenues résistantes aux toxines produites par des variétés génétiquement modifiées pour les combattre, et des ravageurs considérés comme secondaires sont devenus préoccupants. L’impact des intrants est d’autant plus fort que les préconisations ne sont pas toujours respectées. Par exemple, le surdosage de fertilisation azotée, fréquent pour augmenter les rendements des cultures de riz, maïs, canne à sucre…, accroît les infestations. La pression croissante tient aussi à l’intensi- fication des pratiques culturales. Les grandes entreprises, aidées parfois par des institutions de recherche et développement, préconisent des paquets techniques : application systématique de pesticides (insecticides, herbicides, fongi- cides…) ; utilisation d’engrais et de variétés améliorées, génétiquement modifiées ou non. Ces préconisations altèrent le fonctionnement des écosystèmes. Des résistances aux pesticides apparaissent, et certains bioagresseurs ne sont plus maîtrisés. La biodiversité nécessaire en associant industriels et services publics de l’agriculture et de l’environnement. Ils échan- gent des informations sur l’écologie et les dégâts du ravageur, et coordonnent leurs pratiques. Ils ont par exemple décidé de traiter en priorité les champs proches des zones à risques pour éviter la multiplication des foyers d’infestation. Cette démarche collective a bénéficié à l’ensemble des agriculteurs, qui ont économisé des traitements insecticides, et a réduit les risques de pollution des cours d’eau qui se déversent vers la barrière de corail toute proche. Elle a aussi conduit à une réflexion sur les espèces végétales à planter ou à éviter aux abords des plantations de canne à sucre pour diminuer les risques d’infestation : palmiers et ficus ont été ainsi remplacés par des arbres moins attractifs, comme le manguier. Il est donc nécessaire de concevoir et de pratiquer la lutte contre les bioagresseurs à une échelle plus large que la parcelle : le paysage, c’est-à-dire une étendue spatiale, naturelle ou transformée par l’homme, qui présente une identité visuelle ou fonctionnelle. C’est à cette échelle qu’il est pos- sible de tirer parti de la biodiversité présente dans les composantes du paysage : végétation naturelle, plantes cultivées, arbres, corridors forestiers, pâtu- rages, cours d’eau, etc. C’est aussi à cette échelle qu’il est possible de coordonner les actions des agriculteurs et autres acteurs (services de l’État, sociétés de développement…), voire de construire les paysages pour utiliser au mieux les plantes de service, qui attirent ou repoussent les bioagres- seurs des cultures et leurs ennemis naturels. En Afrique du Sud, le foreur de tige Eldana saccharina a envahi les monocultures de canne à sucre dès l’introduction de cette culture au siècle dernier. Seule l’utilisation d’insecticides et de variétés résistantes parvenait à en limi- ter les dégâts. La lutte biologique était difficile, car les parasitoïdes du foreur de tige ne l’ont pas « suivi » lorsqu’il a colonisé la canne sucre à partir de Cyperus papyrus (Cyperacae) et de graminées sauvages. Des recherches ont identifié les plantes de service à utiliser ou à introduire en bordure des champs pour stimuler la régulation naturelle du foreur des tiges : des plantes sauvages comme Cyperus, Erianthus, Pennisetum ou Desmodium, et des plantes cultivées comme le maïs ou le sorgho, qui attirent les parasitoïdes, piègent ou repoussent les bioagresseurs (Fig.1). Des recherches conduites en Australie et en Afrique du Sud (lire encadré p. > Agroécologique, la protection des cultures sera assurée par une équipe transdisciplinaire. > Construire les paysages pour utiliser au mieux les plantes de service. Une pression croissante De plus, la parcelle n’offre qu’une bio- diversité limitée ne permettant ni de tirer parti du pouvoir attractif ou répulsif des plantes sur tel ou tel insecte, ni de stimuler la régulation des populations de bioagresseurs par les auxiliaires. Des agriculteurs de Java (Indonésie) ne s’y sont pas trompés : ils conservent plantes et arbustes naturels (graminées, malvacées, euphorbiacées, figuiers…) autour de leurs parcelles (riz, maïs, canne à sucre, cultures maraîchères, tournesol), car cette végétation naturelle « abrite des insectes variés ». Enfin, l’échelle de la parcelle est insuf- fisante parce qu’elle n’implique que l’agriculteur concerné, alors qu’une lutte efficace suppose une coordination des acteurs à différentes échelles. De plus, l’utilisation systématique d’insecticides a érodé, voire a fait disparaître des savoirs locaux parfois ancestraux, qui avaient pourtant montré leur efficacité. En Afrique de l’Ouest, les agricul- teurs écimaient manuellement les cotonniers qui, en réaction, envoyaient des messages chimiques pour empêcher la ponte des femelles d’Helico- verpa armigera, de Diparopsis watersi et d’Earias, chenilles déprédatrices de la capsule. Si l’écimage pourrait être remis en pratique à la suite de tests probants au Mali, d’autres savoirs ne sont plus mobilisés alors qu’ils pourraient être utiles. C’est le cas de l’épaillage des tiges de canne à sucre pratiqué autrefois à Java (Indonésie) : en ôtant les feuilles bien avant la récolte, il supprime les points d’insertion par lesquels les foreurs pénè- trent dans les tiges. D’autres pratiques culturales augmentent la pression des insectes, comme l’arrachage des plantes naturelles ou le brûlage. L’arrachage des plantes naturelles aux abords des parcelles prive les auxiliaires de nourriture, ce qui crée un désé- quilibre nuisant à la régulation naturelle des populations de bioagresseurs. Quant au brûlage, il est encore pratiqué sur les parcelles de canne à sucre, comme au Soudan ou en Afrique du Sud, La biodiversité nécessaire 4) montrent la pertinence de cette échelle. En Australie, le han- neton ravageur de la canne à sucre, Dermolepida albohirtum, provoque des dégâts dans les parcelles en bordure des forêts abritant des figuiers, des palmiers, des eucalyptus et des acacias. En effet, après s’être développé sur la canne à sucre au stade larvaire, le hanneton adulte vit et se reproduit dans ces arbres, avant d’infester à nouveau les parcelles. Le hanneton ne migrant pas au-delà de 200 m de son habitat d’origine, il suffit d’in- tervenir sur les parcelles à cette distance de la forêt. Intéressés par ce résultat, des agriculteurs du Queensland, qui agissaient chacun sur leurs parcelles, ont décidé de se réunir régulièrement, Plantes de service Habitat naturel divers (réservoirs à parasitoïdes) La gestion des bioagresseurs s’appuie sur la connaissance des interactions au sein de l’agro-écosystème entre d’une part les insectes et leurs ennemis naturels (parasitoïdes, pathogènes, prédateurs…), et d’autre part les plantes hôtes et la végétation naturelle qui abritent les insectes. Mouvement des parasitoïdes Mouvement des bioagresseurs Mouvement des parasitoïdes Plantes de service Mouvement des bioagresseurs Figure 1. Tenir compte des composantes du paysage et des plantes de service pour le contrôle biologique d’Eldana saccharina, ravageur de la canne à sucre en Afrique du Sud. Figure 1. Tenir compte des composantes du paysage et des plantes de service pour le contrôle biologique d’Eldana saccharina, ravageur de la canne à sucre en Afrique du Sud. La gestion des bioagresseurs s’appuie sur la connaissance des interactions au sein de l’agro-écosystème entre d’une part les insectes et leurs ennemis naturels (parasitoïdes, pathogènes, prédateurs…), et d’autre part les plantes hôtes et la végétation naturelle qui abritent les insectes. Source : adapté de Conlong D., Rutherford S., 2010. South African Sugar Research Institute. Source : adapté de Conlong D., Rutherford S., 2010. South African Sugar Research Institute. Habitat naturel divers (réservoirs à parasitoïdes) Habitat naturel divers (réservoirs à parasitoïdes) Quelques mots sur… Quelques mots sur… agressées, ou en précisant les services rendus par les plantes, il sera possible d’identifier la végétation à planter dans et autour des par- celles cultivées pour attirer ou repousser les insectes bioagresseurs ou leurs ennemis naturels. Elles portent aussi sur les processus écologiques au sein des agrosystèmes et au-delà, afin de repenser les pratiques culturales et leur inten- sification, de caractériser et de valoriser les services écosystémiques, et aussi d’utiliser les savoirs locaux. La biodiversité nécessaire Elles portent enfin sur les stra- tégies des acteurs, en vue de la coordination nécessaire pour une protection efficace des cultures. Comme en témoignent ces recherches, mettre au point des systèmes minimisant les infes- tations et favorisant les ennemis naturels des bioagresseurs suppose de prendre en compte les interactions des insectes avec le paysage dans toutes ses composantes : la végétation naturelle, sa localisation, ses caractéristiques ; les parcelles, leur taille et les plantes qui y sont cultivées ; les friches ; les corridors ; et bien sûr les stratégies des agriculteurs et autres acteurs impliqués. Cela suppose donc de les connaître. François-Régis Goebel, est docteur en entomologie appliquée (université Paul Sabatier, Toulouse). Il est spécialisé en protection des plantes et protection intégrée. Au Cirad depuis 1988, il a conduit l’essentiel de ses recherches sur la canne à sucre, en poste à la Réunion, en Afrique du Sud et en Australie, et lors de missions en Indonésie et dans plusieurs pays d’Afrique. Il est actuellement basé à Montpellier dans l’unité de recherche Systèmes de culture annuels, dont il est le directeur adjoint (http://ur-sca.cirad.fr/). francois-regis.goebel @cirad.fr Ces recherches ont donné lieu à des publications parmi lesquelles : 42, rue Scheffer 75116 Paris . France Goebel F.-R., Sallam N., Samson P., Chandler K., 2010. Quantifying spatial movement of the http://www.fdgdon974.fr/IMG/pdf/SCI_may- jun2010_goebel.pdf (accès libre) persp ctive e Directeur de la publication : Patrick Caron, directeur général délégué à la recherche et à la stratégie Coordination : Corinne Cohen, délégation à l’information scientifique et technique Conception graphique/réalisation : Patricia Doucet, délégation à la communication Diffusion : Christiane Jacquet, délégation à la communication Courriel : perspective@cirad.fr www.cirad.fr/publications-ressources/ edition/perspective-policy-brief la transdisciplinarité Agroécologique, la protection des cultures ne sera plus assurée par l’entomologiste ou par l’agronome, mais par une équipe transdis- ciplinaire. Une telle approche est complexe et suppose d’établir des ponts entre les disciplines : entomologie – biologie des insectes, écologie spatiale, écologie des communautés d’insectes, écologie chimique – ; botanique ; agronomie ; sciences humaines et sociales ; modélisation, information spatiale (SIG, télédétection), informatique, etc. Elle suppose aussi d’associer les acteurs dès la conception des recherches. < Des recherches doivent être conduites dans les pays tropicaux, d’où sont originaires de nom- breux ravageurs, sur la biologie des insectes et sur leurs mécanismes de régulation, afin de lutter plus efficacement dans les pays d’origine comme dans les pays de destination. Ces recherches portent tout d’abord sur les insectes et les plantes. Par exemple, en iden- tifiant les messages chimiques (composés vola- tils) libérés par les plantes lorsqu’elles sont greyback cane beetle in sugarcane landscape: data available and research needs. Proceedings of the Aus- tralian Society of Sugar Cane Technologists 32: 71-83 http://www.cabdirect.org/abstracts/20103206832. html Ce Perspective est le fruit de recherches et de réflexions, conduites en équipe et en par- tenariat dans plusieurs pays et régions : l’Afrique du Sud, avec le Sasri (South Afri- can Sugar Research Institute) et l’université du KwaZulu-Natal ; l’Australie, avec le Sugar Research Australia (SRA) Limited – ex-BSES Limited – et l’université du Queensland asso- ciée au Csiro (Commonwealth Scientific and Industrial Research Organisation) ; la Réunion, avec eRCane et la Fédération départementale des groupements de défense contre les orga- nismes nuisibles (FDGDON) ; l’Indonésie, avec l’Isri (Indonesian Sugar Research Institute) ; l’Afrique de l’Ouest. Goebel F.-R., Sallam N., 2011. New pest threats for sugarcane in the new bioeconomy and how to manage them. Current Opinion in Environmental Sustainability 3: 81-89. http://www.sciencedirect.com/science/article/ pii/S1877343510001442 Goebel F.-R., Tabone E., Do Thi Khanh H., Roux E., Marquier M., Frandon J., 2010. Biocontrol of Chilo sacchariphagus (Lepidoptera: crambidae) a key pest of sugarcane: lessons from the past and future prospects. Sugar Cane International 28: 128-132. Organiser e e e t . l r Hunter M.D., 2002. Landscape structure, habitat fragmen- tation, and the ecology of insects. Agricultural and Forest Entomology 4: 159–166. en savoir plus Hunter M.D., 2002. Landscape structure, habitat fragmen- tation, and the ecology of insects. Agricultural and Forest Entomology 4: 159–166. Hunter M.D., 2002. Landscape structure, habitat fragmen- tation, and the ecology of insects. Agricultural and Forest Entomology 4: 159–166. Tscharntke T., Brandl R., 2004. Plant-insect interactions in fragmented landscapes. Annual Review of Entomology 49, 405–430. Bianchi F.J., Booij C.J. & Tscharntke T., 2006. Sustainable Pest Regulation in agricultural landscapes: a review on landscape composition, biodiversity and natural pest control. Proceedings of the Royal Society of Biology 273: 1715-1727. Thies C., Steffan-Dewenter I., Tscharntke T., 2003. Effect of landscape context on herbivory and parasitism at diffe- rent spatial scales. Oïkos 101: 18–25. Thies C., Steffan-Dewenter I., Tscharntke T., 2003. Effect of landscape context on herbivory and parasitism at diffe- rent spatial scales. Oïkos 101: 18–25.
https://openalex.org/W2089311139	https://ria.ua.pt/bitstream/10773/20070/2/Organic-Inorganic%20Eu3%2bTb3%2b%20codoped%20hybrid%20films%20for%20temperature%20mapping%20in%20integrated%20circuits_10.3389fchem.2013.00009.pdf	English	null	Organic–Inorganic Eu3+/Tb3+ codoped hybrid films for temperature mapping in integrated circuits	Frontiers in chemistry	2,013	cc-by	5,487	INTRODUCTION The synthesis and characterization of the Eu3+/Tb3+ co-doped di-ureasil organic/inorganic hybrids has already been described in detail elsewhere (Brites et al., 2010, 2013). The ﬁrst step of the synthesis involves the formation in tetrahydrofuran of an urea cross-linked hybrid precursor (De Zea Bermudez et al., 1999). In the second step, the [Eu(btfa)3(MeOH)(bpeta)] and [Tb(btfa)3(MeOH)(bpeta)] complexes [where btfa−is 4,4,4-triﬂuoro-l-phenyl–1,3-butanedionate, bpeta is 1,2–bis(4– pyridyl)ethane and MeOH methanol] were incorporated as ethanolic solutions together with water and HCl to promote the hydrolysis of the urea cross-linked hybrid precursor. The di-ureasil thermometer, hereafter named UET–1.3, incorporates the Eu3+ and Tb3+ β-diketonate complexes in a 1:3 mass pro- portion, respectively, and is processed as a ﬁlm or a monolith. Although the ﬁlms can be obtained with high thickness control by dip- or spin-coating techniques, here a ∼10 μm thermosen- sitive UET–1.3 layer was deposited over a FR4 printed circuit board (Figure 1A). UET–1.3 monoliths were also employed for photoluminescence characterization and temperature calibration. Miniaturization, integration and the increase of physical com- plexity in electronic devices and circuits tend to generate higher local power densities and localized heating problems (Burzo et al., 2005; Christofferson et al., 2008). The management of the heat ﬂux generated by the several billion of transistors that actually exist in a single chip is one of the main challenges of the modern electronics industry. Thermal management is therefore essen- tial to improves electronics performance and reliability, posing a limitation stronger than the downscaling itself. The temperature distribution across integrated circuits must be, then, accurately mapped with superior spatial resolution (Burzo et al., 2005; Yarimaga et al., 2011; Liu et al., 2012). The well-known limitations of contact thermometers has strengthened the development of non-contact thermometry tech- niques (Brites et al., 2012), such as, infrared (IR) thermography (Meola and Carlomagno, 2004), thermoreﬂectance (Kolodner and Tyson, 1983; Christofferson et al., 2008), optical interferome- try (Kersey and Berkoff, 1992), Raman spectroscopy (Frazão et al., 2009) and luminescence (Aigouy et al., 2005; Brites et al., 2010; Vetrone et al., 2010; Kuzmin et al., 2011; Yarimaga et al., 2011; Benayas et al., 2012). Luminescence methods combine temper- ature sensitivities up to 5.0%·K−1 with spatial resolution below 1 μm and have been used to monitor and map temperature on integrated circuits and electronic devices (Aigouy et al., 2005; Jung et al., 2011; Yarimaga et al., 2011; Brites et al., 2013). ORIGINAL RESEARCH ARTICLE bli h d 08 J l 2013 published: 08 July 2013 doi: 10.3389/fchem.2013.00009 Edited by: Ning Yan, National University of Singapore, Singapore Ning Yan, National University of Singapore, Singapore Ning Yan, National University of Singapore, Singapore Reviewed by: Chaoxian Xiao, Iowa State University, USA Yaping Du, Xi’an Jiaotong University, China Ning Yan, National University of Singapore, Singapore Reviewed by: Chaoxian Xiao, Iowa State University, USA Yaping Du, Xi’an Jiaotong University, China Correspondence: Luís D. Carlos, Departamento de Física, Universidade de Aveiro, Campus de Santiago 3810-193, Aveiro, Portugal e-mail: lcarlos@ua.pt Reviewed by: Chaoxian Xiao, Iowa State University, USA Yaping Du, Xi’an Jiaotong University, China Correspondence: Luís D. Carlos, Departamento de Física, Universidade de Aveiro, Campus de Santiago 3810-193, Aveiro, Portugal e-mail: lcarlos@ua.pt Keywords: organic–inorganic hybrids, lanthanide ions, molecular thermometer, spatio-temporal resolution Carlos D. S. Brites 1, Patrícia P. Lima 1, Nuno J. O. Silva 1, Angel Millán 2, Vitor S. Amaral 1, Fernando Palacio2 and Luís D. Carlos 1* 1 Departamento de Física and CICECO, Universidade de Aveiro, Aveiro, Portugal 2 Departamento de Fisica de la Materia Condensada, Facultad de Ciencias and Instituto de Ciencia de Materiales de Aragón, CSIC–Universidad de Zaragoza, Zaragoza Spain 1 Departamento de Física and CICECO, Universidade de Aveiro, Aveiro, Portugal 2 Departamento de Fisica de la Materia Condensada, Facultad de Ciencias and Instituto de Ciencia de Materiales de Aragón, C Zaragoza Spain Zaragoza, Spain The continuous decrease on the geometric size of electronic devices and integrated circuits generates higher local power densities and localized heating problems that cannot be characterized by conventional thermographic techniques. Here, a self-referencing intensity-based molecular thermometer involving a di-ureasil organic-inorganic hybrid thin ﬁlm co-doped with Eu3+ and Tb3+ tris (β-diketonate) chelates is used to obtain the temperature map of a FR4 printed wiring board with spatio-temporal resolutions of 0.42 μm/4.8 ms. Organic–Inorganic Eu3+/Tb3+ codoped hybrid ﬁlms for temperature mapping in integrated circuits Carlos D. S. Brites 1, Patrícia P. Lima 1, Nuno J. O. Silva 1, Angel Millán 2, Vitor S. Amaral 1, Fernando Palacio2 and Luís D. Carlos 1* ORIGINAL RESEARCH ARTICLE bli h d 08 J l 2013 RESULTS Given the 1/200 detection limit of the detector used in the experiments we can anticipate an ultimate temperature detection limit of δTmin = 0.01 K. Nevertheless, in the experimental conditions used, a temperature uncertainly δT = 0.15 K was estimated from the full-width-at-half-maximum value of the temperature reads Gaussian distribution in the absence of any external heat source (Figure 3B). where IEu and ITb stands for the integrated areas of the 5D0→7F2 and 5D4→7F5 transitions, assigned to Eu3+ and Tb3+, FIGURE 1 \| (A) Geometry of the FR4 printed wiring board where A and B characterize the directions used in temperature mappings. (B) Schematic representation of the setup used. The heating circuit (H) was covered with a UET–1.3 ﬁlm (F) and positioned over a translation stage (T). The temperature was controlled by the source (S). A handheld UV lamp (L) was used to excite the ﬁlm. The optical ﬁber (O) was connected to the portable detector (D) and the emission spectra were processed in the computer (C). A key parameter to evaluate the performance of a luminescent thermometer is the relative sensitivity (S), deﬁned as the relative variation on the thermometric parameter (Q) with temperature: A key parameter to evaluate the performance of a luminescent thermometer is the relative sensitivity (S), deﬁned as the relative variation on the thermometric parameter (Q) with temperature: S = dQ/dT Q . (1) (1) Using a ﬁxed heating current, bi-dimensional temperature proﬁles of the FR4 printed wiring board along the directions A and B in Figure 1A were reconstructed from the emission spectra of UET-1.3 using the 200 and the 450 μm core diam- eter ﬁbers and different scanning steps (ranging from 800 to 50 μm), Figure 4. The spatial resolution of the thermometer was calculated by Equation 2 using the temperature proﬁles along the A direction in Figure 1A (Figure 5A). The results are com- pared in Figure 6. The calculated spatial resolution is consid- erably improved from 10 to 0.5 μm when the scanning step is decreased from 800 μm to values around the ﬁber inner radius (200 and 100 μm, for the 450 and 200 μm ﬁbers, respectively). For lower scanning step values the spatial resolution remains almost constant. RESULTS The emission spectra (Figure 2A), as well as the 5D4 and 5D0 life- time values, are temperature dependent in the 10–330 K range (Brites et al., 2010). This dependence and the room-temperature emission quantum yield value (0.16 ± 0.02 at 365 nm) permit to anticipate that the UET–1.3 thermometer sensitivity in the 290–330 K temperature range is enough to implement a sensor based on the analysis of the emission spectra using commer- cially cost-effective excitation sources and detectors (e.g., hand- held UV lamp and a portable optical-ﬁber connected detector, Figure 1B). The temperature of the UET–1.3 ﬁlm can be accessed measur- ing the emission spectra and using the thermometric parameter , deﬁned as: = I2 Eu −I2 Tb, (4) (4) where IEu and ITb stands for the integrated areas of the 5D0→7F2 and 5D4→7F5 transitions, assigned to Eu3+ and Tb3+, respectively (Figure 2B). Other deﬁnitions for the thermomet- ric parameter are also plausible (namely the ITb/IEu ratiometric form) without losing the generality of the method. Here we use, however, the previously reported thermometric parameter (Brites et al., 2010, 2011, 2013). The UET–1.3 local cali- bration curve (Figure 3A) for the temperature range 290–330 K was computed by three consecutive temperature cycles. A sec- ond order polynomial ﬁt to the experimental values allows the conversion of intensities into temperature. Given the 1/200 detection limit of the detector used in the experiments we can anticipate an ultimate temperature detection limit of δTmin = 0.01 K. Nevertheless, in the experimental conditions used, a temperature uncertainly δT = 0.15 K was estimated from the full-width-at-half-maximum value of the temperature reads Gaussian distribution in the absence of any external heat source (Figure 3B). where IEu and ITb stands for the integrated areas of the 5D0→7F2 and 5D4→7F5 transitions, assigned to Eu3+ and Tb3+, respectively (Figure 2B). Other deﬁnitions for the thermomet- ric parameter are also plausible (namely the ITb/IEu ratiometric form) without losing the generality of the method. Here we use, however, the previously reported thermometric parameter (Brites et al., 2010, 2011, 2013). The UET–1.3 local cali- bration curve (Figure 3A) for the temperature range 290–330 K was computed by three consecutive temperature cycles. A sec- ond order polynomial ﬁt to the experimental values allows the conversion of intensities into temperature. INTRODUCTION The size (250–1000 μm) and geometry of the copper tracks etched over the FR4 printed wiring board determine the temper- ature distribution proﬁle that is controlled by the current feeding the board. The UET–1.3 ﬁlm was excited with a handheld UV lamp (Spectroline E-Series, Aldrich, Model Z169625, operating at 365 ± 25 nm) and the emission was collected with optical ﬁbers of 200 and 450 μm inner diameter. The emission spectra were then analyzed in a portable spectrometer (Ocean Optics, USB-4000 FL) with limit of detection of 1/200. The ﬁber was positioned over the circuit that was moved with a nanomax 3-Axis ﬂexure trans- lation stage from ThorLabs®, with variable steps (ranging from 800 to 50 μm). Here we report the use of a self-referencing intensity- based molecular thermometer involving a di-ureasil organic- inorganic hybrid thin ﬁlm co-doped with Eu3+ and Tb3+ tris (β-diketonate) chelates to map the temperature over a FR4 printed wiring board using commercial detectors and excitation sources. July 2013 \| Volume 1 \| Article 9 \| 1 www.frontiersin.org Hybrid ﬁlms for temperature maping Brites et al. FIGURE 1 \| (A) Geometry of the FR4 printed wiring board where A and B characterize the directions used in temperature mappings. (B) Schematic representation of the setup used. The heating circuit (H) was covered with a UET–1.3 ﬁlm (F) and positioned over a translation stage (T). The temperature was controlled by the source (S). A handheld UV lamp (L) was used to excite the ﬁlm. The optical ﬁber (O) was connected to the portable detector (D) and the emission spectra were processed in the computer (C). RESULTS The higher spatial resolution measured with the 200 μm core diameter ﬁber is 0.42 μm, a value 4.5 times lower than the Rayleigh limit of diffraction (1.89 μm) in the experimental conditions used (see discussion below). Despite the minor changes of the spatial resolution when the scanning step decreases from 200 to 50 μm (Figure 6), the transition from the low to high temperature regions in the proﬁle along the A direction in Figure 1A is clearly much more deﬁned for 50 μm (Figure 4B). The relative sensitivity was proposed as a ﬁgure of merit to com- pare the performance of distinct thermometers (Brites et al., 2012). When the temperature is accessed in different points it is possible to deﬁne the spatial resolution of the measurement (δx) as the minimum distance between points that present tem- perature higher than the temperature uncertainly (δT). It can be estimated using the common expression (Kim et al., 2012): δx = δT ⃗∇T max (2) (2) where ⃗∇T max = dT/dx is the maximum temperature gradient of the mapping. The experimental setup used deﬁnes the temper- ature detection limit and the temperature gradient. The temporal resolution of the measurement (δt) is the minimum time interval between measurements presenting temperature higher than δT: where ⃗∇T max = dT/dx is the maximum temperature gradient of the mapping. The experimental setup used deﬁnes the temper- ature detection limit and the temperature gradient. The temporal resolution of the measurement (δt) is the minimum time interval between measurements presenting temperature higher than δT: δt = δT dT/dt (3) (3) (3) where dT/dt is the temperature change per unit of time. July 2013 \| Volume 1 \| Article 9 \| 2 Frontiers in Chemistry \| Inorganic Chemistry Hybrid ﬁlms for temperature maping Brites et al. FIGURE 2 \| (A) Temperature dependence of the emission spectra of UET–1.3 (excited at 365 nm) in the 10–330K temperature range. The f-f lines corresponding to the 5D4→7F6, 5 (Tb3+) and 5D0→7F2−4 (Eu3+) transitions are identiﬁed. In the area marked with an asterisk there is a superposition between the emission of Eu3+ (5D0→7F0, 1) and Tb3+ (5D4→7F4). (B) Normalized integrated intensity of 5D0→7F2(red) and 5D4→7F5(green) in the temperature range 290–330K. FIGURE 3 \| (A) Local calibration curve of UET–1.3 for the temperature range 290–330K. RESULTS The distribution of temperature (vertical bars) was ﬁtted to a Gaussian (solid line) resulting in a maximum at 295.4K and a full- width-at-half-maximum of 0.15K. FIGURE 3 \| (A) Local calibration curve of UET–1.3 for the temperature range 290–330K. The temperature was cycled three times and the emission spectra recorded at equal time intervals when the temperature increases. For all emission spectra the computed parameter was divided by its value at 303 K (0) to get a thermometric parameter near the unit. The results for all cycles are overlapped and the second For the determination of the temporal resolution limit of the mapping, δt, the heating current was turned on and the ther- malisation of the copper track followed using the UET–1.3 ther- mometer readout (Figure 5B). The temporal resolution achieved, 4.8 ms, is of the same order of magnitude than the detector integration time set (e.g., 10 ms). (tta−stands for thenoyltriﬂuoroacetonate). A 0.01 K tempera- ture resolution is expected based on shot noise of the collected light; however experimental conditions (e.g., electric ﬂuctua- tions of detection system) degrade this limit to the 0.1–1.0 K range. The spatial resolution is limited by the CCD smoothing to 15 μm (Kolodner and Tyson, 1983). Since 2005 there is a sig- niﬁcant number of references reporting the temperature mapping of active electronic devices with high spatial resolution. RESULTS The temperature was cycled three times and the emission spectra recorded at equal time intervals when the temperature increases. For all emission spectra the computed parameter was divided by its value at 303 K (0) to get a thermometric parameter near the unit. The results for all cycles are overlapped and the second degree polynomial ﬁt presented in the Figure (r2 > 0.995) is the local calibration curve. (B) Relative frequency of the UET–1.3 temperature read during 60 s in the absence of current in the heating circuit (inset). The distribution of temperature (vertical bars) was ﬁtted to a Gaussian (solid line) resulting in a maximum at 295.4K and a full- width-at-half-maximum of 0.15K. For the determination of the temporal resolution limit of the (tta−stands for thenoyltriﬂuoroacetonate). A 0.01 K tempera- there is a superposition between the emission of Eu3+ (5D0→7F0, 1) and Tb3+ (5D4→7F4). (B) Normalized integrated intensity of 5D0→7F2(red) and 5D4→7F5(green) in the temperature range 290–330K. there is a superposition between the emission of Eu3+ (5D0→7F0, 1) and Tb3+ (5D4→7F4). (B) Normalized integrated intensity of 5D0→7F2(red) and 5D4→7F5(green) in the temperature range 290–330K. FIGURE 2 \| (A) Temperature dependence of the emission spectra of UET–1.3 (excited at 365 nm) in the 10–330K temperature range. The f-f lines corresponding to the 5D4→7F6, 5 (Tb3+) and 5D0→7F2−4 (Eu3+) transitions are identiﬁed. In the area marked with an asterisk FIGURE 3 \| (A) Local calibration curve of UET–1.3 for the temperature range 290–330K. The temperature was cycled three times and the emission spectra recorded at equal time intervals when the temperature increases. For all emission spectra the computed parameter was divided by its value at 303 K (0) to get a thermometric parameter near the unit. The results for all cycles are overlapped and the second degree polynomial ﬁt presented in the Figure (r2 > 0.995) is the local calibration curve. (B) Relative frequency of the UET–1.3 temperature read during 60 s in the absence of current in the heating circuit (inset). The distribution of temperature (vertical bars) was ﬁtted to a Gaussian (solid line) resulting in a maximum at 295.4K and a full- width-at-half-maximum of 0.15K. degree polynomial ﬁt presented in the Figure (r2 > 0.995) is the local calibration curve. (B) Relative frequency of the UET–1.3 temperature read during 60 s in the absence of current in the heating circuit (inset). DISCUSSION The seminal work of Kolodner and Tyson (1983) demon- strated the potential of non-contact thermometry to map inte- grated circuits reporting the temperature mapping of a MOSFET through the emission of a polymer ﬁlm containing Eu(tta)32H2O Burzo et al. (2005) used the thermoreﬂectance of a MOSFET device to map the temperature in a window of 15 × 50 μm, with uncertainly of 13% (or 2.6 K) and spatial/temporal resolutions of 2.3 μm/1.1 μ s, respectively. Tessier et al. (2007) recognized July 2013 \| Volume 1 \| Article 9 \| 3 www.frontiersin.org www.frontiersin.org Hybrid ﬁlms for temperature maping Brites et al. Brites et al. Hybrid ﬁlms for temperature maping FIGURE 4 \| Pseudocolor temperature maps reconstructed from the emission of UET–1.3 collected with the 200 µm core diameter ﬁber (A) along the directions A and B indicated in Figure 1A, using a scanning step of 200 µm, and (B) along A direction, changing the scanning step from 50 to 200 µm. The shadowed areas correspond to the position of the copper tracks. FIGURE 5 \| (A) Temperature proﬁles along the A direction indicated in Figure 1A for the 200 μm core diameter ﬁber using scanning steps of 50 μm (blue squares) and 200 μm (red circles). The maximum temperature gradient is 0.357 × 106 K·m−1. (B) Temporal dependence of the temperature over the copper track represented in (A) measured with the 450 μm core diameter ﬁber. The arrow marks the instant when the heating current was turned on. FIGURE 4 \| Pseudocolor temperature maps reconstructed from the emission of UET–1.3 collected with the 200 µm core diameter ﬁber (A) along the directions A and B indicated in Figure 1A, using a scanning step of 200 µm, and (B) along A direction, changing the scanning step from 50 to 200 µm. The shadowed areas correspond to the position of the copper tracks. FIGURE 5 \| (A) Temperature proﬁles along the A direction indicated in Figure 1A for the 200 μm core diameter ﬁber using scanning steps of 50 μm (blue squares) and 200 μm (red circles). The maximum temperature gradient is 0.357 × 106 K·m−1. (B) Temporal dependence of the temperature over the copper track represented in (A) measured with the 450 μm core diameter ﬁber. The arrow marks the instant when the heating current was turned on. DISCUSSION FIGURE 5 \| (A) Temperature proﬁles along the A direction indicated in Figure 1A for the 200 μm core diameter ﬁber using scanning steps of 50 μm (blue squares) and 200 μm (red circles). The maximum temperature gradient is 0.357 × 106 K·m−1. (B) Temporal dependence of the temperature over the copper track represented in (A) measured with the 450 μm core diameter ﬁber. The arrow marks the instant when the heating current was turned on. that the thermoreﬂectance of integrated circuits fabricated over silicon can also be used in backside imaging exploiting the trans- parency of this substrate in the near IR region. The technique produces temperature mapping with spatial resolution of 1.7 μm and temporal resolution deﬁned by the camera triggering at 50 ms (Tessier et al., 2007). The small value and the temperature and wavelength dependence of the thermoreﬂectance coefﬁcient is actually the most challenging aspect for thermoreﬂectance based thermometry, making the setup for temperature mapping quite sophisticated (Burzo et al., 2005). et al. (2011) used a small ﬂuorescent Er3+/Yb3+-doped PbF2 nanocrystal as a temperature sensor. The technique presents temperature uncertainly ∼1.0 K, spatial resolution of 0.027 μm, despite the relatively long acquisition times (100 ms per pixel), that invalidates the transient mapping of the device (Saïdi et al., 2011). In fact, the use of conventional optical microscopy for temper- ature mapping set the Rayleigh criterion as the ultimate spatial resolution limit (Tessier et al., 2007; Serrels et al., 2008): δxRL = 1.22yλ D = 1.22 λ NA (5) The use of a scanning thermal microscope (SThM) adapted for ﬂuorescence reads (Benayas et al., 2012) was reported by Aigouy et al. (2005, 2009, 2011) and Saïdi et al. (2009, 2011) to work in the sub-wavelength spatial resolution regime. Regarding high-resolution thermal imaging of integrated circuits, Saïdi (5) where λ is the maximum wavelength, y is the distance to the emitting surface, D is the diameter of the detector and NA is the Frontiers in Chemistry \| Inorganic Chemistry July 2013 \| Volume 1 \| Article 9 \| 4 Hybrid ﬁlms for temperature maping Brites et al. FIGURE 7 \| Spatio-temporal resolution for temperature mapping of integrated circuits using distinct luminescent thermometers. FIGURE 6 \| Spatial resolution of the UET–1.3 thermometers for different scanning steps. The red circles and black squares correspond to the mappings using the 200 and 450μm diameter ﬁber, respectively. DISCUSSION The solid lines are guides for the eyes. The interrupted red and black lines correspond to the Rayleigh spatial resolution limit of the 200 and 450 μm diameter ﬁbers, respectively. FIGURE 7 \| Spatio-temporal resolution for temperature mapping of integrated circuits using distinct luminescent thermometers. FIGURE 6 \| Spatial resolution of the UET–1.3 thermometers for different scanning steps. The red circles and black squares correspond to the mappings using the 200 and 450μm diameter ﬁber, respectively. The solid lines are guides for the eyes. The interrupted red and black lines correspond to the Rayleigh spatial resolution limit of the 200 and 450 μm diameter ﬁbers, respectively. FIGURE 6 \| Spatial resolution of the UET–1.3 thermometers for different scanning steps. The red circles and black squares correspond to the mappings using the 200 and 450μm diameter ﬁber, respectively. The solid lines are guides for the eyes. The interrupted red and black lines correspond to the Rayleigh spatial resolution limit of the 200 and 450 μm diameter ﬁbers, respectively. FIGURE 7 \| Spatio-temporal resolution for temperature mapping of integrated circuits using distinct luminescent thermometers. FIGURE 6 \| Spatial resolution of the UET–1.3 thermometers for different scanning steps. The red circles and black squares correspond to the mappings using the 200 and 450μm diameter ﬁber, respectively. The solid lines are guides for the eyes. The interrupted red and black lines correspond to the Rayleigh spatial resolution limit of the 200 and 450 μm diameter ﬁbers, respectively. FIGURE 7 \| Spatio-temporal resolution for temperature mapping of integrated circuits using distinct luminescent thermometers. temperature with spatio-temporal resolution of 0.42 μm/4.8 ms. The ﬁner temperature spatial resolution reached is below the Rayleigh optical spatial resolution limit, because it results from a spectroscopic measurement. The temperature mapping spatial resolution depends on the scanning step and on the diameter of the optical ﬁber used. The spatial resolution is not improved for scanning step values lower than the ﬁber inner radius; nevertheless narrower ﬁbers produce ﬁner temperature spatial resolution. correspondent numerical aperture. In recent years, a number of ﬂuorescence imaging techniques with sub-diffraction-limit reso- lution have been developed, achieving a spatial resolution until 0.01 μm (Rust et al., 2006). The Rayleigh spatial resolutions applied to the experimental parameters used in this work are presented in Figure 6. DISCUSSION Contrary to the Rayleigh spatial limit that increases for narrower ﬁbers, the temperature spatial resolution as deﬁned here (Equation 2) improves when narrower ﬁbers are used. This conclusion results from the area probed by the ﬁber (ﬁber ﬁeld-of-view) can change signiﬁcantly in a small step (the area change is 1.25% of the ﬁber ﬁeld-of-view for the best spatial resolution value). As the temperature readout results of a spectroscopic measurement (the temperature at each position is averaged over the ﬁeld-of-view of the ﬁber) it is not limited by the Rayleigh criterion. The spatial resolution is limited by the experimental setup used that produces a ﬁeld-of-view averaged temperature change above the sensitivity of the detector. The technique presented here can be easily used for routine temperature maps with cost-effective equipment (e.g., a portable spectrometer and a handheld excitation source) furnishing results of the same order of magnitude of those obtained with more sophisticated setups. We can foresee that the spatio-temporal res- olution values presented here does not constitute the ultimate limit of the technique that predictably will improve both reso- lutions using sensitivity-enhanced materials, for the 290–330 K operating range. Although the UET–1.3 ﬁlm thermometer is a cost effective competitive approach combining equitable spa- tial and temporal resolution we are currently investigating new materials to address these demands. Figure 7 compares the spatio-temporal resolution for temper- ature mapping of integrated circuits using distinct luminescent thermometers. This ﬁgure shows that the choice of high spatial resolution compromises the temporal resolution and vice-versa. To reach spatial resolution in the micrometer range (1–10 μm) the temporal resolution ranges from the microsecond to the frac- tion of the second. Although thermoreﬂectance based technique displays high temporal resolution and SThM based techniques present the highest spatial resolution, the combination of high spatial and temporal resolutions in a single temperature mapping has not been reported yet, showing that there is plenty of room to improve the temperature mappings reported so far. ACKNOWLEDGMENTS We acknowledge Fundação para a Ciência e a Tecnologia (FCT, Portugal), COMPETE and FEDER programs (Pest- C/CTM/LA0011/2011, PTDC/CTM/101324/2008, RECI/CTM- CER/0336/2012) and Integrated Spanish-Portuguese Action PT2009-0131 for ﬁnancial support. The work in Zaragoza has been supported by the grants MAT2011-25991 and CONSOLIDER CSD2007-00010 from the Ministry of Economy and Competitivity. Carlos D. S. Brites (SFRH/BPD/89003/2012) and Patrícia P. Lima (SFRH/BPD/34365/2006) thank FCT for grants. Nuno J. O. Silva acknowledges FCT for Ciência 2008 program. We acknowledge Fundação para a Ciência e a Tecnologia (FCT, Portugal), COMPETE and FEDER programs (Pest- C/CTM/LA0011/2011, PTDC/CTM/101324/2008, RECI/CTM- CER/0336/2012) and Integrated Spanish-Portuguese Action PT2009-0131 for ﬁnancial support. The work in Zaragoza has been supported by the grants MAT2011-25991 and CONSOLIDER CSD2007-00010 from the Ministry of Economy and Competitivity. Carlos D. S. Brites (SFRH/BPD/89003/2012) and Patrícia P. Lima (SFRH/BPD/34365/2006) thank FCT for grants. Nuno J. O. Silva acknowledges FCT for Ciência 2008 program. REFERENCES doi: 10.1088/0957-0233/15/9/R01 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. De De Zea Bermudez, V., Carlos, L. D., and Alcácer, L. (1999). Sol- gel derived urea cross-linked organically modiﬁed silicates. 1. Room temperature mid- infrared spectra. Chem. Mater. 11, 569–580. Benayas, A., Escuder, E., and Jaque, D. (2012). High-resolution confo- cal ﬂuorescence thermal imaging of tightly pumped microchip Nd:YAG laser ceramics. Appl. Phys. B. 107, 697–701. doi: 10.1007/s00340-012- 4879-y Rust, M. J., Bates, M., and Zhuang, X. W. (2006). Sub-diffraction-limit imaging by stochastic optical recon- struction microscopy (STORM). Nat. Methods 3, 793–795. doi: 10.1038/nmeth929 Received: 13 May 2013; paper pend- ing published: 27 May 2013; accepted: 18 June 2013; published online: 08 July 2013. Frazão, O., Correia, C., Santos, J. L., and Baptista, J. M. (2009). Raman ﬁbre Bragg-grating laser sensor with cooperative Rayleigh scattering for strain-temperature measure- ment. Meas. Sci. Technol. 20:045 203. doi: 10.1088/0957-0233/20/4/ 045203 Received: 13 May 2013; paper pend- ing published: 27 May 2013; accepted: 18 June 2013; published online: 08 July 2013. Citation: Brites CDS, Lima PP, Silva NJO, Millán A, Amaral VS, Palacio F and Carlos LD (2013) Organic– Inorganic Eu3+/Tb3+ codoped hybrid ﬁlms for temperature mapping in inte- grated circuits. Front. Chem. 1:9. doi: 10.3389/fchem.2013.00009 Brites, C. D. S., Lima, P. P., Silva, N. J. O., Millán, A., Amaral, V. S., Palacio, F., et al. (2010). A luminescent molecular thermome- ter for long-term absolute tempera- ture measurements at the nanoscale. Adv. Mater. 22, 4499–4504. doi: 10.1002/adma.201001780 Saïdi, E., Babinet, N., Lalouat, L., Lesueur, J., Aigouy, L., Volz, S., et al. (2011). Tuning temperature and size of hot spots and hot- spot arrays. Small 7, 259–264. doi: 10.1002/smll.201001476 Citation: Brites CDS, Lima PP, Silva NJO, Millán A, Amaral VS, Palacio F and Carlos LD (2013) Organic– Inorganic Eu3+/Tb3+ codoped hybrid ﬁlms for temperature mapping in inte- grated circuits. Front. Chem. 1:9. doi: 10.3389/fchem.2013.00009 Jung, W., Kim, Y. W., Yim, D., and Yoo, J. Y. (2011). Microscale surface ther- mometry using SU8/Rhodamine-B thin layer. Sensor Actuat. A-Phys. 171, 228–232. Brites, C. D. S., Lima, P. P., Silva, N. J. O., Millán, A., Amaral, V. S., Palacio, F., et al. (2011). Lanthanide-based luminescent molecular thermome- ters. New. J. Chem. 35, 1177–1183. REFERENCES Kolodner, P., and Tyson, J. A. (1983). Remote thermal imaging with 0.7-mm spatial-resolution using temperature-dependent ﬂuorescent thin-ﬁlms. Appl. Phys. Lett. 42, 117–119. doi: 10.1063/1.93766 thermal imaging of integrated cir- cuits at high spatial resolution. Appl. Phys. Lett. 90, 171112. doi: 10.1063/1.2732179 organic-inorganic hybrid materi- als. J. Lumin. 133, 230–232. doi: 10.1016/j.jlumin.2011.12.050 Aigouy, L., Lalouat, L., Mortier, M., Löw, P., and Bergaud, C. (2011). A scanning thermal probe micro- scope that operates in liquids. Rev. Sci. Instrum. 82, 036106. doi: 10.1063/1.3567794 Burzo, M. G., Komarov, P. L., and Raad, P. E. (2005). Noncontact transient temperature mapping of active electronic devices using the thermoreﬂectance method. IEEE Trans. Compon. Pack. Technol. 28, 637–643. doi: 10.1109/TCAPT. 2005.859738 Vetrone, F., Naccache, R., Zamarron, A., De La Fuente, A. J., Sanz- Rodriguez, F., Maestro, L. M., et al. (2010). Temperature sensing using ﬂuorescent nanothermometers. ACS Nano 4, 3254–3258. doi: 10.1021/nn100244a Kuzmin, A. N., Baev, A., Kachynski, A. V., Fisher, T. S., Shakouri, A., and Prasad, P. N. (2011). Anti-Stokes ﬂuorescence imaging of microscale thermal ﬁelds in thin ﬁlms. J. Appl. Phys. 110, 033512-1–033512-5. doi: 10.1063/1.3606429 Aigouy, L., Saïdi, E., Lalouat, L., Labeguerie-Egea, J., Mortier, M., Löw, P., et al. (2009). AC thermal imaging of a microwire with a ﬂuorescent nanocrystal: inﬂuence of the near ﬁeld on the thermal contrast. J. Appl. Phys. 106, 074301. doi: 10.1063/1. 3233940 Phys. 110, 033512-1–033512-5. doi: 10.1063/1.3606429 Christofferson, J., Maize, K., Ezzahri, Y., Shabani, J., Wang, X., and Shakouri, A. (2008). Microscale and nanoscale thermal charac- terization techniques (Reprinted from Thermal Issues in Emerging Technologies: theory and applica- tion, January, 2007). J. Electron. Packag. 130, 041101–041106. doi: 10.1115/1.2993145 Yarimaga, O., Lee, S., Ham, D. Y., Choi, J. M., Kwon, S. G., Im, M., et al. (2011). Thermoﬂuorescent conjugated polymer sensors for nano- and microscale temper- ature monitoring. Macromol. Chem. Phys. 212, 1211–1220. doi: 10.1002/macp.201100099 Liu, H. X., Sun, W. Q., Xiang, A., Shi, T. W., Chen, Q., and Xu, S. Y. (2012). Towards on-chip time-resolved thermal mapping with micro-/nanosensor arrays. Nanoscale Res. Lett. 7, 1–6. doi: 10.1186/1556-276X-7-484 Aigouy, L., Tessier, G., Mortier, M., and Charlot, B. (2005). Scanning ther- mal imaging of microelectronic cir- cuits with a ﬂuorescent nanoprobe. Appl. Phys. Lett. 87, 184105. doi: 10.1063/1.2123384 Meola, C., and Carlomagno, G. M. (2004). Recent advances in the use of infrared thermography. Meas. Sci. Technol. 15, R27–R58. CONCLUSIONS Here we report the use of Eu3+/Tb3+ co-doped di-ureasil thin ﬁlms to perform temperature mapping of integrated circuits with July 2013 \| Volume 1 \| Article 9 \| 5 www.frontiersin.org www.frontiersin.org Hybrid ﬁlms for temperature maping Brites et al. REFERENCES doi: 10.1039/c0nj01010c Saïdi, E., Samson, B., Aigouy, L., Volz, S., Löw, P., Bergaud, C., et al. (2009). Scanning thermal imaging by near-ﬁeld ﬂuorescence spectroscopy. Nanotechnology 20, 115703–115711. doi: 10.1088/0957- 4484/20/11/115703 This article was submitted to Frontiers in Inorganic Chemistry, a specialty of Frontiers in Chemistry. Kersey, A. D., and Berkoff, T. A. (1992). Fiberoptic Bragg-grating differential-temperature sensor. IEEE Photonics Technol. Lett. 4, 1183–1185. doi: 10.1109/68.163773 Copyright © 2013 Brites, Lima, Silva, Millán, Amaral, Palacio and Carlos. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics etc. Brites, C. D. S., Lima, P. P., Silva, N. J. O., Millán, A., Amaral, V. S., Palacio, F., et al. (2012). Thermometry at the nanoscale. Nanoscale 4, 4799–4829. doi: 10.1039/c2nr30663h Serrels, K. A., Ramsay, E., Dalgarno, P. A., Gerardot, B. D., O’Connor, J. A., Hadﬁeld, R. H., et al. (2008). Solid immersion lens applica- tions for nanophotonic devices. J. Nanophoton. 2, 021854. doi: 10.1117/1.3068652 Kim, K., Jeong, W. H., Lee, W. C., and Reddy, P. (2012). Ultra- high vacuum scanning thermal microscopy for nanometer reso- lution quantitative thermometry. ACS Nano 6, 4248–4257. doi: 10.1021/nn300774n Brites, C. D. S., Lima, P. P., Silva, N. J. O., Millán, A., Amaral, V. S., Palacio, F., et al. (2013). Thermometry at the nano- scale using lanthanide-containing Tessier, G., Bardoux, M., Boue, C., and Fournier, D. (2007). Back side July 2013 \| Volume 1 \| Article 9 \| 6 Frontiers in Chemistry \| Inorganic Chemistry
https://openalex.org/W3162927758	https://www.mdpi.com/2072-6694/13/10/2453/pdf?version=1621391339	English	null	The Significance of Prostate Specific Antigen Persistence in Prostate Cancer Risk Groups on Long-Term Oncological Outcomes	Cancers	2,021	cc-by	6,173	Citation: Milonas, D.; Venclovas, Z.; Sasnauskas, G.; Ruzgas, T. The Signiﬁcance of Prostate Speciﬁc Antigen Persistence in Prostate Cancer Risk Groups on Long-Term Oncological Outcomes. Cancers 2021, 13, 2453. https://doi.org/10.3390/ cancers13102453 Abstract: Objective: To assess the signiﬁcance of prostate-speciﬁc antigen (PSA) persistence at the ﬁrst measurement after radical prostatectomy (RP) on long-term outcomes in different prostate cancer risk groups. Methods: Persistent PSA was deﬁned as ≥0.1 ng/mL at 4–8 weeks after RP. Patients were stratiﬁed into low-, intermediate- and high-risk groups, according to the preoperative PSA, pathological stage, grade group and lymph nodes status. The ten-year cumulative incidence of biochemical recurrence (BCR), metastases, cancer-speciﬁc mortality (CSM) and overall mortality (OM) were calculated in patients with undetectable and persistent PSA in different PCa-risk groups. Multivariate regression analyses depicted the signiﬁcance of PSA persistence on each study endpoint. Results: Of all 1225 men, in 246 (20.1%), PSA persistence was detected. These men had an increased risk of BCR (hazard ratio (HR) 4.2, p < 0.0001), metastases (HR: 2.7, p = 0.002), CRM (HR: 5.5, p = 0.002) and OM (HR: 1.8, p = 0.01) compared to the men with undetectable PSA. The same signiﬁcance of PSA persistence on each study endpoint was found in the high-risk group (HR: 2.5 to 6.2, p = 0.02 to p < 0.0001). In the intermediate-risk group, PSA persistence was found as a predictor of BCR (HR: 3.9, p < 0.0001), while, in the low-risk group, PSA persistence was not detected as a signiﬁcant predictor of outcomes after RP. Conclusions: Persistent PSA could be used as an independent predictor of worse long-term outcomes in high-risk PCa patients, while, in intermediate-risk patients, this parameter signiﬁcantly predicts only biochemical recurrence and has no impact on the outcomes in low-risk PCa patients. 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/). Keywords: prostate cancer; PSA persistence; risk groups; radical prostatectomy; outcomes The Signiﬁcance of Prostate Speciﬁc Antigen Persistence in Prostate Cancer Risk Groups on Long-Term Oncological Outcomes Daimantas Milonas 1,, Zilvinas Venclovas 1, Gustas Sasnauskas 1 and Tomas Ruzgas 2 1 Medical Academy, Department of Urology, Lithuanian University of Health Sciences, 44307 Kaunas, Lithuania; zilvinas.venclovas@stud.lsmu.lt (Z.V.); gustas.sasnauskas@stud.lsmu.lt (G.S.) 2 Department of Applied Mathematics, Kaunas University of Technology, 44249 Kaunas, Lithuania; tomas.ruzgas@ktu.lt g Correspondence: daimantas.milonas@lsmuni.lt Simple Summary: The current prostate cancer guidelines recommend performing the ﬁrst prostate- speciﬁc antigen measurement at three months after radical prostatectomy. However, at an earlier measurement, persistence (≥0.1 ng/mL) of this biomarker could be found in up to 30% of cases, depending on the prostate cancer risk factors. Recent reports have demonstrated an increasing interest in prostate-speciﬁc antigen persistence as a possible additional predictor of disease progression and cancer-speciﬁc survival. However, the data remain scant, with weak evidence. We assessed the relationship between prostate-speciﬁc antigen persistence and long-term oncological outcomes within prostate cancer risk groups. We found that persistence of this biomarker could be used as an independent predictor of worse long-term outcomes in high-risk prostate cancer patients, while in intermediate-risk patients, this parameter signiﬁcantly predicts only biochemical recurrence and has no impact on the outcomes in low-risk patients. Citation: Milonas, D.; Venclovas, Z.; Sasnauskas, G.; Ruzgas, T. The Signiﬁcance of Prostate Speciﬁc Antigen Persistence in Prostate Cancer Risk Groups on Long-Term Oncological Outcomes. Cancers 2021, 13, 2453. https://doi.org/10.3390/ cancers13102453 Academic Editors: Pierre Jean Lamy, Christophe Hennequin, Mathieu Roumiguie and Xavier Rebillard Received: 25 April 2021 Accepted: 15 May 2021 Published: 18 May 2021 Publisher’s Note: MDPI stays neutral cancers cancers cancers 2.1. Patient Population Between 2001 and 2019, 2611 men were treated by RP for clinically localized and locally advanced PCa at the Department of Urology of the Lithuanian University of Health Sciences. Patient data were registered in a PCa database. During the study period, RP was performed by 9 senior urologists. The exclusion criteria were as follows: no follow-up data, a patient did not have PSA measurement within the ﬁrst 2 months after RP, neo- and adjuvant treatments, detected metastases before RP and incomplete clinical or pathological data. Patients were stratiﬁed according to undetectable PSA (PSA < 0.1 ng/mL) vs. persis- tent PSA (PSA ≥0.1 ng/mL) at the ﬁrst measurement 4–8 weeks after RP. Furthermore, patients were divided according to the pathological PCa features into low-risk (pT2, GG1, PSA <10 ng/mL and Nx, pN0); intermediate-risk (pT3a, GG2–3, PSA 10–20 ng/mL and Nx, pN0) and high-risk (pT3b, GG4–5, PSA > 20 ng/mL and pN1) groups. Our study ﬂowchart is presented in Figure 1. Tumor grading was classiﬁed using the revised 2005 International Society of Urologic Pathology (ISUP) Gleason score grading system and the suggested the new GG model after 2014 [16,17]. The university’s ethical committee approved the collection and analysis of the data. 1. Introduction The main limitation of this analysis is the low number of studies analyzing the endpoints of different outcomes. There- fore, to date, there is a lack of evidence on the prognostic signiﬁcance of PSA persistence. Despite presenting, in several reports, a strong association between PSA persistence and unfavorable cancer characteristics [5,7,11], there are no analyses on the importance of PSA persistence in different PCa-risk groups. p g p The aim of our study was to assess the relationship between persistent PSA at the ﬁrst measurement between 4 and 8 weeks after RP and the long-term oncological outcomes within the PCa-risk groups. We hypothesized that the signiﬁcance of PSA persistence on the oncological outcome differed in the PCa-risk groups. Retrospective analyses were per- formed assessing the signiﬁcance of PSA persistence on the 10-year cumulative incidence of BCR, metastases (MTS), CSM and overall mortality (OM) in patients after RP. 1. Introduction Radical prostatectomy (RP) with or without lymph node dissection is an accepted treat- ment modality in patients with localized and locally advanced prostate cancer (PCa) [1–3]. https://www.mdpi.com/journal/cancers Cancers 2021, 13, 2453. https://doi.org/10.3390/cancers13102453 Cancers 2021, 13, 2453 2 of 13 However, a non-negligible number of patients may experience disease recurrence following RP. The European Association of Urology (EAU) prostate cancer (PCa) guidelines recom- mend performing the ﬁrst prostate-speciﬁc antigen (PSA) measurement at three months after RP [4]. Despite that, at early PSA measurements within 4–8 weeks, PSA persistence (≥0.1 ng/mL) could be found in 8–26% of men, depending on the PCa risk factors [5,6]. Recent reports have demonstrated an increasing interest in PSA persistence as a possible additional predictor of disease progression and cancer-speciﬁc survival [5–13]. A very recent systemic review and a meta-analysis included studies assessing the importance of PSA persistence on oncological outcomes after RP and the management of persistently elevated PSA [14,15]. The authors demonstrated the association between PSA persistence and biochemical recurrence (BCR), disease progression and cancer-speciﬁc mortality (CSM) and suggested a beneﬁt from immediate radiotherapy [15]. The main limitation of this analysis is the low number of studies analyzing the endpoints of different outcomes. There- fore, to date, there is a lack of evidence on the prognostic signiﬁcance of PSA persistence. Despite presenting, in several reports, a strong association between PSA persistence and unfavorable cancer characteristics [5,7,11], there are no analyses on the importance of PSA persistence in different PCa-risk groups. However, a non-negligible number of patients may experience disease recurrence following RP. The European Association of Urology (EAU) prostate cancer (PCa) guidelines recom- mend performing the ﬁrst prostate-speciﬁc antigen (PSA) measurement at three months after RP [4]. Despite that, at early PSA measurements within 4–8 weeks, PSA persistence (≥0.1 ng/mL) could be found in 8–26% of men, depending on the PCa risk factors [5,6]. Recent reports have demonstrated an increasing interest in PSA persistence as a possible additional predictor of disease progression and cancer-speciﬁc survival [5–13]. A very recent systemic review and a meta-analysis included studies assessing the importance of PSA persistence on oncological outcomes after RP and the management of persistently elevated PSA [14,15]. The authors demonstrated the association between PSA persistence and biochemical recurrence (BCR), disease progression and cancer-speciﬁc mortality (CSM) and suggested a beneﬁt from immediate radiotherapy [15]. 2.2. Outcomes The study endpoints were the 10-year cumulative incidences of BCR, MTS, CSM and OM. BCR was deﬁned as PSA > 0.2 ng/mL at two consecutive measurements. MTS were deﬁned as skeletal or visceral lesions conﬁrmed by a bone scan, computed tomography (CT) or magnetic resonance imaging (MRI) using RECIST criteria. Local and loco-regional recurrence was histopathologically conﬁrmed by surgery or biopsy or by MRI. MTS-free survival was deﬁned as the time from surgery until the detection of MTS. Death without MTS was considered a competing event. Alive patients without MTS were censored at the last follow-up. OM was deﬁned as the time from surgery until death for any cause. Alive patients were censored. CSM was deﬁned as the time from surgery until cancer-related death. Death for other causes was considered a competing event. Alive patients were 3 of 13 Cancers 2021, 13, 2453 censored as well. Data about patient death were taken from the national healthcare database. All cases of death were rechecked with follow-up data available in the center database. censored as well. Data about patient death were taken from the national healthcare database. All cases of death were rechecked with follow-up data available in the center database. censored as well. Data about patient death were taken from the national healthcare database All cases of death were rechecked with follow-up data available in the center database. Figure 1. Study ﬂowchart. RP—radical prostatectomy, PCa—prostate cancer, GG—grade groups and PSA—prostate- speciﬁc antigen. ﬂowchart. RP—radical prostatectomy, PCa—prostate cancer, GG—grade groups and PSA—prostate- Figure 1. Study ﬂowchart. RP—radical prostatectomy, PCa—prostate cancer, GG—grade groups and PSA—prostate- speciﬁc antigen. 2.3. Statistical Analysis Descriptive statistics included frequencies and proportions for categorical variables, medians with interquartile range and means with 95% conﬁdence intervals (CI) for con- tinuous variables. The chi-square and Mann–Whitney U tests were used to assess the differences between groups. The cumulative incidence function was used to estimate the 10-year BCR, MTS, CSM and OM in men with persistent vs. undetectable PSA in different PCa-risk groups. The Kaplan–Meier method was used to estimate survival in the study groups. A multivariate logistic regression model was used to test the relationship between the clinicopathological covariates and study endpoints. The results were presented as hazard ratios (HR) with 95% CI. Analyses were performed using SAS software (version 9.4 of the SAS System for Windows, by SAS Institute Inc., Cary, NC, USA) with the two-sided signiﬁcance level set at p < 0.05. 3. Results The patient characteristics are presented in Table 1. The median follow-up (interquar- tile range) was 103 (IQR 50–157) months. Of all 1225 studied men, 246 (20.1%) had PSA persistence at the ﬁrst measurement within 4–8 weeks after RP. PSA persistence was detected in 23 of 261 (8.8%), 94 of 705 (13.3%) and 129 of 259 (49.8%) men in the low-, intermediate- and high-risk groups, respectively. Overall, the 10-year cumulative incidence of BCR, MTS, CSM and OM were 39.61% (95% CI: 35.95–43.64), 9.70% (95% CI: 7.67–12.27), 4.81% (95% CI: 3.49–6.61) and 18.15% (95% CI: 16.04–20.53), respectively, with a signiﬁcantly higher incidence in men with PSA persistence (all p < 0.0001, Figure 2). In the multivariate Cancers 2021, 13, 2453 4 of 13 regression analysis, PSA persistence was detected as an independent predictor for each study endpoint (HR: 1.8 to 5.5, p < 0.01 to p < 0.0001, see Table 2). Table 1. Patient characteristics. IQR—interquartile range, PSA—prostate-speciﬁc antigen, LN—lymph nodes, LNI—lymph node invasion, BCR—biochemical recurrence and MTS metastasis range, PSA—prostate-speciﬁc antigen, LN—lymph nodes, LNI—lymph node invasion, BCR—biochemical recurrence asis. 3. Results Parameter All n = 1225 Low-Risk n = 261 Intermediate-Risk n = 705 High-Risk n = 259 p-Value Age (years): median (IQR) 64 (59–68) 63.5 (59–68) 63 (58–68) 65 (60–69) 0.01 PSA (ng/mL): median (IQR) 6.5 (4.8–9.95) 55 (4.3–6.9) 6.3 (4.8–9.6) 10.8 (6.2–21.1) <0.0001 Clinical stage: n, (%) <0.0001 cT1 337 (27.5) 106 (40.6) 197 (27.9) 34 (13.1) cT2 706 (57.6) 144 (55.2) 428 (60.7) 134 (51.7) cT3 178 (14.5) 11 (4.2) 76 (10.8) 91 (35.1) Unknown 4 (0.3) - 4 (0.6) - Biopsy Gleason score: n, (%) <0.0001 6 710 (58) 249 (95.4) 394 (55.9) 67 (25.9) 3 + 4 363 (29.6) 11 (4.2) 270 (38.3) 82 (31.7) 4 + 3 53 (4.3) 1 (0.4) 22 (3.2) 30 (11.6) 8 64 (5.2) - 16 (2.3) 48 (18.5) 9–10 33 (2.7) - 1 (0.1) 32 (12.4) Unknown 2 (0.2) - 2 (0.2) - Pathological Gleason Score: n, (%) <0.0001 6 329 (26.9) 261 (100) 58 (8.2) 10 (3.9) 3 + 4 626 (51.1) - 569 (80.7) 57 (22) 4 + 3 113 (9.2) - 78 (11.1) 35 (13.5) 8 70 (5.7) - - 70 (27.0) 9–10 87 (7.1) - - 87 (33.6) Pathologic stage: n, (%) <0.0001 pT2 776 (63.3) 261 (100) 459 (65.1) 56 (21.6) pT3a 326 (26.6) - 246 (34.9) 80 (30.9) pT3b-pT4 123 (10) - - 123 (47.5) Positive surgical margin: n, (%) 399 (33.7) 41 (16.1) 221 (32.2) 137 (56.1) <0.0001 PLND: n, (%) 489 (39.9) 61 (23.4) 228 (32.3) 200 (77.2) <0.0001 LNI: n, (%) 65 (13.3) - - 65 (32.5) <0.0001 Persistent PSA 246 (20.1) 23 (8.8) 94 (13.3) 129 (49.8) <0.0001 BCR: n, (%) 383 (31.3) 27 (10.3) 179 (25.4) 177 (68.3) <0.0001 MTS: n, (%) 87 (7.1) 4 (1.5) 24 (3.4) 59 (22.8) <0.0001 Death: n, (%) 226 (18.4) 46 (17.6) 113 (16) 67 (25.9) <0.0001 Cancer related death: n, (%) 45 (3.8) 3 (1.1) 11 (1.6) 33 (12.7) <0.0001 IQR—interquartile range, PSA—prostate-speciﬁc antigen, LN—lymph nodes, LNI—lymph node invasion, BCR—biochemical recurrence and MTS metastasis regression analysis, PSA persistence was detected as an independent predictor for each study endpoint (HR: 1.8 to 5.5, p < 0.01 to p < 0.0001, see Table 2). regression analysis, PSA persistence was detected as an independent predictor for each study endpoint (HR: 1.8 to 5.5, p < 0.01 to p < 0.0001, see Table 2). Cancers 2021, 13, 2453 5 of 13 Figure 2. 3. Results Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-specific mortality (CSM) and overall mortality (OM) in all study cohort patients with undetectable vs. persistent prostate-specific antigen (PSA). Figure 2. Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-specific mortality (CSM) and overall mortality (OM) in all study cohort patients with undetectable vs. persistent prostate-specific antigen (PSA). 3.1. Outcomes in the Low-Risk Group 3.1. Outcomes in the Low Risk Group The mean persistent PSA value at the ﬁrst measurement was 0.31 ng/mL (95% CI: 0.16–0.47), and the range was 0.1–1.4 ng/mL. At the 10-year cumulative incidence of BCR, MTS, CSM and OM, the rates in men with undetectable vs. persistent PSA were 11.3% vs. 32.9% (p = 0.03), 0.6% vs. 0.8% (p = 0.4), 1.0% vs. 4.8% (p = 0.2) and 14.1% vs. 26.6% (p = 0.07), respectively (Figure 3). In the multivariate regression analysis, PSA persistence was not detected as a signiﬁcant predictor of BCR, metastases, CSM or OM (Table 3). 6 of 13 Cancers 2021, 13, 2453 Table 2. Multivariable competing risk analysis of the biochemical recurrence (BCR), developing metastases (MTS), overall mortality (OM) and cancer-speciﬁc mortality (CSM) in the whole cohort. Parameter BCR MTS OM CSM HR 95% CI p HR 95% CI p HR 95% CI p HR 95% CI p PSA (ng/mL) 1.1 1–1.02 0.05 0.9 0.97–1.01 0.3 0.9 0.97–1.02 0.6 0.9 0.92–1.01 0.08 Age (years) 1.1 0.99–1.04 0.2 1.1 0.97–1.06 0.6 1.1 1.04–1.12 <0.001 1.1 1.0–1.14 0.04 PSA persistence 4.2 3.06–5.76 <0.001 2.7 1.44–5.09 0.002 1.8 1.13–2.76 0.01 5.5 2.08–14.49 0.006 Stage: pT2 Ref. Ref. Ref. Ref. pT3a 1.7 1.19–2.55 0.004 0.7 0.33–1.54 0.4 0.8 0.48–1.44 0.5 0.7 0.21–2.38 0.6 pT3b-pT4 2.3 1.48–3.59 0.0002 1.4 0.62–2.97 0.5 1.7 0.89–3.07 0.1 2.5 0.84–7.4 0.09 Grade Group: GG 1 Ref. Ref. Ref. Ref. GG 2 1.8 1.11–3.08 0.018 3 0.82–11.09 0.09 1.1 0.58–1.75 0.9 1.2 0.29–5.4 0.8 GG 3 3.6 1.99–6.43 <0.001 10.6 2.71–41.42 0.0007 1.1 0.43–2.59 0.9 0.9 0.09–9.98 0.9 GG 4 2.7 1.45–5.03 0.002 8.1 1.93–33.81 0.004 0.9 0.39–2.18 0.9 3.2 0.69–14.94 0.1 GG 5 3.6 1.99–6.44 <0.001 31.9 8.08–126.2 <0.0001 2.9 1.39–6.03 0.005 5.8 1.29–25.7 0.02 LNI 2.1 1.37–3.22 0.0006 1.4 0.75–2.78 0.3 1.3 0.67–2.57 0.4 2 0.83–5.01 0.1 SM 1.7 1.22–2.25 0.0013 1.8 0.96–3.2 0.07 1.3 0.84–2.04 0.2 2.2 0.89–5.45 0.09 CI—conﬁdence interval, PSA—prostate-speciﬁc antigen, LNI—lymph node invasion and SM—surgical margins. Cancers 2021, 13, 2453 7 of 13 Figure 3. 3. Results Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-speciﬁc mortality(CSM) and overall mortality (OM) in low-risk patients with undetectable vs. persistent prostate-speciﬁc antigen (PSA). 3.2. Outcomes in the Intermediate-Risk Group The mean persistent PSA value in this group was 0.32 ng/mL (95% CI: 0.23–0.41), and the range was 0.1–3.0 ng/mL. Comparing men with undetectable vs. persistent PSA, the 10-year cumulative incidence of BCR was 31.3% vs. 80.8% (p < 0.0001); the incidence of MTS was 4.7% vs. 13.7% (p = 0.03), CSM was 1.35% vs. 4.8% (p = 0.09) and OM was 14.8% vs. 18.7 % (p = 0.2) (Figure 4). The multivariate regression analysis revealed that PSA persistence was a signiﬁcant predictor of BCR (HR: 3.8, p < 0.0001), while signiﬁcance was not detected in the analyses of MTS (HR: 2.6, p = 0.1), CSM (HR: 4.2, p = 0.2) and OM (HR: 1.03, p = 0.9) (Table 3). Figure 3. Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-speciﬁc mortality(CSM) and overall mortality (OM) in low-risk patients with undetectable vs. persistent prostate-speciﬁc antigen (PSA). 3.2. Outcomes in the Intermediate-Risk Group 3.2. Outcomes in the Intermediate-Risk Group The mean persistent PSA value in this group was 0.32 ng/mL (95% CI: 0.23–0.41), and the range was 0.1–3.0 ng/mL. Comparing men with undetectable vs. persistent PSA, the 10-year cumulative incidence of BCR was 31.3% vs. 80.8% (p < 0.0001); the incidence of MTS was 4.7% vs. 13.7% (p = 0.03), CSM was 1.35% vs. 4.8% (p = 0.09) and OM was 14.8% vs. 18.7 % (p = 0.2) (Figure 4). The multivariate regression analysis revealed that PSA persistence was a signiﬁcant predictor of BCR (HR: 3.8, p < 0.0001), while signiﬁcance was not detected in the analyses of MTS (HR: 2.6, p = 0.1), CSM (HR: 4.2, p = 0.2) and OM (HR: 1.03, p = 0.9) (Table 3). 8 of 13 Cancers 2021, 13, 2453 Table 3. Multivariate competing risk analysis of the biochemical recurrence (BCR), developing metastases (MTS), overall mortality (OM) and cancer-speciﬁc mortality (CSM) in the low-, intermediate- and high-risk groups. 3. Results BCR MTS OM CSM Parameter HR 95% CI p HR 95% CI p HR 95% CI p HR 95% CI p Low Risk Group * PSA (ng/mL) 0.94 0.65–1.37 0.7 14.6 0.11− 0.2 1.3 0.88–1.9 0.1 0.9 0.33–2.23 0.7 Age (years) 1.1 0.97–1.26 0.1 1.1 0.62–1.65 0.9 1.1 0.95–1.18 0.2 0.8 0.6–1.18 0.3 Persistent PSA 3.9 0.79–19.84 0.09 21.4 0.1− 0.5 2.3 0.6–8.84 0.2 3.9 <0.0001− 0.9 SM 10.1 2.05–49.68 0.004 <0.1 <0.0001− 0.9 2.1 0.48–9.19 0.3 2.3 <0.0001− 0.9 Intermediate Risk Group PSA (ng/mL) 1.1 0.99–1.11 0.1 0.96 0.82–1.1 0.6 0.9 0.91–1.07 0.8 1.2 0.95–1.51 0.1 Age (years) 1.1 0.99–1.06 0.2 1.1 0.94–1.14 0.5 1.1 1.05–1.19 0.002 1.2 0.93–1.61 0.2 Persistent PSA 3.8 2.16–6.77 <0.0001 2.6 0.74–9.4 0.1 1.1 0.42–2.5 0.9 4.2 0.48–36.11 0.2 Stage pT2 Ref. Ref. Ref. Ref. pT3a 1.6 0.98–2.64 0.06 1.7 0.49–5.8 0.4 0.8 0.38–1.47 0.4 2.8 0.26–31.19 0.4 Grade Group GG 1 Ref. Ref. Ref. Ref. GG 2 0.73–3.07 0.3 1 <0.0001− 0.9 0.8 0.38–1.73 0.6 1 <0.0001− 0.9 GG 3 2.3 0.98–5.6 0.06 1 <0.0001− 0.9 0.9 0.26–3.79 0.9 2.1 <0.0001− 0.9 SM 2.9 1.75–4.88 <0.0001 3.9 1.04–14.5 0.04 1.8 0.93–3.39 0.08 1.3 0.15–11.87 0.8 High Risk Group * PSA (ng/mL) 1.1 0.99–1.02 0.5 0.9 0.97–1.01 0.5 0.9 0.96–1.02 0.4 0.9 0.91–1.01 0.08 Age (years) 1.1 0.98–1.03 0.8 1.1 0.96–1.06 0.7 1.1 1.01–1.12 0.015 1.1 0.99–1.15 0.09 Persistent PSA 5.1 3.31–7.99 <0.0001 2.6 1.2–5.62 0.015 2.7 1.36–5.45 0.005 6.2 1.66–23.06 0.007 Stage pT2 Ref. Ref. Ref. Ref. pT3a 1.8 0.95–3.39 0.07 0.5 0.17–1.32 0.2 0.9 0.33–2.41 0.8 0.3 0.05–1.75 0.2 pT3b–pT4 2.3 1.2–4.24 0.01 1.1 0.42–2.55 0.9 2 0.83–4.83 0.1 1.9 0.56–6.6 0.3 Grade Group GG 1 Ref. Ref. Ref. Ref. GG 2 4.1 0.92–18.23 0.06 1.6 0.18–15.09 0.7 0.8 0.19–2.99 0.7 0.7 0.06–9.04 0.8 GG 3 8.7 1.87–40.14 0.006 4.8 0.53–44.01 0.2 0.9 0.18–4.37 0.9 1.6 0.08–33.07 0.8 GG 4 5 1.12–22.02 0.035 3.9 0.43–35.07 0.2 0.8 0.2–3.21 0.8 2.7 0.25–30.59 0.4 GG 5 6.7 1.53–29.19 0.01 14.3 1.65–123.4 0.016 2.7 0.6–8.6 0.2 5.7 0.56–58.19 0.1 LNI 1.8 1.2–2.78 0.005 1.4 0.74–2.8 0.3 1.1 0.56–2.29 0.7 2.1 0.81–5.39 0.1 SM 1.1 0.77–1.65 0.5 1.6 0.8–3.35 0.2 0.8 0.4–1.48 0.4 1.9 0.66–5.77 0.2 CI—conﬁdence interval, PSA—prostate-speciﬁc antigen, LNI—lymph node invasion and SM—surgical margins. * Multivariate analysis adjusted for preoperative PSA, age, surgical margin status and PSA persistence. Multivariate analysis adjusted for preoperative PSA, age, surgical margin status, stage pT2 vs. 3. Results pT3a, grade group 1 vs. 2 vs.3 and PSA persistence. * Multivariate analysis adjusted for preoperative PSA, age, surgical margin status, stage pT2 vs. pT3a, grade groups, LNI and PSA persistence. Table 3. Multivariate competing risk analysis of the biochemical recurrence (BCR), developing metastases (MTS), overall mortality (OM) and cancer-speciﬁc mortality intermediate- and high-risk groups. Cancers 2021, 13, 2453 9 of 13 Figure 4. Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-speciﬁc mortal- ity (CSM) and overall mortality (OM) in intermediate-risk patients with undetectable vs. persistent prostate-speciﬁc antigen (PSA). Figure 4. Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-speciﬁc mortal- ity (CSM) and overall mortality (OM) in intermediate-risk patients with undetectable vs. persistent prostate-speciﬁc antigen (PSA). 3.3. Outcomes in the High-Risk Group 3.3. Outcomes in the High-Risk Group The mean persistent PSA in this group was 1.9 ng/mL (95% CI: 1.29–2.50), and the range was 0.1–24.6 ng/mL. The cumulative 10-year incidence of BCR, MTS, CSM and OM was signiﬁcantly lower in men with undetectable PSA compared to men with PSA persistence: 50.3% vs. 98% (p < 0.0001), 12.8% vs. 49% (p < 0.0001), 7% vs. 31.9% (p < 0.0001) and 21.2% vs. 43.6% (p = 0.0001), respectively (Figure 5). Differently to other study groups, men with detected PSA persistence had a signiﬁcantly increased risk of BCR (HR: 5.1, p < 0.0001), risk of MTS (HR: 2.6, p = 0.015), CSM (HR: 6.2, p = 0.007) and OM (HR: 2.7, p = 0.005) in the multivariate analysis (Table 3). The mean persistent PSA in this group was 1.9 ng/mL (95% CI: 1.29–2.50), and the range was 0.1–24.6 ng/mL. The cumulative 10-year incidence of BCR, MTS, CSM and OM was signiﬁcantly lower in men with undetectable PSA compared to men with PSA persistence: 50.3% vs. 98% (p < 0.0001), 12.8% vs. 49% (p < 0.0001), 7% vs. 31.9% (p < 0.0001) and 21.2% vs. 43.6% (p = 0.0001), respectively (Figure 5). Differently to other study groups, men with detected PSA persistence had a signiﬁcantly increased risk of BCR (HR: 5.1, p < 0.0001), risk of MTS (HR: 2.6, p = 0.015), CSM (HR: 6.2, p = 0.007) and OM (HR: 2.7, p = 0.005) in the multivariate analysis (Table 3). Cancers 2021, 13, 2453 10 of 13 Figure 5. Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-speciﬁc mortality (CSM) and overall mortality (OM) in high-risk patients with undetectable vs. persistent prostate-speciﬁc antigen (PSA). Figure 5 Cumulative incidence function for biochemica Figure 5. Cumulative incidence function for biochemical recurrence (BCR), metastases (MTS), cancer-speciﬁc mortality (CSM) and overall mortality (OM) in high-risk patients with undetectable vs. persistent prostate-speciﬁc antigen (PSA). 4. Discussion PSA remains the most important surrogate biomarker in the follow-up after radical PCa treatment. Recent studies have demonstrated the importance of PSA persistence in disease progression and CSM after RP [14,15]. However, published data are scant, and there are no reports analyzing the importance of PSA persistence according to PCa risk groups. We investigated the relationship between PSA persistence and long-term oncological outcomes in patients with different risk groups. Among all >1200 patients with available ﬁrst PSA measurements within 4–8 weeks after RP, we found that PSA persistence (>0.1 ng/mL) was a strong predictor of BCR, MTS, CSM and OM. When patients with low-, intermediate- and high-risk PCa features were analyzed separately, we detected different predictive probabilities of PSA persis- tence for the study endpoints. The results presented herein could provide important prognostic information for clinicians and patients very shortly after the initial surgical treatment. A major implication of our study is that the early detection of PSA failure after the initial surgical treatment may assist in more personalized follow-up and additional treatment decision-making. In the low-risk patient cohort, PSA persistence was associated with a worse 10-year cumulative incidence of BCR (33% vs. 11%, p = 0.03) compared to men with undetectable PSA. However, in the multivariate regression analysis, PSA persistence was not found as a signiﬁcant predictor of BCR. Moreover, in low-risk patients, PSA persistence was not associated with a more rapid disease progression and mortality. Our results demonstrate that men with favorable cancer features and PSA persistence will very likely have the Cancers 2021, 13, 2453 11 of 13 11 of 13 same low risk of developing MTS and cancer-related death as men with undetectable PSA. Despite the association with an increased incidence of BCR, PSA persistence in low-risk PCa should not be used as an indicator for early salvage treatment. In the intermediate-risk group, men with PSA persistence demonstrated an increased 10-year cumulative incidence of BCR and MTS, with absolute 51% and 8% differences compared to men with undetectable PSA (p < 0.0001 and p < 0.02, respectively), while the regression analysis revealed that PSA persistence could predict only BCR (HR: 3.9, p < 0.0001). In intermediate-risk patients, PSA persistence demonstrated a signiﬁcant association with BCR. However, this association did not translate to a signiﬁcantly higher risk of MTS or CSM. 4. Discussion Probably, these men are good candidates for an intensiﬁed PSA follow- up and the initiation of additional delayed treatment, according to the PSA dynamics [18], rather than early salvage treatment [19]. Differently to the other groups, in the patients with high-risk PCa features, PSA persistence was associated with a signiﬁcantly higher 10-year cumulative incidence of BCR, MTS, CSM and OM (absolute differences 48%, 39%, 26% and 24%, respectively). Moreover, PSA persistence was found as a signiﬁcant independent predictor at each study endpoint (HR: 2.5–6.2, p < 0.02 to <0.0001). The aggressive nature of high-risk PCa is well- known from previous studies [1]. However, this speciﬁc PCa subset is heterogeneous with different risks of progression and responses to salvage treatments [20,21]. An additional validated biomarker would be very helpful for the recognition of the most aggressive high- risk prostate cancer. The results of the current study show that men with high-risk PCa features and PSA persistence are at a signiﬁcantly higher risk of worse outcomes compared to men with high-risk cancer and undetectable PSA. Moreover, the regression analysis demonstrated that PSA persistence is the most important predictor of CSM. Therefore, PSA persistence could be useful in clinical practice for the early identiﬁcation of patients with potentially rapid disease progression after RP. Similarly, a higher PSA level at 3 months after radiation therapy was found as a signiﬁcant predictor of worse outcomes, mainly in high-risk patients [22]. Taken together, the presented study results suggest that the importance of PSA persis- tence reported in previous studies [5–14] may not be directly translated to each patient: the signiﬁcance of PSA persistence is marginal in low-risk patients, and PSA persistence has the biggest impact on the outcomes in high-risk PCa patients. Patient stratiﬁcation according to unfavorable cancer features is essential for the correct interpretation of PSA persistence. The current study is not devoid of limitations. The retrospective study design and unavailable PSA data within 4–8 weeks after RP may have created a selection bias. The single-center database and relatively low number of ﬁnal events, especially in the low- and intermediate-risk groups, could inﬂuence the signiﬁcance of the presented results. Only men after RP were included into the analysis, while a direct comparison of our ﬁndings with early PSA detection after radiation therapy could conﬁrm the importance of the presented results. 4. Discussion The results of the current study demonstrate the incidence, as well as value, of a persistent PSA increase with a higher PCa-risk group. PSA persistence was associated with a worse 10-year cumulative incidence of BCR, MTS, CSM and OM compared to men with undetectable PSA when analyzing the whole cohort. However, in low-risk patients, a signiﬁcant survival difference was found for BCR; in the intermediate-risk, for BCR and MTS and only in high-risk patients for each outcome. We noted that, in the regression analyses, the predictive probability of PSA persistence was very similar in the high-risk group and in all the study population. Such ﬁndings suggest that the importance of PSA persistence should be interpreted differently in PCa-risk groups, with the highest value of signiﬁcance in cancer with unfavorable features. References 1. Moris, L.; Cumberbatch, M.G.; Van Den Broeck, T.; Gandaglia, G.; Fossati, N.; Kelly, B.; Pal, R.; Briers, E.; Cornford, P.; De Santis, M.; et al. Beneﬁts and Risks of Primary Treatments for High-risk Localized and Locally Advanced Prostate Cancer: An International Multidisciplinary Systematic Review. Eur. Urol. 2020, 77, 614–627. [CrossRef] p y y 2. Joniau, S.G.; Van Baelen, A.A.; Hsu, C.Y.; Van Poppel, H.P. Complications and functional results of prostate cancer. Adv. Urol. 2012, 2012, 706309. [CrossRef] 3. Milonas, D.; Baltrimavicius, R.; Grybas, A.; Gudinaviciene, I.; Trumbeckas, D.; Kincius, M.; Auskalnis, S.; Jievaltas, M. Outcome of surgery in locally advanced pT3a prostate cancer. Cent. Eur. J. Urol. 2011, 64, 209–212. [CrossRef] [PubMed] 4. Mottet, N.; Bellmunt, J.; Bolla, M.; Briers, E.; Cumberbatch, M.G.; De Santis, M.; Fossati, N.; Gross, T.; Henry, A.M.; Joniau, S.; et al. EAU-ESTRO-SIOG Guidelines on Prostate Cancer. Part 1: Screening, Diagnosis, and Local Treatment with Curative Intent. Eur. Urol. 2017, 71, 618–629. [CrossRef] 5. Preisser, F.; Chun, F.K.H.; Pompe, R.S.; Heinze, A.; Salomon, G.; Graefen, M.; Huland, H.; Tilki, D. Persistent Prostate-Speciﬁc Antigen After Radical Prostatectomy and Its Impact on Oncologic Outcomes. Eur. Urol. 2019, 76, 106–114. [CrossRef] 6. Bianchi, L.; Nini, A.; Bianchi, M.; Gandaglia, G.; Fossati, N.; Suardi, N.; Moschini, M.; Dell’Oglio, P.; Schiavina, R.; Montorsi, F.; et al. The Role of Prostate-speciﬁc Antigen Persistence After Radical Prostatectomy for the Prediction of Clinical Progression and Cancer-speciﬁc Mortality in Node-positive Prostate Cancer Patients. Eur. Urol. 2016, 69, 1142–1148. [CrossRef] p y p 7. Kumar, A.; Samavedi, S.; Mouraviev, V.; Bates, A.S.; Coelho, R.F.; Rocco, B.; Patel, V.R. Predictive factors and oncological outcomes of persistently elevated prostate-speciﬁc antigen in patients following robot-assisted radical prostatectomy. J. Robot. Surg. 2017, 11, 37–45. [CrossRef] [PubMed] 8. Bartkowiak, D.; Siegmann, A.; Böhmer, D.; Budach, V.; Wiegel, T. The impact of prostate-speciﬁc antigen persistence after radical prostatectomy on the efﬁcacy of salvage radiotherapy in patients with primary N0 prostate cancer. BJU Int. 2019, 124, 785–791. [CrossRef] [PubMed] 9. Audenet, F.; Seringe, E.; Drouin, S.J.; Comperat, E.; Cussenot, O.; Bitker, M.-O.; Rouprêt, M. Persistently elevated prostate-speciﬁc antigen at six weeks after radical prostatectomy helps in early identiﬁcation of patients who are likely to recur. World J. Urol. 2012, 30, 239–244. [CrossRef] 10. Wiegel, T.; Bartkowiak, D.; Bottke, D.; Thamm, R.; Hinke, A.; Stockle, M.; Rube, C.; Semjonow, A.; Wirth, M.; Storkel, S.; et al. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. 5. Conclusions Persistent PSA could be used as an independent predictor of worse long-term out- comes in the high-risk PCa patients, while in the intermediate-risk patients this parameter Cancers 2021, 13, 2453 12 of 13 12 of 13 signiﬁcantly predicts only biochemical recurrence and has no impact on outcomes in the low-risk PCa patients. Author Contributions: Conceptualization: D.M.; data curation D.M., G.S. and Z.V.; formal analysis T.R.; writing—original draft preparation, D.M. and writing—review and editing D.M., G.S., T.R. and Z.V. 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: This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the University Ethics Committee (Nr. BEC-MF-270 on 20 February 2020). Institutional Review Board Statement: This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the University Ethics Committee (Nr. BEC-MF-270 on 20 February 2020). Informed Consent Statement: Patient consent was waived due to some access restrictions applied to the data underlying the ﬁndings. Informed Consent Statement: Patient consent was waived due to some access restrictions applied to the data underlying the ﬁndings. Data Availability Statement: The data presented in this study are available within the article. y g y 14. Kimura, S.; Urabe, F.; Sasaki, H.; Kimura, T.; Miki, K.; Egawa, S. Prognostic Signiﬁcance of Prostate-Spe after Radical Prostatectomy: A Systematic Review and Meta-Analysis. Cancers 2021, 13, 948. [CrossRe References Prostate-speciﬁc antigen persistence after radical prostatectomy as a predictive factor of clinical relapse-free survival and overall survival: 10-year data of the ARO 96-02 trial. Int. J. Radiat. Oncol. Biol. Phys. 2015, 91, 288–294. [CrossRef] [PubMed] ta of the ARO 96-02 trial. Int. J. Radiat. Oncol. Biol. Phys. 2015, 91, 288–294. [CrossRef] [PubMed] 11. Ploussard, G.; Staerman, F.; Pierrevelcin, J.; Saad, R.; Beauval, J.B.; Roupret, M.; Audenet, F.; Peyromaure, M.; Delongchamps, N.B.; Vincendeau, S.; et al. Predictive factors of oncologic outcomes in patients who do not achieve undetectable prostate speciﬁc antigen after radical prostatectomy. J. Urol. 2013, 190, 1750–1756. [CrossRef] [PubMed] g p y 12. Fossati, N.; Karnes, R.J.; Colicchia, M.; Boorjian, S.A.; Bossi, A.; Seisen, T.; Di Muzio, N.; Cozzarini, C.; Noris Chiorda, B.; Fiorino, C.; et al. Impact of Early Salvage Radiation Therapy in Patients with Persistently Elevated or Rising Prostate-speciﬁc Antigen After Radical Prostatectomy. Eur. Urol. 2018, 73, 436–444. [CrossRef] g y , , [ ] 13. Gandaglia, G.; Boorjian, S.A.; Parker, W.P.; Zaffuto, E.; Fossati, N.; Bandini, M.; Dell’Oglio, P.; Suardi, N.; Montorsi, F.; Karnes, R.J.; et al. Impact of Postoperative Radiotherapy in Men with Persistently Elevated Prostate-speciﬁc Antigen After Radical Prostatectomy for Prostate Cancer: A Long-term Survival Analysis. Eur. Urol. 2017, 72, 910–917. [CrossRef] [PubMed] y g y , , [ ] [ ] 14. Kimura, S.; Urabe, F.; Sasaki, H.; Kimura, T.; Miki, K.; Egawa, S. Prognostic Signiﬁcance of Prostate-Speciﬁc Antigen Persistence after Radical Prostatectomy: A Systematic Review and Meta-Analysis. Cancers 2021, 13, 948. [CrossRef] 13 of 13 13 of 13 Cancers 2021, 13, 2453 15. Ploussard, G.; Fossati, N.; Wiegel, T.; D’Amico, A.; Hofman, M.S.; Gillessen, S.; Mottet, N.; Joniau, S.; Spratt, D.E. Manage- ment of Persistently Elevated Prostate-speciﬁc Antigen After Radical Prostatectomy: A Systematic Review of the Literature. Eur. Urol. Oncol. 2021, 4, 150–169. [CrossRef] 16. Epstein, J.I.; Allsbrook, W.C., Jr.; Amin, M.B.; Egevad, L.L.; Committee, I.G. The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am. J. Surg. Pathol. 2005, 29, 1228–1242. [CrossRef] [PubMed] d, L.; Amin, M.B.; Delahunt, B.; Srigley, J.R.; Humphrey, P.A.; Grading, C. The 2014 International Society of 17. Epstein, J.I.; Egevad, L.; Amin, M.B.; Delahunt, B.; Srigley, J.R.; Humphrey, P.A.; Grading, C. The 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma: Deﬁnition of Grading Patterns and Proposal for a New Grading System. Am. J. References Surg. Pathol. 2016, 40, 244–252. [CrossRef] 18. Van den Broeck, T.; van den Bergh, R.C.N.; Arﬁ, N.; Gross, T.; Moris, L.; Briers, E.; Cumberbatch, M.; De Santis, M.; Tilki, D.; Fanti, S.; et al. Prognostic Value of Biochemical Recurrence Following Treatment with Curative Intent for Prostate Cancer: A Systematic Review. Eur. Urol. 2019, 75, 967–987. [CrossRef] y 19. Fossati, N.; Karnes, R.J.; Cozzarini, C.; Fiorino, C.; Gandaglia, G.; Joniau, S.; Boorjian, S.A.; Goldner, G.; Hinkelbein, W.; Haustermans, K.; et al. Assessing the Optimal Timing for Early Salvage Radiation Therapy in Patients with Prostate-speciﬁc Antigen Rise After Radical Prostatectomy. Eur. Urol. 2016, 69, 728–733. [CrossRef] 20. Joniau, S.; Briganti, A.; Gontero, P.; Gandaglia, G.; Tosco, L.; Fieuws, S.; Tombal, B.; Marchioro, G.; Walz, J.; Kneitz, B.; et al. Stratiﬁcation of high-risk prostate cancer into prognostic categories: A European multi-institutional study. Eur. Urol. 2015, 67, 157–164. [CrossRef] 21. Moris, L.; Devos, G.; Van den Broeck, T.; Milonas, D.; Maarten, M.; Berghen, C.; Meerleer, G.; Devlies, W.; Everaerts, W.; Gevaert, T.; et al. Current and emerging therapies for localized high-risk prostate cancer. Expert Rev. Anticancer Ther. 2020, 21, 267–282. [CrossRef] [PubMed] 22. Bryant, A.K.; D’Amico, A.V.; Nguyen, P.L.; Einck, J.P.; Kane, C.J.; McKay, R.R.; Simpson, D.R.; Mundt, A.J.; Murphy, J.D.; Rose, B.S. Three-month posttreatment prostate-speciﬁc antigen level as a biomarker of treatment response in patients with intermediate-risk or high-risk prostate cancer treated with androgen deprivation therapy and radiotherapy. Cancer 2018, 124, 2939–2947. [CrossRef] [PubMed]
https://openalex.org/W3097033630	https://pure.knaw.nl/ws/files/320667537/7286_Crusciol.pdf	English	null	Nitrogen-Fertilized Systems of Maize Intercropped With Tropical Grasses for Enhanced Yields and Estimated Land Use and Meat Production	Frontiers in sustainable food systems	2,020	cc-by	11,814	Royal Netherlands Academy of Arts and Sciences (KNAW) Link to publication in KNAW Research Portal citation for published version (APA) Crusciol, C. A. C., Mateus, G. P., Momesso, L., Pariz, C. M., Castilhos, A. M., Calonego, J. C., Borghi, E., Costa, C., Franzluebbers, A. J., & Cantarella, H. (2020). Nitrogen-Fertilized Systems of Maize Intercropped With Tropical Grasses for Enhanced Yields and Estimated Land Use and Meat Production. Frontiers In Sustainable Food Systems, 4, Article 544853. https://doi.org/10.3389/fsufs.2020.544853 General rights Copyright and moral rights for the publications made accessible in the public 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. General rights C i ht d ghts for the publications made accessible in the public portal are retained by the authors and/or other copyright owners accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the KNAW public portal for the purpose of pr research s may download and print one copy of any publication from the KNAW public portal for the purpose of private study or search. • You may not further distribute the material or use it for any profit-making activity or commercial gain. • You may freely distribute the URL identifying the publication in the KNAW public portal. You may not further distribute the material or use it for any profit making activity or commercial ga • You may freely distribute the URL identifying the publication in the KNAW public portal. 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. y this document breaches copyright please contact us providing details, and we will remove access to the work immediate ur claim Nitrogen-Fertilized Systems of Maize Intercropped With Tropical Grasses for Enhanced Yields and Estimated Land Use and Meat Production Carlos A. C. Crusciol 1, Gustavo P. Mateus 2, Letusa Momesso 1,3, Cristiano M. Pariz 4, André M. Castilhos 4, Juliano C. Calonego 1, Emerson Borghi 5†, Ciniro Costa 4, Alan J. Franzluebbers 6 and Heitor Cantarella 7 1 Department of Crop Science, College of Agricultural Science, UNESP, São Paulo State University, Botucatu, Brazil, 2 Department of Development Decentralization, Agência Paulista de Tecnologia dos Agronegócios, São Paulo Agency of Agribusiness Technology, Andradina, Brazil, 3 Department of Microbial Ecology, NIOO-KNAW, Netherlands Institute of Ecology, Wageningen, Netherlands, 4 Department of Animal Nutrition and Breeding, School of Veterinary Medicine and Animal Science, UNESP, São Paulo State University, Botucatu, Brazil, 5 College of Agricultural Science, São Paulo State University, Botucatu, Brazil, 6 USDA, Agricultural Research Service, NCSU Campus, Raleigh, NC, United States, 7 IAC, Instituto Agronômico de Campinas, Soils and Environmental Resources Center, Campinas, Brazil Keywords: Brachiaria brizantha, Megathyrsus maximus, Zea mays L., tropical agriculture, intercropping grasses, no-tillage system Reviewed by: Crusciol carlos.crusciol@unesp.br Letusa Momesso letusamomesso@gmail.com †Present address: Emerson Borghi, EMBRAPA, Brazilian Agricultural Research Corporation, Corn and Sorghum Research Center, Sete Lagoas, Brazil Specialty section: This article was submitted to Crop Biology and Sustainability, a section of the journal Frontiers in Sustainable Food Systems Received: 23 March 2020 Accepted: 09 October 2020 Published: 05 November 2020 Citation: Crusciol CAC, Mateus GP, Momesso L, Pariz CM, Castilhos AM, Calonego JC, Borghi E, Costa C, Franzluebbers AJ and Cantarella H (2020) Nitrogen-Fertilized Systems of Maize Intercropped With Tropical Grasses for Enhanced Yields and Estimated Land Use and Meat Production. Front. Sustain. Food Syst. 4:544853. doi: 10.3389/fsufs.2020.544853 Specialty section: This article was submitted to Crop Biology and Sustainability, a section of the journal Frontiers in Sustainable Food Systems Received: 23 March 2020 Accepted: 09 October 2020 Published: 05 November 2020 Reviewed by: Reviewed by: Davey Jones, Bangor University, United Kingdom Muhammad Ali Raza, Sichuan Agricultural University, China Intercropping grain with forage crops bridges the gap between agriculture and sustainability. In tropical regions, forage grasses are increasingly being adopted as winter pasture intercropped and in rotation with maize to maximize food production. However, current recommendations for nitrogen (N) fertilizer application are based on monocropped maize (Zea mays), and the best N management approach for intercropping systems remains unclear. A ﬁeld experiment was carried out in three growing seasons with three intercropping systems [monoculture maize, intercropped with palisadegrass (Urochloa brizantha), and intercropped with guineagrass (Megathyrus maximus)] combined with six different split applications of N to maize (0–0, 100–0, 70–30, 50–50, 30–70, and 0–100 kg N ha−1 at seeding-sidedressing) with four replicates. We measured dry matter (DM) and accumulated N in maize and forage grasses, as well as maize production components and yields. Additionally, land equivalent ratio, relative crowding coefﬁcient, aggressivity of maize with forage grasses, forage crude protein (CP) concentration, estimated animal stocking rate, and estimated meat production and economic outcomes. Greatest maize yield was 8.7 Mg ha−1 for monocropped maize. However, favorable maize yield was also obtained in intercropping systems. Although no difference was observed between intercropping systems, applying all N at sidedressing of maize negatively affected maize and forage yields and, consequently, land use and economic evaluation. For both intercropping systems, estimated meat and land use were 114 and 10% higher when N fertilizer was applied than the control (0–0 kg N ha−1), on average. Maize-forage grass intercropping is a viable alternative production system for improving yields and land use. In addition, estimated meat production and revenue can be enhanced with palisadegrass or guineagrass. At least half of the N fertilizer must be applied early in the growing season of maize to maximize production of the entire system. *Correspondence: Carlos A. C. E-mail address: pure@knaw.nl E-mail address: pure@knaw.nl E-mail address: pure@knaw.nl Download date: 24. Oct. 2024 ORIGINAL RESEARCH published: 05 November 2020 doi: 10.3389/fsufs.2020.544853 Edited by: Edited by: Paulo Mazzafera, Campinas State University, Brazil INTRODUCTION N fertilizer recommendations for intercropping systems with maize-grasses have not been adequately studied for crop yields and meat production. Although agricultural models of possible N fertilizer application in ICLS based on N rates have been documented (Borghi et al., 2014; Mateus et al., 2020), there is a lack of information on how to achieve maximum potential of intercropping systems through fertilizer management. Nitrogen recommendations need to be tested based on suitable application timing. Managing N fertilizer in intercropping systems by dividing the rate into two application timings may promote greater N uptake and yield of maize, as has been shown for intercropped sorghum and forage grasses (Mateus et al., 2016). In addition, split N application may provide suﬃcient N for the high N demand of maize and forage, thus tightening the N cycle and minimizing environmental pollution. Intensive use of agricultural land is a global concern. The challenge of agricultural systems is to increase crop and food production, while reducing land use. A new commercial practice of intercropping grain and forage crops bridges the gap between agriculture and environmental sustainability (Mateus et al., 2016; Martin-Guay et al., 2018). Additionally, fertilizer management, such as N application, beneﬁts these intercropping systems by enhancing yields and minimizing plant competition. However, the appropriate time for N management in intercropping systems remains incompletely studied. A new approach of maize-grass intercropping systems for crop and meat production has been suggested since intercropping with forages results in yield improvements and in satisfactory socioeconomic outcomes for integrated crop-livestock systems (ICLS) with a no-tillage system (NTS) (Derpsch and Friedrich, 2009; Himmelstein et al., 2017; Pariz et al., 2017a). While studies have shown reduction of soil erosion and degradation, stimulation of root growth and increase of forage yields in intercropping systems in relation to sole-cropping systems (Pariz et al., 2017b; Moraes et al., 2019), the potential food supply for livestock and farmers’ proﬁtability do not appear to have been investigated in maize-grasses intercropping systems with proper N management. The aim of this study was to evaluate the eﬀects of split N application to maize- forage intercropping systems on crop yield, land equivalent ratio (LER), crop competition, estimated meat, and overall system revenue. Citation: Crusciol CAC, Mateus GP, Momesso L, Pariz CM, Castilhos AM, Calonego JC, Borghi E, Costa C, Franzluebbers AJ and Cantarella H (2020) Nitrogen-Fertilized Systems of Maize Intercropped With Tropical Grasses for Enhanced Yields and Estimated Land Use and Meat Production. Front. Sustain. Food Syst. 4:544853. doi: 10.3389/fsufs.2020.544853 November 2020 \| Volume 4 \| Article 544853 Frontiers in Sustainable Food Systems \| www.frontiersin.org 1 Nitrogen Management in Intercropping Systems Crusciol et al. INTRODUCTION We hypothesized that N management applied at maize seeding and at sidedressing of maize at V5 growth stage, i.e., the initiation of maize ear development, would (i) increase yields of maize and forage grasses, (ii) increase eﬃciency land use and estimated meat production, (iii) decrease competition between intercropped crops, and (iv) provide high revenue. To test these hypotheses, we used the same N rate of 100 kg ha−1 divided into two applications (seeding + sidedressing) at diﬀerent ratios for maize intercropped with palisadegrass and guineagrass. Intercropping tropical forages and cash crops is an alternative for farmers to develop temporary pasture using ICLS combined with NTS (Pariz et al., 2017b). With ICLS, food production potential (meat and grains) can be enhanced on the same land area and thus limit deforestation of new agricultural areas (Moraes et al., 2019). In the tropical region, forage grasses are being increasingly adopted in ICLS under NTS for winter pasture to maximize system production (Crusciol et al., 2015; Pariz et al., 2016, 2017a,b). Palisadegrass [Urochloa brizantha (=syn. Brachiaria)] and guineagrass [Megathyrsus maximus (=syn. Panicum maximum)] has been suitable species for intercropping with cash crops (Costa et al., 2015; Mateus et al., 2016; Pariz et al., 2016). Importantly, this strategy of intercropping forages with grain crops enhances the success of forage production in the dry winter season with low and irregular rainfall (Borghi et al., 2013a). Therefore, more reliable forage biomass production raises the protein concentration and potential meat production by animals grazing fodder in ICLS (Crusciol et al., 2012, 2014; Moraes et al., 2019). MATERIALS AND METHODS Site Description and Experimental Design A ﬁeld experiment was carried out during three growing seasons (2004–2005, 2005–2006, 2006–2007) in Botucatu, São Paulo, Brazil (48◦26′W, 22◦51′S, 740 m above sea level). The climate is Cwa, i.e., tropical with dry winter and warm, rainy summers, according to the Köppen classiﬁcation. Mean annual precipitation is 1,358 mm and mean annual temperature is 20.7◦C. Precipitation and temperature during the experiment are shown in Figure 1. The soil type was a clayey, kaolinitic, thermic Typic Haplorthox [United States Department of Agriculture (USDA), 2014] with 630, 90, and 280 g kg−1 of clay, silt, and sand, respectively. At the beginning of the experiment, selected chemical properties were determined according to methodology proposed by van Raij et al. (2001) and are shown in Table 1. The soil pH was determined in a 0.01 mol L−1 CaCl2 suspension (1:2:5 soil:solution). Soil organic matter was determined by chromic acid digestion (Heanes, 1984). The total acidity at pH 7.0 (H+Al) was extracted by calcium acetate (0.5 mol L−1 at pH 7.0) and evaluated by titration with 0.025 mol L−1 NaOH solution. The available P and exchangeable basic cations (K+, Ca2+, and Mg2+) Maize-forage grass intercropping has increased as cultivation practice (Sulc and Tracy, 2007; Tracy and Zhang, 2008; Moraes et al., 2019). Intercropping grasses with maize improves soil quality and increases soil organic C and N stocks by promoting deep root systems and better nutrient retention compared with monocrops (Costa et al., 2012, 2015; Cong et al., 2015). Because of the potential of cycling N from soil by plant N uptake and consequent high straw decomposition, providing diversity of residues, and nutrient back to the soil (Pariz et al., 2017b; Martin- Guay et al., 2018). However, maize and grass may compete for N sources at the vegetative growth stages in intercropping systems, since grasses can immobilize N by microbial processes and increase the dependence on N fertilizer for crop yields (Pariz et al., 2011; Mateus et al., 2016), especially during the ﬁrst several years of cultivation in NTS with accumulation of soil organic matter. In addition, N demand by maize is high during early- to mid-season growth (Anghinoni, 2007; Borghi et al., 2014; Garcia et al., 2016). MATERIALS AND METHODS Current recommendations for N application are based on maize monocropping (Cantarella et al., 1997), however, the November 2020 \| Volume 4 \| Article 544853 Frontiers in Sustainable Food Systems \| www.frontiersin.org 2 Crusciol et al. Nitrogen Management in Intercropping Systems FIGURE 1 \| Monthly rainfall, maximum and minimum temperatures during 2004–2005 (A), 2005–2006 (B), and 2006–2007 (C) growing seasons. Nitrogen Management in Intercropping Systems Crusciol et al. FIGURE 1 \| Monthly rainfall, maximum and minimum temperatures during 2004–2005 (A), 2005–2006 (B), and 2006–2007 (C) growing seasons. 2001). The experimental area had been cropped under NTS since 1999 and the historical crop rotation is presented in Supplementary Table 1. were extracted using an ion resin. The Presin concentration was determined colorimetrically (Murphy and Riley, 1962) with a FEMTO 600S spectrophotometer. Exchangeable K+, Ca2+, and Mg2+ in the extracts were determined by an atomic absorption/ﬂame-emission spectrophotometer (Shimadzu AA- 6300). The cation exchange capacity (CEC) was obtained by summing the individual cations (H, Al, K, Ca, and Mg). The base saturation (BS) values were calculated using equivalents exchangeable bases and total acidity results (van Raij et al., The experimental design was a randomized complete block with four replicates on diﬀerent parcels of the same ﬁeld each year. Treatments consisted of monocropped maize, maize intercropped with palisadegrass [Urochloa brizantha (Hochst. Ex A. Rich) R. Webster “Marandu”], and maize intercropped with guineagrass [Megathyrsus maximus (Jacq.) B. K. Simon and S. Frontiers in Sustainable Food Systems \| www.frontiersin.org November 2020 \| Volume 4 \| Article 544853 3 Nitrogen Management in Intercropping Systems Crusciol et al. TABLE 1 \| Soil chemical characteristics at two depths in the experimental areas before initiating the experiment (n = 8). Growing season Depth pH (CaCl2) SOM† (g dm−3) P(resin) (mg dm−3) H+Al (mmolc dm−3) K+ (mmolc dm−3) Ca2+ (mmolc dm−3) SO2− 4 (mmolc dm−3) Mg2+ (mmolc dm−3) CEC‡ (mmolc dm−3) BS§ (%) 2004/2005 0.00–0.20 m 4.7 25 14 52 1.3 20 4.7 10 83 39 0.20–0.40 m 4.4 22 8 76 0.7 15 9.8 8 99 24 2005/2006 0.00–0.20 m 4.5 24 14 49 1.7 19 5.1 9 82 39 0.20–0.40 m 4.3 21 7 73 0.7 11 10.3 8 92 23 2006/2007 0.00–0.20 m 4.8 26 15 47 1.6 18 4.4 12 78 41 0.20–0.40 m 4.6 24 9 66 1.0 14 9.5 9 90 27 †Soil organic matter. ‡Cation exchange capacity. §Base saturation. W. L. Crop Management Soil acidity was ameliorated with dolomite lime application over the soil surface, without soil incorporation. Lime rate was calculated to increase soil base saturation of the surface 0.20 m of soil to 70% (Cantarella et al., 1997) and was applied at concentrations of 3.05, 2.95, and 2.66 Mg ha−1 in August 2004, August 2005 and August 2006, respectively. Dolomitic lime consisted of 400 kg CaO ha−1 and 120 kg MgO ha−1, with 85% eﬀective calcium carbonate equivalence. Pearl millet (Pennisetum glaucum) was sown on 2 Oct. 2004, 5 Oct. 2005, and 3 Oct. 2006 at 0.3-m depth using a no-till drill at a seed density of 20 kg ha−1 to produce crop residues for the ICLS prior to maize in a short-term cultivation. Pearl millet was terminated with glyphosate (1.8 kg ha−1 acid-equivalent), using a spray volume of 250 L ha−1 20 days before maize sowing. Maize (hybrid 30F90) was sown on 15 Dec. 2004, 18 Dec. 2005 and 20 Dec. 2006 at a depth of 0.3 m and a density of 60,000 seeds ha−1 using a no-till drill. Each plot consisted of ten 20-m-long rows of maize and row spacing of 0.45 m. Sampling area was considered within a buﬀer zone of 0.45 m from the perimeter of each plot. MATERIALS AND METHODS Jacobs “Mombaça”] factorially arranged with N applied at seeding and sidedressing of maize: (i) 0–0 (control), (ii) 100– 0, (iii) 70–30, (iv) 50–50, (v) 30–70, and (vi) 0–100 kg N ha−1, respectively (Figures 2A,B). The rate of 100 kg N ha−1 was based on current recommendation and studies in intercropping systems (Cantarella et al., 1997; Mateus et al., 2020). The relatively low rate aimed to reduce environmental impacts from N loss; however, there is currently no speciﬁc recommendation of N fertilizer application for intercropping systems. same time using the same practices. In addition, monocropped palisade grass and guinea grass were seeded at the same time as the forages in intercropping systems using the same practices. The monocropped forages plots were the same size and were only used to calculate the intercropping competition factors. Maize seedlings emerged 5 days after sowing (20 Dec. 2004, 23 Dec. 2005 and 25 Dec. 2006) and forage seedlings emerged 15 days after sowing, on average, for each growing season. Maize and forage were cultivated according to crop needs; atrazine [6-chloro-N2-ethyl-N4-isopropyl-1,3,5-triazine-2,4-diamine] (1.0 kg ha−1 acid-equivalent) using a spray volume of 200 L ha−1 was applied to control emergence of annual broadleaf weeds, deltamethrin [(S)-cyano-(3-phenoxyphenyl)-methyl] (1R,3R)-3-(2,2-dibromoethenyl)-2,2-dimethyl-cyclopropane- 1-carboxylate (5 g ha−1 active ingredient) was used against fall armyworm (Spodoptera frugiperda). Sidedress N fertilization was applied according to the treatments at V5 maize growth stage (ﬁve expanded leaves). Physiological maturity averaged 128, 132, and 130 days after emergence in the 2004–2005, 2005–2006, and 2006–2007, respectively. Maize harvest was 7 days after physiological maturity using a mechanical harvester. Maize, palisadegrass and guineagrass were harvested separately from eight central rows. Frontiers in Sustainable Food Systems \| www.frontiersin.org Sampling and Analyses When 50% of maize plants were in full ﬂowering stage, 20 random leaf samples per plot were collected from the fourth leaf with visible sheath from the apex for nutrition diagnoses (Cantarella et al., 1997). Leaves were washed, dried in forced air circulation at 65◦C for 72 h, ground, and N, P, K, Ca, Mg, and S concentrations in leaves were determined according to Malavolta et al. (1997). The samples were digested with sulfuric acid for N determination and with a nitro-perchloric solution for the other nutrients. The leaf N, P, S concentrations were determined by semi-micro-Kjeldahl distillation, colorimetry, and turbidmetry methods, respectively. The leaf K, Ca, and Mg concentrations were determined by atomic absorption spectrophotometry. Baseline fertilization of maize in the sowing furrows consisted of 84 kg ha−1 P2O5 as triple superphosphate and 48 kg ha−1 K2O as potassium chloride in a 08–28–16 NPK formula for all treatments. At seeding, N application treatments were applied as urea and distributed between 0.5 and 0.10 m next to the seed row by superﬁcial broadcasting. For treatments with intercropping, palisadegrass and guineagrass were simultaneously sown with maize at densities of 15.3 and 15.9 kg ha−1 seed (34% viable seeds), respectively. Palisadegrass and guineagrass were mixed with fertilizer and sown at depths of 0.08 and 0.06 m below soil surface, respectively. Monocropped maize was sown at the Kernel weight was determined and transformed to maize yield ha−1 by correcting to 13% grain moisture. Plant population was determined by counting the number of plants in the four central 5-m rows per plot at harvest. Number of ears per plant, number of November 2020 \| Volume 4 \| Article 544853 4 Nitrogen Management in Intercropping Systems Crusciol et al. FIGURE 2 \| (A) Scheme of cropping systems arrange in the summer season. Grazing by animals was not performed after maize harvest in off season and meat production was estimated using Large Ruminant Nutrition System model. (B) Timing of N application (seeding + sidedressing) in the monocrop and intercrop system in the summer season. FIGURE 2 \| (A) Scheme of cropping systems arrange in the summer season. Grazing by animals was not performed after maize harvest in off season and meat production was estimated using Large Ruminant Nutrition System model. Sampling and Analyses (B) Timing of N application (seeding + sidedressing) in the monocrop and intercrop system i th FIGURE 2 \| (A) Scheme of cropping systems arrange in the summer season. Grazing by animals was not performed after maize harvest in off season and meat production was estimated using Large Ruminant Nutrition System model. (B) Timing of N application (seeding + sidedressing) in the monocrop and intercrop system in the summer season. Intercropping Competition Factors kernels per ear, and 100-kernel weight were determined at harvest and evaluated from 10 plants per plot chosen at random. To study the competition eﬀects between crops and to evaluate intercrop performance, diﬀerent competition functions were calculated: land equivalent ratio (LER), relative crowding (K), and aggressivity index (A). The LER was used to evaluate the land use advantage provided by intercropping (Mead and Willey, 1980): From the time of maize harvest, forage dry matter of palisadegrass and guineagrass were evaluated at 55 days (ﬁrst cut) and 145 days (second cut), in June and September, respectively. Forages were cut at 0.25 m from the soil surface (2 m2 area each area and row spacing = 0.45 m) and removed from the plots. The remainder of plots were cut using a manual mechanical rotary mower to provide faster forage regrowth. Forage dry matter was dried by forced-air circulation at 65◦C for 72 h until constant weight, and weighed. Data were extrapolated to Mg ha−1. A sub-sample of forage dry matter was used to determine total N concentration for crude protein (CP). CP was calculated by formula: CP (%) = total N (%) × 6.25 (Horwitz, 1980). Frontiers in Sustainable Food Systems \| www.frontiersin.org LER = Y1, 2/Y1, 1 + Y2, 1/Y2, 2 LER = Y1, 2/Y1, 1 + Y2, 1/Y2, 2 where Y is the aboveground biomass of crops, and suﬃxes 1 and 2 denote the crops: (1) maize and (2) palisadegrass or guineagrass. Therefore, Y1,2 is the aboveground biomass of maize when grown in a mixture with grasses, Y1,1 is the yield of maize Frontiers in Sustainable Food Systems \| www.frontiersin.org November 2020 \| Volume 4 \| Article 544853 5 Nitrogen Management in Intercropping Systems Crusciol et al. when grown in a monoculture, Y2,1 is the aboveground biomass of the forage (palisadegrass or guineagrass) when grown in a mixture with maize, and Y2,2 is the aboveground biomass of the forage (palisadegrass or guineagrass produced 839 and 1,327 Mg ha−1, respectively) when grown in a monoculture. metabolizable energy and protein gain, since CP of forage was 9.3–14.6%. A animal grazing time was calculated using a method similar to Crusciol et al. (2012), in which a 55 d forage accumulation period occurred after maize harvest followed by two 60-day grazing periods with a 30-day rest period in between grazing periods. Stocking rate was estimated from forage dry matter production, time of animal grazing (days per cut), dry matter intake, and grazing eﬃciency. Total cattle meat produced per hectare was calculated from stocking rate multiplied by the components of ADG, time of animal grazing, and carcass yield (52%). Relative crowding coeﬃcient (K values) is a measure of plant competition theory as an index of the relative competitive abilities between plants in an intercropping system to evaluate and compare the competitive ability of one species to another in a mixture (Zhang et al., 2011). K was calculated according the method of Agegnehu et al. (2006) as follows: Gross revenue ha−1 was calculated by the formula: (price per kg × maize yield) + (price per kg × estimated meat production). Net return per ha was calculated by the formula: (gross revenue – cost ha−1). The Brazilian national average price used was from the last 5 years and values were converted to US$ (Agrolink, 2018). (K)maize = Y1, 2 × Z2, 1 / (Y1, 1 −Y1, 2) × Z1, 2 or (K)forage = Y2, 1 × Z1, 2 / (Y2, 2 −Y2, 1) × Z2, 1 where Y and suﬃxes 1 and 2 denote as described for LER, Z1,2 is the sown proportion of maize, and Z2,1 is the sown proportion of the forage species. LER = Y1, 2/Y1, 1 + Y2, 1/Y2, 2 For this calculation, the plant density of each species was evaluated on the day of maize harvest. Greater K value of one species indicates it is more competitive and dominant than another species in the intercropping system (Li et al., 1999; Wahla et al., 2009). Statistical Analyses All data were initially tested for normality using the Shapiro- Wilk test from the UNIVARIATE procedure using the statistical software R (version 3.5.2) with the package “agricolae” (Mendiburu, 2015). All data were distributed normally (W ≥0.90). Cropping systems, N management treatments, and their interactions were considered ﬁxed eﬀects. Growing season and its interaction with cropping systems and N managements were not signiﬁcant at P < 0.05 for any of the dependent variables. Thus, data were combined across growing seasons. Block was considered a random variable. Analysis of variance (ANOVA) was performed and if the null hypothesis was rejected, means were compared using LSD teste (P ≤0.05). Aggressivity index (A) was calculated to determine relative yield of crop 1 with crop 2 in intercropping (Takim, 2012): (A)maize = (Y1, 2/Y1, 1Y1, 2) −(Y2, 1/Y2, 2Y2, 1) or (A)forage = (Y2, 1/Y2, 2Y2, 1) −(Y1, 2/Y1, 1Y1, 2) where Y and the suﬃxes 1 and 2 denote the same as used in LER and (K). If (A)maize = 0, crops were equally competitive, if (A)forage was negative, then maize dominated, if (A)forage was positive, then forage dominated. Frontiers in Sustainable Food Systems \| www.frontiersin.org Economic Valuation and Estimated Meat Production Plant Nutrition, Agronomic Characteristics, Kernels and Production Attributes of Maize Monocropped maize and maize intercropped with palisadegrass had greater leaf N concentrations than maize intercropped with guineagrass (Supplementary Table 2). Maize intercropped with palisadegrass had greater leaf P, K, and S concentrations than monocropped maize and maize intercropped with guineagrass. Although the interaction of intercropping system and N management was not statistically signiﬁcant for nutrient concentration, all intercropping systems that received N application had greater leaf N, P, and S concentrations than the control without N fertilizer (Supplementary Table 2). Production costs per hectare of monocropped maize and maize intercropped with forages were estimated (CONAB, 2018). Diﬀerences in input costs were forage seed and N fertilizer, as sowing maize monocrop and intercropped forage seeds were the same process. Maize grain yield (kg ha−1) was calculated and multiplied by the value per kg. Although grazing by animals was not carried out for the palisadegrass and guineagrass after maize grain harvest, meat production was calculated using Large Ruminant Nutrition System (LRNS; http://nutritionmodels.tamu.edu/lrns.html) model to estimate grazing performance by animals on tropical perennial grasses. The LRNS model is based on the Net Carbohydrate and Protein System (CNCPS), version 5 (Fox et al., 2004). Energy and protein requirements, performance and dry matter intake by each individual cattle fed in a group were predicted for continuously grazed 450 kg Nellore bulls with 52% carcass yield and 22% Body Fat Grading System. Performance values were predicted from the nutritional composition of palisadegrass and guineagrass and N fertilizer applied. Intercropping system did not inﬂuence plant population, ears per plant, kernels per ear, and 100-kernel weight (Table 2). However, monocropped maize had greater shoot dry matter and grain yield compared with intercropping systems of maize with palisadegrass and guineagrass. Time of N application did not inﬂuence plant population, but all treatments with some N application led to greater number of ears per plant, number of kernels per ear, 100-kernel weight, shoot dry matter, and grain yield of maize compared to the control without N application (Table 2). Shoot dry matter and grain yield of maize were greater in all N management systems with some N applied at seeding Dry matter intake by each individual cattle fed in a group was 9.9–10.0 kg of dry matter day−1. Economic Valuation and Estimated Meat Production Average daily gain (ADG) was used to estimate meat production based on the allowable November 2020 \| Volume 4 \| Article 544853 6 Nitrogen Management in Intercropping Systems Crusciol et al. TABLE 2 \| Agronomic characteristics (plant population and number of ears per plant), kernels attributes (number of kernels per ear and 100-kernel weight) and production attributes (shoot dry matter and grain yield) of maize as affected by intercropping system, N management in the three growing seasons. Agronomic characteristics Kernel attributes Production attributes Treatment Plant population Ears per plant Kernels per ear 100-kernel weight Shoot dry matter Grain Yield Thousand plants ha−1 no. no. g Mg ha−1 Mg ha−1 Intercropping system (IC) Monocropped maize 60.1 a§ 1.13 a 403 a 32 a 17.6 a 8.7 a Maize + palisadegrass 59.7 a 1.14 a 380 a 32 a 16.0 b 8.2 b Maize + guineagrass 59.9 a 1.13 a 370 a 31 a 15.1 b 7.9 b N management (NM)‡ 0–0 59.2 a 0.89 c 333 b 29 b 9.6 c 4.9 c 100–0 60.0 a 1.22 a 381 a 33 a 17.7 a 9.1 a 70–30 60.5 a 1.19 a 399 a 32 a 18.2 a 9.3 a 50–50 59.7 a 1.20 a 404 a 33 a 18.4 a 9.4 a 30–70 60.2 a 1.20 a 396 a 32 a 17.7 a 9.1 a 0–100 59.7 a 1.10 b 391 a 32 a 15.7 b 8.0 b F probability IC 0.313 0.109 0.074 0.563 <0.001 <0.001 NM 0.549 <0.001 <0.001 <0.001 <0.001 <0.001 IC x NM 1.000 0.083 0.656 0.724 0.532 0.963 §Values followed by the same letter are not signiﬁcantly different at P ≤0.05 (LSD test). ‡A rate of 100 kg N ha−1 applied in two-split management at maize seeding and V5 growth stage sidedressing. TABLE 2 \| Agronomic characteristics (plant population and number of ears per plant), kernels attributes (number of kernels per ear and 100-kernel weight) and production attributes (shoot dry matter and grain yield) of maize as affected by intercropping system, N management in the three growing seasons. (i.e., 100–0, 70–30, 50–50, and 30–70 kg N ha−1) than with no N applied at seeding (i.e., 0–0 and 0–100 kg N ha−1). for maize intercropped with both palisadegrass and guineagrass. The aggressivity index (A) showed that maize was less competitive than palisadegrass and guineagrass in all treatments. Maize was more competitive with guineagrass without N fertilizer application. Revenue Not supplying N fertilizer to cropping systems resulted in the lowest estimated net proﬁt (Table 5). When supplying 100 kg N ha−1, net proﬁt was similar among the diﬀerent split N applications, except when no N fertilizer was applied at seeding (0–100 kg N ha−1), which had lower net proﬁt compared to other treatments with N application. Intercropping maize with either forage grass had greater net proﬁt compared to monocropped maize when supplied with N, because of signiﬁcant meat production during the winter/spring. Forage Characteristics and Estimated Meat Production Forage dry matter production, estimated animal stocking rate and estimated meat production were inﬂuenced by intercropping system in the second cut but not in the ﬁrst cut (Table 3). Forage dry matter production in the second cut was 23% greater when maize was intercropped with palisadegrass than with guineagrass, although CP in the second cut was similar in the two intercropping systems. For both intercropping systems, forage dry matter production, CP, estimated animal stocking rate, and estimated meat production were greater when N fertilizer was applied than in the control (0–0 kg N ha-1) in the ﬁrst and second cuts. Frontiers in Sustainable Food Systems \| www.frontiersin.org Maize and Tropical Forage Grass Responses Treatment Forage DM (Mg ha−1) Crude protein (%) Stocking rate (AU ha−1)¶ Meat production (kg ha−1)Ψ First cut† Second cut† First cut Second cut First cut Second cut First cut Second cut Intercropping system (IC) Maize + palisadegrass 2.2 a§ 5.8 a 12.5 a 12.8 a 2.2 a 5.8 a 53.1 a 140.5 a Maize + guineagrass 2.1 a 4.7 b 12.6 a 13.4 a 2.1 a 4.7 b 49.5 a 118.4 b N management (NM)‡ 0–0 1.4 b 3.1 b 11.2 b 10.1 b 1.4 b 5.7 a 30.0 b 59.0 b 100–0 2.3 a 5.7 a 12.8 a 13.8 a 2.3 a 5.8 a 55.7 a 145.2 a 70–30 2.4 a 5.8 a 12.9 a 13.9 a 2.4 a 5.7 a 57.0 a 147.6 a 50–50 2.3 a 5.7 a 12.5 a 13.7 a 2.3 a 5.7 a 54.5 a 140.0 a 30–70 2.3 a 5.7 a 12.8 a 13.8 a 2.3 a 5.6 a 56.3 a 144.5 a 0–100 2.3 a 5.6 a 12.7 a 13.7 a 2.3 a 5.7 a 54.3 a 140.5 a F probability IC 0.148 <0.001 0.127 0.081 0.198 <0.001 0.102 <0.001 NM <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 IC x NM 0.263 0.222 0.387 0.155 0.265 0.175 0.168 0.365 † TABLE 3 \| Forage dry matter (DM) production and crude protein (CP) concentration, estimated animal stocking rate, and estimated meat production as affected by intercropping systems and N management in three growing seasons and ANOVA signiﬁcance. TABLE 4 \| Land equivalent ratio (LER), relative crowding coefﬁcient (K), and aggressivity (A) of maize, palisadegrass, and guineagrass intercropped as a function of N fertilizer applied for maize crop. Maize and Tropical Forage Grass Responses Treatment LER K A Maize† Forage‡ Total‡ Maize Forage Maize Forage Maize ± palisadegrass 0–0Ψ 0.86§ 0.09 0.95 4.00 0.15 −0.0002994 0.0002994 100–0 0.91 0.16 1.06 6.93 0.27 −0.0007091 0.0007091 70–30 0.91 0.16 1.07 7.27 0.27 −0.0007196 0.0007196 50–50 0.91 0.15 1.06 7.30 0.26 −0.0007247 0.0007247 30–70 0.91 0.16 1.07 7.34 0.26 −0.0007070 0.0007070 100–0 0.91 0.14 1.05 7.24 0.24 −0.0006394 0.0006394 Maize ± guineagrass 0 kg N ha−1 0.82 0.06 0.88 3.19 0.09 −0.0001518 0.0001518 100–0 0.86 0.09 0.95 4.13 0.15 −0.0002579 0.0002579 70–30 0.86 0.09 0.95 4.39 0.14 −0.0002684 0.0002684 50–50 0.86 0.09 0.95 4.51 0.14 −0.0002735 0.0002735 30–70 0.86 0.09 0.95 4.24 0.15 −0.0002558 0.0002558 100–0 0.86 0.10 0.96 4.23 0.15 −0.0001883 0.0001883 Ψ First value means the kg N ha−1 applied at seeding and the second value means the kg ha−1 applied sidedressing at maize V6 growth. †Relative to respective monoculture. ‡Relative to respective intercropping system. §Value above 1 means positive impact. N demand of both crops. Furthermore, this research provides as a novel outcome that intercropping systems combined with fertilizer N management showed eﬀectiveness in improving the overall productivity of the whole system especially for enhancing Competition between forage and maize may have been reduced in this study due to the relatively long growing season with the 130-day maturity maize hybrid (Crusciol et al., 2013). Sowing plants with earlier relative maturity may beneﬁt an TABLE 3 \| Forage dry matter (DM) production and crude protein (CP) concentration, estimated animal stocking rate, and estimated meat production as affected by intercropping systems and N management in three growing seasons and ANOVA signiﬁcance. Maize and Tropical Forage Grass Responses p All LER values of maize and forages were lower under intercropping compared with the respective monoculture (Table 4). When combined, LER of maize intercropped with palisadegrass (1.06 average) was more productive than each component separately when receiving N, independent of the split-N ratio. In contrast, LER of maize intercropped with guineagrass (0.95 average) was less productive than individual components grown separately for all N application conditions. Our study provides a novel alternative identifying potential agricultural systems to improve food production by intercropping maize with forage grasses and selecting the proper N management. Currently, recommendations for N fertilizer application consider only monocropping (maize or forage grass) in the summer/fall or fodder in the winter/spring for grain production (Cantarella et al., 1997). Our study shows that split N application timing can increase responses of both maize and forage grasses, while meeting grain crop requirements and high The intercropping competition factor (K) values shown in Table 4 are the interspeciﬁc competitive abilities. Compared with the unfertilized control, Kmaize and Kforage were greater November 2020 \| Volume 4 \| Article 544853 Frontiers in Sustainable Food Systems \| www.frontiersin.org 7 Nitrogen Management in Intercropping Systems Crusciol et al. Crusciol et al. Nitrogen Management in Intercropping Systems TABLE 3 \| Forage dry matter (DM) production and crude protein (CP) concentration, estimated animal stocking rate, and estimated meat production as affected by intercropping systems and N management in three growing seasons and ANOVA signiﬁcance. Maize and Tropical Forage Grass Responses Treatment Forage DM (Mg ha−1) Crude protein (%) Stocking rate (AU ha−1)¶ Meat production (kg ha−1)Ψ First cut† Second cut† First cut Second cut First cut Second cut First cut Second cut Intercropping system (IC) Maize + palisadegrass 2.2 a§ 5.8 a 12.5 a 12.8 a 2.2 a 5.8 a 53.1 a 140.5 a Maize + guineagrass 2.1 a 4.7 b 12.6 a 13.4 a 2.1 a 4.7 b 49.5 a 118.4 b N management (NM)‡ 0–0 1.4 b 3.1 b 11.2 b 10.1 b 1.4 b 5.7 a 30.0 b 59.0 b 100–0 2.3 a 5.7 a 12.8 a 13.8 a 2.3 a 5.8 a 55.7 a 145.2 a 70–30 2.4 a 5.8 a 12.9 a 13.9 a 2.4 a 5.7 a 57.0 a 147.6 a 50–50 2.3 a 5.7 a 12.5 a 13.7 a 2.3 a 5.7 a 54.5 a 140.0 a 30–70 2.3 a 5.7 a 12.8 a 13.8 a 2.3 a 5.6 a 56.3 a 144.5 a 0–100 2.3 a 5.6 a 12.7 a 13.7 a 2.3 a 5.7 a 54.3 a 140.5 a F probability IC 0.148 <0.001 0.127 0.081 0.198 <0.001 0.102 <0.001 NM <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 IC x NM 0.263 0.222 0.387 0.155 0.265 0.175 0.168 0.365 †First and second cut in June and September, respectively. §Values followed by the same letter are not signiﬁcantly different at P ≤0.05 (LSD test). ‡A rate of 100 kg N ha−1 applied in two-split management at maize seeding and V5 growth stage sidedressing. ¶1 AU (animal unit) = 450 kg of body weight. Ψ Estimated meat production = kg of body weight gain (cattle) per ha (estimated) × 52% of carcass yield. TABLE 4 \| Land equivalent ratio (LER), relative crowding coefﬁcient (K), and aggressivity (A) of maize, palisadegrass, and guineagrass intercropped as a function of N fertilizer applied for maize crop. Maize and Tropical Forage Grass Responses Treatment Cost CY¥ Total maize§ Meat productionα Total meat¶ Gross† Net‡ US$ ha−1 Mg ha−1 US$ ha−1 kg ha−1 US$ ha−1 US$ ha−1 US$ ha−1 Monocropped maize 0–0 604 5.3 1,081 0 0 1,081 477 100–0 643 9.5 1,937 0 0 1,937 1,294 70-30 643 9.7 1,978 0 0 1,978 1,335 50-50 643 9.8 1,998 0 0 1,998 1,355 30–70 643 9.5 1,937 0 0 1,937 1,294 0–100 643 8.4 1,713 0 0 1,713 1,070 Maize ± palisadegrass 0–0 626 4.7 958 91 272 1,230 604 100–0 666 9.1 1,856 214 643 2,499 1,833 70–30 666 9.3 1,896 221 664 2,560 1,894 50–50 666 9.4 1,917 206 618 2,535 1,869 30–70 666 9.1 1,856 218 654 2,510 1,844 0–100 666 8.0 1,631 211 634 2,265 1,599 Maize ± guineagrass 0–0 628 4.7 958 87 262 1,220 592 100–0 667 8.7 1,774 187 562 2,336 1,669 70–30 667 8.9 1,815 188 563 2,378 1,711 50–50 667 9.0 1,835 183 548 2,383 1,716 30–70 667 8.7 1,774 184 551 2,325 1,658 0–100 667 7.7 1,570 178 534 2,104 1,437 Mean costs and production costs of monocropped maize and maize intercropped with palisadegrass or guineagrass; the only difference was the forage seeds cost and sidedress nitrogen used for the maize crop. ¥CY is the maize yield. §Total = kg of maize ha−1 × US$ 0.20. αMeat production = kg of body weight gain (cattle) ha−1 (estimate) × 52% of carcass yield (sum of EMP First and Second cuts). ¶Total meat = meat production × US$ 3.00. †Gross is the revenue per ha, which was calculated using the formula: total maize + total meat. ‡Net is the return per ha, which was calculated using the formula (gross ha−1-cost ha−1). d maize, maize intercropped with palisadegrass and maize intercropped with guineagrass as a function of N management appear to limit dry matter in the early growth stages, which were characterized by high N uptake, implying high eﬃciency in intercepting photosynthetically active radiation (Amaral Filho et al., 2005; Sawyer et al., 2010). Grain yield is positively linked to dry matter accumulation and the supply of N and C to kernels (Kowles and Phillips, 1988). Maize and Tropical Forage Grass Responses Treatment LER K A Maize† Forage‡ Total‡ Maize Forage Maize Forage Maize ± palisadegrass 0–0Ψ 0.86§ 0.09 0.95 4.00 0.15 −0.0002994 0.0002994 100–0 0.91 0.16 1.06 6.93 0.27 −0.0007091 0.0007091 70–30 0.91 0.16 1.07 7.27 0.27 −0.0007196 0.0007196 50–50 0.91 0.15 1.06 7.30 0.26 −0.0007247 0.0007247 30–70 0.91 0.16 1.07 7.34 0.26 −0.0007070 0.0007070 100–0 0.91 0.14 1.05 7.24 0.24 −0.0006394 0.0006394 Maize ± guineagrass 0 kg N ha−1 0.82 0.06 0.88 3.19 0.09 −0.0001518 0.0001518 100–0 0.86 0.09 0.95 4.13 0.15 −0.0002579 0.0002579 70–30 0.86 0.09 0.95 4.39 0.14 −0.0002684 0.0002684 50–50 0.86 0.09 0.95 4.51 0.14 −0.0002735 0.0002735 30–70 0.86 0.09 0.95 4.24 0.15 −0.0002558 0.0002558 100–0 0.86 0.10 0.96 4.23 0.15 −0.0001883 0.0001883 Ψ First value means the kg N ha−1 applied at seeding and the second value means the kg ha−1 applied sidedressing at maize V6 growth. †Relative to respective monoculture. ‡Relative to respective intercropping system. §Value above 1 means positive impact. Competition between forage and maize may have been reduced in this study due to the relatively long growing season with the 130-day maturity maize hybrid (Crusciol et al., 2013). Sowing plants with earlier relative maturity may beneﬁt an intercropping system and decrease the competition between N demand of both crops. Furthermore, this research provides as a novel outcome that intercropping systems combined with fertilizer N management showed eﬀectiveness in improving the overall productivity of the whole system, especially for enhancing meat production and revenue for farmers. November 2020 \| Volume 4 \| Article 544853 Frontiers in Sustainable Food Systems \| www.frontiersin.org 8 Nitrogen Management in Intercropping Systems Crusciol et al. TABLE 5 \| Economic evaluation of monocropped maize, maize intercropped with palisadegrass and maize intercropped with guineagrass as a function of N management for maize (average of three growing seasons). Frontiers in Sustainable Food Systems \| www.frontiersin.org Land Use Efﬁciency, Intercropping Competition Factors and Economics Land Use Efﬁciency, Intercropping Competition Factors and Economics (Moraes et al., 2019), and C4 grass residues are more favorable in long-term protection and coverage of the soil under tropical drought conditions than C3 residues due to slower residue decomposition rate (Mateus et al., 2016; Rosolem et al., 2017). (Moraes et al., 2019), and C4 grass residues are more favorable in long-term protection and coverage of the soil under tropical drought conditions than C3 residues due to slower residue decomposition rate (Mateus et al., 2016; Rosolem et al., 2017). Competition Factors and Economics Based on observed yields, LER of maize intercropped with palisadegrass was 1.06. The LER indicates the productivity of land with intercropping relative to sole cropping on separate parcels of land. The value of 1.06 indicated that 6% less land would be needed to achieve the same yield as monocropped maize and palisadegrass separately. These results are in line with those of Meixiua et al. (2020), who found that the average LER in grass/grass (maize/wheat) intercropping was 1.59. Likewise, Pariz et al. (2017b) found that the average LER of maize/palisadegrass intercropping was 1.10. However, maize intercropping with palisadegrass without N application or any intercropping of maize with guineagrass resulted in LER <1, reﬂecting lower productivity of land use. Our study suggests that land saving potential for food production systems in tropical soil can only be obtained in maize-palisadegrass intercropping systems with N management, independent of the type of split-N application. Nitrogen fertilizer application provided an average maize yield of 8.3 Mg ha−1. While N addition increased maize yield, applying no N fertilizer at seeding (0–100 kg N ha−1 applied at maize seeding and sidedressing, respectively) resulted in the lowest ears per plant, shoot dry matter, and grain yield among cropping systems receiving N fertilizer. Similar reductions in maize yield under delayed application of N fertilizer or 100% application from maize growth stages V6-V11 under monocropping have been reported previously (Scharf et al., 2002; Walsh et al., 2012; Muller et al., 2017). Applying the total N rate (100 kg N ha−1) at sidedressing did not match optimum N uptake capabilities of maize, because signiﬁcant N supply is needed during early growth stages. Adequate maize development and N accumulation in the plant are closely associated with metabolism of soluble protein and sugar utilization (Faleiros et al., 1996). Land Use Efﬁciency, Intercropping Competition Factors and Economics Thus, a portion of the N fertilizer must be applied at maize seeding in this NTS with grass cover crop to achieve high yield potential. There was no diﬀerence in maize yield or forage characteristics whether application of N was all at maize seeding or split between seeding + sidedressing. Maize was more competitive (K values) than the forage species; however, the dominant species in the system were forage grasses due to their aggressiveness. The K values of maize were greater than those of the forage species, in agreement with Zarochentseva (2012) and possibly due to the shading eﬀect of maize on forage grass during maize development. Our results showed that maize was able to acquire more resources in the intercropping systems even though the forage grasses were the dominant species. No diﬀerence in forage dry matter production between palisadegrass and guineagrass in the ﬁrst cut may have been related to climate conditions. Low forage growth (2.1 Mg ha−1) occurred with low rainfall and temperatures between 10 and 15◦C in early winter (Mateus et al., 2016). For the second cut, climate conditions could also explain the 23% increase in dry matter for palisadegrass compared with guineagrass. Temperature increased and stimulated the production of forage biomass, apparently with a greater eﬀect on palisadegrass. In general, greater values of estimated animal stocking rate and estimated meat production were obtained with greater forage dry matter production. For LER and K values of intercropping competition, addition of N fertilizer was necessary to enhance competitiveness of maize. Nitrogen management promoted the competitiveness of maize and forage grasses by increasing vegetative growth and providing greater capacity for utilizing limited availability of water (Marschner, 2012; Yang and Udvardi, 2018). Previous studies have shown signiﬁcant diﬀerences among crops in grass/grass intercropping systems, but not among diﬀerent types of N addition (O’Leary and Smith, 1999; Baxevanos et al., 2017). When a species has high competitiveness, the plant acquires more resources and occupies a superior ecological niche (Grace and Tilman, 1990). In addition, the A index values were extremely low for all treatments, indicating a minimum dominance by forage grasses. These ﬁndings highlight the necessity of choosing suitable species for intercropping in maize-forage grass systems to enhance the interspeciﬁc complementarity and reduce interspecies competition (Davis and Woolley, 1993). Maize and Tropical Forage Grass Responses Previous studies have shown that intercropping maize/sorghum with palisadegrass/guineagrass did not aﬀect grain yield or create better conditions for improving sorghum yield (Barducci et al., 2009; Borghi et al., 2013b), which may have been related to lower nutrient demand and diﬀerence in crop hybrid. species (Pariz et al., 2009; Crusciol et al., 2013). Although we observed high N uptake by monocropped maize, other studies have shown that intercropping systems with forage grasses do not impair N uptake by crops (Crusciol et al., 2011; Mateus et al., 2012; Borghi et al., 2013b). Another important ﬁnding of our study is evidence of high N demand, as the leaf N concentration was below the appropriate range for maize (27– 35 g N kg−1) (Cantarella et al., 1997) when there was no N application (control) for all intercropping systems and even for monocropped maize. Despite diﬀerences between intercropping systems, maize was adequately nourished in all treatments. Maize leaf concentrations of P, K, Ca, Mg, and S were within ranges considered adequate (Cantarella et al., 1997), and no nutrition problems were observed. Shoot dry matter and grain yield of maize were strongly related to timing of N fertilizer in the cropping system. Lowest maize yield was a result of insuﬃcient N supply to maize. Grass-grass rotation without N fertilizer addition can result in signiﬁcant N immobilization via competition between plants and microorganisms (Schimel and Bennett, 2004; Kuzyakov and Xu, 2013). Introduction of legumes in the crop rotation can enhance soil N availability with NTS (Boddey et al., 2010). However, cultivation of forage grasses is well-established among farmers Although no diﬀerences in agronomic characteristics and kernel attributes were observed between monoculture and intercropping systems, greater shoot dry matter and grain yield of maize were observed in monocropped maize compared with the other treatments. The lack of competition with tropical forage grasses positively aﬀected maize development and did not November 2020 \| Volume 4 \| Article 544853 9 Nitrogen Management in Intercropping Systems Crusciol et al. Frontiers in Sustainable Food Systems \| www.frontiersin.org Land Use Efﬁciency, Intercropping Competition Factors and Economics Production of dry matter for forage of up to 4 Mg ha−1 is considered good (Borghi et al., 2013a) and was achieved in the second cut, even though air temperature was not ideal for forage development (i.e., 30–35◦C) (Costa et al., 2005). Furthermore, there was no eﬀect of forage species in the intercropping system on CP in the ﬁrst and second cuts. CP is an important parameter of nutritive value. Forage CP averaged 125 g kg−1, which was more than adequate of the 70 g kg−1 minimum required for maintaining rumen microbial eﬃciency in cattle (van Soest, 1994). Intercropping is a sustainable practice of food production to improve quality of pastures and animal carrying capacity. Our results demonstrated that intercropping of tropical forage grasses with maize using NTS is a feasible option for increasing sustainability in tropical areas and can result in higher revenues for farmers due to the productive, economic, and environmental beneﬁts of these systems. Furthermore, these systems can increase global food production from the same land area (Carvalho et al., 2010; FAO—Food and Agriculture Organization of the United Nations, 2010, 2017; Herrero et al., 2010; Franzluebbers and Stuedemann, 2014; Moraes et al., 2019). Therefore, our data indicated that maize intercropped with palisade or guineagrass is a promising approach for farmers, Forage dry matter and CP were at highest levels as long as N was applied, irrespective of timing and split N ratio. As expected, grasses responded to N fertilizer because of high N demand (Boddey et al., 1996; Mateus et al., 2016). The N fertilizer rate of 100 kg ha−1 was considered relatively low for complex intercropping systems, but was compatible with our study’s focus on ﬁnding an eﬃcient N management strategy for enhancing productivity in a sustainable manner. Indeed, we observed greater maize yield and forage dry matter production compared to other studies with application rates of <100 kg N ha−1 (Mateus et al., 2016; Rosolem et al., 2017). November 2020 \| Volume 4 \| Article 544853 10 Nitrogen Management in Intercropping Systems Crusciol et al. especially in the tropical regions of South America, Africa, and parts of Asia, where individuals need additional opportunities to produce food. diversity of plants and soil microorganisms. However, N fertilizer application in these systems is still necessary for maize yields and proﬁtability. CONCLUSION The authors would like to thank the São Paulo Research Foundation for ﬁnancial support (FAPESP, Grant #2003/09914-3 and Grant #2003/01968-7). The ﬁrst, eighth, and tenth authors would like to thank the National Council for Scientiﬁc and Technological Development (CNPq) for an award for excellence in research. Intercropping maize with forage grasses is a promising practice to meet the dual challenges of food production and sustainable development. Since agricultural systems are region- and soil- speciﬁc, variations of intercropping systems may require diﬀerent N fertilization recommendations. Although monocropped maize produced greatest grain yield, intercropping systems were viable in terms of balanced grain and forage yields, land use, and proﬁtability. Estimated meat production and revenue were enhanced with intercropping of palisadegrass or guineagrass with maize. Combining animal production with crop production in an intercropping system can be advantageous not only for farmers, but also for environmental quality and biological Land Use Efﬁciency, Intercropping Competition Factors and Economics At least a portion of total N input should be applied at seeding and the remainder at sidedressing of maize. Application of all N fertilizer at sidedressing was not a productive practice since in this study with pearl millet as previous cover crop under NTS, as it reduced maize yield and revenue. Future studies should examine biodiversity improvements in the soil- plant-microorganism interactions and the negative impacts of N losses and nitrous oxide gasses release to the environment in the short- and long-term intercropping systems. All treatments resulted in net proﬁt, particularly the maize + palisadegrass and maize + guineagrass treatments with N management, because in addition to maize yield in the summer/autumn, farmers can use the forage dry matter production of palisade and guineagrass (Table 3) for animal fodder in the winter/spring. Thus, with maize intercropping, farmers could produce 87–218 kg ha−1 meat, with net proﬁts up to US$ 1,600–1,800 ha−1, depending of N management, which could add an extra US$ 500–600 than monocropped maize. In addition, the need for soil mulch would be satisﬁed in planning for the next crop. SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsufs. 2020.544853/full#supplementary-material AUTHOR CONTRIBUTIONS CACC, GM, CC, and HC designed the experiment. CACC, GM, EB, and JC obtained and processed the data. CACC, LM, CP, AC, and AF analyzed the data. CACC, LM, and CP wrote the paper, with contribution of all co-authors. All authors conﬁrms being contributor of this work and has approved it for publication. CACC, GM, CC, and HC designed the experiment. CACC, GM, EB, and JC obtained and processed the data. CACC, LM, CP, AC, and AF analyzed the data. CACC, LM, and CP wrote the paper, with contribution of all co authors All authors conﬁrms being CACC, GM, CC, and HC designed the experiment. CACC, GM, EB, and JC obtained and processed the data. CACC, LM, CP, AC, DATA AVAILABILITY STATEMENT The original contributions presented in the study are included in the article/Supplementary Materials, further inquiries can be directed to the corresponding authors. Overall, our data suggest that intercropping systems are a great option for the diversiﬁcation on farm and the increase of grain and forage yields. However, these agricultural systems deserve further investigations to assess the disadvantages and impacts of N fertilizer. Our study raises relevant questions about which changes occur in root systems and soil fertility and microbiology in deeper soil layers, as well as the reduction in diseases and pesticide applications, in maize-forage systems receiving N fertilizer, and the eﬀects of fertilizer on N losses to environment and N recovery by plants in short- and long-term of intercropping systems. REFERENCES maximum with corn and nitrogen fertilization. Arch. Zootecn. 58, 211–222. doi: 10.4321/S0004-05922009000200006 maximum with corn and nitrogen fertilization. Arch. Zootecn. 58, 211–222. doi: 10.4321/S0004-05922009000200006 Agegnehu, G., Ghizaw, A., and Sinebo, W. (2006). Yield performance and land-use eﬃciency of barley and fava bean mixed cropping in Ethiopian highlands. Eur. J. Agron. 25, 202–207. doi: 10.1016/j.eja.2006. 05.002 Baxevanos, D., Tsialtas, I. T., Vlachostergios, D. N., Hadigeorgiou, I., Dorda, C., and Lithourgidis, A. (2017). Cultivar competitiveness in pea-oat intercrops under Mediterranean conditions. Field Crop. Res. 214, 94–103. doi: 10.1016/j.fcr.2017.08.024 Agrolink. (2018). Quotations to Farming. Available online at: www.agrolink.com. br (accessed November 01, 2019). Boddey, R. M., Jantalia, C. P., Conceicao, P. C., Zanatta, J. A., Bayer, C., Mielnizuk, J., et al. (2010). Carbon accumulation at depth in Ferralsols under zero till subtropical agriculture. Glob. Change Biol. 16, 784–795. doi: 10.1111/j.1365-2486.2009.02020.x Amaral Filho, J. P. R., Fornasieri Filho, D., Farinelli, R., and Barbosa, J. C. (2005). Row spacing, population density and nitrogen fertilization in maize. R. Bras. Ci. Solo. 29, 467–473. doi: 10.1590/S0100-06832005000300017 Boddey, R. M., Rao, I. M., and Thomas, R. J. (1996). “Nutrient cycling and environmental impact of Brachiaria pastures,” in The Biology, Agronomy, and Improvement of Brachiaria. eds J. W. Miles, B. L. Maass, and C. B. do Valle (Cali: CIAT), 72–86. Anghinoni, I. (2007). “Fertilidade do solo e seu manejo em sistema plantio direto,” in Fertilidade do Solo. ed R. F. Novais (Viçosa: Sociedade Brasileira de Ciência do Solo), 873–928. Borghi, E., Crusciol, C. A. C., Mateus, G. P., Nascente, A. S., and Martins, P. O. (2013a). Intercropping time of corn and palisadegrass or guineagrass Barducci, R. S., Costa, C., Crusciol, C. A. C., Borghi, E., Putarov, T. C., and Sarti, L. M. N. (2009). Production of Brachiaria brizantha and Panicum November 2020 \| Volume 4 \| Article 544853 Frontiers in Sustainable Food Systems \| www.frontiersin.org 11 Nitrogen Management in Intercropping Systems Crusciol et al. aﬀecting grain yield and forage production. Crop Sci. 53, 629–636. doi: 10.2135/cropsci2012.08.0469 aﬀecting grain yield and forage production. Crop Sci. 53, 629–636. doi: 10.2135/cropsci2012.08.0469 FAO—Food and Agriculture Organization of the United Nations (2010). An International Consultation on Integrated Crop-Livestock Systems for Development: The Way Forward for Sustainable Production Intensiﬁcation. 64p (Integrated Crop Management, v.13). (Rome). Borghi, E., Crusciol, C. A. C., Nascente, A. S., Souza, V. V., Martins, P. O., Mateus, G. P., et al. (2013b). REFERENCES Sorghum grain yield, forage biomass production and revenue as aﬀected by intercropping time. Eur. J. Agron. 51, 130–139. doi: 10.1016/j.eja.2013.08.006 Fox, D. G., Tedeschi, L. O., Tylutki, T. P., Russell, J. B., van Amburgh, M. E., Chase, L. E., et al. (2004). The Cornell Net Carbohydrate and Protein System model for evaluating herd nutrition and nutrient excretion. Anim. Feed. Sci. Technol. 112, 29–78. doi: 10.1016/j.anifeedsci.2003.10.006 Borghi, E., Crusciol, C. A. C., Trivelin, P. C. O., Nascente, A. S., Costa, C., and Mateus, G. P. (2014). Nitrogen fertilization (15NH4NO3) of palisadegrass and residual eﬀect on subsequent no-tillage corn. R. Bras. Ci. Solo 38, 1457–1468. doi: 10.1590/S0100-06832014000500011 Franzluebbers, A. J., and Stuedemann, J. A. (2014). Crop and cattle production responses to tillage and cover crop management in an integrated crop– livestock system in the southeastern. USA. Eur. J. Agron. 57, 62–70. doi: 10.1016/j.eja.2013.05.009 Cantarella, H., van Raij, B., and Camargo, C. E. O. (1997). “Cereals,” in Lime and Fertilizer Recommendations for the State of Sao 428 Paulo. 2nd Edn, eds B. Raij, H. van Cantarella, J. A. Quaggio, and A. C. M. Furlani (Campinas: Tech. Bull. 100. Inst. Agronomico), 43–50. Garcia, C. M. P., Costa, C., Andreotti, M., Meirelles, P. R. L., Pariz, C. M., Freitas, L. A., et al. (2016). Wet and dry maize yield under intercrop culivation with marandu grass and/or dwarf pigeon pea and nutritional value of the marandu grass in succession. Austral. J. Crop. Sci. 10, 1564–1571. doi: 10.21475/ajcs.2016.10.11.PNE183 Carvalho, P. C. F., Anghinoni, I., Moraes, A., Souza, E. D., Sulc, R. M., Lang, C. R., et al. (2010). Managing grazing animals to achieve nutrient cycling and soil improvement in no-till integrated systems. Nutr Cycl Agroecosyst. 88, 259–273. doi: 10.1007/s10705-010-9360-x Grace, J. B., and Tilman, D. (1990). On the relationship between plant traits and competitive ability. Perspect. Plant Compet. 1, 51–65. doi: 10.1016/B978-0-12-294452-9.50008-4 CONAB (2018). Cost of Agricultural Production. Available online at: http://www. conab.gov.br/conab/Main.php?MagID=3&MagNo=39 (accessed July 21, 2018). Heanes, D. L. (1984). Determination of total organic-C in soils by an improved chromic acid digestion and spectrophotometric procedure. Commun. Soil Sci. Plant Anal. 10, 1191–1213. doi: 10.1080/00103628409367551 Cong, W. F., Hoﬄand, E., Li, L., Six, J., Sun, J. H., Bao, X. G., et al. (2015). Intercropping enhances organic carbon and nitrogen in soil. Glob. Chang. Biol. 21, 1715–1726. doi: 10.1111/gcb.12738 Costa, K. A. P., Rosa, B., Oliveira, I. P., Custodio, D. P., and Silva, D. C. (2005). REFERENCES Eﬀect of seasonal climate condition on the dry matter production and bromatological composition of Brachiaria brizantha cv. Marandu. Cienc. Anim. Bras. 6, 187–193. Herrero, M., Thornton, P. K., Notenbaert, A. M., Wood, S., Msangi, S., Freeman, H. A., et al. (2010). Smart investments in sustainable food production: revisiting mixed crop-livestock systems. Science 327, 822–825. doi: 10.1126/science.1183725 Costa, N. R., Andreotti, M., Gameiro, R. A., Pariz, C. M., Buzetti, S., and Lopes, K. S. M. (2012). Nitrogen fertilization in the intercropping of corn with two Brachiaria species in a no tillage system. Pesqui. Agropecu. Bras. 47, 1038–1047. doi: 10.1590/S0100-204X2012000800003 Himmelstein, J., Ares, A., Gallagher, D., and Myers, J. (2017). A meta- analysis of intercropping in Africa: impacts on crop yield, farmer income, and integrated pest management eﬀects. Int. J. Agric. Sustain. 15, 1–10. doi: 10.1080/14735903.2016.1242332 Horwitz, W. (1980). Oﬃcial methods of analysis of the association of analytical chemists. Assoc. Oﬀ. Anal. Chem. 60, 414–414. Costa, N. R., Andreotti, M., Lopes, K. S. M., Yokobatake, K. L., Ferreira, J. P., Pariz, C. M., et al. (2015). Soil properties and carbon accumulation in an integrated croplivestock system under no-tillage. Rev. Bras. Ci. Solo. 39, 852–863. doi: 10.1590/01000683rbcs20140269 Kowles, R. V., and Phillips, R. L. (1988). Endosperm development in maize. Int. Rev. Cytol. 112, 97–106. doi: 10.1016/S0074-7696(08)62007-0 Crusciol, C. A. C., Mateus, G. P., Nascente, A. S., Martins, P. O., Borghi, E., and Pariz, C. M. (2012). An innovative crop-forage intercrop system: early cycle soybean cultivars and palisade grass. Agron. J. 104, 1085–1095. doi: 10.2134/agronj2012.0002 Kuzyakov, Y., and Xu, X. (2013). Competition between roots and microorganisms for nitrogen: mechanisms and ecological relevance. New Phytol. 198, 656–669. doi: 10.1111/nph.12235 Li, L., Yang, S., Li, X., Zhang, F., and Christie, P. (1999). Interspeciﬁc complementary and competitive interactions between intercropped maize and faba bean. Plant Soil 212, 105–114. doi: 10.1023/A:1004656205144 Crusciol, C. A. C., Mateus, G. P., Pariz, C. M., Borghi, E., Costa, C., and Silveira, J. P. F. (2011). Nutrition and sorghum hybrids yield with contrasting cycles intercropped with Marandu grass. Pesq. Agropec. Bras. 46, 1234–1240. doi: 10.1590/S0100-204X2011001000017 Malavolta, E., Vitti, G. C., and Oliveira, S. A. (1997). Evaluation of Nutritional Status of Plant: Principles and Applications, 2nd Edn. Piracicaba: Potafos. Marschner, H. (2012). Mineral Nutrition of Higher Plants, 3rd Edn. London: Academic Press. Crusciol, C. A. C., Nascente, A. S., Borghi, E., Soratto, R. P., and Martins, P. O. (2015). REFERENCES Improving soil fertility and crop yield in a tropical region with palisadegrass cover crops. Agron. J. 107, 2271–2280. doi: 10.2134/agronj14.0603 Martin-Guay, M., Paquette, A., Dupras, J., and Rivest, D. (2018). The new green revolution: sustainable intensiﬁcation of agriculture by intercropping. Sci. Total Environ. 615, 767–772. doi: 10.1016/j.scitotenv.2017.10.024 Crusciol, C. A. C., Nascente, A. S., Mateus, G. P., Borghi, E., Leles, E. P., and Santos, N. C. B. (2013). Eﬀect of intercropping on yields of corn with diﬀerent relative maturities and palisadegrass. Agron. J. 105, 599–606. doi: 10.2134/agronj2012.0426 Mateus, G. P., Crusciol, C. A. C., Pariz, C. M., Borghi, E., Costa, C., Martello, J. M., et al. (2016). Sidedress nitrogen application rates to sorghum intercropped with tropical perennial grasses. Agron. J. 108, 433–447. doi: 10.2134/agronj2015.0236 Crusciol, C. A. C., Nascente, A. S., Mateus, G. P., Pariz, C. M., Martins, P. O., and Borghi, E. (2014). Intercropping soybean and palisadegrass for enhanced land use eﬃciency and revenue in a no till system. Eur. J. Agron. 58, 53–62. doi: 10.1016/j.eja.2014.05.001 Mateus, G. P., Crusciol, C. A. C., Pariz, C. M., Costa, N. R., Borghi, E., Costa, C., et al. (2020). Corn intercropped with tropical perennial grasses as aﬀected by sidedress nitrogen application rates. Nutr. Cycl. Agroecosyst. 116, 223–244. doi: 10.1007/s10705-019-10040-1 Davis, J. H. C., and Woolley, J. N. (1993). Genotypic requirement for intercropping. Field Crop. Res. 34, 407–430. doi: 10.1016/0378-4290(93)90124-6 Mateus, G. P., Crusciol, C. A. C., and Borghi, E. (2012). Integrated croplivestock: the new green revolution in the tropics. Pesqui Tecnol. 4, 1–5. Derpsch, R., and Friedrich, T. (2009). “Development and current status of no-till adoption in the world,” in Proceedings on CD 18th Triennial Conference of the International Soil Tillage Research Organization (Izmir: ISTRO), 15–19. Mead, R., and Willey, R. W. (1980). The concepts of a ‘land equivalent ratio’ and advantages in yields from intercropping. Exp. Agric. 16, 217–228. doi: 10.1017/S0014479700010978 Meixiua, T., Fang, G., Stomph, T. J., Jing, W., Wen, Y., Lizhen, Z., et al. (2020). Dynamic process-based modelling of crop growth and competitive water extraction in relay strip intercropping: model development and application to wheat-maize intercropping. Field Crop Res. 246:107613. doi: 10.1016/j.fcr.2019.107613 Faleiros, R. R. S., Seebauer, J. R., and Below, F. E. (1996). Nutritionally induced changes in endosperm of shrunken-1 and brittle-2 maize kernels grown in vitro. Crop Sci. 36, 947–954. doi: 10.2135/cropsci1996.0011183X0036000 40022x FAO—Food Agriculture Organization of the United Nations (2017). Integated Crop-Livestock System (ICLS). REFERENCES H., Ahmad, R., Ahmad, A., and Jabbar, A. (2009). Competitive functions of components crops in some barley based intercropping systems. Int. J. Agric. Biol. 11, 1560–853011. Pariz, C. M., Andreotti, M., Tarsitano, M. A. A., Bergamaschine, A. F., Buzetti, S., and Chioderolli, C. A. (2009). Technical and economical performance of corn intercropped with Panicum and Brachiaria forage in crop-livestock integration system. Pesq. Agropec. Trop. 39, 360–370. Walsh, O., Raun, W., Klatt, A., and Solie, J. (2012). Eﬀect of delayed nitrogen fertilization on maize (Zea mays L.) grain yields and nitrogen use eﬃciency. J. Plant Nutr. 35, 538–555. doi: 10.1080/01904167.2012.644373 Yang, J., and Udvardi, M. (2018). Senescence and nitrogen use eﬃciency in perennial grasses for forage and biofuel production. J. Exp. Bot. 69, 855–865. doi: 10.1093/jxb/erx241 Pariz, C. M., Costa, C., Crusciol, C. A. C., Castilhos, A. M., Meirelles, P. R. L., Roça, R. O., et al. (2017a). Lamb production responses to grass grazing in a companion crop system with corn silage and oversowing of yellow oat in a tropical region. Agr. Syst. 151, 1–11. doi: 10.1016/j.agsy.2016. 11.004 Zarochentseva, O. (2012). Adaptation of Methodology Calculation Relative Crowding Coeﬃcient for Evaluation Competition of Tree Species in Polyculture. Ukraine: Yuriy Fedkovich Chernivtsi National University, 1–2. Zarochentseva, O. (2012). Adaptation of Methodology Calculation Relative Crowding Coeﬃcient for Evaluation Competition of Tree Species in Polyculture. Ukraine: Yuriy Fedkovich Chernivtsi National University, 1–2. Pariz, C. M., Costa, C., Crusciol, C. A. C., Meirelles, P. R. L., Castilhos, A. M., Andreotti, M., et al. (2016). Production and soil responses to intercropping of forage grasses with corn and soybean silage. Agron. J. 108, 2541–2553. doi: 10.2134/agronj2016.02.0082 Zhang, G., Yang, Z., and Dong, S. (2011). Interspeciﬁc competitiveness aﬀects the total biomass yield in an alfalfa and corn intercropping system. Field Crop. Res. 124, 66–73. doi: 10.1016/j.fcr.2011.06.006 Conﬂict of Interest: EB was employed by the company Brazilian Agricultural Research Corporation (EMBRAPA), Brazil. Pariz, C. M., Costa, C., Crusciol, C. A. C., Meirelles, P. R. L., Castilhos, A. M., Andreotti, M., et al. (2017b). Production, nutrient cycling and soil compaction to grazing of grass companion cropping with maize and soybean. Nutr. Cycl. Agroecosyst. 108, 35–54. doi: 10.1007/s10705-016-9821-y The remaining 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. Rosolem, C. A., Ritz, K., Cantarella, H., Galdos, M. V., Hawkesford, M. J., Whalley, W. REFERENCES Available online at: http://www.fao.org/ agriculture/crops/core-themes/theme/spi/scpi-home/managing-ecosystems/ integrated-crop-livestock-systems/en/ (accessed November 1, 2018). Mendiburu, F. D. (2015). Agricolae: Statistical Procedures for Agricultural Research. R Package Version 1.2-3. Available online at: http://CRAN.R-project.org/ package=agricolae Frontiers in Sustainable Food Systems \| www.frontiersin.org Frontiers in Sustainable Food Systems \| www.frontiersin.org November 2020 \| Volume 4 \| Article 544853 12 Nitrogen Management in Intercropping Systems Crusciol et al. Moraes, A., Carvalho, P. C. F., Crusciol, C. A. C., Lang, C. R., Pariz, C. M., Deiss, L., et al. (2019). Chapter 16 - integrated crop-livestock systems as a solution facing the destruction of Pampa and Cerrado biomes in South America by intensive monoculture systems. Agroecosyst. Div. 1, 257–273. doi: 10.1016/B978-0-12-811050-8.00016-9 Schimel, J. P., and Bennett, J. (2004). Nitrogen mineralization: challenges of a changing paradigm. Ecology 85, 591–602. doi: 10.1890/03-8002 Sulc, R. M., and Tracy, B. F. (2007). Integrated crop-livestock systems in the U.S. Corn Belt. Agron. J. 99, 335–345. doi: 10.2134/agronj2006.0086 Corn Belt. Agron. J. 99, 335–345. doi: 10.2134/agronj2006.0086 Takim, F. O. (2012). Advantages of maize-cowpea intercropping over sole cropping through competition indices. J. Agric. Biodivers. Res. 1, 53–59. Muller, S. M., Camberato, J. J., Messina, C., Shanahan, J., Zhang, H., and Vyn, T. J. (2017). Late-split nitrogen applications increased maize plant nitrogen recovery but not yield under moderate to high nitrogen rates. Agron. J. 109, 2689–2699. doi: 10.2134/agronj2017.05.0282 Tracy, B. F., and Zhang, Y. (2008). Soil compaction, corn yield response, and soil nutrient pool dynamics within an integrated crop-livestock system in Illinois. Crop Sci. 48, 1211–1218. doi: 10.2135/cropsci2007.07.0390 Murphy, J., and Riley, J. P. (1962). A modiﬁed single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27, 31–36. doi: 10.1016/S0003-2670(00)88444-5 United States Department of Agriculture (USDA) (2014). Keys to Soil Taxonomy, 12th Edn. Washington, DC, 372. van Raij, B., Andrade, J. C., Cantarella, H., and Quaggio, J. A. (2001). Chemical Analysis for Fertility Evaluation of Tropical Soils. Campinas: Instituto Agronômico (IAC). O’Leary, N., and Smith, M. E. (1999). Breeding corn for adaptation to two diverse intercropping companions. Am. J. Altern. Agric. 14, 158–164. doi: 10.1017/S0889189300008328 van Soest, P. J. (1994). Nutritional Ecology of the Ruminant, 2nd Edn. Ithaca: Cornell University Press. Pariz, C. M., Andreotti, M., Azenha, M. V., Bergamaschine, A. F., Mello, L. M. M., and Lima, R. C. (2011). Corn grain yield and dry mass of Brachiaria intercrops in the crop–livestock integration system. Cienc. Rural 41, 875–882. doi: 10.1590/S0103-84782011000500023 Wahla, I. Copyright © 2020 Crusciol, Mateus, Momesso, Pariz, Castilhos, Calonego, Borghi, Costa, Franzluebbers and Cantarella. 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. Frontiers in Sustainable Food Systems \| www.frontiersin.org November 2020 \| Volume 4 \| Article 544853 REFERENCES R., et al. (2017). Enhanced plant rooting and crop system management for improved N use eﬃciency. Adv. Agron. 146, 205–239. doi: 10.1016/bs.agron.2017.07.002 Copyright © 2020 Crusciol, Mateus, Momesso, Pariz, Castilhos, Calonego, Borghi, Costa, Franzluebbers and Cantarella. 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. Sawyer, J. E., Pedersen, P., Barker, D. W., Diaz, D. A. R., and Albrecht, L. (2010). Intercropping corn and kura clover: response to nitrogen fertilization. Agron. J. 102, 568–574. doi: 10.2134/agronj2009.0392 Scharf, P. C., Wiebold, W. J., and Lory, J. A. (2002). Corn yield response to nitrogen fertilizer timing and deﬁciency level. Agron. J. 94, 435–441. doi: 10.2134/agronj2002.4350 Frontiers in Sustainable Food Systems \| www.frontiersin.org November 2020 \| Volume 4 \| Article 544853 13
https://openalex.org/W2903483051	https://repozytorium.biblos.pk.edu.pl/redo/resources/41498/file/suwFiles/matecconf_icchmt2018_05009.pdf	English	null	Calculations of the pressure drop in the natural circulation boiler evaporator	MATEC web of conferences	2,018	cc-by	2,412	2 Structure of two-phase flows in power boilers Fig. 1. Two-phase flow patterns for water-steam mixture in heated vertical channels [9]. Due to their complex flow structure, multi-phase flows are described by means of many different models Most power boiler evaporators in Poland are made of smooth tubes. The installed units are usually natural circulation boilers. For them, the steam dryness factor x at the risers outlet must not exceed 0.2. Due to that, the circulation multiplicity k cannot be smaller than 5: 1 Introduction 1 x k  (1) (1) Two-phase flows occur in many industrial processes. In the power sector, the change in the state of aggregation takes place primarily in the power boiler evaporator and in the condenser. The appearance of the steam-water mixture involves a change in the thermal and flow properties. The thermal and flow conditions change considerably in the boiler evaporator, where the content of steam in the mixture varies with the amount of heat supplied along the medium flow path. The appearance of the steam-water mixture results in phenomena which cannot be observed in single-phase flows. The pressure drop should be determined taking account of interphase forces, the shape and size of bubbles, the velocity of individual phases (the split ratio) and many other factors [1, 3, 4]. The phenomena occurring in the boiler evaporator involve supercooled boiling and fully-developed nucleate boiling. The steam-water mixture flow structures in vertical channels are shown in Fig. 1. Fig. 1. Two-phase flow patterns for water-steam mixture in heated vertical channels [9]. The pressure drop can be calculated using many mathematical models. In literature, the steam-water mixture is treated as homogenous or it is considered under the phase-slip model assumptions. There are also models using graphical relations intended for the pressure drop determination. The paper presents results of the pressure drop in the OP-210 boiler evaporator determined by means of the homogeneous model, the split-phase models (the Lockhart-Martinelli, the Friedel, the Chisholm model) and the Martinelli-Nelson graphical method [1, 3, 4, 10]. Calculations of the pressure drop in the natural circulation boiler evaporator Sławomir Grądziel1,, Karol Majewski1 stitute of Thermal Power Engineering, Cracow University of Technology, al. Jana Pawła II 37, 31-864 Kraków, Polan Abstract. The paper presents different models used to determine pressure losses in two-phase flows: the homogeneous model, the Lockhart-Martinelli, the Friedel and the Chisholm phase-slip models and the Martinelli-Nelson graphical method. The pressure losses are calculated for the evaporator of an OP-210 boiler with the output of 210×103 kg/h operating in one of the Polish power plants. The results obtained by means of the presented models are compared to each other. MATEC Web of Conferences 240, 05009 (2018) ICCHMT 2018 MATEC Web of Conferences 240, 05009 (2018) ICCHMT 2018 https://doi.org/10.1051/matecconf/201824005009 Corresponding author: gradziel@mech.pk.edu.pl © 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/). * Corresponding author: gradziel@mech.pk.edu.pl © 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/). MATEC Web of Conferences 240, 05009 (2018) MATEC Web of Conferences 240, 05009 (2018) ICCHMT 2018 https://doi.org/10.1051/matecconf/201824005009 In the general case, the total pressure drop is found from the following relation [1-7]: In the general case, the total pressure drop is found from the following relation [1-7]: The friction factor is found using classical equations, e.g. the Blasius formula, based on the Reynolds number found for averaged properties of the mixture [3, 4]. The friction factor is found using classical equations, e.g. the Blasius formula, based on the Reynolds number found for averaged properties of the mixture [3, 4]. f g a dp dp dp dp dz dz dz dz     (1) (1) 3.1 Homogenous model The homogenous model application is related to the assumption that the fluid properties are averaged for the whole mixture and the gaseous and the liquid phase velocities are identical. The equality of the velocities of the two phases means that there is no slip between them. For the homogeneous model, the friction-related pressure drop is found from the following relation:        0 5 2 0 5 2 2 2 2 1 1 1 . n . n n LO Y Bx x x             (8) (8) where: Y2 – the single-phase flow ratio between the gaseous and the liquid phase pressure gradients, B – coefficient dependent on the mass flow (mass flux), n – exponent describing the flow resistance coefficient. For the Blasius formula n=0.25. 2 2 f TP in TP dp f G dz d   (4) (4) The methods of determination of parameters Y2 and B are described in [4]. where: fTP – two-phase friction factor, G – mass flux, kg/(m2s), 3.2 The Lockhart-Martinelli model where: where: The Lockhart-Martinelli model is based on finding two- phase multipliers (the Lockhart-Martinelli multipliers), which make it possible to determine the pressure gradient related to frictional losses arising during the flow of the mixture [4]: f dp dz - frictional of pressure gradient, Pa/m, a dp dz - accelerational of pressure gradient, Pa/m, 2 TP L L dp dp dz dz  (5) g dp dz - gravitational of pressure gradient, Pa/m. (5) The two-phase multiplier used in equation (5) is found from the following relation: In Equation (1), the pressure drops related to changes in momentum and potential energy are determined from the following relation: 2 2 1 1 L C X X   (6) 2 2 1 1 L C X X   (6) (6)     2 2 2 1 1 a L L G x dp d x G dz dz                (2)   1 g G L dp g sin dz         (3) (2) The Martinelli parameters X and C used in equation (6) are described in [4]. The Martinelli parameters X and C used in equation (6) are described in [4]. (3) 3.3 The Friedel model where: where: The Friedel model was developed based on 25 thousand measuring points. It is used for two-phase flows in vertical and horizontal tubes. The two-phase multiplier is found from the following relation [4]: G – mass flux, kg/(m2s), ε – filling ratio, L G ,   – density of the liquid and the gaseous phase, kg/m3, 2 0 045 0 035 3 24 LO . . . FH E Fr We    (7) (7) α – tube inclination angle. Two-phase multiplier 2 LO  concerns the flow of a mixture characterized by properties related to the liquid phase properties. Coefficients E, F, H and the Froude and the Weber numbers (Fr and We, respectively) used in equation (7) are described in [4]. Determination of frictional pressure losses is a complex task. Therefore, many mathematical models have been developed to find the friction-related component. The mixture flow may be treated as homogeneous or as a slip flow. Using slip models involves calculating the two-phase flow multiplier, which makes it possible to find the friction-related pressure drop. 3.4 The Chisholm model The Chisholm model is an empirical method that can be used in a wide range of the steam pressure and dryness factor values. The two-phase multiplier is found from the following formula [4]: where: f ,LO dp dz - frictional losses arising in the flow of a two- phase mixture with properties related to the properties of the liquid phase, Pa/m. 3.5 • The Martinelli-Nelson graphical method TP – two-phase density, kg/m3. 2 MATEC Web of Conferences 240, 05009 (2018) ICCHMT 2018 https://doi.org/10.1051/matecconf/201824005009 The Martinelli-Nelson method was developed on the example of flows of the steam-water mixture through horizontal ducts in the pressure range of 0.689÷20.7 MPa. The two-phase multiplier needed to establish friction-related losses is read from a chart developed based on different values of the steam pressure and dryness factor x (cf. Fig. 2). The medium flow velocity and the degree of evaporation in the OP-210 boiler circulation contour were performed for the heat flux real values. The measurement methodology and the obtained results are presented in [2, 10]. Fig. 3. Evaporator scheme of OP-210 power boiler. Fig. 2. Martinelli-Nelson correlation. Fig. 3. Evaporator scheme of OP-210 power boiler. Heating up to saturation parameters and partial evaporation of boiler water occur in the waterwall tubes. The calculation methodology is based on the evaporator division into a few zones [8]. Evaporation occurs in two zones so that the steam dryness factor at the outlet of the first and the second evaporation section is x1=0.01 and x2=0.07, respectively. A rise in the dryness factor value involves bigger friction-related pressure drops. The resistances obtained using the methodology described in [8], the homogenous model and the phase-slip models are compared in Table 1 and Fig. 4. Fig. 2. Martinelli-Nelson correlation. Fig. 2. Martinelli-Nelson correlation. Fig. 2. Martinelli-Nelson correlation. If the pressure value exceeds the critical point for water, the multiplier is 2 1 LO   [11]. The two-phase multiplier read from the chart enables determination of the friction-related pressure drop by means of the following formula: Table 1. Comparison of results. Section I Section II a dp dz , Pa/m 1546 2399 g dp dz , Pa/m 6019 3879 f dp dz , Pa/m Model 1 142 208 Model 2 174 260 Model 3 318 968 Model 4 211 375 Model 5 194 348 Model 6 110 396 Model 1 – literature-based [8] Model 2 – homogenous model Model 3 – the Lockhart-Martinelli model Model 4 – the Friedel model Model 5 – the Chisholm model Model 6 – the Martinelli-Nelson graphical method Table 1. Comparison of results. 3.5 • The Martinelli-Nelson graphical method Section I Section II a dp dz , Pa/m 1546 2399 g dp dz , Pa/m 6019 3879 f dp dz , Pa/m Model 1 142 208 Model 2 174 260 Model 3 318 968 Model 4 211 375 Model 5 194 348 Model 6 110 396 Model 1 – literature-based [8] Model 2 – homogenous model Model 3 – the Lockhart-Martinelli model Model 4 – the Friedel model Model 5 – the Chisholm model Model 6 – the Martinelli-Nelson graphical method Table 1. Comparison of results. 2 f f ,LO LO dp dp dz dz  (9) (9) 4 Comparison of the OP-210 boiler evaporator pressure drops determined by means of different methods The OP-210 boiler is a single-drum natural circulation boiler. It generates superheated steam with the pressure of 9.8 MPa and temperature of 540°C. The boiler nominal output totals 210 x 103 kg/h. The pressure drop in the boiler evaporator was determined using the methodology described in [8]. The OP-210 boiler analysed circulation system, or contour, (with marked division into evaporation zones) is presented in Fig. 3. 3 3 MATEC Web of Conferences 240, 05009 (2018) ICCHMT 2018 https://doi.org/10.1051/matecconf/201824005009 natural circulation (Wydawnictwo Politechniki Krakowskiej, Kraków, 2012) The calculation results indicate that the friction- related pressure drops take similar values for most of the models under consideration. Some differences occur for the Lockhart-Martinelli method, which was developed based on flow measurements of the air-water and air-oil mixture. 3. Hetsroni G., Handbook of Multiphase System, (Hemisphere Publishing Corporation, Washington, 1982) 4. Hewitt G., Shires G., Bott T., Process heat transfer (CRC Press, Begell House, 1994) Fig. 4. Comparison of frictional losses established by means of different models: 1 – literature-based [8], 2 – homogenous model, 3 – the Lockhart-Martinelli model, 4 – the Friedel model, 5 – the Chisholm model, 6 – the Martinelli-Nelson graphical method 5. P. Ocłoń, M. Nowak, S. Łopata, J. of Therm. Sci., 23, 2, pp. 177-186 (2014) 6. P. Ocłoń, M. Nowak, K. Majewski, AIP Conference Proceedings, 1558, pp. 2419-2422 (2013) 7. P. Ocłoń, M. Nowak, B. Węglowski, T. Nabagło, P. Cisek, M. Jaremkiewicz, K. Majewski, J. of Appl. Comp. Sci., 22, 1, pp. 111-135 (2014) 8. Kuznetsov N.W., Nitor W.W., Dubovski I.E., Karasina E.S., Thermal Calculations of Steam Boilers. Standard Method (Energy, Moscow, 1973) 9. Taler J., Thermal and flow processes in large steam boilers. Modelling and monitoring (PWN, Warszawa, 2011) Fig. 4. Comparison of frictional losses established by means of different models: 1 – literature-based [8], 2 – homogenous model, 3 – the Lockhart-Martinelli model, 4 – the Friedel model, 5 – the Chisholm model, 6 – the Martinelli-Nelson graphical method Fig. 4. Comparison of frictional losses established by means of different models: 1 – literature-based [8], 2 – homogenous model, 3 – the Lockhart-Martinelli model, 4 – the Friedel model, 5 – the Chisholm model, 6 – the Martinelli-Nelson graphical method 10. Zima W., Grądziel S., Simulation of transient processesin heating surfaces of power boilers (LAP LAMBERT Academic Publishing,Germany, 2013) 11. R.C.Martinelli, D.B.Nelson, Trans. Amer. Soc. Mech. Engrs., 70, 6, pp. 5 Conclusions It follows from the calculations that a rise in the gaseous phase content in the flowing mixture involves an increase in friction-related flow resistances. The applied models differ in their approach to the two-phase flow multiplier determination, which results in some differences. However, in the flow case under analysis, the content of the gaseous phase in the mixture is small and, therefore, the discrepancies between individual models are slight. The share of frictional pressure losses in the overall pressure drop is relatively small. For this reason, it does not really matter which model is selected to determine friction- related pressure losses – the selection has little impact on the total pressure drop in the circulation system of natural circulation boilers. The changes in the mixture potential energy and momentum have the dominant effect on the total pressure drop. 4 Comparison of the OP-210 boiler evaporator pressure drops determined by means of different methods 695-702 (1948) The obtained results indicate that frictional losses contribute to the total pressure drop only slightly. This is related to the small value of the steam dryness factor x and to the boiler evaporator vertical structure. Due to that, it is the changes in the mixture potential energy and momentum that have the dominant impact on the level of the total pressure drop. References 1. Dziubiński M., Prywer J., Two-phase fluid mechanics (WNT, Warszawa, 2009) 2. Grądziel S., Modelling of thermal and flow phenomena occurring in evaporator of boiler with 4 4
https://openalex.org/W2920893667	https://link.springer.com/content/pdf/10.1007/s11252-019-00840-2.pdf	English	null	Ungulates in the city: light pollution and open habitats predict the probability of roe deer occurring in an urban environment	Urban ecosystems	2,019	cc-by	8,942	1 Department of Forest Biodiversity, Institute of Forest Ecology and Silviculture, Faculty of Forestry, University of Agriculture, al. 29 Listopada 46, 31-425 Kraków, Poland Urban Ecosystems (2019) 22:513–523 https://doi.org/10.1007/s11252-019-00840-2 Urban Ecosystems (2019) 22:513–523 https://doi.org/10.1007/s11252-019-00840-2 Abstract Although large and medium-sized herbivorous mammals avoid urbanized areas, they have recently begun to colonize towns and cities. In general, ungulates continue to avoid the centres of urban areas, and utilize mainly their thinly built-up outskirts. While extension of urban development is preventing ungulates from penetrating the urban landscape, the influence of noise and light pollution on the occurrence of mammalian herbivores is still poorly understood. Hence, we investigated the hypothesis that habitat availability shapes the distribution of roe deer Capreolus capreolus and artificial lightening discourages them from penetrating the urban landscape. Roe deer was recorded on 37% of randomly selected sample plots (N = 60) located within the city of Kraków (S Poland). The occupied plots contained significantly more open habitats, woodland patches were larger in them, but proximity to rivers, and noise and light pollution were significantly lower. The logistic regression model revealed that an increasing area of open habitats was positively correlated with the probability of roe deer occurring. However, the artificial lighting at night was negatively correlated with the probability of the species occurring: the negative effect of light pollution was mitigated by the greater area of open habitats. Our study highlights the very considerable potential of light pollution as a predictor of the occurrence of large mammals in the urban landscape. We argue that urbanization and the related artificial lighting at night may be a factor preventing ungulates from penetrating potentially suitable habitats in urban areas. Keywords Urbanization . Urban effect . Farmland . Ecological corridors . Artificial lighting success (Kilpatrick and Spohr 2000; Acevedo et al. 2005; Underwood and Kilheffer 2016). In general, large mammalian herbivores avoid strongly urbanized areas (Underwood and Kilheffer 2016; Loro et al. 2016). Small populations of ungu- lates colonize mainly suburban areas (McCarthy et al. 1996; Kilpatrick and Spohr 2000; Mattila and Hadjigeorgiou 2015; Loro et al. 2016), with only small numbers of individuals entering densely built-up areas (Warren 2011). Moreover, the colonization of suburbs by ungulates is often of a transi- tional nature, resulting from overabundances in adjacent areas (Warren 2011; Mattila and Hadjigeorgiou 2015). However, the considerable behavioural plasticity of ungulates (Kilpatrick and Spohr 2000; Acevedo et al. 2011; Morellet et al. 2011; Kušta et al. 2017) may intensify their colonization of towns and cities. It is possible, therefore, that such areas will become significant habitats for this group of mammals. * Michał Ciach michal.ciach@ur.krakow.pl Ungulates in the city: light pollution and open habitats predict the probability of roe deer occurring in an urban environment Michał Ciach1 & Arkadiusz Fröhlich1 Published online: 5 March 2019 # The Author(s) 2019 Published online: 5 March 2019 # The Author(s) 2019 Introduction One of the ecologically more flexible ungulate species is the roe deer Capreolus capreolus, which inhabits wood- lands and open country, and also a mosaic of both (Putman 1986; Morellet et al. 2011; Ewald et al. 2014). The habitats utilized by roe deer are often situated in a landscape matrix that includes areas transformed by humans to varying ex- tents (McCarthy et al. 1996; Acevedo et al. 2005; Torres et al. 2011; Loro et al. 2016). The results of studies done to date on the effects of urbanization on roe deer are equivo- cal: although this species seems to avoid areas with a mod- erate degree of urbanization (Hewison et al. 2001; Loro et al. 2016), it may occasionally prefer the proximity of thinly urbanized areas (Torres et al. 2011). The issue of roe deer encroaching into strongly urbanized areas has not been explored as yet. Furthermore, not many studies of the influence of urbanization on roe deer populations have been conducted during winter, when ungulates can forage on the supplementary food available in human set- tlements (Putman 1986; Kilpatrick and Spohr 2000; Torres et al. 2011). Apart from examining the influence of the urban landcover (Underwood and Kilheffer 2016; Loro et al. 2016) and the proximity of buildings (Hewison et al. 2001; Torres et al. 2011), studies to date have not looked at the effect of the noise and artificial light pro- duced by towns and cities, which may discourages roe deer from entering them, even if potentially suitable habitats are available there. Understanding the effects limiting the ex- ploration by roe deer of urban areas is important in the context of their frequent collisions with road vehicles (Bruinderink and Hazebroek 1996; Found and Boyce 2011), the important role they play in shaping the vegeta- tion (Welch et al. 1991; Putman and Moore 1998) and the contribution they make to the diet of some medium and large predators (Molinari-Jobin et al. 2002; Mattioli et al. 2004; Bateman and Fleming 2012). Human habitations pose a further threat to land animals (McCarthy et al. 1996; Hewison et al. 2001; Underwood and Kilheffer 2016): not only do they cause their habitats to shrink (Underwood and Kilheffer 2016; Loro et al. Introduction 2016), there is increased pressure on the part of humans and their dogs, which frequently flush out wild animals, raising stress levels and inflicting energy losses in the latter (Reimoser 2012; Padié et al. 2015). The encroach- ment of ungulates on to land used by humans often leads to conflicts (Putman and Moore 1998; Warren 2011; Mattila and Hadjigeorgiou 2015): as a consequence, hab- itats may be fenced off (Boone and Hobbs 2004; Harrington and Conover 2006) and the animals culled (Brown et al. 2000; Warren 2011). Yet another danger for terrestrial animals may be light pollution, which can interfere with their orientation abilities and thus with their movements (Longcore and Rich 2004; Beier 2006), while long-term exposure to artificial lighting can disrupt their biological rhythm (Yeates 1949; Lincoln and Guinness 1972; Barber-Meyer 2007); they may therefore tend to avoid strongly illuminated areas (Robert et al. 2015). In the temperate zone, ungulates inhabit mainly woodland and open habitats (farmland, grassland) (Mattioli 2011; Morellet et al. 2011); at the same time, however, they demon- strate a number of adaptations to life in the neighbourhood of human habitations (Kilpatrick and Spohr 2000; Torres et al. 2011; Kušta et al. 2017). Adapting diurnal activity to road traffic helps to avoid collisions with vehicles (Kušta et al. 2017), and diminished activity during hunting periods may limit mortality (Pagon et al. 2013). Moreover, despite the po- tential threat from humans (hunters), ungulates are able to tolerate the presence of human habitations (Mysterud et al. 1999; Hewison et al. 2001) and the associated activities of people (Bonnot et al. 2013; Padié et al. 2015). In certain cases, ungulates can also take advantage of human-dependent food resources, which saves or helps to gain energy (Putman 1986; Kilpatrick and Spohr 2000). Milder temperatures are normal in large cities (Arnfield 2003), so theoretically ungulates may find favourable conditions there in winter, because the thinner snow cover is no obstacle to foraging (Mysterud et al. 1999; Ewald et al. 2014). Another important factor is that popula- tions of the main natural enemies of ungulates – large and medium-sized predators like grey wolf Canis lupus and lynx The aim of this study was to assess the factors governing the winter distribution of roe deer in the urban environment. Introduction The encroachment of urbanization on to natural terrain usually means a loss of species from the latter (Mcdonald et al. 2008; Bateman and Fleming 2012) or their adaptation to this new habitat (McCarthy et al. 1996; Acevedo et al. 2011). As a result, urban areas become an important habitat for certain groups of animals, in which these can achieve high population densities (Mcdonald et al. 2008; Ciach and Fröhlich 2017). Ungulates are animals which have only quite recently begun to colonize urban areas, but so far with rather moderate Ungulates are animals which have only quite recently begun to colonize urban areas, but so far with rather moderate Electronic supplementary material The online version of this article (https://doi.org/10.1007/s11252-019-00840-2) contains supplementary material, which is available to authorized users. * Michał Ciach michal.ciach@ur.krakow.pl 1 Department of Forest Biodiversity, Institute of Forest Ecology and Silviculture, Faculty of Forestry, University of Agriculture, al. 29 Listopada 46, 31-425 Kraków, Poland Electronic supplementary material The online version of this article (https://doi.org/10.1007/s11252-019-00840-2) contains supplementary material, which is available to authorized users. * Michał Ciach michal.ciach@ur.krakow.pl * Michał Ciach michal.ciach@ur.krakow.pl Colonizing the urban landscape, however, is fraught with danger for terrestrial animals (Forman and Alexander 1998; Harrington and Conover 2006; Padié et al. 2015). Expansion of road networks in ungulate habitats hinders movements be- tween habitat patches, which in turn restricts access to food 514 Urban Ecosyst (2019) 22:513–523 Lynx lynx (Molinari-Jobin et al. 2002; Mattioli et al. 2004; Bateman and Fleming 2012) – are not present in urban areas. Predation by humans (hunting) is limited for reasons of safety; there is also greater awareness of / empathy for these animals on the part of the urban populace (Brown et al. 2000; Lee and Miller 2003; Warren 2011). and limits gene exchange, giving rise to a general deterioration in individual condition (Forman and Alexander 1998; Hewison et al. 2009; Seidler et al. 2015). Also, road traffic often leads to collisions with vehicles (Bruinderink and Hazebroek 1996; Forman and Alexander 1998; Zuberogoitia et al. 2014) and/or modifies the animals’ natural diurnal activ- ity (Kušta et al. 2017). Traffic noise also interferes with the animals’ vocal communication and may impact on their court- ship and vigilance (Nemeth et al. 2013; Klett-Mingo et al. 2016). Study area This study was carried out in the city of Kraków (southern Poland, 50°05’ N, 19°55′ E), which covers an area of 327 km2 and has a population density of 2321 persons/km2 (GUS 2016). Built-up areas cover around 6% of city’s overall area and are represented by an urbanization gradient – from the densely built-up city centre, through the suburbs with a mod- erate number of buildings to the scattered buildings typical of a rural landscape. Urban greenery is the predominant form of vegetation in the city (47%), consisting of gardens, squares, road verges and playgrounds, allotments and orchards, parks and cemeteries and other green areas. The next most common group of habitats consists of open areas (37%): arable land (14%), spontaneous vegetation on fallow land (13%), meadows and pastures (8%), wetland vegetation (2%). The remainder of the city’s green areas consists of forests and natural woodland (11%): natural and semi-natural scrub (5%), deciduous and mixed forest (4%) and damp, riparian forest and transformed tree stands (2%). Roads and railway lines make up 4% and surface waters just 1% of the city’s area (Dubiel and Szwagrzyk 2008). The principal waterway in Kraków is the River Wisła (Vistula); six medium-sized tribu- taries and numerous smaller watercourses flow into the Wisła within the city limits (MIIP 2016). The mean number of days Introduction We assessed the role of (a) habitat type within the urban land- scape and (b) noise and light pollution in the shaping of the species’ distribution pattern. We hypothesized that the pre- ferred habitat, which is woodland and/or open habitats in the case of roe deer (Mattioli 2011; Morellet et al. 2011), would be of primary importance. We predicted, however, that noise and light pollution – intense and ever-present in urban environ- ments – would tend to discourage roe deer from entering them and would be a factor governing the probability of their oc- curring in the urban environment. 515 Urban Ecosyst (2019) 22:513–523 Methods with snow cover varies between 60 and 70 days and its mean thickness is 20 cm (maximum recorded was 65 cm); however, both these parameters are showing a long-term decrease indic- ative of systematic climate warming (Falarz 1998). The urban heat island of Kraków is estimated to cover the city area and surrounding regions: there is an up to 1 cm difference in the average depth of snow cover between the city centre and the outskirts, but no differences in the mean number of days with snow cover (Falarz 1998; Arnfield 2003). During the two study years, there was not much snow cover during the winter: it was limited to 12 of winter days in 2013/2014 (mean snow cover thickness = 0.9 cm ±2.5 SD, maximum = 9 cm) and 28 of winter days in 2014/2015 (mean snow cover thickness = 1.5 cm ±2.7 SD, maximum = 13 cm). Selection of sample plots; roe deer counts Sixty sample plots on which roe deer were to be counted (Fig. 1) were selected at random using Quantum GIS software (QGIS 2013). In the first step, Kraków was divided into 389 1 km × 1 km squares from which the sample plots were drawn. The grid of squares was based on a point with coordinates 50°N 20°E. Then, every square was subdivided into four smaller squares of sides 500 m × 500 m (25 ha); one of these four was chosen at random for the counts. On every sample plot two transects each 500 m long were marked out. The ideal transect runs longitudinally or latitudinally, but obstacles in Fig. 1 Study area: distribution of sample plots in Kraków (S Poland) and the main habitat types used for assessing variables that influence the distribution pattern of roe deer Capreolus capreolus in an urban environment Fig. 1 Study area: distribution of sample plots in Kraków (S Poland) and the main habitat types used for assessing variables that influence the distribution pattern of roe deer Capreolus capreolus in an urban environment 516 Urban Ecosyst (2019) 22:513–523 the terrain (existing buildings, fences, walls) caused the real transects to deviate from the ideal. noise level expressed in 9 classes of sound intensity (dB) (1 – <45, 2 – 45-50, 3 – 50.1-55, 4 – 55.1-60, 5 – 60.1-65, 6 – 65.1-70, 7 – 70.1-75, 8 – 75.1-80, 9 – >80). The map was compiled jointly by the Provincial Environmental Conservation Inspectorate in Kraków and the Kraków City Council in 2012 (MIIP 2016). Roe deer were surveyed during winter in 2014 and 2015 and two surveys were carried out each year: early winter (03.01–30.01) and late winter (01.02–28.02). A period of at least two weeks had to elapse between consecutive surveys in a given year. Walking along the transects, the observer scanned a terrain through binocular (10 × 42) recording the presence of roe deer. To enhance the efficiency of detection, observers also searched for tracks leaved on snow/wet soil (if present), which indicates presence of a herd or an individual. The surveys were conducted during the daytime, though only on rain-free, snow-free and windless days (0–1 Beaufort’s scale), between 08:00 and 15:00 h CET. The transects were walked at an average speed of 2 km/h and single survey of a sample plot took around an hour. Selection of sample plots; roe deer counts Each of 60 sample plots was surveyed four times (two survey in each season). All roe deer individuals observed within a sample plot were counted and assigned to one of the distance zones delineated on either side of the transect: (a) 0–100 m and (b) >100 m. Sample plots were defined as occupied by roe deer if at least single individ- ual was recorded during one of the four surveys within a plot limits. Light pollution (LIGHT) was determined on the basis of the Visible Infrared Imaging Radiometer Suite (VIIRS) sup- plied by The Earth Observations Group (EOG 2016). The raster layer contained average radiance using night-time data from the VIIRS during January 2015 and was expressed in nanowatts per square centimetre per steradian (nW/cm2 × SR). The map’s resolution (grid pixel size) was ~460 m (EOG 2016). Sample plots applied in our study do not fit the grid cells of light pollution map, overlapping with up to nine cells of light pollution map. Therefore, the value of this variable was calculated as the average of 100 points located in the regular grid covering each of sample plots. Information on numbers of dogs (DOGS) was obtained during the transect counts. During each field survey all dogs seen within a sample plot limits were counted (free ranging, within enclosed properties or on a leash), and higher value recorded for two surveys (early winter and late winter) was taken as a number of dogs in a given season. Then, we calcu- lated the mean number of dogs based on the values obtained for both seasons (2014 and 2015). Environmental variables The habitat parameters were defined on the basis of existing spatial database resources using GIS tools and the fieldwork. The surface areas (ha) of the largest wooded habitat patch (WOOD_PATCH) and the total area of open habitats (OPEN_HABITAT) within a sample plot limits were calculat- ed using the polygon vector layer of the atlas of the real veg- etation of Kraków (UMK 2012), which is the effect of field- work done in 2006 (Dubiel and Szwagrzyk 2008). The area of largest wooded habitat patch (WOOD_PATCH) included de- ciduous forest, mixed woodland, naturally growing shrubs, parks and cemeteries. Open habitats (OPEN_HABITAT) in- cluded arable land, meadows, pastures, uncultivated and fal- low land, rock vegetation, swards, heaths and the communi- ties of trampled areas. Distance to rivers (RIVERS) was cal- culated by measuring the distance from the sample plot to the nearest line of the layer of rivers located within the study area (WODGiK 2015). The number of buildings (BUILDINGS) was calculated using the polygon vector layer of buildings (WODGiK 2015), summing the number of objects located within a sample plot limits. Results Roe deer were recorded on 36.7% of the randomly selected sample plots (N = 60). The median number of individuals re- corded on a sample plot was 2.5 (quartile range 1–6, range 1– 32) and the modal group size was 1. Occupied plots contained significantly more open habitats, proximity to rivers was sig- nificantly less there, and the area of the largest wooded habitat patch was greater (approaching significance level) (Table 1). The number of dogs did not differ significantly between oc- cupied and unoccupied plots (Table 1). All variables that are attributes of urbanization differed significantly between plots: the number of buildings, the level of noise emission and light pollution were significantly higher on unoccupied than occu- pied plots (Table 1). Since habitat characteristics could potentially affect the ability of researcher to detect roe deer, prior to analyses the effect of selected environmental variables on species detect- ability was tested. We assumed that increasing share of open habitats potentially increase detection distance due to higher visibility range, while increasing area of the largest wooded habitat patch and the number of buildings decrease detection distance due to reduced visibility. To check the influence of habitat characteristics on results of the surveys, we compared environmental variables between sample plots on which roe deer was detected in the different distance zones from the transect line. Based on the data combined from all surveys, sample plots were assigned into three groups: (a) plots on which roe deer was detected exclusively in the distance zone below 100 m, (b) plots on which roe deer was detected exclu- sively in the distance zone above 100 m, and (c) plots on The best supported model (ΔAICc = 0.0) explaining the probability of roe deer occurring contained two variables: area of open habitats and light pollution (Table 2). Logistic regres- sion models run separately for the two variables indicated in the best supported model revealed that the threshold values for a 0.5 occurrence probability were 8.8 ha of open habitats (corresponding to 35.2% of the plot area) and 10.3 nW/cm2 × SR of light pollution (Fig. 2). Data handling and analysis In the first step, mean (±SD) and median (with quartiles) values of each environmental variable were calculated for plots occupied and unoccupied by roe deer, the differences between the two groups being analysed using Mann- Whitney’s U test. Then factors determining the probability of roe deer occurring in an urban environment were investi- gated using a generalized linear model with binomial error distribution (Bolker et al. 2009). The area of open habitats (OPEN_HABITAT), the area of the largest wooded habitat patch (WOOD_PATCH), proximity to rivers (RIVERS), the number of dogs (DOGS), the number of buildings (BUILDINGS), noise emission (NOISE) and light pollution (LIGHT) were taken to be explanatory variables potentially explaining the presence of roe deer. To control for multicollinearity between variables, a correlation matrix was plotted (the correlation was <0.7 for all variable pairs; Table 1S, see Supplementary materials). The noise emission variable (NOISE) was determined from the road noise emission map (MIIP 2016). The mean noise class weighted by its range area was calculated for every sam- ple plot. Average noise levels during the hours of darkness (22:00–06:00 h) were used for these calculations. Noise level during the nighttime is highly correlated with noise level dur- ing the daytime (rS = 0.97, p = 0.000). The map shows the An information theory approach was used to identify and rank the best-supported model, and Akaike’s information cri- terion (AIC) was used for model selection (Burnham and Anderson 2002). Because of the relatively small sample size (n/K ratio <40), a small-sample version of AIC with bias ad- justment (AICc) was applied in the modelling. The resulting models were subsequently ranked in order of increasing AICc. 517 Urban Ecosyst (2019) 22:513–523 The differences between the model and the lowest AICc were calculated (ΔAICc) for each of the resulting models. which roe deer was detected in both zones. All analyzed en- vironmental variables did not differ significantly between dis- tinguished groups (OPEN_HABITAT: F2,19 = 0.05, p = 0.950; WOOD_PATCH: F2,19 = 0.31, p = 0.740; BUILDINGS: F2,19 = 0.79, p = 0.470), indicating that the distance in which animals were detected was not influenced by the habitat characteristics. In order to illustrate directions, estimates (±SE) and 95% confidence intervals (CI) of environmental variables, general- ized linear model with all variables included in the best- supported models (ΔAICc ≤5) were calculated. Data handling and analysis Then, logistic regression models (with plots occupied/unoccupied as a di- chotomous dependent variable) were run for the variables in- dicated in the linear model as being of major importance for the probability of roe deer occurring in order to detect thresh- old values determining the species’ presence. The statistical procedures were performed using Statistica 12 software (StatSoft Inc. 2014). Spatial autocorrelation of plots occupied by roe deer was assessed with Moran’s I tests (Rangel et al. 2010). We detected no evidence of spatial auto- correlation (Moran’s I was close to zero for all separation distances and semi-variance did not increase with lag distance). Finally, the percentage of deviance explained (100 × (deviancenull model – deviancemodel)/deviancenull model) was calculated for the three best supporting models (ΔAICc <5). Thereafter, deviance partitioning analysis (Borcard et al. 1992) was used to partition the deviance of each response vari- able included in the models into the independent effect of a particular predictor (pure effects) and the covarying effect of two or more predictors that cannot be disentangled (joint effects). This procedure allows one to calculate the percentage of the variation explained by each response variable and the joint effect of two or more variables not affected by the collinearity with other factor groups in the model (Legendre and Legendre 1998). Results The pure effect of light pollution was less important in explaining var- iability in occurrence probability (Fig. 4). However, the large deviance was explained by the joint effects of these variables: the probability of roe deer occurring increased with the area of open habitats not polluted by artificial light. analysis showed that the area of open habitats had the greatest effect on the probability of roe deer occurring (Fig. 4), its pure effect accounting for 27% of the total deviance. The pure effect of light pollution was less important in explaining var- iability in occurrence probability (Fig. 4). However, the large deviance was explained by the joint effects of these variables: the probability of roe deer occurring increased with the area of open habitats not polluted by artificial light. regression model (Table 3), a high occurrence probability (P ≥0.8) was recorded on both plots with high light pollution levels (~30 nW/cm2 × SR) but containing a large area of open habitats (~25 ha, corresponding to 100% of the plot area) and plots with a small area of open habitats (~5 ha, 20% of the area) but low light pollution levels (~0 nW/cm2 × SR) (Fig. 3). Two other, less-well supported models (ΔAICc = 0.4 and 3.6, see Table 2) indicated that the area of the largest wooded habitat patch was another potentially useful predictor for explaining the probability of roe deer occurring (Table 3). Results According to the logistic Table 1 Descriptive statistics and Mann-Whitney’s U-test results for variables analysed in sample plots unoccupied and occupied by roe deer Capreolus capreolus in an urban environment (Kraków, S Poland; for variables, see Methods) Variable Unoccupied Occupied Zc p Mean SD Median Quartile range Mean SD Median Quartile range OPEN_ HABITAT 2.3 3.4 0.3 0.0–4.4 16.0 6.9 19.2 9.8–21.9 −5.94 0.000 WOOD_PATCH 5.5 11.2 0.0 0.0–2.4 12.1 20.0 1.2 0.0–15.0 −1.81 0.070 RIVERS 0.7 0.8 0.3 0.0–1.2 0.2 0.4 0.0 0.0–0.1 2.68 0.007 DOGS 4.9 4.1 4.5 1.5–7.0 4.2 4.0 3.5 1.0–6.0 0.71 0.475 BUILDINGS 82.6 45.6 72.0 54.0–122.0 34.9 35.3 21.0 8.0–49.0 3.99 0.000 NOISE 2.2 0.9 2.0 1.5–2.9 1.7 0.8 1.3 1.0–2.0 2.81 0.005 LIGHT 25.2 13.7 25.1 12.4–33.0 6.6 5.8 4.4 2.9–8.7 5.43 0.000 518 Urban Ecosyst (2019) 22:513–523 Table 3 Estimates (±SE) and 95% confidence intervals (CI) of all en- vironmental variables included in the best-supported models (ΔAICc ≤5) predicting the probability of roe deer Capreolus capreolus occurring in an urban environment (Kraków, S Poland; for variables, see Methods) Table 2 Sets of candidate models explaining the probability of roe deer Capreolus capreolus occurring in an urban environment (Kraków, S Poland; for variables, see Methods). Akaike’s information criterion for small samples (AICc) and the difference between the given model and the most parsimonious model (ΔAICc) are reported for each model Table 2 Sets of candidate models explaining the probability of roe deer Capreolus capreolus occurring in an urban environment (Kraków, S Poland; for variables, see Methods). Akaike’s information criterion for small samples (AICc) and the difference between the given model and the most parsimonious model (ΔAICc) are reported for each model Variable Estimate SE Wald’s stat. 95% CI p INTERCEPT −2.17 1.98 1.21 −6.05 to 1.70 0.271 OPEN_ HABITAT 0.48 0.20 5.61 0.08 to 0.88 0.018 LIGHT −0.23 0.13 2.98 −0.49 to 0.03 0.084 WOOD_PATCH 0.04 0.03 1.63 −0.02 to 0.11 0.201 Model AICc ΔAICc OPEN_HABITAT + LIGHT 29.4 0.0 OPEN_HABITAT + WOOD_PATCH + LIGHT 29.8 0.4 OPEN_HABITAT + WOOD_PATCH 33.0 3.6 OPEN_HABITAT 34.4 5.0 ALL VARIABLES 36.2 6.8 LIGHT 46.5 17.2 BUILDINGS 65.9 36.5 RIVERS 75.6 46.2 NOISE 76.4 47.0 WOOD_PATCH 80.4 51.1 INTERCEPT 80.9 51.5 DOGS 82.7 53.3 analysis showed that the area of open habitats had the greatest effect on the probability of roe deer occurring (Fig. 4), its pure effect accounting for 27% of the total deviance. Discussion 3 Logistic regression model predicting the probability of roe deer Capreolus capreolus occurring based on the area of open habitat (ha) and light pollu- tion (nW/cm2 × SR) in an urban environment (Kraków, S Poland; for variables, see Methods). The occurrence probability function is p = e(0.37y-0.24x-0.61)/(1 + e(0.37y- 0.24x-0.61)), where y and x repre- sent the area of open habitat and light pollution, respectively d are lacking, roe deer will avail them- hedgerows (Morellet et al. 2011), which alment, particularly in thickly populated rud et al. 1999; Bonnot et al. 2013). season, wooded areas are probably less important to roe deer, because leafless trees and shrubs do not provide effective camouflage (Putman 1986; Mysterud et al. 1999; Hewison et al. 2009; Bonnot et al. 2013). Moreover, during spells of bad weather in winter, roe deer have greater energy requirements, so they may spend more time foraging in n model y of roe s rea of ht pollu- n urban Poland; ds). The nction is (0.37y- repre- itat and ely important to roe deer, because leafless trees and shrubs do not provide effective camouflage (Putman 1986; Mysterud et al. 1999; Hewison et al. 2009; Bonnot et al. 2013). Moreover, during spells of bad weather in winter, roe deer have greater energy requirements, so they may spend more time foraging in Fig. 3 Logistic regression model predicting the probability of roe deer Capreolus capreolus occurring based on the area of open habitat (ha) and light pollu- tion (nW/cm2 × SR) in an urban environment (Kraków, S Poland; for variables, see Methods). The occurrence probability function is p = e(0.37y-0.24x-0.61)/(1 + e(0.37y- 0.24x-0.61)), where y and x repre- sent the area of open habitat and light pollution, respectively important to roe deer, because leafless trees and shrubs do not provide effective camouflage (Putman 1986; Mysterud et al. 1999; Hewison et al. 2009; Bonnot et al. 2013). Moreover, during spells of bad weather in winter, roe deer have greater energy requirements, so they may spend more time foraging in important to roe deer, because leafless trees and shrubs do not provide effective camouflage (Putman 1986; Mysterud et al. 1999; Hewison et al. 2009; Bonnot et al. 2013). Fig. 3 Logistic regression model predicting the probability of roe deer Capreolus capreolus occurring based on the area of open habitat (ha) and light pollu- tion (nW/cm2 × SR) in an urban environment (Kraków, S Poland; for variables, see Methods). The occurrence probability function is p = e(0.37y-0.24x-0.61)/(1 + e(0.37y- 0.24x-0.61)), where y and x repre- sent the area of open habitat and light pollution, respectively Discussion This study has shown that the probability of roe deer occurring in an urban matrix is positively correlated with an increasing area of open habitats (grassland and arable land). Roe deer, originally a woodland species (Mattioli 2011), exhibits con- siderable flexibility where habitat choice is concerned: it oc- curs in both woodland and in open terrain, and also in the ecotone, i.e. the boundary zone in between (Hewison et al. 2001; Morellet et al. 2011; Lovari et al. 2017). In predomi- nantly open habitats, roe deer tend to remain close to clumps of trees (Hewison et al. 2001; Morellet et al. 2011; Lovari et al. 2017), the presence of which determines whether this species will inhabit farmland or not (Acevedo et al. 2005). Where The best supported model (ΔAICc = 0.0) of the probability of roe deer occurring, including the area of open habitats and light pollution, explained 70.9% of the deviance in the dataset, a value that increased to 73.3% when the area of the largest wooded habitat patch was added. Deviance partitioning Fig. 2 Relationship between the area of open habitat (left; ha) and light pollution (right; nW/cm2 × SR) and the probability of roe deer Capreolus capreolus occurring in an urban environment (Kraków, S Poland; for variables, see Methods). The occurrence probability functions are: p = e(0.41y-3.62)/(1 + e(0.41y-3.62)) and p = e(−0.25x + 2.57)/(1 + e(−0.25x + 2.57)), where y and x represent the area of open habitat and light pollution, respectively. Grey shading indicates 95% confidence intervals e(0.41y-3.62)/(1 + e(0.41y-3.62)) and p = e(−0.25x + 2.57)/(1 + e(−0.25x + 2.57)), where y and x represent the area of open habitat and light pollution, respectively. Grey shading indicates 95% confidence intervals e(0.41y-3.62)/(1 + e(0.41y-3.62)) and p = e(−0.25x + 2.57)/(1 + e(−0.25x + 2.57)), where y and x represent the area of open habitat and light pollution, respectively. Grey shading indicates 95% confidence intervals Fig. 2 Relationship between the area of open habitat (left; ha) and light pollution (right; nW/cm2 × SR) and the probability of roe deer Capreolus capreolus occurring in an urban environment (Kraków, S Poland; for variables, see Methods). The occurrence probability functions are: p = 519 Urban Ecosyst (2019) 22:513–523 Fig. Discussion Moreover, during spells of bad weather in winter, roe deer have greater energy requirements, so they may spend more time foraging in patches of woodland are lacking, roe deer will avail them- selves of shrubs and hedgerows (Morellet et al. 2011), which provide shelter/concealment, particularly in thickly populated urban areas (Mysterud et al. 1999; Bonnot et al. 2013). Outside the growing season, wooded areas are probably less Fig. 4 Deviance partitioning analysis for the probability of roe deer Capreolus capreolus occurring in an urban environment (Kraków, S Poland); the percentage of variability in the probability of occurrence is explained by the pure (A – area of open habitat, B – light pollution, C – largest wooded habitat patch) and joint effects of habitat variables 520 Urban Ecosyst (2019) 22:513–523 budget (Foster and Provencio 1999; Biebouw and Blumstein 2003; Barber-Meyer 2007) or causing them to avoid areas affected by artificial light (Robert et al. 2015). Our work has shown that light pollution could be used as an additional pre- dictor of ungulate occurrence: there is therefore an urgent need for further studies of the influence of artificial light on the ecology of this group of animals. farmland, which is less secure but richer in food (Putman 1986; Mysterud et al. 1999). During this season, therefore, farmland may be a key habitat, supporting the viability of a population. p p Artificial light is often used as a predictor for the occurrence/density of animals, particularly birds and bats (Azam et al. 2016; Ciach and Fröhlich 2017; Froidevaux et al. 2017). Because they are frequently active at night, un- gulates often enter areas that are affected by artificial lighting (Hewison et al. 2009; Pagon et al. 2013), so their populations are potentially affected by it (Beier 2006; Hölker et al. 2010). To the best of our knowledge, however, the present study is the first one to assess the correlation between artificial lighting at night and the distribution of ungulate population. The oc- currence of roe deer was less probable in areas with high intensities of artificial lighting: this parameter explained the occurrence of this species better than the number of buildings or noise levels. Fig. 5 Examples of sample plots located within an urban environment illustrating potential habitats of roe deer Capreolus capreolus in an urban environment (Kraków, S Poland); plot selection based on threshold values for 0.5 occurrence probability – 8.8 ha for open habitats (corresponding to 35.2% of the plot area) and 10.3 nW/ cm2 × SR for light pollution; sat- ellite images taken from Google Earth, supplied by Digital Globe – the respective coordinates of the centres of the images are: 50°3′ 57.72″N 19°52′19.08″E, 50°4′ 30.54″N 19°51′53.58″E, 50°2′ 36.94″N 19°53′31.84″E, 50°1′ 47.93″N 19°51′51.01″E; eye alti- tudes 1.1 km Discussion Although the mechanism of avoiding artificial light has not been well documented, experimental modifica- tions of the photoperiod have shown that artificial light can lead to hormonal changes in ungulates that alter their repro- ductive cycles as well as moult strategies for the maintenance of suitable body temperatures (Yeates 1949; Lincoln and Guinness 1972). The adverse effects of light pollution have also been demonstrated in other mammals: the extended pho- toperiod in urban areas lit up at night led to changes in their diurnal activity, thereby significantly affecting their time The logistic regression model presented in this paper indicates that the probability of roe deer occurring on farm- land not affected by artificial lighting is increasing (Fig. 5). Although the drop in quality of agricultural habitats ex- posed to artificial lighting has been confirmed with respect to some other species (De Molenaar and Sanders 2006; Bennie et al. 2017; Froidevaux et al. 2017), knowledge of this subject with regard to large mammals is still lacking. Since farmland is intensively utilized by humans during the growing season, ungulates tend to avoid this habitat at that time, but do make use of it in winter, when woodlands do not supply sufficient food resources (Putman 1986; Hewison et al. 2009). In the vicinity of human habitations, however, ungulates are mostly active at night, which is probably a strategy helping to avoid contact with humans (Mysterud et al. 1999; Bonnot et al. 2013; Pagon et al. 2013). Animals may equate the presence of artificial light- ing in farming habitats with the presence of humans: per- ceiving this as a threat, they will tend to avoid illuminated areas (De Molenaar and Sanders 2006; Robert et al. 2015). Fig. 5 Examples of sample plots located within an urban environment illustrating potential habitats of roe deer Capreolus capreolus in an urban environment (Kraków, S Poland); plot selection based on threshold values for 0.5 occurrence probability – 8.8 ha for open habitats (corresponding to 35.2% of the plot area) and 10.3 nW/ cm2 × SR for light pollution; sat- ellite images taken from Google Earth, supplied by Digital Globe – the respective coordinates of the centres of the images are: 50°3′ 57.72″N 19°52′19.08″E, 50°4′ 30.54″N 19°51′53.58″E, 50°2′ 36.94″N 19°53′31.84″E, 50°1′ 47.93″N 19°51′51.01″E; eye alti- tudes 1.1 km Fig. Discussion 5 Examples of sample plots located within an urban environment illustrating potential habitats of roe deer Capreolus capreolus in an urban environment (Kraków, S Poland); plot selection based on threshold values for 0.5 occurrence probability – 8.8 ha for open habitats (corresponding to 35.2% of the plot area) and 10.3 nW/ cm2 × SR for light pollution; sat- ellite images taken from Google Earth, supplied by Digital Globe – the respective coordinates of the centres of the images are: 50°3′ 57.72″N 19°52′19.08″E, 50°4′ 30.54″N 19°51′53.58″E, 50°2′ 36.94″N 19°53′31.84″E, 50°1′ 47.93″N 19°51′51.01″E; eye alti- tudes 1.1 km 521 Urban Ecosyst (2019) 22:513–523 The results of our study indicate that there was a lower intensity of noise in the areas colonized by roe deer, al- though this was not a key factor in the population distribu- tion model. Noise, the main source of which in urban areas is road traffic, interferes with the propagation of sound waves and seriously affects the distribution of animal popu- lations in which individuals communicate vocally with one another (Nemeth et al. 2013; Ciach and Fröhlich 2017) and foraging (Siemers and Schaub 2011; Mason et al. 2016) relies on acoustic signals. However, the results of studies conducted on small mammals indicate that traffic noise is not considered a factor leading to avoidance of roads (McGregor and Bender 2008). Roads and the traffic they carry are more likely to bar the movements of terrestrial mammals (Forman and Alexander 1998; Seidler et al. 2015), which, moreover, often fall victim to collisions with vehicles (Bruinderink and Hazebroek 1996; Zuberogoitia et al. 2014). The presence of roads does also mean, however, that ungulates limit their activities near them during periods of heavy traffic (Kušta et al. 2017). This mechanism pro- vides evidence that ungulates have to some extent adapted to inhabiting areas with a dense road network (McCarthy et al. 1996; Acevedo et al. 2005; Underwood and Kilheffer 2016; Loro et al. 2016; Kušta et al. 2017), and that noise in itself is probably not the most important factor governing the distri- bution of their populations (Figs. 5 and 6). Our study has shown that the areas colonized by roe deer lie close to rivers. Rivers and their associated riparian forests act as migration corridors (Romanowski 2007; Rodriguez- Iturbe et al. 2009). Discussion This strongly suggests, in turn, an indirect effect of light pollution on environment, where changes in species distributions can give rise to changes in the habitats that are utilized or avoided. Our results demonstrate that light pollution may be an urban effect shaping the spatial distribution of large mammals. Bateman PW, Fleming PA (2012) Big city life: carnivores in urban envi- ronments. J Zool 287:1–23 Beier P (2006) Effects of artificial night lighting on terrestrial mammals. In: Rich C, Longcore T (eds) Ecological consequences of artificial night lighting. Island press, Washington. Covelo, London, pp 19–42 Bennie J, Davies TW, Cruse D et al (2017) Artificial light at night alters grassland vegetation species composition and phenology. J Appl Ecol. https://doi.org/10.1111/1365-2664.12927 Biebouw K, Blumstein DT (2003) Tammar wallabies ( Macropus eugenii ) associate safety with higher levels of nocturnal illumination. Ethol Ecol Evol 15:159–172 Bolker BM, Brooks ME, Clark CJ et al (2009) Generalized linear mixed models: a practical guide for ecology and evolution. Trends Ecol Evol 24:127–135 Bonnot N, Morellet N, Verheyden H (2013) Habitat use under predation risk: hunting, roads and human dwellings influence the spatial be- haviour of roe deer. Eur J Wildl Res 59:185–193 Boone RB, Hobbs NT (2004) Lines around fragments: effects of fencing on large herbivores. African J Range Forage Sci 21:147–158 Acknowledgements We wish to express our gratitude to members of the Foresters’ Scientific Club of the Faculty of Forestry in Kraków for their help with the fieldwork. Financial support for this study was provided by the Polish Ministry of Science and Higher Education by statutory grant (DS 3421). Borcard D, Legendre P, Drapeau P (1992) Partialling out the spatial com- ponent of ecological variation. Ecology 73:1045–1055 Brown T, Decker D, Riley S, Enck J (2000) The future of hunting as a mechanism to control white-tailed deer populations. Wildl Soc 28: 797–807 Author contributions statement AF conceived the idea and coordinated the collection of the data, AF and MC designed the study and participated in the fieldworks; MC analyzed the data; MC and AF wrote the manuscript. Bruinderink GWTAG, Hazebroek E (1996) Ungulate traffic collisions in Europe. Conserv Biol 10:1059–1067 Burnham KP, Anderson DR (2002) Model selection and multi-model inference: a practical-theoretical approach. Springer, New York Ciach M, Fröhlich A (2017) Habitat type, food resources, noise and light pollution explain the species composition, abundance and stability of a winter bird assemblage in an urban environment. Discussion Despite the widespread and far-reaching transformation of riparian environments, watercourses in highly urbanized environments are often the only corridors along which animals can penetrate them (May 2006; Ignatieva et al. 2011). Maintaining urban watercourses in as natural a state as possible will therefore probably favour the colonization of towns and cities by large mammals. This research has also shown that the number of dogs was unimportant as a factor affecting the probability of roe deer occurring. The dog is most closely related to the wolf, the natural predator of medium-sized and large ungulates (Mattioli et al. 2004). The barking of dogs scares off ungulates, which can affect the time budget of the latter and increase their energy expenditure (Reimoser 2012; Padié et al. 2015). Most dogs in urban areas are kept indoors or within enclosed properties, while those moving about the city are on a leash. Since there are no feral and/or free-ranging dogs, ungulates to some extent simply ignore the presence of these animals (Padié et al. 2015); again, this is probably another adaptation to living near human habita- tions, where pet dogs are ubiquitous (McCarthy et al. 1996; Kilpatrick and Spohr 2000; Loro et al. 2016). We have shown that the probability of roe deer occurring in urban environments is correlated with the presence of open Fig. 6 Winter habitats of roe deer Capreolus capreolus in an urban environment (Kraków, S Poland) Fig. 6 Winter habitats of roe deer Capreolus capreolus in an urban environment (Kraków, S Poland) 522 Urban Ecosyst (2019) 22:513–523 Barber-Meyer SM (2007) Photopollution impacts on the nocturnal be- haviour of the sugar glider (Petaurus breviceps). Pacific Conserv Biol 13:171–176 habitats and light pollution. To date, studies of the influence of urbanization on ungulates analysed only the area built up or the distance to buildings (Hewison et al. 2001; Torres et al. 2011; Underwood and Kilheffer 2016; Loro et al. 2016). This paper is the first to demonstrate that light pollution can sub- stantially affect the probability of roe deer occurring; conse- quently, artificial lighting is likely to be an additional factor discouraging ungulates from colonizing potentially suitable habitats in urban areas. The roe deer is one of the commonest grazing animals in the temperate zone and its presence/ absence may therefore affect the vegetation structure. Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest. De Molenaar J, Sanders M (2006) Road lighting and grassland birds: local influence of road lighting on a black-tailed godwit population. In: Rich C, Longcore T (eds) Ecological Consequences of Artificial Night Lighting. Island Press, Washington D.C., pp 114–136 Ethical approval All applicable institutional and national guidelines for the care and use of animals were followed. Ethical approval All applicable institutional and national guidelines for the care and use of animals were followed. Night Lighting. Island Press, Washington D.C., pp 114–136 Dubiel E, Szwagrzyk J (2008) Atlas Roślinności Rzeczywistej Krakowa. In: Urząd Miasta Krakowa. Wydział Kształtowania Środowiska, Kraków 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. EOG (The Earth Observations Group) (2016) Nighttime VIIRS Day/ Night Band Composites. http://ngdc.noaa.gov/eog/viirs/download_ monthly.html. Accessed 28 Mar 2015 Ewald M, Dupke C, Heurich M et al (2014) LiDAR remote sensing of Forest structure and GPS telemetry data provide insights on winter habitat selection of European roe deer. Forests 5:1374–1390 Falarz M (1998) Long-term variability of snow cover in Cracow and suburban areas. Folia Geogr Phys 3:473–481 Discussion Urban Ecosyst 20:547–559 References Chronobiol Int 30: 772–785 Hölker F, Wolter C, Perkin EK, Tockner K (2010) Light pollution as a biodiversity threat. Trends Ecol Evol 25:681–682 Ignatieva M, Stewart GH, Meurk C (2011) Planning and design of eco- logical networks in urban areas. Landsc Ecol Eng 7:17–25 Putman RJ (1986) Foraging by roe deer in agricultural areas and impact on arable crops. J Appl Ecol 23:91–99 Kilpatrick H, Spohr S (2000) Spatial and temporal use of a suburban land- scape by female white-tailed deer. Wildl Soc Bull 28:1023–1029 Putman RJ, Moore NP (1998) Impact of deer in lowland Britain on agricul- ture, forestry and conservation habitats. Mamm Rev 28:141–164 Klett-Mingo IJ, Pavón I, Gil D (2016) Great tits, Parus major, increase vigilance time and reduce feeding effort during peaks of aircraft noise. Anim Behav 115:29–34 QGIS (Quantum GIS Development Team) (2013) Quantum GIS geographic information system. In: Open source Geospatial Foundation project Rangel TF, Diniz-Filho JAF, Bini LM (2010) SAM: a comprehensive appli- cation for spatial analysis in macroecology. Ecography 33:46–50 Kušta T, Keken Z, Ježek M et al (2017) The effect of traffic intensity and animal activity on probability of ungulate-vehicle collisions in the Czech Republic. Saf Sci 91:105–113 Reimoser S (2012) Influence of anthropogenic disturbances on activity, behavior and heart rate of roe deer (Capreolus capreolus) and Red Deer (Cervus elaphus), in context of their daily and yearly patterns. In: Cahler AA, Marsten JP (eds) Deer: Habitat. Behavioral and Conservation. Nova Science Publisher, New York, pp 1–95 Lee M, Miller R (2003) Managing elk in the wildland-urban interface: attitudes of Flagstaff. Arizona residents Wildl Soc Bull 31:185–191 Legendre P, Legendre L (1998) Numerical ecology. 3rd Edn. Elsevier publishers, Amsterdam, Holland Robert KA, Lesku JA, Partecke J, Chambers B (2015) Artificial light at night desynchronizes strictly seasonal reproduction in a wild mam- mal. Proc Royal Soc B 282. https://doi.org/10.1098/rspb.2015.1745 Lincoln GA, Guinness FE (1972) Effect of altered photoperiod on delayed implantation and moulting in roe deer. J Reprod Fertil 31:455–457 Longcore T, Rich C (2004) Ecological light pollution. Front Ecol Environ 2:191–198 Rodriguez-Iturbe I, Muneepeerakul R, Bertuzzo E et al (2009) River networks as ecological corridors: a complex systems perspective for integrating hydrologic, geomorphologic, and ecologic dynamics. Water Resour Res. https://doi.org/10.1029/2008WR007124 Loro M, Ortega E, Arce RM, Geneletti D (2016) Assessing landscape resis- tance to roe deer dispersal using fuzzy set theory and multicriteria anal- ysis: a case study in Central Spain. References Landsc Ecol Eng 12:41–60 Water Resour Res. https://doi.org/10.1029/2008WR007124 Romanowski J (2007) Vistula river valley as the ecological corridor for mammals. Polish J Ecol 55:805–819 Lovari S, Serrao G, Mori E (2017) Woodland features determining home range size of roe deer. Behav Process 140:115–120 Seidler RG, Long RA, Berger J et al (2015) Identifying impediments to long-distance mammal migrations. Conserv Biol 29:99–109 Mason TJ, McClure CJW, Barber JR (2016) Anthropogenic noise impairs owl hunting behavior. Biol Conserv 199:29–32 Siemers BM, Schaub A (2011) Hunting at the highway: traffic noise reduces foraging efficiency in acoustic predators. Proc Biol Sci 278:1646–1652 Mattila M, Hadjigeorgiou I (2015) Conservation and management of fallow deer (Dama dama dama L.) on Lemnos Island, Greece. Turkish J Vet Anim Sci 39:560–567 StatSoft Inc. (2014) Statistica (data analysis software system), version 10 Mattioli S (2011) Family Cervidae (deer). In: Wilson DE, Mittermeier RA (eds) Handbook of the mammals of the world, vol 2. Hoofed Mammals. Lynx edicions, Barcelona, pp 350–443 Torres RT, Carvalho JC, Panzacchi M et al (2011) Comparative use of forest habitats by roe deer and moose in a human-modified land- scape in southeastern Norway during winter. Ecol Res 26:781–789 Mattioli L, Capitani C, Avanzinelli E et al (2004) Predation by wolves (Canis lupus) on roe deer (Capreolus capreolus) in north-eastern Apennine, Italy. J Zool 264:249–258 UMK (Urząd Miasta Krakowa) (2012) Mapa roślinności rzeczywistej miasta Krakowa. Urząd Miasta Krakowa, Kraków Underwood H, Kilheffer CR (2016) Modeling the effects of land cover and use on landscape capability for urban ungulate populations. In: Francis RA, Millington JDA, Chadwick MA (eds) Urban landscape ecology: science. Policy and Practice. Routledge, London and New York, pp 181–208 p y May R (2006) BConnectivity^ in urban rivers: conflict and convergence between ecology and design. Technol Soc 28:477–488 McCarthy A, Baker A, Rotherham I (1996) Urban-fringe deer manage- ment issues-a South Yorkshire case study. Br Wildl 8:12–19 Mcdonald RI, Kareiva P, Forman RTT (2008) The implications of current and future urbanization for global protected areas and biodiversity conservation. Biol Conserv 141:1695–1703 Warren R (2011) Deer overabundance in the USA: recent advances in population control. Anim Prod Sci 51:259–266 Welch D, Staines BW, Scott D et al (1991) Leader browsing by red and roe deer on young Sitka spruce trees in Western Scotland I. Damage rates and the influence of habitat factors. References Forman RTT, Alexander LE (1998) Roadfs and their major ecological effects. Annu Rev Ecol Syst 29:207–231 Acevedo P, Delibes-Mateos M, Escudero MA et al (2005) Environmental constraints in the colonization sequence of roe deer (Capreolus capreolus Linnaeus, 1758) across the Iberian Mountains, Spain. J Biogeogr 32:1671–1680 Foster RG, Provencio I (1999) The regulation of vertebrate biological clocks by light. In: Djamgoz MBA, Loew ER et al (eds) Archer SN. Adaptive Mechanisms in the Ecology of Vision. Springer, Dordrecht, pp 223–243 Acevedo P, Farfán MÁ, Márquez AL et al (2011) Past, present and future of wild ungulates in relation to changes in land use. Landsc Ecol 26:19–31 Found R, Boyce MS (2011) Predicting deer–vehicle collisions in an ur- ban area. J Environ Manag 92:2486–2493 Arnfield AJ (2003) Two decades of urban climate research: a review of turbulence, exchanges of energy and water, and the urban heat is- land. Int J Climatol 23:1–26 Froidevaux JSP, Boughey KL, Barlow KE, Jones G (2017) Factors driv- ing population recovery of the greater horseshoe bat (Rhinolophus ferrumequinum) in the UK: implications for conservation. Biodivers Conserv 26:1601–1621 Azam C, Le Viol I, Julien J-F et al (2016) Disentangling the relative effect of light pollution, impervious surfaces and intensive agriculture on bat activity with a national-scale monitoring program. Landsc Ecol 31:2471–2483 GUS (Główny Urząd Statystyczny) (2016) Powierzchnia i ludność w przekroju terytorialnym w 2015 r. Główny Urząd Statystyczny, Warszawa 523 Urban Ecosyst (2019) 22:513–523 Harrington JL, Conover MR (2006) Characteristics of ungulate behavior and mortality associated with wire fences. Wildl Soc Bull 34:1295–1305 Nemeth E, Pieretti N, Zollinger SA et al (2013) Bird song and anthropo- genic noise: vocal constraints may explain why birds sing higher- frequency songs in cities. Proc R Soc B Biol Sci 280:2012–2798 Hewison AM, Vincent JP, Joachim J et al (2001) The effects of woodland fragmentation and human activity on roe deer distribution in agri- cultural landscapes. Can J Zool 79:679–689 Padié S, Morellet N, Cargnelutti B et al (2015) Time to leave? Immediate response of roe deer to experimental disturbances using playbacks. Eur J Wildl Res 61:871–879 Hewison AJM, Morellet N, Verheyden H et al (2009) Landscape fragmenta- tion influences winter body mass of roe deer. Ecography 32:1062–1070 Pagon N, Grignolio S, Pipia A et al (2013) Seasonal variation of activity patterns in roe deer in a temperate forested area. References Forestry 64:61–82 McGregor R, Bender D (2008) Do small mammals avoid roads because of the traffic? J Appl Ecol 45:117–123 MIIP (Małopolska Infrastruktura Informacji Przestrzennej) (2016) Geoportal MIIP. http://miip.geomalopolska.pl/imap/. Accessed 28 Mar 2016 WODGiK (Wojewódzki Ośrodek Dokumentacji Geodezyjnej i Kartograficznej) (2015) Baza danych obiektów topograficznych. http://www.geomalopolska.pl/. Accessed 28 Mar 2015 Molinari-Jobin A, Molinari P, Breitenmoser-Würsten C, Breitenmoser Molinari-Jobin U (2002) Significance of lynx Lynx lynx predation for roe deer Capreolus capreolus and chamois Rupicapra rupicapra mortality in the Swiss Jura Mountains. Wildlife Biol 8:109–115 Yeates N (1949) The breeding season of the sheep with particular reference to its modification by artificial means using light. J Agric Sci 39:1–43 Zuberogoitia I, del Real J, Torres JJ et al (2014) Ungulate vehicle colli- sions in a Peri-urban environment: consequences of transportation infrastructures planned assuming the absence of ungulates. PLoS One 9:e107713 Morellet N, Van Moorter B, Cargnelutti B et al (2011) Landscape com- position influences roe deer habitat selection at both home range and landscape scales. Landsc Ecol 26:999–1010 Mysterud A, Lian L-B, Hjermann DØ (1999) Scale-dependent trade-offs in foraging by European roe deer ( Capreolus capreolus ) during winter. Can J Zool 77:1486–1493
https://openalex.org/W4306741921	https://www.mdpi.com/2227-9032/10/10/2071/pdf?version=1666664624	English	null	The Relationship between Mustard Import and COVID-19 Deaths: A Workflow with Cross-Country Text Mining	Healthcare	2,022	cc-by	8,055	healthcare healthcare healthcare healthcare Ge Zhan 1 , Fuming Yang 2, Liangbo Zhang 3,* and Hanfeng Wang 2 Ge Zhan 1 , Fuming Yang 2, Liangbo Zhang 3,* and Hanfeng Wang 2 1 AI Data Analytics Lab, Beijing Normal University-Hong Kong Baptist University (BNU-HKBU United International College), Zhuhai 519087, China 2 Division of Science & Technology, Beijing Normal University-Hong Kong Baptist University (BNU-HKBU United International College), Zhuhai 519087, China g ) 3 School of Economics and Management, Harbin Institute of Technology Shenzhen, Shenzhen 518000, China d b h d 3 School of Economics and Management, Harbin Institute of Technology Shenzhen, Shenzhen 518000, China * Correspondence: 20b957003@stu.hit.edu.cn Abstract: We developed a workﬂow for the search and screening of natural products by drawing from worldwide experiences shared by online platform users, illustrated how to cope with COVID- 19 with a text-mining approach, and statistically tested the natural product identiﬁed. We built a knowledge base, which consists of three ontologies pertaining to 7653 narratives. Mustard emerged from texting mining and knowledge engineering as an important candidate relating to COVID-19 outcomes. The ﬁndings indicate that, after controlling for the containment index, the net import of mustard is related with reduced total and new deaths of COVID-19 for the non-vaccination time period, with considerable effect size (>0.2). Keywords: text-mining; knowledge graph; COVID-19; natural foods Citation: Zhan, G.; Yang, F.; Zhang, L.; Wang, H. The Relationship between Mustard Import and COVID-19 Deaths: A Workﬂow with Cross-Country Text Mining. Healthcare 2022, 10, 2071. https:// doi.org/10.3390/healthcare10102071 Academic Editor: Guanghua Han Received: 10 September 2022 Accepted: 8 October 2022 Published: 18 October 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. 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/) 1. Introduction Many developing economies have inadequate doses to vaccinate their populations [1]. Countries with high vaccination rates are experiencing a new round of the outbreak, largely caused by new COVID-19 variants such as Delta and Lambda. Recent research indicates the potential for COVID-19 variants to escape from neutralizing humoral immunity [2,3], and the effectiveness of vaccination has been found to be lower among people with the Delta variant than those with the Alpha variant [4]. There is an urgent need for non-vaccination interventions against this disease [1]. g Drug-repurposing opportunities identiﬁed in previous studies need to be evaluated over time, and clinical trials are still in progress [5,6]. While there are a huge amount of COVID-19 research works published on drug discovery, few studies have suggested exploiting potential natural products [7]. Previous works on natural product identiﬁcation have proposed approaches mining genome and metabolomics data [8,9]. However, such methods are limited by the scope of their search and are usually time-consuming. A more efﬁcient or novel way of gaining insights from data in healthcare research is mining and analyzing online text with natural language processing (NLP) or text analytics [10]. For example, researchers of recent healthcare studies have learned public opinion and sentiment on topics such as COVID-19 vaccine boosters [11] and intimate partner violence [12] by mining textual data from social media. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. 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/). In this study, we developed a new workﬂow for the search and screening of natural products from 7653 narratives that can be potentially used to cope with COVID-19. To illustrate this approach, we explored online narrative texts collected from a large tourism platform, identiﬁed a potential natural food with a knowledge graph approach, and tested it statistically. We hypothesize that natural food consumption is related to COVID-19 outcomes. We then test the data on COVID-19 deaths and cases drawn from John Hopkins University (JHU) database. https://www.mdpi.com/journal/healthcare Healthcare 2022, 10, 2071. https://doi.org/10.3390/healthcare10102071 Healthcare 2022, 10, 2071 2 of 10 eas des nto the The number of medicine discovery projects on the basis of big data and data mining has been growing in recent years [13,14]. 2.1. A Workﬂow with NPN Approach tion such as closing schools and community transmission during C 2.1. A Workﬂow with NPN Approach tion such as closing schools and community transmission during C As the workﬂow (Figure 1) is used to isolate the best cases of countries from natural products narratives (NPN), which may provide us hints on natural products, we deleted countries with highly restrict responses to COVID-19. As successful government inter- vention such as closing schools and banning gatherings are highly effective at controlling community transmission during COVID-19 [19–21], countries with a lower containment and health index are considered more promising for the detection of natural solutions. We used the Containment and Health Index, which was developed by the Oxford Coronavirus Government Response Tracker (OxCGRT) project, and a higher value of the index indicates a stricter response to the pandemic (100 for strictest response). The index is based on thirteen policy-response indicators including school closures, workplace closures, travel bans, testing policy, contact tracing, face coverings, and vaccine policy. If countries with low containment observe fewer COVID-19 deaths or cases, there might be some unknown causes. Countries that did not achieve 50% in the COVID Data Transparency Index (total- analysis.com/Covid19/TAIndex, accessed on 5 August 2021) were deleted in data mining, as the numbers of deaths and conﬁrmed cases in these countries might not be accurate. community transmission during COVID-19 [19–21], countries with a lower containm and health index are considered more promising for the detection of natural solutions used the Containment and Health Index, which was developed by the Oxford Cor virus Government Response Tracker (OxCGRT) project, and a higher value of the in ndicates a stricter response to the pandemic (100 for strictest response). The index is b on thirteen policy-response indicators including school closures, workplace closu travel bans, testing policy, contact tracing, face coverings, and vaccine policy. If coun with low containment observe fewer COVID-19 deaths or cases, there might be some known causes. Countries that did not achieve 50% in the COVID Data Transparency In totalanalysis.com/Covid19/TAIndex, accessed on 5 August 2021) were deleted in mining, as the numbers of deaths and confirmed cases in these countries might no accurate. Figure 1. A workflow for exploring natural products narratives (NPN). Yellow boxes represen use of cross-country data for the screening and selection of “best case” countries, e.g. those Figure 1. A workﬂow for exploring natural products narratives (NPN). 1. Introduction However, most studies so far built their databases by collecting published COVID-19 studies [15,16]. We compiled a dataset by visiting a large Chinese tourism website wherein a large population of users upload and share a large amount of travel writings that state their experiences with various overseas destinations. The use of online travel writing helps us to efﬁciently gain insights into the consumption of some “special” local food or consumption culture for each destination [17,18]. The content downloaded includes 7653 travel writings, as well as the upload date and destination. The data were grouped and combined according to destination, and each document consists of all the travel writings of a particular country. In total, we compiled a database representing 209 countries, covering all major regions and cultures in the world. The dataset reﬂects local condition within a time period before the pandemic, so this type of data was considered more appropriate than Wikipedia, wherein we cannot efﬁciently learn in-depth information on natural products in each location and can not specify a particular time period. The content downloaded includes 7653 travel writings, as well as the upload date destination. The data were grouped and combined according to destination, and each ument consists of all the travel writings of a particular country. In total, we compil database representing 209 countries, covering all major regions and cultures in the wo The dataset reflects local condition within a time period before the pandemic, so this of data was considered more appropriate than Wikipedia, wherein we cannot efficie earn in-depth information on natural products in each location and can not specify a icular time period. 2. Methods 2.1. A Workflow with NPN Approach As the workflow (Figure 1) is used to isolate the best cases of countries from nat products narratives (NPN), which may provide us hints on natural products, we del 2.1. A Workﬂow with NPN Approach tion such as closing schools and community transmission during C Yellow boxes represent the use of cross-country data for the screening and selection of “best case” countries, e.g. those with lower conﬁrmed cases and deaths, and for statistical testing. Blue boxes show a knowledge engineering process that transforms text into structured data and then visualizes promising nodes (natural products) in the knowledge graph. The orange box indicates domain knowledge in the knowledge engineering process. The green box shows experiments needed before conﬁrming the identiﬁcation of natural products. Figure 1. A workflow for exploring natural products narratives (NPN). Yellow boxes represen use of cross-country data for the screening and selection of “best case” countries, e.g. those Figure 1. A workﬂow for exploring natural products narratives (NPN). Yellow boxes represent the use of cross-country data for the screening and selection of “best case” countries, e.g. those with lower conﬁrmed cases and deaths, and for statistical testing. Blue boxes show a knowledge engineering process that transforms text into structured data and then visualizes promising nodes (natural products) in the knowledge graph. The orange box indicates domain knowledge in the knowledge engineering process. The green box shows experiments needed before conﬁrming the identiﬁcation of natural products. Healthcare 2022, 10, 2071 3 of 10 We end up with a ﬁnal list of countries with both a low containment index and low total COVID-19 conﬁrmed cases (Table 1). Although African countries lack effective containment and health facilities, many of them have surprisingly low level of total cases per million (<1000), e.g. Burundi, Burkina Faso, Congo Dem. Rep., Madagascar, Niger, Senegal, Somalia, Sudan, and Tanzania. S1 ﬁgures (supporting information [22]) show that most of the selected countries observe less COVID-19 deaths than the world average. We end up with a ﬁnal list of countries with both a low containment index and low total COVID-19 conﬁrmed cases (Table 1). Although African countries lack effective containment and health facilities, many of them have surprisingly low level of total cases per million (<1000), e.g. Burundi, Burkina Faso, Congo Dem. Rep., Madagascar, Niger, Senegal, Somalia, Sudan, and Tanzania. S1 ﬁgures (supporting information [22]) show that most of the selected countries observe less COVID-19 deaths than the world average. Table 1. Countries ranked by containment index. Table 1. Countries ranked by containment index. Table 1. Countries ranked by containment index. 2.1. A Workﬂow with NPN Approach tion such as closing schools and community transmission during C In group A, countries have big variance in total cases, ranging from 427 to 56,119. As the aim of data mining was to explore “best-case” countries that may provide hints of natural foods against COVID-19, we deleted countries with total conﬁrmed cases above 10,000. Singapore and Korea Rep. have demonstrated good administrative practice in pandemic control, so both countries were removed from the sample. In group B where countries have smaller variance, we deleted those with total cases above 1000. 2.2. Data Analysis Knowledge engineering as a sub-ﬁeld of artiﬁcial intelligence involves transferring human knowledge into a database and representing expert knowledge and reasoning [23]. We adopt knowledge engineering logic to develop a system to sort text data, isolate knowledge, and compile a domain-speciﬁc knowledge base. The knowledge base consists of three ontologies: food, drink, and smell. While natural food and drink might serve as medicine, substances in the air may inﬂuence the process by which a virus spreads from infected people to others. NLTK's (Natural Language Toolkit) tokenizers were used to convert narrative text into structured data. All ﬁles were then merged on the basis of these ontologies. This approach enabled us to connect concepts among sub-datasets and identify natural products spanning several countries, which has been a common limitation of previous knowledge mining studies [24]. 2.1. A Workﬂow with NPN Approach tion such as closing schools and community transmission during C Country Name Containment Index Total Cases/M Country Name Containment Index Total Cases/M Nicaragua 12.50 881 Somalia 38.69 280 Tanzania 16.37 9 Finland 44.94 4595 Burundi 17.26 58 Denmark 48.21 14,238 Yemen Rep. 18.45 70 Norway 50.60 6750 Afghanistan 23.81 1195 United Arab Emirates 52.62 17,203 Central African Rep. 24.40 1018 Netherlands 52.98 31,289 Congo Dem. Rep. 24.70 144 Singapore 52.98 9953 Sudan 26.19 416 South Africa 55.36 13,359 Mauritius 26.79 397 Sweden 57.14 25,819 Niger 27.08 66 Australia 57.14 1095 Mauritania 28.57 1873 Korea Rep. 58.63 686 Burkina Faso 28.57 140 Bahrain 59.52 51,209 New Zealand 32.14 427 Belgium 60.71 49,977 Syria 33.04 456 Lithuania 61.31 22,963 Congo Rep. 35.71 1046 Germany 61.31 13,065 Eswatini 36.90 5553 Qatar 61.90 48,246 Tajikistan 36.90 1282 Morocco 62.20 9749 Senegal 36.90 962 United Kingdom 63.10 24,265 Madagascar 36.90 626 Luxembourg 63.10 56,119 Haiti 38.10 815 Spain 63.69 35,428 Note: Total cases/m is total cases per million. Countries in the table with both a low containment index and low total COVID-19 conﬁrmed cases are included in analysis. Countries that failed to manage data transparently (<50% in the Covid Data Transparency Index) were deleted in data mining. Group A lists countries with good data transparency (≥50%), and group B includes countries whence transparency data are not available. In both groups, the top 20 countries with the lowest containment index were listed. In group A, countries have big variance in total cases, ranging from 427 to 56,119. As the aim of data mining was to explore “best-case” countries that may provide hints of natural foods against COVID-19, we deleted countries with total conﬁrmed cases above 10,000. Singapore and Korea Rep. have demonstrated good administrative practice in pandemic control, so both countries were removed from the sample. In group B where countries have smaller variance, we deleted those with total cases above 1000. Note: Total cases/m is total cases per million. Countries in the table with both a low containment index and low total COVID-19 conﬁrmed cases are included in analysis. Countries that failed to manage data transparently (<50% in the Covid Data Transparency Index) were deleted in data mining. Group A lists countries with good data transparency (≥50%), and group B includes countries whence transparency data are not available. In both groups, the top 20 countries with the lowest containment index were listed. 3.1. Corpus Development and Knowledge Engineering If the nutrition and consumption of this food are related to the prevention, or cure, of this disease, then the net import of this mustard should negatively inﬂuence the number of COVID-19 deaths. The ﬁrst COVID-19 vaccination took place in the mid of December 2020. To test the pure effect of mustard, we collected worldwide COVID-19 data from 1 March to 10 December 2020, a time period before vaccination. able [27,28]. We tested the relationship between net import of mustard and COVID-19 deaths worldwide. If the nutrition and consumption of this food are related to the prevention, or cure, of this disease, then the net import of this mustard should negatively influence the number of COVID-19 deaths. The first COVID-19 vaccination took place in the mid of December 2020. To test the pure effect of mustard, we collected worldwide COVID-19 data from 1 March to 10 December 2020, a time period before vaccination. p y p p p g y We tested the relationship between net import of mustard and COVID-19 deaths worldwide. If the nutrition and consumption of this food are related to the prevention, or cure, of this disease, then the net import of this mustard should negatively inﬂuence the number of COVID-19 deaths. The ﬁrst COVID-19 vaccination took place in the mid of December 2020. To test the pure effect of mustard, we collected worldwide COVID-19 data from 1 March to 10 December 2020, a time period before vaccination. 3.1. Corpus Development and Knowledge Engineering We developed a corpus of Chinese words related to eat, drink, food, smell, dish, cooking, taste, etc. The corpus was employed to screen out the sentences that do not consist of food, drink, and smell information. These invalid sentences were dropped from the knowledge base. Then, seven research assistants (college students) were recruited and trained. They manually annotated each valid sentence with tags indicating the actual Healthcare 2022, 10, 2071 4 of 10 product name used among local people and classiﬁed the tags under each ontology. The annotated data were checked and veriﬁed until inter-rater reliability was over 90%. y The purpose of the knowledge engineering was to identify similarities from the sample countries in terms of food-consumption culture. The more documents connected to a node, the more important the factor should be. A notable challenge with the text-mining approach was isolating good candidates from the large amount of nodes. We removed those nodes that are commonly consumed globally. A good candidate of potential factors identiﬁed from the data is mustard, which emerges as an important node linking multiple documents/countries from different angles (Figure 2). Mustard has been mentioned frequently in narratives about Moroccan food and its dye industry. It has also been noted in other narratives as common ingredients in Senegal’s Yassa (local dish), hot dog in Norway, and brunch in Finland and Australia. 5 of 12 Figure 2. Knowledge graph of nodes connected to mustard. The graph was made by VOSviewer [25,26]. Circle size reflects the number of links. Nodes from the same country are clustered with the same color. Figure 2. Knowledge graph of nodes connected to mustard. The graph was made by VOSviewer [25,26]. Circle size reﬂects the number of links. Nodes from the same country are clustered with the same color. Figure 2. Knowledge graph of nodes connected to mustard. The graph was made by VOSviewer [25,26]. Circle size reflects the number of links. Nodes from the same country are clustered with the same color. Figure 2. Knowledge graph of nodes connected to mustard. The graph was made by VOSviewer [25,26]. Circle size reﬂects the number of links. Nodes from the same country are clustered with the same color. Mustard consumption is measured by net import, which is defined as imports minus exports of mustard. 3.1. Corpus Development and Knowledge Engineering Net import data (product category: 210,330–mustard flour and meal and prepared mustard, by country) were provided by World Integrated Trade Solution (WITS), which was co-developed by the World Bank and the United Nations Conference on Trade and Development (UNCTAD). Net import is a good proxy for national con- sumption, particularly when consumption data of a specific product category is not avail- able [27 28] Mustard consumption is measured by net import, which is deﬁned as imports minus exports of mustard. Net import data (product category: 210,330–mustard ﬂour and meal and prepared mustard, by country) were provided by World Integrated Trade Solution (WITS), which was co-developed by the World Bank and the United Nations Conference on Trade and Development (UNCTAD). Net import is a good proxy for national consumption, particularly when consumption data of a speciﬁc product category is not available [27,28]. Mustard consumption is measured by net import, which is defined as imports minus exports of mustard. Net import data (product category: 210,330–mustard flour and meal and prepared mustard, by country) were provided by World Integrated Trade Solution (WITS), which was co-developed by the World Bank and the United Nations Conference on Trade and Development (UNCTAD). Net import is a good proxy for national con- sumption, particularly when consumption data of a specific product category is not avail- able [27,28]. We tested the relationship between net import of mustard and COVID-19 deaths worldwide. If the nutrition and consumption of this food are related to the prevention, or cure, of this disease, then the net import of this mustard should negatively influence the number of COVID-19 deaths. The first COVID-19 vaccination took place in the mid of December 2020. To test the pure effect of mustard, we collected worldwide COVID-19 data from 1 March to 10 December 2020, a time period before vaccination. Mustard consumption is measured by net import, which is deﬁned as imports minus exports of mustard. Net import data (product category: 210,330–mustard ﬂour and meal and prepared mustard, by country) were provided by World Integrated Trade Solution (WITS), which was co-developed by the World Bank and the United Nations Conference on Trade and Development (UNCTAD). Net import is a good proxy for national consumption, particularly when consumption data of a speciﬁc product category is not available [27,28]. We tested the relationship between net import of mustard and COVID-19 deaths worldwide. 3.2. Hypothesis Testing 3.2. Hypothesis Testing yp g The fruits or vegetables mentioned in the narratives pertaining to countries with low levels of COVID deaths and containment index include cassava, pepper, cabbage, papaya, banana, orange, avocado, olive, cocoa, coconut, pineapple, mango, apple, watermelon, al- d t t b h ll t f l t t h b i i The fruits or vegetables mentioned in the narratives pertaining to countries with low levels of COVID deaths and containment index include cassava, pepper, cabbage, papaya, banana, orange, avocado, olive, cocoa, coconut, pineapple, mango, apple, watermelon, Healthcare 2022, 10, 2071 5 of 10 5 of 10 almonds, tomato, cucumber, shallot, fennel, pepper, potato, cherry, beans, corn, ginger, cinnamon, calyx pear, chamomile, sisal, eggplant, radish, red pomelo, tricholoma matsu- take, jujube, jackfruit, avocado, litchi, plum, sugarcane, polo, baobab fruit, sakya, papaya, kiwifruit, and cactus fruit. We checked and compared with the WITS food list and tested those foods for which we could ﬁnd valid data from the WITS database. Both import and export data in 2018 were downloaded. The food category consisting of multiple types of food was not considered (e.g. “080450—Fruit, edible; guavas, mangoes and mangosteens, fresh or dried”), as one can not tell which particular food might cause an effect. Some veg- etables or fruits indicate non-trival but small negative associations (correlation coefﬁcient around 0.1) on COVID deaths, such as “cucumbers/gherkins” and “coconuts”, but none of these have a correlation coefﬁcient with COVID deaths greater than 0.2 (see Table 2). Table 2. Correlation tests between selected foods and COVID-19 outcomes. 3.2. Hypothesis Testing 3.2. Hypothesis Testing Potatoes Cucumbers and Gherkins Fruits of the Genus Capsicum or of the Genus Pimenta 70,190 70,700 70,960 N = 14,757 N = 12,300 N = 14,621 Total cases 0.0864 −0.0672 −0.0302 New cases 0.0946 −0.0694 −0.0465 Total deaths 0.1272 −0.106 −0.05 New deaths 0.0735 −0.0729 −0.0459 mushrooms and trufﬂes coconuts almonds 71,230 80,110 80,212 N = 14,191 N = 14,378 N = 14,312 Total cases −0.0294 −0.0201 0.231 New cases 0.0636 −0.0801 0.1922 Total deaths −0.0426 −0.0628 0.3168 New deaths −0.0023 −0.1166 0.2081 pineapples oranges citrus fruit 80,430 80,510 80,590 N = 13,420 N = 14,534 N = 10,325 Total cases 0.2359 −0.014 −0.0767 New cases 0.1707 −0.0022 −0.0579 Total deaths 0.2916 −0.0838 −0.0662 New deaths 0.1785 −0.074 −0.0522 grapes papaws plums and sloes 80,620 80,720 80,940 N = 13,837 N = 8381 N = 12,994 Total cases 0.2066 0.1354 0.0083 New cases 0.227 0.1258 0.0238 Total deaths 0.2766 0.1708 0.0039 New deaths 0.2479 0.1062 0.0275 apples 81,330 N = 11,232 Total cases 0.1649 New cases 0.1726 Total deaths 0.2379 New deaths 0.1767 Note: WITS product codes and correlation coefﬁcients are shown. Table 2. Correlation tests between selected foods and COVID-19 outcomes. We then analyzed other vegetables and fruits from the WITS trade database and ended up with 41 other types of foods for correlation tests. The results of grape indicate a small but considerable relationship with COVID deaths (correlation coefﬁcient > 0.1 for total Healthcare 2022, 10, 2071 6 of 10 and new deaths). Again, mustard shows a much stronger relationship with COVID deaths compared with the 41 types of foods. We used regression models in statistical tests with robust standard errors. As multiple data, i.e. daily deaths, were observed from each country, the determination of statistical signiﬁcance was based on clustered-robust standard errors (clustered by country). This is more conservative, as well as more accurate usually, for hypothesis-testing [29]. We hypothesize that mustard consumption is associated with COVID-19 outcomes. The dependent variables that we investigate in the statistical test include total and new deaths of COVID-19. Correlation analyses indicate considerable associations [30] between net import of mustard and total deaths (Pearson’s r = −0.24, p < 0.05), and between net import of mustard and new deaths (Pearson’s r = −0.21, p < 0.05). The direction of the relationships is consistent with our hypothesis. We developed two models in testing the hypothesis (Table 3). 3.2. Hypothesis Testing 3.2. Hypothesis Testing Model 2: number of obs. = 13,412; F(12, 72) = 18.90; Prob > F = 0.000; r2 = 0.431; Root MSE = 1.480. We tested if the net import of mustard has a negative relationship with total deaths. The results indicate that Model 1 predicts a good proportion of total deaths (r2 = 0.514) and a signiﬁcant and negative effect of net import on total deaths (p = 0.020) after controlling for all of the confounding variables above. We next tested with Model 2 if the net import of mustard could predict new deaths caused by COVID-19. The same set of variables were controlled. Again, the net import of mustard shows a signiﬁcant and negative relationship with new deaths (p = 0.034). 3.2. Hypothesis Testing 3.2. Hypothesis Testing As healthy economics and demographics may assist nations with combatting the COVID-19 pandemic [31,32], we selected the following confounders: population, life expectancy, GDP per capita, car- diovascular death rate, diabetes prevalence, percent of population aged 70+, population density, and number of hospital beds (per thousand), which were drawn from Our World in Data (Global Change Data Lab). As national infrastructure may also facilitate an efﬁcient reaction to COVID-19, we also controlled confounding factors, such as electricity and mobile subscriptions, by drawing data from World Bank. Table 3. Mustard net import and deaths caused by COVID-19. Table 3. Mustard net import and deaths caused by COVID-19. Total Deaths (Model 1) New Deaths (Model 2) Coef. SE t p > t Coef. SE t p > t Net import −0.011 0.005 −2.37 0.020 −0.011 0.005 −2.17 0.034 Containment index 0.010 0.011 0.90 0.371 0.014 0.010 1.48 0.144 Log(population) 1.051 0.088 11.91 0.000 0.692 0.067 10.32 0.000 Life expectancy −0.146 0.063 −2.33 0.023 −0.173 0.036 −4.74 0.000 Log(GDP per capita) 1.181 0.357 3.31 0.001 0.315 0.227 1.39 0.169 Cardiovasc death rate −0.002 0.001 −1.70 0.094 −0.003 0.001 −3.69 0.000 Diabetes prevalence −0.053 0.040 −1.33 0.189 −0.032 0.033 −0.96 0.340 Aged 70 0.097 0.051 1.90 0.062 0.089 0.032 2.78 0.007 Log (population density) 0.008 0.129 0.06 0.953 −0.115 0.093 −1.24 0.220 Hospital beds (1000) −0.144 0.059 −2.46 0.016 −0.140 0.039 −3.55 0.001 Electricity 0.054 0.019 2.81 0.006 0.082 0.015 5.62 0.000 Mobile subscriptions −0.023 0.008 −3.07 0.003 −0.008 0.004 −1.83 0.072 Note: Unit of net import: million USD. Total and new deaths are log transformed. The p values gained from a two-sided t test; SE clustered and adjusted for 73 countries; p values < 0.05 indicates statistical signiﬁcance. Model 1: number of obs. = 19,530; F(12, 72) = 91.93; Prob > F = 0.000; r2 = 0.514; Root MSE = 1.981. Model 2: number of obs. = 13,412; F(12, 72) = 18.90; Prob > F = 0.000; r2 = 0.431; Root MSE = 1.480. Table 3. Mustard net import and deaths caused by COVID-19. Note: Unit of net import: million USD. Total and new deaths are log transformed. The p values gained from a two-sided t test; SE clustered and adjusted for 73 countries; p values < 0.05 indicates statistical signiﬁcance. Model 1: number of obs. = 19,530; F(12, 72) = 91.93; Prob > F = 0.000; r2 = 0.514; Root MSE = 1.981. 3.3. Sensitivity Analysis for Additional COVID-19 Outcomes COVID-19 outcomes can be alternatively measured with the number of conﬁrmed cases. After controlling for all of the above confounding factors, the results indicate that the net import of mustard is negatively and signiﬁcantly related to total (p = 0.046) and new cases (p = 0.037) (Table 4). Correlation analysis results indicate a considerable effect size for Healthcare 2022, 10, 2071 7 of 10 the associations between the net import of mustard and total cases (Pearson’s r = −0.19) and between the net import of mustard and new cases (Pearson’s r = −0.23). the associations between the net import of mustard and total cases (Pearson’s r = −0.19) and between the net import of mustard and new cases (Pearson’s r = −0.23). the associations between the net import of mustard and total cases (Pearson’s r = −0.19) and between the net import of mustard and new cases (Pearson’s r = −0.23). Table 4. Sensitivity analysis: conﬁrmed cases as Covid-19 outcomes. Total Cases (Model 3) New Cases (Model 4) Coef. SE t p > t Coef. SE t p > t Net import −0.010 0.005 −2.030 0.046 −0.017 0.008 −2.120 0.037 Containment index 0.012 0.010 1.210 0.228 0.007 0.018 0.370 0.710 Log(population) 0.942 0.079 11.900 0.000 0.763 0.141 5.420 0.000 Life expectancy −0.047 0.069 −0.680 0.502 −0.159 0.072 −2.210 0.030 Log(GDP per capita) 1.502 0.348 4.310 0.000 1.383 0.417 3.310 0.001 Cardiovasc death rate 0.000 0.001 0.240 0.813 −0.001 0.001 −0.390 0.701 Diabetes prevalence −0.020 0.041 −0.490 0.627 −0.028 0.052 −0.540 0.593 Aged 70 −0.002 0.046 −0.050 0.963 0.011 0.058 0.190 0.848 Log(population density) 0.008 0.081 0.100 0.920 −0.009 0.119 −0.080 0.940 Hospital beds (1000) −0.074 0.054 −1.380 0.172 −0.088 0.067 −1.320 0.192 Electricity 0.011 0.019 0.600 0.553 0.039 0.020 1.990 0.050 Mobile subscriptions −0.014 0.006 −2.130 0.036 −0.014 0.009 −1.500 0.137 Note: Unit of net import: million USD. Total and new cases are log transformed. The p values gained from a two-sided t test; SE clustered and adjusted for 74 countries; p values < 0.05 indicates statistical signiﬁcance. Model 1: number of obs. = 20,853; F(12, 73) = 42.48; Prob > F = 0.000; r2 = 0.349; Root MSE = 2.358. Model 2: number of obs. = 18,889; F(12, 73) = 13.39; Prob > F = 0.000; r2 = 0.287; Root MSE = 2.173. Table 4. Sensitivity analysis: conﬁrmed cases as Covid-19 outcomes. Note: Unit of net import: million USD. 3.3. Sensitivity Analysis for Additional COVID-19 Outcomes Total and new cases are log transformed. The p values gained from a two-sided t test; SE clustered and adjusted for 74 countries; p values < 0.05 indicates statistical signiﬁcance. Model 1: number of obs. = 20,853; F(12, 73) = 42.48; Prob > F = 0.000; r2 = 0.349; Root MSE = 2.358. Model 2: number of obs. = 18,889; F(12, 73) = 13.39; Prob > F = 0.000; r2 = 0.287; Root MSE = 2.173. As shown in previous studies, deaths and conﬁrmed cases are strongly inﬂuenced by governmental factors such as lockdown, stringency, testing policy, the extent of contact tracing, requirements to wear face coverings, and policies around vaccine rollout [33]; we did additional tests by dividing all sample countries into two groups with a high (scored above 70) and low Containment Index (≤70). Figure 3a–d indicates that the net import of mustard consistently have negative relationships with total (new) deaths and total (new) cases for both groups of countries. The negative relationship is stronger for countries with a high containment index, with r ranging from −0.749 to −0.688. 022, 10, x 9 of 12 Figure 3. Factors correlated with the net import of mustard. (A,B) Total and new deaths caused by COVID-19 are negatively related to net import (log transformed). Brown lines represent a high con- tainment index (>70), while blue lines represent a low containment index (≤70). (C,D) Total and new COVID-19 confirmed cases are negatively related to net import. The correlations are compared with different levels of the containment index. 4. Discussion The proposed workflow can accelerate the identification of potential natural prod- Figure 3. Factors correlated with the net import of mustard. (A,B) Total and new deaths caused by COVID-19 are negatively related to net import (log transformed). Brown lines represent a high containment index (>70), while blue lines represent a low containment index (≤70). (C,D) Total and new COVID-19 conﬁrmed cases are negatively related to net import. The correlations are compared with different levels of the containment index. Figure 3. Factors correlated with the net import of mustard. (A,B) Total and new deaths caused by COVID-19 are negatively related to net import (log transformed). Brown lines represent a high con- tainment index (>70), while blue lines represent a low containment index (≤70). (C,D) Total and new COVID-19 confirmed cases are negatively related to net import. 4. Discussion The proposed workﬂow can accelerate the identiﬁcation of potential natural products against COVID-19. This novel method has been illustrated by drawing from experiences and writings relating to natural products on over 200 countries. Although narratives pertaining to a particular country would be contextual, a cross-country comparison of such data may bring some new and natural solutions against the disease. Since both cross-country NPN and product data can be collected from online or public sources, our approach is a general and time-efﬁcient one, without additional inputs from metabolic engineering or gene expression. The case of mustard provides a proof-of-concept. The net import of mustard has been found to be associated with reduced deaths, particularly when nations manage to improve the containment index to a level of above 70 (r becomes around −0.7). The ﬁndings are robust to the use of conﬁrmed case data. These ﬁndings provide new explanations on why some countries, although less resourceful in terms of vaccine, crisis containment, and healthcare facilities, are not harmed seriously by the pandemic. By examining the relationships between the net import of other foods and COVID-19 outcomes, we found several notable positive correlations (>0.2) between COVID-19 total deaths and the net import of almond (0.317), pineapple (0.292), grapes (0.277), and apples (0.238). These fruits, although not associated with a reduced number of COVID-19 deaths, might provide new hints on the factors relating to the growing number of COVID-19 deaths. Future research could address this issue by investigating the nutrition or chemical mechanisms underlying the connections between these fruits and COVID-19. This study was limited by the scope of the destination data. Although we collected a fairly good number of travel writings on 209 nations or regions, the dataset does not cover every country in the world and does not convey a complete picture of local consumption and culture. Future studies could investigate the molecular mechanism pertaining to mus- tard, particularly pertaining to mustard seed, which has been used to make sauce and dye in our sample countries. Mustard has been found to be useful in the treatment of pneumo- nia, asthma, or cough and in relieving pain symptoms, such as headaches and neuralgia. Since fermentation with Lactobacillus Plantarum can enhance the anti-inﬂammatory activity of mustard leaves, mustard leaves fermented by this microorganism may be helpful in the treatment of inﬂammation [34]. The SARS-CoV-2 protein 3CLPro is essential for successful viral replication. 3.3. Sensitivity Analysis for Additional COVID-19 Outcomes The correlations are compared with different levels of the containment index. 4. Discussion Figure 3. Factors correlated with the net import of mustard. (A,B) Total and new deaths caused by COVID-19 are negatively related to net import (log transformed). Brown lines represent a high containment index (>70), while blue lines represent a low containment index (≤70). (C,D) Total and new COVID-19 conﬁrmed cases are negatively related to net import. The correlations are compared with different levels of the containment index. Figure 3. Factors correlated with the net import of mustard. (A,B) Total and new deaths caused by COVID-19 are negatively related to net import (log transformed). Brown lines represent a high con- tainment index (>70), while blue lines represent a low containment index (≤70). (C,D) Total and new COVID-19 confirmed cases are negatively related to net import. The correlations are compared with different levels of the containment index. 4. Discussion Th d kfl l t th id tifi ti f t ti l t l d Figure 3. Factors correlated with the net import of mustard. (A,B) Total and new deaths caused by COVID-19 are negatively related to net import (log transformed). Brown lines represent a high containment index (>70), while blue lines represent a low containment index (≤70). (C,D) Total and new COVID-19 conﬁrmed cases are negatively related to net import. The correlations are compared with different levels of the containment index. Healthcare 2022, 10, 2071 8 of 10 8 of 10 Informed Consent Statement: Not applicable. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. References 1. Boum, I.Y.; Ouattara, A.; Torreele, E.; Okonta, C. How to ensure a needs-driven and community-centred vaccination strategy for COVID-19 in Africa. BMJ Glob. Health. 2021, 6, e005306. [CrossRef] [PubMed] 2. Garcia-Beltran, W.F.; Lam, E.C.; St Denis, K.; Nitido, A.D.; Garcia, Z.H.; Hauser, B.M.; Feldman, J.; Pavlovic, M.N.; Gregory, D.J.; Poznansky, M.C.; et al. Multiple SARS-CoV-2 variants escape neutralization by vaccine-induced humoral immunity. Cell 2021, 184, 2372–2383.e9. [CrossRef] [PubMed] 3. Zhou, D.; Dejnirattisai, W.; Supasa, P.; Liu, C.; Mentzer, A.J.; Ginn, H.M.; Zhao, Y.; Duyvesteyn, H.M.E.; Tuekprakhon, A.; Nutalai, R.; et al. Evidence of escape of SARS-CoV-2 variant B.1.351 from natural and vaccine-induced sera. Cell 2021, 184, 2348–2361.e6. [CrossRef] [PubMed] 4. Lopez Bernal, J.; Andrews, N.; Gower, C.; Gallagher, E.; Simmons, R.; Thelwall, S.; Stowe, J.; Tessier, E.; Groves, N.; Dabrera, G.; et al. Effectiveness of Covid-19 vaccines against the B.1.617.2 (Delta) variant. N. Engl. J. Med. 2021, 385, 585–594. [CrossRef] 5. Gaziano, L.; Giambartolomei, C.; Pereira, A.C.; Gaulton, A.; Posner, D.C.; Swanson, S.A.; Ho, Y.-L.; Iyengar, S.K.; Kosik, N.M.; Vujkovic, M.; et al. Actionable druggable genome-wide Mendelian randomization identiﬁes repurposing opportunities for COVID-19. Nat. Med. 2021, 27, 668–676. [CrossRef] 6. Guy, R.K.; DiPaola, R.S.; Romanelli, F.; Dutch, R.E. Rapid repurposing of drugs for COVID-19. Sc [CrossRef] 7. Shyr, Y.; Berry, L.D.; Hsu, C.Y. Scientiﬁc rigor in the age of COVID-19. JAMA Oncol. 2021, 7, 171–172. [CrossRef] 8. Doroghazi, J.R.; Albright, J.C.; Goering, A.W.; Ju, K.S.; Haines, R.R.; Tchalukov, K.A.; Labeda, D.P.; Kellehe, N.L.; Metcalf, W.W. A roadmap for natural product discovery based on large-scale genomics and metabolomics. Nat. Chem. Biol. 2014, 10, 963–968. [CrossRef] 9. Kersten, R.D.; Yang, Y.L.; Xu, Y.; Cimermancic, P.; Nam, S.J.; Fenical, W.; Fischbach, M.A.; Moore, B.S.; Dorrestein, P.C. A mass spectrometry–guided genome mining approach for natural product peptidogenomics. Nat. Chem. Biol. 2011, 7, 794–802. [CrossRef] 10. Elbattah, M.; Arnaud, É.; Gignon, M.; Dequen, G. The role of text analytics in Healthcare: A review of recent developments and applications. In Proceedings of the 14th International Joint Conference on Biomedical Engineering Systems and Technologies, Vienna, Austria, 11–13 February 2021; pp. 825–832. [CrossRef] y pp 11. Ong, S.-Q.; Pauzi, M.B.M.; Gan, K.H. Text mining and determinants of sentiments towards the COVID-19 vaccine booster of twitter users in Malaysia. Healthcare 2022, 10, 994. [CrossRef] 12. Xu, H.; Zeng, J.; Tai, Z.; Hao, H. [ ] 21. Liang, L.L.; Tseng, C.H.; Ho, H.J.; Wu, C.Y. Covid-19 mortality is negatively associated with test number and government effectiveness. Sci. Rep. 2020, 10, 12567. [CrossRef] References Public attention and sentiment toward intimate partner violence based on Weibo in China: A text mining approach. Healthcare 2022, 10, 198. [CrossRef] 13. Nagashima, T.; Shirakawa, H.; Nakagawa, T.; Kaneko, S. Prevention of antipsychotic-induced hyperglycaemia by vitamin D: A data mining prediction followed by experimental exploration of the molecular mechanism. Sci. Rep. 2016, 6, 26375. [CrossRef] 13. Nagashima, T.; Shirakawa, H.; Nakagawa, T.; Kaneko, S. Prevention of antipsychotic-induced hyperglycaemia by vitamin D: A data mining prediction followed by experimental exploration of the molecular mechanism. Sci. Rep. 2016, 6, 26375. [CrossRef] 14. Sarangdhar, M.; Tabar, S.; Schmidt, C.; Kushwaha, A.; Shah, K.; Dahlquist, J.E.; Jegga, A.G.; Aronow, B.J. Data mining differential clinical outcomes associated with drug regimens using adverse event reporting data. Nat. Biotechnol. 2016, 34, 697–700. [CrossRef] 15. Michel, F.; Gandon, F.; Ah-Kane, V.; Bobasheva, A.; Cabrio, E.; Corby, O.; Gazzotti, R.; Giboin, A.; Marro, S.; Mayer, T.; et al. Covid-on-the-Web: Knowledge graph and services to advance COVID-19 research. In Proceedings of the International Semantic Web Conference Athens Greece 1–6 November 2020; pp 294–310 13. Nagashima, T.; Shirakawa, H.; Nakagawa, T.; Kaneko, S. Prevention of antipsychotic-induced hyperglycaemia by vitamin D: A data mining prediction followed by experimental exploration of the molecular mechanism. Sci. Rep. 2016, 6, 26375. [CrossRef] 14. Sarangdhar, M.; Tabar, S.; Schmidt, C.; Kushwaha, A.; Shah, K.; Dahlquist, J.E.; Jegga, A.G.; Aronow, B.J. Data mining differential g g p y yp g y y data mining prediction followed by experimental exploration of the molecular mechanism. Sci. Rep. 2016, 6, 26375. [CrossRef] 14. Sarangdhar, M.; Tabar, S.; Schmidt, C.; Kushwaha, A.; Shah, K.; Dahlquist, J.E.; Jegga, A.G.; Aronow, B.J. Data mining differential clinical outcomes associated with drug regimens using adverse event reporting data. Nat. Biotechnol. 2016, 34, 697–700. [CrossRef] 15. Michel, F.; Gandon, F.; Ah-Kane, V.; Bobasheva, A.; Cabrio, E.; Corby, O.; Gazzotti, R.; Giboin, A.; Marro, S.; Mayer, T.; et al. 14. Sarangdhar, M.; Tabar, S.; Schmidt, C.; Kushwaha, A.; Shah, K.; Dahlquist, J.E.; Jegga, A.G.; Aronow, B.J. Data mining differential clinical outcomes associated with drug regimens using adverse event reporting data. Nat. Biotechnol. 2016, 34, 697–700. [CrossRef] 14. Sarangdhar, M.; Tabar, S.; Schmidt, C.; Kushwaha, A.; Shah, K.; Dahlquist, J.E.; Jegga, A.G.; Aronow, B.J. Data mining differential clinical outcomes associated with drug regimens using adverse event reporting data. Nat. Biotechnol. 2016, 34, 697–700. [CrossRef] 15. 4. Discussion A recent study found that a glucosinolate derivative found in mustard seeds is a potent inhibitor of SARS-CoV-2 3CLPro [35]. Future work may also try using other data-driven or algorithm-driven approaches in the identiﬁcation of foods related to COVID 19. While our workﬂow is developed on the basis of a knowledge graph that is exploratory in nature and built on domain-speciﬁc knowledge, deep learning applications in NLP, particularly BERT models, have good poten- tial for the efﬁcient identiﬁcation of food-related words or terms from large social network sites, such as Twitter. For example, in a recent healthcare study, researchers developed a NLP algorithm and accurately extracted sleep parameters from polysomnography text noted in electronic medical records [36]. Author Contributions: G.Z. developed the idea and research. G.Z. wrote the ﬁrst draft of the manuscript, and all other authors discussed the method and results. F.Y. collected and validated the text data. L.Z. edited the manuscript. G.Z., L.Z., and H.W. performed the analyses. H.W. generated all ﬁgures. All authors have read and agreed to the published version of the manuscript. Funding: G.Z. acknowledge support by the National Natural Science Foundation of China (Key Program, grant code 71832015), Higher Education Enhancement Plan by the Guangdong Education Department (UICR0400011-21), UIC Research Grant (R202027), and Joint Research Project from the Guangdong Planning Ofﬁce of Philosophy and Social Science (GD20XGL55). Institutional Review Board Statement: Not applicable. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. 9 of 10 9 of 10 Healthcare 2022, 10, 2071 Data Availability Statement: Sample data and aggregated statistics for replication and academic research purposes are available from the corresponding author on reasonable request. Conﬂicts of Interest: The authors declare no conﬂict of interest. References Network graph repres g g g p p 9 scientiﬁc publications to aid knowledge discovery. BMJ Health Care Inform. 2021, 28, e100254. [CrossRef] p g y J f , , [ ] 25. Van Eck, N.J.; Waltman, L. Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics 2010, 84, 523–538. [CrossRef] [ ] 26. Van Eck, N.J.; Waltman, L. Text mining and visualization using VOSviewer. arXiv 2011, arXiv:1109.2058. 27. Feuerstein, S. Do coffee roasters beneﬁt from high prices of green coffee? Int. J. Indus. Org. 2002, 20, 89–118. [CrossRef] 28. Michael, S.W. Application of a dynamic panel data estimator to cross-country coffee demand: A tale of two eras. J. Econ. Dev. 2009, 34, 1–17. [CrossRef] 29. Cameron, A.C.; Gelbach, J.B.; Miller, D.L. Robust inference with multiway clustering. J. Bus. Econ. Stat. 2011, 29, 238–249. [CrossRef] 30. Ellis, P.D. The Essential Guide to Effect Sizes: Statistical Power, Meta-Analysis, and the Interpretation of Research Results; Cambridge University Press: Cambridge, UK, 2010. 31. Gaye, B.; Khoury, S.; Cene, C.W.; Kingue, S.; N’Guetta, R.; Lassale, C.; Baldé, D.; Diop, I.B.; Dowd, J.B.; Mills, M.C.; et al. Socio-demographic and epidemiological consideration of Africa’s COVID-19 response: What is the possible pandemic course? Nat. Med. 2020, 26, 996–999. [CrossRef] [PubMed] 32. Sharma, A.; Borah, S.B.; Moses, A.C. Responses to COVID-19: The role of governance, healthcare infrastructure, and learning from past pandemics. J. Bus. Res. 2021, 122, 597–607. [CrossRef] [PubMed] 33. Haider, N.; Osman, A.Y.; Gadzekpo, A.; Akipede, G.O.; Asogun, D.; Ansumana, R.; Lessells, R.J.; Khan, P.; Hamid, M.M.A.; Yeboah-Manu, D.; et al. Lockdown measures in response to COVID-19 in nine sub-Saharan African countries. BMJ Glob. Health 2020, 5, e003319. [CrossRef] [PubMed] 34. Le, B.; Anh, P.T.N.; Yang, S.H. Enhancement of the anti-inﬂammatory effect of mustard kimchi on RAW 264.7 macrophages by the Lactobacillus plantarum fermentation-mediated generation of phenolic compound derivatives. Foods 2020, 9, 181. [CrossRef] [PubMed] [ ] 35. Guijarro-Real, C.; Plazas, M.; Rodríguez-Burruezo, A.; Prohens, J.; Fita, A. Potential in vitro inhibition of selected plant extracts against SARS-CoV-2 chymotripsin-like protease (3CLPro) activity. Foods 2021, 10, 1503. [CrossRef] j , ; , ; g , ; , J ; , ainst SARS-CoV-2 chymotripsin-like protease (3CLPro) activity. Foods 2021, 10, 1503. [CrossRef] 36. Rahman, M.; Nowakowski, S.; Agrawal, R.; Naik, A.; Sharafkhaneh, A.; Razjouyan, J. Validation of a natural language processing algorithm for the extraction of the sleep parameters from the polysomnography reports. Healthcare 2022, 10, 1837. References Michel, F.; Gandon, F.; Ah-Kane, V.; Bobasheva, A.; Cabrio, E.; Corby, O.; Gazzotti, R.; Giboin, A.; Marro, S.; Mayer, T.; et al. Covid-on-the-Web: Knowledge graph and services to advance COVID-19 research. In Proceedings of the International Semantic Web Conference, Athens, Greece, 1–6 November 2020; pp. 294–310. g g g p g 15. Michel, F.; Gandon, F.; Ah-Kane, V.; Bobasheva, A.; Cabrio, E.; Corby, O.; Gazzotti, R.; Giboin, A.; Marro, S.; Mayer, T.; et al. Covid-on-the-Web: Knowledge graph and services to advance COVID-19 research. In Proceedings of the International Semantic Web Conference, Athens, Greece, 1–6 November 2020; pp. 294–310. 16. Steenwinckel, B.; Vandewiele, G.; Rausch, I.; Heyvaert, P.; Taelman, R.; Colpaert, P.; Simoens, P.; Dimou, A.; De Turck, F.; Ongenae, F. Facilitating the analysis of COVID-19 literature through a knowledge graph. In Proceedings of the International Semantic Web Conference, Athens, Greece, 1–6 November 2020; pp. 344–357. pp 17. Tsai, C.-T.S. Memorable tourist experiences and place attachment when consuming local food. Int. J. Tour. Res. 2016, 18, 536–548. [CrossRef] 18. Kim, S.; Choe, J.Y.; King, B.; Oh, M.; Otoo, F.E. Tourist perceptions of local food: A mapping of cultural values. Int. J. Tour. Res. 2021; in print. [CrossRef] 19. Brauner, J.M.; Mindermann, S.; Sharma, M.; Johnston, D.; Salvatier, J.; Gavenˇciak, T.; Stephenson, A.B.; Leech, G.; Altman, G.; Mikulik, V.; et al. Inferring the effectiveness of government interventions against COVID-19. Science 2021, 371, eabd9338. [CrossRef] 20. Hsiang, S.; Allen, D.; Annan-Phan, S.; Bell, K.; Bolliger, I.; Chong, T.; Druckenmiller, H.; Huang, L.Y.; Hultgren, A.; Krasovich, E.; et al. The effect of large-scale anti-contagion policies on the COVID-19 pandemic. Nature 2020, 584, 262–267. [CrossRef] [ ] 21. Liang, L.L.; Tseng, C.H.; Ho, H.J.; Wu, C.Y. Covid-19 mortality is negatively associated with test number and government effectiveness. Sci. Rep. 2020, 10, 12567. [CrossRef] Healthcare 2022, 10, 2071 10 of 10 22. Coronavirus Resource Center, Data Notes by Regions, by Johns Hopkins University & Medicine. Available online: https: //coronavirus.jhu.edu/region (accessed on 5 August 2021). j g g 23. Shi, Z. Intelligence Science: Leading the Age of Intelligence; Tsinghua University Press: Beijing, China, 2021; pp. 1–31. g y j g Cernile, G.; Heritage, T.; Sebire, N.J.; Gordon, B.; Schwering, T.; Kazemlou, S.; Borecki, Y. Network gra COVID-19 scientiﬁc publications to aid knowledge discovery. BMJ Health Care Inform. 2021, 28, e100254. [C G.; Heritage, T.; Sebire, N.J.; Gordon, B.; Schwering, T.; Kazemlou, S.; Borecki, Y. References [CrossRef]
https://openalex.org/W4376275029	https://www.mdpi.com/1424-8220/23/10/4714/pdf?version=1684138556	English	null	Automatic Monitoring Alarm Method of Dammed Lake Based on Hybrid Segmentation Algorithm	Sensors	2,023	cc-by	9,015	Article Automatic Monitoring Alarm Method of Dammed Lake Based on Hybrid Segmentation Algorithm Ziming Cai 1,†, Liang Sun 2,† , Baosheng An 3,, Xin Zhong 2, Wei Yang 3, Zhongyan Wang 3, Yan Zhou 2,4, Feng Zhan 1, and Xinwei Wang 2,4,* ,†, Liang Sun 2,† , Baosheng An 3,, Xin Zhong 2, Wei Yang 3, Zhongyan Wang 3, Yan Zhou 2,4 and Xinwei Wang 2,4,* 1 School of Resources, Environment and Materials, Guangxi University, Nanning 530004, China 1 School of Resources, Environment and Materials, Guangxi University, Nanning 530004, China 2 Optoelectronic System Laboratory, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China 3 I tit t f Tib t Pl t R h Chi A d f S i B iji 100101 Chi 1 School of Resources, Environment and Materials, Guangxi University, Nanning 530004, China 2 Optoelectronic System Laboratory, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China 3 Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100101, China 4 School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China * Correspondence: anbaosheng@itpcas.ac.cn (B.A.); phy_idea@outlook.com (F.Z.); wangxinwei@semi.ac.cn (X.W.) † These authors contributed equally to this work. 1 School of Resources, Environment and Materials, Guangxi University, Nanning 530004, China 2 Optoelectronic System Laboratory, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China 3 Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100101, China 4 School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China * Correspondence: anbaosheng@itpcas.ac.cn (B.A.); phy_idea@outlook.com (F.Z.); wangxinwei@semi.ac.cn (X.W.) † These authors contributed equally to this work. * Correspondence: anbaosheng@itpcas.ac.cn (B.A.); phy_idea@outlook.com (F.Z.); wangxinwei@semi ac cn (X W ) * Correspondence: anbaosheng@itpcas.ac.cn (B.A.); phy_idea@outlook.com (F.Z.); wangxinwei@semi.ac.cn (X.W.) † These authors contributed equally to this work. Abstract: Mountainous regions are prone to dammed lake disasters due to their rough topography, scant vegetation, and high summer rainfall. By measuring water level variation, monitoring systems can detect dammed lake events when mudslides block rivers or boost water level. Therefore, an automatic monitoring alarm method based on a hybrid segmentation algorithm is proposed. The algorithm uses the k-means clustering algorithm to segment the picture scene in the RGB color space and the region growing algorithm on the image green channel to select the river target from the segmented scene. The pixel water level variation is used to trigger an alarm for the dammed lake event after the water level has been retrieved. sensors sensors sensors Citation: Cai, Z.; Sun, L.; An, B.; Zhong, X.; Yang, W.; Wang, Z.; Zhou, Y.; Zhan, F.; Wang, X. Automatic Monitoring Alarm Method of Dammed Lake Based on Hybrid Segmentation Algorithm. Sensors 2023, 23, 4714. https://doi.org/ 10.3390/s23104714 Keywords: dammed lake; automatic monitoring alarm; k-means clustering algorithm; region growing algorithm; water level recognition Academic Editors: Goncalo Jesus and Anabela Oliveira Academic Editors: Goncalo Jesus and Anabela Oliveira Article Automatic Monitoring Alarm Method of Dammed Lake Based on Hybrid Segmentation Algorithm In the Yarlung Tsangpo River basin of the Tibet Autonomous Region of China, the proposed automatic lake monitoring system was installed. We pick up data from April to November 2021, during which the river experienced low, high, and low water levels. Unlike conventional region growing algorithms, the algorithm does not rely on engineering knowledge to pick seed point parameters. Using our method, the accuracy rate is 89.29% and the miss rate is 11.76%, which is 29.12% higher and 17.65% lower than the traditional region growing algorithm, respectively. The monitoring results indicate that the proposed method is a highly adaptable and accurate unmanned dammed lake monitoring system. 1. Introduction Received: 13 March 2023 Revised: 6 May 2023 Accepted: 8 May 2023 Published: 12 May 2023 Global warming causes the occurrence of extreme weather and the rapid melting of glaciers, affecting the environment and the safety of humans [1,2]. As the world’s third pole, the Qinghai-Tibet Plateau is more sensitive to climate change and has a more delicate ecosystem [3–6]. In October 2018, the Grand Canyon on the Yarlung Tsangpo River (YTR) in the Sedongpu Basin experienced two successive glacier collapses and river-blocking incidents, and they caused a huge disaster and posed a potential threat to the residents and transport lines upstream and downstream from Paizhen Town and in the area along the river banks in Medog County. The incidents produced a disaster chain process of glacier collapse, glacial debris ﬂow, river blockage, dammed lake, and outburst ﬂood [7]. With the climate warming on the Tibetan Plateau, such disasters will continue or even intensify in the future. To ensure the safety of human lives and property and to better comprehend the law of glacier movement, it is of considerable practical importance to conduct research on the automatic dammed lake event alarm. 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/ 4.0/). https://www.mdpi.com/journal/sensors Sensors 2023, 23, 4714. https://doi.org/10.3390/s23104714 Sensors 2023, 23, 4714 2 of 12 It is feasible to detect whether a dammed lake event is caused by observing the river water level because variations in water level are typically associated with river obstacles. Currently, numerous efforts have been made to monitor the variation in water level, such as pressure sensors [8–11], ultrasonic meters [12–16], and optical vision monitoring [17,18]. Pressure sensors must be calibrated, and they are extremely sensitive to any lateral move- ment at mounting points. Continuous water pressure can cause sensor breakdown, re- quiring frequent calibration and replacement. Ultrasonic sensors use ultrasonic pulses to measure the time it takes for a signal to travel from an emitter to a receiver, but their lifetime is short and the returned sound wave values are sensitive to temperature, precipitation, and snowfall. Optical image monitoring applications are affordable and provide a vast monitor- ing area, compared with conventional ﬁxed-point sensors, to detect water levels [19]. In terms of system maintenance, the maintenance of pressure sensors and ultrasonic meters is frequently a time-consuming and resource-intensive endeavor, especially for the disaster monitoring devices installed in uninhabited areas, due to a variety of environmental and technical reasons. Optical imaging monitoring is stable, but it generates a large number of images that require manual judgment and consumes a lot of manpower. In order to save costs and liberate manpower, the realization of automatic monitoring has become an urgent problem. g p Ground-based monitoring systems are well suited for long-term ﬁxed-point automatic monitoring of speciﬁc small areas, such as wild rivers, urban rivers, drainage ditches, and so on. Optical satellite and aerial monitoring systems are often used for large-scale geographical analysis because their accuracy is not as high as that of ground monitoring due to long observation ranges, and their monitoring of ﬁxed sites is not continuous. Water level identiﬁcation methods for automatic monitoring are classiﬁed into two types based on whether or not manual marking is used. Using artiﬁcial markers, such as a water gauge, can determine the variation in water level, and recognizing the variation in the river boundary also can determine the variation in water level. 2 2 Hybrid Segmentation Algorithm 2.2. Hybrid Segmentation Algorithm 2.2. Hybrid Segmentation Algorithm The proposed algorithm is a combination of RG and k-means clustering algorithms. The goal of RG is to separate the greenish river pixels, and the goal of k-means clustering is to classify all pixels, so green channel gray image and full-channel gray image are used, respectively. The input RGB images are divided into two types of gray images. The gray value of full-channel gray images adopts 30% red channel, 59% green channel, and 11% blue channel. The gray value of green channel gray images adopts 100% green channel. Step 1 shows the process of selecting the region of interest (ROI) and the detection line. Step 2 shows the process of using the RG algorithm to obtain the RG line as the auxiliary line on the green channel of an image. Step 3 shows the process of using the k-means clustering algorithm on denoised images. Step 4 shows the process of obtaining water level by combining the skeleton thinning algorithm based on k-means clustering with the RG-line. The region of interest (ROI) is a manually selected rectangular area of the image. Step 5 shows the process of making alarm judgments by using pixel water level data of The proposed algorithm is a combination of RG and k-means clustering algorithms. The goal of RG is to separate the greenish river pixels, and the goal of k-means clustering is to classify all pixels, so green channel gray image and full-channel gray image are used, respectively. The input RGB images are divided into two types of gray images. The gray value of full-channel gray images adopts 30% red channel, 59% green channel, and 11% blue channel. The gray value of green channel gray images adopts 100% green channel. Step 1 shows the process of selecting the region of interest (ROI) and the detection line. Step 2 shows the process of using the RG algorithm to obtain the RG line as the auxiliary line on the green channel of an image. Step 3 shows the process of using the k-means clustering algorithm on denoised images. Step 4 shows the process of obtaining water level by combining the skeleton thinning algorithm based on k-means clustering with the RG-line. The region of interest (ROI) is a manually selected rectangular area of the image. Step 5 shows the process of making alarm judgments by using pixel water level data of the waterline. the waterline. When the monitoring site has a water gauge marker, the intensity information on both sides of the water level will be different, and the variation of the water level can be obtained by extracting the reﬂection parameter of the river water bodies from the image [20,21]. In addition, when the image template is extracted from the marker in advance, the part of the template that can be matched by the multi-template matching technology is the area that has not been covered by the water, allowing the variation in water level to be obtained. [22,23]. The manual placement of water gauges is highly limited because it is often inaccessible at many monitoring sites. However, automatic monitoring has grown in popularity [24–27] and the key to achieving automatic monitoring is the variation in water level. When there is no gauge marker, the variation in river boundary can provide information on the variation in water level. So, the extraction of the river boundary is critical for automatic water level identiﬁcation. Canny edge detection is a simple and effective image processing technology proposed for application in real-world environments [28,29]. However, the bank lines of natural rivers without artiﬁcial embankments are fuzzy, and canny edge detection cannot identify lines precisely. The region growing (RG) algorithm is often used to identify the region of water due to the uniformity of watercolor in favor of seed growth. The hybrid algorithm achieves relatively higher precision in ﬂood area identiﬁcation compared with growth cutting and RG alone [30]. Other hybrid algorithms can also identify ﬂood zone with relatively high levels of accuracy, such as threshold value and RG [31]. To achieve a good identiﬁcation effect in a complex river scene, the above algorithms require regular parameter modiﬁcation, which is not conducive to automatic monitoring. Therefore, this paper proposes a hybrid segmentation algorithm for automatic monitoring alarm systems in no man’s land that does not rely on artiﬁcial markers. y The remainder of this paper is organized as follows. Section 2 describes the automatic monitoring alarm system and hybrid segmentation algorithm in detail. Section 3 shows the experimental results of the proposed method, and Section 4 gives conclusions. Sensors 2023, 23, 4714 3 of 12 hows 3 of 12 hows 2.1. Automatic Monitoring Alarm System 2.1. Automatic Monitoring Alarm System Lhasa Earth System Multi Dimensio Lhasa Earth System Multi-Dimension Observatory Network established a long-term automatic monitoring and early warning system focused on glacier collapse disaster chains in the Grand Canyon on the YTR [7]. In the system, the automatic monitoring alarm sub-system (AMAS) is an image-based unattended monitoring unit that decides whether a dammed lake event has occurred at a monitoring site. The AMAS analyzes variations in water level by combining images captured on-site with a back-end image processing module. When the water level rises, the system sends an alarm indicating that the lake has been dammed. Figure 1 shows that the system consists of a camera, satellite, and back-end image processing module. Images are transmitted by satellite to the back-end image processing module, and the proposed algorithm is used to detect the water level. When an alarm happens, the user will receive a short messaging service (SMS) that a dammed lake has occurred. Lhasa Earth System Multi-Dimension Observatory Network established a long-term automatic monitoring and early warning system focused on glacier collapse disaster chains in the Grand Canyon on the YTR [7]. In the system, the automatic monitoring alarm sub-system (AMAS) is an image-based unattended monitoring unit that decides whether a dammed lake event has occurred at a monitoring site. The AMAS analyzes variations in water level by combining images captured on-site with a back-end image processing module. When the water level rises, the system sends an alarm indicating that the lake has been dammed. Figure 1 shows that the system consists of a camera, satellite, and back- end image processing module. Images are transmitted by satellite to the back-end image processing module, and the proposed algorithm is used to detect the water level. When an alarm happens, the user will receive a short messaging service (SMS) that a dammed lake has occurred. Figure 1. Automatic monitoring alarm sub-system for dammed lake. Figure 1. Automatic monitoring alarm sub-system for dammed lake. Figure 1. Automatic monitoring alarm sub-system for dammed lake. Figure 1. Automatic monitoring alarm sub-system for dammed lake. 2. Methodology e odo ogy 2 1 Automatic Mon 2.1. Automatic Monitoring Alarm System 2.1. Automatic Monitoring Alarm System Lha a Ea th Sy te Multi Di e io 2 2 Hybrid Segmentation Algorithm 2.2. Hybrid Segmentation Algorithm The proposed method’s ﬂow diagram is shown below, Figure 2b excludes include The proposed method’s ﬂow diagram is shown below, Figure 2b excludes include step 5, which is the alarm strategy. ep 5, which Step 1: Region of interest (ROI) is selected to avoid unneeded information interference. It reduces the inﬂuence of low-resolution regions and improves the performance of image seg- mentation. The detection line gives additional scene information helping the RG algorithm to segment the scene. g The details of the processing are shown in Figure 2 based on an image from the YTR on 24 April 2021. The a priori knowledge image in Figure 2b shows the ROI and detection line selection. Due to the depth of ﬁeld span of the image being large, we select the right 1/3 area as the monitoring ROI for better resolution of river boundary and larger river area. g y g The water level is low in April and high in August according to the hydrological characteristics of the central basin of the YTR. As shown in the a priori knowledge image in Figure 2b, we positioned the detection line at the position of the yellow line where the Sensors 2023, 23, 4714 4 of 12 water level is believed to be unreachable. In August, the detection line was personally conﬁrmed. When the water level exceeds the detection line, move the water level up. When the detection line is not exceeded by the water level, keep it unchanged. 4 of 12 Figure 2. (a) Flow diagram of the proposed method. (b) Flow chart of the proposed algorithm using the real data on 24 April 2021 from the YTR. The color is used to distinguish between diﬀerent types of pixels. The pixels of the image have been divided into 4 categories using K- means clustering. The red box of the priori knowledge image is a region of interest. The red box of skeleton pixel image is an enlarged image of the area. Step 1: Region of interest (ROI) is selected to avoid unneeded information interference. It d th i ﬂ f l l ti i d i th f f i Figure 2. (a) Flow diagram of the proposed method. (b) Flow chart of the proposed algorithm using the real data on 24 April 2021 from the YTR. The color is used to distinguish between different types of pixels. 2 2 Hybrid Segmentation Algorithm 2.2. Hybrid Segmentation Algorithm The pixels of the image have been divided into 4 categories using K-means clustering. The red box of the priori knowledge image is a region of interest. The red box of skeleton pixel image is an enlarged image of the area. Step 2: RG algorithm is the process of aggregating pixels or subregions into larger areas di t d ﬁ d it i Th l ti f th d i t d th th it i ure 2. (a) Flow diagram of the proposed method. (b) Flow chart of the proposed algorithm ng the real data on 24 April 2021 from the YTR. The color is used to distinguish between ﬀerent types of pixels. The pixels of the image have been divided into 4 categories using K- ans clustering. The red box of the priori knowledge image is a region of interest. The red box of leton pixel image is an enlarged image of the area. Figure 2. (a) Flow diagram of the proposed method. (b) Flow chart of the proposed algorithm using the real data on 24 April 2021 from the YTR. The color is used to distinguish between different types of pixels. The pixels of the image have been divided into 4 categories using K-means clustering. The red box of the priori knowledge image is a region of interest. The red box of skeleton pixel image is an enlarged image of the area. ep 1: Step 2: The K-mean clustering algorithm is achieved by iteratively updating the initial estimate of the class center, using the procedures below: (1) Initialize class gravity center µi with i = 1 · · · n; (2) Assign each pixel to the nearest center of class C; (2) Assign each pixel to the nearest center of class C; (3) The update center is the mean of the gray value of pixels that are assigned to a particular class; (3) The update center is the mean of the gray value of pixels that are assigned to a particular class; p Repeat (2) and (3) until the algorithm converges. p (4) Repeat (2) and (3) until the algorithm converges. The K-means algorithm minimizes as much as possible the variance V between classes. The variance V is given by V = k ∑ i=1 ∑ xj∈Ci xj −µi 2 (2) (2) where, xj represents the jth pixel, 1 ≤j ≤n. µi represents the mean of the ith cluster, 1 ≤i ≤m. Ci represents the ith cluster. The K-means clustering image in Figure 2b shows the result of the clustering of all pixels in ROI. ROI has three primary objectives including the river and both sides of the river. The parameter K in the k-means clustering algorithm is related to the number of primary targets. So, we select K = 4. As a transition category, a new category has been added. Step 4: The principle of the traditional skeleton thinning algorithm is to extract the center contour of the object on the image. The algorithm uses 3 × 3 pixel windows to thin out the target from the target boundary to the target center until it is corroded to the point where it cannot be corroded (width of a single-layer pixel) in binary images. Then the image skeleton is obtained. Our proposed boundary-based skeleton thinning algorithm is to extract the boundary skeleton instead of the center skeleton. Image skeleton extraction based on the boundary is divided into two steps: (1) Extract the contour of the target and record these contour points. (2) Detect whether x pixels below these contour points are all non-target pixels in turn (in a binary image, the non-target pixel is represented by a color difference with the contour point pixel). When all pixels are non-target pixels, the contour point is a river boundary point pixel. ep 1: Step 2: Region of interest (ROI) is selected to avoid unneeded information interference. It duces the inﬂuence of low-resolution regions and improves the performance of image gmentation. The detection line gives additional scene information helping the RG gorithm to segment the scene. The details of the processing are shown in Figure 2 based on an image from the YTR on April 2021. The a priori knowledge image in Figure 2b shows the ROI and detection line RG algorithm is the process of aggregating pixels or subregions into larger areas according to predeﬁned criteria. The location of the seed point and the growth criterion are the parameters involved in the RG algorithm. The algorithm evaluates the relationship between each seed point and its eight neighboring areas. Similar points are used as the seed points for the next growth. The algorithm stops growing when no similar seed points exist. dge image in Figure 2b shows the ROI and detection line span of the image being large, we select the right 1/3 area olution of river boundary and larger river area p(x1, y1) −p(x2, y2) > G (1) a (1) Sensors 2023, 23, 4714 5 of 12 where, p(x2, y2) is the pixel gray value of seed points, p(x1, y1) is the gray value of adjacent pixel points, and G is the gray difference threshold. where, p(x2, y2) is the pixel gray value of seed points, p(x1, y1) is the gray value of adjacent pixel points, and G is the gray difference threshold. In this study, the location of the seed point is obtained by automatically spreading seed points in a square grid. Each seed point produces an RG image. Overlay all RG images that do not overlap the detection line. The max-connected domain is derived from the overlay graph, while RG-line is derived from the water level-side boundary information of the max-connected domain graph. The growth criterion is the gray value difference on both sides of the water level. The RG image in Figure 2b shows the maximum connectivity domain and adjunction region. The average pixel width of the river is roughly estimated as the side length of the square grid and half the width of the adjunction region. g Step 3: K-means clustering is a classiﬁcation algorithm that classiﬁes pixels in an image into K classes, while the Non-local mean (NLM) algorithm is used to obtain the denoised image. ep 1: Step 2: When all pixels are not non-target pixels, the contour point is not a river boundary point pixel. Where, x is related to the estimated average river width, generally one-tenth of the average river width is appropriate. We use the RG-line as a supplementary line to help remove noisy pixels from the skeleton thinning graph. The waterline is derived from the image boundary of the processed skeleton thinning graph on the water level side. The intermediate process image in Figure 2b shows the position of the adjunction region in the k-means clustering graph. The skeleton pixel image in Figure 2b shows the result of thinning skeleton of four types of pixels (red, green, blue, and yellow) and the process of ﬁltering in the Sensors 2023, 23, 4714 6 of 12 6 of 12 adjunction region of the skeleton graph. The skeleton graph has a linear shape, and the number of pixels within a connected component reﬂects the continuity of this boundary. Firstly, compare the size of the number of pixels in the maximum connected components for each pixel type. Secondly, compare the size of the number of pixels in the second connected component. Finally, compare the size of the number of pixels in the third connected com- ponent. In each comparison, the pixel type with the most pixels gets one point. When the score is the same, the pixel type with more pixels in the maximum connected component is preferred. p The blue pixel image in Figure 2b shows that the selected pixel categories are those with the highest score. The resulting image in Figure 2b shows the process of removing all pixels below the dividing line by taking the RG line as the dividing line to obtain the waterline. g Step 5: The monitoring system analyzes the difference in pixel water level to determine whether a dammed lake event has happened. The user set the water level alarm threshold, and the system alarm when the difference exceeds the threshold. The water level threshold used in this paper is 2 pixels. The pixel water level is the number of pixels within the image at the water level height under a ﬁxed ﬁeld of view. The difference in pixel water level equates to the difference in water level between two images captured at adjacent monitoring times. g Figure 3 shows the analysis of the difference in water level. ep 1: Step 2: Figure 3a shows the statistical differences between the pixel water level on 1 May 2021 and 24 April 2021. Figure 3b shows a magniﬁed image of the top three peaks. The value of pixel water level difference of −1, 0, and 1 is regarded as the numerical ﬂuctuation caused by algorithm resolution, so these values are not considered. In the top three peaks, when the number of negative values exceeds the number of positive values, the water level is regarded as rising, and when the number of positive values exceeds the number of negative values, the water level is regarded as falling. Considering the possibility of no variation in water level, a ratio of 0 over 50% is considered as reﬂecting no variation in water level. 7 of 12 Figure 3. Diﬀerence in water level analysis process. The waterline of 24 April 2021 and 1 May 2021 is used for example analysis. (a) Statistical results of diﬀerence in pixel water level at diﬀerent positions. (b) Enlarged graph of the diﬀerence in pixel water level at diﬀerent positions of the top three peaks in the red region. A cross indicates that the data there has been removed due to numerical ﬂuctuation. Figure 3. Difference in water level analysis process. The waterline of 24 April 2021 and 1 May 2021 is used for example analysis. (a) Statistical results of difference in pixel water level at different positions. (b) Enlarged graph of the difference in pixel water level at different positions of the top three peaks in the red region. A cross indicates that the data there has been removed due to numerical ﬂuctuation. Figure 3. Diﬀerence in water level analysis process. The waterline of 24 April 2021 and 1 May 2021 s used for example analysis. (a) Statistical results of diﬀerence in pixel water level at diﬀerent positions. (b) Enlarged graph of the diﬀerence in pixel water level at diﬀerent positions of the top hree peaks in the red region. A cross indicates that the data there has been removed due to numerical ﬂuctuation. Figure 3. Difference in water level analysis process. The waterline of 24 April 2021 and 1 May 2021 is used for example analysis. (a) Statistical results of difference in pixel water level at different positions. (b) Enlarged graph of the difference in pixel water level at different positions of the top three peaks in the red region. 3. Results and Discussion 3.1. Dataset 3. Results and Discussion 3.1. Dataset The proposed AMAS was used in Nyingchi, Yarlung Tsangpo River Basin, Tibet Autonomous Region, China. The river is frequently blocked by glacial debris ﬂows, and causing a signiﬁcant risk of the dammed lake. The format of the image is png, and the capture rate of the image is one frame per hour. The AMAS software interface consists of reference images, input images, and parameter settings. The reference image is the initial image of the water level, and the input image is the image that needs to be evaluated. When water levels exceed a predetermined threshold, the AMAS shows an alarm screen. The proposed AMAS was used in Nyingchi, Yarlung Tsangpo River Basin, Tibet Autonomous Region, China. The river is frequently blocked by glacial debris ﬂows, and causing a signiﬁcant risk of the dammed lake. The format of the image is png, and the capture rate of the image is one frame per hour. The AMAS software interface consists of reference images, input images, and parameter se1ings. The reference image is the initial image of the water level, and the input image is the image that needs to be evaluated. When water levels exceed a predetermined threshold, the AMAS shows an alarm screen. The river region is saturated with water vapor and sunlight, providing a stable tem- perature inversion. Hence, fog is common in the morning. Due to the viewing angle, the camera tends to produce overexposed images during midday. An hour after noon, when the sunlight shifts to the right angle is a good window for observation. Notably, noon in Tibet is two to three hours later than Beijing time because of the difference in longitude. In conclusion, we choose 4 pm as the viewing window. Figure 4 shows the clear image, the halo image, and the fog image from left to right. p The river region is saturated with water vapor and sunlight, providing a stable temperature inversion. Hence, fog is common in the morning. Due to the viewing angle, the camera tends to produce overexposed images during midday. An hour after noon, when the sunlight shifts to the right angle is a good window for observation. Notably, noon in Tibet is two to three hours later than Beijing time because of the diﬀerence in longitude. In conclusion, we choose 4 pm as the viewing window. 3. Results and Discussion 3.1. Dataset 3. Results and Discussion 3.1. Dataset Figure 4 shows the clear image, the halo image, and the fog image from left to right. (b) (c) Figure 4. Pictures of monitoring scenarios. (a) Image that can be recognized by the algorithm. (b) and (c) Occluded Image that were not recognized by the algorithm due to halo or fog. Figure 4. Pictures of monitoring scenarios. (a) Image that can be recognized by the algorithm. (b) and (c) Occluded Image that were not recognized by the algorithm due to halo or fog. (a) (c) (b) (a) (b) (c) Figure 4. Pictures of monitoring scenarios. (a) Image that can be recognized by the algorithm. (b) and (c) Occluded Image that were not recognized by the algorithm due to halo or fog. Figure 4. Pictures of monitoring scenarios. (a) Image that can be recognized by the algorithm. (b) and (c) Occluded Image that were not recognized by the algorithm due to halo or fog. ep 1: Step 2: A cross indicates that the data there has been removed due to numerical ﬂuctuation. 7 of 12 he top 7 of 12 he top Sensors 2023, 23, 4714 3.2. Results of the Proposed Scheme p , g g p target river is obscured by a hole or fog in the image. The green position indicates a normal photograph. The blue box is the viewing window at 16:00. Figure 5. Distribution of observed data across time from April to November. The observation window is in the blue box. Figure 5. Distribution of observed data across time from April to November. The observation window is in the blue box. 9 of recognize river border pixels accurately. Figure 7 shows the visual comparison results diﬀerent methods. Figure 5. Distribution of observed data across time from April to November. The observation window is in the blue box. Figure 5. Distribution of observed data across time from April to November. The observation window is in the blue box. recognize river border pixels accurately. Figure 7 shows the visual comparison results diﬀerent methods. Figure 6 shows the monthly variations in the water level identiﬁed by the proposed algorithm. In the selected area, the black lines represent the boundary lines identiﬁed by our method, and the red dots are the transverse centers of the boundary lines. The ﬂuctuating water level follows a pattern of initially rising, then lowering and then rising, with the highest water level occurring in August. Signiﬁcant gray diﬀerences created by white water waves will have a negative eﬀect on the RG algorithm, which uses gray diﬀerences as a growth criterion, causing the algorithm to terminate growth before the region of the target is completely detected. In addition, noise interference still presents in the image after denoising, resulting in the identiﬁcation of the non-river region as a river region. K-means clustering employs Euclidean distance based on pixel gray value as the judging criteria to classify pixels across the whole image. Compared to the local gray diﬀerence comparison of the RG algorithm, the global gray diﬀerence comparison is able to distinguish between the regions of white water waves and rivers more accurately. The proposed method is able to Figure 6. Monthly variation of water level from 24 April 2021 to 24 November 2021. The red do are the transverse centers of the boundary lines and Red lines make visual changes easier to notic Figure 6. Monthly variation of water level from 24 April 2021 to 24 November 2021. The red dots are the transverse centers of the boundary lines and Red lines make visual changes easier to notice. 3.2. Results of the Proposed Scheme In order to estimate the boundary recognition performance, the images identiﬁed by the proposed algorithm were compared with the RG algorithm. The results show that the proposed algorithm has a higher accurate ability for recognizing boundaries than the RG algorithm. Figure 5 shows the statistical distribution of all image data collected during the test. The y-coordinate shows the time change over the course of a day, and the x-coordinate shows the time change over the course of a month. The vacant position indicates that the camera lacks power, resulting in failure to take images. The red position indicates that the target river is obscured by a hole or fog in the image. The green position indicates a normal photograph. The blue box is the viewing window at 16:00. Figure 6 shows the monthly variations in the water level identiﬁed by the proposed algorithm. In the selected area, the black lines represent the boundary lines identiﬁed by our method, and the red dots are the transverse centers of the boundary lines. The ﬂuctuating water level follows a pattern of initially rising, then lowering and then rising, with the highest water level occurring in August. Signiﬁcant gray differences created by white water waves will have a negative effect on the RG algorithm, which uses gray differences as a growth criterion, causing the algorithm to terminate growth before the region of the target is completely detected. In addition, noise interference still presents in the image after denoising, resulting in the identiﬁcation of the non-river region as a river region. K-means clustering employs Euclidean distance based on pixel gray value as the judging criteria to classify pixels across the whole image. Compared to the local gray difference comparison of the RG algorithm, the global gray difference comparison is able to distinguish between the regions of white water waves Sensors 2023, 23, 4714 8 of 12 he test. rdinate and rivers more accurately. The proposed method is able to recognize river border pixels accurately. Figure 7 shows the visual comparison results of different methods. p , g g p target river is obscured by a hole or fog in the image. The green position indicates a normal photograph. The blue box is the viewing window at 16:00. and rivers more accurately. The proposed method is able to recognize river border pixels accurately. Figure 7 shows the visual comparison results of different methods. 3.2. Results of the Proposed Scheme Figure 6. Monthly variation of water level from 24 April 2021 to 24 November 2021. The red do are the transverse centers of the boundary lines and Red lines make visual changes easier to notic Figure 6. Monthly variation of water level from 24 April 2021 to 24 November 2021. The red dots are the transverse centers of the boundary lines and Red lines make visual changes easier to notice. Sensors 2023, 23, 4714 9 of 12 ed dots 9 of 12 ed dots Figure 7. Visual comparison. Observations are made every seven days from 24 April 2021 to 24 November 2021, and the color red denotes the water level identified by RG. The color blue is the water Figure 7. Visual comparison. Observations are made every seven days from 24 April 2021 to 24 November 2021, and the color red denotes the water level identified by RG. The color blue is the water level identified by the proposed algorithm. The color green indicates the overlap of water level identified by the two algorithms. (a) Visual comparison of 24 April 2021 (b) Visual comparison of 1 May 2021 (c) Visual comparison of 1 May 2021 (d) Visual comparison of 1 May 2021 to 24 November 2021. Figure 7. Visual comparison. Observations are made every seven days from 24 April 2021 to 24 November 2021, and the color red denotes the water level identified by RG. The color blue is the water Figure 7. Visual comparison. Observations are made every seven days from 24 April 2021 to 24 November 2021, and the color red denotes the water level identified by RG. The color blue is the water level identified by the proposed algorithm. The color green indicates the overlap of water level identified by the two algorithms. (a) Visual comparison of 24 April 2021 (b) Visual comparison of 1 May 2021 (c) Visual comparison of 1 May 2021 (d) Visual comparison of 1 May 2021 to 24 November 2021. Figure 7a,b and c, respectively, show the enlarged visual comparison of the water level on 24 April 2021, 1 May 2021, and 9 May 2021. Figure 7d shows a visual comparison of water levels from 17 May 2021 to 24 November 2021. The boundary pixels in the contrast image underwent a 5 × 5 convolution core operation for better visual presentation. 3.2. Results of the Proposed Scheme The accuracy is the proportion of samples correctly identiﬁed by the classiﬁer out of the total number of samples. Although accuracy can evaluate the overall correctness of a classiﬁer, it is not a very efﬁcient measuring indicator when the categories in the total samples have a skewness distribution. Even if the classiﬁer’s classiﬁcation accuracy is greater than 99%, this does not necessarily indicate that the model is good. Hence, the classiﬁer will frequently use the accuracy rate and recall rate to measure. We use the binary model index to compare the alarm methods using different segmentation algorithms. The indexes of the binary model are accuracy (ACC), precision (PRE), recall (TPR, True Positive Rate), false alarm (FPR, False Positive Rate), and miss rate (FNR, True Negative Rate). ACC = (TP + TN)/(TP + FN + FP + TN) (3) PRC = TP/(TP + FP) (4) TPR = TP(TP + FP) (5) (3) (4) (5) Sensors 2023, 23, 4714 10 of 12 10 of 12 FPR = FP(FP + TN) (6) FNR = FN(TP + FN) (7) 𝑇𝑃𝑅= 𝑇𝑃ሺ𝑇𝑃൅𝐹𝑃ሻ (6) 𝐹𝑃𝑅= 𝐹𝑃ሺ𝐹𝑃൅𝑇𝑁ሻ (7) 𝐹𝑁𝑅= 𝐹𝑁ሺ𝑇𝑃൅𝐹𝑁ሻ (8) (6) ) ) (7) ) where, True positive (TP) means that the positive sample is predicted to be positive, False negative (FN) means that a positive sample is predicted to be negative, False positive (FP) means that a negative sample is predicted to be positive, and True negative (TN) means that a negative sample is predicted to be negative. The results show that the proposed algorithm is superior to the traditional RG algorithm in accuracy, precision, recall, false alarm, and miss rate. where, True positive (TP) means that the positive sample is predicted to be positive, False egative (FN) means that a positive sample is predicted to be negative, False positive (FP) means that a negative sample is predicted to be positive, and True negative (TN) means hat a negative sample is predicted to be negative. The results show that the proposed gorithm is superior to the traditional RG algorithm in accuracy, precision, recall, false arm, and miss rate. When all samples were evaluated to be negative, the accuracy index remained ex- tremely high if the majority of samples were negative. The precision index suffers from the same issue, when a False positive (FP) is low, the precision index will be high. Therefore, we measured the recall (TPR) of statistical results. 3.2. Results of the Proposed Scheme Figure 8 shows the statistics and analysis of dammed lake alarm. Figure 8a shows the statistics of the alarm. Figure 8b shows the comparison of indicators using different methods. The proposed algorithm improved in these three dimensions by 29.12%, 27.08%, and 17.65%, respectively. The proposed algorithm decreases false alarm and miss rates by 45.46% and 17.65%, respectively. The proposed algorithm is superior to the RG algorithm in all ﬁve dimensions. When all samples were evaluated to be negative, the accuracy index remained xtremely high if the majority of samples were negative. The precision index suffers from he same issue, when a False positive (FP) is low, the precision index will be high. Therefore, e measured the recall (TPR) of statistical results. Figure 8 shows the statistics and analysis f dammed lake alarm. Figure 8a shows the statistics of the alarm. Figure 8b shows the omparison of indicators using different methods. The proposed algorithm improved in hese three dimensions by 29.12%, 27.08%, and 17.65%, respectively. The proposed gorithm decreases false alarm and miss rates by 45.46% and 17.65%, respectively. The roposed algorithm is superior to the RG algorithm in all five dimensions. (a) (a) (b) Figure 8. Alarm situation and indicator comparison. We counted the alarms from 1 May 2021 to 24 November 2021 and compared the corresponding indicators. (a) Alarm distribution. Green indicates that the system does not automatically alarm, and red indicates that the system automatically Figure 8. Alarm situation and indicator comparison. We counted the alarms from 1 May 2021 to 24 November 2021 and compared the corresponding indicators. (a) Alarm distribution. Green indicates that the system does not automatically alarm, and red indicates that the system automatically alarms. The three cells in a monitor, from left to right, are labeled as follows: RG, Ours, and truth. For simpler visual comparison, the true values are divided by a dotted line. (b) Indicator comparison. Comparison between RG and Ours in ACC, PRE, TPR, FPR, and FNR. (b) (b) igure 8. Alarm situation and indicator comparison. We counted the alarms from 1 May 2021 to 24 ovember 2021 and compared the corresponding indicators. (a) Alarm distribution. Green indicates hat the system does not automatically alarm, and red indicates that the system automatically Figure 8. Alarm situation and indicator comparison. We counted the alarms from 1 May 2021 to 24 November 2021 and compared the corresponding indicators. (a) Alarm distribution. 3.2. Results of the Proposed Scheme Green indicates that the system does not automatically alarm, and red indicates that the system automatically alarms. The three cells in a monitor, from left to right, are labeled as follows: RG, Ours, and truth. For simpler visual comparison, the true values are divided by a dotted line. (b) Indicator comparison. Comparison between RG and Ours in ACC, PRE, TPR, FPR, and FNR. 4. Conclusions Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data is unavailable due to privacy. Data Availability Statement: The data is unavailable due to privacy. Conﬂicts of Interest: The authors declare no conﬂict of interest. References 1. Skea, J.; Shukla, P.; Kılkı¸s, ¸S. Climate Change 2022: Mitigation of Climate Change; Cambridge University Press: Cambridge, MA, USA, 2022. 2. Shi, Y.F.; Shen, Y.P.; Kang, E.; Li, D.L.; Ding, Y.J.; Zhang, G.W.; Hu, R.J. Recent and future climate change in Northwest China. Clim. Change 2007, 80, 379–393. [CrossRef] Xiao, T.; He, J. Dammed lake bursting and ﬂood routing in the Yarlung Tsangpo Grand Canyon in October , 583, 124603. [CrossRef] 3. Chen, C.; Zhang, L.; Xiao, T.; He, J. Dammed lake bursting and ﬂood routing in the Yarlung Tsangp 2018. J. Hydrol. 2020, 583, 124603. [CrossRef] y 4. Zhuang, Y.; Yin, Y.; Xing, A.; Jin, K. Combined numerical investigation of the Yigong rock slide-debris avalanche and subsequent dam-break ﬂood propagation in Tibet, China. Landslides 2020, 17, 2217–2229. [CrossRef] p p g 5. Jiang, F.; Dai, X.; Xie, Z.; Xu, T.; Yin, S.; Qu, G.; Yang, S.; Zhang, Y.; Yang, Z.; Xu, J. Flood inundation evolution of Dammed lake and evaluation of regional ecological spatiotemporal response—A case study of Sichuan-Tibet region. Environ. Sci. Pollut. Res. 2022, 29, 71290–71310. [CrossRef] [PubMed] [ ] [ ] 6. Tian, S.; Chen, N.; Rahman, M.; Hu, G.; Peng, T.; Zhang, Y.; Liu, M. New insights into the occurrence of the catastrophic Zhaiban slope debris ﬂow that occurred in a dry valley in the Hengduan Mountains in southwest China. Landslides 2022, 19, 647–657. [CrossRef] 7. An, B.; Wang, W.; Yang, W.; Bai, L.; Zhang, F.; Zeng, C.; Wang, L.; Zhou, J.; Li, X.; Li, J.; et al. Process, mechanisms, and early warning of glacier collapse-induced river blocking disasters in the Yarlung Tsangpo Grand Canyon, southeastern Tibetan Plateau. Sci. Total Environ. 2021, 816, 151652. [CrossRef] 8. Zheng, G.; Zong, H.; Zhuan, X.; Wang, L. High-accuracy surface-perceiving water level gauge with self-calibration for hydrogra- phy. IEEE Sens. J. 2010, 10, 1893–1900. [CrossRef] 9. Hall, A.C.; Schumann, G.J.P.; Bamber, J.L.; Bates, P.D.; Trigg, M.A. Geodetic corrections to Amazon River water level gauges using ICESat altimetry. Water Resour. Res. 2012, 48, 6. [CrossRef] 9. Hall, A.C.; Schumann, G.J.P.; Bamber, J.L.; Bates, P.D.; Trigg, M.A. Geodetic corrections to Amazon R ICESat altimetry. Water Resour. Res. 2012, 48, 6. [CrossRef] Geodetic aspects of water-level gauge elevations/elevation changes and gauge set-points in coastal waters sch. 2012, 56, 257–275. 10. Weiss, R.; Sudau, A. Geodetic aspects of water-level gauge elevations/elevation changes and gauge Hydrol. Wasserbewirtsch. 2012, 56, 257–275. 11. 4. Conclusions In this paper, a hybrid algorithm based on RG and k-means clustering is proposed for an automatic monitoring alarm sub-system for dammed lake events. Firstly, ROI and detection lines are used to provide prior knowledge of the scenario. Then, a hybrid segmentation algorithm is proposed to identify river boundaries. The results of the visual comparison show that the proposed algorithm has more accurate boundary recognition ability than the RG algorithm. Lastly, the identiﬁed boundaries are analyzed numerically. The results show that the proposed algorithm performs well in quantitative measurement compared with other methods. Regarding accurate alarms, the performance of the proposed method is better than that of the RG method in accuracy (89.29%), precision (93.75%), and recall (88.23%). In terms of its ability to reduce errors, the proposed method is also superior to the RG method in false alarm (9.09%) and miss rate (11.76%), hence decreasing resource waste caused by false alarms. In the future, the proposed method will be tested in different geographical regions to strengthen the algorithm’s robustness, and at the same time, image enhancement will be carried out in image preprocessing to ﬁlter halo and fog and obtain Sensors 2023, 23, 4714 11 of 12 11 of 12 clear images to realize better performance of the proposed method. This research proposes an automatic monitoring alarm sub-system as a solution to the problem of dammed lake monitoring. Future applications of this method include ﬂood warnings and seasonal river state judgment. clear images to realize better performance of the proposed method. This research proposes an automatic monitoring alarm sub-system as a solution to the problem of dammed lake monitoring. Future applications of this method include ﬂood warnings and seasonal river state judgment. Author Contributions: Conceptualization, X.W. and B.A.; methodology, Z.C., L.S. and X.W.; software, L.S. and X.Z.; validation, Z.C., L.S. and X.W.; formal analysis, Z.C. and L.S.; funding acquisition, X.W. and B.A.; resources, W.Y., Z.W. and X.W.; investigation, W.Y. and Z.W.; data curation, Z.C. and L.S.; writing—original draft preparation, Z.C. and L.S.; writing—review and editing, X.W., Y.Z. and F.Z.; visualization, Z.C., L.S. and X.W.; supervision, X.W. and B.A. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the Lhasa Earth System Multi-Dimension Observatory Network, LEMON, Youth Innovation Promotion Association of the Chinese Academy of Sciences (Y2021044). Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. 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. References Heiner, B.; Barfuss, S.L.; Johnson, M.C. Conditional assessment of ﬂow measurement accuracy. J. Irrig. Drain. Eng. 2011, 137, 367–374. [CrossRef] 12. Chen, Y.-C. Flood discharge measurement of a mountain river—Nanshih River in Taiwan. Hydrol. Earth Syst. Sci. 2013, 17, 1951–1962. [CrossRef] 13. Hoque, R.; Nakayama, D.; Matsuyama, H.; Matsumoto, J. Flood monitoring, mapping and assessing capabilities using RADARSAT remote sensing, GIS and ground data for Bangladesh. Nat. Hazards 2011, 57, 525–548. [CrossRef] 14. Ji, Y.; Zhang, M.; Wang, Y.; Wang, P.; Wang, A.; Wu, Y.; Xu, H.; Zhang, Y. Microwave-photonic sensor for remote water-level monitoring based on chaotic laser. Int. J. Bifurc. Chaos 2014, 24, 1450032. [CrossRef] 15. Boon, J.D. Reducing Wave-Induced Microwave Water-Level Measurement Error with a Least Squares–Designed Digital Filter. J. Atmos. Ocean. Technol. 2014, 31, 491–502. [CrossRef] 16 Korostynska O ; Mason A ; Al Shamma’a A Microwave sensors for the non invasive monitoring of industrial and medical 16. Korostynska, O.; Mason, A.; Al-Shamma’a, A. Microwave sensors for the non-invasive monitoring of industrial and medical applications. Sens. Rev. 2014, 34, 182–191. [CrossRef] 12 of 12 12 of 12 Sensors 2023, 23, 4714 17. Lin, Y.T.; Lin, Y.C.; Han, J.Y. Automatic water-level detection using single-camera images with varied poses. Measurement 2018, 127, 167–174. [CrossRef] , [ ] 18. Lo, S.W.; Wu, J.H.; Lin, F.P.; Hsu, C.H. Cyber Surveillance for Flood Disasters. Sensors 2015, 15, 2369–2387. [CrossRef] y 19. Arshad, B.; Ogie, R.; Barthelemy, J.; Pradhan, B.; Verstaevel, N.; Perez, P. Computer Vision and IoT Monitoring and Mapping: A Systematic Review. Sensors 2019, 19, 5012. [CrossRef] g pp g y 20. Udomsiri, S.; Iwahashi, M. Design of FIR Filter for Water Level Detection. World Acad. Sci. Eng. Technol. Int. J. Electr. Comput. Energetic Electron. Commun. Eng. 2008, 2, 2663–2668. g g 21. Lo, S.W.; Wu, J.H.; Lin, F.P.; Hsu, C.H. Visual Sensing for Urban Flood Monitoring. Sensors 2015, 15, 20006–20029. [CrossRef] 1. Lo, S.W.; Wu, J.H.; Lin, F.P.; Hsu, C.H. Visual Sens 22. Dou, G.; Chen, R.S.; Han, C.T.; Liu, Z.W.; Liu, J.F. Research on Water-Level Recognition Method Based Convolutional Neural Networks. Water 2022, 14, 1890. [CrossRef] 22. Dou, G.; Chen, R.S.; Han, C.T.; Liu, Z.W.; Liu, J.F. Research on Water-Level Recognition Method Based on Image Processing and Convolutional Neural Networks. Water 2022, 14, 1890. [CrossRef] 23. Chen, G.J.; Bai, K.J.; Lin, Z.C.; Liao, X.X.; Liu, S.P.; Lin, Z.Y.; Zhang, Q.; Jia, X.P. References Method on water level based on image processing. Signal Image Video Process. 2021, 15, 33–41. [CrossRef] g p g g g 24. Martinis, S.; Twele, A.; Strobl, C.; Kersten, J.; Stein, E. A Multi-Scale Flood Monitoring System Based on Fully Automatic MODIS and TerraSAR-X Processing Chains. Remote Sens. 2013, 5, 5598–5619. [CrossRef] g 25. Yu, Z.B.; Chang, R.C.; Chen, Z. Automatic Detection Method for Loess Landslides Based on GEE and an Improved YOLOX Algorithm. Remote Sens. 2022, 14, 4599. [CrossRef] 26. Shah-Hosseini, R.; Homayouni, S.; Safari, A. Environmental monitoring based on automatic change detection from remotely sensed data: Kernel-based approach. J. Appl. Remote Sens. 2015, 9, 21. [CrossRef] pp pp 27. Khan, A.A.; Vrabie, V.; Mars, J.I.; Girard, A.; D’Urso, G. Automatic Monitoring System for Singularity Detection in Dikes by DTS Data Measurement. IEEE Trans. Instrum. Meas. 2010, 59, 2167–2175. [CrossRef] pp pp 27. Khan, A.A.; Vrabie, V.; Mars, J.I.; Girard, A.; D’Urso, G. Automatic Monitoring System for Singularity Detection in Dikes by DTS Data Measurement. IEEE Trans. Instrum. Meas. 2010, 59, 2167–2175. [CrossRef] .; Han, J.Y.; Su, Y.F.; Jang, J.H. Quantifying Flood Water Levels Using Image-Based Volunteered Geographic Sens. 2020, 12, 706. [CrossRef] 28. Lin, Y.T.; Yang, M.D.; Han, J.Y.; Su, Y.F.; Jang, J.H. Quantifying Flood Water Levels Using Image-Ba Information. Remote Sens. 2020, 12, 706. [CrossRef] 29. Liu, H.; Jezek, K.C. Automated extraction of coastline from satellite imagery by integrating Canny boundary detection and locally adaptive thresholding methods. Int. J. Remote Sens. 2004, 25, 937–958. [CrossRef] 29. Liu, H.; Jezek, K.C. Automated extraction of coastline from satellite imagery by integrating Canny boundary detection and locally adaptive thresholding methods. Int. J. Remote Sens. 2004, 25, 937–958. [CrossRef] p g J , , [ ] 30. Lo, S.W.; Wu, J.H.; Chen, L.C.; Tseng, C.H.; Lin, F.P.; Hsu, C.H. Uncertainty Comparison of Visual Sensing in Adverse Weather Conditions. Sensors 2016, 16, 1125. [CrossRef] p g J 30. Lo, S.W.; Wu, J.H.; Chen, L.C.; Tseng, C.H.; Lin, F.P.; Hsu, C.H. Uncertainty Comparison of Visual Sensing in Adverse Weather Conditions. Sensors 2016, 16, 1125. [CrossRef] 31. Muhadi, N.A.; Abdullah, A.F.; Bejo, S.K.; Mahadi, M.R.; Mijic, A. Image Segmentation Methods for Flood Monitoring System. Water 2020, 12, 1825. [CrossRef] 31. Muhadi, N.A.; Abdullah, A.F.; Bejo, S.K.; Mahadi, M.R.; Mijic, A. Image Segmentation Methods for Flood Monitoring System. Water 2020, 12, 1825. [CrossRef] ullah, A.F.; Bejo, S.K.; Mahadi, M.R.; Mijic, A. References Image Segmentation Methods for Flood Monitoring System [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/W2087804854	http://www.jafs.com.pl/pdf-66309-7261?filename=Genetic relationships.pdf	English	null	Genetic relationships among time of egg formation, clutch traits and traditional selection traits in laying hens	Journal of Animal and Feed Sciences	2,010	cc-by	3,919	Journal of Animal and Feed Sciences, 19, 2010, 452-459 Journal of Animal and Feed Sciences, 19, 2010, 452-459 5 Corresponding author: e-mail: tomasz@jay.up.poznan.pl Genetic relationships among time of egg formation, clutch traits and traditional selection traits in laying hens A. Wolc 1,2, M. Bednarczyk 3, M. Lisowski 4 and T. Szwaczkowski lPoznan University of Life Sciences, Department of Genetics and Animal Breeding Wofynska 33, 60-637 Poznah, Poland 2Iowa State University, Department of Animal Science Ames, Iowa 50011-3150, USA 3 University of Technology and Life Sciences, Department of Animal Biotechnology Ks. Kordeckiego 20, 85-225 Bydgoszcz, Poland 4National Research Institute of Animal Production, Department of Animal Reproduction Biotechnology 32-083 Bailee, Poland lPoznan University of Life Sciences, Department of Genetics and Animal Breeding Wofynska 33, 60-637 Poznah, Poland 2Iowa State University, Department of Animal Science Ames, Iowa 50011-3150, USA 3 University of Technology and Life Sciences, Department of Animal Biotechnology Ks. Kordeckiego 20, 85-225 Bydgoszcz, Poland 4National Research Institute of Animal Production, Department of Animal Reproduction Biotechnology 32-083 Bailee, Poland INTRODUCTION It is well known that long term selection in laying hens led to substantial decrease of genetic variability in traditional selection traits, for instance initial egg production, egg weight, body weight and age at first egg (Szwaczkowski, 2003). Therefore, other groups of traits describing egg production such as clutch traits were considered as a potential selection criterion (Sheldon and Yoo, 1993). Formation of clutches results from longer than 24 h time needed for egg formation (Bednarczyk et al., 2000; Chen and Tixier-Boichard, 2003a). Therefore analysis of time interval between consecutively laid eggs seems closer to the physiology of the egg production process. Bird ovulatory cycle is thought to be controlled by a circadian rhythm, entrained by the daily light-dark cycle, that governs the timing of the preovulatory surge of luteinizing hormone and by the growth and maturation of the follicles. More details on physiological background of egg formation are given by Lillpers (1993). The laying rhythm has been studied for several species of domestic fowls, for instance laying hens (Lillpers and Wilhelmson, 1993; Luc et al., 1996, Bednarczyk et al., 2000), ducks (Simmons and Hetzel, 1983), geese (Rosinski et al., 2006) and turkeys (Pyrzak and Siopes, 1989). In general, the results seem to be promising. Miandmients et al. (1993) found positive relationship between egg production and clutch length. However, majority of above mentioned studies were based on evaluation of phenotypic relationship between 'new' and 'classical' egg production traits. It should be stressed that the data from the population used in this study were analysed by Bednarczyk et al. (2000) using the sire-dam model without integrating connections and additional relationships between individuals. Thus, it could have led to overestimation of residual variances and, in consequence, bias in the estimates of genetic parameters of the studied traits. Nowadays, the advantages of animal model with full relationship matrix over sire-dam model to analyse genetic and phenotypic parameters of multigenerational populations seem to be evident. The objectives of this study were to estimate the heritability of egg formation, clutch characters and traditional selection traits as well as the genetic and phenotypic correlations between them via multitrait animal model. ABSTRACT In a population of Rhode Island White hens heritability of egg formation, clutch characters and traditional selection traits as well as the genetic and phenotypic correlations between them were estimated via multitrait animal model. Over 1300 birds and about 4000 birds were recorded in two consecutive generations for oviposition time and traditional traits, respectively. The heritability estimates obtained for age at first egg (h 2=0.42), egg weight (h 2=0.50) and body weight (h 2=0.42) were considerably higher than those for initial egg production (h 2=0.22), clutch traits (h 2 between 0.11 and 0.23) and oviposition time (h 2 between 0.13 and 0.19). Both genetic and phenotypic correlations between clutch traits and traditional selection traits were low, except for initial egg production and maximal clutch length (rg=0.40 and rp=0.38). As expected, negative correlations were registered for number of clutches and average clutch length. It indicates an opportunity of selection aimed at improvement of egg production persistence by an increase in the average clutch size. Oviposition time was favourably correlated with traditional selection criteria. KEY WORDS: laying hens, oviposition time, clutch traits, heritability, correlations 5 Corresponding author: e-mail: tomasz@jay.up.poznan.pl 453 WOLC A. ETAL. WOLC A. ETAL. MATERIAL AND METHODS Two consecutive generations ofthe A22 line (Rhode Island White) were recorded. The genetic improvement programme applied is based on classical selection index including five traits: initial egg production, rate of initial egg production, body weight, average egg weight and age at first egg. More details on the selection SELECTION TRAITS IN LAYING HENS 454 index applied were described by Wezyk and Szwaczkowski (1997). The study was undertaken at Poultry Research Branch of National Research Institute of Animal Production at Zakrzewo near Poznan (Poland). From eighteenth week of life birds were kept in individual cages with 750 cm 2 per hen. After stimulation period lightdark time ratio was 14:10 h. Compound feed in mash form was provided ad libitum. Temperature and humidity were automatically controlled. A total of 720 cages were connected to a computer device that automatically recorded time and cage in which an egg was laid. Each bird from such cage was characterized for within clutch time of egg formation from first oviposition until the 64 th week of age using an electronic data collection system described by Bednarczyk et al. (1997). Bar code and hen number expressed in digits placed in front of cage identified each hen. At oviposition the data read from the mark were recorded by the LAG- 950 Laser Reader, coded in the memory of BCP-601 terminal and then transferred to the computer for further analysis. The records were used to define the following traits: mean oviposition time (MT), within bird variance of oviposition time (VT), minimum oviposition time (MINT), maximum oviposition time (MAXT) for hens in the experimental setting. Other production traits were also measured for all hens from the given generation. Based on individual daily egg records number of clutches (CN), average (CS) and maximal (MCS) clutch length and initial egg production (IEP-38 th week of life) were determined. Additionally traits included in the selection index were measured: body weight (BW) at thirty weeks, age at first egg (AFE) and egg weight (EW). Statistical description of these data is given in Table 1. Table 1. y = Xb + Za + e y = Xb + Za + e where: y - the (N x 1) vector of observations on t traits with following form: y = (y >x f y 2 . , t y j J;b - the (pt x 1) vector of fixed effects (two years and eight hatch periods) with following form: b = (b\ ,b 2 ,...,b^) ; a - the (qt x 1) vector of random additive genetic effects with following form: a = (a\, a 2,..., a t ) ; e - the (N x 1) vector of random errors with following form: e = ( e t , e 2 e < ) ; X and Z are N x pt and N x qt these incidence matrices for fixed and random effects, respectively. Whereas X = (I t ® X j ) and Z = (I t ® Z.) with X. and Z{ are the incidence matrices for single trait, I t is diagonal matrix and <8> denotes Kronecker product. where: y - the (N x 1) vector of observations on t traits with following form: y = (y >x f y 2 . , t y j J;b - the (pt x 1) vector of fixed effects (two years and eight hatch periods) with following form: b = (b\ ,b 2 ,...,b^) ; a - the (qt x 1) vector of random additive genetic effects with following form: a = (a\, a 2,..., a t ) ; e - the (N x 1) vector of random errors with following form: e = ( e t , e 2 e < ) ; X and Z are N x pt and N x qt these incidence matrices for fixed and random effects, respectively. Whereas X = (I t ® X j ) and Z = (I t ® Z.) with X. and Z{ are the incidence matrices for single trait, I t is diagonal matrix and <8> denotes Kronecker product. Traits were analysed in groups of seven due to computing demands with the same model used for all traits. REML approach was applied to estimate variance components in the DXMUX programme which belongs to the DFREML package (Meyer, 2001). The following parameters were estimated: heritability, genetic and phenotypic correlations. Standard deviations of heritability estimates were derived from the average information matrix (Meyer, 2001). WOLC A. ETAL. WOLC A. ETAL. 455 The following multitrait animal model was employed: MATERIAL AND METHODS Description of the data set Trait 1 Number of recorded Mean SD Trait 1 individuals Mean SD AFE, days 4131 147.44 10.59 IEP, psc 4081 112.33 14.23 EW,g 3976 58.57 4.15 BW, g 4129 1940.0 187.7 CN 4118 32.47 12.36 CS, psc 4122 9.10 5.07 MCS, psc 4116 49.19 24.68 MT, hours 1369 24.09 0.25 VT, squared hours 1350 1.59 1.20 MINT, hours 1370 21.03 1.36 MAXT, hours 1373 27.28 1.36 1 AFE - age at first egg, IEP - initial egg production, EW - egg weight, BW - body weight, CN - number of clutches, CS - average clutch lenght, MCS - maximal clutch length, MT - mean oviposition time, VT - variance of oviposition time, MINT - minimal oviposition time, MAXT - maximal oviposition time Table 1. Description of the data set RESULTS AND DISCUSSION Estimates of heritability of the studied traits as well as genetic and phenotypic correlations between them are listed in Table 2. Generally, the heritability estimates obtained for classical characters (especially AFE, EW and BW) were considerably higher than those for clutch traits. It should be recalled that these traits and IEP have been included in selection index applied to the studied population. A number of reports on heritabilites of the so-called classical traits are available in the literature (see Szwaczkowski, 2003). The effectiveness of the current genetic improvement programme in the population can be implied from the estimates of genetic parameters of classical traits. Body weight, egg weight and age at first egg were moderately heritable, whereas for traits directly related to egg production lower estimates of heritability were obtained. A wide range of heritability estimates of traditional selection traits are available in the literature. For instance, heritability of body weight varies for both layer and broiler chicken (Bednarczyk et al., 2000; Szydlowski and Szwaczkowski, 2001; Zieba et al., 2003). Egg weight was found to be moderately heritable in the study herein. Similar estimates were obtained by Nurgiartiningsih et al. (2004) for SELECTION TRAITS IN LAYING HENS 456 White Leghorns as well as by Zieba et al. (2003) for Rhode Island strains. Body weight was positively correlated with egg weight, however the correlation was not very high (0.25 at the phenotypic and 0.21 at the genetic level). It corresponds to the results published so far (Szwaczkowski, 2003). Table 2. RESULTS AND DISCUSSION Estimates of heritabilities and their standard deviations in the diagonal, genetic correlations (above diagonal) and phenotypic correlations (below diagonal) between traits 1 studied Traits AFE IEP EW BW CN CS MCS MT VT MINT MAXT AFE 0.42 -o.m 0.013 0.031 -0.160 0.110 -0.1193 0.1562 -0.4249 0.2831 -0.0957 (±0.04) IEP -0.595 0.22 -0.028 0.038 -0.079 0.129 0.4001 -0.3672 0.3391 -0.3007 0.1802 (±0.03) EW -0.018 0.044 0.50 0.215 -0.027 0.040 -0.1301 -0.1635 -0.4141 0.1581 -0.0889 (±0.04) BW -0.195 0.147 0.253 0.42 -0.126 0.124 -0.0976 0.1048 0.1135 -0.0107 0.0224 (±0.04) CN -0.036 -0.294 -0.036 -0.159 0.23 -0.989 -0.813 0.577 0.195 0.257 0.163 (±0.04) CS -0.030 0.329 0.042 0.152 -0.791 0.23 0.885 -0.558 -0.175 -0.283 -0.112 (±0.03) MCS 0.022 0.382 -0.004 0.011 -0.518 0.572 0.11 -0.323 -0.210 -0.240 -0.030 (±0.02) MT 0.162 -0.178 -0.011 -0.130 0.265 -0.220 -0.139 0.13 -0.478 0.580 -0.178 (±0.05) VT 0.048 -0.025 -0.014 -0.023 0.077 -0.092 -0.037 0.047 0.18 -0.827 0.615 (±0.02) MINT -0.121 -0.016 -0.006 0.047 0.112 -0.104 -0.190 0.256 -0.626 0.19 -0.534 (±0.06) MAXT 0.158 -0.035 -0.012 -0.103 -0.001 0.012 0.145 0.238 0.646 -0.544 0.19 (±0.07) 'AFE - age at first age, IEP - initial egg production, EW - egg weight, BW - body weight, CN - number of clutches, CS - average clutch length, MCS - maximal clutch length, MT - mean oviposition time, VT - variance of oviposition time, MINT - minimal oviposition time, MAXT - maximal oviposition time Table 2. Estimates of heritabilities and their standard deviations in the diagonal, genetic (above diagonal) and phenotypic correlations (below diagonal) between traits 1 studied 'AFE - age at first age, IEP - initial egg production, EW - egg weight, BW - body weight, CN - number of clutches, CS - average clutch length, MCS - maximal clutch length, MT - mean oviposition time, VT - variance of oviposition time, MINT - minimal oviposition time, MAXT - maximal oviposition time As it was already mentioned, initial egg production was characterized by lower heritability. It is caused by long term selection on this trait. It is well known that estimates from commercial populations are usually lower than from experimental or local breeds. Because of decreased heritability and genetic progress for initial or total egg production, new measurements were tested for their usefulness in the breeding programmes. The first group of these traits were clutch traits. WOLC A. ETAL. WOLC A. ETAL. for CN and CS. As expected, the heritability estimates reported by other authors varied depending on the type of population under study and statistical model. Chen and Tixier-Boichard (2003b) concluded that average clutch size can be effectively improved by selection in dwarf laying hens whereas the selection response was not considerably affected by naked neck gene. The heritability estimates by Chen and Tixier-Boichard (2003b) were relatively higher than ones estimated in the present study, but it should be mentioned that they were obtained for transformed data. It is well know that data transformation often leads to higher heritability estimates since an empirical distribution of given trait better approximates normality. On the other hand, high estimates for untransformed observations of clutch traits in two lines of laying hens were found by Luc et al. (1996). Various estimates of heritability (from sire, dam and sire+dam components) for laying rhythm in geese were estimated by Rosinski et al. (2006). It may indicate a more complex genetic background of these traits. Both genetic and phenotypic correlations between clutch traits and traditional selection traits were low, except for initial egg production and maximal clutch length (rg=0.40 and rp=0.38). As expected, negative correlations were registered for number of clutches and average clutch length. It indicates an opportunity of selection aimed at improvement of egg production persistence by an increase in the average clutch size. Clutch traits were extensively evaluated over the last decades. The different approaches included sire+dam model (Bednarczyk, 2000), and animal model estimates confirmed by selection experiment (Chen and Boichard, 2003b). Selection experiment proved that it is possible to obtain significant progress in clutch length and that correlated response can be achieved in egg number. The studies on oviposition times were already carried out in the 1970s (McClung et al., 1976); however, they were newly undertaken in the XXI century (Lewis et al., 2004) due to reduced selection response in the traditional selection criteria applied in layer populations. Average time between eggs in the studied population was 24.09 h; however, the average minimal time was slightly above 21 h, which suggests a high range for further improvement. The heritability of minimal and maximal times between ovipositions was almost 0.2 which is similar to the level estimated for initial egg production in the studied population. Temporal laying organization was also studied in quails (Houdelier et al., 2002). RESULTS AND DISCUSSION In the studied population 23% of the variation of number of clutches and average clutch length was attributed to genetic factors. However, these estimates are negligibly higher than obtained by Bednarczyk et al. (2000) via sire+dam model for this population. Heritability estimate for maximal clutch length was lower (0.11) than these found 457 ACKNOWLEDGEMENTS The authors would like to thank Dr. Jesus Arango for helpful comments on the draft version of the manuscript. CONCLUSIONS Generally, substancial amount of genetic vriation was found for egg formation and traditional selection traits. However, it should be recalled that both genetic and phenotypic influences on the performance traits are always estimated for the given population. Hence, generalization of the results for oviposition time and clutch size should be done carefully, especially considering small population size. On the other hand, the obtained results seem to be promising and stimulating for further studies on including the clutch and oviposition traits in genetic improvement programmes in laying hens. WOLC A. ETAL. As it was already mentioned the heritability estimate for mean oviposition time was relatively low (0.13), whereas the parameters for traits describing variability of time of egg formation were higher. Low heritability (h 2=0.09) of time interval was recently reported in Leghorn hens by Icken et al. (2008). However, Lillpers (1991) reported heritability estimates of time of oviposition ranging from 0.38 to 0.78. The genetic and phenotypic relationships between these traits and classical ones are basically favourable, resulting from the biological background of egg formation. For instance, when an increase in initial egg production is genetically associated with a decrease of average time of oviposition SELECTION TRAITS IN LAYING HENS 458 (rg= -0.37). Also a favourable genetic correlation was estimated between average egg weight and mean time of oviposition. The results suggest a possibility of effective selection for improvement of egg production traits; but, from economical perspective, other important characters should be included into the analysis. It is well known, that egg quality and reproductive traits are determined by a number of factors. An investigation carried out by Tumova et al. (2007) indicated a significant phenotypic relationship between oviposition time and egg quality. So far, Tumova and Ebeid (2005) reported that eggs laid in the morning had a slightly higher ratio of yolk than eggs laid in the afternoon. Effects of oviposition time on reproductive traits have been examined by Zakaria et al. (2009). Bednarczyk M., Lisowski M., Kielczewski K., Skornicki G., 1997. Construction of optimal selection indexes on basis electronic control of egg production in chicken (in Polish). Report Z-5/96 COBRD, Poznah (Poland) Chen C.F., Tixier-Boichard M., 2003b. Estimation of genetic variability and selection response for clutch length in dwarf brown-egg layers carrying or not the naked neck gene. Genet. Sel. Evol. 35, 533-537 Chen C.F., Tixier-Boichard M , 2003a. Correlated responses to long-term selection for clutch length in dwarf brown-egg layers carrying or not carrying the naked neck gene. Poultry Sci. 82, 709-720 Bednarczyk M., Kielczewski K., Szwaczkowski T., 2000. Genetic parameters of the traditional selection traits and some clutch traits In a commercial line of laying hens. Arch. Gefliigelk. 64, 129-133 Chen C.F., Tixier-Boichard M , 2003a. Correlated responses to long-term selection for clutch length in dwarf brown-egg layers carrying or not carrying the naked neck gene. Poultry Sci. 82, 709-720 Chen C.F., Tixier-Boichard M., 2003b. Estimation of genetic variability and selection response for WOLC A. ETAL. 459 Houdelier C, Guyomarch C, Luniineau S., 2002. Daily temporal organization of laying in Japanese quail: variability and heritability. Chronobiol. Int. 19, 377-392 Icken W., Cavero D., Schmutz M., Thurner S., Wendl G., Preisinger R., 2008. Analysis of the time interval within laying sequences in a Transponder nest. In: Proceedings of XXII World Poultry Congress. Brisbane (Australia) Lewis P.D., Backhouse D., Gous R.M., 2004. Photoperiod and oviposition time in broiler breeders. Brit. Poultry Sci. 45,561-564 Lillpers K., 1991. Genetic variation in the time of ovipostion in the laying hens. Brit. Poultry Sci. 32, 303-312 Lillpers K., 1993. Oviposition patterns and egg production in domestic hens. Dissertation. Swedish University of Agricultural Sciences, Department of Animal Breeding and Genetics, Uppsala Lillpers K., Wilhelmson M.W., 1993. Genetic and phenotypic parameters for oviposition pattern traits in three selection lines of laying hens. Brit. Poultry Sci. 34, 297-308 Luc K.M., Miyoshi S., Suzuki M., Mitsumoto T., 1996. Heritability estimates of some clutch traits in laying hens. Jpn. Poultry Sci. 33, 23-28 McClung M.R., Wang A.B.S., Jones W.T., 1976. Response to selection for time interval between oviposition time in the hen. Poultry Sci. 33, 23-28 Meyer K., 2001. DFREML ver. 3.1. University of New England (Australia) Miandmients P., Hiammal J., Rummiel K., 1993. Length of eggs' clutches and time intervals in white and brown layers (in Russian). Pticevodstvo 42, 14-15 Nurgiartiningsih V.M.A., Mielenz N., Preisinger R., Schmutz M., Schueler L., 2004. Estimation of genetic parameters based on individual and group mean records in laying hens. Brit. Poultry Sci. 45,604-610 Pyrzak P., Siopes T.D., 1989. Characteristics of oviposition patterns of turkey hens and the influence of different wavelenght of light. Brit. Poultry Sci. 31, 871-879 Rosinski A., Nowaczewski S., Kontecka H., Bednarczyk M., Elminowska-Wenda G., Bielinska H., Maczynska A., 2006. Analysis of the laying rhythm and reproductive traits of geese. Folia Biol. -Krakow 54, 145-152 Sheldon B.L., Yoo B.H., 1993. Egg genetics - future prospects. In: Proceedings of 10 th International Symposium on Current Problems of Avian Genetics. Nitra (Slovakia), pp. 49-54 Simmons G.S., Hetzel D.J.S., 1983. The relationships between oviposition, ovulation and egg formation in Khaki Campbell ducks. Brit. Poultry Sci. 24, 21-29 Szwaczkowski T., 2003. Use of mixed model methodology in poultry breeding: Estimation of genetic parameters. In: W.M. Muir, S.E. Aggrey (Editors). Poultry Genetics, Breeding and Biotechnology. CAB Int. 11, 165-201 gy p y g genetic parameters. In: W.M. Muir, S.E. Aggrey (Editors). REFERENCES Bednarczyk M., Kielczewski K., Szwaczkowski T., 2000. Genetic parameters of the traditional selection traits and some clutch traits In a commercial line of laying hens. Arch. Gefliigelk. 64, 129-133 Bednarczyk M., Lisowski M., Kielczewski K., Skornicki G., 1997. Construction of optimal selection indexes on basis electronic control of egg production in chicken (in Polish). Report Z-5/96 COBRD, Poznah (Poland) Chen C.F., Tixier-Boichard M , 2003a. Correlated responses to long-term selection for clutch length in dwarf brown-egg layers carrying or not carrying the naked neck gene. Poultry Sci. 82, 709-720 Chen C.F., Tixier-Boichard M., 2003b. Estimation of genetic variability and selection response for clutch length in dwarf brown-egg layers carrying or not the naked neck gene. Genet. Sel. Evol. 35, 533-537 WOLC A. ETAL. WOLC A. ETAL. Poultry Genetics, Breeding and Biotechnology. CAB Int. 11, 165-201 gy Szydlowski M., Szwaczkowski T., 2001. Bayesian segregation analysis of production traits in two strains of laying chicken. Poultry Sci. 80, 125-131 Tumova E., Ebeid T., 2005. Effect of time of oviposition on egg quality characteristics in cages and in a litter housing system. Czech J. Anim. Sci. 50, 129-134 Tumova E., Zita L., Hubeny M., Skrivan M., Ledvinka Z., 2007. The effect of oviposition time and genotype on egg quality characteristics in egg type hens. Czech J. Anim. Sci. 52, 26-30 Wezyk S., Szwaczkowski T., 1997. Application of mixed model methodology in breeding strategies for laying fowl. World Poultry Sci. J. 53, 325-336 Zakaria A.H., Plumstead P.W., Romero-Sanches H., Leksrisompong N., Brake J., 2009. The effects of oviposition time on egg weight loss during storage and incubation, fertility, and hatchability of broiler hatching eggs. Poultry Sci. 88, 2712-2717 Zieba G., Lukaszewicz M., Twardowska M., Witkowski A., 2003. Genetic trends of laying merit in maternal (M55) and paternal (V44) strains of hens. Anim. Sci. Pap. Rep. 21, 241-249
https://openalex.org/W4368372507	https://www.researchsquare.com/article/rs-2874508/latest.pdf	English	null	Improved Functionnectome by dissociating the contributions of white matter fiber classes to functional activation	Research Square (Research Square)	2,023	cc-by	9,043	Improved Functionnectome by dissociating the contributions of white matter fiber classes to functional activation Victor Nozais GIN-IMN, UMR5293 CNRS, Université de Bordeaux Guillaume Theaud Sherbrooke Connectivity Imaging Lab, Université de Sherbrooke Maxime Descoteaux Sherbrooke Connectivity Imaging Lab, Université de Sherbrooke Michel Thiebaut Schotten GIN-IMN, UMR5293 CNRS, Université de Bordeaux Laurent Petit (  laurent.petit@u-bordeaux.fr ) GIN-IMN, UMR5293 CNRS, Université de Bordeaux Research Article Keywords: task-related fMRI, diffusion MRI, tractography, human brain Posted Date: May 4th, 2023 DOI: https://doi.org/10.21203/rs.3.rs-2874508/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Research Article License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Page 1/24 Page 1/24 Page 1/24 Abstract Integrating the underlying brain circuit's structural and functional architecture is required to explore the functional organization of cognitive networks properly. In that regard, we recently introduced the Functionnectome. This structural-functional method combines an fMRI acquisition with tractography- derived white matter connectivity data to map cognitive processes onto the white matter. However, this multimodal integration faces three significant challenges: 1) the definition of the interface between grey and white matter, as the tractography streamlines may fail to reach the grey matter properly; 2) the scrambling effect of crossing fibers on functional signal, as a single voxel in such regions can be structurally connected to several cognitive networks with heterogeneous functional signals; and 3) the difficulty of interpretation of the resulting cognitive maps, as crossing and overlapping white matter tracts can obscure the organization of the studied network. In the present study, we tackled these problems by developing a streamline-extension procedure and dividing the white matter anatomical priors between association, commissural, and projection fibers. This approach significantly improved the characterization of the white matter involvement in the studied cognitive processes. The new Functionnectome priors produced are now readily available, and the analysis workflow highlighted here should also be generalizable to other structural-functional approaches. Introduction Magnetic resonance imaging (MRI) is one of the best tools to study the human brain in vivo (Bandettini 2012; Lerch et al. 2017). It offers a wealth of modalities, leveraging different properties of matter to infer and contrast brain tissues' state and inform us of their structural and functional characteristics. Notably, the blood-oxygen-level-dependent (BOLD) signal of functional MRI (fMRI) (Ogawa et al. 1990) has shown tremendous success in mapping cognitive areas and networks in the grey matter (Yarkoni et al. 2011). Analysis of fMRI data traditionally involves leveraging the link between the BOLD signal and hemodynamic response to physiological processes. It reveals brain areas in which the BOLD signal displays a specific pattern related to cognition or other functional organization of the brain. Likewise, diffusion MRI (dMRI) data, in tandem with tractography algorithms (Conturo et al. 1999; Jeurissen et al. 2019), have been harnessed to tackle the challenge of charting the white matter tracts defining the brain’s structural connectivity. Diffusion MRI yields information about the local directions of water diffusion across the brain (Le Bihan et al. 1986). Because water tends to diffuse in the same direction as axonal fibers in the white matter, tractography uses dMRI data to try and reconstruct fiber bundles – or “streamlines” – that can be employed to estimate structural connectivity in the brain (Jeurissen et al. 2019). Over the past decade, the exploration of the brain’s function using MRI has shifted from unimodal fMRI to the combination of functional and structural information (Douaud et al. 2011; Hermundstad et al. 2013). One domain that has particularly benefited from this is the functional study of white matter in the brain. While there is emerging evidence that the white matter holds meaningful BOLD signals (Gore et al. 2019; Li et al. 2019), it is too weak in comparison to the grey matter. Yet, the white matter plays a major role in Page 2/24 Page 2/24 Page 2/24 cognition: it is the structural foundation of the functional networks classically observed in grey matter (for a review see (Thiebaut de Schotten and Forkel 2022)). Assessing the functional involvement of the white matter in cognitive processes is thus crucial to better understand the brain and to shed light on how diseases and damage to the white matter affect it. Traditionally, this aspect has mainly been investigated through lesion-symptom mapping, correlating cognitive deficit with the location of brain lesions. Introduction In this framework, and more generally in the study of brain connectivity, the white matter fibers are classified among three groups: association, commissural, and projection fibers (Catani and Thiebaut de Schotten 2012). The association fibers correspond to the axons connecting different parts of the cortex in the same hemisphere; the commissural fibers correspond to the axons connecting the cortex of the two hemispheres (crossing the interhemispheric fissure); and projection fibers correspond to axons connecting the cortex with subcortical structures. This categorization of the brain connections reflects the functional organization of the cognitive networks they support and is very useful in the description and clinical study of these networks (Aralasmak et al. 2006; Catani and Thiebaut de Schotten 2012; Hasan et al. 2010). Yet, relying solely on the clinical approach to study the function of the white matter greatly limits our insights into the subject as symptoms cannot reflect all the brain's cognitive processes. By undertaking a multimodal approach and complementing the grey matter functional information with structural connectivity derived from tractography, it is possible to map cognitive processes onto the white matter of the healthy human brain. In this regard, we recently introduced the Functionnectome (Nozais et al. 2021), a structural-functional method – and an open-source program – that combines an fMRI acquisition with whole-brain, tractography-derived, white matter connectivity data. It has been successfully applied to task fMRI (t-fMRI) – mapping white matter domains linked to specific cognitive tasks – and to resting-state fMRI (rs-fMRI) – mapping the white matter and grey matter joint contribution to resting-state networks (Nozais et al. 2023). However, integrating fMRI and tractography data is not trivial, and multiple factors may impact the results and their interpretability. Among them, three potential pitfalls stand out: 1) The proper definition of the interface between grey and white matter is crucial to effectively combine the grey matter functional signal with the related white matter connectivity. The resulting structural-functional relationship between the two could be underestimated if the streamlines do not fully reach the grey matter. 2) The structural-functional integration in white matter areas with crossing fibers can be problematic. In these regions, a single voxel can be structurally connected to several functional networks with widely different cognitive roles, effectively reducing its apparent involvement with any specific cognitive process (i.e., locally lowering the signal-to-noise ratio). Introduction 3) Probably the most important point lies in the interpretability of the brain maps derived from structural-functional integration. Indeed, using whole-brain tractograms to estimate the structural connectivity necessarily leads to maps involving the whole studied networks without distinctions between which fiber tract is being activated. This is a problem when these tracts are close or even overlapping, as it would lead to ambiguity regarding the architecture of the studied functional network. Page 3/24 Page 3/24 In the present study, we explored the effect of these three points on structural-functional analysis using the Functionnectome. More specifically, we first implemented and tested a procedure to allow streamlines to reach deeper into grey matter to help improve the interface between grey and white matter. Second, we adopted the framework mentioned above, categorizing the connectivity data by fiber class (association, commissural, and projection fibers). By filtering the tractograms using these categories, we could tackle the two remaining issues together: it reduced the occurrence of fiber-crossing areas and allowed the examination of the cognitive networks across their different components (and fiber tracts) separately. The results significantly improved the Functionnectome priors and the analysis workflow. They also offer some general advice for the field of structural-functional analysis. Dataset, acquisition parameters, and preprocessing The dataset used in this study comes from the young-adult cohort of Human Connectome Project (HCP) (Van Essen et al. 2013). It consists of 3T MRI scans with anatomical, diffusion (dMRI), and functional (fMRI) MRI acquisitions. The anatomical and diffusion scans are from a subset of 100 randomly selected participants and were used to generate normative white matter priors (containing structural connectivity information) used in the Functionnectome analyses. The functional scans are from a subset of 46 participants, the same participants as the one used in the original Functionnectome paper (Nozais et al. 2021) and independent from the 100 participants used for the priors. The HCP data was acquired by the WU-Minn Consortium with IRB approval, all participants gave their informed consent, and the WU-Minn HCP Consortium Open Access Data Use Terms were respected in the study. The detailed acquisition parameters for each scan can be found on the HCP website (https://www.humanconnectome.org/hcp-protocols-ya-3t-imaging) and were discussed in an HCP article (Ugurbil et al. 2013). We briefly summarize them here. Anatomical scans T1-weighted (T1w) acquisitions (3D MPRAGE), with TR of 2400 ms, TE of 2.14 ms, T1 of 1 ms, flip angle of 8°, 0.7 x 0.7 x 0.7 mm3 isotropic voxels, and a bandwidth of 210 Hz/Px. Methods The processing and analysis workflow is summarized in Fig. 1 below. Functional scans Gradient-echo EPI acquisitions, with TR of 720 ms, TE of 31.1 ms, flip angle of 52°, 2 x 2 x 2 mm3 isotropic voxels, Echo spacing of 0.58 ms, and bandwidth of 2290 Hz/Px. Gradient-echo EPI acquisitions, with TR of 720 ms, TE of 31.1 ms, flip angle of 52°, 2 x 2 x 2 mm3 isotropic voxels, Echo spacing of 0.58 ms, and bandwidth of 2290 Hz/Px. All the data were downloaded from the HCP website in their preprocessed form. This preprocessing was done using the HCP’s Minimal Preprocessing Pipelines (MPP) (Glasser et al. 2013). Briefly, the structural preprocessing consisted of linear and non-linear registration to a standard MNI space, and a brain segmentation using FreeSurfer (Reuter et al. 2010); the dMRI preprocessing consisted of b0 intensity normalization, correction of EPI distortions, correction of Eddy currents and motion effects, correction of gradient non-linearity, and a registration to the T1w image in the participant’s space; the fMRI preprocessing consisted of gradient distortion correction, motion correction, EPI distortion correction, and registration to MNI space. Tractography The tractography was performed using a probabilistic tractography approach. We used TractoFlow (Theaud et al. 2020), an automated dMRI processing pipeline for probabilistic tractography. TractoFlow was run with its two tracking algorithms: a classical local tracking and the Particle Filter Tracking (PFT) (Girard et al. 2014). The streamlines generated by the two probabilistic tracking algorithms were combined to form each whole-brain probabilistic tractogram. Most of the settings used for the tracking were the default settings proposed by TractoFlow: the shape factor for the elongated symmetric diffusion tensor of the FRF was 1.5 x 10− 3 mm2.s− 1; spherical harmonics order was 8; all the shells were used (b = 1000, 2000, and 3000) for the fODF; step-size was 0.5 mm; and maximal angle between steps was 20°. The seeding was done on the whole white matter, with 10 seeds per voxel for the PFT and 5 seeds per voxel for the local tracking. Each tractogram was then registered to the MNI space by applying the linear and nonlinear transformations used to register the anatomical scan. Diffusion scans Spin-echo EPI (Echo-planar imaging) acquisitions, with TR of 5520 ms, TE of 89.5 ms, flip angle of 78°, 1.25 x 1.25 x 1.25 mm3 isotropic voxels, a Multiband factor of 3, echo spacing of 0.78 ms, bandwidth of 1488 Hz/Px, and three shells of b-values of 1000, 2000, and 3000 s/mm2. In total, 90 diffusion directions were acquired per shell and 18 b = 0 images for a total of 288 diffusion-weighted images. Page 4/24 Page 4/24 Functional scans Gradient-echo EPI acquisitions, with TR of 720 ms, TE of 31.1 ms, flip angle of 52°, 2 x 2 x 2 mm3 isotropic voxels, Echo spacing of 0.58 ms, and bandwidth of 2290 Hz/Px. Streamline filtering and categorization When combining functional information directly on the white matter using connectivity information from whole-brain tractography, as the Functionnectome does, two potential issues arise: First, the resulting activation maps do not differentiate between fiber tracts, which limits our ability to interpret the data and precisely identify the tracts involved in the circuit. Second, whole-brain tractograms present crossing- fibers regions (i.e., regions of the brains where multiple white matter tracts, part of different functional networks, intersect). Using structural-functional combination methods in these areas can mix unrelated functional signals, artificially lowering the local signal-to-noise ratio and thus the ability to precisely estimate the involvement of the white matter in specific cognitive functions. When combining functional information directly on the white matter us whole-brain tractography, as the Functionnectome does, two potential i activation maps do not differentiate between fiber tracts, which limits o precisely identify the tracts involved in the circuit. Second, whole-brain fibers regions (i.e., regions of the brains where multiple white matter tra networks, intersect). Using structural-functional combination methods functional signals, artificially lowering the local signal-to-noise ratio an estimate the involvement of the white matter in specific cognitive funct To circumvent these problems, each tractogram was split into three categories of streamlines: association, commissural, and projection streamlines (Catani and Thiebaut de Schotten 2012). To further reduce the occurrence of crossing fibers, association streamlines were also split into three length categories: “short fibers” with streamlines below 40 mm in length, “medium fibers” with streamlines between 40 and 80 mm, and “long fibers” with streamlines longer than 80 mm (Schüz and Braitenberg 2002; Shastin et al. 2022). This filtering process was performed using ExTractor_Flow (Petit et al. 2023), a filtering method that first automatically classifies the streamlines of a whole brain tractogram as being either anatomically plausible or implausible. It is a rule-based automatic pipeline filtering out false-positive streamlines by following brain neuroanatomy organizational principles (Meynert 1885; Dejerine and Dejerine-Klumpke 1901; Ludwig and Klingler 1956; Crosby et al. 1962; Schüz and Braitenberg 2002; Schmahmann and Pandya 2006; Nieuwenhuys et al. 2008), that can be summarized as follow. First, every cortical area is linked with other cortical and subcortical regions by pathways grouped into three distinct categories: association, commissural, and projection fibers. Streamline extension When combining volumetric grey matter data with tractography, a significant risk lies in the streamlines not properly reaching the grey matter voxels. More specifically, the stopping criteria of the tractography algorithm (i.e., the criteria defining when and where a streamline should stop) may not allow the streamlines to reach deep enough into the grey matter. Note that we are not discussing the challenge of properly connecting the whole grey matter using tractography. This topic and the related gyral bias (Rheault et al. 2020; St-Onge et al. 2018) are mentioned in the discussion but they go beyond the current proposed improvements. In the present study, we focus on the termination location of streamlines and their relation to structural-functional coupling. Indeed, in methods combining grey matter functional and tractography (like the Functionnectome), this coupling could be underestimated if the streamlines stop before properly reaching the grey matter voxel they should be related to. To test this effect, we devised a simple post-hoc lengthening procedure that was applied on each tractogram, effectively allowing the streamlines to fully reach the grey matter if it was close to its termination points. The extension procedure Page 5/24 Page 5/24 is as follows: all streamlines are linearly extended by 5 mm at both ends in the same direction as the streamline’s ending segments; if the extension doesn’t reach the grey matter, the extension is canceled; if the extension reaches the grey matter but changes area (as per the FreeSurfer grey matter parcellation) or gets out of the grey matter, the extended part is cut, keeping only the part in the first grey matter area (Fig. 2). This last point reduces the risk that the procedure extends the streamline to a different functional domain. Such a procedure is similar to tractography toolboxes when building connectomes, like the tck2connectome function from MRtrix3 (Smith et al. 2015; Tournier et al. 2019). However, the current method differs from these approaches as they usually only allow the association of a streamline to a whole grey matter region, while our method was specifically tailored to our needs by associating the streamlines to the grey matter voxels they reach. It was necessary for testing the effect of such extension on Functionnectome analyses. Streamline filtering and categorization Second, the commissural, projection, and long-range association fibers travel within the central WM part of the core of a gyrus, namely the gyral stem, and therefore never from the side of a gyrus by crossing a sulcus. Third, the shorter U-shaped association fibers connect adjacent gyri by running in a thin band immediately beneath the GM, constituting the most Page 6/24 Page 6/24 external part of the gyral stem. Following these anatomical principles, regions of interest (ROIs) from the Johns Hopkins University (JHU) template (Oishi et al. 2009) are used to define ROI-based sequential filtering conditions. ExTractorflow starts, therefore, by considering a streamline as anatomically- implausible either when shorter than 20 mm, making a 360° loop, terminating along the ventricles, or even being broken within the deep WM structures as defined in (Oishi et al. 2009), e.g., the streamlines ending within the corpus callosum, corona radiata, internal capsule, external capsule, parts of association bundles, etc. Interestingly, the JHU template includes both cortical grey matter and underlying superficial white matter parts for each gyrus. By manipulating these different ROIs, it is possible to create additional ROIs that optimize the filtering of anatomically plausible fibers while applying the principles of WM fiber organization emanating from a given gyrus. Two additional ROIs were thus used for each gyrus and each hemisphere. The first one corresponds to the outermost part of the gyrus, namely its GM shell. The second corresponds to the "entrance/exit door" of the gyrus, namely its WM stem, which was manually drawn between the shell boundaries corresponding to the fundus of the sulci delimiting the gyrus. An essential step in ExTractorflow is to consider as plausible fibers only those ending in a gyrus while having passed through the stem and not crossing the shell. A second stage of anatomical Boolean rules uses the JHU template ROIs to keep anatomically plausible streamlines strictly. Finally, the different anatomical rules used in ExTractorflow were also used in the context of the current study to split the remaining plausible streamlines into the association, commissural, and projection streamlines (Fig. 3). Functionnectome, white matter priors, and activation maps To evaluate the effect of splitting the tractograms by fiber class on a structural-functional analysis, we used each set of tractograms in Functionnectome analyses (Nozais et al. 2021). The whole procedure is described in the original paper, but we summarized it below. First, the white matter priors are computed using the tractograms (n = 100). These priors correspond to maps, one per brain voxel, giving the probability of structural connectivity from this voxel to the other brain voxels. These priors can then be used by the Functionnectome program (available at http://www.bcblab.com) to project grey matter BOLD signals onto the related white matter. Essentially, for a given white matter voxel, the Functionnectome uses the structural connectivity information from the white matter priors to do a weighted average of the BOLD time-series from the grey matter voxels connected to the said white matter voxel. This results in a new functional volume (with 4 dimensions), with functional time-series on the white matter voxels, that can then be statistically analyzed in the same way as a classical fMRI volume. Here, we used the Functionnectome program to project the signal of two task-fMRI datasets onto the white matter (finger- tapping dataset, n = 46; working-memory dataset, n = 45). The functionnectome volumes of each participant were then statistically analyzed by generalized linear modeling (GLM) to reveal the corresponding activation maps. Both first- and second-level analyses were done using FEAT (FMRI Expert Analysis Tool, v6.00), from the FMRIB Software Library (FSL), yielding a group-level activation z-map per task. Page 7/24 In the present study, we created one set of white matter priors per set of tractograms, for a total of eight sets of priors, namely: two whole-brain sets, extended and non-extended; four association extended sets (all fibers, and separately short, medium, and long association extended fibers); one commissural extended set; and one projection extended set. Together, the association, commissural, and projection priors are referred to as the split priors, in opposition to the whole-brain priors. Note that all the split priors we generated with tractograms that underwent the streamline-extension procedure. The functional analyses described above (for the two functional datasets) were repeated for each of these 8 sets of priors. Functionnectome, white matter priors, and activation maps In the present study, we created one set of white matter priors per s sets of priors, namely: two whole-brain sets, extended and non-exte (all fibers, and separately short, medium, and long association exte extended set; and one projection extended set. Together, the assoc priors are referred to as the split priors, in opposition to the whole-b we generated with tractograms that underwent the streamline-exte analyses described above (for the two functional datasets) were re priors. Maps analyses and statistics All the activation z-maps were analyzed and displayed after applying a white matter mask to keep only white matter voxels, and using an arbitrarily high threshold of z = 3 (equivalent to p < 0.0013, uncorrected). We kept this uncorrected threshold to make the displayed map clearer and easier to read, as applying a false discovery rate correction did not change the significance of the results with the chosen threshold. The white matter mask was defined as the voxels attributed to white matter in at least 10% of the study’s participants. All the presented maps are z-maps, displaying voxelwise z-scores for the task activation being studied. The slices shown in the results were selected to display the main white matter functional pathways revealed by the Functionnectome activation mapping. The unthresholded z-maps are freely available on Neurovault (https://identifiers.org/neurovault.collection:13538). All the activation z-maps were analyzed and displayed after applying a white matter mask to keep only white matter voxels, and using an arbitrarily high threshold of z = 3 (equivalent to p < 0.0013, To compare the results obtained from using whole-brain priors and those obtained using split priors, the corresponding z-maps (association, commissural, and projection) were combined, keeping, for each voxel, the maximum z-score across the three maps. These maps are referred to as the z-max maps. Similarly, the activation z-maps generated using the three association priors split by length (short, medium, and long) were recombined in the same manner to create association z-max maps. Note that the whole-brain z-max map mentioned above uses these association z-max maps instead of the basic association z-map. The statistical comparison between the maps was done on the voxels with z-scores above 3 in at least one of the maps, to only compare regions that are considered activated. The p-values displayed are the results from the two-sided paired t-test on the selected voxels. Enhancement of the activity detection in the white matter We first explored the effect of extending the WM streamlines in the GM on the structural-functional analysis. We therefore analyzed the involvement of the white matter in the right-hand finger-tapping task by contrasting the activation maps from the Functionnectome analysis using the priors generated with and without streamline extension. The resulting maps (Fig. 4A-B) showed that using the extension method significantly improved the statistics of the activated regions (i.e., with z > 3), resulting in an Page 8/24 Page 8/24 average increase of 12% (p < 0.0001) of the z-values across the map. The analysis was repeated with the working-memory dataset (Fig. 5A-B), which also demonstrated a significant increase of 4% (p < 0.0001) in z-values across the map when using the extended tractograms. average increase of 12% (p < 0.0001) of the z-values across the map. The analysis was repeated with the working-memory dataset (Fig. 5A-B), which also demonstrated a significant increase of 4% (p < 0.0001) in z-values across the map when using the extended tractograms. We interpreted these increases as proof that using our simple method to extend the streamlines improved the structural-function coupling in the analysis. As a result, we used the extension method in all the following analyses. While the increase in z-values from the extension procedure is relatively modest, the improvement from splitting the white matter priors and recombining the activation maps are quite striking. First, compared to the z-maps from the extended priors, we note an 83% increase (p < 0.0001) in the mean z-values of the z- max map for the finger-tapping task (Fig. 4B-C), and a 31% increase (p < 0.0001) for the working memory task (Fig. 5B-C). The z-max map from the finger-tapping task also shows a clear recovery of the signal in areas of crossing fibers, very distinct from the disappearance of the “gaps” in the activation map. This is especially obvious around the centrum semiovale, where the superior longitudinal fasciculus (SLF, not involved in the task) intersects with callosal fibers and the cortico-spinal tract (both involved in the task). Likewise, when focusing on the split priors for the association fibers, splitting them again by length category allowed reducing further the effect of the remaining crossing regions. In the z-maps from the finger-tapping task (Fig. 6), using the z-max approach with the three split association priors increases the z-values by 103% on average (p < 0.0001). Enhancement of the activity detection in the white matter For the working-memory task, we observe a 34% increase on average (p < 0.0001) with the z-max approach. Also, similarly to what is shown above with the whole- brain z-max map on the finger-tapping task, the association z-max map displays a recovery of the activation in fiber-crossing regions, filling the visible gap where the SLF intersects with shorter association fibers. Clarifying the structure of the networks Using the split priors also allows us to differentiate between the functional integration mediated by different classes of fibers and, thus, different tracts. In the case of the finger-tapping task (Fig. 7), using the split priors helps differentiate the functional involvement of multiple fiber tracts, both near grey matter hubs and in the deep white matter. For example, we can see that while most of the fibers involved in activations around the thalamic nuclei are from the projection category, some of this activity is also related to association fibers connecting the insula with the motor cortex. Similarly, this separation helps determine that the strong activation observed near the motor cortex from the left hemisphere is mostly related to association fibers. Likewise, in the activation maps from the working-memory task (Fig. 8), we can better differentiate between the strong activations related to either the association or commissural fibers. Most interestingly, the split also reveals the involvement of the projection fibers, which is not obvious when using whole- brain priors. Page 9/24 Page 9/24 Similarly, to the split between fiber classes, splitting the association fibers by length category reveals the different scales in connectivity that govern the functional organization of the cognitive networks. In the case of the finger-tapping task (Fig. 9A-B-C), we observe that most of the strong activation near the motor cortex is related to short and medium-length (U-shaped) fibers. Also, the connectivity of the insula to the other parts of the network is mostly mediated by medium-length fibers, but also in part by long association fibers. As for the working-memory task (Fig. 9D-E-F), the split reveals the strong activations related to small and medium-length fibers, especially in the frontal lobe, likely driving the local integration of functional information. It also unveils the exact connectivity profile specifically related to the long-distance association in the network, with both fronto-parietal and parieto-temporal connectivity. Discussion We explored two approaches to significantly enhance structural-functional analysis to study the function of white matter and improved the Functionnectome framework accordingly. Our results highlight three main factors that influence the analyses and their interpretation: elongating the streamlines at the interface between grey matter voxels and white matter tracts improves the structural-functional integration; splitting the fibers by category upstream of the analysis, then recombining the resulting maps vastly increase the ability to detect the involvement of the white matter in cognition; and the same split also strikingly clarify the structural-functional architecture of the studied networks. Moreover, these improvements should be equally applicable to other structural-functional methods (besides the Functionnectome), which would give our approach a global impact on the domain. Structural-functional analyses in neuroimaging, combining cortical functional information and tractography data, have been successfully used to explore the structure-function relationship between the brain’s grey and white matter. However, the definition of the interface between the two is not a trivial step, and two notable approaches have been used. One consists in using functionally defined ROI as a seed for the tractography (Javad et al. 2014), while the other consists in doing whole brain tractography and then selecting the streamlines terminating in the studied grey matter areas (Wang et al. 2015). The method defining the termination point of the streamlines also plays a major role in shaping the interface. Here, three main methods exist: a streamline can stop when it reaches the edge of a predefined white matter mask (Theaud et al. 2020); it can stop when it reaches a voxel associated with specific diffusion metrics (e.g., fractional anisotropy lower than a given threshold) (Thiebaut de Schotten et al. 2011), or can be guided to the cortical surface by a geometrical model when approaching the grey matter (St-Onge et al. 2018; St-Onge et al. 2021). In each case, there is no guarantee that the streamline will actually fully reach the grey matter it should be linked to. Even in the case when streamlines are guided to the cortical surface, which will improve the estimation of cortical connectivity (St-Onge et al. 2021), the streamlines may not reach all the functionally related grey matter voxels. This is especially true for voxels on the outer side of the cortex or deep inside subcortical nuclei. Discussion Page 10/24 Page 10/24 Page 10/24 In our study, we showed that a simple elongation of the streamline (constricted by grey matter boundaries) is enough to correct this bias and to significantly improve the results of a structural- functional analysis. While the increase in the significance of the activations was somewhat modest, it should be noted that the tractography used in the present study was very “aggressive”, attempting to maximize the estimation of the brain connectivity by generating a vast number of streamlines. It is entirely possible that the streamline-extension procedure would have a much stronger effect on less dense tractograms. This method has thus the potential to become a standard step applied to tractograms when used in this framework. It should be noted, however, that for the same region the connectivity changes according to the layer of the cortex studied (Pandya et al. 2015). While such estimation is beyond the resolution available nowadays, future endeavors based on higher resolution diffusion weighted imaging datasets should adapt streamline elongation to the anatomical principles of the cortical architecture. Additionally, to ensure a proper streamline elongation, the orientation of the terminal segments of a streamline must be correct. This may not always be the case, as can be illustrated by the existence of the gyral bias (Schilling et al. 2018). Ideally, this elongation step should be part of the tractography itself and coupled with a method that ensures the proper orientation of the streamlines (e.g., (St-Onge et al. 2018; St-Onge et al. 2021) with surface-enhanced tractography) and not done post-hoc. The other avenue we explored to improve structural-functional analyses was to split the analysis by fiber class, independently studying association, commissural and projection fibers. Depending on the method used to perform the structural and functional integration, a detection bias may arise from fibers crossing and mixing unrelated functional signals together. Splitting the analysis by fiber category and reassembling the resulting maps can correct this bias. In the case of the Functionnectome analysis, which directly combines the functional information from grey matter sources onto white matter voxels, we showed that this correction effect is present and very strong. Here, splitting the analysis enabled us to both fill the gap in the activation maps located in regions of crossing fibers, and to increase the significance of the studied activation globally. Discussion These two points are very beneficial from a technical standpoint and a clear improvement in our ability to map the involvement of white matter in cognitive processes fully. Nevertheless, the aspect that may be the most important when splitting the analysis by fiber category is the ability it gives us to disentangle different fibers' roles in a white matter activation. Indeed, each set of split priors, whether by fiber class or fiber length, revealed details about the organization of the studied cognitive circuits, which would have been obscured in the analyses using whole-brain priors. For example, the working memory task revealed projection fibers connecting the frontal lobe with thalamic nuclei. This activity was effectively hidden by the cortico-cortical parts of the circuit in the whole-brain analysis, and is extremely interesting, especially in the light of the mounting evidence of the involvement of the thalamus in working memory (Piras et al. 2010; Roy et al. 2022; Watanabe and Funahashi 2012). We expect the improvement derived from splitting the tractograms to be generalizable to other volumetric structural-functional methods that integrate information directly on white matter voxels, such as track- weighted imaging (Calamante 2017) or white-matter interpolation of fMRI (Tarun et al. 2020). However, fiber crossing can still happen in split tractograms, especially with association fibers (e.g., the SLF Page 11/24 Page 11/24 crossing the frontal aslant tract). To further improve this aspect, one conceivable approach would be to use an approach akin to using fixels (Dhollander et al. 2021) instead of voxels in the analysis. With a type of volumetric data that keeps the information about the direction of the fibers from which the signal was extracted, it should help separate unrelated signals when integrating them onto the white matter, which in turn would remove the need to split the tractograms. However, this type of analysis would require a complete remake of the current methods, is more complicated to implement, would likely be much more computationally intensive, and its results would definitely be harder to interpret. The seven sets of extended priors (whole-brain, association, commissural, projection, and short-/medium-/long-association fibers), are all freely available online and are now part of the updated Functionnectome. These new priors, and especially the split priors, will help disentangle the involvement of different fiber tract populations in cognitive functions. Conclusion We showed that structural-functional analyses combining fMRI and tractography can be dramatically improved through two relatively easy-to-implement steps: By ensuring a proper reach of each streamline in the grey matter and by grouping streamlines into classes that minimize streamline crossings. In the present study, we used the Functionnectome as an example of structural-functional analysis, but we believe our results are generalizable to many other similar methods and could thus help improve results obtained in this field. Discussion The priors can be easily selected and downloaded from the Functionnectome interface, then the analysis has to be run separately for each set of priors. Acknowledgements: This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 818521). Data were provided by the Human Connectome Project, WU-Minn Con- sortium (Principal Investigators: David Van Essen and Kamil Ugurbil; 1U54MH091657) funded by the 16 NIH Institutes and Centers that support the NIH Blueprint for Neuroscience Research and by the McDonnell Center for Systems Neuroscience at Washington University. Victor Nozais received financial support for this project from the “Consulat Général de France à Québec” and the “Conseil Franco-Québécois de Coopération Universitaire” as part of the Samuel de Champlain Program for the Development of Strategic Partnerships in Education and Research. References Page 12/24 Page 12/24 Page 12/24 1. Aralasmak A, Ulmer JL, Kocak M, Salvan CV, Hillis AE, Yousem DM (2006) Association, commissural, and projection pathways and their functional deficit reported in literature. J Comput Assist Tomogr 30 (5):695–715. doi:10.1097/01.rct.0000226397.43235.8b 2. Bandettini PA (2012) Twenty years of functional MRI: the science and the stories. Neuroimage 62 (2):575–588. doi:10.1016/j.neuroimage.2012.04.026 3. Calamante F (2017) Track-weighted imaging methods: extracting information from a streamlines tractogram. MAGMA 30 (4):317–335. doi:10.1007/s10334-017-0608-1 4. Catani M, Thiebaut de Schotten M (2012) Atlas of Human Brain Connections. Oxford University Press, 5. Conturo TE, Lori NF, Cull TS, Akbudak E, Snyder AZ, Shimony JS, McKinstry RC, Burton H, Raichle ME (1999) Tracking neuronal fiber pathways in the living human brain. Proc Natl Acad Sci U S A 96 (18):10422–10427. doi:10.1073/pnas.96.18.10422 6. Crosby E, Humphrey T, Lauer EW (1962) Correlative anatomy of the nervous system. The MacMillan Company, New York 7. Dejerine J, Dejerine-Klumpke A (1901) Anatomie des centres nerveux. To 8. Dhollander T, Clemente A, Singh M, Boonstra F, Civier O, Duque JD, Egorova N, Enticott P, Fuelscher I, Gajamange S, Genc S, Gottlieb E, Hyde C, Imms P, Kelly C, Kirkovski M, Kolbe S, Liang X, Malhotra A, Mito R, Poudel G, Silk TJ, Vaughan DN, Zanin J, Raffelt D, Caeyenberghs K (2021) Fixel-based Analysis of Diffusion MRI: Methods, Applications, Challenges and Opportunities. Neuroimage 241:118417. doi:10.1016/j.neuroimage.2021.118417 9. Douaud G, Filippini N, Knight S, Talbot K, Turner MR (2011) Integration of structural and functional magnetic resonance imaging in amyotrophic lateral sclerosis. Brain 134 (Pt 12):3470–3479. doi:10.1093/brain/awr279 10. Girard G, Whittingstall K, Deriche R, Descoteaux M (2014) Towars quantitative connectivity analysis: Reducing tractography biaises. NeuroImage 98:266–278 11. Glasser MF, Sotiropoulos SN, Wilson JA, Coalson TS, Fischl B, Andersson JL, Xu J, Jbabdi S, Webster M, Polimeni JR, Van Essen DC, Jenkinson M, Consortium WU-MH (2013) The minimal preprocessing pipelines for the Human Connectome Project. Neuroimage 80:105–124. doi:10.1016/j.neuroimage.2013.04.127 12. Gore JC, Li M, Gao Y, Wu TL, Schilling KG, Huang Y, Mishra A, Newton AT, Rogers BP, Chen LM, Anderson AW, Ding Z (2019) Functional MRI and resting state connectivity in white matter - a mini- review. Magn Reson Imaging 63:1–11. doi:10.1016/j.mri.2019.07.017 13. Hasan KM, Kamali A, Abid H, Kramer LA, Fletcher JM, Ewing-Cobbs L (2010) Quantification of the spatiotemporal microstructural organization of the human brain association, projection and commissural pathways across the lifespan using diffusion tensor tractography. Brain Struct Funct 214 (4):361–373. References doi:10.1007/s00429-009-0238-0 14. Hermundstad AM, Bassett DS, Brown KS, Aminoff EM, Clewett D, Freeman S, Frithsen A, Johnson A, Tipper CM, Miller MB, Grafton ST, Carlson JM (2013) Structural foundations of resting-state and 14. Hermundstad AM, Bassett DS, Brown KS, Aminoff EM, Clewett D, Freeman S, Frithsen A, Johnson A, Tipper CM, Miller MB, Grafton ST, Carlson JM (2013) Structural foundations of resting-state and Page 13/24 Page 13/24 task-based functional connectivity in the human brain. Proc Natl Acad Sci U S A 110 (15):6169– 6174. doi:10.1073/pnas.1219562110 task-based functional connectivity in the human brain. Proc Natl Acad Sci U S A 110 (15):6169– 6174. doi:10.1073/pnas.1219562110 15. Javad F, Warren JD, Micallef C, Thornton JS, Golay X, Yousry T, Mancini L (2014) Auditory tracts identified with combined fMRI and diffusion tractography. Neuroimage 84:562–574. doi:10.1016/j.neuroimage.2013.09.007 16. Jeurissen B, Descoteaux M, Mori S, Leemans A (2019) Diffusion MRI fiber tractography of the brain. NMR Biomed 32 (4):e3785. doi:10.1002/nbm.3785 17. Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M (1986) MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 161 (2):401–407. doi:10.1148/radiology.161.2.3763909 18. Lerch JP, van der Kouwe AJ, Raznahan A, Paus T, Johansen-Berg H, Miller KL, Smith SM, Fischl B, Sotiropoulos SN (2017) Studying neuroanatomy using MRI. Nat Neurosci 20 (3):314–326. doi:10.1038/nn.4501 19. Li M, Newton AT, Anderson AW, Ding Z, Gore JC (2019) Characterization of the hemodynamic response function in white matter tracts for event-related fMRI. Nat Commun 10 (1):1140. doi:10.1038/s41467-019-09076-2 20. Ludwig E, Klingler J (1956) Atlas cerebri humani. S. Karger, Basel 20. Ludwig E, Klingler J (1956) Atlas cerebri humani. S. Karger, Basel 21. Meynert T (1885) Psychiatry: Clinical Treatise on the Diseases of the Fore-Brain, trans. B. Sachs. GP Putnam, New York & London 22. Nieuwenhuys R, Voogd J, van Huijzen C (2008) The human central nervous system. 4th edn. Springer-Verlag, Berlin 23. Nozais V, Forkel SJ, Foulon C, Petit L, Thiebaut de Schotten M (2021) Functionnectome as a framework to analyse the contribution of brain circuits to fMRI. Commun Biol 4 (1):1035. doi:10.1038/s42003-021-02530-2 24. Nozais V, Forkel SJ, Petit L, Talozzi L, Corbetta M, Thiebaut de Schotten M, Joliot M (2023) Atlasing white matter and grey matter joint contributions to resting-state networks in the human brain. bioRxiv:2022.2001.2010.475690. doi:10.1101/2022.01.10.475690 25. Ogawa S, Lee TM, Kay AR, Tank DW (1990) Brain magnetic resonance imaging with contrast dependent on blood oxygenation. References Proc Natl Acad Sci U S A 87 (24):9868–9872. doi:10.1073/pnas.87.24.9868 26. Oishi K, Faria A, Jiang H, Li X, Akhter K, Zhang J, Hsu JT, Miller MI, van Zijl PCM, Albert M, Lyketsos CG, Woods R, Toga AW, Pike GB, Rosa-Neto P, Evans A, Mazziotta J, Mori S (2009) Atlas-based whole brain white matter analysis using large deformation diffeomorphic metric mapping: Application to normal elderly and Alzheimer's disease participants. NeuroImage 46 (2):486–499 27. Pandya D, Petrides M, Cipolloni PB (2015) Cerebral Cortex: Architecture, Connections, and the Dual Origin Concept. Oxford University Press, 27. Pandya D, Petrides M, Cipolloni PB (2015) Cerebral Cortex: Architecture, Connections, and the Dual Origin Concept. Oxford University Press, Page 14/24 28. Petit L, Ali KM, Rheault F, Bore A, Cremona S, Corsini F, De Benedictis A, Descoteaux M, Sarubbo S (2023) The structural connectivity of the human angular gyrus as revealed by microdissection and Page 14/24 28. Petit L, Ali KM, Rheault F, Bore A, Cremona S, Corsini F, De Benedictis A, Descoteaux M, Sarubbo S (2023) The structural connectivity of the human angular gyrus as revealed by microdissection and diffusion tractography. Brain Struct Funct 228 (1):103–120. doi:10.1007/s00429-022-02551-5 diffusion tractography. Brain Struct Funct 228 (1):103–120. doi:10.1007/s00429-022-02551-5 29. Piras F, Caltagirone C, Spalletta G (2010) Working memory performance and thalamus microstructure in healthy subjects. Neuroscience 171 (2):496–505. doi:10.1016/j.neuroscience.2010.09.006 30. Reuter M, Rosas HD, Fischl B (2010) Highly accurate inverse consistent registration: a robust approach. Neuroimage 53 (4):1181–1196. doi:10.1016/j.neuroimage.2010.07.020 31. Rheault F, Poulin P, Valcourt Caron A, St-Onge E, Descoteaux M (2020) Common misconceptions, hidden biases and modern challenges of dMRI tractography. J Neural Eng 17 (1):011001. doi:10.1088/1741-2552/ab6aad 32. Roy DS, Zhang Y, Aida T, Shen C, Skaggs KM, Hou Y, Fleishman M, Mosto O, Weninger A, Feng G (2022) Anterior thalamic circuits crucial for working memory. Proc Natl Acad Sci U S A 119 (20):e2118712119. doi:10.1073/pnas.2118712119 33. Schilling K, Gao Y, Janve V, Stepniewska I, Landman BA, Anderson AW (2018) Confirmation of a gyral bias in diffusion MRI fiber tractography. Hum Brain Mapp 39 (3):1449–1466. doi:10.1002/hbm.23936 34. Schmahmann JD, Pandya DN (2006) Fiber pathways of the brain. Oxford University Press, New York 34. Schmahmann JD, Pandya DN (2006) Fiber pathways of the brain. Oxfor 35. Schüz A, Braitenberg V (2002) The human cortical white matter: Quantitative aspects of cortico- cortical long-range connectivity. In: Schuz A, Miller R (eds) Cortical areas: Unity and diversity. References Conceptual Advances in Brain Research. Taylor & Francis, London, pp 377–386 36. Shastin D, Genc S, Parker GD, Koller K, Tax CMW, Evans J, Hamandi K, Gray WP, Jones DK, Chamberland M (2022) Surface-based tracking for short association fibre tractography. Neuroimage 260:119423. doi:10.1016/j.neuroimage.2022.119423 37. Smith RE, Tournier JD, Calamante F, Connelly A (2015) The effects of SIFT on the reproducibility and biological accuracy of the structural connectome. Neuroimage 104:253–265. doi:10.1016/j.neuroimage.2014.10.004 38. St-Onge E, Al-Sharif N, Girard G, Theaud G, Descoteaux M (2021) Cortical Surfaces Integration with Tractography for Structural Connectivity Analysis. Brain Connect 11 (7):505–517. doi:10.1089/brain.2020.0930 39. St-Onge E, Daducci A, Girard G, Descoteaux M (2018) Surface-enhanced tractography (SET). Neuroimage 169:524–539. doi:10.1016/j.neuroimage.2017.12.036 40. Tarun A, Behjat H, Bolton T, Abramian D, Van De Ville D (2020) Structural mediation of human brain activity revealed by white-matter interpolation of fMRI. Neuroimage 213:116718. doi:10.1016/j.neuroimage.2020.116718 41. Theaud G, Houde JC, Bore A, Rheault F, Morency F, Descoteaux M (2020) TractoFlow: A robust, efficient and reproducible diffusion MRI pipeline leveraging Nextflow & Singularity. Neuroimage 218:116889. doi:10.1016/j.neuroimage.2020.116889 Page 15/24 42. Thiebaut de Schotten M, Ffytche DH, Bizzi A, Dell'Acqua F, Allin M, Walshe M, Murray R, Williams SC, Murphy DG, Catani M (2011) Atlasing location, asymmetry and inter-subject variability of white matter tracts in the human brain with MR diffusion tractography. Neuroimage 54 (1):49–59. doi:10.1016/j.neuroimage.2010.07.055 matter tracts in the human brain with MR diffusion tractography. Neuroimage 54 (1):49–59. doi:10.1016/j.neuroimage.2010.07.055 43. Thiebaut de Schotten M, Forkel SJ (2022) The emergent properties of the connected brain. Science 378 (6619):505–510. doi:10.1126/science.abq2591 43. Thiebaut de Schotten M, Forkel SJ (2022) The emergent properties of the connected brain. Science 378 (6619):505–510. doi:10.1126/science.abq2591 44. Tournier JD, Smith R, Raffelt D, Tabbara R, Dhollander T, Pietsch M, Christiaens D, Jeurissen B, Yeh CH, Connelly A (2019) MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation. Neuroimage 202:116137. doi:10.1016/j.neuroimage.2019.116137 44. Tournier JD, Smith R, Raffelt D, Tabbara R, Dhollander T, Pietsch M, Christiaens D, Jeurissen B, Yeh CH, Connelly A (2019) MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation. Neuroimage 202:116137. doi:10.1016/j.neuroimage.2019.116137 45. References Ugurbil K, Xu J, Auerbach EJ, Moeller S, Vu AT, Duarte-Carvajalino JM, Lenglet C, Wu X, Schmitter S, Van de Moortele PF, Strupp J, Sapiro G, De Martino F, Wang D, Harel N, Garwood M, Chen L, Feinberg DA, Smith SM, Miller KL, Sotiropoulos SN, Jbabdi S, Andersson JL, Behrens TE, Glasser MF, Van Essen DC, Yacoub E, Consortium WU-MH (2013) Pushing spatial and temporal resolution for functional and diffusion MRI in the Human Connectome Project. Neuroimage 80:80–104. doi:10.1016/j.neuroimage.2013.05.012 45. Ugurbil K, Xu J, Auerbach EJ, Moeller S, Vu AT, Duarte-Carvajalino JM, Lenglet C, Wu X, Schmitter S, Van de Moortele PF, Strupp J, Sapiro G, De Martino F, Wang D, Harel N, Garwood M, Chen L, Feinberg DA, Smith SM, Miller KL, Sotiropoulos SN, Jbabdi S, Andersson JL, Behrens TE, Glasser MF, Van Essen DC, Yacoub E, Consortium WU-MH (2013) Pushing spatial and temporal resolution for functional and diffusion MRI in the Human Connectome Project. Neuroimage 80:80–104. doi:10.1016/j.neuroimage.2013.05.012 46. Van Essen DC, Smith SM, Barch DM, Behrens TE, Yacoub E, Ugurbil K, Consortium WU-MH (2013) The WU-Minn Human Connectome Project: an overview. Neuroimage 80:62–79. doi:10.1016/j.neuroimage.2013.05.041 46. Van Essen DC, Smith SM, Barch DM, Behrens TE, Yacoub E, Ugurbil K, Consortium WU-MH (2013) The WU-Minn Human Connectome Project: an overview. Neuroimage 80:62–79. doi:10.1016/j.neuroimage.2013.05.041 47. Wang Z, Dai Z, Gong G, Zhou C, He Y (2015) Understanding structural-functional relationships in the human brain: a large-scale network perspective. Neuroscientist 21 (3):290–305. doi:10.1177/1073858414537560 48. Watanabe Y, Funahashi S (2012) Thalamic mediodorsal nucleus and working memory. Neurosci Biobehav Rev 36 (1):134–142. doi:10.1016/j.neubiorev.2011.05.003 49. Yarkoni T, Poldrack RA, Nichols TE, Van Essen DC, Wager TD (2011) Large-scale automated synthesis of human functional neuroimaging data. Nat Methods 8 (8):665–670. doi:10.1038/nmeth.1635 Figure 1 Graphical summary of the processing and analysis. In orange, each processing step. In green, the new priors now provided with the Functionnectome to improve the integration of functional and structural information on the white matter. Here, “class” refers to the classes of white matter fibers included in the tractograms (as listed below the “Filtered tractograms” box). Graphical summary of the processing and analysis. In orange, each processing step. In green, the new priors now provided with the Functionnectome to improve the integration of functional and structural information on the white matter. Here, “class” refers to the classes of white matter fibers included in the tractograms (as listed below the “Filtered tractograms” box). Figure 2 Streamline lengthening procedure. A: All the streamlines are lengthened by a specific amount (5 mm here). B: The parts of the extension considered incorrect are removed. a: Correct extension. b: Part of the Figures Page 16/24 Page 16/24 Page 16/24 Figure 1 Figure 2 Streamline lengthening procedure. A: All the streamlines are lengthened by a specific amount (5 mm here). B: The parts of the extension considered incorrect are removed. a: Correct extension. b: Part of the Page 17/24 extension leaves the grey matter. c: Part of the extension crosses to another grey matter parcel. d: The extension doesn’t reach the grey matter. Figure 3 Schematic illustration of the three different classes of white matter fibers (and streamlines). Association fibers link different parts of the same hemisphere. Commissural fibers link the two hemispheres. Projection fibers comprise all the fibers connected to subcortical structures and the brainstem. Figure 3 Schematic illustration of the three different classes of white matter fibers (and streamlines). Association fibers link different parts of the same hemisphere. Commissural fibers link the two hemispheres. Projection fibers comprise all the fibers connected to subcortical structures and the brainstem. Schematic illustration of the three different classes of white matter fibers (and streamlines). Association fibers link different parts of the same hemisphere. Commissural fibers link the two hemispheres. Projection fibers comprise all the fibers connected to subcortical structures and the brainstem. Page 18/24 Figure 4 Effect of streamline extension and fiber categorization on the group-level functionnectome z-map of the right-hand finger-tapping task. A: Z-map generated using whole-brain priors from non-extended tractograms. B: Z-map generated using whole-brain priors from extended tractograms. C: Z-max map generated by recombining the z-maps obtained from using the priors split by fiber class. Figure 4 Effect of streamline extension and fiber categorization on the group-level functionnectome z-map of the Figure 4 Figure 4 Effect of streamline extension and fiber categorization on the group-level functionnectome z-map of the right-hand finger-tapping task. A: Z-map generated using whole-brain priors from non-extended tractograms. B: Z-map generated using whole-brain priors from extended tractograms. C: Z-max map generated by recombining the z-maps obtained from using the priors split by fiber class. Page 19/24 Page 19/24 Figure 5 Effect of streamline extension and fiber splitting on the group-level functionnectome z-map of th working memory task. A: Z-map generated using whole-brain priors from non-extended tractogra map generated using whole-brain priors from extended tractograms. C: Z-max map generated by recombining the z-maps obtained from using the priors split by fiber class. Figure 5 Figure 5 Effect of streamline extension and fiber splitting on the group-level functionnectome z-map of the working memory task. A: Z-map generated using whole-brain priors from non-extended tractograms. B: Z- map generated using whole-brain priors from extended tractograms. C: Z-max map generated by recombining the z-maps obtained from using the priors split by fiber class. Page 20/24 Figure 6 Effect of categorization by the length of association fibers on the group-level functionnectome z-map of the right-hand finger-tapping task. A: Z-map generated using association priors. B: Z-max map generated by recombining the z-maps obtained from using the association priors split according to fiber length (short, medium, and long association fibers). Figure 6 Effect of categorization by the length of association fibers on the group-level functionnectome z-map of the right-hand finger-tapping task. A: Z-map generated using association priors. B: Z-max map generated by recombining the z-maps obtained from using the association priors split according to fiber length (short, medium, and long association fibers). Effect of categorization by the length of association fibers on the group-level functionnectome z-map of the right-hand finger-tapping task. A: Z-map generated using association priors. B: Z-max map generated by recombining the z-maps obtained from using the association priors split according to fiber length (short, medium, and long association fibers). Page 21/24 Page 21/24 Page 21/24 Page 21/24 Figure 7 White matter activation maps for each class of fibers in the right-hand finger-tapping task. A: Z-max map generated by recombining the z-maps generated using the three sets of association priors (split by length), the asterisk () indicates that this is a Z-max map, contrary to the other two maps. B: Z-map generated using commissural priors. C: Z-map generated using projection priors. Figure 7 Figure 7 White matter activation maps for each class of fibers in the right-hand finger-tapping task. A: Z-max map generated by recombining the z-maps generated using the three sets of association priors (split by length), the asterisk () indicates that this is a Z-max map, contrary to the other two maps. B: Z-map generated using commissural priors. C: Z-map generated using projection priors. White matter activation maps for each class of fibers in the right-hand finger-tapping task. A: Z-max map generated by recombining the z-maps generated using the three sets of association priors (split by length), the asterisk () indicates that this is a Z-max map, contrary to the other two maps. B: Z-map generated using commissural priors. C: Z-map generated using projection priors. Page 22/24 Figure 8 White matter activation maps for each class of fibers in the working-memory task. A: Z-max map generated by recombining the z-maps generated using the three sets of association priors (split by length), the asterisk () indicates that this is a Z-max map, contrary to the other two maps. B: Z-map generated using commissural priors. C: Z-map generated using projection priors. floatimage1.png Figure 9 White matter activation maps for each length category of association fibers in the right-hand finger- tapping task and working-memory task. A-B-C: Finger-tapping Z-maps generated using priors for the short, medium, and long fibers category, respectively. D-E-F: Working-memory Z-maps generated using priors for the short, medium, and long fibers category, respectively. Figure 8 White matter activation maps for each class of fibers in the working-memory task. A: Z-max map generated by recombining the z-maps generated using the three sets of association priors (split by length), the asterisk () indicates that this is a Z-max map, contrary to the other two maps. B: Z-map generated using commissural priors. C: Z-map generated using projection priors. White matter activation maps for each class of fibers in the working-memory task. A: Z-max map generated by recombining the z-maps generated using the three sets of association priors (split by length), the asterisk () indicates that this is a Z-max map, contrary to the other two maps. B: Z-map generated using commissural priors. C: Z-map generated using projection priors. Page 23/24 Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. Page 24/24
https://openalex.org/W4229947286	https://e-journal.unair.ac.id/JBE/article/download/1633/2541	Indonesian	null	Correlation Between Oral Contraceptives Use and Sexual Activity with Cervical Cancer	Jurnal Berkala Epidemiologi/Jurnal berkala epidemiologi	2,017	cc-by-sa	7,030	HUBUNGAN FAKTOR RISIKO PENGGUNAAN KONTRASEPSI ORAL DAN AKTIVITAS SEKSUAL DENGAN KEJADIAN KANKER SERVIKS Association between Oral Contraceptives Use and Sexual Activity with Cervical Cancer Vita Wulandari FKM Universitas Airlangga, fita.theone@gmail.com Alamat Korespondensi: Departemen Epidemiologi Fakultas Kesehatan Masyaraka Surabaya, Jawa Timur, Indonesia Vita Wulandari FKM Universitas Airlangga, fita.theone@gmail.com Departemen Epidemiologi Fakultas Kesehatan Masyarakat Universitas Airlangga Surabaya, Jawa Timur, Indonesia ABSTRAK World Health Organization pada tahun 2012 menunjukkan bahwa dari total kasus baru kanker serviks di dunia, 85% kasus ditemukan di negara berkembang. Tahun 2015 Kota Malang dan Kabupaten Malang merupakan daerah waspada kanker serviks. Penelitian ini bertujuan menganalisis hubungan antara faktor risiko penggunaan kontrasepsi oral dan aktivitas seksual dengan kejadian kanker serviks pada pasien di Poli Obstetri dan Ginekologi Rumah Sakit Umum Daerah (RSUD) Dr. Saiful Anwar Malang. Penelitian ini menggunakan rancangan case control. Populasi penelitian yaitu pasien rawat jalan poli obstetri dan ginekologi RSUD Dr. Saiful Anwar Malang bulan November 2015. Sampel kasus sebanyak 37 pasien kanker serviks dan sampel kontrol sebanyak 111 pasien bukan kanker serviks. Pengambilan sampel menggunakan systematic random sampling. Analisis data menggunakan uji Chi-square. Variabel bebas yang diteliti adalah penggunaan kontrasepsi oral dan aktivitas seksual yang meliputi usia pertama kali berhubungan seksual < 18 tahun, usia pertama kali hamil < 18 tahun, dan riwayat abortus. Pengumpulan data primer menggunakan wawancara sedangkan sekunder menggunakan rekam medik pasien. Hasil penelitian menunjukkan ada hubungan antara usia pertama kali berhubungan seksual < 18 tahun (p=0,0225147014; OR=2,3194; 95% CI=1,0854–4,9561), usia pertama kali hamil < 18 tahun (p=0,0236276656; OR=2,3388; 95% CI 1,0890–5,0230), dan riwayat abortus (p=0,0038911219; OR=3,2653; 95%CI=1,4593–7,3063) dengan kanker serviks. Kesimpulan penelitian ini yaitu usia pertama kali berhubungan seksual, usia pertama kali hamil dan abortus merupakan faktor risiko kejadian kanker serviks. Kata kunci: kontrasepsi, oral, abortus, hubungan seks, kanker serviks. ©2016 FKM_UNAIR All right reserved. Open access under CC BY– SA license doi: 10.20473/jbe.v4i3. 2016. 432–443 Received 23 March 2016, received in revised form 8 December 2016, Accepted 23 December 2016, Published online: 21 January 2017 ABSTRACT ABSTRACT The World Health Organization in 2012 showed that of the total new cases of cervical cancer in the world, 85% of cases were found in developing countries. In 2015, Malang City and Malang District were alert for cervical cancer. This study aims to analyze the relationship between risk factors for use of oral contraceptives and sexual activity with the incidence of cervical cancer in patients in the Obstetrics and Gynecology Clinic of the Regional Public Hospital (RSUD) Dr. Saiful Anwar Malang. This study uses a case-control design. The study population was poly obstetrics and gynecology outpatients Dr. Saiful Anwar Malang in November 2015. Case samples were 37 cervical cancer patients and 111 control samples were non-cervical cancer patients. Sampling using systematic random sampling. Data analysis using the Chi- square test. The independent variables studied were the use of oral contraceptives and sexual activity which included the age of first sexual intercourse <18 years, the age of first pregnancy <18 years, and a history of abortion. Primary data collection uses interviews while secondary uses the patient's medical record. The results showed there was a relationship between the age of first sexual intercourse <18 years (p = 0.0225147014; OR = 2.3194; 95% CI = 1.0854–4.9561), age of first pregnancy <18 years (p = 0.0236276656; OR = 2.33388; 95% CI 1.0890-5.0230), and history of abortion (p = 0.0038911219; OR = 3.2653; 95% CI = 1.4593–7.3063) with cervical cancer. The conclusion of this study is the age of first sexual intercourse, age at first pregnancy and abortion are risk factors for cervical cancer. Keywords: contraceptives, oral, abortion, sexual intercourse, cervical cancer PENDAHULUAN kanker serviks disebabkan oleh Human Papilloma Virus (HPV). Kanker serviks dapat dicegah dengan vaksinasi HPV, skrining, serta pengobatan lesi pra-kanker (WHO, 2014). WHO menyatakan Kanker serviks merupakan suatu keganasan pada serviks (Kemenkes RI, 2010). Sebagian besar ©2016 FKM_UNAIR All right reserved. Open access under CC BY– SA license doi: 10.20473/jbe.v4i3. 2016. 432–443 Received 23 March 2016, received in revised form 8 December 2016, Accepted 23 December 2016, Published online: 21 January 2017 Vita Wulandari, Hubungan Faktor Risiko Penggunaan Kontrasepsi ... 433 bahwa pada tahun 2012 dari 528.000 kasus baru kanker serviks di dunia 85% diantaranya terjadi di Negara berkembang. Pada tahun yang sama, terdapat 266.000 ibu meninggal karena kanker serviks. Hampir di seluruh dunia 9 dari 10 kasus tersebut sekitar 231.000 perempuan hidup dan mati di negara pendapatan menengah ke bawah. Sebaliknya, 1 dari 10 kasus tersebut atau 35.000 wanita hidup dan mati di negara dengan pendapatan tinggi (WHO, 2014). Data kasus kanker serviks yang diperoleh dari Surveilans Terpadu Penyakit (STP) Rumah Sakit Sentinel Jawa Timur pada tahun 2011–2014 selalu mengalami peningkatan setiap tahunnya. Kasus kanker serviks di Jawa Timur pada tahun 2013 sebesar 3.971 kasus sedangkan pada tahun 2014 sebesar 4.094 kasus. bahwa pada tahun 2012 dari 528.000 kasus baru kanker serviks di dunia 85% diantaranya terjadi di Negara berkembang. Pada tahun yang sama, terdapat 266.000 ibu meninggal karena kanker serviks. Hampir di seluruh dunia 9 dari 10 kasus tersebut sekitar 231.000 perempuan hidup dan mati di negara pendapatan menengah ke bawah. Sebaliknya, 1 dari 10 kasus tersebut atau 35.000 wanita hidup dan mati di negara dengan pendapatan tinggi (WHO, 2014). Data kasus kanker serviks yang diperoleh dari Surveilans Terpadu Penyakit (STP) Rumah Sakit Sentinel Jawa Timur pada tahun 2011–2014 selalu mengalami peningkatan setiap tahunnya. Kasus kanker serviks di Jawa Timur pada tahun 2013 sebesar 3.971 kasus sedangkan pada tahun 2014 sebesar 4.094 kasus. 2014). Hal ini menunjukkan bahwa kontrasepsi oral atau pil masih banyak diminati, sedangkan kontrasepsi oral sendiri merupakan salah satu faktor risiko dari kanker serviks, apalagi jika penggunaannya semakin lama atau lebih dari 5 tahun. Aktivitas seksual seperti usia pertama kali berhubungan seksual, usia pertama kali hamil dan riwayat abortus juga merupakan faktor risiko kanker serviks. Namun wanita yang memiliki riwayat usia pertama kali berhubungan seksual pada usia < 18 tahun di Indonesia juga masih banyak. Semakin muda usia melakukan hubungan seksual pertama kali memengaruhi besarnya risiko terjadinya kanker serviks. PENDAHULUAN Selain usia pertama kali berhubungan seksual, usia pertama kali hamil juga merupakan faktor risiko kanker serviks yang dapat meningkatkan insiden kanker serviks. Semakin muda usia pertama kali hamil atau < 18 tahun maka usia pertama kali berhubungan seksual juga semakin muda sehingga serviks lebih mudah terpapar HPV. Wanita yang pernah memiliki riwayat abortus memiliki peningkatan risiko kanker serviks dikarenakan terjadi perlukaan pada uterus dan serviks. Alasan tersebut yang mendasari peneliti ingin menganalisis hubungan antara faktor risiko penggunaan kontrasepsi oral dan aktivitas seksual seperti usia pertama kali berhubungan seksual, usia pertama kali hamil dan riwayat abortus dengan kejadian kanker serviks. Kota Malang pada tahun 2011 merupakan penyumbang terbesar jumlah kasus kanker serviks di Jawa Timur yaitu sebanyak 747 kasus yang merupakan jumlah tertinggi di Provinsi Jawa Timur (Republika, 2013). Pada tahun 2012 menurut data STP Rumah Sakit Sentinel Jawa Timur ditemukan kasus tertinggi yaitu di RSUD Dr. Saiful Anwar. Tren kasus kanker serviks di RSUD Dr. Saiful Anwar Malang pada tahun 2011–2013 selalu mengalami peningkatan. Tahun 2015 Dinas Kesehatan Provinsi Jawa Timur mencatat ada tujuh kota dan kabupaten yang masuk katagori daerah waspada bahaya kanker serviks diantaranya terdapat Kabupaten Malang dan Kota Malang (Surya, 2015). Kanker serviks seringkali tidak menunjukkan gejala. Kebanyakan pasien diketahui positif kanker serviks sudah pada stadium lanjut, hal ini dikarenakan sebagian besar wanita tidak mengetahui faktor risiko kanker serviks, tanda maupun gejalanya sehingga mereka terlambat untuk melakukan skrining kanker serviks. METODE Penelitian ini merupakan penelitian analitik observasional dengan menggunakan rancangan case control. Populasi pada penelitian ini terbagi menjadi dua kelompok yakni populasi kasus dan populasi kontrol. Populasi kasus yaitu pasien rawat jalan ruang onkologi poli obstetri dan ginekologi RSUD Dr. Saiful Anwar Malang yang memeriksakan diri pada bulan November 2015 dan tercatat pada rekam medik yang didiagnosa kanker serviks. Sedangkan populasi kontrol yaitu pasien rawat jalan ruang ginekologi, KB, fertilitas, nifas dan hamil poli obstetri dan ginekologi RSUD Dr. Saiful Anwar Malang yang memeriksakan diri pada bulan November 2015 dan tercatat pada rekam medik tidak didiagnosa kanker serviks. Warga Indonesia sekarang ini sebagian besar menggunakan kontrasepsi untuk membatasi dan menjaga jarak kelahiran anaknya. Semakin meningkatnya jumlah akseptor KB ini dikarenakan adanya program pemerintah untuk mencegah peledakan penduduk dimulai pada masa Orde Baru sampai saat ini. Data SDKI menunjukkan tren prevalensi penggunaan kontrasepsi di Indonesia sejak tahun 1991–2012 cenderung meningkat. Data BKKBN menunjukkan bahwa pada tahun 2013 dari 8.500.247 PUS yang merupakan peserta KB baru sebesar 48,56% atau hampir separuhnya menggunakan metode kontrasepsi suntikan, dan 26,6% menggunakan pil sebagai kontrasepsi pilihan kedua setelah suntikan (Pusat Data dan Informasi, Penelitian ini menggunakan perbandingan 1:3 dengan besar sampel kasus 37 dan sampel kontrol sebesar 111. Pada penelitian ini menggunakan kriteria inklusi dan eksklusi agar tidak terjadi bias. 434 Jurnal Berkala Epidemiologi, Vol. 4 No. 3, September 2016: 432–442 kepada responden menggunakan lembar wawancara dan data sekunder yaitu data rekam medik pasien. Kriteria inklusi kelompok kasus dan kontrol pada penelitian ini yaitu pernah atau sedang menggunakan kontrasepsi oral kombinasi atau pil kombinasi dan pernah hamil. Kriteria eksklusi untuk kelompok kasus yaitu pernah menggunakan kontrasepsi hormonal selain kontrasepsi oral kombinasi. Sedangkan kriteria eksklusi untuk kelompok kontrol yaitu pernah menggunakan kontrasepsi oral kombinasi dan juga didiagnosa oleh dokter menderita kanker dan tercatat di rekam medik. Pengolahan dan analisis data dikerjakan setelah peneliti melakukan pengumpulan data. Analisis data dilakukan secara deskriptif dan analitik. Analisis deskriptif menggunakan tabulasi silang. Sedangkan analisis secara analitik dengan menggunakan uji statistik Chi Square. Hubungan antara penggunaan kontrasepsi oral dan aktivitas seksual dengan kejadian kanker serviks dapat diketahui melalui uji Chi Square dengan kemaknaan statistik sebesar 0,05. Besarnya risiko penggunaan kontrasepsi oral dan aktivitas seksual untuk menjadi kanker serviks dapat diketahui dengan menggunakan perhitungan Odds Ratio dengan 95% CI. Penentuan sampel pada penelitian ini dengan cara acak menggunakan sistematic random sampling karena total populasi tidak pasti. Karakteristik Responden Tabel 1 menunjukkan bahwa responden yang menderita kanker serviks sebanyak 54,05% berusia > 50 tahun, begitu juga responden yang tidak menderita kanker serviks sebagian besar yaitu sebesar 75,7% berusia 50 tahun. Pendidikan responden yang menderita kanker serviks sebagian besar berpendidikan SD yaitu sebanyak 48,6%. Sedangkan responden yang tidak menderita kanker serviks sebanyak 36% memiliki pendidikan terakhir SMA. Responden yang menderita kanker serviks sebagian besar bekerja sebagai ibu rumah tangga yaitu sebanyak 51,35%. Begitu juga responden yang tidak menderita kanker serviks sebagian besar bekerja sebagai ibu rumah tangga yaitu sebanyak 51,35%. Lokasi penelitian di poli obstetri dan ginekologi RSUD Dr. Saiful Anwar Malang pada bulan Maret sampai November 2015. Sedangkan pengambilan data dilakukan pada bulan November 2015. Sumber data yang digunakan berasal dari data primer yang didapatkan melalui wawancara secara langsung Tabel 1. Karakteristik Responden berdasarkan Usia, Pendidikan dan Pekerjaan Variabel Kasus (n = 37) Kontrol (n = 111) Usia < 50 tahun > 50 tahun 17 (45,95%) 20 (54,05%) 27 (24,3%) 84 (75,7%) Pendidikan Tidak Tamat SD SD SMP SMA PT 6 (16,2%) 18 (48,6%) 6 (16,2%) 6 (16,2%) 1 (2,7%) 2 (1,8%) 32 (28,8%) 32 (28,8%) 40 (36%) 5 (4,5%) Pekerjaan IRT Pedagang atau Wiraswasta Swasta PNS Petani 19 (51,35%) 2 (5,41%) 8 (21,62%) 1 (2,7%) 7 (18,92%) 64 (57,7%) 2 (1,8%) 38 (34,2%) 3 (2,7%) 4 (3,6%) Tabel 1. Karakteristik Responden berdasarkan Usia, Pendidikan dan Pekerjaan METODE Pada penelitian ini setelah diketahui besar sampel maka dilakukan pengambilan sampel dengan cara menentukan interval antar sampel terlebih dahulu. Setelah didapatkan interval maka diambil sampel pertama secara acak dari angka 1 sampai dengan interval antar sampel. Sampel selanjutnya dipilih secara sistematik berdasarkan interval yang telah ditetapkan sampai didapatkan jumlah yang sesuai dengan besar sampel. Pada penelitian ini didapatkan interval antar sampel penelitian yaitu 2. Rumus interval antar sampel menurut Supranto (2007) yaitu total populasi dibagi dengan jumlah sampel. Hubungan Usia Pertama Kali Berhubungan Seksual dengan Kejadian Kanker Serviks Tabel 3 menunjukkan bahwa responden yang menderita kanker serviks sebanyak 59,5% melakukan hubungan seksual pertama kali pada usia < 18 tahun. Sedangkan responden yang tidak menderita kanker serviks sebagian besar melakukan hubungan seksual pertama kali pada usia > 18 tahun yaitu sebesar 61,3 %. Nilai p pada Chi Square yaitu sebesar 0,045 dengan α = 0,05. Dapat disimpulkan bahwa nilai p < dari α, sehingga terdapat hubungan yang signiﬁ kan antara usia pertama kali berhubungan seksual dengan kejadian kanker serviks. Analisis data menghasilkan OR 2,319 dengan nilai 95% CI (1,085 < OR < 4,956). Angka besar risiko tersebut bermakna secara epidemiologi. Hal ini menunjukkan bahwa wanita yang pertama kali melakukan hubungan seksual pada usia < 18 tahun berisiko menderita kanker serviks sebesar 2,3 kali Usia Pertama Kali Hamil Tabel 3 menunjukkan bahwa responden yang menderita kanker serviks sebanyak 51,4% memiliki riwayat pertama kali hamil pada usia > 18 tahun. Begitu juga responden yang tidak menderita kanker serviks sebagian besar (71,2%) memilki riwayat pertama kali hamil pada usia > 18 tahun. Nilai p pada Chi Square sebesar 0,045. Nilai p tersebut lebih kecil dari α (0,05). Sehingga secara statistik menunjukkan bahwa terdapat hubungan yang signiﬁ kan antara usia pertama kali hamil dengan kejadian kanker serviks. Analisis data menunjukkan hasil OR 2,339 dengan 95% CI (1,089 < OR < 5,023). Hal ini menunjukkan bahwa angka besar risiko tersebut bermakna secara epidemiologi. Sehingga dapat disimpulkan bahwa wanita yang hamil pertama kali pada usia < 18 tahun berisiko menderita kanker serviks sebesar 2,3 kali bila dibandingkan dengan wanita yang hamil pertama kali pada usia > 18 tahun. Hubungan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks Responden pada penelitian ini sudah dipilih berdasarkan kriteria inklusi dan eksklusi sampel penelitian yaitu yang pernah menggunakan kontrasepsi oral kombinasi. Tabel 2 menunjukkan bahwa responden yang menderita kanker serviks sebagian besar pernah menggunakan kontrasepsi oral kombinasi selama < 5 tahun yaitu sebanyak 56,76 %, begitu juga responden yang tidak menderita kanker serviks sebagian besar (63,06 %) pernah menggunakan kontrasepsi oral kombinasi selama < 5 tahun. Nilai p pada Chi Square yaitu sebesar 0,626 dengan α = 0,05. Nilai p tersebut > α yang menunjukkan bahwa H0 diterima. Sehingga secara statistik dapat disimpulkan bahwa tidak Vita Wulandari, Hubungan Faktor Risiko Penggunaan Kontrasepsi ... 435 Tabel 2. Hubungan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks Penggunaan Kontrasepsi Oral Kasus Kontrol OR (95% CI) > 5 tahun < 5 tahun 16 (43,24 %) 21 (56,76 %) 41 (36,94 %) 70 (63,06 %) 1,3008 (0,611–2,771) Total 37 (100 %) 111 (100 %) Tabel 2. Hubungan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks bila dibandingkan dengan wanita yang pertama kali melakukan hubungan seksual pada usia > 18 tahun. ada hubungan yang signiﬁ kan antara penggunaan kontrasepsi oral dengan kejadian kanker serviks. Analisis data menghasilkan OR 1,301 dengan nilai 95% CI (0,611 < OR < 2,771). Angka besar risiko tersebut tidak bermakna secara epidemiologi. Hal ini berarti penggunaan kontrasepsi oral > 5 tahun tidak meningkatkan risiko kanker serviks. Karakteristik Responden Hasil penelitian ini menunjukkan bahwa responden yang menderita kanker serviks maupun yang tidak menderita kanker serviks sebagian besar berusia > 50 tahun. Hal ini menunjukkan bahwa sebagian besar responden berada pada usia menopause. Seringkali kanker serviks tidak menunjukkan gejala, sebagian besar pasien kanker serviks terdiagnosa pada stadium lanjut dan pada usia tua. Perkembangan prekanker menjadi kanker serviks invasif membutuhkan waktu sekitar 10 tahun lebih. Penelitian yang dilakukan oleh Lestari (2012) pada ibu rumah tangga pasangan usia subur yang datang berkunjung ke Puskesmas Jaten II Kabupaten Karanganyar menunjukkan bahwa terdapat hubungan yang signiﬁ kan antara tingkat pendidikan dengan perilaku deteksi dini kanker serviks baik IVA ataupun pap smear. Hasil perhitungan Chi Square menghasilkan nilai p = 0,017 dengan α = 0,05. Melakukan skrining lebih awal memungkinkan untuk dilakukan penanganan lebih cepat jika terdeteksi hasil pemeriksaan positif sehingga insiden kanker serviks dapat menurun. Saat ini program pemerintah untuk melakukan skrining pre kanker dengan IVA dilakukan pada wanita usia subur yaitu pada usia 30–50 tahun. Semakin awal dan tepat diagnosa diharapkan semakin cepat pemberian therapy yang diberikan. Kemenkes RI (2010) menyatakan bahwa walaupun wanita usia 40 tahun merupakan kelompok usia yang paling sering terserang kanker serviks, namun apabila dilakukan skrining pada 10 tahun sebelumnya atau lebih mungkin dapat mencegah perubahan displasia tingkat tinggi untuk tumbuh menjadi kanker serviks yang biasa disebut Cervical Intraephitelial Neoplasia (CIN) II atau III. Pada penelitian yang dilakukan Winship Cancer Institute of Emory University (2014) menunjukkan hasil bahwa sangat sedikit wanita di bawah usia 20 tahun yang telah didiagnosa kanker serviks dan setengah dari mereka didiagnosa pada usia antara 35 dan 55 tahun. Risiko berkurang setelah usia 55, tetapi 20% dari kasus menetap pada wanita usia 60 tahun lebih. Terlihat Jenis pekerjaan responden baik yang menderita kanker serviks maupun tidak menderita kanker serviks sebagian besar bekerja sebagai ibu rumah tangga. Jenis pekerjaan atau sosio- ekonomi berpengaruh pada kemampuan seseorang untuk memeriksakan diri ke fasilitas kesehatan apabila terdapat tanda dan gejala sakit, ataupun untuk skrining kanker serviks. Hasil penelitian ini sama dengan penelitian yang dilakukan oleh Syatriani (2011), di RSU Pemerintah Dr. Wahidin Sudirohusodo Makassar yang menunjukkan bahwa sebagian besar responden yang menderita kanker serviks bekerja sebagai ibu rumah tangga (88,73%), begitu juga responden yang tidak menderita kanker serviks bekerja sebagai ibu rumah tangga (61,97%) dengan OR sebesar 4,087. Riwayat Abortus Tabel 3 menunjukkan bahwa responden yang menderita kanker serviks sebanyak 56,8% tidak pernah mengalami abortus, begitu juga responden yang tidak menderita kanker serviks sebagian besar Tabel 3. Hubungan Aktivitas Seksual dengan Kejadian Kanker Serviks Variabel Aktivitas Seksual Kasus Kontrol OR (95% CI) Usia Pertama Kali Berhubungan Seksual < 18 tahun > 18 tahun 22 (59,5 %) 15 (40,5 %) 43 (38,7 %) 68 (61,3 %) 2,319 (1,085–4,956) Usia Pertama Kali Hamil < 18 tahun > 18 tahun 18 (48,6 %) 19 (51,4 %) 32 (28,8 %) 79 (71,2 %) 2,339 (1,089–5,023) Riwayat Abortus Ya Tidak 16 (43,2 %) 21 (56,8 %) 21 (18,9 %) 90 (81,1 %) 3,265 (1,459– ,306) Total 37 (100 %) 111 (100 %) Tabel 3. Hubungan Aktivitas Seksual dengan Kejadian Kanker Serviks Jurnal Berkala Epidemiologi, Vol. 4 No. 3, September 2016: 432–442 436 tidak pernah mengalami abortus yaitu sebesar 81, 1%. Nilai p pada Chi Square sebesar 0,006 dengan α = 0,05. Sehingga dapat disimpulkan bahwa terdapat hubungan yang signiﬁ kan antara riwayat abortus dengan kejadian kanker serviks. Analisis data menunjukkan nilai OR sebesar 3,265 dengan 95% CI (1,459 < OR < 7,306). Hal ini menunjukkan bahwa angka besar risiko tersebut bermakna secara epidemiologi. Sehingga dapat disimpulkan bahwa wanita yang mempunyai riwayat abortus berisiko menderita kanker serviks sebesar 3,2 kali bila dibandingkan dengan wanita yang tidak pernah abortus. pola yang disebabkan dua faktor berlawanan yaitu perubahan pada perilaku seksual dan kecenderungan mutasi genetik dari waktu ke waktu. Data tingkat pendidikan pada penelitian ini diperoleh dari wawancara secara langsung kepada responden tentang pendidikan terakhirnya. Hasil penelitian menunjukkan bahwa tingkat pendidikan responden yang menderita kanker serviks sebagian besar yaitu SD, sedangkan untuk responden yang tidak menderita kanker serviks sebagian besar memiliki tingkat pendidikan SMA. Tingkat pendidikan memengaruhi pola pikir seseorang, sehingga semakin tinggi tingkat pendidikan seseorang maka pengetahuan seseorang untuk menerima informasi tentang kanker serviks, tanda, gejala, faktor risiko dan cara pencegahannya akan semakin baik. Wanita yang memiliki tingkat pendidikan tinggi biasanya memiliki wawasan yang luas dan cara pandang tentang kesehatan berbeda dengan yang berpendidikan rendah. Wanita dengan pendidikan tinggi lebih mudah untuk menerima informasi tentang kesehatan, sehingga mereka lebih mudah merubah perilaku kesehatannya dan lebih cepat untuk melakukan skrining penyakit yang dapat dicegah seperti kanker serviks. Hubungan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks Penelitian ini menunjukkan hasil bahwa baik responden yang menderita kanker serviks maupun tidak sebagian besar menggunakan kontrasepsi oral kombinasi < 5 tahun. Penelitian ini menunjukkan bahwa tidak ada hubungan antara penggunaan kontrasepsi oral dengan kejadian kanker serviks. Penggunaan kontrasepsi oral selama > 5 tahun tidak meningkatkan risiko kanker serviks. Penelitian case control pada 2182 pasien Rumah Sakit di Afrika Selatan yang menderita kanker serviks yang dilakukan oleh Urban et al. (2012), juga menunjukkan hasil yang berbeda dengan penelitian ini. Pada penelitian tersebut wanita yang memiliki riwayat penggunaan kontrasepsi oral mengalami peningkatan risiko kanker serviks sebesar 1,01 kali. begitu juga penelitian yang dilakukan oleh Vecchia et al. (2014), menunjukkan bahwa wanita yang memiliki riwayat penggunaan kontrasepsi oral selama > 5 tahun atau yang baru saja menggunakan kontrasepsi oral memiliki peningkatan risiko kanker serviks sebesar 2 kali lebih besar. Penelitian ini menunjukkan bahwa kelompok kasus atau responden dengan diagnosa kanker serviks sebagian besar memiliki riwayat penggunaan kontrasepsi oral kombinasi selama < 5 tahun. Hasil ini didapatkan dengan menanyakan riwayat penggunaan kontrasepsi oral kombinasi, karena semua responden yang didiagnosa kanker serviks saat ini tidak ada yang menggunakan kontrasepsi, hanya beberapa saja yang menggunakan metode kontrasepsi Metode Operasi Wanita (MOW). Saifuddin et al. (2010), membagi kontrasepsi hormonal menjadi dua yaitu kontrasepsi kombinasi dan kontrasepsi progestin. Untuk kontrasepsi kombinasi yang berisi hormon estrogen dan progesteron dibagi menjadi dua yaitu pil kombinasi dan suntikan kombinasi. Sedangkan untuk kontrasepsi progestin yang hanya berisi hormon progesterone saja dibagi menjadi beberapa macam, Diantaranya yaitu kontrasepsi suntikan progestin, kontrasepsi pil progestin (minipil), kontrasepsi implant, dan Alat Kontrasepsi Dalam Rahim (AKDR) dengan progestin. Andrews (2010), menyatakan bahwa pil kontrasepsi oral kombinasi mengandung 2 hormon yaitu estrogen dan progestogen. Pil ini bekerja dalam tiga cara yaitu menghentikan ovulasi, menebalkan mukus serviks untuk menghentikan sperma masuk uterus dan membantu mencegah terjadinya implantasi dengan mengubah endometrium. Penelitian ini sama dengan penelitian case control yang dilakukan oleh Shields et al. (2004), yang menunjukkan hasil bahwa penggunaan kontrasepsi oral selama 5–10 tahun tidak memiliki hubungan yang signiﬁ kan dengan kejadian kanker serviks, hasil analisis uji multivariat menunjukkan bahwa penggunaan kontrasepsi oral selama 5–10 tahun dengan kejadian kanker serviks didapatkan OR 1,9 (95% CI = 0,9 < OR < 4,1) sehingga angka besar risiko tersebut tidak bermakna secara epidemiologi. Sehingga dapat disimpulkan bahwa penggunaan kontrasepsi oral selama 5–10 tahun tidak meningkatkan risiko kanker serviks. Karakteristik Responden Faktor sosial ekonomi memiliki hubungan dengan stadium kanker serviks, hal ini dikarenakan kurangnya kemampuan untuk melakukan skrining Pap smear (Ibﬂ et et al., 2012). Vita Wulandari, Hubungan Faktor Risiko Penggunaan Kontrasepsi ... 437 Alliance for Cervical Cancer Prevention (2004) menjelaskan bahwa status sosial ekonomi rendah merupakan faktor risiko dari kebanyakan masalah kesehatan termasuk kanker serviks. Wanita dengan sosial ekonomi yang rendah seringkali tidak memiliki pendapatan, membatasi ke akses pelayanan kesehatan, kekurangan gizi, dan tingkat kesadaran tentang pencegahan perilaku dan persoalan kesehatan yang rendah. Semua itu merupakan faktor yang dapat membuat mereka sangat mudah diserang penyakit seperti kanker serviks yang merupakan penyakit yang dapat dicegah. tidak bermakna secara epidemiologi. Sehingga penggunaan kontrasepsi oral selama 5–8 tahun tidak meningkatkan risiko kanker serviks bila dibandingkan dengan yang menggunakan kontrasepsi oral kombinasi selama lebih dari 8 tahun. Penelitian yang lain menunjukkan hasil berbeda dengan penelitian ini. Penelitian yang dilakukan Paramita et al. (2010), menunjukkan bahwa wanita yang menggunakan kontrasepsi oral selama 5–25 tahun lebih berisiko menderita kanker serviks 4,17 kali bila dibandingkan dengan yang menggunakan kontrasepsi oral selama 1–4 tahun ataupun yang tidak pernah menggunakan kontrasepsi oral. Penelitian cross sectional yang dilakukan oleh Dewi et al. (2014), juga menunjukkan bahwa penggunaan kontrasepsi oral memiliki hubungan yang signiﬁ kan dengan stadium kanker serviks di RSUD Kota Semarang. Hubungan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks Penelitian yang dilakukan oleh Vaisy et al .(2014), di Iran juga menunjukkan hasil bahwa penggunaan kontrasepsi oral kombinasi selama 5–8 tahun tidak berhubungan dengan kejadian kanker serviks, nilai OR 2,4 (95% CI 0,54 < OR < 3,9) yang membuktikan bahwa nilai OR tersebut Pil kontrasepsi oral kombinasi mengandung dua hormon yaitu estrogen dan progestogen. Pil 438 Jurnal Berkala Epidemiologi, Vol. 4 No. 3, September 2016: 432–442 kontrasepsi oral kombinasi memiliki implikasi terhadap kanker serviks, kemungkinan karena esterogen yang terdapat dalam pil tersebut membuat ektropian pada serviks menjadi lebih luas, akibatnya terbentuk area yang lebih luas tempat metaplasia menjadi lebih rentan terhadap HPV (Andrews, 2010). Sedangkan Robboy et al (2009), menyebutkan bahwa ektropian pada serviks disebabkan oleh progesterone. Sama seperti multiparitas, penggunaan kontrasepsi oral yang lama dapat meningkatkan risiko kanker serviks melalui peningkatan kadar hormon pada tubuh wanita. Kandungan progesteron pada kontrasepsi oral mungkin menyebabkan ektropian serviks yang dapat meningkatkan paparan HPV yang diperoleh dari mekanisme trauma pada daerah serviks. Penggunaan kontrasepsi oral juga berpotensi meningkatkan risiko kanker serviks yang disebabkan oleh peningkatan paparan bahan karsinogen seperti HPV. Kemenkes RI (2010), menjelaskan istilah ektropian yaitu tampilan serviks yang terjadi akibat paparan terhadap hormon esterogen dan progesteron. Efek atau pengaruh tersebut ditunjukkan dengan bertambahnya jaringan kelenjar pada permukaan bagian luar serviks. Goldman et al (2013), setuju bahwa penggunaan kontrasepsi oral kombinasi mungkin menjadi penting dalam etiologi tumor sel serviks skuamosa invasif jika digunakan pada waktu genting dalam masa perkembangan reproduksi wanita yaitu usia < 17 tahun. Kuie (2009), juga setuju bahwa penggunaan kontrasepsi oral pada masa pubertas dapat meningkatkan risiko kanker serviks. Esterogen mempunyai dampak yang sangat besar pada serviks selama masa pubertas dan kehamilan dengan merangsang metaplasia skuamosa. Dampak yang sama pasti meningkat dalam serviks pada wanita yang menggunakan pil kontrasepsi oral kombinasi dalam waktu lama. Dalam teori, metaplasia skuamosa menimbulkan kondisi serviks mudah diserang bahan karsinogen. Sebagian besar kelompok kasus memiliki riwayat menggunakan kontrasepsi oral kombinasi selama < 5 tahun. Hal ini mungkin disebabkan oleh faktor risiko kanker serviks yang lain sehingga dapat meningkatkan risiko kanker serviks. Faktor risiko kanker serviks yang juga dapat meningkatkan risiko kanker serviks yaitu perilaku seksual berganti- ganti pasangan seksual, menikah > 2 kali, memiliki riwayat pasangan seksual suami yang pernah terinfeksi HPV, ataupun pernah melakukan hubungan seksual pertama kali pada usia < 18 tahun. Hubungan Usia Pertama Kali Berhubungan Seksual dengan Kejadian Kanker Serviks Hubungan Usia Pertama Kali Berhubungan Seksual dengan Kejadian Kanker Serviks Pada penelitian ini didapatkan hasil bahwa responden yang menderita kanker serviks sebagian besar memiliki riwayat pertama kali melakukan hubungan seksual pada usia < 18 tahun. Sedangkan responden yang tidak menderita kanker serviks sebagian besar melakukan hubungan seksual pertama kali pada usia > 18 tahun. Penelitian ini menunjukkan adanya hubungan antara usia pertama kali berhubungan seksual dengan kejadian kanker serviks. Melakukan hubungan pertama kali pada usia < 18 tahun dapat meningkatkan risiko kanker serviks. p p Semakin muda usia remaja untuk berhubungan seksual maka risiko untuk menjadi kanker serviks akan meningkat dikarenakan serviks pada usia pubertas masih belum matang sehingga serviks rentang terpapar HPV. Berhubungan seksual pada usia pubertas (kurang dari 18 tahun) membuat Squamo Columnar Junction (SCJ) bergeser sehingga dapat menimbulkan zona transformasi, yaitu zona yang terjadi akibat terjadi pergeseran SCJ asli menjadi SCJ baru. Zona transformasi tersebut merupakan tempat yang sering sekali menjadi asal displasia dan kanker. Pada awalnya remaja yang berhubungan seksual pada usia muda rentang terpapar HPV yang dapat menjadi pre kanker pada usia muda, dan akan terus berkembang dengan semakin bertambahnya usia untuk menjadi kanker. Penelitian yang dilakukan oleh Yuniar et al (2009), di Puskesmas Karanganyar Kabupaten Purbalingga juga menyebutkan bahwa terdapat hubungan antara usia pertama kali berhubungan seksual dengan kejadian kanker serviks. Pada penelitian tersebut menyebutkan bahwa usia pertama kali berhubungan seksual < 20 tahun dan > 35 tahun berisiko 14,3 kali menyebabkan kanker serviks bila dibandingkan dengan wanita yang melakukan hubungan seksual pada usia 20–35 tahun di mana nilai 95% CI = 1,747 < OR < 117,058. Kemenkes RI (2010), menjelaskan bahwa pada masa puber yang ditandai dengan meningkatnya hormon perempuan (esterogen dan progesteron), dan berlanjut sampai bertahun-tahun pada masa subur. Sel-sel kolumnar di dalam Sambungan Skuamo Kolumnar (SSK) secara bertahap digantikan oleh sel-sel skuamosa yang baru berkembang yang disebut metaplasia skuamosa yang terjadi di zona transformasi. Perubahan serviks yang tidak biasa (abnormal) seperti displasia dan kanker hampir selalu muncul di zona transformasi. Penelitian yang dilakukan oleh Makuza et al. (2015), juga menunjukkan hasil yang sama dengan penelitian ini yaitu wanita dengan usia pertama kali berhubungan seksual < 20 tahun memiliki peningkatan risiko menderita kanker serviks sebesar 1,75 bila dibandingkan dengan wanita yang melakukan hubungan seksual pertama kali pada usia > 20 tahun dengan nilai 95% CI (1,01 < OR < 3,03). Hubungan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks g Penelitian ini menunjukkan hasil bahwa penggunaan kontrasepsi oral kombinasi > 5 tahun tidak meningkatkan risiko kejadian kanker serviks bila dibandingkan dengan yang menggunakan kontrasepsi oral kombinasi selama < 5 tahun. Sebagian besar kelompok kasus atau yang menderita kanker serviks memiliki riwayat menggunakan kontrasepsi oral kombinasi selama < 5 tahun, dan memiliki riwayat berhubungan seksual pertama kali pada usia < 18 tahun dan saat ini berusia > 50 tahun atau usia menopause. Hal ini membuktikan bahwa responden dengan kanker serviks sebagian besar menikah pada usia < 18 tahun dan pernah memakai kontrasepsi oral kombinasi pada usia remaja. Kekurangan peneliti pada penelitian ini yaitu peneliti tidak membagi interval lama penggunaan kontrasepsi oral secara proporsional sehingga terjadi bias pada hasil penelitian. Selain itu peneliti tidak menggali lebih dalam lagi tentang riwayat kontrasepsi yang pernah digunakan oleh responden selain kontrasepsi oral kombinasi, peneliti hanya membatasi penelitian ini pada responden yang pernah menggunakan kontrasepsi oral kombinasi dan mengeluarkan anggota populasi yang pernah menggunakan kontrasepsi suntik kombinasi. Peneliti tidak menanyakan apakah responden pernah menggunakan metode kontrasepsi lain selain kontrasepsi oral kombinasi seperti metode kontrasepsi barrier. Metode kontrasepsi barrier diantaranya yaitu kondom, spermisida, dan diafragma. Penggunaan kontrasepsi oral menyebabkan seseorang untuk enggan memakai metode kontrasepsi barrier seperti kondom untuk mencegah penularan infeksi HPV secara langsung. Pada masa orde baru pemerintah memiliki program keluarga berencana untuk menekan laju pertumbuhan penduduk dengan cara pil kotrasepsi dan AKDR, dikarenakan masyarakat masih belum bisa menerima program MOW pada saat itu. Penggunaan kontrasepsi oral pada usia muda < 18 tahun dapat meningkatkan risiko terjadinya kanker serviks yang disebabkan oleh belum matangnya serviks sehingga mudah terpapar HPV. Pre kanker dapat terjadi akibat penggunaan kontrasepsi oral kombinasi pada usia remaja, sehingga semakin bertambah tahun pre kanker tersebut akan semakin parah dan berubah menjadi kanker. Robboy et al. (2009), juga menyatakan bahwa metode kontrasepsi barrier seperti diafragma dan kondom mungkin dapat melindungi epithelium serviks dari agent penularan seksual seperti HPV pada sperma, penemuan ini berarti metode Vita Wulandari, Hubungan Faktor Risiko Penggunaan Kontrasepsi ... 439 tahun bila dibandingkan dengan wanita yang tidak berhubungan seksual atau dibandingkan wanita yang berhubungan seksual pada usia > 22 tahun (Merrill, 2010). kontrasepsi barrier dapat menurunkan risiko kanker serviks, namun laporan penelitian ini sudah lama sebelum berkembangnya tes HPV yang valid. DNA dan RNA HPV dapat ditemukan pada plasma sperma dan pada sel sperma. Hubungan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks Rasjidi (2010) memaparkan bahwa risiko kanker serviks akan meningkat lebih dari 10 kali bila hubungan seks pertama kali dilakukan di bawah usia 15 tahun atau bila memiliki pasangan seksual lebih dari 6 orang. Berhubungan seksual dengan pria yang berisiko tinggi mengidap kondiloma akuminata akan meningkatkan risiko kanker serviks. Sel kolumnar serviks yang lebih peka terhadap metaplasia selama usia dewasa menyebabkan wanita yang berhubungan seksual sebelum usia 18 tahun akan berisiko terkena kanker serviks sampai 5 kali lipat. Hubungan Usia Pertama Kali Hamil dengan Kejadian Kanker Serviks Abortus merupakan kehamilan yang berakhir sebelum janin atau buah kehamilan tersebut mampu hidup di luar kandungan, tanpa mempersoalkan penyebabnya, selain itu biasanya abortus terjadi jika lama kehamilan < 20 minggu atau berat badan janin < 500 gram (Sastrawinata et al., 2004). Abortus memiliki dampak bagi kesehatan dan keselamatan hidup wanita. Infeksi pada daerah rahim dan serviks dapat terjadi akibat proses abortus. Hasil penelitian ini menunjukkan baik kelompok kasus maupun kontrol sebagian besar hamil pertama kali pada usia > 18 tahun. Penelitian ini menunjukkan adanya hubungan antara usia pertama kali hamil dengan kejadian kanker serviks. Riwayat pertama kali hamil pada usia < 18 tahun dapat meningkatkan risiko kanker serviks. Hasil penelitian ini sama dengan penelitian yang dilakukan oleh Makuza et al. (2015), di Rwanda yang menunjukkan hasil bahwa usia pertama kali hamil yaitu < 20 tahun berisiko menderita kanker serviks sebesar 2,1 bila dibandingkan dengan wanita yang memiliki riwayat usia pertama kali hamil pada usia > 20 tahun dengan nilai 95% CI (1,20 < OR < 3,67). Hasil penelitian ini menunjukkan bahwa baik responden yang menderita kanker serviks maupun tidak menderita kanker serviks sebagian besar tidak pernah memiliki riwayat abortus. Penelitian ini menunjukkan adanya hubungan antara riwayat abortus dengan kejadian kanker serviks. Riwayat abortus seorang wanita dapat meningkatkan risiko kanker serviks. Wanita yang pertama kali hamil pada usia pubertas atau < 18 tahun berhubungan dengan serviks yang belum matang, selain itu kehamilan berhubungan dengan penurunan imunitas saat hamil sehingga wanita hamil usia muda lebih muda terpapar virus HPV dibandingkan mereka yang hamil pertama kali pada usia dewasa. Wanita yang pertama kali hamil pada usia < 18 tahun menunjukkan bahwa wanita tersebut melakukan hubungan seksual pertama kali pada usia yang lebih muda daripada usia sewaktu melakukan hubungan seksual pertama kali. Sedangkan berhubungan seksual pada usia pubertas atau < 18 tahun memicu timbulnya zona transformasi, yaitu zona yang terbentuk dari pergesaran Squamo Columnar Junction (SCJ) asli menuju SCJ baru. Zona transformasi sangat jelas terlihat pada wanita yang mengalami displasia. Tipe proses persalinan juga memiliki hubungan dengan kanker serviks. Wanita yang pernah melahirkan dengan section caesaria (SC) tidak berbeda dengan wanita yang tidak pernah mengalami persalinan. Melahirkan secara pervaginam memiliki risiko kanker serviks sebesar 2,6 kali. Wanita yang pernah mengalami persalinan secara pervaginam dan juga section caesaria memiliki risiko sebesar 2,2 kali. Hubungan Usia Pertama Kali Berhubungan Seksual dengan Kejadian Kanker Serviks Penelitian pada wanita kota di Indian menyebutkan bahwa 84% pasien kanker serviks berhubungan seksual pertama kali sebelum usia 16 tahun, dan risiko kanker serviks ini meningkat secara signifikan dengan semakin muda usia pertama kali berhubungan seksual. Bagitu juga penelitian case control yang lain menyebutkan bahwa terdapat peningkatan risiko kanker serviks sebesar 2,4 kali (95% CI = 1,1 < OR < 5,3) pada wanita yang berhubungan seksual sebelum usia 18 Goldman et al (2013), memaparkan bahwa usia yang lebih muda saat pertama kali berhubungan seksual diketahui merupakan faktor risiko kanker serviks. Penelitian pada 45.000 wanita menyatakan bahwa risiko kanker serviks meningkat di antara wanita yang berumur < 24 tahun saat berhubungan seksual pertama kali, dan risiko meningkat dengan semakin mudanya usia saat berhubungan seksual di bawah 17 tahun. Umur yang lebih muda saat berhubungan seksual pertama kali menunjukkan 440 Jurnal Berkala Epidemiologi, Vol. 4 No. 3, September 2016: 432–442 peningkatan lamanya HPV yang terus menerus, dan juga sebagai faktor risiko dari prevalensi HPV. hamil secara proporsional. Peneliti hanya membagi berdasarkan usia pertama kali hamil < 18 tahun dan > 18 tahun, sehingga pembagian kriteria usia pertama kali hamil ini dapat menyebabkan bias pada hasil penelitian. peningkatan lamanya HPV yang terus menerus, dan juga sebagai faktor risiko dari prevalensi HPV. Kekurangan peneliti pada penelitian ini yaitu peneliti tidak membagi interval usia pertama kali berhubungan seksual secara proporsional, sehingga hal ini dapat menyebabkan bias pada hasil penelitian. j g g p Kekurangan peneliti pada penelitian ini yaitu peneliti tidak membagi interval usia pertama kali berhubungan seksual secara proporsional, sehingga hal ini dapat menyebabkan bias pada hasil penelitian. Hubungan Usia Pertama Kali Hamil dengan Kejadian Kanker Serviks Sebaliknya, terdapat hubungan yang signiﬁ kan antara riwayat abortus sebanyak > 2 kali dengan kanker serviks (tanpa melihat secara induksi atau spontan) dengan OR sebesar 0,6 bila dibandingkan dengan wanita yang tidak pernah mengalami persalinan, mengalami persalinan secara section caesaria ataupun keduanya (Guttmacher Institute, 2002). Walaupun usia mens pertama kali dan menopause tidak berpengaruh pada risiko kanker serviks, memiliki riwayat hamil pertama kali di usia muda dan jumlah kehamilan dapat meningkatkan risiko kanker serviks, manajemen persalinan yang tidak tepat juga dapat meningkatkan risiko kanker serviks (Rasjidi, 2010). Penelitian yang dilakukan pada 17.047 wanita di Taipei, Taiwan pusat pelayanan perencanaan keluarga pada tahun 1991–1992 sebanyak 55% memiliki hasil Pap smear normal, 44% ditemukan hasil Pap smear tidak normal dan hanya 0,9% yang menderita displasia. Tren hasil pap smear positif ditemukan di antara wanita yang memiliki peningkatan frekuensi abortus secara induksi dan insiden prekanker serviks (P < 0,01) (Dong, et al., 1995). Abortus dapat terjadi secara Kekurangan peneliti pada penelitian ini yaitu sama dengan variabel usia pertama kali berhubungan seksual, peneliti tidak membagi usia pertama kali Vita Wulandari, Hubungan Faktor Risiko Penggunaan Kontrasepsi ... 441 kanker serviks, tanda, gejala, faktor risiko dan cara pencegahannya. Selain itu responden diharapkan lebih memperhatikan kesehatan reproduksinya, khususnya untuk responden yang tidak menderita kanker serviks namun memiliki faktor risiko diharapkan untuk selalu melakukan skrining secara rutin setiap tahunnya atau sesuai rujukan dokter. Untuk responden yang menderita kanker serviks diharapkan dapat meningkatkan kualitas hidup dengan treatment dan olahraga secara teratur serta mengonsumsi makanan bergizi agar kondisinya tidak semakin memburuk. spontan maupun secara induksi. Abortus dengan induksi memiliki hubungan dengan peningkatan risiko kanker serviks dikarenakan saat melakukan abortus terjadi perlukaan rahim untuk membersihkan sisa hasil konsepsi. Namun yang terjadi tidak hanya perlukaan rahim, perlukaan pada serviks juga bisa terjadi sehingga semakin sering wanita mengalami peningkatan frekuensi abortus secara induksi maka risiko untuk menderita kanker serviks akan semakin meningkat bila dibandingkan dengan wanita yang tidak memiliki riwayat abortus. Simpulan Alliance for Cervical Cancer Prevention. 2004. Risk Factors for Cervical Cancer: Evidence to Date, Washington: Alliance for Cervical Cancer Prevention. Penggunaan kontrasepsi oral pada pasien di RSUD Dr. Saiful Anwar Malang tidak memiliki hubungan yang signiﬁ kan dengan kejadian kanker serviks. Aktivitas seksual yang diteliti pada penelitian ini meliputi usia pertama kali berhubungan seksual, usia pertama kali hamil, dan riwayat abortus. Usia pertama kali berhubungan seksual, usia pertama kali hamil dan riwayat abortus memiliki hubungan yang signiﬁ kan dengan kejadian kanker serviks. Andrews, G. 2010. Women’s Sexual Health. 2nd Edition ed. Jakarta: EGC. Dewi, N.K., Rejeki, S. & Istiana, S. 2014. Hubungan Lama Penggunaan Kontrasepsi Oral pada Wanita Usia Lebih dari 35 Tahun dengan Stadium Kanker Serviks di RSUD Kota Semarang. Jurnal Kebidanan Unimus. 4(1), pp. 31–38. Saran Muhimpundu, M.A., Pace, L.E., Ntaganira, J. & Riedel, D.J. 2015. Prevalence and Risk Factors for Cervical Cancer abd Pre-cancerous lesions in Rwanda. The Pan African Medical Journal, Volume 23, pp. 22–26. Surya. 2015. Jatim Darurat Kanker Serviks, Ini Tujuh Daerah di Jatim Yang Masuk Peta Waspada Kanker Serviks. [Online] Available at: http:// surabaya.tribunnews.com/2015/03/09/ini-tujuh- daerah-di-jatim-yang-masuk-peta-waspada- kanker-serviks [Accessed 7 Juli 2015]. Merrill, R.M. 2010. Reproductive Epidemiology Principle and Methods. Canada: Jones and Bartlett Publishers. Urban, M., Banks, E., Egger, S., Canfell, K., O’Connell, D., Beral, V. & Sitas, F. 2012. Injectable and Oral Contraceptive Use and Cancers of the Breast, Cervix, Ovary, and Endometrium in Black South African Women: Case–Control Study. PLos Med, 9(3), pp. 1–11. Paramita, S., Soetomo, S., Widodo, M.A. & Sumitro, S.B. 2010. High Parity and Hormonal Contraception Use as Risk Factors for Cervical Cancer in East Kalimantan. Medical Journal of Indonesia, 19(4), pp. 268–72. Pusat Data dan Informasi. 2014. Situasi dan Analisis Keluarga Berencana, Jakarta Selatan: s.n. Vaisy, A., Lotﬁ nejad, S. & Zhian, F. 2014. Risk of Cancer with Combined Oral Contraceptive Use among. Asian Pac J Cancer Prev, 15(14), pp. 5517–5522. Rasjidi, I. 2010. 100 Questions & Answer Kanker pada Wanita. Jakarta: PT Elex Media Komputindo. Republika. 2013. Ratusan Perempuan di Malang Menderita Kanker Serviks. [Online] Available at: http://www.republika.co.id/berita/nasional/ jawa-timur/13/12/22/my7tcq-ratusan-perempuan- di-malang-menderita-kanker-serviks [Accessed 14 Maret 2015]. Vecchia, L., Carlo, Boccia & Stefania. 2014. Oral Contraceptives, Human Papillomavirus and Cervical Cancer. European Journal of Cancer Prevention, 23(2), pp. 110–112. WHO. 2014. Comprehensive Cervical Cancer Control: a Guide to Essential Practice Second Edition. Switzerland: WHO Press. Robboy, S.J., Mutter, G.L., Prat, J., Bentley, R.C., Russell, P. & Anderson, M.C. 2009. Pathology of the Female Reproductive Tract. 2nd ed. British: Churchill Livingstone Elsevier. Winship Cancer Institute of Emory University. 2014. Cervical Cancer: Risk Factors. [Online] Available at: http://www.cancerquest.org/cervical-cancer- risks.html [Accessed 22 Maret 2015]. Saifuddin, A.B. Affandi, B., Baharuddin, M. & Soekir, S., 2010. Buku Panduan Praktis Pelayanan Kontrasepsi. 2 ed. Jakarta: PT. Bina Pustaka Sarwono Prawirohardjo. Wisconsin Cancer Data Bulletin. 2014. Cervical Cancer in Wisconsin. Wisconsin: Wisconsin Cancer Reporting System. Wulandari, V. 2016. Hubungan Paritas, Usia Pertama Kali Berhubungan Seksual dan Penggunaan Kontrasepsi Oral dengan Kejadian Kanker Serviks (Studi di Poli Obstetri dan Ginekologi RSUD Dr. Saiful Anwar Malang). Skripsi. Surabaya: Universitas Airlangga. Sastrawinata, S., Martaadisoebrata, D. & Wirakusumah, F. 2004. Ilmu Kesehatan Reproduksi: Obstetri Patologi. 2 ed. Jakarta: EGC. Satyarini, S. 2011. Saran Dong, P.D., Lin, R.S. & Royal, J. 1995. Induced Abortion in Taiwan. Social Health, 115(2) pp. 100–108. Bagi tempat penelitian diharapkan hasil penelitian ini dapat dijadikan bahan informasi penting bagi pihak PKRS RSUD Dr. Saiful Anwar Malang dalam upaya penyebaran informasi tentang kanker serviks, selain itu sebagai upaya promotif dan preventif untuk menurunkan kejadian kanker serviks. Goldman, M. B., Troisi, R. & Rexrode, K. M. 2013. Women & Health 2nd Edition. London: Academic Press. Guttmacher Institute. 2002. Long-Term Pill Use, High Parity Raise Cervical Cancer Risk Among Women with Human Papillomavirus Infection. International Family Planning Perspectives, 28(3), pp. 176–181. Bagi peneliti selanjutnya diharapkan hasil penelitian ini dapat dijadikan sebagai bahan rujukan penelitian epidemiologi tentang kanker serviks. Selain itu diharapkan peneliti selanjutnya lebih banyak membaca referensi agar penelitian selanjutnya lebih baik lagi, dan tidak membatasi penelitian tersebut hanya pada faktor risiko yang sama pada penelitian ini. Diharapkan peneliti selanjutnya lebih memperdalam pada saat menggali informasi kepada responden seperti metode melahirkan secara normal atau section caesaria, riwayat penggunaan metode kontrasepsi lain selain kontrasepsi oral kombinasi untuk mencegah penularan HPV secara langsung seperti metode kontrasepsi barrier. Hal ini dimaksudkan agar tidak terjadi bias dalam penelitian selanjutnya. ( ) pp Ibﬂ et, E., Kjaer, S. K., Johansen, C., Hogdal, C., Jessen, M.S., Frederiksen, K., Frederiksen, B.L., Osler, M. & Dalton, S.O. 2012. Socioeconomic Position and Stage of Cervical Cancer in Danish Woman Diagnosed 2005 to 2009. American Association for Cancer Research, Volume 21, pp. 835–842. Kemenkes RI. 2010. Buku Acuan Pencegahan Kanker Leher Rahim dan Kanker Payudara. Jakarta: Dinas Kesehatan Provinsi Jawa Timur Seksi P3PMK Tahun Anggaran 2012. Kuie, T.S. 2009. Cervical Cancer its Causes and Prevention. Singapore: Marshall Cavendish Editions. Hasil penelitian ini diharapkan dapat dijadikan bahan informasi tambahan untuk responden tentang 442 Jurnal Berkala Epidemiologi, Vol. 4 No. 3, September 2016: 432–442 A Case-Control Study of Risk Factors for Invasive Cervical Cancer among U.S. Women Exposed to Oncogenic Types of Human Papillomavirus. Cancer Epidemiol Biomarker, 13(10), pp. 1574–1582. A Case-Control Study of Risk Factors for Invasive Cervical Cancer among U.S. Women Exposed to Oncogenic Types of Human Papillomavirus. Cancer Epidemiol Biomarker, 13(10), pp. 1574–1582. Lestari, S. 2012. Hubungan Tingkat Pendidikan, Pengetahuan dan Sikap Ibu Rumah Tangga dengan Perilaku Deteksi Dini Kanker Serviks Metode IVA di Puskesmas Jaten II Kabupaten Karanganyar. Thesis. Surakarta: Universitas Sebelas Maret. Supranto, J. 2007. Statistik untuk Pemimpin Berwawasan Global. 2 ed. Jakarta: Salemba Empat. Makuza, J.D., Nsanzimana, S. Saran Faktor Risiko Kanker Serviks di Rumah Sakit Umum Pemerintah Dr. Wahidin Sudirohusodo Makassar, Sulawesi Selatan. Jurnal Kesehatan Masyarakat Nasional, 5(6), pp. 283–288. Yuniar, I., Saryono & Rohani, F. 2009. Faktor-faktor yang memengaruhi kejadian kanker serviks di Puskesmas Karanganyar. Jurnal Ilmiah Kesehatan Keperawatan, 5(2) pp. 109–118. Shields, T.S., Brinton, L.A., Burk, R.D., Wang, S.S., Weinstein, S.J., Ziegler, R.G., Studentsov, Y.Y., McAdams, M. & Schiffman, M. 2004.
https://openalex.org/W1981180756	https://europepmc.org/articles/pmc4044527?pdf=render	English	null	Unsupervised free-breathing 3-dimensional imaging of morphology, function and flow in congenital heart disease under 30 minutes: pilot study	Journal of cardiovascular magnetic resonance	2,014	cc-by	1,243	© 2014 Krishnamurthy 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. 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. Background Cardiac MRI for congenital heart disease (CHD) is an operator dependent and time-intensive examination requiring real-time decision making regarding choice of sequences, planes, and acquisition parameters to adapt to unique morphological and functional variables in a given patient. Krishnamurthy et al. Journal of Cardiovascular Magnetic Resonance 2014, 16(Suppl 1):W8 http://www.jcmr-online.com/content/16/S1/W8 Krishnamurthy et al. Journal of Cardiovascular Magnetic Resonance 2014, 16(Suppl 1):W8 http://www.jcmr-online.com/content/16/S1/W8 Krishnamurthy et al. Journal of Cardiovascular Magnetic Resonance 2014, 16(Suppl 1):W8 http://www.jcmr-online.com/content/16/S1/W8 Results All FB 3D sequences were technically feasible in all 5 patients. Average time for completion of 5 FB 3D sequences was 29 minutes. Average score for first-pass MRA was 1.9/5. Average score for equilibrium MRA was 1.3/5. Clinical scores for 2D SSFP were consistently better than 3D-SSFP, but 3D SSFP images were ade- quate for recognition of pathology in all cases (2D vs 3D: 1.5 ± 0.5 vs 1.6 ± 0.9) and had better inter-slice alignment (1.4 ± 0.5 vs 1 ± 0). Average percentage dif- ference between 2D and 3D cine SSFP volumetric data is shown in table 3, and Figure 2. Comparative flow ana- lysis between 2D PC and 4D PC data revealed broad correlation (Figure 2, table 3) though the stroke volume, forward and backward flows through the aorta were not statistically different (p > 0.35; paired Student’s t-test) To evaluate technical feasibility, image quality and quanti- tative integrity of a free-breathing (FB) protocol following administration of blood pool contrast agent, utilizing 3-dimensional (3D) imaging of morphology, function, and flow without physician supervision in a cohort of patients with CHD. Methods Five patients with CHD were included in this pilot study (table 2 in Figure 2). The FB MR studies were per- formed on a Philips Acheiva 1.5T magnet using a 5-channel phased array coil (see Table 1 in Figure 1) 1. Respiratory synchronized [1], time-resolved MRA 2. Equilibrium phase MRA 3. 3D cine SSFP 4.4D phase contrast (PC) flow imaging 5.3D whole-heart single phase SSFP (coronary) Comparative data was obtained using conventional 2D cine RT SSFP sequences [2] in the VLA, 4 chamber and short axis planes, and 2D PC ima- ging. Data Analysis: Image quality assessment and quan- titative volumetric and flow analysis were performed by three blinded, experienced users. MRA images were graded using a semi-quantitative scale from 1-5 for Rajesh Krishnamurthy1, Ramkumar Krishnamurthy1, Elijah Bolin2, LaDonna Malone1, Myriam E Almeida-Jones1, Amol Pednekar3 From 17th Annual SCMR Scientific Sessions New Orleans, LA, USA. 16-19 January 2014 relevant imaging targets in CHD [1], with 1: excellent, no limitations, and 5: non-diagnostic. The clinical scoring system for 2D and 3D cine SSFP was based on blood- myocardial contrast, endocardial edge definition and inter-slice alignment [2]. Paired t-test analysis was per- formed on LV and RV volumes obtained by an experi- enced observer using the same software 1Radiology, Texas Children’s Hospital, Houston, Texas, USA Full list of author information is available at the end of the article Unsupervised free-breathing 3-dimensional imaging of morphology, function and flow in congenital heart disease under 30 minutes: pilot study Rajesh Krishnamurthy1, Ramkumar Krishnamurthy1, Elijah Bolin2, LaDonna Malone1, Myriam E Almeida-Jones1, Amol Pednekar3 Conclusions The free breathing first pass MRA, equilibrium MRA, 3D cine SSFP, and 3D single-phase SSFP exhibit significant clinical utility. We demonstrate the feasibility of perform- ing an observer independent comprehensive CMR in CHD utilizing FB 3D acquisitions for morphology, func- tion and flow within 30 minutes using a 5-channel phased-array coil. Better acquisition hardware (eg., 32 ch coil) will lead to superior image quality. 1Radiology, Texas Children’s Hospital, Houston, Texas, USA Full list of author information is available at the end of the article Krishnamurthy et al. Journal of Cardiovascular Magnetic Resonance 2014, 16(Suppl 1):W8 http://www.jcmr-online.com/content/16/S1/W8 Page 2 of 3 Figure 1 Representative images of patients acquired using the FB 3D protocol. First pass MRA 3D MIP acquired immediately after administration of blood pool contrast agent is shown in (A). (B) and (C) are equilibrium MRA images acquired ~2-3 minutes after contrast injection. Additional vasculature is clearly see in equilibrium MRA wrt first pass MRA images. (D) is a 2D SSFP 4-chamber cine image; (E) and (F) are reconstructed 3D images obtained in a similar imaging plan. Regurgitant jet is clearly seen in (F) that could not be clearly visualized using 2D acquisition. (G), (H) and (I) demonstrates feasibility of capturing complex anatomic details/flow (pulmonary stenosis) using 4D flow imaging. Whole heart SSFP imaging (not shown) post contrast also demonstrated significant clinical utility. pp http://www.jcmr-online.com/content/16/S1/W8 Figure 1 Representative images of patients acquired using the FB 3D protocol. First pass MRA 3D MIP acquired immediately after administration of blood pool contrast agent is shown in (A). (B) and (C) are equilibrium MRA images acquired ~2-3 minutes after contrast injection. Additional vasculature is clearly see in equilibrium MRA wrt first pass MRA images. (D) is a 2D SSFP 4-chamber cine image; (E) and (F) are reconstructed 3D images obtained in a similar imaging plan. Regurgitant jet is clearly seen in (F) that could not be clearly visualized using 2D acquisition. (G), (H) and (I) demonstrates feasibility of capturing complex anatomic details/flow (pulmonary stenosis) using 4D flow imaging. Whole heart SSFP imaging (not shown) post contrast also demonstrated significant clinical utility. Figure 1 Representative images of patients acquired using the FB 3D protocol. First pass MRA 3D MIP acquired immediately after administration of blood pool contrast agent is shown in (A). (B) and (C) are equilibrium MRA images acquired ~2-3 minutes after contrast injection. Conclusions Additional vasculature is clearly see in equilibrium MRA wrt first pass MRA images. (D) is a 2D SSFP 4-chamber cine image; (E) and (F) are reconstructed 3D images obtained in a similar imaging plan. Regurgitant jet is clearly seen in (F) that could not be clearly visualized using 2D acquisition. (G), (H) and (I) demonstrates feasibility of capturing complex anatomic details/flow (pulmonary stenosis) using 4D flow imaging. Whole heart SSFP imaging (not shown) post contrast also demonstrated significant clinical utility. Page 3 of 3 Funding None. Authors’ details 1Radiology, Texas Children’s Hospital, Houston, Texas, USA. 2Radiology, Baylor College of Medicine, Houston, Texas, USA. 3Clinical Science, Philips Healthcare, Houston, Texas, USA. References 1. JCMR 2010, 12(Suppl 1):O31. 2. JCMR 2013, 15(Suppl 1):O98. doi:10.1186/1532-429X-16-S1-W8 Cite this article as: Krishnamurthy et al.: Uns 3-dimensional imaging of morphology, funct heart disease under 30 minutes: pilot study. J Magnetic Resonance 2014 16(Suppl 1):W8. Figure 2 dren’s Hospital Houston Texas USA 2Radiology Baylor References 1. JCMR 2010, 12(Suppl 1):O31. 2. JCMR 2013, 15(Suppl 1):O98. doi:10.1186/1532-429X-16-S1-W8 Cite this article as: Krishnamurthy et al.: Unsupervised free-breathing 3-dimensional imaging of morphology, function and flow in congenital Funding None. Authors’ details 1Radiology, Texas Children’s Hospital, Houston, Texas, USA. 2Radiology, Baylor College of Medicine, Houston, Texas, USA. 3Clinical Science, Philips Healthcare, Houston, Texas, USA. P bli h d 16 J 2014 References 1. JCMR 2010, 12(Suppl 1):O31. 2. JCMR 2013, 15(Suppl 1):O98. doi:10.1186/1532-429X-16-S1-W8 Cite this article as: Krishnamurthy et al.: Unsupervised free-breathing 3-dimensional imaging of morphology, function and flow in congenital heart disease under 30 minutes: pilot study. Journal of Cardiovascular Magnetic Resonance 2014 16(Suppl 1):W8. Funding Published: 16 January 2014
https://openalex.org/W1862972370	https://figshare.com/articles/thesis/Wandering_wives_or_foreign_fillies_The_women_of_archaic_Greek_colonisation/17006014/1/files/31459189.pdf	English	null	Wandering wives or foreign fillies? The women of archaic Greek colonisation	null	2,021	cc-by	52,411	THE WOMEN OF ARCHAIC GREEK COLONISATION By Harriet Kerr By Harriet Kerr Jacques-Louis David, ‘The Intervention of the Sabine Women’, 1799, oil on canvas, 385 x 522 cm, Musée du Louvre. Jacques-Louis David, ‘The Intervention of the Sabine Women’, 1799, oil on canvas, 385 x 522 cm, Musée du Louvre. A thesis submitted to Victoria University of Wellington in fulfilment of the requirements for the degree of Master of Arts in Classics Victoria University of Wellington 2013 A thesis submitted to Victoria University of Wellington in fulfilment of the requirements for the degree of Master of Arts in Classics 2013 Cover image sourced: ARTstor Slide Gallery: ARTSTOR_103_41822001085339, http://library.artstor.org/library/secure/ViewImages?id=8CJGczI9NzldLS1WEDhzTnkrX3gqdVB6 eSw%3D&userId=gDhPcjIh&zoomparams, accessed 11/02/2013 Cover image sourced: ARTstor Slide Gallery: ARTSTOR_103_41822001085339, http://library.artstor.org/library/secure/ViewImages?id=8CJGczI9NzldLS1WEDhzTnkrX3gqdVB6 eSw%3D&userId=gDhPcjIh&zoomparams, accessed 11/02/2013 1 1 ABSTRACT Greek colonisation in the archaic period encompassed an enormous geographical area. But for all its prevalence, the textual evidence is limited in both quantity and quality and the archaeological evidence goes only some way towards helping decipher social change and ethnicity. These issues become even more apparent when considering the position of women in the new city foundations. Did Greek colonists take their own wives with them to their new homes? Were Greek women sent out at a later date once the colony had become established? Did Greek colonists intermarry with indigenous women on arrival? Or did something else happen, including a mix of these options? The weight of scholarly opinion currently falls in favour of intermarriage, though frequently little evidence is proffered to support this view. This thesis focuses on this hypothesis and examines the evidence (or lack thereof) to support this conclusion. Chapter One examines the problems associated with archaic Greek colonisation generally, particularly those issues connected with the ‘language of colonisation’. The study of Greek colonisation has been complicated by imprecise and ambiguous terminology, which frequently draws comparison with more modern (although altogether different) instances of the phenomenon. A major repercussion of this is the tendency to overlook both women and any indigenous peoples. The opening chapter also examines the various reasons behind the foundation of colonies, as well as the different types of settlements, so that an assessment can be made as to whether Greek women might have been more likely to accompany colonising expeditions in some instances over others. Chapter Two looks at the concept of intermarriage more closely and assesses Greek attitudes towards foreign women. It also evaluates the evidence typically called upon by scholars to argue for and against intermarriage in Greek colonisation. Chapter Three assesses the evidence for the presence of women in ten different colonies. Presented roughly in chronological order, these colonies were selected for their geographical scope, covering different regions from the Western Mediterranean, Magna Graecia, North Africa, and the Black Sea. This discussion explores both the literary and archaeological evidence (where possible) for each of these colonies and assesses the potential for intermarriage. This thesis demonstrates that broad conclusions about intermarriage as a widespread practice are unsustainable and concludes that colonisation in the archaic period cannot be considered a uniform phenomenon. 2 TABLE OF CONTENTS TABLE OF CONTENTS Acknowledgements ...................................................................................................... 4 Abbreviations ............................................................................................................... 5 Introduction ................................................................................................................. ABSTRACT 6 Chapter One – The Prolegomena: Women and Greek Colonisation ................... 10 Introduction .......................................................................................................................... 10 Practical considerations ........................................................................................................ 16 Reasons for founding colonies ............................................................................................. 20 Different types of settlement ................................................................................................ 23 Conclusion ............................................................................................................................ 29 Chapter Two – Intermarriage: A Normal Practice? .............................................. 31 ‘Intermarriage’ as a concept ................................................................................................. 31 Intermarriage versus taking Greek wives ............................................................................. 34 Conclusion ............................................................................................................................ 38 Chapter Three – A Survey of Selected Greek Colonies: Greek Wives or Indigenous Women? .................................................................................................. 39 Introduction .......................................................................................................................... 39 Western Italy ........................................................................................................................ 39 Pithekoussai ..................................................................................................................... 39 Sicily ..................................................................................................................................... 45 Syracuse ........................................................................................................................... 45 Megara Hyblaea ............................................................................................................... 50 Morgantina ....................................................................................................................... 57 Southern Italy ....................................................................................................................... 62 Taras ................................................................................................................................. 62 Locri Epizephyrii ............................................................................................................. 66 Metapontum ..................................................................................................................... 68 North Africa ......................................................................................................................... 71 Cyrene .............................................................................................................................. 71 Western Mediterranean ........................................................................................................ 81 Massalia ........................................................................................................................... 81 Black Sea .............................................................................................................................. 86 Conclusion .................................................................................................................. 92 Appendix 1 – Figures ................................................................................................. 99 Bibliography ............................................................................................................. 107 3 ACKNOWLEDGEMENTS It’s a well-recognised fact that writing a thesis is a challenging task. A few people deserve special mention here for their help and support at various stages along the way. It’s a well-recognised fact that writing a thesis is a challenging task. A few people deserve special mention here for their help and support at various stages along the way. I’m grateful for the guidance and expertise of my supervisor, Professor Matthew Trundle. Thanks to the Classics Department at Victoria University of Wellington, for their support, feedback, and encouragement: in particular, Dr Diana Burton, for going above and beyond, and dedicating hours of her time to help with Greek; Dr Judy Deuling, for her help with fibulae; Chris de L’isle, for always having or finding answers to my (almost) endless conundrums; and Alex Wilson, for help on all things technological. I’m also grateful for the advice, assistance, and moral support of Emily Simons, Geoff Ardell, Jess Casbolt, Dan Knox, and Dan Diggins. Thanks to Victoria University of Wellington for providing financial support in the form of a Master’s Scholarship, and a research grant to attend ASCS 34 in Sydney. Finally, thanks to my parents, Jenny and Max, not only for their years of proofreading, but also for their unwavering support and encouragement. 4 4 ( ) 2 In this thesis, I spell all forms of the verb ‘to colonise’ with New Zealand English spelling, according to the Oxford New Zealand Dictionary. The only exceptions to this are in quotations where I keep any original spelling. 3 3 Throughout this thesis, I will use the term ‘indigenous’ as the lesser of evils, but I do not wish to imply in doing so that any people described as indigenous necessarily had permanent habitation. It is extremely difficult to describe any population already living on, around, or inhabiting the land that the Greek colonists chose to settle on. All the terminology is problematic. Whitehouse and Wilkins (1989) 115 discuss the issues involved. ‘Local’ is too imprecise (and I would add that it does not include nomadic peoples); ‘indigenous’ implies that the peoples have had permanent habitation since the beginning of time; ‘native’ has modern day colonial overtones (an issue that will be discussed in depth in Chapter One). ABBREVIATIONS CPS: Rosenbaum, E. (1960) A Catalogue of Cyrenaican Portrait Sculpture, London: Oxford University Press. IG: Lewis, D., Jeffery, L., Erxleben, E. & Hallof, K. (eds) (1981, 1994, 1998) Inscriptiones Graecae I: Inscriptiones Atticae Euclidis anno anteriores, Berlin: Berlin-Brandenburgische Akademie der Wissenschaften, 3rd ed. LSJ: Liddell, H. G. & Scott, R. (1968) A Greek-English Lexicon, Oxford: Claredon Press, 9th ed, revised by Henry Stuart Jones. ML: Meiggs, R. & Lewis, D. (eds) (1969) Selection of Greek Historical Inscriptions to the end of the Fifth Century BC, Oxford: Claredon. OCD: Hornblower, S. & Spawforth, A. (2012) The Oxford Classical Dictionary, Oxford: Oxford University Press, 4th ed. (unless otherwise specified as being a different edition). OGIS: Dittenberger, W. (1903-05) Orientis Graeci Inscriptiones Selectae, Leipzig: Herzel. OLD: Glare, P. G. W. (1968) Oxford Latin Dictionary, Oxford: Oxford University Press. SEG: Supplementum Epigraphicum Graecum. SEG: Supplementum Epigraphicum Graecum. SEG: Supplementum Epigraphicum Graecum. 5 5 1 Graham (1981-82) 293. p p ) 4 The term was coined by Dunbabin (1948), and is a term continuously used to describe the Greeks living in Italy and Sicily (see for example, David Ridgway’s (1992) publication The First Western Greeks). 1 Graham (1981-82) 293. 2 In this thesis, I spell all forms of the verb ‘to colonise’ with New Zealand English spelling, according to the Oxford New Zealand Dictionary. The only exceptions to this are in quotations where I keep any original spelling. 3 Throughout this thesis, I will use the term ‘indigenous’ as the lesser of evils, but I do not wish to imply in doing so that any people described as indigenous necessarily had permanent habitation. It is extremely difficult to describe any population already living on, around, or inhabiting the land that the Greek colonists chose to settle on. All the terminology is problematic. Whitehouse and Wilkins (1989) 115 discuss the issues involved. ‘Local’ is too imprecise (and I would add that it does not include nomadic peoples); ‘indigenous’ implies that the peoples have had permanent habitation since the beginning of time; ‘native’ has modern day colonial overtones (an issue that will be discussed in depth in Chapter One). 4 The term was coined by Dunbabin (1948), and is a term continuously used to describe the Greeks living in Italy and Sicily (see for example, David Ridgway’s (1992) publication The First Western Greeks). 1 Graham (1981-82) 293. INTRODUCTION “The historian who tries to recover the facts about Greek colonization in the archaic period is confronted by many problems.”1 So wrote A. J. Graham, the father of Greek colonisation studies,2 in 1981-82. These problems include a limited number of written sources and unreliable, sometimes contradictory, archaeological evidence. The difficulties increase, however, when the position of women in these colonies is considered. Did Greek colonists take their own wives with them to their new homes? Were Greek women sent out at a later date once the colony had become established? Did Greek colonists intermarry with indigenous women on arrival?3 Or did something else happen, including a mix of these options? In the second half of the twentieth century and continuing into the twenty-first, the weight of the scholarly opinion supports the view that only men founded colonies and therefore intermarried after arrival. This thesis focuses on this hypothesis and examines the evidence (or lack thereof) to support this conclusion. The chronological and geographical scope of archaic Greek colonisation is often unspecific and accordingly there is little consensus among scholars about it. This thesis encompasses the period of the eighth to sixth centuries BC. Geographically, Greek colonisation reached almost to the end of the known world. Previous studies of Greek colonisation often limit their examinations to a single region, most commonly the ‘Western Greeks’ of Magna Graecia.4 For a more rounded discussion, this thesis endeavours to examine colonies ranging from the Western Mediterranean to the Black Sea (see Figure 1). The importance of these archaic 6 colonies is emphasised by the fact that by the fourth century BC, they may have accounted for some 40% of all Greeks.5 This thesis employs a mixture of literary and archaeological evidence. Much of the textual evidence has a clear grounding in myth, making its historicity questionable.6 Some accounts are clearly aetiological ones, created to explain an earlier event or practice. Further, this literary evidence comes from a period (or rather, different periods) considerably after the period of initial colonisation. This factor has implications for our interpretation of the evidence. A second or third- hand account will never be as accurate as a first-hand one. Later accounts are frequently influenced by popular thought of their time, or shaped to fit contemporary political intentions. Women do not feature prominently in the textual evidence, but the absence of women from the Greek sources is deceptive. g ( ) ; ( ) 6 The distinction between ‘myth’ and ‘reality’ was not necessarily recognised by the Greeks. See Dougherty (1993a) 15. 5 De Angelis (2011) 18; Scheidel (2003) 134-135. 6 g y ( ) 7 Dewald (1980) 12. Obviously, not all of these mentions concern women in archaic Greek colonies. 8 Dougherty (1993a) 15. 7 Dewald (1980) 12. Obviously, not all of these mentions concern women in archaic Greek colonies. 8 Wiedemann (1983) 165. colonies. 8 Wiedemann (1983) 165. 8 Wiedemann (1983) 165. INTRODUCTION It would be easy to assume that their absence from sources was due to their physical absence during the process of colonisation. Such an argumentum e silentio is unacceptable in this instance: most Greek historical sources in other contexts make very little mention of women. Herodotus provides a great deal of evidence for this thesis because he mentions women more frequently than other writers. He refers to a wide variety of women including wives, daughters, mothers, queens, priestesses, and prostitutes (among others), with a total of 375 mentions throughout the Histories.7 By contrast Thucydides, for example, mentions women eight times less frequently than Herodotus.8 Like women, indigenous populations are largely ignored in the literary record. Non-Greek sources are virtually non-existent, so we must be mindful of this in our interpretation of the available evidence. The archaeological evidence also presents challenges. Many excavations were conducted in the nineteenth or early twentieth centuries when the interpretation of the problems of Greek colonisation was very different. As a result, any analysis of the presence of women, indigenous peoples, or skeletal evidence is extremely 7 7 limited. Ethnicity and identity are big factors in such a project. Typically, we could expect to distinguish between Greeks and various indigenous populations through archaeology – in particular through burial customs and grave goods. However, it is becoming increasingly obvious that this is not straightforward. Many sites reveal a mixture of customs, and also the adoption and adaptation by some groups of the practices of others.9 In general, much of the archaeological evidence is open to more than one interpretation and leaves ample room for conjecture and speculation. By the same token, distinguishing male from female in the necropoleis is becoming increasingly more difficult as the common methods of sex association cannot be relied upon (that is, we cannot definitively conclude that graves containing objects typically associated with women also belonged to females). Therefore archaeology provides only limited indications of ethnicity and identity, and must be used with caution. Where possible, the archaeological evidence should be combined with the literary, so that a more rounded picture may be achieved. What follows is in three parts. Chapter One examines the problems associated with archaic Greek colonisation generally, particularly those issues connected with the ‘language of colonisation’. p y , ( ) p , y y , “which may be actively played up or down according to local needs”. See also Hall (1997). 9 This is partly because, as Owen (2005) 8 points out, ethnicity is a socially-constructed notion, 9 This is partly because, as Owen (2005) 8 points out, ethnicity is a socially-constructed notion, “which may be actively played up or down according to local needs”. See also Hall (1997). 9 This is partly because as Owen (2005) 8 points out ethnicity is a socially-constructed notion INTRODUCTION The study of Greek colonisation has been complicated by imprecise and ambiguous terminology, which frequently draws comparison with more modern (although altogether different) instances of the phenomenon. A major repercussion of this is the tendency to overlook both women and any indigenous peoples in the scholarship. Chapter One also examines the various reasons behind the foundation of colonies, possible methods of transport, as well as the different types of settlements, so that an assessment can be made about whether Greek women might have been more likely to accompany colonising expeditions in some instances over others. Chapter Two looks at the concept of intermarriage more closely and assesses the Greek attitude towards foreign women. It also evaluates the evidence typically called upon by scholars to argue for and against intermarriage in Greek colonisation. Chapter Three assesses the evidence of women in ten different colonies. Presented roughly in chronological order, these colonies were selected for their geographical scope, 8 covering different regions from the Western Mediterranean, Magna Graecia, North Africa, and the Black Sea. I examine both the literary and archaeological evidence from each colony (where possible), and assess the potential for intermarriage at each place. Overall, I argue that contrary to the practices of much of the scholarship, archaic Greek colonisation cannot and should not be considered as a uniform phenomenon. Each colony needs to be considered individually, as each colony had a different mother city, a different foundation date, different foundation stories, and accordingly each appears to have had different practices in sourcing its women. Therefore, I argue that based on the available evidence, broad conclusions about intermarriage in the colonies are as risky as broad conclusions about Greek women accompanying men on the colonial voyage. 9 Introduction Colonisation is a difficult concept. It is subject to a variety of interpretations, laden with a wide variety of connotations, and applied to a variety of peoples – from the ancient Greeks to the Romans and Vikings, from the Spanish, French, and British to the Russians expanding the Soviet Union. In addition, colonisation covers more than people and land: it is also applied to plants, bacteria, and a wide variety of animals. In English today we talk about the colonisation of space, ant colonies, bee colonies, mushroom colonies, and even nudist colonies. Consequently, it is necessary to examine the way in which the terminology associated with colonisation has been used to describe the Greek settlements established in the archaic period, and to identify how this use has impacted on the scholarship. The language that nineteenth and twentieth century scholars have used to discuss Greek colonisation is often imprecise and sometimes ambiguous. Throughout the twentieth century, there has been a tendency to use terms arbitrarily, with little regard to resulting repercussions. This has led to many of the complications surrounding the study of Greek colonisation. One of the primary issues in the discussion of colonisation is the association and conflation with contemporary interpretations of the concept. This is particularly evident in western scholarship, where our own experiences of British colonial systems have greatly influenced the way in which we may interpret the colonial experience of others. Many scholars have drawn direct comparisons between Greek colonisation and British colonisation. For example, Australian-born T. J. Dunbabin stated in his preface to the foundational text, The Western Greeks: I have drawn much on the parallel to the relations between colonies and mother country provided in Australia and New Zealand. Here political independence is combined with almost complete cultural dependence, on which the colonials pride themselves. Difference in 10 manner of life is due to difference of material circumstances, and is not enough to destroy the essential unity. ( ) ; ( ) 15 Isaac (2004) argues against this, though he uses little evidence to back up his assertions, and the purpose of his work, in fact, appears to be to absolve modern society of the creation of racism. 10 Dunbabin (1948) vii. Dunbabin goes on to analogise the economic life of ancient colonies with modern examples of raw material production. 11 Osborne (1998) 252 ( ) 12 Beazley in Dunbabin (1957) 5. See Ridgway (1990) 62: “‘primitive’ is not an adjective that I would willingly apply today to the Italian Iron Age”. 13 J (1940) 27 ( ) 14 Boardman (1980) 7; Tsetskhladze (2006c) lii. ( ) 14 Boardman (1980) 7; Tsetskhladze (2006c) lii. 15 Introduction This unity is the pride of most colonials; so probably in antiquity.10 By comparing ancient with modern colonisation, notions of cultural and political control are insinuated in the colonisation process; in particular, the idea that there are elements of power and domination over any indigenous inhabitants.11 For example, in his foreword to Dunbabin’s The Greeks and their Eastern Neighbours (1957), John Beazley wrote: “In the West the peoples with whom the Greeks came into contact were at a more primitive stage of development than they themselves”.12 In 1940, A. H. M. Jones described the Greek colonies in Black Sea as “mere islets of civilization in a sea of barbarism”.13 These quotations clearly demonstrate the prevailing attitudes of the early twentieth century, deeming the indigenous peoples to be lesser than the Greeks: less advanced, less developed, and perhaps less intelligent as well. The prejudices against the Aborigines of Australia and the Māori of New Zealand following colonial contact (and arguably continuing to this day) are primarily the result of a colonial view that the colonising group is superior to the colonised, due to perceptions that the latter has a primitive way of life. However, many scholars are beginning to recognise that the Greek outlook on other peoples was, perhaps, very different. The term βάρβαρος simply referred to a person who did not speak Greek.14 It was not a term tied to any particular people or race, and did not have the same wild and savage connotations that the term carries today.15 It appears that the distinction between the Greeks and the barbaroi emerged only following the Persian invasion of Greece in 480/479 BC; many scholars argue that there were no 11 (or few) feelings of contempt or racially motivated hostility prior to this.16 It follows that a comparison between Greek colonisation and modern colonisation is not a fair or accurate comparison, particularly because of the very different ways the two approached indigenous populations. The overtones of political and social control associated with modern colonisation were not necessarily present in the settlement process of the ancient Greeks. More recent scholarship increasingly acknowledges this distinction. 16 Tuplin (1999) 54; Antonaccio (2007) 204; Roebuck (1959) 33. Certainly, following the Persian war, previous enemies such as the Trojans gained a retrospective status as “contemptible barbarians” (Tuplin (1999) 55). Hall (1989) argues for the importance of tragedy as a vehicle for inventing and defining ‘the barbarian’. ( ), 19 Osborne (1998) 269. g g 17 Tsetskhladze (2006c) xxvii. Also recently Snodgrass (2005) 45; Owen (2005); Shepherd (1999) 271. See also Eikeland (2006) 23; Hodos (2006) 10; and in particular, De Angelis (1998) for the critical analysis of Dunbabin’s approach. 18 See Gosden (2004) and Lyons & Papadopoulos (2002) for examples Osborne (1998) 269. 20 Osborne (1998) 268. De Angelis (2011) 21 argues that we need to coin new terms to use, such as ‘apoikiazation’ instead of colonisation. 21 Whitley (2001) 125. y pp 18 See Gosden (2004), and Lyons & Papadopoulos (2002) for examples. 19 22 Graham, A. J. ‘Colonization, Greek’ excerpt from OCD2 (1970), 264. Introduction Tsetskhladze argues: “colonisation is essentially a modern Anglophone concept, based on examination and interpretation of the imperial activity of the European powers of our era, transported back and forced onto ancient Greece”.17 Yet, some recent publications continue to link ancient and modern colonisation – not just conflating their interpretation, but also presenting comparative approaches as though they were different versions of the same phenomenon.18 This reflects how difficult it is to suppress the instinctual need to relate past events and practices to modern experience, as a way of better understanding those events and practices. But we need to be wary of this, and bear in mind that, other than the term itself, there is little in common between ancient and modern colonisation. In an influential paper, Osborne called for expunging the term ‘colonisation’ from books on early Greece.19 The justification for such a bold appeal was that otherwise “we will go on calling the settlements ‘colonies’ and will go on mistaking both the causes and the nature of settlement in the West by invoking a colonization model”.20 While on the one hand, I agree that the conflation of ancient with modern colonisation must cease, I also agree with Whitley’s observation, that “we have to call this process something, and colonisation is as good a term as any”.21 As long as scholars are both aware of the pitfalls of interpretation in previous scholarship, and 12 recognise the differences between the ancient and modern processes, then scholarship will continue to move forward. The ever-changing nature of our understanding of colonisation has almost certainly affected the scholarship on the Greek settlement process in the archaic period. This change is particularly evident in the discussion of the definitions of ‘colonisation’ in the Oxford Classical Dictionary. The Second Edition of the Oxford Classical Dictionary, published in 1970, defines Greek colonisation as: Colonization was always a natural activity for Greeks, living in a poor country. Mycenaean colonies of the Late Bronze Age have been revealed by archaeologists (e.g. at Miletus), the coast of Asia Minor and the islands off it were settled at the beginning of the Iron Age, and there was much colonization in Asia under Alexander and in the Hellenistic period. Nevertheless, the greatest colonizing achievement, by which Greek cities were spread round the coasts of the Mediterranean and Pontus, is that of the archaic period, c. 750-c. Introduction 550.22 The most recent edition of the same work, published in 2012, saw a drastic change in the definition of Greek colonisation: ‘Colonization’, in the language of a former imperial power, is a somewhat misleading definition of the process of major Greek expansion that took place between c. 734 and 580 B.C. In fact, the process itself was not so much ‘Greek’ as directed in different ways and for different reasons by a number of independent city-states. This at least emerges with relative clarity from both the historical and the archaeological evidence. For the rest, the mass of general and particular information that has accumulated under these two headings is only rarely susceptible to a single uncontroversial interpretation. Although the position has greatly improved since the 1930s, it is still only too true that archaeologists and ancient historians do not always appreciate each other’s aims and methods – a problem that is 13 exacerbated by the fact that on the subject of colonization ancient no less than modern authors are more than usually influenced by their own political agenda and accordingly more than usually liable to project the priorities, practices, and terminology of their own times onto the much earlier events they purport to describe.23 This simple comparison shows the radical change in scholarly thinking within just forty years. There is a much greater emphasis placed on terminology; and the problems associated with the subject are more clearly outlined. Importantly, the 2012 edition acknowledges that the settlement process was initiated by different city states, each for different reasons, and most importantly that not all colonies were the same. This is a fundamental point, crucial to avoiding the ‘Colonisation Model’ which Osborne sought to eradicate.24 Modern scholarship often falls into the trap of talking about an ‘age of colonisation’.25 It is undeniable that migration was a prevalent feature of the ancient world as it remains today. A recurring theme in the Homeric poems concerns men who wander abroad. In Greece, migration took place in the late eleventh to tenth centuries BC, with the Ionians, Dorians, and Aeolians shifting around the Aegean and into Asia Minor, as well as the Mycenaeans’ movement prior to this.26 Hesiod’s father, for example, migrated and chose to settle in Boiotia (Works and Days 630-640). The Hellenistic period was also a period of great migration, into Syria, Egypt, and the East. 23 Ridgeway, D.‘Colonization, Greek’ excerpt from OCD4 (2012), 348. 24 g y, , 4 Osborne (1998) as discussed above. ( ) 25 For example, Murray (1993) 102, cf. Graham (1982) 83. 26 p y ( ) ( ) 26 Tsetskhladze (2006c) xxiii. ( ) 27 Murray (1993) 102. Osborne (1998) as discussed above. 25 For example, Murray (1993) 102, cf. Graham (1982) 83. 26 p , y ( 26 Tsetskhladze (2006c) xxiii. 27 6 Tsetskhladze (2006c) xxiii. 7 Murray (1993) 102. ( ) 27 Murray (1993) 102. 32 Consequently, a distinction can be drawn between public and private ventures. For example, Graham (1964) 7-8 (re-emphasised Graham (1982) 143-146) draws a distinction between the foundation of Cyrene as a state act, versus the abortive, and private, expedition of Dorieus (Herodotus 5.42-48). The reasons for these ventures will be discussed later in this chapter. Osborne (1998) shifted the debate significantly, arguing that private ventures alone should account for archaic Greek colonisation in the west. Most recently Morakis (2011) assessed the language used by Thucydides (6.3-5) to argue that the first generation colonists of Sicily were of a private character, while the later expeditions had a more ‘state-guided character’. For example, Morakis (2011) 467 argues: “at 6.3.1, where Thucydides says Ἑλλήνων δὲ πρῶτοι Χαλκιδῆς ἐξ Εὐβοίας πλεύσαντες µετὰ Θουκλέους οἰκιστοῦ Νάξον ᾤκισαν, the subjects of the colonising movement are Thoucles the oikist and Χαλκιδῆς but not the state/polis of Chalcis. The ethnic of the colonists (Chalcidians) is used in an indefinite context alongside Thoucles’ name and only to denote the geographical origin of the colonists (that is the reason for the ἐξ).” These arguments are heavily tied up with the reasons why colonies were founded, to be discussed shortly, and whether or not the polis was yet in existence. It is also difficult that these debates have a hugely western focus with rare regard to how the situation in the Black Sea may have tallied or differed. 33 Th ἰ ή ( iki ) i li i d ‘ iki ’ , j p y gy 30 See Wilson (2006) 25 for the following list of points of ‘long established certainties’. 31 ( ) 29 Graham (1981-82) 294 argues that the Greeks did not distinguish between the colonisation that took place in the migratory period and the colonisation of the archaic period. While this may be true, a major theme in this chapter is that nearly all terminology is modern. 30 ( ) g p g 31 For example, Cumae appears to have had more than one mother city. See Graham (1964) 16 32 28 Wilson (2006) 27. 29 28 Wilson (2006) 27. 29 Graham (1981-82) 294 argues that the Greeks did not distinguish between the colonisation that took place in the migratory period and the colonisation of the archaic period. While this may be true, a major theme in this chapter is that nearly all terminology is modern. 30 See Wilson (2006) 25 for the following list of points of ‘long established certainties’. 31 For example, Cumae appears to have had more than one mother city. See Graham (1964) 16. 32 Consequently, a distinction can be drawn between public and private ventures. For example, Graham (1964) 7-8 (re-emphasised Graham (1982) 143-146) draws a distinction between the foundation of Cyrene as a state act, versus the abortive, and private, expedition of Dorieus (Herodotus 5.42-48). The reasons for these ventures will be discussed later in this chapter. Osborne (1998) shifted the debate significantly, arguing that private ventures alone should account for archaic Greek colonisation in the west. Most recently Morakis (2011) assessed the language used by Thucydides (6.3-5) to argue that the first generation colonists of Sicily were of a private character, while the later expeditions had a more ‘state-guided character’. For example, Morakis (2011) 467 argues: “at 6.3.1, where Thucydides says Ἑλλήνων δὲ πρῶτοι Χαλκιδῆς ἐξ Εὐβοίας πλεύσαντες µετὰ Θουκλέους οἰκιστοῦ Νάξον ᾤκισαν, the subjects of the colonising movement are Thoucles the oikist and Χαλκιδῆς but not the state/polis of Chalcis. The ethnic of the colonists (Chalcidians) is used in an indefinite context alongside Thoucles’ name and only to denote the geographical origin of the colonists (that is the reason for the ἐξ).” These arguments are heavily tied up with the reasons why colonies were founded, to be discussed shortly, and whether or not the polis was yet in existence. It is also difficult that these debates have a hugely western focus with rare regard to how the situation in the Black Sea may have tallied or differed. 33 The οἰκιστής (oikistes) is anglicised as ‘oikist’. 33 The οἰκιστής (oikistes) is anglicised as ‘oikist’. Introduction The movement that took place in the archaic period, however, is frequently singled out as the so-called ‘age of colonisation’. Certainly, the colonies of this period stand apart from other periods of migration and settlement partly in their scale and the extent of their geographical reach. Indeed, such was the extent of this colonisation that it remained unsurpassed until the late fourth century BC with Alexander the Great’s conquests (which were of a different nature).27 However, talking about an ‘age of colonisation’ is dangerous, as it implies that all the colonies from that age should 14 be treated as a single movement or phenomenon.28 This in turn deceptively suggests that all of the colonies from within that movement were founded for the same reasons, with matching intentions. A careful distinction must also be made between colonisation and migration. Migration occurs when the entire population of a settlement uproots itself (for whatever reason) and relocates to another area to establish a new settlement. By contrast, with colonisation, only a portion of a population relocates while the original settlement (mother city) continues.29 The elements required for Greek colonisation are usually discussed within the parameters of some so-called ‘long established certainties’.30 Some of these certainties are now open to fresh questioning. It is usually assumed, for example, that all colonies were founded by a state (polis) and that consequently, each colony was designed to be a mirror image of its mother city. In fact, colonies could have multiple mother cities as sometimes colonists were collected from all over the Greek world.31 In addition, the impetus for the foundation of a colony was not necessarily found in the polis, but could come from a private individual.32 Colonising parties are typically described as being led by a single individual, the oikistes.33 He was apparently responsible for setting up the social, religious, and political aspects of the new community, and he was usually worshipped as a hero after his death, with his burial in a prominent place in the agora. However, there are instances where there was more than one oikist (such as Metapontum), or 15 where there is no record of the oikist at all (such as the settlement at Pithekoussai). It is usually assumed that the Delphic oracle was consulted prior to setting out on the colonial expedition to gain approval, advice, or geographical aid. Introduction In some instances, such as the foundation of Cyrene from Thera, Delphi was consulted a number of times (Herodotus 4.150-151;155). In other instances, however, Delphi may not have been consulted at all (such as Dorieus’ activities in North Africa).34 Finally, it is usually assumed that the process of settlement in a new land involved the violent expulsion of the indigenous population. This point is most fundamental to the focus of this thesis and, as we will see, it was not always the case. Some colonies were settled on previously uninhabited land, and conversely, there is strong evidence for others that relations between the Greek colonists and indigenous population were relatively peaceful (such as Megara Hyblaea). Thus, it now appears that little is certain on the list of ‘long established certainties’. One of the primary problems with such a list is that, as we will see, it fails to acknowledge that every single Greek colony differed from the next, even those which had the same mother city. Such ‘long established certainties’ fit within the ‘age of colonisation’ argument, and consequently they do not seem to often coincide with the reality of Greek colonisation. Furthermore, Malkin points out that many of the expectations or conditions associated with the foundation of Greek colonies are in fact ones that date to the classical period, and therefore are not applicable to those of the archaic period.35 34 Interestingly, at least in this instance, the foundation’s inability to succeed has been directly linked to their neglect to consult Delphi – see Herodotus (5.42). 35 M lki (1987) 155 g p ( ) 35 Malkin (1987) 155. 35 Malkin (1987) 155. ( ) 38 See for example, Dunbabin (1948) 454; Murray (1993) 113; Graham (1964) 43. Furthermore, I thank Arthur Pomeroy for pointing out that these two pentekonters could have been for military support, and not the vessels containing the colonists. 39 ( ) ( ) 40 See Cawkwell (1992) 290 for other possibilities, such as the addition of a sigma after ἄνδρας to show that it was 200 men. 37 Wilson (2006) 39. 38 41 Tsetskhladze (1994) 114; Petropoulos (2005) 17. 36 All translations and bolded emphasis are my own. 37 41 Tsetskhladze (1994) 114; Petropoulos (2005) 17. 36 All translations and bolded emphasis are my own. 37 il ( ) p y 37 Wilson (2006) 39. 38 37 Wilson (2006) 39. o Van Wees (2004) 203. 43 Carpenter (1948). Carpenter chooses his access date, citing Thucydides (1.13), who states that there were improved ship building techniques at Corinth during the same period. Carpenter misinterprets Thucydides, who also states that the Corinthians were the first of the Greeks to have triremes (not pentekonters), and that Ameinocles made four of these ships for the Samians (and so is typically seen as being the first Greek to build triremes – see Graham (1958) 26-27). Furthermore, little extensive archaeological exploration had been conducted at Carpenter’s time of writing. 44 Practical considerations As an alternative to this ‘one size fits all’ approach, it is useful to consider in practical terms what was required to make a colonial expedition work. In the first few years of existence, the new settlements would have needed to have been relatively self-sufficient. Food, water, and shelter were all essential to survival. With a view towards more long-term survival, certainly an ability to reproduce was crucial and this obviously required women. Homer’s Odyssey (9.116-24) gives us some indication of what colonists looked for in an area to be potentially settled: 16 νῆσος ἔπειτα λάχεια παρὲκ λιµένος τετάνυσται, γαίης Κυκλώπων οὔτε σχεδὸν οὔτ᾽ ἀποτηλοῦ, ὑλήεσσ᾽ ἐν δ᾽ αἶγες ἀπειρέσιαι γεγάασιν ἄγριαι... νῆσος ἔπειτα λάχεια παρὲκ λιµένος τετάνυσται, γαίης Κυκλώπων οὔτε σχεδὸν οὔτ᾽ ἀποτηλοῦ, ὑλήεσσ᾽ ἐν δ᾽ αἶγες ἀπειρέσιαι γεγάασιν ἄγριαι... ...οὔτ᾽ ἄρα ποίµνῃσιν καταΐσχεται οὔτ᾽ ἀρότοισιν, ἀλλ᾽ ἥ γ᾽ ἄσπαρτος καὶ ἀνήροτος ἤµατα πάντα ἀνδρῶν χηρεύει, βόσκει δέ τε µηκάδας αἶγας. There is a fertile, wooded island stretching before the harbour, neither near the land of the Cyclopes, nor far away, in which countless goats live wild…used neither for flocks, nor corn-fields [ploughing], but in fact is unsown and unploughed for all the days, it lacks men feeding only bleating goats.36 The island is currently uninhabited and is fertile. Though technically Odysseus is not looking to found a colony, he is aware of its potential for settlement and consequently this passage is commonly seen to reflect the ideals of colonisation.37 The number of colonists required for each expedition is not known, but presumably this could vary considerably. Herodotus (4.153) records that two pentekonters voyaged from Thera to Cyrene; many scholars have presumed, incorrectly, that this meant 200 people went to the new colony.38 Morrison and Williams’ study, Greek Oared Ships demonstrates that a pentekonter typically carried 50 oarsmen and two officers.39 By contrast, Herodotus, elsewhere, allows for 80 men in each pentekonter (7.184.3).40 The debate surrounding which kinds of vessels were used on colonial voyages becomes particularly salient for the Black Sea colonies; specifically the date that the Greeks were able to sail into the Black Sea, and therefore the earliest possible date of settlement. y g 47 Graham (1971) 39. This is of course extremely contentious. It is not yet clear whether the territory surrounding the Black Sea was heavily populated or not; an issue complicated by the nomadic character of the indigenous peoples. See Tsetskhladze (1998b) 44. (This is an issue that will be discussed in the Black Sea section of Chapter Three). Drews (1976) supported this argument a few years later. Graham (1958) 29-30. 46 It would be highly unlikely that a group of colonists would depart their mother city to settle in new lands without sufficient provisions to last them past the initial settlement period (however long that may have been). As such, the pentekonter, with its limited storage space, seems less likely to have been used. Hence, the ships that were used by the colonists would, in all probability, have been able to access the Black Sea by sailing into it. 44 Labaree (1957) 31. 45 g y 48 Graham (1990) 53-54. 45 Graham (1958) 29-30. 46 42 Carpenter (1948). 43 Practical considerations There are references to the region in mythological traditions (for example, the Argonauts’ voyage to Colchis in search of the golden fleece), though scholars have struggled to place these instances chronologically.41 In 1948 17 Carpenter became the leading proponent in this debate, publishing an article which concluded that the Black Sea was closed to the Greeks before c. 680 BC, due to the strong Bosporus current.42 Carpenter argued that access to the Black Sea was tied to the invention of the pentekonter, thus explaining why there is little archaeological evidence before this time.43 It was necessary for crews to row pentekonters through the entrance to the Black Sea, as sailing through was not possible. In 1957 Labaree conducted a more technical analysis of the capabilities of the pentekonter,44 and a similar study was undertaken by Graham in 1958 to assess the effects of the wind and currents on access to the Black Sea, whether by sailing or rowing.45 Both concluded that the Greeks were in fact able to sail into the Black Sea, and thus the date of the Greek penetration of the Black Sea did not rely on technical improvements in rowing vessels.46 Graham reasserted his opinion in 1971, and added that the absence of Greek colonies (particularly with reference to the northern shores of the Black Sea) was due to the ‘warlike strength’ of the indigenous population.47 Graham re-emphasised his position in 1990, attempting to use some pottery as evidence.48 The limitations of this approach were criticised by Boardman in 1991: whether there is any archaeological evidence for earlier exploration or settlement is another matter, but Graham has pressed claims which, as whether there is any archaeological evidence for earlier exploration or settlement is another matter, but Graham has pressed claims which, as I hope to show, cannot be upheld since the dating of the pottery or its settlement is another matter, but Graham has pressed claims which, as I hope to show, cannot be upheld since the dating of the pottery or its I hope to show, cannot be upheld since the dating of the pottery or its 18 pedigree are either wrong or too dubious to be taken seriously, however tempting they may seem.49 There have been few significant developments in this debate since. Tsetskhladze readdressed the arguments in 1994, but did not produce any new conclusions. ( ) 51 See Labaree (1957) 31 for a cogent description of the pentekonter. There is frequent conflation of the pentekonter and the trireme in the aforementioned arguments – see Van Wees (2004) 202- 206 for clarification. At this time, such ships were typically owned by private individuals; there was no state-owned war fleet such as that possessed by Athens in the fifth century BC. This point could potentially impact on the debate surrounding whether colonial voyages were public or private ventures. 52 ( ) 50 Tsetskhladze (1994) 111-112. 52 Koromila (1991) 24. 3 49 Boardman (1991) 387. ( ) 53 Graham (1990) 45. Boardman (1991) 387. 0 Tsetskhladze (1994) 111-112. ( ) 51 See Labaree (1957) 31 for a cogent description of the pentekonter. There Practical considerations However, arguably more importantly, he did outline the reasons behind the long- running controversy: most fundamentally, acknowledging the dearth of any substantial archaeological evidence on which to base an opinion, coupled with a lack of any strict chronology to support an assessment.50 From this examination of vessels, it seems that pentekonters were an unlikely means for transporting Greek women to colonies. They were designed specifically as vessels that were not necessarily dependent on winds but could be manipulated by the sheer force of rowers.51 As such, their design departed drastically from the old, sailing merchant ships. Instead of having a broad, deep hull to allow for maximum storage space, the pentekonter was long and narrow, with a shallow draught to allow for high-speed propulsion.52 Consequently, there was little room for storage as the available space was taken up by the multiple rowers required to propel the ship. This, however, does not seem to correspond to many textual references to pentekonters. For example, Herodotus (1.163.1-2) states that the Phocaeans were the first Greeks to make long sea voyages and he specifically claims that these were made in pentekonters: ἐναυτίλλοντο δὲ οὐ στρογγύλῃσι νηυσὶ ἀλλὰ πεντηκοντέροισι ‘they sailed not in round ships but pentekonters’. This is difficult to reconcile with the fact that the Phocaeans were sailing to locations famous as trading ports, such as Tartessus, and they would presumably therefore require space in their vessels to transport both goods to sell and goods purchased.53 In another instance, Herodotus (1.164) describes how during the Phocaean evacuation of their city: οἱ Φωκαιέες ἐν τούτῳ κατασπάσαντες τὰς 19 πεντηκοντέρους, ἐσθέµενοι τέκνα καὶ γυναῖκας καὶ ἔπιπλα πάντα ‘the Phocaeans, in the meantime, pulled down the pentekonters, and put in the children, and women, and all moveable property’. For colonial voyages, it must be assumed that, at the minimum, some crops were transported for planting in the new settlement. 54 In the Black Sea, for example, Pashkevich’s study demonstrates that the Greek colonists did not plant crops that were indigenous to the Black Sea, but instead brought crops with them from the Greek mainland. S p ( ) 56 Morrison & Williams (1968) 39 point out that, due to issues of perspective, it is difficult to distinguish between a ship with two levels of rowers, and a ship where both sides of the ship are depicted. See also Kirk (1949); Popham (1987); Pomey (1996). Practical considerations In the case of Cyrene, too, Herodotus (4.151-152) reports that the Theraean colonists left Corobius, their guide, on an island (Platea) with enough supplies for a number of months, which would seem to be far more than a pentekonter could have carried.55 This is clearly not a straightforward issue, nor one that offers an easy answer. It seems that the textual references refer to a different type of ‘pentekonter’ from our understanding of it as a long, shallow vessel with 50 rowers. Visual representations are also of little help given the stylised images characteristic of the Geometric style.56 For colonial voyages with even a modicum of planning prior to departure, some storage space to transport goods would have been imperative. If women were to be sent out to the colony at a later date, possibly with another wave of second-generation colonists, some holding space in their transport vessel would have been necessary. It is inconceivable that women could have rowed the ships themselves. Consequently, we must conclude that the ships used for colonial voyages, typically called pentekonters in the literary evidence, had some provision for storage space, and were therefore not pentekonters of the type known from later sources. 54 Pashkevich (2001). ( ) 55 See Petropoulos (2005) 124. 56 54 Pashkevich (2001). 54 Pashkevich (2001). Pashkevich (2001). 55 See Petropoulos (2005) 124. Camp (1979). 60 Holloway (1981) 144. Holloway also points out that other factors could have caused this phenomenon such as a local instability of ground water in the agora. 61 Whitley (2001) 125-126. ( ) 63 Overpopulation has also been proffered as a reason for the foundation of secondary colonies, such as Selinus, a secondary colony of Megara Hyblaea. De Angelis (1994) presents a very compelling study showing that Megara Hyblaea was using only a fraction of its available land so it is unlikely that Selinus was founded due to overpopulation. 64 ( ); ( ) 58 Scheidel (2003) 121; 131 argues that it is unlikely that historical populations ever reached full saturation level. 59 Camp (1979) 57 See for example Gwynn (1918); Littman (1974); Snodgrass (1980); Van Compernolle (1983); Waters (1974); Owen (2005). 58 y 64 Van Compernolle (1983) 1037-1038. y ( ) 62 Tsetskhladze (2006c) xxviii. 63 y p p 64 Van Compernolle (1983) 1037-1038. Whitley (2001) 125 126. 2 Tsetskhladze (2006c) xxviii. 3 Reasons for founding colonies The reasons behind overseas Greek settlement in the archaic period have been a much loved topic for debate. The approach to this debate, however, usually assumes that there could be only one reason why all of the colonies were founded. In reality though, it appears that each colony differed from the next. Some 20 possible reasons for the overseas expansion will be outlined below, and these will be further synthesised in the case studies in Chapter Three. For a long time, historians championed overpopulation as the primary reason for the foundation of Greek colonies.57 This may have stemmed from the influence of later Greeks, such as Plato, who wrote that founding colonies was a cure for overpopulation (Laws 740e), and compared it to the way a swarm of bees settles a new hive (Laws 708b). Certainly, logically, overpopulation would provide good reason to create new settlements. However, although difficult to document, there is little evidence that overpopulation was ever an issue.58 Camp concluded that the closing of many of the wells in the Athenian agora in the late eighth century BC, coupled with the increase in graves at the time, were signs of the conditions that could lead to colonisation – drought and overpopulation.59 Such regional arguments are highly problematic however. Camp’s hypothesis is based on a small area of Attica,60 and moreover, Attica (and the Argolid) is well-recognised as being one of the only areas not to send out any colonies.61 Furthermore, several regions of mainland Greece remained uninhabited well into the archaic period, and therefore overseas expansion was not necessarily required to solve overpopulation. These factors are incongruent with Camp’s conclusions. 68 Cawkwell (1992) 296. For example, the hugely successful trading port Tartessus was located on the south coast of the Iberian peninsula through the ‘Pillars of Herakles’ (see Herodotus 4.152). 67 Blakeway (1932-1933) 202. Osborne (1998) 268-269 sees this as an inappropriate discussion as talk of a ‘flag’ is anachronistic. While this is certainly true, the argument in its essence still holds true: did trade come before settlement? 65 Cawkwell (1992) 301. 66 Crielaard (1992-93) 242. See also Dougherty (2001) 137; Dougherty (1993b) for the violence associated with colonisation. 67 65 Cawkwell (1992) 301. 66 65 Cawkwell (1992) 301. 66 ( ) 66 Crielaard (1992-93) 242. See also Dougherty (2001) 137; Dougherty (1993b) for the violence associated with colonisation. 67 Blakeway (1932-1933) 202. Osborne (1998) 268-269 sees this as an inappropriate discussion as talk of a ‘flag’ is anachronistic. While this is certainly true, the argument in its essence still holds true: did trade come before settlement? 68 Cawkwell (1992) 296. For example, the hugely successful trading port Tartessus was located on the south coast of the Iberian peninsula through the ‘Pillars of Herakles’ (see Herodotus 4 152) Reasons for founding colonies By the same token, those areas which were prolific colonisers, such as Euboea, Corinthia, or the Megarid, yield no hard evidence that their population was rising.62 Therefore it seems unlikely that strict overpopulation by itself was a primary cause for colonisation.63 As a variant of this concept, Van Compernolle suggests that overpopulation was caused by uneven land distribution and that the colonising party was made up of people who had no access to resources for survival in their homelands.64 Such 21 regional problems may have been exacerbated by some sort of natural or climatic disaster and this has also been linked to the overpopulation argument.65 Certainly, Herodotus (4.151) reports that the Theraeans were suffering from an extensive drought, which eventually compelled them to seek advice from Delphi, who in turn advised them to found a colony in Libya (and indeed, this is presumably the genesis of the overpopulation argument). The reasons outlined above each affected the entire mother city population. Accordingly, if a colony was the product of overpopulation or some sort of natural disaster, it would seem likely that a wider cross section of the community would leave to settle in new lands. Consequently, there was potential for women to accompany men on these voyages. Other factors driving overseas settlement made the participation of women less likely. One of these, the simple desire to acquire more land, could have been a strong motivating factor for landless men to found new colonies. Thucydides (1.15.1) states: ἐπιπλέοντες γὰρ τὰς νήσους κατεστρέφοντο, καὶ µάλιστα ὅσοι µὴ διαρκῆ εἶχον χώραν ‘for sailing to the islands, they subdued them, as they did not have sufficient land of their own’. Crielaard adds that part of the attraction of colonisation could have been simply that ‘conquering’ territory appealed to the aristocratic warrior ethos.66 The desire for new land was also intimately tied up with trade and wealth. 69 For example, evidence of Mycenaean contacts with Magna Graecia can be seen in the archaeological record, and Antonaccio (2005) 97 argues that contacts between Italy and Greece probably remained uninterrupted. Consequently, knowledge of the attractiveness of Magna Graecia (at least) was likely to have been reasonably widespread. 70 Whitley (2001) 126 Reasons for founding colonies Indeed, Blakeway famously questioned whether ‘the flag followed trade’.67 Certainly this could explain why many Greek colonies (for example, Pithekoussai, Massalia, Heraclea Pontica, to name a few) tended to be settled on the fringes of the Greek world, rather than close to contemporary settlement; trade vessels were potentially more likely to stop at the terminus of a trade route than halfway en route.68 Knowledge of fertile land or the existence of 22 raw resources may have also been a motivating factor for setting out colonies.69 Simple desire to gain more agricultural land appears as a trait of the later colonies, and so ‘land hunger’, as it were, may well have been fuelled by the knowledge and experience gained by the early Greek colonies overseas.70 Civil strife at home presents another potential reason why Greeks founded colonies. Factional disputes could lead to certain individuals or groups being ordered (or choosing) to leave the mother city. Crielaard connects this factor, in particular, with the aristocracy – the socio-political and religious functions were largely carried out by the elite, and the impact on these of the changes associated with the development of the polis could have motivated many aristocrats to seek alternative settlements, in the form of new colonies overseas.71 Such ventures could well have fallen into the ‘private’ ventures category, rather than the ‘public’. The colony Locri Epizephyrii appears to have been an example of this. Although previously, scholars promoted overpopulation as the sole reason behind the foundation of archaic Greek colonies, there is little evidence to support such a theory. Instead, colonies seem to have been founded for a variety of different reasons. Accordingly, each colony had the potential to differ from the next. y ( ) 71 Crielaard (1992-93) 240-241. 2 ( ) 72 See, for example, Whitley (2001); Hansen & Nielsen (2004); Wilson (2006); Wilson (1997); Hansen (1997); Demetriou (2011). 73 LSJ sv ἀποικία See, for example, Whitley (2001); Hansen & Nielsen (2004); Wilson (2006); Wilson (1997); Hansen (1997); Demetriou (2011). 73 LSJ sv. ἀποικία. ( ) 73 LSJ sv. ἀποικία. ( ) ( ) 77 Wilson (1997) 199. See also Van Compernolle (1983) 1037; Malkin (1994b) 1; Ridgway (1992) 108. The debate surrounding the formation of the polis is vast and ongoing. With regard to Greek colonies, poleis are typically thought to have had fortification walls; however, at the early period of development of the polis concurrent with the foundation of colonies, there were no city walls. Karlsson (1989) traces the first defensive walls in Sicily to the period of tyranny. See also Fischer- Hansen (1996) 319. By contrast, de Polignac (2005) 47 argues that some primitive enclosures separated the interior urbanised area from the ‘outside’ in the archaic period, primarily following so-called ‘accidents of topography’ (that is, naturally occurring phenomena such as ridges, watercourses, or the edge of a plateau). ( ) 79 Murray (1993) 102-3; Hodos (2006) 19-20; Wilson (1997) 205; Ridgway (1992) 107 also suggests that Pithekoussai, in particular, encouraged the polis to the shape for which it was later known. Holloway (1981) 147: “urbanism and the emergence of the polis also seem to accompany rather than to precede the stimulus of colonization”. 74 Hodos (2006) 19. This is course is an ideal, and as we will see in Chapter Three, the reality was often quite different. 75 Wil (1997) 205 ( ) 76 Wilson (1997) 199; Hodos (2006) 19. Wilson (1997) 205. 76 Wilson (1997) 199; Hodos (2006) 19. q 75 Wilson (1997) 205. 75 Wilson (1997) 205. 76 Wilson (1997) 205. 76 Wilson (1997) 199; Hodos (2006) 19. 78 Wilson (1997) 205. Different types of settlements During the archaic period, Greek settlements are typically described as falling into two categories: the apoikia and the emporion. This distinction has previously been of central focus in the scholarship.72 I will examine each in turn to determine whether the settlement type had any bearing on the presence of women. The first type of Greek settlement is the apoikia. The Greek lends itself to a definition; literally meaning ‘a home away from home’,73 an apoikia is a settlement 23 that is theoretically designed as a mirror image of its mother city.74 The settlement was set up in such a way that it could fulfil a popular ideal of the time, autarkeia, or self-sufficiency.75 Presumably then, agriculture was an essential part of these communities in order to produce food to ensure survival in a self-sufficient way. However, as discussed above with the ‘long established certainties’, the ideals attached to these new settlements did not always line up in reality. Scholars have claimed that an apoikia was itself a polis, which reflected the polis from which it originated.76 This, however, is problematic. In the eighth and early seventh centuries BC, when many overseas colonies were established, the ideology of the polis was not yet fully developed, and consequently, as Wilson puts it, was “certainly incapable of spawning a child of its own”.77 Some arguments, such as that canvassed above, that an apoikia was a settlement striving to achieve autarkeia, do not hold up: autarkeia as a concept came into play much later, during the fourth century BC, as a response to the increasing dependence of the polis on foreign trade.78 Following this line of reasoning, some scholars have argued that the so-called ‘age of colonisation’ was an essential part of the development of the polis, as colonies became more independent from their mother cities.79 Thus, it is hard to support the idea that the polis was necessary in order for colonisation to succeed. Malkin aptly states: the act of sending away entire groups to colonize abroad was not necessarily evidence of a well-organized, coherently formed political community (polis) at home. Sometimes, it was the very act of 24 separation and colonization which formed, homogenized, and consolidated the “mother city” as polis.80 Another important distinction to note contrasts the Greek apoikia with the Latin colonia. ( ) 87 Hodos (2006) 19. This is largely because the Greek emporion has been assimilated with Karl Polanyi’s (1963) famous ‘port of trade’ concept, which was intended to be a universal institution. However, as Figueira’s (1984) publication argued, the port of trade was a heuristic concept that did not translate through to the Greek word, emporion. See Polanyi (1963); Demetriou (2011) 256. ( ) g ( ) y ( ) 82 Colonia, -ae derives from colonus -i, the tiller farmer, and is also linked to the verb, colo, colere, meaning to till, cultivate, or inhabit (OLD sv. colonia). Gosden (2004) 1-2 has drawn attention to how in modern times, the ‘terra nullius’ law denied the indigenous population rights to their own land because they did not till it. The concept of this law is bound up with the Latin term, colonia. 83 Hodos (2006) 19. 85 De Angelis (1998) 529; Hodos (2006) 19. 86 Malkin (1994b) 2. 81 Osborne (1998) 252; De Angelis (1998) 539; Finley (1976) 173. ( ) 81 Osborne (1998) 252; De Angelis (1998) 539; Finley (1976) 173. 82 80 Malkin (1994b) 2. 86 Hansen & Nielsen (2004) 150. ( ) 84 Wilson (2006) 28. 83 Hodos (2006) 19. Hodos (2006) 19. 84 Wilson (2006) 28. Different types of settlements Various scholars have demonstrated that the two terms have become virtually synonymous in modern scholarship.81 However, they are in fact very different. The Latin term incorporates a reference to agriculture.82 Roman colonisation also had a distinctly imperial character. Consequently, when the two terms are paired, Greek colonisation, too, is given an imperial character which, in reality, it never had.83 While colonia was in fact a technical term used by the Romans, the Greek ‘equivalent’, apoikia, cannot be seen as an official term employed by the Greeks. Rather, it was a generic description used for a variety of settlements.84 The conflation of the two terms has been traced back to the fifteenth century AD, when Lorenzo Valla translated Thucydides’ Greek apoikia as the Latin colonia.85 Obviously, the English term, ‘colony’ directly stems from colonia, and consequently, those same notions of imperialism were transferred. Thus, it is easy to see the confusion when English terms are used to describe the Greek process. Hansen and Nielsen argue that ‘emigrant community’ would be a better translation for apoikia than colony.86 A second type of settlement is the emporion. Traditionally, emporia are thought to have had a distinctly economic function, primarily relying on the import and export of goods for a reason for existence and continued survival.87 The characteristics of an emporion can be found in Herodotus’ (2.178-180) description of the Egyptian settlement, Naucratis. These features included a harbour, a quay, 25 warehouses, administrative buildings, and a food market.88 Because an emporion did not necessarily have a permanent population, but instead had traders coming and going, it is perhaps less likely that Greek wives would have lived within the settlement, or that much intermarriage occurred. Alternatively, hetairai are noted as being present.89 For example, at Naucratis, a settlement which was famed for its hetairai according to Herodotus (2.135.5), Rhodopis is mentioned in a number of sources as the most well known hetaira.90 In a settlement that was seemingly designed around trade, where better for hetairai to pursue their own business. Like the apoikia, the emporion has also become assimilated with the concept of the polis. Settlements which have been argued to be emporia are, more often than not, also described as poleis (sometimes in addition to an emporion) by ancient sources. ( ) 92 This settlement will be further examined in Chapter Three. 93 89 Moller (2001) 152. 90 Rhodopis is mentioned by Herodotus (2.134-5); Strabo (17.1.33); Athenaeus (13.596c); Sappho (fr. 202). 91 Wilson (1997) 204 ( ) 89 Moller (2001) 152. ( ) 91 Wilson (1997) 204. 88 Hodos (2006) 19. 89 94 Hind (1997) 109. ( ) 95 Wilson (1997) 204. 93 Wilson (1997) 204. on (1997) 204. settlement will be further examined in Chapter Three. on (1997) 204 91 Wilson (1997) 204. Different types of settlements Herodotus makes nine references to emporia.91 One of these is the Milesian colony, Olbia, in the Black Sea.92 The physical location of this settlement on the mouth of the River Bug would have placed it ideally for trade with the neighbouring Scythians, and this was arguably its initial purpose. However, by the sixth and early fifth centuries BC, the town was minting its own (unusual, dolphin- shaped) coinage, and had an elaborate town grid plan laid out, neither of which is a trait of a community solely constructed for commerce.93 Looking at the written evidence, Herodotus (4.17.1) states that Olbia is the emporion of Borysthenes.94 However, slightly later, Herodotus (4.18.1) recognises that the inhabitants of Olbia called themselves Ὀλβιοπολίτας, which therefore implies that the people of Olbia believed their city to be a polis in its own right. Clearly Herodotus’ use of the terms emporion and polis demonstrates that the two are not mutually exclusive. Instead, Wilson suggests that when Herodotus describes a polis as being an emporion, he is merely stressing the significance of trade to that city.95 26 This would certainly appear to be the case for settlements such as Naucratis, a settlement which is largely agreed to be an emporion,96 which lacked both a mother city and a permanent Greek settlement, and which instead focused its efforts on trade. Accordingly, it would be easy to assume that Naucratis was definitely not an apoikia, and was solely an emporion. For example, Herodotus (2.179) calls Naucratis an emporion. However, Naucratis is called a polis in the first line of a fourth century BC honorary decree (OGIS 120): ἡ πόλις ἡ Ναυκρατιτ̣[ῶν]. Thus, despite the fact that this settlement that appears exclusively to fit the definition of an emporion, ancient sources (though admittedly much later than the period directly being dealt with here) do not follow suit so closely. Again, we are left to conclude that the emporion and the polis are not mutually exclusive terms and that their meanings overlap. It is crucial to note that the word emporion itself does not in fact occur in any extant archaic text.97 Herodotus, though one of the earliest sources available to modern historians, was writing in the fifth century BC. The oldest extant inscriptions in which the word ‘emporion’ occurs also date from the mid-fifth century BC. 96 As opposed to settlements such as Pithekoussai where there is continued debate as to its emporion status. See especially Ridgway (1973) 107-120; also Wilson (2006) 34; Graham (1982) 103; D’Agostino (1996) 535-536; Mertens & Greco (1996) 243. ; g ( ) ; ( ) 97 Hansen (1997) 84. This article was reprinted in Tsetskhladze (2006a), with the primary difference being that in the latter publication, Hansen explicated his belief that emporia did exist in the archaic period, making his argument much clearer. 98 p , g g 98 These epigraphical examples were suggested by Hansen (1997) 84. 99 ( ) p g p 99 Hansen (1997) 94. Different types of settlements For example, two boundary stones found in the Piraeus at Athens are both inscribed ἐµπορί[ο] καὶ hοδο̑ hόρος. (IG I³ 1101 A and B). The word emporion is also found in a fragmentary decree presumably regulating the foundation of a colony: τοις εµπορι[οις] (IG I³ 47 A.7).98 Apart from this evidence from the fifth century BC, we have no extant material to indicate that the Greeks had developed the concept of an emporion prior to this. The one exception to this is the name of the archaic colony called ‘Emporion’ (Spanish: Ampurias) in Spain.99 Founded in c. 550 BC, the name of the settlement, Emporion, seems to be the earliest reference to the term (here used as a proper noun), roughly a century before it appears in Herodotus. Thus, we are left with a ‘chicken and egg’ conundrum; did the term come from the colony, or was the 27 colony named after the term? Demetriou’s analysis of the cadastral studies of Emporion helps answer some of these questions.100 Originally, as has been discussed above, scholars assumed that emporia would have had much smaller areas of territory than apoikia, relying on trade rather than agriculture for survival. Certainly at Emporion, it appeared that the settlement itself was small (roughly four hectares) and had a quite limited territory.101 However, the cadastral studies present a different story. Roads ran outwards from the settlement to the country. Furthermore, much of the surrounding country was divided up into kleroi (parcels of land), which according to Demetriou were marked out following the same measurements used in the mid-sixth century BC in mainland Greece.102 Given the corresponding measurements, the kleroi at Emporion may have existed since the founding of the settlement.103 It is not so clear however, whether the indigenous population were included in the distribution, nor how long the kleroi were maintained.104 Hodos claims that the Greeks were not consistent in their use of the terms apoikia and emporion. 105 This appears to have been because the Greeks did not make significant distinctions between the various population movements.106 This in turn suggests that the Greeks did not give the same importance to exact definitions and applications of the two terms. If that is so, the confusion in the ancient language is in fact more recent, and is heavily tied in to the confusion in the modern definitions as well. ( ) 103 We could perhaps assume from this that each of the colonists started on an even par at the beginning of the settlement. Because we know so little about how Greek colonists were selected for colonial voyages, we know little about the status of colonists. Further exploration in this area, using such cadastral studies, could be a good place to make headway in this debate. 104 ( ) 106 Wilson (2006) 29. 100 Demetriou (2011) 263-264. 101 Demetriou (2011) 264. 102 Demetriou (2011) 264. 103 We could perhaps assume from this that each of the colonists started on an even par at the beginning of the settlement. Because we know so little about how Greek colonists were selected for colonial voyages, we know little about the status of colonists. Further exploration in this area, using such cadastral studies, could be a good place to make headway in this debate. 104 For example, contrast to the Spartan land tenure system, where we are equally uncertain about how long the kleros system remained unchanged following the land reforms of Lycurgus. We know that they must have changed at some stage, as some Spartans were in possession of more than one kleros, even if they maintained their original proportions. As with the Spartan system, such a land tenure system does not necessarily provide that everyone will receive land of similar quality, even if the quantities were roughly equal. 105 Hodos (2006) 19. 106 Wilson (2006) 29. 107 T t khl d (2006 ) iii 104 For example, contrast to the Spartan land tenure system, where we are equally uncertain about how long the kleros system remained unchanged following the land reforms of Lycurgus. We know that they must have changed at some stage, as some Spartans were in possession of more than one kleros, even if they maintained their original proportions. As with the Spartan system, such a land tenure system does not necessarily provide that everyone will receive land of similar quality, even if the quantities were roughly equal. 107 Tsetskhladze (2006c) xxviii. 105 Hodos (2006) 19. 101 Demetriou (2011) 264. 102 102 Demetriou (2011) 264. 103 100 Demetriou (2011) 263-264. 101 Demetriou (2011) 264. 102 Demetriou (2011) 264. 103 Different types of settlements Tsetskhladze sensibly remarks: “terminology can never reflect the full reality – it can illuminate or distort in equal measure”.107 For this reason, the distinction between an apoikia and an emporion does little to illuminate the position of women. 28 Conclusion The foregoing demonstrates that colonisation remains a difficult concept to define. This is partly because it is not entirely clear what the Greek colonisation process involved, but also largely because the term has become entwined with other meanings over a long period of use. It has become a multifaceted term which can be attached to a number of different processes. Greek colonisation has become conflated with our modern concepts of colonisation, especially that undertaken by the British. This has given Greek colonisation an imperialist flavour, and coloured views about the way women and any indigenous population are considered. As a result, the study of Greek colonisation has undergone many changes in emphasis and interpretation throughout the twentieth century. Though we are not entirely clear what the Greek colonisation process involved, we can be certain that food, water, and shelter would have been necessary to ensure survival. Similarly, an ability to reproduce would have been crucial to the continuation of any colony. There is significant uncertainty about the type of transport used to reach the new settlements. The textual evidence, in addition to the various studies completed on the Black Sea, and arguments from logic, strongly suggest that pentekonters were used to transport colonists. However, this does not correspond with our understanding of the pentekonter as a long, shallow vessel with no storage space. Until this issue is resolved, we can only speculate about whether (or perhaps, how) ships may have transported women and basic possessions to the colonies. There also appear to be many diverse reasons for founding colonies. No one reason can be adduced from the evidence as dominant, and in all likelihood, each colony had a different founding impetus from the next one. Thus, some explanations are consistent with the presence of women in colonial voyages, whereas others indicate that women would more likely be found from within indigenous populations. Similarly, the different types of settlement – apoikia versus the emporion – could determine whether women were involved and how their involvement played out. In reality, however, it appears that the distinction commonly drawn between these different settlement types is one that did not necessarily exist in the ancient world. 108 Of course, a mixture of these options could have occurred – some Greek wives accompanied their husbands, while other men intermarried. They are not mutually exclusive, though they do tend to be presented in the scholarship as such. 109 INTERMARRIAGE: A NORMAL PRACTICE? This chapter introduces the concept of intermarriage and assesses whether the Greeks considered intermarriage a normal practice, both in the context of colonisation and outside. The evidence typically put forward to argue both for and against intermarriage in the archaic Greek colonies will then be considered. Conclusion Accordingly, the different settlement types 29 may not have been a strong determinant of the presence of women, and in that case, an assessment of settlement types does little to further the issue of whether Greek colonists brought Greek women or intermarried with the indigenous population. may not have been a strong determinant of the presence of women, and in that case, an assessment of settlement types does little to further the issue of whether Greek colonists brought Greek women or intermarried with the indigenous population. 30 p p 109 Graham (1988) 304. p 109 Graham (1988) 304. See above footnote 16. 112 Rouge (1970) 307. See Just’s (1989) 40-75 chapter on ‘Marriage and the State’ for the distinctions between different types of women. We must be mindful however, that many of the terminological distinctions only became prevalent in the classical period. g y p p 113 Herodotus provides evidence for this view, explaining much of his Histories as avenging or retaliating against insults towards women (1.1-5, 1.7-12, 1.61, 3.3, 3.50, and 5.92 to name but a few examples). The most widely known of all Greek wars, the Trojan War, was initially caused by the capture of Helen by Paris. This action, according to Herodotus (1.1-3), was inspired by a series of previous abductions of Io, Europa, and Medea. 114 Rihll (1993) 79 110 Shepherd (1999) 292. p ( ) 111 See above footnote 16. 112 g y p p 113 Herodotus provides evidence for this view, explaining much of his Histories as avenging or retaliating against insults towards women (1.1-5, 1.7-12, 1.61, 3.3, 3.50, and 5.92 to name but a few examples). The most widely known of all Greek wars, the Trojan War, was initially caused by 110 Shepherd (1999) 292. 111 See above footnote 16. ‘Intermarriage’ as a concept For the purposes of this thesis, I am choosing to view foundation scenarios in relatively black and white terms to simplify an already very complex issue. Specifically, I will consider whether male Greek colonists brought their own wives with them (or by extension other Greek females), or whether male Greek colonists intermarried with indigenous females.108 Crucially, I am assuming that the majority of Greek colonists would have always been male, and therefore, that intermarriage only took place between Greek males and indigenous females, rather than the other way around. My focus is largely on first generation colonists. ‘Intermarriage’ encompasses both consensual and non-consensual relationships, or rather, marriages produced and recognised in peaceful circumstances as well as those that were a product of force. It is also important to acknowledge that the nature of such marriages could be considerably more casual than our modern definition of marriage. A study of the relationships between the Greek colonists and the various indigenous peoples is essential to this discussion. This is not always straightforward. Graham aptly states: of the many deficiencies in our evidence for Greek colonization in the archaic period, one of the worst is the lack of good information about the native peoples among whom the Greeks established their new settlements.109 31 Nevertheless, it is useful to consider whether relations between the Greeks and the indigenous populations were hostile or friendly, particularly on the arrival of the colonists, to indicate whether intermarriage could have been an option. Of course, as Shepherd points out: “good and bad relations with the indigenous population do not necessarily facilitate or preclude intermarriage”.110 It is also important to consider whether the Greeks themselves perceived intermarriage as being a normal practice. As was briefly discussed in Chapter One, many scholars argue that the Greeks did not attach any great stigma to the barbaroi, or see them through a racist lens.111 Consequently, it is unlikely that Greek men would have felt any reluctance towards taking a non-Greek wife.112 Further, Greek traditions long recognised that wars were often started over women, and that women were frequently part of the plunder.113 As Rihll argues: “the Greeks considered taking, by force if necessary, to be a normal and legitmate method of acquisition”.114 Taking women by force was simply an extension of this behaviour. This attitude is made clear in Homer’s Odyssey (9.39-42). Ἰλιόθεν µε φέρων ἄνεµος Κικόνεσσι πέλασσεν, Ἰσµάρῳ. ‘Intermarriage’ as a concept ἔνθα δ᾽ ἐγὼ πόλιν ἔπραθον, ὤλεσα δ᾽ αὐτούς: ἐκ πόλιος δ᾽ ἀλόχους καὶ κτήµατα πολλὰ λαβόντες δασσάµεθ᾽, ὡς µή τίς µοι ἀτεµβόµενος κίοι ἴσης. From Troy the wind carried me and brought me to Ismarus of the Cicones. There I sacked the city and I destroyed the men. Having taken the wives and many possessions from the city, we divided them so that no one was cheated of an equal share. From Troy the wind carried me and brought me to Ismarus of the Cicones. There I sacked the city and I destroyed the men. Having taken the wives and many possessions from the city, we divided them so that no one was cheated of an equal share. Other ancient peoples shared this attitude. For example Dionysius of Halicarnassus (Antiquitates Romanae 2.30.5) states that when Romulus and his companions seized and married the Sabine virgins (usually referred to as the ‘Rape 32 of the Sabine Women’), these actions were justified on the basis that this was a traditional method used by the Greeks for acquiring wives.115 τῇ δ᾽ ἑξῆς ἡµέρᾳ προαχθεισῶν τῶν παρθένων, παραµυθησάµενος αὐτῶν τὴν ἀθυµίαν ὁ Ῥωµύλος, ὡς οὐκ ἐφ᾽ ὕβρει τῆς ἁρπαγῆς ἀλλ᾽ ἐπὶ γάµῳ γενοµένης, Ἑλληνικόν τε καὶ ἀρχαῖον ἀποφαίνων τὸ ἔθος καὶ τρόπων συµπάντων καθ᾽ οὓς συνάπτονται γάµοι ταῖς γυναιξὶν ἐπιφανέστατον. The next day, having brought forward the maidens, Romulus comforted them in their despair, as they had not been seized out of wanton violence but of becoming wedded. He pointed out that this was an ancient Greek custom and that of all methods of contracting marriage for women, it was the most distinguished. Greaves has suggested that: “it may be not a marriage ceremony that was the Greek tradition which Romulus adopted but rather the violent abduction of wives as part of an act of colonization”.116 Livy’s (Ab Urbe 1.9.14-15) account records how Romulus approached his neighbours to ask for rights of intermarriage so that he could ensure Rome’s future beyond the current generation.117 When they refused, he arranged to trick the Sabine women to come into the city to watch the Consualia games where they were seized by the men.118 sed ipse Romulus circumibat docebatque patrum id superbia factum, qui conubium finitimis negassent; illas tamen in matrimonio, in societate fortunarum omnium civitatisque, et quo nihil carius humano generi sit, liberum fore. 115 Greaves (1998) 572. See cover image for a graphic representation of this event by Jacques- Louis David, painted in 1799. I am not suggesting that this event was necessarily historical, however, the perceived attitude conveyed through the myth about the Greeks’ behaviour is interesting to note. 116 G (1998) 573 118 rapio, rapere meaning ‘to seize’ rather than ‘to rape’ which is the terminology normally given to the story (OLD sv. rapio). 119 This suggests that it was not a mixed marriage, per se, but a legally valid partnership resulting in citizen children. Greaves (1998) 573. 117 Van Compernolle (1983) 1042 suggests that Livy draws upon a Greek colonial tradition. 118 g 116 Greaves (1998) 573. 11 116 Greaves (1998) 573. 117 Van Compernolle (1983) 1042 suggests that Livy draws upon a Greek colonial tradition. 116 Greaves (1998) 573. 117 V C ll (1983) 1042 t th t Li d G k l i l t diti ‘Intermarriage’ as a concept But Romulus himself went around them [the women] and explained that the pride of their parents had caused this deed, when they had refused their neighbours intermarriage, but that, they would be in a state of marriage, in a partnership of all fortunes and citizenship, and, what is dearer than anything to the human race, this would be a free union.119 Violence associated with the act of marriage was not uncommon. For example, Plutarch (Lycurgus 15.3), reports how the Spartan marriage ceremony involved the 121 Dougherty (1993a) 63. Dougherty also points out the similarities in the depiction of marriage and death in the iconographic tradition of Greek vase painting. Indeed, as a process, marriage can be seen to follow Van Gennep’s (1960) ‘Rite of Passage’ tripartite structure, that can also be applied to other passages in life such as birth or death (see Morris (1992) 9-10). Marriage generally can be interpreted as a method of integration of two people – the ἀνακαλυπτήρια where the groom lifts the bride’s veil and sees her face for the ‘first’ time marks how marriage as a ritual brings together two strangers (see Sissa (1990) 94-99; Oakley & Sinos (1993) 25). 120 Greaves (1998) 572-573 argued that there were strong similarities between Romulus’ justification in the Rape of Sabines and the Spartan marriage ceremony. Although Plutarch wrote after Dionysius, it is possible that both authors had access to a common source which could account for the similarities. g y ( ) 123 In his preface to The Western Greeks, Dunbabin (1948) vi wrote “I am inclined to stress the purity of Greek culture in the colonial cities” thus demonstrating his reluctance to recognise any indigenous influence in the colonies. Violence associated with the act of marriage was not uncommon. For example, Plutarch (Lycurgus 15.3), reports how the Spartan marriage ceremony involved the 33 formal capture of the wife by the bridegroom.120 The myth about the abduction of Persephone by Hades is commonly referred today as ‘the Rape of Persephone’, and can be seen to establish the archetype of marriage as being synonymous with violence and sexuality.121 Pindar (Pythian 9.1-60) describes the rape of the nymph Cyrene by Apollo. This tradition is often described as symbolising the Greek colonisation of the city by the same name.122 From this variety of sources, we can be fairly certain that the Greeks were not averse to intermarriage. g 124 Gwynn (1918) 109: “For in every sphere of their colonising activity, the Greeks met races which, though socially and intellectually their inferiors, were still, in feature and colour, of the same general type.” g g g 122 Dougherty (2001) 132. 123 125 Dunbabin (1948) 186. Intermarriage versus taking Greek wives Early modern scholarship, however, was frequently reluctant to acknowledge that Greeks abroad might have interacted, let alone intermarried, with the local populations, beyond engaging in warfare.123 The few that did proffer it as an option seemed to feel the need to apologise for the Greeks’ behaviour. For example, Gwynn in 1918 excused intermarriage by citing extenuating circumstances due to remoteness, and stated that it was tolerable because there was an absence of skin-colour differences.124 Dunbabin in 1948 made allowances for intermarriage occurring in those colonies which had a mixture of Greeks from various parts of the mainland in them: “this might dispose them to hold less strongly to the traditions of their mother country…and might also dispose them more readily to intermarry with the natives”.125 With the rise of feminist and multi-cultural debates and theories in the later twentieth century, scholars began to acknowledge women and the possibility of intermarriage more readily (and 125 Dunbabin (1948) 186. 34 reasonably), although frequently their assertions were not supported by any evidence.126 Those who do use evidence typically call on one of two passages from Herodotus, depending on whether they are arguing for or against intermarriage. However, as we will see, neither passage is capable of swaying the argument decisively one way or the other. The first passage of Herodotus (1.146) is interpreted as evidence in favour of intermarriage:127 The first passage of Herodotus (1.146) is interpreted as evidence in favour of intermarriage:127 The first passage of Herodotus (1.146) is interpreted as evidence in favour of intermarriage:127 τούτων δὴ εἵνεκα καὶ οἱ Ἴωνες δυώδεκα πόλιας ἐποιήσαντο: ἐπεὶ ὥς γέ τι µᾶλλον οὗτοι Ἴωνες εἰσὶ τῶν ἄλλων Ἰώνων ἢ κάλλιόν τι γεγόνασι, µωρίη πολλὴ λέγειν: τῶν Ἄβαντες µὲν ἐξ Εὐβοίες εἰσὶ οὐκ ἐλαχίστη µοῖρα, τοῖσι Ἰωνίης µέτα οὐδὲ τοῦ οὐνόµατος οὐδέν, Μινύαι δὲ Ὀρχοµένιοί σφι ἀναµεµίχαται καὶ Καδµεῖοι καὶ Δρύοπες καὶ Φωκέες ἀποδάσµιοι καὶ Μολοσσοὶ καὶ Ἀρκάδες Πελασγοὶ καὶ Δωριέες Ἐπιδαύριοι, ἄλλα τε ἔθνεα πολλὰ ἀναµεµίχαται: οἱ δὲ αὐτῶν ἀπὸ τοῦ πρυτανηίου τοῦ Ἀθηναίων ὁρµηθέντες καὶ νοµίζοντες γενναιότατοι εἶναι Ἰώνων, οὗτοι δὲ οὐ γυναῖκας ἠγάγοντο ἐς τὴν ἀποικίην128 ἀλλὰ Καείρας ἔσχον, τῶν ἐφόνευσαν τοὺς γονέας. διὰ τοῦτὸν δὲ τὸν φόνον αἱ γυναῖκες αὗται νόµον θέµεναι σφίσι αὐτῇσι ὅρκους ἐπήλασαν καὶ παρέδοσαν τῇσι θυγατράσι, µή κοτε ὁµοσιτῆσαι τοῖσι ἀνδράσι µηδὲ οὐνόµατι βῶσαι τὸν ἑωυτῆς ἄνδρα, τοῦδε εἵνεκα ὅτι ἐφόνευσαν σφέων τοὺς πατέρας καὶ ἄνδρας καὶ παῖδας καὶ ἔπειτα ταῦτα ποιήσαντες αὐτῇσι συνοίκεον. 126 See for example, Finley (1968) 18: “it is hardly likely that an adequate number [of women] (if any) were brought from Greece”; Dougherty (1993a) 67: “there is little doubt that intermarriage took place, despite the reticence of the Greeks to mention it”; Freeman (1999) 70: “intermarriage with natives would have been inevitable if the settlement was to endure”. p y ( ) ( ) 128 Herodotus’ use of apoikia demonstrates the inconsistency and the fluidity of the concept discussed in the previous chapter. 127 See for example, Pomeroy (1995) 34; Hodos (1999) 66 127 See for example, Pomeroy (1995) 34; Hodos (1999) 66. 128 Intermarriage versus taking Greek wives For the following reason, the Ionians built twelve cities. It is foolish to say that they did so because they are more Ionian than other Ionians, or in some way better born. The Abantians from Euboia are not the least part of the Ionians, and don’t even have the name of Ionians. The Minyae of Orchomenus have mixed with them, and the Cadmeians, and the Dryopes, and the breakaway Phocaeans, and the Molossians, and the Pelasgian Arcadians, and the Dorians from Epidauros, and many other nations are mixed together. As for those from the prytaneum of Athens, who think themselves to be the purest of Ionians, they took no women to the colony [of Miletus], but took Carian women and killed their parents. Because of this slaughter, the women made a law, swearing an oath, and handed it down to their daughters, never to take meals with their husbands, nor call their husbands by name, on account of the fact that they had killed their 35 fathers and husbands and sons, and after they had done these things, they lived with them.129 This passage clearly describes the foundation of Miletus by Ionian settlers. The foundation of Miletus took place prior to the colonisation of the archaic period and therefore this evidence describes a practice from an earlier occasion. Furthermore, the passage can be interpreted as an aetiological story characteristic of Herodotean writing. The aetiological nature of the story is clearly demonstrated by the oddity of the tale requiring further explanation and justification. The assertion that the law was imposed and enforced by the women themselves and passed on to their own daughters gave Herodotus the necessary connection between the unusual customs which were practised in his own day and events which purported to have happened long before. In this way, Herodotus was able to link the two tales and use the aition to provide explanation. While on the one hand, the very nature of an aetiological tale means that it cannot be taken as factual historical evidence and was certainly at least in part created to explain contemporary customs, on the other hand, it could be argued that the story presents elements of fact or general testimony about Ionian practices. 129 Pausanias (7.2.6) also reports this in his account of the foundation of Miletus; however it is likely that his work is based on Herodotus’, as it includes the same information. 130 Coldstream (1993) 98 Coldstream (1993) 98. 131 Graham (1981-82) 294. ( ) 131 Graham (1981-82) 294. y 130 Coldstream (1993) 98. Intermarriage versus taking Greek wives This passage, Coldstream argues, was intended to demonstrate the extent to which the Ionians of Asia Minor were a considerably mixed population.130 With reference to women specifically, Graham recommends that we look for any indication in the words that Herodotus uses as to whether he perceived the actions of the Ionian settlers to be normal or unusual. The words: οὗτοι δὲ οὐ γυναῖκας ἠγάγοντο ἐς τὴν ἀποικίην ‘they took no women to the colony, but took Carian women and killed their parents’, seem to “imply clearly enough that in Herodotus’ opinion it was normal for colonists to take women with them”. 131 That is, Herodotus felt the need to mention that they did not take women in this case because that was contrary to customary practice. By itself, however, this one passage is not sufficient evidence to conclude that Greek colonists intermarried with the indigenous population. 36 The scholarship frequently uses another passage, also from Herodotus (1.164), as evidence against intermarriage, and in favour of the theory that Greek colonists took Greek wives with them:132 ὁ δὲ Ἅρπαγος ὡς ἐπήλασε τὴν στρατιήν, ἐπολιόρκεε αὐτούς, προισχόµενος ἔπεα ὥς οἱ καταχρᾷ εἰ βούλονται Φωκαιέες προµαχεῶνα ἕνα µοῦνον τοῦ τείχεος ἐρεῖψαι καὶ οἴκηµα ἓν κατιρῶσαι. οἱ δὲ Φωκαιέες περιηµεκτέοντες τῇ δουλοσύνῃ ἔφασαν θέλειν βουλεύσασθαι ἡµέρην µίαν καὶ ἔπειτα ὑποκρινέεσθαι: ἐν ᾧ δὲ βουλεύονται αὐτοί, ἀπαγαγεῖν ἐκεῖνον ἐκέλευον τὴν στρατιὴν ἀπὸ τοῦ τείχεος. ὁ δ᾽ Ἅρπαγος ἔφη εἰδέναι µὲν εὖ τὰ ἐκεῖνοι µέλλοιεν ποιέειν, ὅµως δὲ σφι παριέναι βουλεύσασθαι. ἐν ᾧ ὦν ὁ Ἅρπαγος ἀπὸ τοῦ τείχεος ἀπήγαγε τὴν στρατιήν, οἱ Φωκαιέες ἐν τούτῳ κατασπάσαντες τὰς πεντηκοντέρους, ἐσθέµενοι τέκνα καὶ γυναῖκας καὶ ἔπιπλα πάντα, πρὸς δὲ καὶ τὰ ἀγάλµατα τὰ ἐν τῶν ἱρῶν καὶ τὰ ἄλλα ἀναθήµατα, χωρὶς ὅ τι χαλκὸς ἢ λίθος ἢ γραφὴ ἦν, τὰ δὲ ἄλλα πάντα ἐσθέντες καὶ αὐτοὶ εἰσβάντες ἔπλεον ἐπὶ Χίου. τὴν δὲ Φωκαίην ἐρηµωθεῖσαν ἀνδρῶν ἔσχον οἱ Πέρσαι. Harpagus, having driven his army against them [the Phocaeans] and besieged them, offered a proposal; that it would satisfy him if the Phocaeans were willing to demolish one battlement of the wall, and to consecrate one house. But the Phocaeans, aggrieved at the thought of slavery, said they wanted to deliberate for one day, and then they would give an answer. While they were deliberating, they asked him to withdraw his army from the wall. Harpagus knew what they were about to do, nevertheless, he allowed them to deliberate. p g 133 Hall (2004) 80; Rouge (1970) 312. 132 Graham (1981-82) 300. See also Hall (2004) 40 and Shepherd (1999) 270 who discuss this passage but discount it for the same reasons I do. 133 132 Graham (1981-82) 300. See also Hall (2004) 40 and Shepherd (1999) 270 who discuss this p g 133 Hall (2004) 80; Rouge (1970) 312. Intermarriage versus taking Greek wives While Harpagus led his army from the wall, the Phocaeans, in the meantime, pulled down the pentekonters, and put in the children, and women, and all moveable property, as well as the statues from the temples, and the other dedications, except for what was made of bronze, or stone, or painted, and once they had loaded everything else, they themselves embarked and sailed for Chios. The Persians took Phocaea, stripped bare of men. This passage is the only one we have where women are specifically described as being present on the boat leaving to a new homeland. However, this passage in fact describes the total evacuation of the city of Phocaea in c. 540 BC following the Persian invasion.133 The Phocaeans chose emigration over submission, wishing to avoid becoming Persian slaves or subjects. Therefore, the primary intention underlying the move to a new land (Elea in Italy) was to avoid enslavement and this drove the need for a complete evacuation and migration of the entire 37 population. This diverges strongly from the usual practice of Greek colonisation, where instead of the entire population departing, the colonists were limited (in some way or another) to a smaller section or group of the population. Therefore, this evidence is not wholly applicable to colonisation, and should not be used as the sole basis of an argument against intermarriage or in favour of wives and husbands travelling together to found colonies. CHAPTER THREE CHAPTER THREE Introduction This chapter examines ten different Greek colonies of the archaic period, grouped by region. The primary focus is on the foundation of these colonies with attention given to the first (or earliest) generation of colonists so that we might better establish whether Greek women were present from the outset, or whether intermarriage seems more likely. Where possible, both literary and archaeological evidence is considered hand in hand. In addition to providing general background, a brief analysis of the various types of settlements as well as the reasons colonies were founded could help contribute to assessing whether women were taken on the initial voyage to found the colony. Thus, if a settlement was focused on trade above all else, how necessary would women have been (or at all, even once the settlement had become established)? If colonists set out with the specific intention of gaining land or resources, surely their settlement process would have been more territorially aggressive, and thus more like warlike conquests in which it would not be appropriate to involve women. 134 Buchner’s (1966) article sets out to prove that Pithekoussai is the oldest Greek colony in the 134 Buchner’s (1966) article sets out to prove that Pithekoussai is the oldest Greek colony in th western Mediterranean. See also Hodos (1999) 61. 135 B h (1966) 4 Conclusion In the second half of the twentieth century and continuing into the twenty-first, the weight of the scholarly opinion supported the view that only men founded colonies and therefore intermarried with indigenous women on arrival. Often, however, scholars asserted bald statements supported by little evidence. Alternatively, two passages from Herodotus are frequently called upon to argue for (1.146) or against (1.164) intermarriage. An examination of these two sources illustrates that neither source is capable of conclusively resolving the argument one way or another (despite many attempts to do so). 38 ( ) 135 Buchner (1966) 4. 135 Buchner (1966) 4. ( ) g ( ) ( ) 138 Livy (8.22.5-6) claims that the colonists first settled at Pithekoussai, then moved to the mainland to establish Cumae, whereas Strabo (5.4.4) argues that a separate group of colonists left to found Cumae directly from Euboea. See Hodos (1999) 62; Graham (1982) 101. y ( ) ; ( ) 139 Becker (1995) 273-274; Buchner (1966) 5; Klein (1972) 34. 140 Cu ae d ec y o uboea. See odos ( 999) 6 ; G a a ( 98 ) 139 Becker (1995) 273-274; Buchner (1966) 5; Klein (1972) 34. Graham (1982) 99. 137 See especially Ridgway (1973) 107-120; also Wilson (2006) 34; Cook (1962) 113-114; Graham (1982) 103; D’Agostino (1996) 535-536; Mertens & Greco (1996) 243. ( ) ; ( ) 141 Shepherd (1999) 276; Becker (1995) 276. p ( ) 142 Buchner (1966) 5. Becker (1995) 273 274; Buchner (1966) 5; 140 Becker (1995) 273; Coldstream (1994) 51. ( ) ; ( ) ; ( ) 140 Becker (1995) 273; Coldstream (1994) 51. ( ) 143 These are published in Italian by Buchner & Ridgway (1993). See also Boardman (1994) 95. 144 Ridgway (1992) 101 136 Graham (1982) 99. 137 ( ) 143 These are published in Italian by Buchner & Ridgway (1993). See also Board 144 ( ) 143 These are published in Italian by Buchner & Ridgway (1993). See also Boardman (1994) 95. 144 Ridgway (1992) 101. 144 Ridgway (1992) 101. 136 Graham (1982) 99. 137 ( ) ; ( ) 41 Shepherd (1999) 276; Becker (1995) 276. 42 B h (1966) 5 Shepherd (1999) 276; Becker (1995) 276. 142 Buchner (1966) 5. 141 Shepherd (1999) 276; Becker (1995) 276. 142 Buchner (1966) 5. Pithekoussai Pithekoussai, an island in the Bay of Naples, has been described as the oldest Greek colony in the Western Mediterranean.134 According to Strabo (5.4.9), it was founded in the eighth century BC by colonists from the two main cities on the island of Euboea: Eretria and Chalcis.135 The site has not revealed any Iron Age material which suggests that the island (or at least the area of settlement) was 39 uninhabited prior to the arrival of the Greeks.136 The reasons behind Pithekoussai’s foundation are not obvious. Further, there has been continued debate over whether Pithekoussai was an emporion, simply a trading station, rather than an apoikia.137 Pithekoussai could have been an island settlement that laid the groundwork for the establishment of Cumae on the mainland a generation later, as was a common practice for colonies in their initial years.138 Although the sites of the ancient town, cemetery, and acropolis were identified by a local scholar in the nineteenth century, controlled archaeological excavations did not begin until 1952.139 Intensive research was conducted by Giorgio Buchner for a number of years. Excavations began in the cemetery, and from very early on, it was apparent that the soil conditions at Pithekoussai have negatively affected the majority of the inhumed skeletons. Heat from thermal activity has reduced most bones to powder,140 so that scholars have been largely unable to carry out any osteological analysis, particularly cranial morphology which is a typical method of assessing gender.141 This has not deterred excavation of burial receptacles and grave goods. The main period of excavation ran from 1952-1961, and in that time, an area of approximately a thousand square metres was examined.142 However, this makes up less than 10% of the whole cemetery at San Montano, Pithekoussai. A total of 1,300 graves has been excavated, but information on only a portion of these, 723, has been fully published.143 Furthermore, of the latter, only 493 are Geometric – roughly a mere 2.5% of the total cemetery.144 To complicate matters further, to answer the question about the origins of women in Greek colonies, only very early graves should be examined (those potentially belonging to the first generation of 40 colonists).145 How reliable might it be, however, to base a hypothesis on such a small proportion of the entire cemetery?146 Much of the focus of these excavations has been on the fibulae found in many of the graves. 145 Ridgway (1992) 101-103 estimates that the total population of Pithekoussai in the late eighth century BC could have been between 5,000-10,000. If this was indeed the case, it suggests that there were both a relatively large number of colonists in the initial colonising party, as well as multiple groups joining the settlement. Osborne (1998) 258 argues towards a more mobile population model such as that of Naucratis, with a focus on profit and trade. 146 p y ( ) 147 These Italic fibulae have been found in Italy contemporary to those of Pithekoussai. See Toms (2000). Other fibulae types existed in Greece, but these were primarily used in votive contexts. 148 Hodos (1999) 64. p p , p 146 This was one of the primary concerns of Coldstream (1993) 92. Buchner’s well known, widely accepted hypothesis reads: ( ) 149 Hodos (1999) 69. Pithekoussai Women in mainland Greece are typically described as wearing the Doric peplos which was fastened with two straight pins holding the dress on each shoulder. These pins were usually made of bronze or gold (see Figure 2). The head of the straight pin was often extensively decorated. By contrast, the Italic147 fibulae found in the excavations at Pithekoussai, were made up of a pin, a hinge or spring, and a bow (see Figure 3). These fibulae resemble the modern safety pin. Like the straight pin, they were typically made out of bronze, iron, gold, or a combination of these. The bow provided an increased scope for decoration, and as such, was often inlaid with semi-precious stones, bone, or glass. These Italic fibulae can be divided up into several types: the leech, nevicella, and bone-and- amber types are associated with females, while the serpentine type is associated with males. Examples of each of these types have been found at Pithekoussai. The discovery of these Italic fibulae types at Pithekoussai was significant for two main reasons. First, no fibulae of the same form as those found at Pithekoussai have been found on the Greek mainland.148 Second, at Pithekoussai, so far only a total of 12 straight pins have been discovered. This suggests that the women who lived at Pithekoussai did not wear the Doric peplos that was common among the Greeks, and in turn, it has therefore been argued that those women were not Greek.149 Buchner’s well known, widely accepted hypothesis reads: 41 It cannot possibly have been the men who set the fashion for indigenous personal ornaments…it must have been the women. It follows that most, if not all, of the colonists’ women were not Greeks but natives – who were not prepared to abandon the haberdashery to which they were accustomed.150 This is too firm a conclusion for the evidence that we have. There is no source of metal anywhere on the island of Pithekoussai. Strabo (5.4.9) claimed that gold mines were present and that these were one of the initial motivating factors for the Euboean colonists to settle there. We now know that this is geologically impossible, and no other metal is found there either.151 However, Buchner’s excavations of the area of habitation in the Mazzola area uncovered a number of significant buildings, connected in some way with metal working (see Figure 4). 150 Buchner (1979) 135. This quotation comes from a later publication of Buchner’s, slightly adapted from the original and translated into English (hence is the version I use here). His original hypothesis can be found in Buchner, G. (1975) ‘Nuovi aspetti e problemi posti dagli di Pitecusa’, 59-86. 151 151 Cook (1962) 114. ( ) 152 Buchner (1970-71) 66. 1 3 153 Buchner (1970-71) 66. ( ) 154 Buchner (1979) 135. ( ) 160 See Becker (1992) and (1995). ‘Sex’ and ‘gender’ tend to be used as interchangeable terms (as indeed is demonstrated by Becker). This, however, is incorrect. Sex is a biological category, genetically controlled, whereas gender is a cultural construct that can vary between social classes, as well as across time and space – a social identity. See Liston (2012) 127-128; Walker & Cook (1998) 255-256; White & Folkens (2005) 385. Pithekoussai The building known as Structure III has been described as a blacksmith’s workshop. Half was covered by a roof, with the other half acting as an open courtyard. The middle of this ‘courtyard’ appeared to be heavily burnt, which led Buchner to conclude that it was the site of a forge.152 Structure IV also has a forge, as well as a number of bits of iron and other metals which had accumulated in and outside the building.153 One of the most important finds in Structure IV was a miscast fibula (see Figure 5). This evidence suggests that these Italic fibulae were manufactured on Pithekoussai itself.154 The simple fact that there was no source of metal on the island, however, suggests that metal must have been sourced elsewhere and subsequently processed or manufactured into fibulae locally. This would have involved trade, which points to some sort of relationship between the Greek colonists at Pithekoussai and one or more indigenous populations in Italy, to enable the continued production of fibulae 42 locally.155 We cannot directly conclude from this, however, that intermarriage occurred at Pithekoussai. I further question whether a mere sex-association is enough to draw the conclusion Buchner does; that is, because the fibulae were supposedly used by women, it was women who dictated the types used. A recent study by Shepherd of the fibulae has emphasised that they more commonly occur in the graves of children rather than adult women.156 She has also demonstrated that fibulae were often found in quantity, and as such could be considered as grave offerings rather than every day functional objects used with indigenous dress. On the basis of Shepherd’s study, it is not possible to reach the same definite conclusion as Buchner. Clearly fibulae served not only a practical purpose, but could also be used as grave offerings.157 Furthermore, the accumulation of fibulae in some graves suggests that they could be interpreted more as indicators of class, status, or age than as markers of ethnicity or identity.158 Coldstream also suggests that a case could be made for Greek women “making do with what was locally available”.159 This is certainly plausible, and again calls into question the problematic connection between object and ethnicity. 157 I acknowledge that fibulae were used in a votive way, even on mainland Greece, but in this thesis, I am solely considering fibulae in the context of the grave. Accordingly, I am assuming their practical use through their positioning relative to the deceased, so where pins were found on the shoulders, it is possible to conclude that the body was buried wearing the straight pins. This contrasts to graves where fibulae were found in multiple numbers and were not necessarily positioned on the body, and so in those instances, the deceased were almost certainly not wearing the pins, and accordingly it cannot be said they had a practical purpose but instead served another role as grave goods. 158 ( ) 156 Shepherd (2005) 117. This study and Shepherd (1999) admittedly take into consideration fibulae throughout the Greek colonies, rather than those just at Pithekoussai. This is an important consideration, however, which will be discussed again. 159 Coldstream (1993) 93. 160 g g 158 Hodos (1999) 67. Hodos (1999) 73. 156 Shepherd (2005) 117. This study and Shepherd (1999) admittedly take into consideration fibulae throughout the Greek colonies, rather than those just at Pithekoussai. This is an important consideration, however, which will be discussed again. 157 I acknowledge that fibulae were used in a votive way, even on mainland Greece, but in this thesis, I am solely considering fibulae in the context of the grave. Accordingly, I am assuming their practical use through their positioning relative to the deceased, so where pins were found on the shoulders, it is possible to conclude that the body was buried wearing the straight pins. This contrasts to graves where fibulae were found in multiple numbers and were not necessarily positioned on the body, and so in those instances, the deceased were almost certainly not wearing the pins, and accordingly it cannot be said they had a practical purpose but instead served another role as grave goods. 158 d (1999) 6 155 Hodos (1999) 73. Pithekoussai In the 1990s, Becker carried out various studies evaluating the sex of the deceased in the necropolis of Pithekoussai.160 He undertook a new approach which involved looking at the pieces of bone surviving the cremation process, coupled with objects associated with the burial, primarily grave goods and the type of 43 burial receptacle. An assessment of sex was first made solely based on bone analysis, and then again solely based on grave goods. Finally, both assessments were combined and compared. In the majority of cases, the two methods of sex analysis produced the same outcome: overall from these findings, Becker’s study gave a sex ratio of adult males to adult females of approximately 40:60.161 It is not clear from his published work, however, whether the cremations analysed were linked only to the very early graves. Becker has proposed that this higher ratio of females to males could suggest a high incidence of polygyny or the presence of numerous female slaves, though certainly more research is required in this area.162 As we will see later, Pithekoussai is not the only colony where this sex imbalance exists in favour of women. Becker’s study was hugely important for confirming the sex of those deceased, previously made solely on the basis of fibulae types. Consistent with wider practice (particularly Etruscan), the serpentine fibulae found at Pithekoussai were (at least in the majority of cases) found with men. Interestingly, I have not yet seen any scholar suggest, in line with Buchner’s hypothesis, that these could be indigenous men. Shepherd asks: “are these the off- spring of a mixed society? Or do we have a case of indigenous wives organising their Greek husbands’ wardrobes?”163 This further demonstrates the weaknesses of Buchner’s hypothesis, and shows that inferring ethnicity and identity from objects is risky. In the case of Pithekoussai, the archaeological discoveries have been interpreted as evidence of intermarriage. The plethora of Italic fibulae which were clearly manufactured locally from metal that must have been obtained by trade with other populations, are certainly an indication of Greek to indigenous contact. 161 Becker (1995) 276. ( ) 163 Shepherd (1999) 294. ( ) 162 Becker (1995) 276. p ( ) 164 Buchner (1979) 135. Pithekoussai These have been used by Buchner to argue that, given their Italic type, they must have been used by Italic (and therefore indigenous) women.164 The strength of this reasoning, however, is weakened by several factors: only a small part of the cemetery has been excavated and so other graves may yet yield counter-evidence; in some cases the fibulae appear to be used as grave goods rather than explicitly to fasten clothing; finally, there are instances of fibulae being found in the graves of 44 children rather than adult women, adding to the argument that they were not necessarily items used only for a practical purpose. Therefore, although it is not possible to apply Buchner’s hypothesis, there is not strong nor sufficient evidence that Greek women were taken to Pithekoussai either. On balance, it seems more likely that intermarriage occurred at Pithekoussai but we can in no way definitively conclude this from the present evidence. Shepherd (1995) 52. 166 Graham (1982) 105; Wilson (1981-82) 82. Sicily has three different indigenous populations (see Pseudo-Skylax 13.2) of which the majority appear to be the Sicels, occupying the eastern side of the island. Syracuse Syracuse, located on the eastern side of Sicily, was a colony of Corinth founded c. 734/3 BC.165 Thucydides (6.3.2) states that Archias, a Heraclid from Corinth, founded Syracuse. 165 Shepherd (1995) 52. 167 Wilson (1987-88) 111. 165 Shepherd (1995) 52. ( of the island. 167 of the island. 167 Wilson (1987-88) 111. Syracuse, located on the eastern side of Sicily, was a colony of Corinth founded c. 734/3 BC.165 Thucydides (6.3.2) states that Archias, a Heraclid from Corinth, founded Syracuse. Συρακούσας δὲ τοῦ ἐχοµένου ἔτους Ἀρχίας τῶν Ἡρακλειδῶν ἐκ Κορίνθου ᾤκισε, Σικελοὺς ἐξελάσας πρῶτον ἐκ τῆς νήσου ἐν ᾗ νῦν οὐκέτι περικλυζοµένῃ ἡ πόλις ἡ ἐντός ἐστιν. The following year, Archias of the Heraclids of Corinth, founded Syracuse, first driving the Sicels off the island on which the city is now, but is no longer surrounded by water. This has been largely confirmed by archaeological excavation, where the first houses of the colonists have been found placed directly on top of the remains of the Sicel village.166 However, a Siculan hut was uncovered more recently which appears to have been in use in the late eighth century BC.167 This perhaps suggests that the expulsion of the indigenous Sicels may not have been as thorough as Thucydides suggests. Thucydides makes very clear that the Sicels living on Ortygia, the offshore island of Syracuse, were driven away by the Corinthian colonists on their arrival. However, from Herodotus (7.155.2) it is evident that, at some time, many of these Sicels were also subjugated and turned into killyrioi: 45 µετὰ δὲ τοῦτο τὸ εὕρηµα τοὺς γαµόρους καλεοµένους τῶν Συρηκοσίων ἐκπεσόντας ὑπό τε τοῦ δήµου καὶ τῶν σφετέρων δούλων, καλεοµένων δὲ Κυλλυρίων, ὁ Γέλων καταγαγὼν τούτους ἐκ Κασµένης πόλιος ἐς τὰς Συρηκούσας ἔσχε καὶ ταύτας: ὁ γὰρ δῆµος ὁ τῶν Συρηκοσίων ἐπιόντι Γέλωνι παραδιδοῖ τὴν πόλιν καὶ ἑωυτόν. The gamoroi had been driven out by the demos and their own slaves, called the killyrioi. After this piece of good luck, Gelon led them from Casmene city into Syracuse, and took possession of it. For, the demos of the Syracusans surrendered the city and themselves to Gelon as he approached. Herodotus reports on events in the fifth century BC when Syracuse was governed by a small group of landed aristocrats (gamoroi). Their slaves, the killyrioi, appear to have been the original indigenous population that was driven away by Archias (as mentioned by Thucydides above). As such, the killyrioi have been compared in status to the Spartan helots (that is, having an intermediary status between free men and chattel slaves),168 and probably also outnumbered their masters.169 Finley assumes (without citing any evidence) that some (female) Sicels were not subjugated, and instead were taken by the colonists as wives.170 Does this mean we should assume that some indigenous women were taken as wives and others as killyrioi? If so, how were the two differentiated? p , y ( ) ; ( ) 172 Obviously this view did not take into consideration the various reasons why colonies were founded. It shows a clear influence of modern colonisation, where the British colonies, in particular, were designed as small slivers of Britain. g p 170 Finley (1968) 20. This argument shows evidence of the period in which it was written with no acknowledgment that seizing wives could be a form of subjugation. To simplify matters, I am assuming that Finley solely differentiates between the end result of a wife versus a slave. p p 169 Garlan (1988) 95. See also Aristotle Constitution of Syracuse (fr. 586 Rose), who compared killyrioi with the helots of Sparta, penestai of Thessaly, and clarotai of Crete. Garlan assumes that such ‘rural slaves’ existed in other colonies (Taras, Locri Epizephyrii, for example) though no evidence confirming their presence has been found. 170 168 Rutter (2002) 138; that being said, the terminology used by Herodotus τῶν δούλων suggests these people were more slave-like. 169 g y y 171 For example, see Holloway (1991) 64; Boardman (1980) 173. 172 173 Shepherd (1995) 52-56. Shepherd’s study argues that the burial practices of colonies generally were not the same as their mother cities. She uses three colonies, Syracuse, Megara Hyblaea, and Gela to demonstrate this. Shepherd (1995) 54. 176 To the contrary, Shepherd (1995) 55 points out that Syracuse has an excellent stone supply so it would have take very little effort to continue using sarcophagi, had the Syracusans wanted. 177 Hodos (1999) 68. 174 Shepherd (1995) 52. This method was used from the eighth to sixth centuries 175 Gela to demonstrate this. 174 Shepherd (1995) 52. This method was used from the eighth to sixth centuries. 175 Sh h d (1995) 54 ( ) 178 Shepherd (1995) 54. p ( ) 179 Hodos (1999) 68. p ( ) 175 Shepherd (1995) 54. ( ) 178 Shepherd (1995) 54. 179 Hodos (1999) 68. Syracuse, located on the eastern side of Sicily, was a colony of Corinth founded c. 734/3 BC.165 Thucydides (6.3.2) states that Archias, a Heraclid from Corinth, founded Syracuse. There is little evidence remaining on the killyrioi, so it is difficult to even begin to speculate. For many years, colonies were assumed to have conformed with the same burial practices of their mother city.171 This followed the notion that each colony was designed as a mirror image of its mother city.172 In 1995, Shepherd produced an important sudy which demonstrated that Syracuse’s burial practices departed 46 radically from those of the mother city, Corinth.173 In the North Cemetery at Corinth, the Corinthians typically buried the dead (of all ages), contracted, in monolithic sarcophagi.174 This contrasts dramatically with the Fusco Cemetery in Syracuse, where Shepherd’s study reveals that, “well over half the adult burials made in the first century or so of the colony’s existence were not in monolithic sarcophagi, and the rate of use of the sarcophagus declines significantly in later periods, resulting in only a very small minority in the later sixth century”.175 Instead, the deceased of Syracuse were typically buried in rock cut fossa graves, or trench graves, covered by stone slabs, though a variety of other methods were also used. The reasons for these changes are not clear. There appear to be no practical (environmental or geological) reasons that could have prevented a continuation of mother city burial practices.176 Hodos argues that Syracuse was attempting to assert cultural independence from the mother city and a departure from familiar burial customs was one way to demonstrate such independence.177 This is not a particularly convincing argument without good reason for the colony to assert its independence (such as bad relations with the mother city or the exile of the colonists). Shepherd tentatively proposed that, for Syracuse and Corinth, the difference in burial methods could be explained if Syracuse retained the burial method that had been used previously (in the Geometric period) at Corinth.178 This lack of consistency could also suggest that the colonists may not have solely originated in Corinth, and instead, brought with them burial practices of their own cities. 183 Buchner (1979) 135. 184 Graham (1981-82) 301; Lorimer (1950) 338; Shepherd (1999) 227. 185 i (19 0) 338 uc e ( 979) 35. 184 Graham (1981-82) 301; Lorimer (1950) 338; Shepherd (1999) 227. 185 185 Lorimer (1950) 338. Lorimer (1950) 338. 186 Graham (1981-82) 301. 181 Holloway (1991) 52. 182 182 Graham (1981-82) 301. 183 y ( ) 182 Graham (1981-82) 301. ( ) 186 Graham (1981-82) 301. ( ) 185 Lorimer (1950) 338. ( ) 186 Graham (1981-82) 301. 182 Graham (1981-82) 301. 183 Buchner (1979) 135 184 Graham (1981-82) 301; Lorimer (1950) 338; Shepherd (1999) 227. 185 Lorimer (1950) 338. 186 G h (1981 82) 301 Syracuse, located on the eastern side of Sicily, was a colony of Corinth founded c. 734/3 BC.165 Thucydides (6.3.2) states that Archias, a Heraclid from Corinth, founded Syracuse. To complicate matters further, it is not even possible to argue that the colonists borrowed burial customs from their indigenous neighbours due to the fact that these differed in tradition yet again.179 47 In the Fusco cemetery, infants’ graves have been found in the tombs of the assumed first (or earliest) generation colonists.180 There are a number of possibilities for this: that the indigenous population was still within the settlement and therefore these people and their infants could be buried within the same cemetery; that Greek women and children also accompanied the male colonists to the settlement; or possibly that intermarriage had occurred and these were the children of Greek men and indigenous women. One example in particular, in seemingly the earliest recorded tomb, revealed a child’s burial, together with a Greek Geometric amphora, and a small bronze horse. Holloway has suggested that considering such objects were usually votive offerings, the child was probably that of an aristocratic Greek.181 The presence of children in a cemetery can be taken as an indication that women were certainly present. Other discoveries in the cemeteries have included the bodies of many (apparent) women who were buried with two straight pins on their shoulders.182 If we were to follow Buchner’s argument, discussed earlier with regard to Pithekoussai, this would surely indicate that these women wore the Doric peplos and that they were therefore Greek women in traditional dress.183 However, a few instances of fibulae of Italic types, associated with indigenous women, have also been discovered in some of the graves.184 By contrast, these pins were not found at the shoulders of the deceased and so a practical purpose seems less likely; rather they appear to have solely served the role of grave offerings.185 Furthermore, many of these fibulae occur in the graves of juveniles, or in a few graves of adults (sex undetermined), seemingly serving as replacements for the straight pins.186 It is therefore very difficult to reconcile these findings with Buchner’s theory about indigenous women, proposed with Pithekoussai in mind. ( ) 188 Hodos (1999) 73 argues that some metal working must have taken place in these colonies to obtain and maintain weapons and other tools, which have not yet been discovered. The metal was almost certainly sourced from other indigenous populations. The lack of a source of metal is an issue throughout Sicily. 189 o e ( 9 ) . 190 Shepherd (1999) 283; Angel (1972) 97. Indeed, Liston (2012) 126 points out that Kurtz & Boardman’s (1971) foundational study, Greek Burial Customs, lacks entries in the index relating to ‘skeleton’ or ‘body’, illustrating the relatively recent shift in importance. 187 Graham (1981-82) 301. 187 Graham (1981-82) 301. Syracuse, located on the eastern side of Sicily, was a colony of Corinth founded c. 734/3 BC.165 Thucydides (6.3.2) states that Archias, a Heraclid from Corinth, founded Syracuse. Graham attempted to solve this problem by stating: “it seems clear that the manufacture of these objects in the West was quickly so organized that even for Greek men, or for a population that certainly contained Greek women, the only 48 fibulae and other metal personal ornaments available were those of Italic type”.187 We cannot be sure that Syracuse ‘certainly contained Greek women’, largely for the same reasons as were raised in the analysis of Pithekoussai. The simple fact that fibulae were found in the graves of children weakens any argument linking them to women alone. Instead of being indicators of ethnic identity (whether indigenous or Greek), it seems more appropriate to regard the fibulae as markers of age and class. A source of metal for the Syracusan fibulae (albeit few that they are) has not yet been discovered, nor indeed have many areas for metal working.188 Drawing firm conclusions about the evidence in the cemeteries is more difficult because of the early date of the excavation of Syracuse. Syracuse was one of the many sites uncovered by Paolo Orsi, the ‘godfather’ of Sicilian excavation, in the late nineteenth century (1891-1895).189 Excavation methods have changed significantly since then and more emphasis has been placed on examining, retaining, and cataloging every single thing that is uncovered. Typical for his period, Orsi regularly discarded bones that were uncovered, so that later analysis using updated or new technologies is now not possible.190 Furthermore, analysis was primarily based on the grave goods found, which required drawing correlations between goods ‘meant’ for women, children, and men. More recent excavations have demonstrated that such assumptions are dangerous. For example fibulae, which women were assumed to have used solely in a practical sense (in a grave context), also appear to have been markers of wealth or status for children, and some men, as well as also serving these functions for women. As a result, our knowledge of Syracuse is more limited than perhaps we might expect. It is difficult to reach an overall conclusion about Syracuse. The literary evidence gives us limited insight into the supposed banishment of the indigenous population, though it is also clear that many were subjugated. Other evidence rests heavily on archaeology, in particular the departure from mother city burial 49 practices. As at Pithekoussai, fibulae have been found though in conjunction with the discovery of the straight pins. Syracuse, located on the eastern side of Sicily, was a colony of Corinth founded c. 734/3 BC.165 Thucydides (6.3.2) states that Archias, a Heraclid from Corinth, founded Syracuse. A plausible explanation is that the fibulae were indicators of class or status, but if that is the case, there is little to support arguments either way about the women who lived in Syracuse and were buried with them. The discovery of an early child burial, potentially of an aristocratic Greek, is strong evidence pointing to the presence of Greek children and therefore Greek women. But this is only one example, however, and so little can be concluded about the practices of the wider population. Accordingly, for Syracuse the evidence is inconclusive as to whether intermarriage took place. 191 De Angelis (2003a) 12. Thucydides (6.4) states that the settlement at Megara Hyblaea existed for 245 years before being destroyed by Syracuse. From Herodotus (7.156), we are able to work out the destruction would have fallen between 485-480 BC. Working backwards from here, we get a foundation date of c. 728. 192 192 De Angelis (2003a) 47-48. See IG vii 52 for Orsippos’ epitaph inscription. 191 De Angelis (2003a) 12. Thucydides (6.4) states that the settlement at Megara Hyblaea existed ( ) 197 As was discussed in Chapter One, the oikistes became the central figure of a founder-cult, and accordingly was worshipped as a hero after his death and accorded burial in a prominent place in the agora (see Wilson (2006) 25). For example, Pindar Pythian (5.92-5) claims that Battus was buried at the edge of the Cyrenaean agora and was worshipped as a hero. For more, see Malkin (1987); Dougherty (1993a). Of course, as was pointed out in the introduction, it must be acknowledged that there are many exceptions to these rules. Megara Hyblaea Megara Hyblaea in Sicily was founded by colonists from Megara in c. 728 BC.191 It is not entirely clear why the colony was founded, in part due to the limited knowledge of the mother city, Megara. De Angelis has suggested that Megara Hyblaea was founded during a period of significant territorial change. The synoecism that seems to have ended the Megarian monarchy in the mid-eighth century BC, and the loss and recovery of land referred to in the epitaph of the Olympic victor, Orsippos of 720 BC, both indicate that the period was one of great change and transformation.192 Thucydides (6.4.1-2) describes the many tribulations that the colonists endured, making various unsuccessful attempts to settle before they reached their final destination: κατὰ δὲ τὸν αὐτὸν χρόνον καὶ Λάµις ἐκ Μεγάρων ἀποικίαν ἄγων ἐς Σικελίαν ἀφίκετο, καὶ ὑπὲρ Παντακύου τε ποταµοῦ Τρώτιλόν τι ὄνοµα χωρίον οἰκίσας, καὶ ὕστερον αὐτόθεν τοῖς Χαλκιδεῦσιν ἐς Λεοντίνους ὀλίγον χρόνον ξυµπολιτεύσας καὶ ὑπὸ αὐτῶν ἐκπεσὼν καὶ Θάψον οἰκίσας αὐτὸς µὲν ἀποθνῄσκει. At this time, Lamis arrived in Sicily leading a colonial party from Megara. He founded a place called Trotilus on the river Pantacyas, and later, for a short time, they lived as fellow citizens with the Chalcidians at Leontini. Having been driven out by them, he founded Thapsus and then died. At this time, Lamis arrived in Sicily leading a colonial party from Megara. He founded a place called Trotilus on the river Pantacyas, and later, for a short time, they lived as fellow citizens with the Chalcidians at Leontini. Having been driven out by them, he founded Thapsus and then died. 50 Throughout the many vicissitudes faced by the colonists, they were led by their oikistes, Lamis. However, Thucydides also reveals that Lamis died at Thapsus shortly after its foundation. An archaeological exploration of this area by Paolo Orsi brought to light an area of cemetery, with one grave in particular situated at a higher level. Graham (1988) 310. 196 Graham (1981-82) 299. 193 Graham (1982) 106; Graham (1988) 309. 194 ( ) 196 Graham (1981-82) 299. 197 194 Graham (1982) 107. 195 Graham (1988) 310 ( ) 195 Graham (1988) 310. ( ) 196 Graham (1981-82) 299. 193 Graham (1982) 106; Graham (1988) 309. 194 G h (1982) 107 ( ) 194 Graham (1982) 107. Graham (1982) 107. 195 Graham (1988) 310. (1988) 310. (1981-82) 299. di d i Ch t O th iki b th t l fi f f d lt d Megara Hyblaea This grave contained two skeletons, two Corinthian Late Geometric cups, and a pair of bronze tweezers.193 It has been suggested that this could be the grave of Lamis himself, due to the dating of the grave to the same period coupled with the evidence in Thucydides about the location of Lamis’ death.194 It has further been suggested that the second body in the grave could be that of Lamis’ (Greek) wife.195 This could be interpreted as evidence that the colonists from Megara brought wives with them rather than intermarried with the local population. Such an interpretation is very problematic. Like Syracuse, Megara Hyblaea also suffers from early excavation in the late nineteenth century when there was little importance placed on the analysis of bones.196 The skeletons in Lamis’ supposed grave have long been discarded, therefore there is no way of carrying out any sex analysis on the bones. We can merely postulate that in cases where there are two bodies in one grave, they are frequently husband and wife, and so this grave in Thapsus may also fit this model. However, even this is problematic. If the grave is indeed that of Lamis, it seems unlikely that his wife would have died shortly afterwards (or before) to allow her to be buried in the same grave as her husband before the colonists moved on to their final destination of Megara Hyblaea. In addition, if we take into consideration the status that an oikistes appears to have held, it does not seem probable that Lamis would have been granted such a simple burial going by the accompanying grave goods.197 These are complicated issues for conjecture, but ultimately, the example of a single, prominent figure is not sufficient evidence to sustain a conclusion that Greek wives accompanied all (or ( ) 195 Graham (1988) 310. 51 some of) the colonists travelling to Megara Hyblaea. Furthermore, it is possible that the assumed Greek wife was sourced en route to Thapsus, for example at the temporary settlement Trotilus, and thus was not Greek at all.198 The series of foundations leading up to the ultimate colony of Megara Hyblaea create such implications for our interpretation of the evidence. The site of the final settlement, Megara Hyblaea, gives an indication of the relationship between the colonists and Sicels. p ( ) 200 Thucydides does not make clear who drove the Megarian colonists from Thapsus, but we can recognise from the verb ἀναστάντες , meaning ‘to force to migrate, transplant’ (LSJ sv. ἀνίστηµι), that they were indeed expelled. De Angelis (2003a) 13 believes that it was Syracuse who drove the Megarian colonists away. DeVoto (2005) 90 assumes that the Megarians submitted to the Sicels at Thapsus. 198 This is a hugely important point that Graham (1982, 1988) overlooks. 199 g y p 199 Shepherd (1999) 290. 200 Megara Hyblaea Often, as is the case for Cumae (Pithekoussai), Syracuse (Ortygia), Cyrene (Platea), to name just a few examples, colonists are more comfortable first settling on an island before moving to the mainland. This appears to have been so that the colonists could assess the indigenous population from an easily defensible place, before they negotiated or forced their way onto the mainland. By contrast, Megara Hyblaea was founded on the mainland from the outset, on a site which is low lying and has no natural defences. Therefore, it seems likely that the site can only have been settled with the permission of the neighbouring indigenous population.199 This hypothesis is further confirmed with evidence from Thucydides (6.4). He states: οἱ δ᾽ ἄλλοι ἐκ τῆς Θάψου ἀναστάντες Ὕβλωνος βασιλέως Σικελοῦ προδόντος τὴν χώραν καὶ καθηγησαµένου Μεγαρέας ᾤκισαν τοὺς Ὑβλαίους κληθέντας. καὶ ἔτη οἰκήσαντες πέντε καὶ τεσσαράκοντα καὶ διακόσια ὑπὸ Γέλωνος τυράννου Συρακοσίων ἀνέστησαν ἐκ τῆς πόλεως καὶ χώρας. πρὶν δὲ ἀναστῆναι, ἔτεσιν ὕστερον ἑκατὸν ἢ αὐτοὺς οἰκίσαι, Πάµιλλον πέµψαντες Σελινοῦντα κτίζουσι, καὶ ἐκ Μεγάρων τῆς µητροπόλεως οὔσης αὐτοῖς ἐπελθὼν ξυγκατῴκισεν. But the others were forced to leave Thapsus200 and they founded the place called Hyblaea. Hyblon, the king of the Sicels gave the Megarians the land and he led the way. They lived there for 245 years, before they were compelled to migrate from the city and land by Gelon, the tyrant of Syracuse. Before this happened, 100 years after they founded there [Hyblaea], they sent Pamillus, who had come from the mother city Megara to join them in colonising, and founded Selinus. 52 No evidence in archaeological examinations has revealed any Sicel settlement on the site prior to the arrival of the Megarians.201 Therefore, the colonists were not compelled to drive out the indigenous population, which would have put a strain on the relationship between local and immigrant population. It is feasible, then, that good relations existed between the Greek colonists and the indigenous Sicels, and that these were conducive to opportunities for intermarriage. Relations between the two peoples were, perhaps, so good that the settlement, Megara Hyblaea, gained its name from the Sicel king Hyblon. LSJ sv. προδίδωµι. 205 Graham (1988) 311-313. Graham cites other uses of the verb (Xenophon, Hellenica 1.5.7 and Polybius 31.27.5) where he argues they should too be translated as ‘to betray’. Malkin (2002b) 220-221 agrees with Graham’s translation as ‘to betray’. Neither scholar, however, discusses what sort of betrayal Thucydides implies. 201 Shepherd (1999) 290. p ( ) 202 Malkin (2002b) 220. 203 LSJ sv. καθηγέοµαι. 204 204 LSJ sv. προδίδωµι. Megara Hyblaea Supporting this idea, Malkin has put forward an interesting theory that because of the death of the oikistes Lamis, the Sicel king Hyblon helped fill this role.202 The language used by Thucydides supports this interpretation, with the verb καθηγέοµαι meaning to ‘act as a guide’ or to ‘lead the way’.203 At the very least, the Sicel king had good relations with the Greek colonists, and Hyblon’s relations with others can probably be extended to encompass the peoples over whom he ruled (that is, the neighbouring Sicels). Of course, good relations were not imperative to intermarriage taking place, but if they were good, there is a stronger chance that it could have occurred. The interpretation of the verb προδόντος, used by Thucydides, is crucial to understanding the relationship between the Megarian colonists and Hyblon. Like many Greek words, the verb προδίδωµι can be translated in a number of different ways, and in this instance, the interpretation can significantly change the implications of the action.204 Graham argues that προδίδωµι should be translated in its most commonly used capacity, meaning ‘to betray, forsake’. 205 He argues that Thucydides, probably personally, considered the transference of land from indigenous to Greek hands a betrayal, due to the recent hostility in neighbouring colonies against the Sicels (Syracuse and Leontini) in a strictly Greek versus Sicel approach. Graham also suggests that Hyblon could have belonged to a larger Sicel alliance against the Greeks which he betrayed, or more specifically that Hyblon 53 betrayed his own people. These suggestions are problematic. It does not seem possible to interpret the betrayal as that of Hyblon betraying his own people, chiefly because the land, τὴν χώραν, is quite clearly the accusative object of the verb, ‘betray’. Therefore, it seems unlikely that the verb, προδίδωµι, should be translated as ‘to betray’. Alternative translations of προδίδωµι include ‘to pay in advance’ or simply ‘to give to’. Paying in advance suggests that Hyblon gave the colonists the land as a part of a negotiated deal whereby the colonists would gain the land and would agree not to subjugate the Sicels. 206 Graham (1988) 312 states that “in the Thucydidean passage too emendation, to παραδόντος, was tentatively proposed by Classen in his first edition (Berlin, 1875). It is unfortunate that later editions promoted the emendation into the text, and that they were followed by the Loeb edition.” Clearly, therefore, Classen thought that ‘to give, hand out’ was a valid reading (obviously Graham disagrees with this). g ) 207 Admittedly, we do not know the exact location of Hyblon’s throne, but Megara Hyblaea is only 20 kilometres from Syracuse, so it would be highly unlikely that Hyblon was not aware of the consequences of the presence of other Greek colonists in Sicily. 208 q p 208 Shepherd (1995) 56. 208 Shepherd (1995) 56. Megara Hyblaea On the other hand, giving the land implies that Hyblon simply handed the colonists the land.206 Taking into account the fact that neighbouring Sicels had been driven off their lands at Syracuse and Leontini, it seems curious that Hyblon would give away land to a people he knew had caused problems for his neighbours.207 For that reason, it seems more likely that an offer of land was a precaution on the part of Hyblon. Of course there are many limitations to such a conjecture. We cannot be sure of the numbers of Greek colonists (as discussed in Chapter One) and even less so, the numbers of Sicels connected with Hyblon. Consequently, we cannot argue one way or another as to whether Hyblon was able to drive the colonists away and chose not to. Furthermore, we cannot rule out the option that friendly relations between the two peoples existed for their own sake. An examination of burial sites of Megara Hyblaea has, like Syracuse, demonstrated that not all colonies followed their mother city’s burial practices. Shepherd conducted a study in 1995, and although she freely admitted that there was little accessible information about the relevant date of the burials of Megara, it appears that they did not differ greatly from those of Corinth – and so these were used as an alternative comparison.208 In this period, the burial practices of both Corinth and Megara typically used slab-lined cist graves and monolithic 54 sarcophagi. By contrast, at Megara Hyblaea inhumation or cremation in amphorae was the usual burial practice.209 This could therefore indicate that the colonists were influenced by their indigenous neighbours, and adopted local practices for their own. It could also imply that some Sicels joined the Megarian colonists in settling at Megara Hyblaea, bringing with them indigenous burial practices. Again, there is no apparent reason which could have prevented a continuation of mother city burial practices. Therefore, this change was likely a conscious decision. S ep e d ( 995) 57. 211 Shepherd (1995) 59. Shepherd reaches this conclusion by directly comparing the burial practices of Megara Hyblaea with Syracuse. This theory seems convincing, given that a certain amount of competition and hostility existed between the two colonies, before Syracuse destroyed Megara Hyblaea completely in 485-480 BC. 212 ( ) 214 Shepherd (1999) 290-291. At the date of publication, Shepherd states that only 20 fibulae had been uncovered. p ( ) 213 Buchner (1979) 135. p ( ) 210 Shepherd (1995) 57. 209 Shepherd (1995) 56. 212 Shepherd (2005) 118. 212 Shepherd (2005) 118. 213 Buchner (1979) 135 209 Shepherd (1995) 56. 210 Shepherd (1995) 57. Megara Hyblaea Interestingly, by about the second half of the seventh century BC, there seems to have been a revival of mother city practices, especially with monolithic sarcophagi, and vessel burials became less common.210 Any influence from the mother city in this change seems unlikely (unless there was a massive influx of new Megarian colonists), so perhaps the change came as Megara Hyblaea had become more settled and was beginning to prosper, and burial was just one way in which citizens could demonstrate their prosperity.211 Certainly in other Sicilian colonies the use of monolithic sarcophagi appears to have become more and more an elite burial type and thus a good indicator of wealth and status.212 Comparing evidence from Megara Hyblaea to another colony discussed earlier, Pithekoussai, brings an interesting difficulty in the interpretation of this evidence. Buchner’s explanation for the fibulae found in the cemeteries draws a correlation between the origin of the fibulae and the origin of those who possessed them; that is, fibulae of Italic types must have been used by indigenous women.213 The various problems with this conclusion have already been outlined. If this hypothesis is applied to Megara Hyblaea, it becomes even more problematic. Though not all of the burials at Megara Hyblaea have been fully published, very few fibulae have been found.214 By contrast, straight pins occur much more 55 frequently, with a total of 156 discovered and recorded in publication.215 In addition, the pins were usually found in pairs, and were located on the shoulders of the deceased, suggesting they served a practical purpose rather than acting solely as grave offerings. If Buchner’s hypothesis from Pithekoussai is applied here, it would appear that the women of Megara Hyblaea were almost entirely of Greek origin rather than Sicel. This, however, is difficult to reconcile with the literary evidence (particularly of Thucydides) which implies that the colonists had unusually good relations with the indigenous population (though of course, this does not necessarily mean that intermarriage occurred). Therefore it seems that the presence of fibulae is no more an argument for intermarriage than the presence of straight pins is against it. Furthermore, as not all the straight pins belong to the same date (as indeed not all the fibulae do), it is entirely possible that by even the sixth century BC, ethnic distinctions may have been less apparent. 215 Shepherd (1999) 291. As above, the number 156 was correct at the date of Shepherd’s publication. publication. 216 De Angelis (2003a) 54. 215 Shepherd (1999) 291. As above, the number 156 was correct at the date of Shepherd’s publication. 216 De Angelis (2003a) 54. Megara Hyblaea Such a comparison further highlights the problematic nature of basing a theory of intermarriage on any single type of object found. It also demonstrates that when a hypothesis seems plausible for one colony, it is not necessarily possible to extend it to others. The final crucial point to note for Megara Hyblaea is that around the mid-seventh century BC, a little under a century after foundation, the indigenous sites within the surrounding area were abandoned. This is particularly noticeable at Pantalica and Villasmundo. De Angelis suggests that this phenomenon indicates demographic transfer; either the indigenous populations moved away altogether, or they moved into the urban centre of Megara Hyblaea, and were subsumed into its population.216 The indigenous influences within the site, coupled with the evidence discussed from Thucydides, support the likelihood that the indigenous populations surrounding Megara Hyblaea became included among the colonists. There is some evidence to support an argument that the colonists who settled in Megara Hyblaea were accompanied by Greek wives. The possibility that Lamis and his wife were buried in the same grave on their journey from Megara to the new colony, as well as the discovery of a higher proportion of straight pins in the 56 graves than Italic fibulae, lend support for this interpretation. These are not strong arguments, however, and the weight of the evidence, notably good relations between the colonists and the indigenous population (or at least between the colonists and Hyblon) set out in the literary tradition, and the apparent integration of two groups living within the settlement within 100 years of establishment, support the view that intermarriage was highly likely. 218 Tsakirgis (1995) 123. 219 4 217 Eikeland (2006) 121. 218 g ( ) 219 OCD4 (2012) 968. ( ) 220 Lyons (1996b) 177. y ( ) 221 Eikeland (2006) 126. Morgantina Morgantina lies almost at the geographical centre of Sicily, about 48 kilometres north of the south coast and about 54 kilometres inland from the east coast.217 The settlement is located on a steep ridge, approximately three kilometres in length, known today as Serra Orlando, with the main settlement located on the so-called Cittadella to the east of Serra Orlando.218 Both Strabo (6.1.6) and Dionysius of Halicarnassus (Antiquitates Romanae 1.12.3) relate the foundation myth of Morgantina, according to which the founder of the settlement was a man named Morges (after whom the settlement was more than likely named). Pliny (Natural History 3.10) describes how the settlers were from Bruttium, though according to an analysis of pottery and masonry styles, the settlers were from Catana or Leontini, settling c. 560 BC.219 As such, Morgantina does not have a mother city per se, but instead should be considered as a secondary colony. The settlement has a long history of occupation, from the Iron Age, right down to the first century AD. 220 This, coupled with the fact that the town lacks an orthogonal plan, suggests that colonisation of Morgantina was a more gradual process and may not have taken place through a formal act of foundation.221 The first controlled archaeological excavations were conducted by Paolo Orsi in 1912, and have been continued at a number of different stages by various people and groups since. The majority of tombs were uncovered in 1969-1970 as part of the Princeton University excavations. The Princeton University team established that the most common tomb type in the archaic period was the chamber tomb, 217 Eikeland (2006) 121. 218 57 57 accounting for 67% of the total burials.222 The fossa grave was the next most common; a rectangular trench surrounded by a shallow ledge, covered by terracotta roof tiles or stone slabs.223 Children were typically inhumed in fossa graves; adults could also be buried in fossa graves, though these tended to be more modest burials with very few grave goods accompanying the burial. Therefore, the majority of burials at Morgantina fit within an indigenous burial tradition. The Princeton University study acknowledged that many of the burials at Morgantina had been disturbed particularly by grave robbers during the initial excavation and cleaning of the chamber tombs making them difficult to assess. 222 Lyons (1996a) 15. 223 Lyons (1996a) 21. 224 Sjoqvist (1958) 158. 225 Shepherd (2005) 117. 226 Sjoqvist (1973) 35. 227 Lyons (1996a) 179. 228 Eikeland (2006) 94; Sjoqvist (1958) 156. y ( ) 224 Sjoqvist (1958) 158. 225 Shepherd (2005) 117. j q ( ) 227 Lyons (1996a) 179. 223 Lyons (1996a) 21. 224 p ( ) 226 Sjoqvist (1973) 35. y ( ) 228 Eikeland (2006) 94; Sjoqvist (1958) 156. 222 Lyons (1996a) 15. Morgantina Furthermore, Sjoqvist recorded in his Preliminary Report of the Second Season of Excavation that many of the tombs had flooded, with water causing objects to float around, thus upsetting them from their original placement.224 Despite this, archaeological examinations of the graves have revealed a curious mixture of cultural features. For example, one burial reveals a Greek-style sarcophagus found inside a traditional Sicel chamber tomb.225 In another example, within Tomb 4, one burial presents a Greek burial rite – a cremation with an ash urn sunk into a pit in the rock – but the ashes from the cremation were collected in a Sicel vessel.226 Such a mixed typology of the tombs supports an interpretation of Greeks and Sicels living side by side, with both peoples using the necropolis. This view is reinforced by the fact that there was no segregation between the various tomb types or burial rites used; that is, different areas of the cemetery were not restricted to one burial type, but instead all were mixed in together.227 The grave goods also give this impression; again, a diverse mixture of finds has been revealed. There are four main kinds of archaic pottery: Attic imports, late Corinthian imports, indigenous Siculo ware, and finally a Siculo ware which imitates Attic forms. 228 This mixture of indigenous Siculo and Greek Geometric pottery could be a further indication of the Sicels and Greeks living side by side. On the other hand, it is easy to imagine that Sicels themselves could have used 58 their own traditional pottery, together with Greek pottery, which they also found occasion to imitate. However, a closer analysis of some of the pottery gives a better indication that there probably was a Greek presence at Morgantina explaining the presence of Greek wares. Various items associated with drinking parties (symposia) were found in multiple graves, including those thought to belong to women.229 Eikeland accordingly suggests that women were also allowed to participate in symposia at Morgantina, while they were typically restricted to men (and hetairai) back in mainland Greece.230 The sheer number of drinking vessels found seems greater than an assortment of objects collected through trade and more likely to reflect an adaptation of Greek social practices.231 Other objects found include fibulae, jewellery, and weapons. p ( ) 232 Lyons (1996a) 109 admits that weaponry is found in indigenous necropoleis in abundance in the Iron Age, though it begins to decrease slightly in frequency in the seventh and sixth centuries BC. This could be because of an adoption and adaptation of Greek practices, which meant that burials with weaponry became less popular or common. 233 229 Most bones uncovered in the excavations were in an extremely poor state of preservation, and due to the periodic opening and rearrangement of the tombs many of the bones were mixed together, with others not saved at all for later studies. Taking this into consideration, coupled with the fact that many of the grave goods had been shifted from their original location, it is difficult to make any definitive determinations of sex. y 230 Eikeland (2006) 94. Eikeland does not take into consideration that these objects associated with females could have also belonged to hetairai at Morgantina. 231 231 Shepherd (2005) 117. 233 Lyons (1996a) 109. Morgantina Because many of the burials had been disturbed, it is difficult to assess whether the fibulae served a practical purpose, or were merely grave goods. Making a reliable assessment of jewellery finds is equally difficult. Weapons are not commonly found in the cemeteries of the Greek colonies in the archaic period. Even in colonies already discussed which were founded amidst violence, such as Syracuse, there is a distinct absence of any weaponry in the cemeteries. This contrasts with indigenous Italic settlements where weapons are frequently found in graves.232 Morgantina is different again, with some weaponry being revealed. Most spectacularly, the Princeton University excavation identified a chamber tomb with two warriors buried in complete armour.233 Each warrior was equipped with a Corinthian-style helmet, greaves, and a sword. A large shield was also discovered within the tomb. The high quality of this weaponry is clearly an anomaly, though other bronze and iron weapons of varying quality have been 59 found in other graves at Morgantina.234 Lyons has suggested that this weaponry, coupled with some significant deposits of jewellery, could point towards the existence of some sort of an elite class within Morgantina.235 This suggestion is further reinforced with the multiple findings associated with the symposium discussed above. Clearly, the burials and associated finds are very different from both indigenous and Greek practice. This makes it difficult to conclude one way or another whether the women found in some of the graves were likewise Greek, or indigenous. The Princeton University study concludes that the diversity of evidence that is so apparent at Morgantina suggests an increasingly stratified community with access to a greater range of imported goods (hence the mixed nature of finds).236 If we agree with this hypothesis, does it mean that we turn away from a theory of intermarriage? Antonaccio argues in this vein, stating that neither an influential Greek population, nor mixed marriages would necessarily be required to explain the presence of Greek features at Morgantina. 237 However, the sheer number of Greek objects lends itself more to the impression of an extended physical presence, than mere trade encounters. On its own though, this does not help clarify whether there were Greek women present in the settlement. 234 It is likely that this weaponry was not used strictly for military purposes, but instead was used for hunting. 235 L (1996 ) 109 ( ) 240 Antonaccio & Neils (1995) 265. Antonaccio (1997) 180. 238 Antonaccio (1993) 352; Antonaccio & Neils (1995) 261. 239 ( ) 240 Antonaccio & Neils (1995) 265. ( ) ; 239 Antonaccio & Neils (1995) 261. Antonaccio (1993) 352; Antonaccio & Neils (1995) 261. 239 Antonaccio & Neils (1995) 261. y ( ) 237 Antonaccio (1997) 180. ( 241 Antonaccio (1997) 191. 35 Lyons (1996a) 109. 36 Lyons (1996a) 133. 37 Antonaccio (1997) 180 235 Lyons (1996a) 109. 236 Lyons (1996a) 133. 39 Antonaccio & Neils (1995) 261. 40 Antonaccio & Neils (1995) 265. 237 Antonaccio (1997) 180. 238 Antonaccio (1993) 352; Antonaccio & Neils (1995) 261. 236 Lyons (1996a) 133. 237 238 Antonaccio (1993) 352; Antonaccio & Neils (1995) 261. 239 Antonaccio & Neils (1995) 261. 235 Lyons (1996a) 109. 236 Lyons (1996a) 133. 234 It is likely that this weaponry was not used strictly for military purposes, but instead was used for hunting. 235 Lyons (1996a) 109. 236 Lyons (1996a) 133 Antonaccio (1997) 180. Antonaccio (1993) 352; Antonaccio & Neils (1995) 261. Antonaccio & Neils (1995) 261. y ( ) 237 Antonaccio (1997) 180. 239 Antonaccio & Neils (1995) 261. Antonaccio (1997) 180. Antonaccio (1993) 352; Antonaccio & Neils (1995) 261. A i & N il (1995) 261 Morgantina One important find from Morgantina was a piece of graffito, combining both Greek and Sicel elements – a Greek script and language used to record a Sicel name (Figure 6).238 In 1990, a single large sherd was discovered in a grave, making up part of the neck and rim of a black-slipped Laconian ware krater.239 The fragment is 18 centimetres long and 13 centimetres high which suggests that originally the vessel probably had a 38 centimetre diameter and stood about 30 centimetres in height making it a standard size.240 Both the vase shape and the forms of the letters incised suggest a date in the second half of the sixth century BC.241 A rare example of writing is recorded on the sherd, stating: κυπαρας εµι, 234 It is likely that this weaponry was not used strictly for military purposes, but instead was used for hunting. 235 Lyons (1996a) 109. 60 meaning “I belong to Kupara”. The practice of naming objects was not uncommon; in other colonies there are the occasional instances of it, such as the well-known eighth century BC ‘Nestor’s Cup’ of Pithekoussai.242 Antonaccio and Neils’ study demonstrates that the Morgantina graffito employs a standard formula, with a name recorded in the genitive (indicating the owner or dedicand), followed by a verb, εἰµί (or in this case, ἐµί), used in this situation to mean “I belong to so-and-so”.243 Their study acknowledged that the -ρας ending was indeed the genitive of a α-declension, and the owner was therefore likely to be a woman.244 The name found on the fragment, Kupara, is not Greek. It does, however, appear to be a feminine name and is attested elsewhere in Sicily. For example, the name is paralleled on an inscription on a lead tablet in the north-west of Sicily, on a record discharging a debt of one archon and his children, Saiso and Kupura, to a goddess.245 Instances of the name Kupara are also found on what are probably votive offerings in a sanctuary at Terravecchia di Cuti.246 Kupara has also been attested in Syracuse as an alternative Sicel name for the spring Arethousa. 242 ‘Nestor’s Cup’ is significant for being one of the earliest inscriptions in archaic Greek alphabet, as well as being one of the few pieces of original poetry from the eighth century BC. It displays a verse inscription in three retrograde lines, stating: “Nestor’s cup was fine to drink from. But whoever drinks from this cup will immediately be seized by the desire of fair-crowned Aphrodite.” (translation: Graham (1982) 99). See Hansen (1976) for a general discussion on ‘Nestor’s Cup’, and Watkins (1976) for a focused analysis of the inscription; also Malkin (1998) 156-158. 243 Antonaccio & Neils (1995) 268 272 , g y, p 245 Antonaccio (1997) 181; Antonaccio & Neils (1995) 269. ( ) y p ; ( 243 Antonaccio & Neils (1995) 268, 272. ( ) 247 Antonaccio (1997) 181. 243 Antonaccio & Neils (1995) 268, 272. 244 ( ) 246 Antonaccio (1997) 181. Antonaccio & Neils (1995) 268, 272. 244 H li i ti ll l i ibl i th d l i ( ) , owever, linguistically, a male owner is possible in the Morgantina At Syracuse she was personified as a nymph on coinage, and that same image has also been found on coinage in Morgantina during its period of Syracusan control toward the end of the fourth century BC.247 Therefore, it is possible that the Kupara on the Morgantina graffito was in fact dedicated to a nymph rather than belonging to an actual woman named Kupara. Although we cannot be certain either way, it seems less likely that the Kupara on the fragment refers to the nymph, given that the evidence (coinage) used to back up such a suggestion is from two centuries after the krater with the inscription was used. The dating of this sherd in the second half of the sixth century BC seems to coincide with the beginning of a more significant Greek presence at Morgantina. 61 While on the one hand, it would be easy to assume that this sherd was a clear indication of intermarriage (given the sherd belonged to an indigenous woman, Kupara), on the other hand, we must be mindful that there was already a very established indigenous presence at Morgantina, in addition to a minor Greek one, when the Greeks seem to have first arrived. Consequently, this sherd does not make it possible to definitively rule out the possibility that Greek wives accompanied the colonists to Morgantina, nor does it confirm that intermarriage took place. It does, however, further confirm that Greeks and Sicels coexisted at the settlement, and accordingly, it seems likely that intermarriage would have occurred (even if not straight away). At Morgantina it is very difficult to firmly conclude whether the early colonial women were Greek or indigenous. To do this calls for distinguishing between a Greek settlement with a likely case of intermarriage and a settlement with a very high degree of Hellenisation. Certainly, the indigenous population had considerable contact with the Greeks and they appear to have lived side by side in a mixed community. The sheer number of Greek objects reinforces the impression of deeper contacts than simply trading encounters, and the evidence also points to Greek customs and practices being modified and taking on some indigenous characteristics. All of this suggests the possibility of intermarriage at Morgantina between Greek men and indigenous women. 248 Spartans tended to expand outwards into Messenia and the wider Peloponnese, rather than overseas. overseas. 249 Astour (1985) 24. 249 Astour (1985) 24. overseas. 250 Pembroke (1970) 1241; Brauer (1986) 3. Strabo (6.3.2-3) reports two different accounts. The first is of a fifth century BC Syracusan historian Antiochus; the second is of fourth century BC Ephorus of Cyme in Aeolia, Asia Minor. The two versions are relatively substantial (and so are not being covered here in any depth). They also differ in many aspects, especially regarding who the partheniai were. This is not important to my argument, but it is significant to observe that both versions make pertinent that the colonists (the partheniai) were not in any complete sense ‘Spartans’. See Brauer (1986) or Pembroke (1970) for further explanation. 251 p ( 251 Malkin (1994a) 141. ( ) 252 Redfield (2003) 293. Spartans . See Brauer (1986) or Pembroke (1970) for further explanation. 251 Malkin (1994a) 141. p ( ) ( ) p 251 Malkin (1994a) 141. 252 253 Malkin (1994a) 122. Malkin’s discussion deems the oracle to be authentic; however he also recognises that, regardless, it can be seen as ‘thematically authentic’ in its reflection of early attitudes toward territory and the indigenous peoples. Graham (1981-82) 298-299 disagrees, and argues that “oracles of this kind are a notoriously concocted element in our traditions about colonial foundations, and this one must clearly be rejected as unhistorical.” Shepherd (1999) 270 also argues that the narrative is so fabricated that the presence of a Greek wife is unlikely. Malkin (1994a) 126 further suggests that these attitudes contained in the oracle could simply be because Taras was a Spartan colony, and given their recent conflict with the Messenians, the same hostile attitudes were projected onto the new environment where territorial appropriation was necessary. 254 M lki (1994 ) 121 Taras Taras, located in south-east Italy, was founded in c. 706 BC by colonists from Sparta. It appears to have been one of very few colonies founded by Sparta in the archaic period.248 Taras is located in the modern day Gulf of Tarentum and was situated between an inner and outer natural harbour.249 The site has been constantly occupied since indigenous habitation just prior to the establishment of the Greek colony, later becoming Roman Tarentum, then modern Taranto. 62 Consequently, extensive excavation work is almost impossible, so our knowledge of the site is heavily reliant on the literary tradition. Tradition has it that the colonists concerned with Taras’ foundation were known as the partheniai. Historians have concluded that these were probably the sons of Spartan women and helot men, who were illegitmate and born during the first Messenian war when the husbands of the women were away fighting.250 The motivating factors behind leaving Sparta and founding a new colony are not entirely clear. Malkin has postulated that because the foundation took place following the first Messenian war and seems to have involved non-integrated groups of Spartans (the partheniai), it is likely that some sort of stasis involving land and/or status was behind the foundation.251 The relationship between the partheniai as colonists and the Spartan state itself is therefore ambiguous. While on the one hand, the partheniai were apparently banished from the mother city almost as if they were a threat, on the other hand they were representing Sparta as a colony and, as Redfield puts it, were “to recapitulate Spartan experience”.252 It is tempting to conclude that women accompanied the men to Italy given their compromised status. Clearly the label of the group as ‘partheniai’ suggests a separate status, and a status predicated on the female line. Another curious aspect about the foundation story of Taras is the oracle given to the oikistes, Phalanthus, by the Delphic oracle, recorded by Strabo (6.3.2): ὁ δ᾽ ἔχρησε: ‘Σατύριόν τοι δῶκα Τάραντά τε πίονα δῆµον οἰκῆσαι, καὶ πῆµα Ἰαπύγεσσι γενέσθαι’. The oracle proclaimed: ‘I gave both Satyrion and the rich land Taras to you to inhabit, and to be a calamity to the Iapygians’. Not only does the oracle reveal the exact location where the colonists were to settle, but it appears to command Phalanthus to make war on the indigenous population, the Iapygians. ( ) 255 Whitehouse & Wilkins (1989) 105. p j 254 Malkin (1994a) 121. 2 Taras Whether the oracle recorded in Strabo is authentic or 63 not is irrelevant to this discussion; whether it was the answer to an enquiry at Delphi or was a fiction, it characterises the attitudes the colonists took to Taras.253 The oracle also states that the Spartan colonists first settled at Satyrion (modern Leporano) before they relocated to Taras.254 This has been confirmed archaeologically and excavations have revealed Greek material from c. 750 BC.255 As we go through each case study of each colony, it is becoming increasingly clear that it was common practice for colonists to settle on an island or another easily defensible place, so that they could assess the territory and establish relations (whether hostile or friendly) with the indigenous population. Another interesting tale associated with the foundation of Taras is reported to us by Pausanias (10.10.6-8): Τάραντα δὲ ἀπῴκισαν µὲν Λακεδαιµόνιοι, οἰκιστὴς δὲ ἐγένετο Σπαρτιάτης Φάλανθος. στελλοµένῳ δὲ ἐς ἀποικίαν τῷ Φαλάνθῳ λόγιον ἦλθεν ἐκ Δελφῶν: ὑετοῦ αὐτὸν αἰσθόµενον ὑπὸ αἴθρᾳ, τηνικαῦτα καὶ χώραν κτήσεσθαι καὶ πόλιν. τὸ µὲν δὴ παραυτίκα οὔτε ἰδίᾳ τὸ µάντευµα ἐπισκεψάµενος οὔτε πρὸς τῶν ἐξηγητῶν τινα ἀνακοινώσας κατέσχε ταῖς ναυσὶν ἐς Ἰταλίαν: ὡς δέ οἱ νικῶντι τοὺς βαρβάρους οὐκ ἐγίνετο οὔτε τινὰ ἑλεῖν τῶν πόλεων οὔτε ἐπικρατῆσαι χώρας, ἐς ἀνάµνησιν ἀφικνεῖτο τοῦ χρησµοῦ, καὶ ἀδύνατα ἐνόµιζέν οἱ τὸν θεὸν χρῆσαι: µὴ γὰρ ἄν ποτε ἐν καθαρῷ καὶ αἰθρίῳ τῷ ἀέρι ὑσθῆναι. καὶ αὐτὸν ἡ γυνὴ ἀθύµως ἔχοντα —ἠκολουθήκει γὰρ οἴκοθεν—τά τε ἄλλα ἐφιλοφρονεῖτο καὶ ἐς τὰ γόνατα ἐσθεµένη τὰ αὑτῆς τοῦ ἀνδρὸς τὴν κεφαλὴν ἐξέλεγε τοὺς φθεῖρας: καί πως ὑπὸ εὐνοίας δακρῦσαι παρίσταται τῇ γυναικὶ ὁρώσῃ τοῦ ἀνδρὸς ἐς οὐδὲν προχωροῦντα τὰ πράγµατα. προέχει δὲ ἀφειδέστερον τῶν δακρύων καὶ ἔβρεχε γὰρ τοῦ Φαλάνθου τὴν κεφαλήν συνίησί τε τῆς µαντείας ὄνοµα γὰρ δὴ ἦν Αἴθρα τῇ γυναικί καὶ οὕτω τῇ ἐπιούσῃ νυκτὶ Τάραντα τῶν βαρβάρων εἷλε µεγίστην καὶ εὐδαιµονεστάτην τῶν ἐπὶ θαλάσσῃ πόλεων. 64 Taras is a colony of the Spartans, and its oikistes was Phalanthus, a spartiates. When setting out for the colony, an oracle came to Phalanthus from Delphi, [declaring] that when he felt rain under a cloudless sky, he would acquire land and a city. At first he neither reviewed the oracle himself, nor consulted one of his advisors, but he continued to Italy with his ships. But although he conquered the barbarians, he neither seized any city nor ruled over any land. p ( ) 257 Brauer (1986) 11. 256 Shepherd (2012) 220. 257 Taras Further, the literary evidence, particularly Pausanias (10.5-8), seems to have a solid foundation in myth and accordingly cannot be relied upon for its historicity. The dearth of any archaeological evidence also leaves us with little to justify and verify the literary evidence in the way it is possible to do (at least to a certain extent) in other Greek colonies. This makes it difficult to assess the practices of the colonists once they had settled into their new colony, and especially, their relations with the indigenous inhabitants of the area after ‘being a plague’ to them. Further, the literary evidence, particularly Pausanias (10.5-8), seems to have a solid foundation in myth and accordingly cannot be relied upon for its historicity. The dearth of any archaeological evidence also leaves us with little to justify and verify the literary evidence in the way it is possible to do (at least to a certain extent) in other Greek colonies. If we solely consider the literary evidence discussed above, it appears that intermarriage was unlikely at Taras. The Spartan colonists seem to have driven out all the indigenous inhabitants, the Iapygians, on their arrival (though there was potentially opportunity for them to seize wives during this process). Furthermore, while there is evidence, albeit heavily grounded in myth, that at least one wife accompanied her husband, it does not speak to the practices of the other colonists. Her presence in the story does perhaps indicate that the presence of a wife or woman was not so unusual that she appears out of place. Accordingly, at Taras, it is very difficult to reach any firm conclusion about whether or not intermarriage occurred. Taras He called to mind the oracle and considered that the god proclaimed an impossibility. For it could never rain from a clear and cloudless sky. His wife, while he was dispirited – for she had accompanied him from home – treated him affectionately and placed her husband’s head between her knees and picked out the lice. And out of affection, she cried seeing her husband’s affairs coming to nothing. As her tears poured forth, she wetted Phalanthus’ head, and then he understood the oracle, for his wife’s name was Aethra. And thus on that night, he seized Taras from the barbarians, the largest and most wealthy city by the sea. This passage about the foundation of Taras is well-known for the riddle it contains about ‘rain under a cloudless sky’. More significantly, it also reveals, according to the tale, that the oikistes’ wife, Aethra, accompanied him on his journey, and indeed that she played a crucial role in the foundation of Taras. This is comparable to the evidence discussed regarding Megara Hyblaea’s founder, Lamis, and the grave that potentially contains his body and that of his own wife. On the other hand, however, it is not credible to extrapolate the evidence on the oikistes to every single colonist who took part in the foundation, just as the practices of monarchs or other minority groups cannot be used by themselves as evidence for the practices of the majority population. In other words, Aethra, because of the crucial role she plays in the foundational narrative, appears to have been an exceptional member of the colonial group.256 Even so, it is reassuring to note that there was a role for individual women. Furthermore, it is worth noting that the presence of women was not so unusual that Aethra seems out of place in the story. An assessment of the likelihood of intermarriage at Taras is made extremely difficult by the fact that any conclusion is based almost solely on literary evidence. While the literary evidence for the foundation of the colony is very rich, our evidence of the early years of the colony after foundation is almost non-existent.257 65 This makes it difficult to assess the practices of the colonists once they had settled into their new colony, and especially, their relations with the indigenous inhabitants of the area after ‘being a plague’ to them. ( ) 259 Pembroke (1970) 1250. 258 Greco (2006) 173. ( ) 260 Redfield (2003) 253. Locri Epizephyrii Taras is frequently analysed alongside another southern Italian colony, Locri, or Locri Epizephyrii. This is primarily because the foundation myths of both colonies include women, which in turn aids consideration of the social function that women served in wider archaic society, and particularly colonial society.258 Locri was founded in c. 673 BC by the Locrians of eastern Greece. The colony is located in south-eastern Italy, in the ‘heel’ close to Taras. The city is unique among Greek colonies in that it never appears to have had a proper name.259 Instead, the city, Locri, was named after the people who lived in it (οἱ Λοκροί).260 This contrasts with the usual practice where Greek peoples are named after their ( ) 260 Redfield (2003) 253. 66 city (for example: the Corinthians of Corinth; the Athenians of Athens; the Thebans of Thebes). Locri is especially significant as the literary tradition states that there were female settlers involved in the settlement process. Polybius (12.5.6-8) writes: πρῶτον µὲν ὅτι πάντα τὰ διὰ προγόνων ἔνδοξα παρ᾽ αὐτοῖς ἀπὸ τῶν γυναικῶν, οὐκ ἀπὸ τῶν ἀνδρῶν ἐστιν, οἷον εὐθέως εὐγενεῖς παρὰ σφίσι νοµίζεσθαι τοὺς ἀπὸ τῶν ἑκατὸν οἰκιῶν λεγοµένους: ταύτας δ᾽ εἶναι τὰς ἑκατὸν οἰκίας τὰς προκριθείσας ὑπὸ τῶν Λοκρῶν πρὶν ἢ τὴν ἀποικίαν ἐξελθεῖν...τούτων δή τινας τῶν γυναικῶν συνεξᾶραι µετὰ τῆς ἀποικίας, ὧν τοὺς ἀπογόνους ἔτι νῦν εὐγενεῖς νοµίζεσθαι καὶ καλεῖσθαι τοὺς ἀπὸ τῶν ἑκατὸν οἰκιῶν. Firstly they said that all ancestral reputation comes through the female line not the male. These belonged to ‘the hundred families’, selected by the Locrians before sending out the colony...some of these women assisted in raising the colony, their descendants are still now thought to be noble and called the ‘men of the hundred families’. This passage describes the crucial role that women supposedly played in the foundation of Locri. These women were the bearers of the noble bloodlines of the so-called Hundred Houses, the leading families of Locris in Greece. Because the Locrians are reported to follow their descent through women, it has been argued that this narrative is a fabrication – possibly an aetiology dating to the fifth century BC, probably created to explain the matrilineal succession of Locri.261 This tradition, therefore, is heavily rooted in myth and its historicity should not be taken as a given. However, the passage does show that there could have been potential for women to fulfill an important role. 261 Shepherd (2012) 219; Pembroke (1970) 1252; Cawkwell (1992) 295; Hall (2004) 40. Locri Epizephyrii The majority of the graves published by Orsi were those which were perceived to have interesting contents (grave goods). These tended to be items such as toys, and objects associated with women, such as needle cases, mirrors, perfume jars, spindles, tweezers, or ointment boxes. As a result, it is difficult to use archaeology to further explore the issue of intermarriage at Locri. Our understanding of the early settlement of Locri, as with Taras, relies overwhelmingly on literary evidence. The literary evidence has a solid base in myth and seems to have served as an aetiological tale to explain the matrilineal succession of Locri. Whatever the caveats, if such evidence is not contradicted by other material, the picture we have is that the Greek colonists included women as well as men and that the local population was driven away so that indigenous women were simply not available. This conclusion can only be, at best, tentative. ( ) 263 Redfield (2003) 220. ( ) 264 Whitehouse & Wilkins (1989) 106. 262 Redfield (2003) 207. Locri Epizephyrii Whether their presence in this tradition was considered unusual or typical is unclear today. Polybius (12.5.10) also states that, as in other colonies, the Locrians expelled the indigenous peoples: διότι καθ᾽ ὃν καιρὸν τοὺς Σικελοὺς ἐκβάλοιεν τοὺς κατασχόντας τὸν τόπον τοῦτον τῆς Ἰταλίας ‘they drove out the Sicels who occupied this place in Italy’. It is not clear whether or not some indigenous women were kept for marriage and reproductive purposes, but it should not be ruled out. 67 67 Locri has been excavated more extensively than Taras, yet the finds are of little use to our analysis. Locri (the cemeteries and sanctuaries in particular) were excavated by Paolo Orsi, beginning in 1890.262 As the excavation was so early, the excavation techniques were not as advanced as they have become, and as at other colonies such as Pithekoussai, Megara Hyblaea, and Syracuse, the bones from the graves were not considered important, and consequently, no attempts were made to determine the sex of the deceased. Furthermore, Orsi’s excavations of Locri were not thoroughly published. For example, at the Lucifero cemetery, 1675 graves were excavated but only a mere 162 were published.263 As a result of this selectivity, the sex ratios of the deceased appear skewed. The majority of the graves published by Orsi were those which were perceived to have interesting contents (grave goods). These tended to be items such as toys, and objects associated with women, such as needle cases, mirrors, perfume jars, spindles, tweezers, or ointment boxes. As a result, it is difficult to use archaeology to further explore the issue of intermarriage at Locri. Locri has been excavated more extensively than Taras, yet the finds are of little use to our analysis. Locri (the cemeteries and sanctuaries in particular) were excavated by Paolo Orsi, beginning in 1890.262 As the excavation was so early, the excavation techniques were not as advanced as they have become, and as at other colonies such as Pithekoussai, Megara Hyblaea, and Syracuse, the bones from the graves were not considered important, and consequently, no attempts were made to determine the sex of the deceased. Furthermore, Orsi’s excavations of Locri were not thoroughly published. For example, at the Lucifero cemetery, 1675 graves were excavated but only a mere 162 were published.263 As a result of this selectivity, the sex ratios of the deceased appear skewed. 262 Redfield (2003) 207. 263 ( ) ( ) 266 Carter (2006) 81. As such, Metapontum probably did not have a ‘formal foundation’ per se (see below). 267 y ( ) 269 Whitehouse & Wilkins (1989) 115. 268 Whitley (2001) 125. 269 265 Carter (1998) 5; Carter (1981) 167. 266 265 Carter (1998) 5; Carter (1981) 167. ( ) 270 Malkin (1994a) 121. 270 Malkin (1994a) 121. 267 Carter (2006) 81. 268 Whitley (2001) 125 ( ) 271 Carter (1998) 6. 267 Carter (2006) 81. 265 Carter (1998) 5; Carter (1981) 167. 266 Carter (2006) 81. As such, Metapontum probably did not have a ‘formal foundation’ per se (see ) 267 Carter (2006) 81. 267 Carter (2006) 81. ( ) 271 Carter (1998) 6. Whitley (2001) 125. 269 Whitehouse & Wilkins (1989) 115. ( ) , p p y p ( below). 267 Carter (2006) 81. Metapontum The Greek colony of Metapontum, modern Metaponto, was founded in c. 650 BC by colonists from Achaia and Troizen in the northern Peloponnese.264 Located about 40 kilometres south west of Taras, the settlement covered an extensive area. The entire chora of Metapontum has been thoroughly assessed, aided by the fact 68 that the ancient city was abandoned in the fifth century BC and never reoccupied due to an ongoing threat of malaria.265 Like the other South Italian colonies examined in this thesis, Metapontum’s foundation myths are complicated and difficult to follow. Strabo (6.1.15) states that the Pylians, sailing home from the Trojan War, founded the city of Metapontum with king Nestor. He also records an alternative tradition whereby Metapontum was founded by an Achaian, Leukippus, who acquired the site through tricking the Tarantines. It is not imperative to analyse these foundation myths for the purposes of this thesis, but it is significant to note that the literary evidence does not allow for a specific group of Greeks to be identified as the ‘first’ Greeks in Metapontum.266 Carter believes that this is because the Greeks of Metapontum represented a variety of mother cities.267 The reasons for the foundation of Metapontum are not immediately obvious. Whitley points out that the settlement is not positioned near any trade routes, that it has only an average harbour, and is not situated near any natural resources which could have motivated Greek settlement.268 These factors, coupled with the knowledge that the Metapontine coin would later display an ear of wheat (symbolising agricultural prosperity),269 strongly suggest that Metapontum was settled simply to gain more agricultural land. It also seems to have been, at least in part, settled with the intention of preventing further Tarentine expansion southward.270 Metapontum was occupied prior to Greek arrival and the establishment of the colony. ( ) 274 Shepherd (2012) 226. This applied to individuals between the ages of 15 and 49 and from the period 500-301 BC (the majority of burials come from this period). 275 ( ) 277 Carter (1998) 509. p ( j y p ) 275 Carter (1998) 145. It seems unlikely that half of every generation of males were consistently killed in warfare. See Carter (1998) 509. 278 See Henneberg & Henneberg (2001); Henneberg (1998). ( ) 276 Carter (2006) 41-42; Shepherd (2012) 226. 277 Metapontum Mycenaean pottery has been uncovered suggesting that there was contact between the indigenous population and some Greeks (if not something more permanent) as far back as the fourteenth to thirteenth centuries BC.271 Greek pottery dating to the eighth and seventh centuries BC has also been discovered, though it is not clear whether there was a permanent Greek presence at this time, 69 and if there was, what effect it might have had on the indigenous population.272 The excavation of an Iron Age village at Incoronata (within Metapontum’s chora) has demonstrated that the indigenous settlement certainly had extensive contact with Greek traders, and probably also had some Greek inhabitants living within it.273 When the Greek colonists arrived subsequently, the indigenous population and the earlier Greeks already appear to have been coexisting peacefully, and there is little reason to assume that this would suddenly have changed. If this was already the norm in the area, intermarriage was a definite possibility open to the colonists. The Pantanello Necropolis within the chora of Metapontum demonstrates some unusual features. In particular, the focus of scholarship has centred on the skewed sex ratio: women outnumber men by 2:1.274 Biologically, males and females survive into adulthood in roughly equal numbers (taking into consideration exposure, death in warfare, death in childbirth, for example), so there must be another explanation for such an imbalance.275 One possibility is that men were buried in another, as yet unidentified, location away from their families.276 This skewed ratio has little relevance to the period of focus in this thesis; oligandria is given as a reason for the shortage of males in the cemetery, but this only became a problem in the fifth and fourth centuries BC, thus at least a century after the foundation. Carter states that in the initial period, c. 600-501 BC, the sex ratio seems to be 1:1, though he admits that the sample size is extremely small and so it is not possible to get an accurate result.277 Consequently, while this is interesting to speculate about, it is important not to get sidetracked and to acknowledge that this has little to do with solving the questions at the heart of this thesis. ( ) 273 Carter (1998) 6. 274 ( ) 276 Carter (2006) 41-42; Shepherd (2012) 226. 277 Carter (1998) 509. 278 272 Carter (1998) 6. 2 3 ( ) p ( ) 277 Carter (1998) 509. 278 272 Carter (1998) 6. 273 Carter (1998) 6. 274 Shepherd (2012) 226. This applied to individuals between the ages of 15 and 49 and from the period 500-301 BC (the majority of burials come from this period). 275 Carter (1998) 145. It seems unlikely that half of every generation of males were consistently killed in warfare. See Carter (1998) 509. 276 Carter (2006) 41-42; Shepherd (2012) 226. 277 Carter (1998) 509. 278 See Henneberg & Henneberg (2001); Henneberg (1998). Ca te ( 998) 509. ee Henneberg & Henneberg (2001); Henneberg (1998). Carter (1998) 6. 273 Carter (1998) 6. 274 ( ) 281 The area surrounding the settlement of Cyrene is frequently called ‘Cyrenaica’, but as Austin (2008) 187 (footnote 1) points out, this term is anachronistic for it was not used before the Augustan period. Herodotus (for example, at 4.199.1) uses the term ‘Cyrenaea’ to describe the territory of Cyrene. 279 Henneberg (1998): that is, metric being physical measurements, and non-metric looking at other characteristics that may be present in teeth (for example, the Carabelli Cusp, or the so-called Etruscan upper lateral incisor). This study acknowledged that metric and non-metric characteristics could well be controlled by different genetic factors, or be influenced by various environmental factors, affecting results. See Carter (2006) 82. 280 Metapontum Recently, intensive study from a scientific angle has been conducted on two of the populations within the chora of Metapontum (Pantanello and Crucinia).278 One study is of particular relevance, in which Henneberg assessed both the metric and 70 non-metric characteristics of teeth, and how these characteristics could be used as indicators of ethnicity and sexual dimorphism.279 In brief, the overall results of this analysis demonstrated that there was indeed a biological relationship between the Greeks living in Metapontum and the indigenous Italics, and that this relationship was more prevalent among those living within the chora of Metapontum than those living in the city itself.280 This suggests that intermarriage was more common in the country than in the urban environment. This could in turn indicate that some Greek women were present in the city to allow for the less pronounced combination of characteristics. We must recognise that this study looks at individuals not just from the period of foundation, but over a number of centuries. This still remains relevant, however, as it is possible to determine the extent of intermarriage over the years, and the evidence allows us to assume that intermarriage was also a feature of the earliest generations of analysis (the first generations of colonists). At Metapontum, therefore, there was a long period of interaction prior to the archaic colonisation of the Greeks. After settlement there was a greater concentration of the newcomers in the countryside, easing their presence in a way that more forceful takeover in the chora might not. , g ( ) 280 Carter (2006) 82-83. , g 280 Carter (2006) 82-83. 281 282 Osborne (2009) 12 argues that it is unsurprising that divergent accounts exist; it was advantageous for the Theraeans to uphold their links with their colony because it had become more prosperous than them (as the mother city). The Cyrenaeans, by contrast, had no reason to maintain links with Thera. Accordingly, Osborne argues “it is vain to seek historical truth from either account [in Herodotus]. Neither account was interested in a full and frank record of what happened; both parties were telling stories of the past to suit the present, stories that were not only selective in their memories but free in their elaboration.” While I agree with this, I would add that the textual evidence of Cyrene is really no different from any other colony in that the historicity of these sources should not be taken for granted. Foundation tales, in particular, are heavily grounded in myth as the Greeks enjoyed speculating about their beginnings (see Dougherty (1993a)). It is also important to note that the Greeks did not necessarily distinguish between ‘myth’ and ‘reality’. 283 This is the only instance I have identified where we are told exactly how the colonising party was selected. This could be because the Cyrenaean expedition was particularly well organised, while other colonising ventures were put together in a more ‘hodge-podge’ way. 284 C k ll (1992) 291 Cawkwell (1992) 291. 285 This decree dates from the fourth century BC, but it appears to quote from the original seventh century BC decree of Thera on the foundation of Cyrene. There has been much debate about the authenticy of the decree, though this is not especially relevant to discuss here. See Austin (2008) 190 Also Graham (1960/2001) and Jeffery (1961) for discussion in favour of authenticity and ( ) 285 This decree dates from the fourth century BC, but it appears to quote from the original seventh century BC decree of Thera on the foundation of Cyrene. There has been much debate about the authenticy of the decree, though this is not especially relevant to discuss here. See Austin (2008) 190. Also Graham (1960/2001) and Jeffery (1961) for discussion in favour of authenticity, and Dusanic (1978) for discussion against. 286 ( ) 286 Marshall (2004) 128. ( ) 287 Graham (1981-82) 301. Cyrene Cyrene is located in North Africa, modern day Shahhat – Jabal al Akhdar.281 The settlement was founded by colonists from Thera in c. 630 BC. Cyrene is one of the most well known of the Greek colonies, particularly due to an extensive discussion by Herodotus, who uses the Persian expedition to Libya in c. 514 BC as an opportunity to discuss his other knowledge of Libya. The foundation of Cyrene 71 is particularly well documented, with Herodotus reporting two separate versions – one of the Theraeans (4.150-153) and one of the Cyrenaeans (4.154-156) – as well as a fourth century BC inscription purporting to be the foundation decree.282 There are also a number of references to intermarriage in later source material. Herodotus (4.151.1) reports the Theraeans’ belief behind the impetus for the foundation of Cyrene: ἑπτὰ δὲ ἐτέων µετὰ ταῦτα οὐκ ὗε τὴν Θήρην, ἐν τοῖσι τὰ δένδρεα πάντα σφι τὰ ἐν τῇ νήσῳ πλὴν ἑνὸς ἐξαυάνθη. χρεωµένοισι δὲ τοῖσι Θηραίοισι προέφερε ἡ Πυθίη τὴν ἐς Λιβύην ἀποικίην. For seven years following this, there was no rain at Thera and in that time all the trees on the island, except one, dried up. The Pythia, who the Theraeans consulted, proposed a colony in Libya. Herodotus clearly states that the Theraeans believed the colonists set out from Thera because of drought. If this was the case, it would be more likely that a cross section of the community would have left, including women, if not children as well, rather than just men alone. But as will be shortly discussed, there are few indications that women and children accompanied the colonists. Herodotus (4.156) tells us that in the Cyrenaean version of the story, when the colonists failed to settle in Libya, and attempted to return to Thera, the Theraeans refused to allow them back into the city, and indeed would not even let them land their boats. This indicates that the relations between the colonists and the Theraeans were frosty. Accordingly, if we are to follow the Cyrenaean version, it is more likely that there was some sort of stasis in Thera, and that a group (possibly led by Battus) was, in effect, exiled. This view is reflected in the scholiast to Pindar, Pythian Odes 4.10a (Menecles of Barca, FGrHist 270 F 6) where civil disruption caused the faction led by Battus to be driven from Thera. authenticy of the decree, though this is not especially relevant to discuss here. See Austin (2008) 190. Also Graham (1960/2001) and Jeffery (1961) for discussion in favour of authenticity, and Dusanic (1978) for discussion against. 286 g 284 Cawkwell (1992) 291. 285 Cyrene 72 Herodotus (4.153) states that, accordingly to the Theraean version, the members of the colonising party were drawn by lot:283 Herodotus (4.153) states that, accordingly to the Theraean version, the members of the colonising party were drawn by lot:283 Θηραίοισι δὲ ἕαδε ἀδελφεόν τε ἀπ᾽ ἀδελφεοῦ πέµπειν πάλῳ λαγχάνοντα καὶ ἀπὸ τῶν χώρων ἁπάντων ἑπτὰ ἐόντων ἄνδρας. The Theraeans decided to send brother from brother, obtained by lot, men from all seven regions [in Thera]. Cawkwell argues that the position of the word ἄνδρας in the sentence is emphatic and thus could be indicating that it was only men who went on the colonising expedition.284 The Foundation Decree of Cyrene (ML 5.27-30, or SEG 9.3.28- 31)285 similarly states that: τᾶι ὁµοίαι πλὲν κατὰ τὸν οἶκον, υἱὸν δὲ ἕνα,καταλ[έ]\|γεσθαί τ[ε ἀπὸ τῶγ χώρων ἁπάντων] τοὺς ἡβῶντας, καὶ τῶν [ἄλǁ‖λ]ων Θηραίων ἐλευθέρος, [ὅ καλῆι], πλέν… …They are to sail according to house, following both equal and similar terms, with one son picked out from each and free persons of the other Theraeans are to sail… Although no mention is made of intermarriage in either source, Marshall has suggested that in line with the exclusively masculine nature of the colonisation implied in the sources, marriage with indigenous women was a fundamental part of the process of the foundation.286 I am less convinced that such an argumentum e silentio provides a full answer. That is to say, the lack of any mention of women does not necessarily mean that they were not involved in the process of foundation.287 While the colonists seem to have blundered their way through the foundation, spending their first two years on the island of Platea (Herodotus 4.157) and a 73 further six years at a settlement called Aziris (Herodotus 4.158), their relations with the indigenous Libyans following this seem to have been friendly. The interaction between the Greeks (Cyrenaeans) and indigenous populations (Libyans) is made clear in a number of sources. Herodotus (4.189.1) presents this interaction as a two-way process stating, for example, that: τὴν δὲ ἄρα ἐσθῆτα καὶ τὰς αἰγίδας τῶν ἀγαλµάτων τῆς Ἀθηναίης ἐκ τῶν Λιβυσσέων ἐποιήσαντο οἱ Ἕλληνες. The Greeks copied the clothing and aegis of the statues of Athena from the Libyans. The Greeks copied the clothing and aegis of the statues of Athena from the Libyans. 288 Stucchi (1989) 83. Stucchi also analyses various representations of the quadriga in Cyrene, which back up the evidence in Herodotus, particularly their use as transport vehicles (rather than racing or war chariots). 289 g ) 289 Applebaum (1979) 13. acing or war chariots). 89 Applebaum (1979) 13. Cyrene In addition, Herodotus (4.189.3) also states that the Libyans taught the Greeks to harness and drive the four-horse chariot.288 On the other hand, Herodotus acknowledges that the Greeks taught the Libyans many things, and the adoption and adaptation of customs was mutual.289 The Libyan tribe, the Asbystae, for example, is marked out by Herodotus (4.170) in the way that their general way of life imitated that of Cyrene closely. Indeed, such was the relationship between the colonists and the Libyans that Herodotus (4.158) reports that the Libyans physically led the blundering colonists to the place of ultimate settlement. Admittedly, the Libyans seem to have ulterior motives for this action: they led the Theraeans in the dark so that they would be able to pass through the best parts of the country without them seeing it. There is further literary evidence specifically pointing to a strong tradition of intermarriage between the Greeks and the Libyan women at Cyrene which may have begun at the inception of the colony. The third century BC poet, Callimachus, who came from Cyrene, claimed in his Hymn to Apollo (85-86): ὅτε ζωστῆρες Ἐνυοῦς ἀνέρες ὠρχήσαντο µετὰ ξανθῇσι Λιβύσσαις. ὅτε ζωστῆρες Ἐνυοῦς ἀνέρες ὠρχήσαντο µετὰ ξανθῇσι Λιβύσσαις. The warrior-belted men of war [the colonists] danced with the golden- haired Libyans. Scholars typically quote this short passage as evidence of some sort of interaction between the male colonists and the indigenous female Libyans (that could be Scholars typically quote this short passage as evidence of some sort of interaction between the male colonists and the indigenous female Libyans (that could be 74 explained by intermarriage).290 However, prostitution could also account for the interaction described, or the passage could be a military reference. Pindar’s Pythian (9.105-125) also illustrates a tradition (albeit rooted in myth) of intermarriage in Cyrene. Pindar describes the marriage between the Cyrenaean Alexidamus and the unnamed daughter of the Libyan Antaeus. Many Cyrenaeans had come hoping to marry the girl. Antaeus decided to organise a race to decide which man would marry his daughter. She was placed on the finish line, and the first man to touch her robes would become her husband. Alexidamus won the race and the resulting wedding is described in order to glorify the city. This story not only reports this instance of intermarriage, but could also be symbolic of intermarriage between Cyrenaean men and Libyan women more generally.291 The late fourth century (321 BC) edict of Ptolemy I which established the constitution of Cyrene, proclaimed that the sons of Cyrenaean men and Libyan women should have full citizen rights (SEG 9.1.1-3). This demonstrates that by the late fourth century BC intermarriage certainly occurred commonly, and indeed occurred despite restrictions prior to the passing of the edict. On the other hand, however, it also suggests that prior to Ptolemy’s constitution, intermarriage only produced non-citizens. Ptolemy’s edict could also, perhaps, indicate a continuous tradition of intermarriage at Cyrene. This would appear to contrast with other Greek colonies, at least in terms of the limitations of scholarly discussion, where intermarriage is usually only suggested as a short-term solution to a lack of women, rather than an ongoing practice as is indicated in this edict. In 1960, Rosenbaum conducted a study of Cyrenaean portrait sculpture, comparing distinctive features of Greek figures versus Libyan. ( ) ; ( ) ; y ( ) ; g 291 Marshall (2004) 129; Cawkwell (1992) 291; Calame (1990) 302-303. 292 b ( ) 290 Marshall (2004) 129; Graham (1981-82) 296; Murray (1993) 115; Rouge (1970) 316. 292 Rosenbaum (1960) 21. ( ) ; 292 Rosenbaum (1960) 21. ὅτε ζωστῆρες Ἐνυοῦς ἀνέρες ὠρχήσαντο µετὰ ξανθῇσι Λιβύσσαις. For example, the features of CPS 1 (Figure 7), a Libyan man, are compared to non-Libyans CPS 188, 189, 190, 191, and 199 (Figures 8 and 9).292 These examples each have features obviously not conforming to Greek, Roman, or Egyptian ethnic types, such as tight curly hair, full lips, and high, prominent cheekbones, that are instead apparent in CPS 1. From this, Rosenbaum concludes that: “that we find these traits at all is a proof of 75 75 the mixture of races that had taken place”.293 Many of these busts, however, date considerably after the period of foundation. CPS 1 dates to the second half of the fourth century BC, while the majority of others (CPS 188, 189, 190, 191, and 199) date as late as the first century AD. Therefore, like the edict of Ptolemy I, these busts are good evidence for intermarriage occurring at these times but we cannot necessarily conclude that intermarriage took place between the Theraean colonists and Libyans at the time of the colony’s foundation. A limestone relief, dating to the Hellenistic period, depicting a female deity and a pastoral scene, also combines a mixture of Greek and Libyan elements which “surely reflects the good relations between the Greek and native populations of Cyrene” (Figure 10).294 While both the busts and the limestone relief are far too late to be conclusive evidence of intermarriage occurring at the time of foundation, they are indicative of a continuous practice of intermarriage (also indicated by Ptolemy’s edict, discussed above). On the other hand, however, conclusions based on physical features for ‘racial profiles’ are not necessarily sound. Herodotus (4.186) describes how the women of Cyrene observed certain food taboos, which were the same as those in Libya.295 οὕτω µὲν µέχρι τῆς Τριτωνίδος λίµνης ἀπ᾽ Αἰγύπτου νοµάδες εἰσὶ κρεοφάγοι τε καὶ γαλακτοπόται Λίβυες, καὶ θηλέων τε βοῶν οὔτι γευόµενοι, διότι περ οὐδὲ Αἰγύπτιοι, καὶ ὗς οὐ τρέφοντες. βοῶν µέν νυν θηλέων οὐδ᾽ αἱ Κυρηναίων γυναῖκες δικαιοῦσι πατέεσθαι διὰ τὴν ἐν Αἰγύπτῳ Ἶσιν, ἀλλὰ καὶ νηστηίας αὐτῇ καὶ ὁρτὰς ἐπιτελέουσι. Thus between Egypt and the Lake Tritonus are Libyan nomads who eat meat and drink milk, though they don’t eat cows nor rear swine, for the same reasons as the Egyptians. The women of Cyrene don’t allow the consumption of cows because of Isis of Egypt, and also they honour her with fasts and festivals. Wanis (1992) 41. 295 Boardman (1980) 155; Van Compernolle (1983) 1044. ( ) 294 Wanis (1992) 41. ( 294 Wanis (1992) 41. 293 Rosenbaum (1960) 22. 293 Rosenbaum (1960) 22. 294 293 Rosenbaum (1960) 22. 294 293 Rosenbaum (1960) 22. 294 Wanis (1992) 41. 295 d (1980) 1 C ll (1983) 1044 LSJ sv. βατταρίζω. 301 Applebaum (1979) 13. Malkin (1987) 63 however, draws to attention to the Greek tradition (still in practice today) of naming the first born son after their paternal grandfather – Battus’ grandson was indeed Battus, so-called the Fortunate – which implies the use of ‘Battus’ was more of a name than a title. 298 Laronde (1990) 178. 299 White (1987) 79. Parke & Wormell (1956) 74 use this point to argue that the oracle given to Battus must be anachronistic. 300 LSJ sv βατταρίζω 296 Marshall (2004)130. ( ) 297 Marshall (2004) 127-128. 298 Laronde (1990) 178. 300 LSJ sv. βατταρίζω. 301 ὅτε ζωστῆρες Ἐνυοῦς ἀνέρες ὠρχήσαντο µετὰ ξανθῇσι Λιβύσσαις. This, of course, appears as a retrospective aetiological tale, characteristic of Herodotean writing, to explain why such a practice would exist in his own time. Despite this, by the fifth century BC when Herodotus was writing, the custom, whatever its origins, must have been practised by enough Libyan women that they 76 were able to influence Cyrenaean religious customs applying to eating rules and taboos.296 The custom is clear evidence of cross-cultural contact. Onomastic studies can further provide insight into the relationship between Cyrenaeans and Libyans. The occurrence of a significant number of Libyan names in Cyrene’s inscriptions suggests, firstly, that the Libyans were integrated within Cyrenaean society, and secondly, that the Cyrenaeans and Libyans did indeed have close relations with one another, possibly including intermarriage. 297 We should be mindful however, as Laronde puts it, “not everyone with a Libyan name is necessarily Libyan, nor does a Greek name imply a purely Greek origin of its bearer”.298 Herodotus (4.155) had learned that ‘battus’ meant ‘king’ in the Libyan language and suggests accordingly that this may have been a title imposed on Battus later in his life (that is, once he was king in Cyrene).299 However, he also reports that Battus’ name could have been engendered from his supposed stammer, alluding to the verb βατταρίζειν meaning ‘to stammer’.300 As Applebaum points out, the first option, that Battus gained his name through his role as king, is in accordance with other sources (Diodorus 8.29; Pindar Pythian 5.87; SEG 9.189) who state that Battus originally possessed the name Aristoteles.301 Another example of using onomastics to assess the relations between the Cyrenaeans and the Libyans is King Alazeir of Barca (another Greek colony in Libya). Due to his name, he was originally thought to be a Libyan, but Herodotus (4.164.4) states that his daughter was the wife of Arcesilaus III (reigning 530-515 BC) of Cyrene, to whom she was also related by blood, thus making her a member of the Battiad dynasty. This suggests that either Alazeir was a Libyan and his daughter’s Battiad descent came from her mother; or that Alazeir was a Battiad 77 with a Libyan name. 302 Graham (1981-82) 296-297. ( ) 303 Laronde (1990) 178-179. 305 Laronde (1990) 179. ( ) 306 Laronde (1990) 179. 304 Laronde (1990) 179. 302 Graham (1981-82) 296-297. 303 Laronde (1990) 178-179. 304 Laronde (1990) 179. 305 Laronde (1990) 179. 306 Laronde (1990) 179. 307 Compare this to Thasos, where Archilochus (F 102 West) described how ‘the scum of all Greece flocked to Thasos’ so that they might compete against the Thracian tribes for the gold mines on the island. Herodotus (6.46-47) and Thucydides (1.100.2) attest to the affluence of these mines. See Van Wees (2004) 29, 261. These are examples of so-called ‘proto-colonists’ joining an already established colony. ὅτε ζωστῆρες Ἐνυοῦς ἀνέρες ὠρχήσαντο µετὰ ξανθῇσι Λιβύσσαις. The latter suggests that intermarriage could have occurred in an earlier generation, hence the Libyan name.302 A second or first century BC list of 90 names found at el-Gubba, 44 kilometres east of Cyrene, gives us further insight into this issue.303 The names in the list, laid out in SEG 9.348, are probably too many to correspond to a group of magistrates, and instead, Laronde suggests that they are related to the optimum size of a sedentary population that was able to live off the surrounding land.304 More central to our discussion, both the names and patronyms of individuals are given in this list. The result of an analysis of these names leaves us with only seven individuals of indigenous (Libyan) origin, 19 individuals of Cyrenaean (Greek) origin, and the remainder indistinguishable.305 Even with such a large number of the total names being indistinguishable, this list provides a number of insights. Firstly, it demonstrates that the community was indeed a mixed one. Certainly Cyrenaeans and Libyans were living side by side, and it would not take much imagination to suppose some intermarriage may have taken place. The origin of this list is important as it demonstrates that the Greek presence in Libya had dispersed significantly by the second century BC, at least from Cyrene to el-Gubba, and consequently it is possible to postulate that interpenetration with the indigenous population took place.306 Despite all the evidence discussed above, it is difficult to draw any firm conclusions about intermarriage as a widespread practice at Cyrene’s inception. The evidence discussed so far covers dates from at least 200 years after the Greek colonists first arrived. Although much of this presents intermarriage as likely, there are also indications in Herodotus that relations between the Greeks and Libyans may not have been as amicable as they appeared, discussed further below. 78 Herodotus (4.159.2-4) decribes the influx of new settlers to Libya and states that the newcomers were invited to take part in the division of land:307 ἐπὶ δὲ τοῦ τρίτου, Βάττου τοῦ εὐδαίµονος καλεοµένου, Ἕλληνας πάντας ὥρµησε χρήσασα ἡ Πυθίη πλέειν συνοικήσοντας Κυρηναίοισι Λιβύην. ἐπεκαλέοντο γὰρ οἱ Κυρηναῖοι ἐπὶ γῆς ἀναδασµῷ, ἔχρησε δὲ ὧδε ἔχοντα, “ὃς δέ κεν ἐς Λιβύην πολυήρατον ὕστερον ἔλθῃ γᾶς ἀναδαιοµένας, µετὰ οἷ ποκα φαµὶ µελήσειν”. συλλεχθέντος δὲ ὁµίλου πολλοῦ ἐς τὴν Κυρήνην, περιταµνόµενοι γῆν πολλὴν οἱ περίοικοι Λίβυες. y 308 Stucchi (1989) 75; Mitchell (2000) 86. Mitchell dates this battle to shortly before 570 BC. As a result of the defeat, Apries lost his throne to Amasis, who became pharaoh in 570 BC. Herodotus (2.161-163) tells this story in his account of Egypt. 309 ( ) y gyp 309 Mitchell (2000) 87. ( ) 309 Mitchell (2000) 87. ὅτε ζωστῆρες Ἐνυοῦς ἀνέρες ὠρχήσαντο µετὰ ξανθῇσι Λιβύσσαις. In the time of the third ruler, known as Battus the Fortunate, the Pythia gave an oracle urging all the Greeks to sail [to Cyrene] so as to dwell with the Cyrenaeans in Libya. For the Cyrenaeans called them in to a land distribution. [The Pythia] proclaimed this: “whoever should go to much-loved Libya late [after the land had been distributed], I say to him that he will regret it forever”. A great multitude gathered at Cyrene, cutting off much land from the neighbouring Libyans. Herodotus goes on to describe how relations between the Cyrenaean Greeks and the Libyans began to break down, with the loss of the latter’s land. The Libyans, resenting the loss of their territory appealed both to their king, Adicran, and the Egyptian king Apries. A battle at Irasa followed, and the Libyans and Egyptians were badly defeated by the Cyrenaeans.308 Such a breakdown of relations between the Cyrenaeans and Libyans could imply that intermarriage was less likely to occur at this time. Herodotus goes on to describe how relations between the Cyrenaean Greeks and the Libyans began to break down, with the loss of the latter’s land. The Libyans, resenting the loss of their territory appealed both to their king, Adicran, and the Egyptian king Apries. A battle at Irasa followed, and the Libyans and Egyptians were badly defeated by the Cyrenaeans.308 Such a breakdown of relations between the Cyrenaeans and Libyans could imply that intermarriage was less likely to occur at this time. Cyrene appears to have had a rapid population growth following this. Herodotus (4.160) states that in the battle at Leucon, the Cyrenaeans lost as many as 7000 hoplites. While we must be wary of the accuracy of Herodotus’ figures, it is plausible that this could reflect a very large population expansion in the reign of Battus II and the early stages of the reign of his son, Arcesilaus II.309 Possibly this could be due to intermarriage, although for such a dramatic increase in such a short period of time, a considerable number of Cyrenaean Greek men would have married Libyan women near the start of Battus II’s twenty-odd year reign, so that 79 the offspring of these unions could be old enough to fight as hoplites in the Battle at Leucon.310 This does seem plausible. 310 Intermarriage to produce ‘citizen’ offspring could only be between Cyrenaean Greek men and Libyan women; unions between Cyrenaean women and Libyan men would not be acknowledged as marriages. See Mitchell (2000) 99. 310 Intermarriage to produce ‘citizen’ offspring could only be between Cyrenaean Greek men and ib i b d ib ld b k l d d Intermarriage to produce citizen offspring could only be between Cyrenaean Greek men and Libyan women; unions between Cyrenaean women and Libyan men would not be acknowledged as marriages. See Mitchell (2000) 99. 311 Whitley (2001) 125 describes Massalia as the ‘last major Greek colony’. This description assumes that the Greek colonies of the Black Sea were almost part of a different ‘movement’; a view that is not subscribed to in this thesis. There is also debate over whether Massalia and Emporion were founded concurrently, or whether Massalia was founded first, before the colonists moved south to Emporion. See Mierse (1994) 790-794. 312 ὅτε ζωστῆρες Ἐνυοῦς ἀνέρες ὠρχήσαντο µετὰ ξανθῇσι Λιβύσσαις. Another way in which the population of Cyrene could have dramatically increased is by the inclusion of Libyans (males) within their own population count. This seems unlikely due to the breakdown of relations between the two peoples during the reign of Battus II, leading to the Battle at Irasa. A further shipment of Greeks from the mainland during the reign of Battus II may have augmented Cyrene’s population. The influx of mainland Greeks certainly was considerable enough to make an impact on the surrounding Libyans who were displaced from their land. Thus, it seems reasonable to suppose that these new colonists could have significantly boosted the Cyrenaean population, to the extent that they had 7000 hoplites who died at Leucon. In particular, an even greater number of mainland Greeks could have been absorbed into the population if they had settled into the wider territory surrounding Cyrene, not just the city itself. The fact that the Greeks were invited to share in land distribution and that the oracle proclaimed that settlers should go to Libya (rather than Cyrene), could suggest that the Greeks were not merely contained within the city, Cyrene. It would appear that the influx of Greek colonists could have boosted the population significantly. Of course, it is more likely that the population expansion was the result of a combination of some intermarriage together with a large number of settlers from Greece. At first glance, the evidence suggests that intermarriage between colonists and the indigenous women took place at Cyrene – the possibility that it was men alone who went on the initial colonising venture, the impacts of each society on the customs and practices of the other, the literary evidence, and the edict of Ptolemy I which may have been an attempt to regularise existing practice. All these points argue in favour of intermarriage. To counter that, the sudden increase in the number of colonists during the reign of Battus II and the subsequent breakdown in relationship between the two groups points, at the least, to a less favourable time 80 for intermarriage to occur. On balance, however, and over time, intermarriage does seem rather more likely than not. p 312 Dietler & Lopez-Ruiz (2009) 10. 312 Dietler & Lopez-Ruiz (2009) 10. ( ) 314 Little is known about Justin, other than that he was an historian who lived under the Roman Empire. His exact dates are uncertain. Nevertheless, even the beginning of the Roman Empire is considerably after the foundation of Massalia and so our analysis must take into consideration his lateness as a source. 313 Shefton (1994) 61. 314 Massalia Massalia, modern day Marseilles, is located in France. Both Thucydides (1.13) and Strabo (4.1.4) tell us that Massalia was founded by the Phocaeans. It appears to have been settled c. 600 BC, around the same time as Emporion (in Spain, discussed briefly in Chapter One).311 Massalia would later become one of the most important centres for trade and was the largest city in the Western Mediterranean (indigenous or Greek) until after the Roman conquest.312 The literary evidence offers a clear example of intermarriage between the Greeks and the indigenous population. Athenaeus (13.576a-b), quoting from Aristotle’s (now lost) Constitution of the Massaliotes, records the intermarriage between the indigenous king’s daughter and the Greek leader: Φωκαεῖς οἱ ἐν Ἰωνίᾳ ἐµπορίᾳ χρώµενοι ἔκτισαν Μασσαλίαν. Εὔξενος δὲ ὁ Φωκαεὺς Νάννῳ ὁ τῷ βασιλεῖ τοῦτο δ᾽ ἦν αὐτῷ ὄνοµα ἦν ξένος, οὗτος ὁ Νάννος ἐπιτελῶν γάµους τῆς θυγατρὸς κατὰ τύχην παραγενόµενον τὸν Εὔξενον παρακέκληκεν ἐπὶ τὴν θοίνην. ὁ δὲ γάµος ἐγίγνετο τόνδε τὸν τρόπον ἔδει µετὰ τὸ δεῖπνον εἰσελθοῦσαν τὴν παῖδα φιάλην κεκερασµένην ᾧ βούλοιτο δοῦναι τῶν παρόντων µνηστήρων ᾧ δὲ δοίη, τοῦτον εἶναι νυµφίον. ἡ δὲ παῖς εἰσελθοῦσα δίδωσιν εἴτε ἀπὸ τύχης εἴτε καὶ δι᾽ ἄλλην τινὰ αἰτίαν τῷ Εὐξένῳ: ὄνοµα δ᾽ ἦν τῇ παιδὶ Πέττα…ἔλαβεν ὁ Εὔξενος γυναῖκα καὶ συνῴκει µεταθέµενος τοὔνοµα Ἀριστοξένην. καὶ ἔστι γένος ἐν Μασσαλίᾳ ἀπὸ τῆς ἀνθρώπου µέχρι νῦν Πρωτιάδαι καλούµενον. Πρῶτις γὰρ ἐγένετο υἱὸς Εὐξένου καὶ τῆς Ἀριστοξένης. The Phocaeans of Ionia were traders and founded Massalia. Euxenus of Phocaea was a guest-friend of King Nannus, which was actually his name. Nannus was celebrating his daughter’s wedding, and since Euxenus happened to be there, he invited him to the feast. The wedding was organised in this way: after the meal, going in, the girl was to give mixed wine to whichever suitor there she might want, and whoever she gave it to would be her bridegroom. Having entered the 81 room, either by chance, or by some other reason, the girl gave it to Euxenus: her name was Petta… Euxenus married the girl and set up dwelling, changing her name to Aristoxene. And there is a family in Massalia now descended from her, called the Protidae, for Protis was the son of Euxenus and Aristoxene. room, either by chance, or by some other reason, the girl gave it to Euxenus: her name was Petta… Euxenus married the girl and set up dwelling, changing her name to Aristoxene. Massalia And there is a family in Massalia now descended from her, called the Protidae, for Protis was the son of Euxenus and Aristoxene. In this tradition, it is apparent that the Greek oikistes was a guest-friend of the indigenous leader. This suggests that firstly, the area had been previously explored prior to Massalia’s establishment; and secondly, that the relationship between the oikistes and the indigenous king was a very friendly one, so much so that land was given to him to settle in the area that came to be known as Massalia. Archaeological exploration has not revealed any evidence of prior settlement on the site, so it appears that the land given to the Greeks was virgin.313 This is comparable to the founding of Megara Hyblaea in Sicily, discussed above. Similarly, summarising Pompeius Trogus, Justin (43.3.8-11) reports: Similarly, summarising Pompeius Trogus, Justin (43.3.8-11) reports: Duces classis Simos et Protis fuere. Itaque regem Segobrigiorum, Nannum nomine, in cuius finibus urbem condere gestiebant, amicitiam petentes conveniunt. Forte eo die rex occupatus in apparatu nuptiarum Gyptis filiae erat, quam more gentis electo inter epulas genero nuptum tradere illic parabat. Itaque cum ad nuptias invitati omnes proci essent, rogantur etiam Graeci hospites ad convivium. Introducta deinde virgo cum iuberetur a patre aquam porrigere ei, quem virum eligeret, tunc omissis omnibus ad Graecos conversa aquam Proti porrigit, qui factus ex hospite gener locum condendae urbis a socero accepit. The leaders of the fleet were Simos and Protis. And so they met the king of the Segobrigi, whose name was Nannus, in whose territory they were eager to found a city, seeking friendship. By chance that day, the king was busy with preparation for the wedding of his daughter Gyptis, whom by the custom of his race he was prepared to give as bride to one chosen from among the banquet guests as a son in law. And thus, since all the suitors invited to the wedding were there, the Greeks were asked as guests to the feast. Then, the girl was led in and commanded by her father to pass the water to whomever of the men she chose. When she ignored everyone and went over to the Greeks, she passed the water to Protis, who, having gone from stranger to son-in-law, received a place from his father-in-law for his city to be founded.314 82 This passage has clear parallels with Athenaeus. p y 316 As such, this has led Lyons (2000) 89 to describe the union as ‘legitimate intermarriage’. I, however, am not making a point of distinguishing between forced intermarriage and willing. 317 Lyons (2000) 89 315 Morel (2006) 365 accepts that even with the limitations of these sources, the stories probably contain a ‘core of plausibility’. Lyons (2000) 89. 318 Van Compernolle (1983) 1040. y ( ) 318 Van Compernolle (1983) 1040. , 317 Lyons (2000) 89. 318 Massalia While the names differ (with Euxenus now called Protis, and Petta called Gyptis), the basic events are the same.315 It is interesting to note that the marriage between Protis and Gyptis does not appear to be forced; that is, she was not physically seized by him (compare, for example, the Sabine women).316 In the evidence it appears that Gyptis picked her own husband. Furthermore, Justin clearly reports that all the colonists were invited, and so cannot be considered unwelcome. Lyons has interpreted this marriage as a symbolic transmission of power, legitimising the Phocaeans’ claim to land surrounding their new settlement.317 In direct contrast to this literary evidence indicating that intermarriage took place right from the start of the foundation of Massalia, we are also told by Strabo (4.1.4) that at least one woman was indeed taken from the homeland to the new settlement. ἀπαίρουσι γὰρ τοῖς Φωκαιεῦσιν ἐκ τῆς οἰκείας λόγιον ἐκπεσεῖν φασιν ἡγεµόνι χρήσασθαι τοῦ πλοῦ παρὰ τῆς Ἐφεσίας Ἀρτέµιδος λαβοῦσι…Ἀριστάρχῃ δὲ τῶν ἐντίµων σφόδρα γυναικῶν παραστῆναι κατ᾽ ὄναρ τὴν θεὸν καὶ κελεῦσαι συναπαίρειν τοῖς Φωκαιεῦσιν ἀφίδρυµά τι τῶν ἱερῶν λαβούσῃ: γενοµένου δὲ τούτου καὶ τῆς ἀποικίας λαβούσης τέλος, τό τε ἱερὸν ἱδρύσασθαι καὶ τὴν Ἀριστάρχην τιµῆσαι διαφερόντως ἱέρειαν ἀποδείξαντας. For, they say that when they were leaving their home, [the Pythia?] sent an oracle to the Phocaeans commanding them to take from Artemis of Ephesus a leader for the expedition… the goddess appeared in a dream to Aristarcha, [one] of the most honourable women, and told her to sail away with the Phocaeans, taking an image of the holy things: when these things had happened and the colony was founded, they set up the temple and honoured Aristarcha, appointing her priestess above all. This passage from Strabo clearly reveals that Aristarcha travelled with the colonists to Massalia. However, Aristarcha is described as being ‘of the most honourable women’ thus distinguishing her from other women, most probably by her status and influence.318 Graham points out her exceptional position is emphasised by the fact that her name has been preserved in the record; other than 83 the names of the oikistai, it is rare for names to have been recorded.319 It also appears that Aristarcha was notable in this tale for being the only woman who accompanied the colonists on the voyage. p ( ) ; p ( ) 321 Graham (1995) 13; see also Schaps (1979) 73 for the importance of women in Greek religion. 322 G h (1981 82) 304 ( ) 320 Shepherd (2012) 220; Shepherd (1999) 270. 322 Graham (1981-82) 304. Graham (1995) 13; see also Schaps (1979) 73 for the importance of women in Greek religion. 322 Graham (1981-82) 304. p ( ) ; p ( ) 321 Graham (1995) 13; see also Schaps (1979) 73 for the importance of women in Gre 322 21 Graham (1995) 13; see also Schaps (1979) 73 for the importance of women in Greek religion. 22 Graham (1981-82) 304. ( ) 320 Shepherd (2012) 220; Shepherd (1999) 270. 319 Graham (1981-82) 302-303. 320 319 Graham (1981-82) 302-303. ( ) ; p ( ) p 322 Graham (1981-82) 304. Massalia Indeed, Shepherd argues that these records of the activities of such females neither preclude nor include similar participation by other women.320 Either way, however, the actions of one woman (especially those of a priestess) cannot be extrapolated to all females. Graham argues that priestesses would always have been transported to the new colonies: “all Greek colonies would need Greek women in order to ensure a proper relation between the community and its gods, and it’s absurd to imagine that native women, who did not even speak Greek, could be entrusted with these important tasks.”321 In addition to Massalia, there is evidence that at Thasos a priestess named Kleoboia was also transported to the new colony. Pausanias (10.28.3) describes a painting at Delphi by Polygnotus, depicting the girl in Charon’s boat holding a chest associated with Demeter on her lap, bolstering the opinion that Kleoboia was the first to bring the rites of Demeter from Paros to Thasos. Graham argues that it was common for females to serve female deities, and males to serve male deities (with exceptions, of course), hence the importance of having Greek women in the colonies.322 That said, however, it does not mean that every single woman in a colony would need to be Greek – rather, one Greek priestess could be transported when required for religious purposes, and indigenous women used for marriage, reproductive, or other purposes. Further, this line of argument does not consider the potential for syncretic religions, which would not necessarily require Greek women in the capacity Graham suggests. The literary evidence relating to Massalia thus leaves us with a dichotomy. There is evidence in favour of intermarriage (Justin and Aristotle), and evidence revealing that women, or rather, at least one woman, was transported with the colonists to the new settlement (Strabo). Dominguez suggests that the stories contained in Justin and Aristotle reflect the initial arrival of the Phocaeans in Massalia, whereas Strabo’s account relates to the later arrival of more Greeks once the new settlement had become more established (likely Phocaeans, following the 84 84 Persian conquest of their city; compare to Herodotus 1.164).323 On the other hand, however, Malkin states that Strabo’s use of the genitive absolute, τῆς ἀποικίας λαβούσης τέλος, is more indicative of the original foundation. 323 Dominguez (1999) 77. Hodos (1999) 66 has suggested that priestesses joining protocolonists (that is, the second wave of colonists following the establishment of the settlement) was more common than just this instance. 324 M lki (1987) 70 324 Malkin (1987) 70. 325 Rouge (1970) 312; Shepherd (1999) 270. See also Van Compernolle (1983) 1040. The connection between Massalia and the cult of Artemis of Ephesos is noted again by Strabo (4.1.4). 326 Shefton (1994) 61. 327 l ( ) ( ) 328 Tsetskhladze (2006c) lxv. ( ) 327 Morel (2006) 366. 326 Shefton (1994) 61. 324 Malkin (1987) 70. p p 332 Heinen (2001) 3. Massalia 324 It has also been argued that the Aristarcha of this tale was a later invention to account for the presence of the Temple of Artemis of Ephesis at Massalia, and the consequent transference of cult.325 It is perhaps most significant that she was to serve a specific role in the new colony, as a priestess. Many Greek priestesses were virgins, and as such, we could conjecture that Aristarcha was to remain a virgin, and cannot consequently be considered as a ‘Greek wife’ or a woman who could later become a wife, relevant to the question at the heart of this thesis. The archaeological exploration of Massalia has been hampered by modern settlement. However, Shefton reports that archaeological investigation has demonstrated that prior to the Greeks’ arrival, the site was uninhabited (though there is evidence of an earlier Etruscan presence in the area dating to around the mid-seventh century BC).326 Archaeologically, it also appears that the Greek colonists and the indigenous population were fairly integrated. Morel reports that local or indigenous pottery accounts for about 30% of the pottery thus far uncovered.327 This mix, according to Tsetskhladze, is due to the small chora of Massalia which encouraged close relations to ensure prosperity and survival of the colony.328 For Massalia, then, the main evidence is literary and this evidence suggests intermarriage, but that at least one woman, a priestess, accompanied the colonists. The weight of this evidence, however, reinforced by archaeological finds, leans in favour of intermarriage. 85 331 Graham (1961) 194. Scholars are beginning to make more of an attempt to combine their western and eastern efforts; this is particularly visible in the Colloquia Pontica series used in this thesis: Tuplin (ed) (2003); Tsetskhladze (ed) (2001); Solovyov (1999); and the work conducted by ( ) 330 Tsetskhladze (1994) 112. The eastern scholars had no, or very limited, access to western scholarship until the dissolution of the Soviet Union in 1991. 331 p ( ) ( ); ( ) ( ); y ( ); y the Danish National Research Foundation’s Centre for Black Sea Studies: see http://www.pontos.dk/. 332 ( ) 333 Petropoulos (2005) 6. 329 Heinen (2001) 3. 329 Heinen (2001) 3. 330 Black Sea Greek colonisation of the Black Sea region was highly prolific and its inclusion in an examination of archaic Greek colonisation should be imperative. Yet, the Black Sea is also one of the most problematic areas of the Greek world to study. Until very recently geographical, social, and political factors have prevented extensive study of the Black Sea. The Soviet Union closely controlled its territory which precluded many opportunities for excavation particularly by western scholars.329 Some limited excavation was carried out by eastern scholars, though this was performed without the ‘classics lens’ that one would usually expect of excavations of classical sites.330 In addition, there is little analysis of the archaeological evidence which makes interpretation challenging. Furthermore, there was (and to a certain extent, still is) a very limited crossover of scholarship between eastern and western scholars, and any crossover is made more challenging by language barriers.331 Many western scholars do not read Russian, Ukranian, or Georgian (for example), and so there are few who are able to move seamlessly between the eastern and western scholarship. Language differences and an apparently different (cultural) mentality make the countries surrounding the Black Sea seem even further away for western scholars. They have been described by Heinen as ‘a different world’, leading to a tendency for many western scholars to ignore the Black Sea region altogether.332 Others tend to juxtapose it against the colonisation which took place in the west, as if it were a completely different process.333 To complicate matters further, there are also archaeological difficulties in studying the Black Sea area, even now that it has been opened up for study. On the Bulgarian and Romanian coasts, all the Greek cities (with the exception of Histria) 86 lie under modern cities making extensive excavation near impossible.334 On the Colchian (modern-day Georgian) coast, the Greek cities are yet to even be located archaeologically. p ( ) 336 Shilik (1997) 115: “the global eustatic transgression resulted in a considerable rise in sea-level during Late Quaternary period by no less than 110-120 m over the past 20 ka, from the Last Glacial Maximum to the present day. Minor oscillations are distinguishable against the background of this general trend, their frequency being 1500/2000 years and their amplitude exceeding 10 m.” Such oscillations are clearly visible in the Black Sea. Tsetskhaladze (1998b) 19. 338 Some limited underwater exploration has been conducted; for example Lapin in 1961 led a team from Moscow in underwater excavation of Berezan. However, Nazarov (1997) 133 reveals that it was limited by insufficient equipment and experience, coupled with poor visibility, leaving the team to primarily establish sea depth and sediment thickness. Some further exploration has been conducted since but the results of this are limited to those who have an understanding of Cyrillic languages. Black Sea In addition, many of the excavations that have been completed have never been published (for example, many of the results from Berezan).335 The area covered by the Black Sea itself has also fluctuated considerably over time.336 For example, at Olbia on the northern coast, the modern coastline is 400- 500 metres back from the ancient one, meaning parts of the ancient settlement are underwater.337 Multiple other settlements are also now submerged and require underwater exploration and excavation.338 This makes archaeological exploration more difficult, especially in instances where settlements (or parts of settlements) have fallen away into the water, effectively destroying the opportunity to study them in situ.339 The reasons for the archaic colonisation of the Black Sea area are not immediately obvious. Unlike the majority of other colonies discussed in this thesis, there are virtually no literary sources for the Black Sea that might proffer evidence to better assess this issue. Herodotus has a limited focus in his description of Scythia, aiming to familiarise his audience with the physical area, rather than the Greek colonies or indigenous peoples (see, for example, 4.85-86 on the physical dimensions of the Black Sea, Bosporus, and Hellespont).340 Other authors, too, make passing references, such as Polybius’ (4.56) description of the city Sinope on 340 Petropoulos (2005) 1. Herodotus’ discussion on the Black Sea has been questioned, in terms of whether he claimed he went there, and whether he did in reality. Armayor (1978) 62 argues that “if Herodotus went to the Black Sea at all, his narrative bears little or no relation to whatever his travels may have been…Herodotus drew heavily on previous Greek traditions of the north when he came to build these claims”. See also West (2007). 340 Petropoulos (2005) 1. Herodotus’ discussion on the Black Sea has been questioned, in terms of whether he claimed he went there, and whether he did in reality. Armayor (1978) 62 argues that “if Herodotus went to the Black Sea at all, his narrative bears little or no relation to whatever his travels may have been…Herodotus drew heavily on previous Greek traditions of the north when he came to build these claims”. See also West (2007). 340 Petropoulos (2005) 1. Herodotus’ discussion on the Black Sea has been questioned, in terms of whether he claimed he went there, and whether he did in reality. 339 For example, at Berezan, some archaeological deposits are visible in the sides of the sea cliff and other finds have been made by local fisherman off the coast. See Nazarov (1997). ( ) 335 Petropoulos (2005) 3. 336 337 Tsetskhaladze (1998b) 19. 338 y 337 Tsetskhaladze (1998b) 19. 338 334 Tsetskhladze (1998b) 18. 334 Tsetskhladze (1998b) 18. 335 Petropoulos (2005) 3. 336 Shilik (1997) 115: “the global eustatic transgression resulted in a considerable rise in sea-level during Late Quaternary period by no less than 110-120 m over the past 20 ka, from the Last Glacial Maximum to the present day. Minor oscillations are distinguishable against the background of this general trend, their frequency being 1500/2000 years and their amplitude exceeding 10 m.” Such oscillations are clearly visible in the Black Sea. 337 Tsetskhaladze (1998b) 19. 338 Some limited underwater exploration has been conducted; for example Lapin in 1961 led a team from Moscow in underwater excavation of Berezan. However, Nazarov (1997) 133 reveals that it was limited by insufficient equipment and experience, coupled with poor visibility, leaving the team to primarily establish sea depth and sediment thickness. Some further exploration has been conducted since but the results of this are limited to those who have an understanding of Cyrillic languages. 339 For example, at Berezan, some archaeological deposits are visible in the sides of the sea cliff, and other finds have been made by local fisherman off the coast. See Nazarov (1997). 340 Petropoulos (2005) 1. Herodotus’ discussion on the Black Sea has been questioned, in terms of whether he claimed he went there, and whether he did in reality. Armayor (1978) 62 argues that “if Herodotus went to the Black Sea at all, his narrative bears little or no relation to whatever his travels may have been…Herodotus drew heavily on previous Greek traditions of the north when he came to build these claims”. See also West (2007). 340 Petropoulos (2005) 1. Herodotus’ discussion on the Black Sea has been questioned, in terms of whether he claimed he went there, and whether he did in reality. Armayor (1978) 62 argues that “if Herodotus went to the Black Sea at all, his narrative bears little or no relation to whatever his travels may have been…Herodotus drew heavily on previous Greek traditions of the north when he came to build these claims”. See also West (2007). 334 Tsetskhladze (1998b) 18. 335 Petropoulos (2005) 3. ( ) 345 Petropoulos (2005) 58. Petropoulos suggests that there may have only been six or seven permanent settlements throughout the entire Black Sea until c. 580 BC. 346 Hi d (1998) 131 133 Roebuck (1959) 116 130. 343 Tsetskhladze & Treister (1995); Tsetskhladze (1995); Tsetskhladze (2006c) xxix. 344 Pashkevich (2001) ( ) 342 Roebuck (1959) 116-130. ( ) 347 Petropoulos (2005) 26. 346 Hind (1998) 131-133. 347 Petropoulos (2005) 26. 341 Roebuck (1959) 87-103. p 346 Hind (1998) 131-133. 346 Hind (1998) 131-133. 347 Petropoulos (2005) 26 43 Tsetskhladze & Treister (1995); Tsetskhladze (1995); Tsetskhladze (2006c) xxix. 44 Pashkevich (2001). 45 342 Roebuck (1959) 116-130. Roebuck (1959) 87-103. 42 Roebuck (1959) 116-130. Black Sea Petropoulos notes that modern Scythian specialists argue against the existence of a permanent (settled) indigenous population along the northern coast of the Black Sea.347 Herodotus (4.127.1-2) addresses this issue, 88 citing a message from Idanthyrsus, the Scythian king, given to the Persian king Darius: πρὸς ταῦτα ὁ Σκυθέων βασιλεὺς Ἰδάνθυρσος λέγει τάδε. ‘οὕτω τὸ ἐµὸν ἔχει, ὦ Πέρσα. ἐγὼ οὐδένα κω ἀνθρώπων δείσας ἔφυγον οὔτε πρότερον οὔτε νῦν σὲ φεύγω, οὐδέ τι νεώτερον εἰµὶ ποιήσας νῦν ἢ καὶ ἐν εἰρήνη ἐώθεα ποιέειν. ὅ τι δὲ οὐκ αὐτίκα µάχοµαι τοι, ἐγὼ καὶ τοῦτο σηµανέω. ἡµῖν οὔτε ἄστεα οὔτε γῆ πεφυτευµένη ἐστί, τῶν πέρι δείσαντες µὴ ἁλῷ, ἢ καρῇ ταχύτερον ἂν ὑµῖν συµµίσγοιµεν ἐς µάχην… To this the Scythian king, Idanthyrsus replied: ‘Persian, this is my practice. I have never yet fled from any man in fear, not before, nor do I flee from you now. I am not doing anything now other than I am accustomed to. As to why I don’t fight you forthwith, I will tell you. We have neither towns, nor cultivated land, [so that] fear of losing it or [seeing it] ravaged, might bring us together to hasty battle.’ Herodotus notes that the Scythians (and indeed the Thracians and Getae too) were essentially a nomadic people; seemingly even more so than those of Libya.348 They did not appear to form any sort of homogenous ethnic group.349 Herodotus (4.81.1) himself seems confused about their numbers, and even about who counted as a Scythian and who did not: Herodotus notes that the Scythians (and indeed the Thracians and Getae too) were essentially a nomadic people; seemingly even more so than those of Libya.348 They did not appear to form any sort of homogenous ethnic group.349 Herodotus (4.81.1) himself seems confused about their numbers, and even about who counted as a Scythian and who did not: πλῆθος δὲ τὸ Σκυθέων οὐκ οἷος τε ἐγενόµην ἀτρεκέως πυθέσθαι, ἀλλὰ διαφόρους λόγους περὶ τοῦ ἀριθµοῦ ἤκουον: καὶ γὰρ κάρτα πολλοὺς εἶναι σφέας καὶ ὀλίγους ὡς Σκύθας εἶναι. I was not able to learn exactly how numerous the Scythians are, but I heard different accounts about their numbers: for [some say] they are very numerous, and [some say] they are few, as Scythians. The interpretation of the Scythians as nomads is not straightforward. 348 See also Herodotus (4.2.2) where in reference to the Scythian treatment of their slaves and prisoners of war, he comments that: οὐ γὰρ ἀρόται εἰσὶ ἀλλὰ νοµάδες. 349 p , γ ρ ρ 349 See Herodotus (4.17-27; 99-117) for some distinctions. 350 p , γ ρ ρ µ ς 349 See Herodotus (4.17-27; 99-117) for some distinctions. 350 P l (2005) 124 Black Sea Armayor (1978) 62 argues that “if Herodotus went to the Black Sea at all, his narrative bears little or no relation to whatever his travels may have been…Herodotus drew heavily on previous Greek traditions of the north when he came to build these claims”. See also West (2007). 87 the southern coast. Unfortunately, none of these various pieces of literary evidence look at Greek colonies or colonisation directly, or even indirectly. Roebuck argued in 1959 that the main reason behind Greek colonisation in the Black Sea area was an interest in metal particularly in the south and east,341 and an interest in the grain of the north.342 This argument was held to be true for many years, until a study by Tsetskhladze and Treister in 1995 established that the area surrounding the Black Sea was not in fact rich in metals, and crucially, that the Milesians (who instigated several colonies in the region) had access to various natural resources much closer to home.343 The presumed interest in grain has also been discounted. Pashkevich demonstrated that the Greek colonists did not plant crops that were indigenous to the Black Sea but instead brought familiar crops with them from the mainland.344 Furthermore, it is unlikely that any large exports of grain were made until a much later period, as the total population of the various Greek colonies must have been very small.345 Therefore, it seems unlikely that the push to colonise the Black Sea region was due to the desire to exploit resources. We know that the situation in the homeland was often the impetus for colonisation. Megara, for example, suffered the loss of some land to their dominating neighbours, the Corinthians and Athenians, in the seventh and sixth centuries BC.346 It is possible that the main driving force for the Megarians to colonise was to satisfy their land hunger and replace lost land, most notably by settling at Heraclea Pontica. Without more detailed literary evidence, however, it is near impossible to become more certain of the reasons behind the colonisation of the Black Sea. It is also difficult to assess the relationship between the Greek colonists and the indigenous populations. 350 Petropoulos (2005) 124. 348 See also Herodotus (4.2.2) where in reference to the Scythian treatment of their slaves and ( ; 350 Petropoulos (2005) 124. 351 Excavations were begun at the end of the nineteenth century; however, these have been halted a number of times for political reasons, and many of the results have never been published (see in particular, Solovyov (1999) 19-27; also Petropoulos (2005) 3; Solovyov (2001) 118). Most recently, see Petersen (2010) 41-120 for a succinct summary of all the excavation of Olbia, particularly of the cemetery, though much of this material is too late for the purposes of this thesis. Petersen (2010) 115 also points out that the majority of Berezan’s excavation reports remain unpublished, and now have been separated from the actual burial finds. Black Sea Petropoulos suggests it is likely that the Greeks brought wives with them to this area, given that the contact between the settling Greek colonists and the roaming nomads was presumably minimal.350 This is an attractive proposal, although we cannot be sure of the degree of contact one way or another. 89 One of the best excavated areas in the Black Sea region includes Olbia and modern day Berezan, both located in the north.351 Berezan is thought to be the Milesian colony, Borysthenes, and is often described as the oldest colony in the northern Black Sea.352 Both settlements are notable for a number of ‘dugout’ buildings (Figure 11). These were each 100-262 square metres and appear to have contained a number of domestic areas (both work and living) all centred around an interior courtyard.353 Previously these have been interpreted as temporary living structures for the new colonists.354 However, Solovyov has postulated that it is unlikely that every single one of these structures was used by the colonists, given their relatively great number; instead, he believes that only two or three were occupied by the Greeks.355 This is hard to reconcile with the notion conveyed in the literary evidence that the indigenous peoples of this area were nomads.356 However, it is possible that the area and its inhabitants were sufficiently unknown to the Greeks (at least in so far as they convey to us in the extant literary evidence, for example Herodotus 4.81.1 discussed above), and plausible that the indigenous population 353 Solovyov (2001) 129. These buildings were probably single storey, but as Solovyov points out, it is not possible to exclude the presence of a second storey. 354 353 Solovyov (2001) 129. These buildings were probably single storey, but as Solovyov points out, it is not possible to exclude the presence of a second storey. 354 354 See especially Kuznetsov’s (1999) chapter which asseses this issue; also Tsetskhladze (1998b) 20, 44; Kryzhitskii (2007) 19; 355 354 See especially Kuznetsov’s (1999) chapter which asseses this issue; also Tsetskhladze (1998b) 20, 44; Kryzhitskii (2007) 19; 355 355 Solovyov (2001) 126-128; 131. This view is also formed by later changes taking place in the third quarter of the sixth century BC when many of the dugouts were filled in, having been recently surrendered by their occupants. This occurred concurrently with other changes to the cultural face of Berezan. 352 Solovyov (2001) 117. There is, however, some debate as to the ancient name of this settlement. Solovyov (1998) 205, for example, uses the name Borysthenes interchangeably for both Berezan and Olbia. Berezan is now an island, but it is thought that it was a peninsula in antiquity (see Boardman (1998) 201; Kryzhitskii (1997) 101; Hind (1983-1984) 79), with Olbia opposite. There is also confusion over the order in which Berezan and Olbia were settled; that is, whether Berezan was settled first, and then colonists moved to Olbia, or whether the two settlements were established concurrently. The archaeological evidence does not particularly help solve this issue. Boardman (1998) 202 has suggested that the foundation dates were close which would explain why the two cannot be distinguished archaeologically. Graham (1982) 125 argues that Olbia and Berezan were settled concurrently, however his (1994/2001) publication acknowledges that his argument has been universally rejected by Russian scholars. 351 Excavations were begun at the end of the nineteenth century; however, these have been halted a number of times for political reasons, and many of the results have never been published (see in particular, Solovyov (1999) 19-27; also Petropoulos (2005) 3; Solovyov (2001) 118). Most recently, see Petersen (2010) 41-120 for a succinct summary of all the excavation of Olbia, particularly of the cemetery, though much of this material is too late for the purposes of this thesis. Petersen (2010) 115 also points out that the majority of Berezan’s excavation reports remain unpublished, and now have been separated from the actual burial finds. 352 Solovyov (2001) 117. There is, however, some debate as to the ancient name of this settlement. Solovyov (1998) 205, for example, uses the name Borysthenes interchangeably for both Berezan and Olbia. Berezan is now an island, but it is thought that it was a peninsula in antiquity (see Boardman (1998) 201; Kryzhitskii (1997) 101; Hind (1983-1984) 79), with Olbia opposite. There is also confusion over the order in which Berezan and Olbia were settled; that is, whether Berezan was settled first, and then colonists moved to Olbia, or whether the two settlements were established concurrently. The archaeological evidence does not particularly help solve this issue. Boardman (1998) 202 has suggested that the foundation dates were close which would explain why the two cannot be distinguished archaeologically. Graham (1982) 125 argues that Olbia and Berezan were settled concurrently, however his (1994/2001) publication acknowledges that his argument has been universally rejected by Russian scholars. Black Sea This ‘urbanisation’, Solovyov believes, can only have taken place as a result of mass immigration and the organisation of the organs of government. Kryzhitskii (2006) 108 states that these dugouts cannot be used as indicators of ethnicity given their status as “structures of a transient character”. 355 Solovyov (2001) 126-128; 131. This view is also formed by later changes taking place in the third quarter of the sixth century BC when many of the dugouts were filled in, having been recently surrendered by their occupants. This occurred concurrently with other changes to the cultural face of Berezan. This ‘urbanisation’, Solovyov believes, can only have taken place as a result of mass immigration and the organisation of the organs of government. Kryzhitskii (2006) 108 states that these dugouts cannot be used as indicators of ethnicity given their status as “structures of a transient character”. 356 Buyskikh (2007) 24-26 considers Solovyov’s above revision of the scholarship to be “absurd”, and argues that Solovyov used an unreliable methodology and ignored the broader picture where such dugouts had been found in other early Greek settlements in the wider region. 356 Buyskikh (2007) 24-26 considers Solovyov’s above revision of the scholarship to be “absurd”, and argues that Solovyov used an unreliable methodology and ignored the broader picture where such dugouts had been found in other early Greek settlements in the wider region. 90 was not a homogenous group, so that some were nomads and others were not. If we are to agree with Solovyov’s interpretation of the dugouts, it seems that there was certainly some interaction between the indigenous peoples living in the dugouts and the Greek colonists.357 Furthermore, Petropoulos reports that some indigenous handmade pottery, dating from the second half of the sixth century BC, was found adjacent to Greek pottery. This could suggest that indigenous Scythians were living alongside the Greek colonists. The interpretation of the dugouts, in particular, can only be tentative, as like the fibulae, they are poor indicators of ethnicity. 357 Buyskikh (2007) 27 admits that the potential coexistence of Greeks and non-Greeks is still a controversial view that eastern scholars are reluctant to subscribe to. 358 P l (2005) 27 p ( ) 359 Petropoulos (2005) 28. 358 Petropoulos (2005) 27. 358 Petropoulos (2005) 27. 359 Petropoulos (2005) 28. Black Sea Petropoulos postulates that given the relatively early period of contact, it is possible to assume that the relations between the Greek colonists and the various indigenous peoples were friendly, or at least not actively hostile.358 The traces of fire found at some Greek settlements in the Black Sea, dating to the beginning of the fifth century BC, are thought to indicate the breakdown of relations between the indigenous population and the Greek settlers.359 Lack of literary evidence and limited access to eastern scholarship relating to archaeological discoveries mean that it is very difficult to reach an informed view about the nature of the relationsip between the Greek colonists and the indigenous populations in the Black Sea, and so even harder to assess the likelihood that intermarriage would have occurred. Clearly, the colonies of this region present a rich resource for further study, not only of the relationship between Greek settlers and the indigenous peoples, but of the region more generally as an important centre of the ancient world in its own right. 91 362 Domanskij & Marcenko (2003). ( ) 361 See footnote 138 above. 362 360 Graham (1971) 36. 360 Graham (1971) 36. 362 Domanskij & Marcenko (2003). CONCLUSION The picture of Greek colonisation that we can achieve will, therefore, always be drawn in rather broad lines, and the task of the historian is continually to try to improve the quality and validity of these rather general constructions.360 – A. J. Graham This statement, made by Graham in 1971, demonstrates strikingly clear perception of the issues for the time it was made. It remains a fitting explanation of the struggles surrounding the study of Greek colonisation and an accurate description of the role that scholars must play in order to further scholarship and understanding. Consequently, in this conclusion I will focus on drawing together some common threads from my consideration of the relationship between Greeks and indigenous peoples in general, and in relation to the individual colonies discussed in the previous chapter. It is crucial to note, however, that each colony must continue to be assessed on its own merits; for the differences between each colony, each region, each mother city, and so on, are endless, ruling out any idea that the ‘Greek colonies’ can be considered a coherent and uniform phenomenon. The tendency for new colonists to first settle on an island or another easily defensible place before relocating to their final settlement was common among the colonies discussed in this survey. This was observed definitely at Syracuse with Ortygia as the preliminary settlement; at Cyrene with the island Platea prior; and at Taras with Satyrion. Arguments have also been made that Pithekoussai was an island settlement before colonists moved to the Italian mainland to settle at Cumae.361 The situation with Borysthenes (or the island Berezan) and Olbia in the Black Sea is less clear, although some scholars have argued in favour of this process there too.362 We may only speculate about the reasons for these actions. It seems likely that these colonists first chose to settle on an island (with the exception of Satyrion/Taras) because islands are inherently easily defensible places. Logically, 92 92 it seems likely that this was a good opportunity to assess the surrounding territory, and locate and assess an appropriate place for permanent settlement. Importantly, it could also provide a suitable means to establish relations with the indigenous population inhabiting the area. Perhaps it also enabled a first occasion to intermarry. This may have involved forcing the indigenous population from its territory, or establishing some sort of friendly relationship conducive to coexistence. 363 Diodorus Siculus (20.41.1), writing about the Hellenistic period, associates the presence of women and children in an army as being like a colonising expedition, and their absence as a sign of a military campaign. This is not necessarily good evidence for the colonies of the archaic period. CONCLUSION The process often appears obscure. In the case of Ortygia/Syracuse, the indigenous population was driven out of its lands (Thucydides 6.3.2), whereas at Platea/Cyrene, the indigenous population helped the Theraean colonists find the location of their final settlement (Herodotus 4.158). Both colonies began life on an island but their subsequent histories differed dramatically. Accordingly, it is very difficult to reach any general conclusion about any impact that settling on an island prior to the mainland could have had on the likelihood of intermarriage taking place. Instead, the process of initially settling an easily defensible place before establishing the colony proper seems to speak more to the type of venture that the colonists were undertaking. Often such island settlements provided reconnaisance. The Greeks made their first moves toward settlement understanding that some sort of military action might be necessary. If so, can we assume that women would be less likely to accompany these expeditions and therefore that Greek wives (or more generally, Greek women) did not accompany the men on the journey?363 For example, at Syracuse (although the evidence is not entirely clear), it appears that the relationship between the Corinthian colonists and the indigenous population was considerably more hostile leading to the expulsion of the indigenous population and/or their subsequent subjugation by the newcomers. Obviously there are many limitations to this hypothesis. Is it really possible that the colonists knew whether their venture would involve military action prior to leaving their mother city? This would require a great deal of knowledge about their new destination. Some prior knowledge was not an unreasonable expectation, particularly with the colonies that appear to have been founded with the intention of utilising natural resources nearby or those established for trade purposes. On 93 the other hand, however, some colonies such as Cyrene appear to have been founded with little or no prior knowledge of the new lands, or even of where exactly they were setting sail for. Some colonies were settled on so-called ‘virgin land’ while others were established where there were already existing indigenous settlements. The precise reasons behind this difference may never be known (or indeed if there was a specific reason beyond convenience or exigency), but it is interesting to reflect on what impact these various actions may or may not have had on the indigenous populations and how they affected the Greek-indigenous relationship, and accordingly the likelihood of intermarriage. CONCLUSION When colonies were founded on virgin land, it is important to identify whether or not that land was given to the Greeks by the indigenous population. Pithekoussai is relatively obscure on this point. Megara Hyblaea, Massalia, and Cyrene, however, were each definitely settled on virgin land. In the case of Megara Hyblaea there is debate over exactly how the verb προδίδωµι used by Thucydides (6.4) to describe the Sicel king Hyblon’s action should be translated. The various options have already been extensively discussed, but I interpret Hyblon’s actions as completely willing on his part though with the acknowledgement that there may have been ulterior motives behind the gift alongside sheer generosity. Furthermore, it seems that the Greek colonists and the indigenous Sicels had a good relationship that probably involved intermarriage. Massalia, too, has fairly convincing literary evidence (Athenaeus 13.576a-b; Justin 43.3.5-11) which reports the marriage between the Greek oikistes and the indigenous king’s daughter, and states that land was also given as a part of this arrangement. Finally, at Cyrene (the colony with the most compelling evidence in favour of intermarriage) the Theraean colonists were physically led to the site of their settlement by Libyan guides (Herodotus 4.158). Accordingly, it appears that in many cases where the Greek colonies were settled on virgin land, it was given to them by the indigenous population (or their leader). Furthermore, the gift of land to the Greeks seems to have encouraged a good relationship between the colonists and indigenous peoples and consequently led to intermarriage between the two. But can the opposite scenario also be said to be true? 94 Morgantina, Metapontum, Syracuse, Taras, and Locri were all colonies founded on land already occupied by another population. Morgantina and Metapontum stand apart from the others due to the fact they do not seem to have ever had a ‘formal foundation’. That is, while there is a definite record of the continued occupation of their respective sites, it is not clear when the various Greek groups first arrived. By contrast, the arrival of colonists in Syracuse, Taras, and Locri is explicit in the literary record. More importantly, at each of these colonies, the newly arriving colonists are said to have driven away or ejected the respective indigenous populations who had been living on the land. 364 Graham (1981-82) strongly follows this line of argument; that the Greek colonists must have taken Greek women with them otherwise there would have been cult roles left unfulfilled. Although a role for women in religion should not be discounted, I do not accept the argument as it stands (or, in its entirety), especially as the colonies were seen to have taken on characteristics (including religious ones) of their new environment. 364 Graham (1981-82) strongly follows this line of argument; that the Greek colonists must have taken Greek women with them otherwise there would have been cult roles left unfulfilled. Although a role for women in religion should not be discounted, I do not accept the argument as it stands (or, in its entirety), especially as the colonies were seen to have taken on characteristics (including religious ones) of their new environment. stands (or, in its entirety), especially as the colonies were seen to have taken on characteristics (including religious ones) of their new environment. taken Greek women with them otherwise there would have been cult roles left unfulfilled. Although a role for women in religion should not be discounted, I do not accept the argument as it t d ( i it ti t ) i ll th l i t h t k h t i ti 364 Graham (1981-82) strongly follows this line of argument; that the Greek colonists must have taken Greek women with them otherwise there would have been cult roles left unfulfilled. CONCLUSION On the surface, at Taras and Locri in particular, there is some evidence pointing to the presence of Greek women (or in the case of Taras, a woman) at the foundation of the colony. This conflicts with the assumption previously raised that women were less likely to accompany men on ventures which involved military action. On the other hand, however, the evidence for both Taras and Locri seems to have its tradition rooted in myth and certainly in the case of Locri (Polybius 12.5.3-11) the tradition may have served a later political purpose. Furthermore, Taras is linked to the presence of only one woman at the foundation (Pausanias 10.5-8) – a woman of exceptional status being the oikistes’ wife. Consequently, these examples can neither be used to argue in favour of many Greek women being present, nor to definitely rule out the possibility of intermarriage. The passage of time has obscured much about the specifics of each colonial venture. With the exception of Cyrene, we are left to piece together small fragments of various traditions to try and develop a fuller picture. We lack names in the record other than the occasional reference to an oikistes. Curiously, of the few names that we do have, only two of these belong to women and both of these were priestesses – Aristarcha of Massalia recorded in Pausanias 10.28.3, discussed extensively in the previous chapter; and Kleoboia of Thasos recorded in Strabo 4.1.4. Above all, this points to an argument about the importance of women in religion, and more specifically, of Greek women in Greek religion.364 95 Many of the foundation tales which we use as evidence are almost certainly invented at a later date (that is, long after the colony’s foundation), invented to give the ‘mixed bag’ of peoples a common and precise origin.365 The precise reason for creating the tales is seldom obvious but we must assume that such stories always had the potential to hold an ulterior motive or to have undergone transformation as they were passed down the generations. The literary evidence is thus perhaps more disappointing and limited than we might have initially imagined. For this reason, archaeological evidence has acquired new importance in the second half of the twentieth century. ( ) 366 Desborough (1976) 27-28. Within the oikos, Nevett (1994) 99 demonstrates that evidence of gender separation is very limited. Nevett (1999) 127-153 argues that the Greek house in Sicily was considerably less structured than those in the mainland, with most rooms being multifunctional, and as a result, there is little to differentiate archaeologically the separation between the male and female parts of the household. 367 365 Malkin (2002b) 196. 366 p 367 Buchner (1979) 135. CONCLUSION Continued excavation, and the application of modern analytical techniques will be important to strengthen our knowledge of the colonies, in conjuction with the various pieces of literary evidence. Much more study also needs to be undertaken of indigenous populations and their practices, so that we might better compare them with the Greeks in the colonies. The pitfalls of a heavy reliance on archaeology, however, were revealed in Chapter Three. Excavations tend to largely centre on the necropoleis of settlements, due to the fact that domestic environments tend to be less well preserved and distinguishing female from male is even less clear.366 Within the necropoleis, the fibulae associated in particular with the Italic colonies of Pithekoussai, Syracuse, and Megara Hyblaea demonstrated the dangers associated with trying to use archaeology as an indicator of identity or ethnicity. Buchner’s well-known and well-regarded hypothesis argued that the presence of Italic fibulae as opposed to the typical Greek straight pins indicated that the wearers were no longer wearing the traditional Greek Doric peplos and therefore that they were not Greek women.367 This argument is questionable even for the colony for which it was initially proposed, Pithekoussai, but becomes even more problematic in relation to the fibulae of other colonies. At both Syracuse and Megara Hyblaea fibulae do feature, but they appear less frequently than the traditional straight pins. Furthermore, they seem to have had less of a practical function than their straight 96 counterparts which were often found in pairs in graves on the shoulders of the deceased indicating that they were buried wearing the pins. Buchner’s hypothesis would lead to the conclusion that the women at Syracuse and Megara Hyblaea were in fact Greek, rather than indigenous. While the evidence at Syracuse is inconclusive, at Megara Hyblaea there is fairly convincing evidence that the relations between the colonists and the indigenous Sicels were friendly, and as such that intermarriage more than likely occurred, which is difficult to reconcile with Buchner’s hypothesis. The serpentine fibulae associated with males are also difficult to fit within Buchner’s theory. No one appears to have suggested that these could have belonged to indigenous males, but how else are we to interpret them if we follow Buchner’s hypothesis to its logical conclusion? The existence of serpentine fibulae emphasises how important it is not to use fibulae to determine ethnicity or identity. 368 Hodos (1999) 74: “intermarriage can be dismissed in favour of trade as an explanation of how the Greeks in Sicily obtained their fibulae”. I do not wish to argue that trade and intermarriage mutually exclude each other, but I suggest that trade was certainly a factor in the presence of these fibulae being found at colonial sites. 368 Hodos (1999) 74: “intermarriage can be dismissed in favour of trade as an explanation of how CONCLUSION The presence of fibulae can be no more an argument in favour of intermarriage than the occurrence of straight pins is against it. Other than the evidence pointing to the manufacture of fibulae at Pithekoussai, there is little else to suggest that the fibulae could not have been sourced from elsewhere through trade.368 This appears to be more likely until we discover other metal depositories in Sicily and evidence of metal-working. This is not to say, however, that we should rule out the possibility that both Greeks and indigenous people could have lived in the same settlements concurrently, accounting for the presence of both fibulae and straight pins. The unsuccessful attempt to use archaeology as an indicator of ethnicity is also apparent at Morgantina and Metapontum. At both these sites a plethora of Greek pottery has been uncovered along with considerable amounts of indigenous ware. On the one hand, this mixture of pottery could be an indicator of the indigenous peoples and Greeks living side by side. On the other hand, however, the presence of Greek pottery is not necessarily indicative of the presence of Greek people as such objects can be easily acquired in great numbers through trade. Without the 97 use of DNA testing, archaeology is therefore a poor indicator of ethnicity and identity. While Osborne’s (1998) call for the eradication of colonisation from history books seems drastic, his underlying point is cogent and ought to be taken seriously. Colonisation cannot continue to be considered a uniform phenomenon and neither can the settlement of archaic colonies be evaluated in a similar way to eighteenth and nineteenth century colonisation. In the same way that each colony has different foundation dates, different mother cities, different foundation stories, and is settled in different locations for different reasons, so perhaps too the source of women for each colony also differed. Given these differences and uncertainties, broad conclusions about intermarriage as a widespread practice are as risky as broad conclusions about women and men travelling together in a joint colonial enterprise of archaic Greece. 98 APPENDIX 1: FIGURES Base map sourced: http://d-maps.com/ APPENDIX 1: FIGURES Figure 1: Map of major archaic Greek colonies discussed APPENDIX 1: FIGURES Figure 1: Map of major archaic Greek colonies discussed Base map sourced: http://d-maps.com/ Base map sourced: http://d-maps.com/ 99 Figure 2: Example of straight pin from Greece Inv. No: 86.009, The University of Queensland Collection, c. CONCLUSION seventh-sixth century BC, 10 mm diameter, 135 mm length. Figure 2: Example of straight pin from Greece Inv. No: 86.009, The University of Queensland Collection, c. seventh-sixth century BC, 10 mm diameter, 135 mm length. Figure 3: Italic fibulae types – Nos 1-3: navicella fibulae; no 4: leech fibula; no 5: serpentine fibula; no 6: animal fibula; nos 7-8: bone and amber fibulae (drawings by Orsi 1895, based on fibulae from Syracuse) Figure 3: Italic fibulae types – Nos 1-3: navicella fibulae; no 4: leech fibula; no 5: serpentine fibula; no 6: animal fibula; nos 7-8: bone and amber fibulae (drawings by Orsi 1895, based on fibulae from Syracuse) Image from: Shepherd, G. (1999) ‘Fibulae and Females’, in Tsetskhladze, G. R. (ed) Ancient Greeks West and East, Leiden: Brill, p. 279. Image from: Shepherd, G. (1999) ‘Fibulae and Females’, in Tsetskhladze, G. R. (ed) Ancient Greeks West and East, Leiden: Brill, p. 279. 100 Figure 4: Area of inhabitation at Pithekoussai RECENT WORK AT PITHEKOUSSAI (ISCHIA), 1965-71 ??C.P pST R.TRX SR g.STR. II ILI Ii L • . . I -- III I J STR. M PITHEKOUSSAI MEZZAVIA: RRE9 /MYRZZO1-, FIG 5 G. BUCHNER tion-perhaps by earthquake and falling s. ke Structure I, the other buildings are all con- in one way or another with metal-working. re III, modified more than once, was a black- workshop. The post-hole pattern shows that e western side was covered: at the other side open courtyard, the middle of which was burnt-which suggests that here was the site forge. The two successive floor-levels pro- many pieces of bloom and iron-slag, and the floor surfaces were impregnated with in- ble tiny iron fragments. cture IV originally had an oval plan in the period; it was later reconstructed, partly re- he same foundations, as a rectangular building. oo we seem to have some sort of forge rather to suppose that it is to be associated with the earl period-in fact to the last phase of the building function as a workshop, where precious metals mig well have been worked. It should be pointed out that, even though t Mazzola site seems to be a metal-working quart metal must also have been worked on the acropo as well. The dump on the east slope of Monte Vico produced a quantity of blooms and iron-sla and a few bellows-mouthpieces ('tuyeres': Dialoghi Archeologia iii: 1-2, 1969, fig. al that can be assigned to the pre-hellenic on Age (as found at Castiglione d'Ischia: is almos in which th Figure 5: Miscast fibula found in Structure IV umulated outside this building included snippets of bronze sheet and wire, a small joins betwee Considere on Age (as found at Castiglione d'Ischia: -37, 65 ff.). e second of the three terraces which make up a southerly retaining wall delimits the area towards the steep slope of the hill; to the haotic mass of boulders is probably the result uakes during the site's history. A second wall (north-west-south-east) separates this errace from the third and lowest. ce of building has been found on the upper- ace. During the 1969 and 1970 campaigns, mplete structural units were revealed on the rrace, and one on the lowest terrace. These ged to the earliest attested period at the site. still on the lowest terrace, we found a further of the eighth century, and part of another be dated to the first half of the sixth century of these are shown on the plan, Fig. 5). h-century building is made of more or less red blocks of green Epomeo tufa-a material nique quite unknown in the earlier period. in which was also embedded a whole roug pot: these homely circumstances suggest Figure 5: Miscast fibula found in Structure IV building s snippets of bronze sheet and wire, a small joins widely separated Considered as a whole, the pottery from Maz and from the dump on Monte di Vico is nota different from that found in the Valle San Mont cemetery. Certain types that occur regularly in cemetery are virtually absent outside it: and versa. Mazzola and Monte di Vico have produ a minuscule quantity of the small closed perfume v (including squat lekythoi) that are the hall-mark LG II in the cemetery: on the other hand, the 'domestic' sites abound in fragments of large pain and frequently figured, craters and ampho These in turn are comparatively rare in the cemet and in any case are so far confined to the crema tombs: the painted craters are smaller, and pain amphoras are used only very exceptionally for enchytrismos burials (normally deposited in amph and pithoi of coarse unpainted impasto). The image of the pottery in use at Pithekou during the second half of the eighth century has been changed out of all recognition. In particu the existence has been revealed of a whole new of painted Geometric pottery, frequently figured d t LG I d ith t d bt FIG. CONCLUSION 26)-as well as a pie of iron mineral in its natural state (pure hematit that can definitely be assigned to the Rio Mari deposit on the island of Elba (analysis by Profess G. Marinelli, University of Pisa). The pottery. Most of the vast quantity of potte found at Mazzola does not appear to be connecte Figure 4: Area of inhabitation at Pithekoussai RECENT WORK AT PITHEKOUS uperficial damage caused by planting vines The Greek settlement at Mazzola was by a Bronze Age Apennine Culture settle- in the dump on Monte di Vico, there is also the only building so far that shows subsequent transformation or reconstruc ments of a number of more or less com including the crater, Fig. 6 (the unillustra Image from: Buchner, G. (1970-71) ‘Recent Work at Pithekoussai (Ischia), 1965- 71’, Archaeological Reports 17, p. 65. belongs phase ructure, and near it were found in situ two polished pieces of hard phonolith that look ls. In addition to iron, other metals were here as well: especially bronze. The detritus bro from the nearby inhabited areas along with intended to raise the levels of the floors in the su sive periods of the work-shops. This accounts the otherwise disturbingly frequent phenomenon perficial damage caused by planting vines The Greek settlement at Mazzola was by a Bronze Age Apennine Culture settle- in the dump on Monte di Vico, there is also the only building so far that shows subsequent transformation or reconstruct ments of a number of more or less com including the crater, Fig. 6 (the unillustra Image from: Buchner, G. (1970-71) ‘Recent Work at Pithekoussai (Ischia), 1965- 71’, Archaeological Reports 17, p. 65. belongs phase ructure, and near it were found in situ two polished pieces of hard phonolith that look s. In addition to iron, other metals were ere as well: especially bronze The detritus bro from the nearby inhabited areas along with intended to raise the levels of the floors in the su sive periods of the work-shops. This accounts the otherwise disturbingly frequent phenomenon al that can be assigned to the pre-hellenic on Age (as found at Castiglione d'Ischia: is almos in whic th Figure 5: Miscast fibula found in Structure IV umulated outside this building included snippets of bronze sheet and wire, a small joins betwee Considere al that can be assigned to the pre-hellenic on Age (as found at Castiglione d'Ischia: is almos in which th Figure 5: Miscast fibula found in Structure IV umulated outside this building included snippets of bronze sheet and wire, a small joins betwee Considere 7 ngot, drops of greenish vitreous slag, pieces nd a miscast fibula, with casting seams, because the foot was too short (Fig. 7). i fi f id bl Image from: Buchner, G. (1970-71) ‘Recent Work at Pithekoussai (Ischia), 1965- 71’, Archaeological Reports 17, p. 66. be dated to t e first a o century of these are shown on the plan, Fig. 5). h-century building is made of more or less red blocks of green Epomeo tufa-a material i it unknown in the earlier i d image o t e pottery during the second half of the eighth century has been changed out of all recognition. In partic the existence has been revealed of a whole new of painted Geometric pottery frequently figured 7 ngot, drops of greenish vitreous slag, pieces nd a miscast fibula, with casting seams, because the foot was too short (Fig 7) Image from: Buchner, G. (1970-71) ‘Recent Work at Pithekoussai (Ischia), 1965- 71’, Archaeological Reports 17, p. 66. 101 ic of th h a dou bly wit Figure 6: Kupara graffito from Morgantina Figure 6: Kupara graffito from Morgantina TAFEL V -#?' ;if^m 'i* > ; < "" -' . " '. ^r??isr' >s8C >%%.; .j?r^''' Image from: Antonaccio, C. M. & Neils, J. (1995) ‘A New Graffito from Archaic Morgantina’, Zeitschrift für Papyrologie und Epigraphik 105, p. 277. Image from: Antonaccio, C. M. & Neils, J. (1995) ‘A New Graffito from Archaic Morgantina’, Zeitschrift für Papyrologie und Epigraphik 105, p. 277. 102 Figure 7: CPS 1 – Libyan bust Figure 7: CPS 1 – Libyan bust Figure 8: CPS 189, 190 – Cyrenaean busts showing a mixture of Greek and Libyan features Figure 8: CPS 189, 190 – Cyrenaean busts showing a mixture of Greek and Libyan features Libyan features Libyan features 103 Figure 9: CPS 191 – Cyrenaean bust showing a mixture of Greek and Libyan features Figure 9: CPS 191 – Cyrenaean bust showing a mixture of Greek and Libyan f t 104 Figure 10: Cyrenaean relief showing a mixture of Greek and Libyan aspects Image from: Wanis, S. (1992) ‘A New Relief from Cyrene with a Liby Libyan Studies 23, p. 42. Unpublished, 77cm high, 45cm wide, 22cm Image from: Wanis, S. (1992) ‘A New Relief from Cyrene with a Libyan Scene’, Libyan Studies 23, p. 42. Unpublished, 77cm high, 45cm wide, 22cm deep. Image from: Wanis, S. al that can be assigned to the pre-hellenic on Age (as found at Castiglione d'Ischia: is almos in which th Figure 5: Miscast fibula found in Structure IV umulated outside this building included snippets of bronze sheet and wire, a small joins betwee Considere (1992) ‘A New Relief from Cyrene with a Libyan Scene’, Libyan Studies 23, p. 42. Unpublished, 77cm high, 45cm wide, 22cm deep. 105 Figure 11: Examples of dugouts in the Black Sea Image from: Kuznetsov, V. D. (1999) ‘Early Types of Greek Dwelling Houses in the North Black Sea’ in Tsetskhladze, G. R. (ed) Ancient Greeks West and East, Leiden: Brill, p. 543. Figure 11: Examples of dugouts in the Black Sea Image from: Kuznetsov, V. D. (1999) ‘Early Types of Greek Dwelling Houses in the North Black Sea’ in Tsetskhladze, G. R. (ed) Ancient Greeks West and East, Leiden: Brill, p. 543. 106 BIBLIOGRAPHY Albanese-Procelli, R. M. (1996) ‘Greeks and Indigienous People in Eastern Sicily: Forms of Interaction and Acculturation’ in Leighton, R. (ed) Early Societies in Sicily: New Developments in Archaeological Research, London: Accordia Research Centre, 167-176. Angel, J. L. (1972) ‘Ecology and Population in the Eastern Mediterranean’, World Archaeology 4 (1), 88-105. Antonaccio, C. (1993) ‘Writing and Naming at Archaic Morgantina’, American Journal of Archaeology 97, 352. Antonaccio, C. (1997) ‘Urbanization at Archaic Morgantina’, Acta Hyperborea 7, 167-193. Antonaccio, C. (2001) ‘Ethnicity and Colonization’ in Malkin, I. (ed) Ancient Perceptions of Greek Ethnicity, Washington D.C.: Center for Hellenic Studies, 113-158. Antonaccio, C. (2003) ‘Hybridity and the Cultures within Greek Culture’ in Dougherty, C & Kurke, L (eds) The Cultures Within Ancient Greek Culture: Contact, Conflict, Collaboration, Cambridge: Cambridge University Press, 57-74. Antonaccio, C. (2005) ‘Excavating Colonization’ in Hurst, H & Owen, S (eds) Ancient Colonizations: Analogy, Similarity and Difference, London: Duckworth, 97-113. Antonaccio, C. (2007) ‘Colonization: Greece on the Move, 900-480’ in Shapiro, H. A. (ed) The Cambridge Companion to Archaic Greece, New York: Cambridge University Press, 201-224. Antonaccio, C. & Neils, J. (1995) ‘A New Graffito from Archaic Morgantina’, Zeitschrift für Papyrologie und Epigraphik 105, 261-277. Applebaum, S. (1979) Jews and Greeks in Ancient Cyrene, Leiden: Brill. Armayor, O. K. (1978) ‘Did Herodotus Ever Go to the Black Sea?’, Harvard Studies in Classical Philology 82, 45-62. Astour, M. C. (1985) ‘Ancient Greek Civilization in Southern Italy’, Journal of Aesthetic Education 19 (1), 23-37. Attema, P. (2008) ‘Conflict or Coexistence? Remarks on Indigenous Settlement and Greek Colonization in the Foothills and Hinterland of the Sibaritide (Northern Calabria, Italy)’ in Bilde, P. G. & Petersen, J. H. (eds) Meetings 107 of Cultures – Between Conflicts and Coexistence, Aarhus: Aarhus University Press, 67-100. Aubet, M. E. (1993) Phoenicians and the West: Politics, Colonies, and Trade, Cambridge: Cambridge University Press. Austin, M. (2008) ‘The Greeks in Libya’ in Tsetskladze, G. R. (ed) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Volume Two, Leiden: Brill, 187-218. Becker, M. J. (1986) ‘An Ethnographical and Archaeological Survey of Unusual Mortuary Procedures as a Reflection of Cultural Diversity: Some Suggestions for the Interpretation of the Human Skeletal Deposits from Excavations at Entella, Sicily, Italy’, La Parola del Pasato: Rivista di Studi Antichi 226, 31-56. Becker, M. J. BIBLIOGRAPHY (1992) ‘A Summary of the Analysis of Cremated Human Skeletal Remains from the Greek Colony of Pithekoussai at Lacco Ameno, Ischia, Italy’, Old World Archaeology Newsletter 16 (1), 15-23. Becker, M. J. (1995) ‘Human Skeletal Remains from the Pre-Colonial Greek Emporium of Pithekoussai on Ischia (NA): Culture Contact in Italy from the Early VIII to II Century B.C.’ in Christie, N. (ed) Settlement and Economy in Italy 1500 B.C. – A.D. 1500, Oxford: Oxbow Books, 273-282. Becker, M. J. (1996) ‘The Human Skeletons from the Archaic Cemeteries’ in Lyons, C. The Archaic Cemeteries. Morgantina Studies, Volume 5, New Jersey: Princeton University Press, 227-237. Blakeway, A. A. (1932-3) ‘Prolegomena to the Study of Greek Commerce with Italy, Sicily and France in the Eighth and Seventh centuries BC’, British School at Athens 33, 170-208. Boardman, J. (1980) The Greeks Overseas, London: Thames & Hudson Ltd. Boardman, J. (1991) ‘Early Greek Pottery on Black Sea Sites?’, Oxford Journal of Archaeology 10 (3), 387-390. Boardman, J. (1994) ‘Orientalia and Orientals on Ischia’ in D’Agnostino & Ridgway, D. ΑΠΟΙΚΙΑ – Scritti in onore di Giorgio Buchner, Naples: Istituto Universitario Orientale, 95-100. Boardman, J. (1998) ‘Olbia and Berezan: The Early Pottery’ in Tsetskhladze, G. R. (ed) Greek Colonisation of the Black Sea Area, Stuttgart: Steiner, 201- 204. Boardman, J. (2001) ‘Aspects of “Colonization”’, Bulletin of the American Schools of Oriental Research 322, 33-42. 108 Boardman, J. & Hammond, N. G. L. (eds) (1982) The Cambridge Ancient History. Volume III, Part 3. The Expansion of the Greek World, Eighth to Sixth Centuries B.C., 2nd ed, Cambridge: Cambridge University Press. Bowden, H. (1996) ‘The Greek Settlement and Sanctuaries at Naukratis: Herodotus and Archaeology’ in Hansen, M. H. & Raaflaub, K. (eds) More Studies in the Ancient Greek Polis, Copenhagen: Franz Steiner Verlag, 17- 38. Bradley, G. & Wilson, J. P. (eds) (2006) Greek and Roman Colonization: Origins, Ideologies, and Interactions, Swansea: The Classical Press of Wales. Brauer, G. C. Jr. (1986) Taras: Its History and Coinage, New York: Caratzas Publishing. Braund, D. (1998) ‘Writing and Re-Inventing Colonial Origins: Problems from Colchis and the Bosporus’ in Tsetskhladze, G. R. (ed) Greek Colonisation of the Black Sea Area, Stuttgart: Steiner, 287-296. Buchner, G. (1966) ‘Pithekoussai, Oldest Greek Colony in the West’, Expedition 8 (4), 4-12. Buchner, G. (1970-71) ‘Recent Work at Pithekoussai (Ischia), 1965-1971’, Archaeological Reports 17, 63-67. Buchner, G. BIBLIOGRAPHY (1979) ‘Early Orientalizing Aspects of the Euboean Connection’ in Ridgway, D & F. (eds) Italy Before the Romans, London: Academic Press Inc. Ltd., 129-144. Buchner, G. & Ridgway, D. (1993) Pithekoussai 1: La Necropoli. Tombe 1-723 Scanate dal 1952 al 1961, Rome: Monumenti Antichi. Burstein, S. M. (1974) Outpost of Hellenism: The Emergence of Heraclea on the Black Sea, California: University of California Press. Buyskikh, S. B. (2007) ‘Contacts Between Greeks and Non-Greeks on the Lower Bug in the Sixth and Fifth Centuries BC’ in Braund, D. & Kryzhitskii, S. D. (eds) Classical Olbia and the Scythian World: From the Sixth Century BC to the Second Century AD, Oxford: Oxford University Press, 23-36. Calame, C. (1990) ‘Narrating the Foundation of a City: The Symbolic Birth of Cyrene’ in Edmunds, L. (ed) Approaches to Greek Myth, Baltimore: John Hopkins University Press, 275-341. Camp, J. M. (1979) ‘A Drought in the Late Eighth Century B.C.’, Hesperia 48 (4), 397-411. 109 Carpenter, R. (1948) ‘The Greek Penetration of the Black Sea’, American Journal of Archaeology 52 (1), 1-10. Carter, J. C. (1981) ‘Rural Settlement at Metaponto’ in Barker, G. & Hodges, R. (eds) Archaeology and Italian Society: Prehistoric, Roman and Medieval Studies, Oxford: BAR International Series 102, 167-178. Carter, J. C. (1993) ‘Taking Possession of the Land: Early Colonization in Southern Italy’ in Scott, R. & Scott, A. R. (eds) Eius Virtutis Studiosi: Classical and Post-Classical Studies in Memory of Frank Edward Brown, Washington D.C.: National Gallery of Art, 342-67. Carter, J. C. (1998) The Chora of Metaponto: The Necropoleis. Volumes One and Two, Austin: University of Texas Press. Carter, J. C. (2006) Discovering the Greek Country at Metaponto, Ann Arbor: University of Michigan Press. Cawkwell, G. L. (1992) ‘Early Colonisation’, Classical Quarterly 42 (2), 289-303. Chapman, J. & Dolukhanov, P. (eds) (1997) Landscapes in Flux: Central and Eastern Europe in Antiquity, Oxford: Oxbow Books. Coja, M. (1990) ‘Greek Colonists and Native Populations in Dobruja (Moesia Inferior): The Archaeological Evidence’ in Descoeudres, J. P. (ed) Greek Colonists and Native Populations, Oxford: Oxford University Press, 157- 168. Coldstream, J. N. (1993) ‘Mixed Marriages at the Frontiers of the Early Greek World’, Oxford Journal of Archaeology 12, 89-108. Coldstream, J. N. (1994) ‘Prospectors and Pioneers: Pithekoussai, Kyme and Central Italy’ in Tsetskhladze, G. R & de Angelis, F (eds) The Archaeology of Greek Colonisation, Oxford: Oxford University Committee for Archaeology, 47-59. Coldstream, J. N. BIBLIOGRAPHY (1995) ‘Euboean Geometric Imports from the Acropolis of Pithekoussai’, The Annual of the British School at Athens 90, 251-267. Coldstream, J. N. (2003) Geometric Greece 900-700 BC, Oxford: Routledge, 2nd ed. Cook, R. M. (1962) ‘Reasons for the Foundation of Ischia and Cumae’, Historia 11 (1), 113-114. Crielaard, J. P. (1992-1993) ‘How the West was Won: Euboeans vs. Phoenicians’, Hamburger Beiträge zur Archäologie 19/20, 235-260. 110 d’Agnostino, B. (1996) ‘The Impact of the Greek Colonies on the Indigenous Peoples of Campania’ in Pugliese Carratelli, G. (ed) The Western Greeks. Classical Civilization in the Western Mediterranean, London: Thames & Hudson, 533-540. d’Agnostino, B. & Ridgway, D. (1994) ΑΠΟΙΚΙΑ – Scritti in onore di Giorgio Buchner, Naples: Istituto Universitario Orientale. Dalby, A. (1992) ‘Greeks Abroad: Social Organisation and Food Among the Ten Thousand’, Journal of Hellenic Studies 112, 16-30. Danner, P. (1997) ‘Megara, Megara Hyblaea, and Selinus: The Relationship Between the Town Planning of a Mother City, a Colony, and a Sub-Colony in the Archaic Period’, Acta Hyperborea 7, 143-165. De Angelis, F. (1994) ‘The Foundation of Selinous: Overpopulation or Opportunities?’ in Tsetskhladze, G. R & de Angelis, F (eds) The Archaeology of Greek Colonisation, Oxford: Oxford University Committee for Archaeology, 87-110. De Angelis, F. (1998) ‘Ancient Past, Imperial Present: The British Empire in T. J. Dunbabin’s The Western Greeks’, Antiquity 72, 539-549. De Angelis, F. (2003a) Megara Hyblaia and Selinous: The Development of Two Greek City-States in Archaic Sicily, Oxford: Oxford University School of Archaeology. De Angelis, F. (2003b) ‘Equations of Culture: The Meeting of Natives and Greeks in Sicily (ca 750-450 B.C.)’, Ancient West and East 2 (1), 19-50. De Angelis, F. (2009) ‘Colonies and Colonization’ in Boys-Stones, G, Graziosi, B, & Vasunia, P. (eds) The Oxford Handbook of Hellenic Studies, Oxford: Oxford University Press, 48-64. De Angelis, F. (2011) ‘Ancient Greek Colonization in the 21st Century: Some Suggested Directions’, Bollettino di Archeologia. Proceedings of the 17th International Congress of Classical Archaeology, Rome, September 22-26 2008, 18-30. Demetriou, D. (2011) ‘What is an Emporion? A Reassessment’, Historia 60 (3), 255-272. De Polignac, F. (2005) ‘Forms and Processes: Some Thoughts on the Meaning of Urbanization in Early Archaic Greece’ in Osborne, R. & Cunliffe, B. (eds) Mediterranean Urbanization 800-600 BC, Oxford: Oxford University Press, 45-69. 111 Desborough, V. (1976) ‘The Background to Euboean Participation in Early Greek Maritime Enterprise’ in Emmison, F. & Stephens, R. (eds) Tribute to an Antiquary. BIBLIOGRAPHY Essays Presented to Marc Fitch, London: Leopard’s Head Press Limited, 27-40. Descoeudres, J. P. (ed) (1990) Greek Colonists and Native Populations, Oxford: Oxford University Press. DeVoto, J. G. (2005) ‘Megarian Colonies in Sicily’, Ancient World 36 (1), 90-106. Dewald, C. (1980) ‘Biology and Politics: Women in Herodotus’ Histories’, Pacific Coast Philology 15, 11-18. Dietler, M. (2010) Archaeologies of Colonialism: Consumption, Entanglement, and Violence in Ancient Mediterranean France, Berkeley: University of California Press. Dietler, M. & Lopez-Ruiz, C. (eds) (2009) Colonial Encounters in Ancient Iberia: Phoenicians, Greek, and Indigenous Relations, Chicago: The University of Chicago Press. Dittenberger, W. (1903-05) Orientis Graeci Inscriptiones Selectae, Leipzig: Herzel. Di Vita, A. (1996) ‘Urban Planning in Ancient Sicily’ in Pugliese Carratelli, G. (ed) The Western Greeks. Classical Civilization in the Western Mediterranean, London: Thames & Hudson, 263-307. Docter, R. F. & Niemeyer, H. G. (1994) ‘Pithekoussai: the Carthaginian Connection on the Archaeological Evidence of Euboeo-Phoenician Partnership in the 8th and 7th Centuries B.C.’ in D’Agnostino & Ridgway, D. (eds) ΑΠΟΙΚΙΑ – Scritti in onore di Giorgio Buchner, Naples: Istituto Universitario Orientale, 101-116. Domanskij, J. V. & Marcenko, K. K. (2003) ‘Towards Determining the Chief Function of the Settlement of Borysthenes’ in Bilde, P. G., Hojte, J. M. & Stolba, V. F. (eds) The Cauldron of Ariantas, Studies Presented to A. N. Sceglov on the Occasion of his 70th Birthday, Aarhus: Aarhus University Press, 29-36. Dominguez, A. J. (1999) ‘Ephesos and Greek Colonization’ in Krinzinger, K. & Friesinger, H. (eds) 100 Jahre oi Sterreichische Forschungen in Ephesos, Vienna, 75-80. Dominguez, A. J. (2006) ‘Greeks in the Iberian Peninsula’ in Tsetskladze, G. R. (ed) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Volume One, Leiden: Brill, 429-505. 112 Dougherty, C. (1992) ‘When Rain Falls From the Clear Blue Sky: Riddles and Colonization Oracles’, Classical Antiquity 11 (1), 28-44. Dougherty, C. (1993a) The Poetics of Colonization: From City to Text in Archaic Greece, Oxford: Oxford University Press. Dougherty, C. (1993b) ‘It’s Murder to Found a Greek Colony’ in Dougherty, C. & Kurke, L. Cultural Poetics in Archaic Greece: Cult, Performance, Politics, Cambridge: Cambridge University Press, 178-198. Dougherty, C. (1994) ‘Archaic Greek Foundation Poetry: Questions of Genre and Occasion’, Journal of Hellenic Studies 114, 35-46. Dougherty, C. (2001) The Raft of Odysseus: The Ethnographic Imagination of Homer’s Odyssey, Oxford: Oxford University Press. Drews, R. BIBLIOGRAPHY (1976) ‘The Earliest Greek Settlements on the Black Sea’, Journal of Hellenic Studies 96, 18-31. Dunbabin, T. J. (1948) The Western Greeks, Oxford: Oxford University Press. Dunbabin, T. J. (1957) The Greeks and their Eastern Neighbours, London: Society for the Promotion of Hellenic Studies. Dusanic, S. (1978) ‘The ὅρκιον τῶν οἰκιστήρων and Fourth-Century Cyrene’, Chiron 8, 55-76. Eikeland, K. S. (2006) Indigenous Households: Transculturation of Sicily and Southern Italy in the Archaic Period, Goteborg: Goteborg University Press. Figueria, T. J. (1984) ‘Karl Polanyi and Ancient Greek Trade: The Port of Trade’, Ancient World 10, 15-30. Finley, M. I. (1962) ‘The Black Sea and Danubian Regions and the Slave Trade in Antiquity’, Klio 40, 51-59. Finley, M. I. (1968) Ancient Sicily to the Arab Conquest, London: Chatto & Windus. Finley, M. I. (1976) ‘Colonies – An Attempt at a Typology’, Transactions of the Royal Historical Society 26, 167-188. Fischer-Hansen, T. (1996) ‘The Earliest Town Planning of the Western Greek Colonies With Special Regard to Sicily’ in Hansen, M. H. (ed) Introduction to an Inventory of Poleis, Copenhagen: Munksgaard, 317-373. Fitzjohn, M. (2007) ‘Equality in the Colonies: Concepts of Equality in Sicily during the Eighth to Six Centuries BC’, World Archaeology 39 (2), 215- 228. 113 Frederiksen, R. (1999) ‘From Death to Life: The Cemetery of Fusco and the Reconstruction of Early Colonial Society’ in Tsetskhladze, G. (ed) Ancient Greeks West and East, Leiden: Brill, 229-266. Freeman, C. (1999) The Greek Achievement. The Foundation of the Western World, New York: Penguin. Garlan, Y. (1988) Slavery in Ancient Greece, Ithaca: Cornell University Press. Gjerstad, E. (1979) ‘A Cypro-Greek Royal Marriage in the 8th Century BC’ in Karageorghis, V. (ed) Studies Presented in Memory of Porphyrios Dikaios, Cyprus: Lions Club of Nicosia, 89-93. Glare, P. G. W. (1968) Oxford Latin Dictionary, Oxford: Oxford University Press. Gomez, F. J. (1993) ‘Iberia as a Barbarian Land: Perception of a Cultural Stereotype’, The Ancient World 24 (2), 131-142. Gosden, C. (2004) Archaeology and Colonialism: Cultural Contact from 5000 BC to the Present, Cambridge: Cambridge University Press. Graham, A. J. (1958) ‘The Date of the Greek Penetration of the Black Sea’, Bulletin of the Institute of Classical Studies 5, 25-42. Graham, A. J. (1960) ‘The Authenticity of the ὅπκιον τῶν οἰκιστήρων of Cyrene’, Journal of Hellenic Studies 80, 94-111. Graham, A. J. (1964) Colony and Mother City in Ancient Greece, Manchester: Manchester University Press. Graham, A. J. Graham, A. J. (1988) ‘Megara Hyblaea and the Sicels’ in Lordkipanidze, O. (ed) Local Ethnopolitical Entities of the Black Sea Littoral in the Seventh to Fourth Centuries BC, Tbilisi, 304-321. BIBLIOGRAPHY (1971) ‘Patterns in Early Greek Colonisation’, Journal of Hellenic Studies 91, 35-47. Graham, A. J. (1981-82) ‘Religion, Women and Greek Colonization’ in Religione e citta nel mondo antica, Atti, Centro ricerche e documentazione sull’antichita classica 11, 293-314. Graham, A. J. (1982) ‘The Colonial Expansion of Greece’ in Boardman, J. & Hammond, N. G. L. (eds) The Cambridge Ancient History, 2nd ed, Volume III, Part 3. The Expansion of the Greek World, Eighth to Sixth Centuries B.C., Cambridge: Cambridge University Press, 83-162. Graham, A. J. (1984) ‘Commercial Interchanges between Greeks and Natives’, Ancient World 10, 3-10. Graham, A. J. (1988) ‘Megara Hyblaea and the Sicels’ in Lordkipanidze, O. (ed) Local Ethnopolitical Entities of the Black Sea Littoral in the Seventh to Fourth Centuries BC, Tbilisi, 304-321. 114 Graham, A. J. (1990) ‘Pre-Colonial Contacts: Questions and Problems’ in Descoeudres, J. P. (ed) Greek Colonists and Native Populations, Oxford: Oxford University Press, 45-60. Graham, A. J. (1994/2001) ‘Greek and Roman Settlements on the Black Sea Coasts: Historical Background’ in Graham. A. J. Collected Papers on Greek Colonization, Leiden: Brill, 139-148. Graham, A. J. (1995) ‘The Odyssey, History, and Women’ in Cohen, B. (ed) The Distaff Side: Representing the Female in Homer’s Odyssey, Oxford: Oxford University Press, 3-16. Graham, A. J. (2001) Collected Papers on Greek Colonization, Leiden: Brill. Graham, H. F. (1961) ‘Soviet Classical Scholarship and the Black Sea Region’, The Classical Journal 56 (5), 194-202. Greaves, A. M. (1998) ‘Dionysius of Halicarnassus, Antiquitates Romanae 2.30 and Herodotus 1.146’, Classical Quarterly 48 (2), 572-574. Greco, E. (2006) ‘Greek Colonisation in Southern Italy: A Methodological Essay’ in Tsetskhasze, G. R. (ed) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Volume One, Leiden: Brill, 169- 200. Gwynn, A. (1918) ‘The Character of Greek Colonisation’, The Journal of Hellenic Studies 38, 88-123. Hackens, T., Holloway, N. D., and Holloway, R. R. (eds) (1984) Crossroads of the Mediterranean, Providence: Brown University. Hall, E. (1989) Inventing the Barbarian: Greek Self-Definition Through Tragedy, Oxford: Oxford University Press. Hall, J. M. (1997) Ethnic Identity in Greek Antiquity, Cambridge: Cambridge University Press. Hall, J. M. (2004) ‘How “Greek” were the Early Western Greeks?’ in Lomas, K. (ed) Greek Identity in the Western Mediterranean. Papers in Honour of Brian Shefton, Leiden: Brill, 35-54. Hall, J. M. (2006) ‘Foundation Stories’ in Tsetskhladze, G. R. (ed) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Hall, J. M. (2007) A History of the Archaic Greek World, ca. 1200-479 BCE, Massachusetts: Blackwell Publishing Ltd. Hodos, T. (2006) Local Responses to Colonization in the Iron Age Mediterranean, Oxford: Routledge. BIBLIOGRAPHY Volume Two, Leiden: Brill, 383-426. Hall, J. M. (2007) A History of the Archaic Greek World, ca. 1200-479 BCE, Massachusetts: Blackwell Publishing Ltd. 115 Hansen, M. H. (ed) (1996) Introduction to an Inventory of Poleis, Copenhagen: The Royal Danish Academy of Sciences and Letters. Hansen, M. H. (1997) ‘Emporion: A Study of the Use and Meaning of the Term in the Archaic and Classical Periods’ in Nielsen, T. H. (ed) Yet More Studies in the Ancient Greek Polis, Stuttgard: Steiner Verlag, 83-105. Hansen, M. H. & Nielsen, T. H. (eds) (2004) An Inventory of Archaic and Classical Poleis, New York: Oxford University Press. Hansen, P. A. (1976) ‘Pithecusan Humour. The Interpretation of ‘Nestor’s Cup’ Reconsidered’, Glotta 54, 25-44. Heinen, H. (2001) ‘Greeks, Iranians and Romans on the Northern Shore of the Black Sea’ in Tsetskhladze, G. R. (ed) North Pontic Archaeology: Recent Discoveries and Studies, Leiden: Brill, 1-23. Henneberg, M. & Henneberg, R. J. (2001) ‘Analysis of Human Skeletal and Dental Remains from Metaponto (7th – 2nd C BC)’, AttiTaranto 40, 461- 474. Henneberg, R. J. (1998) Dental Health and Affiliations of Inhabitants of the Ancient Greek Colony in Metaponto, Italy (6th – 3rd century BC), Johannesburg: University of Witwatersrand. Hind, J. (1983-84) ‘Greeks and Barbarian Peoples on the Shores of the Black Sea’, Archaeological Reports 30, 71-97. Hind, J. (1992-93) ‘Archaeology of the Greeks and Barbarian Peoples around the Black Sea (1982-1992)’, Archaeological Reports 39, 82-112. Hind, J. (1997) ‘Colonies and Ports-of-Trade on the Northern Shores of the Black Sea: Borysthenes, Kremnoi and the “Other Pontic Emporia” in Herodotos’ in Nielsen, T. H. (ed) Yet More Studies in the Ancient Greek Polis, Stuttgard: Steiner Verlag, 107-116. Hind, J. (1998) ‘Megarian Colonisation in the Western Half of the Black Sea (Sister and Daughter Cities of Herakleia)’ in Tsetskhladze, G. R. (ed) Greek Colonisation of the Black Sea Area, Stuttgart: Steiner, 131-152. Hodge, A. T. (1982) ‘Massalia – The Truth’, Classical Views 1, 164-173. Hodos, T. (1999) ‘Intermarriage in the Western Greek Colonies’, Oxford Journal of Archaeology 18 (1), 61-78. Hodos, T. (2006) Local Responses to Colonization in the Iron Age Mediterranean, Oxford: Routledge. 116 Hojte, J. M. (2008) ‘The Cities That Never Were: Failed Attempts at Colonization in the Black Sea’ in Bilde, P. G. & Petersen, J. H. (eds) Meetings of Cultures – Between Conflicts and Coexistence, Aarhus: Aarhus University Press, 149-162. Holloway, R. BIBLIOGRAPHY (1981) ‘Motives for Colonisation’ in Holloway, R. (ed) Italy and the Aegean 3000-700 BC, Louvain-la-Neuve: Institut Superieur d’Archeologie et d’Histoire de l’Art, College Erasme, 133-154. Holloway, R. R. (1991) The Archaeology of Ancient Sicily, London: Routledge. Hornblower, S. & Spawforth, A. (2012) The Oxford Classical Dictionary, Oxford: Oxford University Press, 4th ed. Humphreys, S. C. (1978) Anthropology and the Greeks, London: Routledge & Kegan Paul. Hurst, H. & Owen, S. (eds) (2005) Ancient Colonizations: Analogy, Similarity and Difference, London: Duckworth. Isaac, B. (2004) The Invention of Racism in Classical Antiquity, New Jersey: Princeton University Press. Jeffery, L. H. (1961) ‘The Pact of the First Settlers at Cyrene’, Historia 10 (2), 139-147. Jones, A. H. M. (1940) The Greek City from Alexander to Justinian, Oxford: Oxford University Press. Jones, S. (1997) The Archaeology of Ethnicity, London: Routledge. Just, R. (1989) Women in Athenian Law and Life, London: Routledge. Karlsson, L. (1989) ‘Some Notes on the Fortifications of Greek Sicily’, Opuscula Romana 17 (6), 77-89. Kirk, G. S. (1949) ‘Ships on Geometric Vases’, Annual of the British School at Athens 44, 93-153. Kirkwood, G. (1981) ‘Pythian 5.72-6, 9.90-2, and the Voice of Pindar’, Illinois Classical Studies 6 (1), 12-23. Klein, J. (1972) ‘A Greek Metalworking Quarter of the Eighth Century BC: Excavations on Ischia’, Expedition 14, 34-39. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. Kryzhitskii, S. D. (1997) ‘The Landscape of the North Pontic City-States – a Case Study From Olbia’ in Chapman, J. & Dolukhanov, P. (eds) (1997) 117 Landscapes in Flux: Central and Eastern Europe in Antiquity, Oxford: Oxbow Books, 101-114. Kryzhitskii, S. D. (2006) ‘The Rural Environs of Olbia: Some Problems of Current Importance’ in Bilde, P. G. & Stolba, V. F. (eds) Surveying the Greek Chora. The Black Sea Region in a Comparative Perspective, Aarhus: Aarhus University Press, 99-114. Kryzhitskii, S. D. (2007) ‘Criteria for the Presence of Barbarians in the Population of Early Olbia’ in Braund, D. & Kryzhitskii, S. D. (eds) Classical Olbia and the Scythian World: From the Sixth Century BC to the Second Century AD, Oxford: Oxford University Press, 17-22. Kurtz, D. C. & Boardman, J. (1971) Greek Burial Customs, New York: Cornell University Press. Kuznetsov, V. D. (1999) ‘Early Types of Greek Dwelling Houses in the North Black Sea’ in Tsetskhladze, G. R. (ed) Ancient Greeks West and East, Leiden: Brill, 531-564. La Rosa, V. (1996) ‘The Impact of the Greek Colonies on the Non-Hellenic Inhabitants of Sicily’ in Pugliese Carratelli, G. (ed) The Western Greeks. Classical Civilization in the Western Mediterranean, London: Thames & Hudson, 523-532. Labaree, B. W. (1957) ‘How the Greeks Sailed into the Black Sea’, American Journal of Archaeology 61 (1), 29-33. Laronde, A. (1990) ‘Greeks and Libyans in Cyrenaica’ in Descoeudres, J. P. (ed) Greek Colonists and Native Populations, Oxford: Oxford University Press, 169-180. Leighton, R. (ed) (1996) Early Societies in Sicily: New Developments in Archaeological Research, London: Accordia Research Centre. Leighton, R. (1999) Sicily Before History: An Archaeological Survey from the Palaeolithic to the Iron Age, New York: Cornell University Press. Lewis, D., Jeffery, L., Erxleben, E. & Hallof, K. (eds) (1981, 1994, 1998) Inscriptiones Graecae I: Inscriptiones Atticae Euclidis anno anteriores, Berlin: Berlin-Brandenburgische Akademie der Wissenschaften, 3rd ed. Liddell, H. G. & Scott, R. (1968) A Greek-English Lexicon, Oxford: Claredon Press, 9th ed, revised by Jones, H. S. Liston, M. A. (2012) ‘Reading the Bones: Interpreting the Skeletal Evidence for Women’s Lives in Ancient Greece’ in James, S. L. & Dillon, S. (eds) A 118 Companion to Women in the Ancient World, West Sussex: Blackwell Publishing Ltd, 125-140. Companion to Women in the Ancient World, West Sussex: Blackwell Publishing Ltd, 125-140. Littman, R. J. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. (1974) The Greek Experiment: Imperialism and Social Conflict 800- 400 BC, London: Thames and Hudson. Lomas, K. (ed) (2004a) Greek Identity in the Western Mediterranean: Papers in Honour of Brian Shefton, Leiden: Brill. Lomas, K. (2004b) ‘Introduction’ in Lomas, K. (ed) Greek Identity in the Western Mediterranean. Papers in Honour of Brian Shefton, Leiden: Brill, 1-14. Lorimer, H. L. (1950) Homer and the Monuments, London: Macmillan & Co Ltd. Luke, J. (2003) Ports of Trade, Al Mina and Geometric Greek Pottery in the Levant, BAR International Series 1100, Oxford: Archaeopress. Lyons, C. L. (1996a) The Archaic Cemeteries. Morgantina Studies, Volume 5, New Jersey: Princeton University Press. Lyons, C. L. (1996b) ‘Sicel Burials at Morgantina: Defining Social and Ethnic Identities’ in Leighton, R. (ed) Early Societies in Sicily: New Developments in Archaeological Research, London: Accordia Research Centre, 177-188. Lyons, C. L. (2000) ‘Gender and Burial in Early Colonial Sicily: The Case of Morgantina’ in Donald, M. & Hurcombe, L. (eds) Representations of Gender from Prehistory to the Present, Great Britain: Macmillan Press Ltd, 87-103. Lyons, C. L. & Papadopoulos, J. K. (eds) (2002) The Archaeology of Colonialism: Issues and Debates, Los Angeles: Getty Research Institute. Maddoli, G. (1996) ‘Cults and Religious Doctrines of the Western Greeks’ in Pugliese Carratelli, G. (ed) The Western Greeks. Classical Civilization in the Western Mediterranean, London: Thames & Hudson, 481-498. Malkin, I. (1987) Religion and Colonization in Ancient Greece, Leiden: Brill. Malkin, I. (1994a) Myth and Territory in the Spartan Mediterranean, Cambridge: Cambridge University Press. Malkin, I. (1994b) ‘Inside and Outside: Colonisation and the Formation of the Mother City’ in D’Agnostino & Ridgway, D. ΑΠΟΙΚΙΑ – Scritti in onore di Giorgio Buchner, Naples: Istituto Universitario Orientale, 1-9. Malkin, I. (1998) The Returns of Odysseus: Colonization and Ethnicity, California: University of California Press. 119 Malkin, I. (2002a) ‘A Colonial Middle Ground: Greek, Etruscan, and Local Elites in the Bay of Naples’ in Lyons, C. L. & Papadopoulos, J. K. (eds) The Archaeology of Colonialism, Los Angeles: Getty Research Institute, 151- 181. Malkin, I. (2002b) ‘Exploring the Validity of the Concept of ‘Foundation’: A Visit to Megara Hyblaia’ in Gorman, V. B. & Robinson, E. W. (eds) Oikistes: Studies in Constitutions, Colonies, and Military Power in the Ancient World. Offered in Honor of A. J. Graham, Leiden: Brill, 195-225. Malkin, I. (2004) ‘Postcolonial Concepts and Ancient Greek Colonization’, Modern Language Quarterly 65 (3), 341-364. Marshall, E. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. (1999) ‘Sex and Paternity: Gendering the Foundation of Kyrene’ in Foxhall, L. & Salmon, J. (eds) When Men were Men, London: Routledge, 98-110. Marshall, E. (2004) ‘Women and the Transmission of Libyan Culture’ in McHardy, F. & Marshall, E. (eds) Women’s Influence on Classical Civilisation, London: Routledge, 127-136. Meiggs, R. & Lewis, D. (eds) (1969) Selection of Greek Historical Inscriptions to the end of the Fifth Century BC, Oxford: Claredon. Mertens, D. & Greco, E. (1996) ‘Urban Planning in Magna Graecia’ in Pugliese Carratelli, G. (ed) The Western Greeks. Classical Civilization in the Western Mediterranean, London: Thames & Hudson, 243-262. Mierse, W. E. (1994) ‘Ampurias: The Urban Development of a Graeco-Roman City on the Iberian Coast’, Latomus 54, 790-805. Mitchell, B. J. (2000) ‘Cyrene: Typical or Atypical?’ in Brock, R. & Hodkinson, S. (eds) Alternatives to Athens, New York: Oxford University Press, 82- 102. Moller, A. (2000) Naukratis: Trade in Archaic Greece, New York: Oxford University Press Inc. Moller, A. (2001) ‘Naukratis, or How to Identify a Port of Trade’ in Tandy, D. W. (ed) Prehistory and History: Ethnicity, Class and Political Economy, Montreal: Black Rose Books, 145-158. Morakis, A. (2011) ‘Thucydides and the Character of Greek Colonisation in Sicily’, Classical Quarterly 61 (2), 460-492. Morel, J. P. (1984) ‘Greek Colonization in Italy and the West (Problems of Evidence of Interpretation)’ in Hackens, T, Holloway, N. D, and Holloway, 120 R. R. (eds) Crossroads of the Mediterranean, Providence: Brown University, 123-162. Morel, J. P. (2006) ‘Phocaean Colonisation’ in Tsetskhasze, G. R. (ed) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Volume One, Leiden: Brill, 358-428. Morgan, C. & Hall, J. (1996) ‘Achaian Poleis and Achaian Colonisation’ in Hansen, M. H. (ed) Introduction to an Inventory of Poleis, Copenhagen: Munksgaard, 164-232. Morris, I. (1987) Burial and Ancient Society. The Rise of the Greek City-State, Cambridge: Cambridge University Press. Morris, I. (1992) Death-Ritual and Social Structure in Classical Antiquity, Cambridge: Cambridge University Press. Morris, I. (1996) ‘The Absolute Chronology of the Greek Colonies in Sicily’, Acta Archaologica Supplementum 67, 51-59. Morrison, J. S. & Williams, R. T. (1968) Greek Oared Ships, Cambridge: Cambridge University Press. Murray, O. (1993) Early Greece, London: Fontana, 2nd ed. Nazarov, V. V. (1997) ‘The Ancient Landscape of Berezan Island’ in Chapman, J. & Dolukhanov, P. (eds) (1997) Landscapes in Flux: Central and Eastern Europe in Antiquity, Oxford: Oxbow Books, 131-136. Nevett, L. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. (1994) ‘Separation or Seclusion? Towards an Archaeological Approach to Investigating Women in the Greek Household in the Fifth to Third Centuries BC’ in Pearson, M. P. & Richards, C. (eds) Architecture and Order: Approaches to Social Space, London: Routledge, 98-112. Nevett, L. (1999) House and Society in the Ancient Greek World, Cambridge: Cambridge University Press. Nielsen, T. H. (ed) (1997) Yet More Studies in the Ancient Greek Polis, Stuttgard: Steiner Verlag. Oakley, J. H. & Sinos, R. H. (1993) The Wedding in Ancient Athens, Wisconsin: The University of Wisconsin Press. Osborne, R. (1998) ‘Early Greek Colonization? The Nature of Greek Settlement in the West’ in Fisher, N. & Van Wees, H. (eds) Archaic Greece: New Approaches and New Evidence, London: Duckworth & The Classical Press of Wales, 251-269. 121 Osborne, R. (2009) Greece in the Making 1200-479 BC, Oxford: Routledge, 2nd ed. Owen, S. (2003) ‘Of Dogs and Men: Archilochos, Archaeology and the Greek Settlement of Thasos’, Proceedings of the Cambridge Philological Society 49, 1-18. Owen, S. (2005) ‘Analogy, Archaeology and Archaic Greek Colonization’ in Hurst, H. & Owen, S. (eds) Ancient Colonizations: Analogy, Similarity and Difference, London: Duckworth, 5-22. Parke, H. W. & Wormell, D. E. W. (1956) The Delphic Oracle: Volume 1, The History, Oxford: Basil Blackwell. Pashkevich, G. A. (2001) ‘Archaeobotanical Studies on the Northern Coast of the Black Sea’, Eurasia Antiqua 7, 511-567. Pembroke, S. (1970) ‘Locres et Tarente: le role des femmes dans la fondation de deux Colonies Grecques’, Annales (Economies, Societes, Civilisations) 25 (5), 1240-1270. Petersen, J. H. (2010) Cultural Interaction and Social Strategies on the Pontic Shores: Burial Customs in the Northern Black Sea Area c. 550-270 BC, Aarhus: Aarhus University Press. Petropoulos, E. K. (2005) Hellenistic Colonization in Euxeinos Pontos: Penetration, Early Establishment and the Problem of the ‘Emporion’ Revisited, BAR International Series 1394, Oxford: Archaeopress. Polanyi, K. (1963) ‘Ports of Trade in Early Societies’, Journal of Economic History 23 (1), 30-45. Pomeroy, S. (1995) Goddesses, Whores, Wives, and Slaves, New York: Schocken Books. Pomey, P. (1996) ‘Navigation and Ships in the Age of Greek Colonization’ in Pugliese Carratelli, G. (ed) The Western Greeks. Classical Civilization in the Western Mediterranean, London: Thames & Hudson, 133-140. Popham, M. (1987) ‘An Early Euboean Ship’, Oxford Journal of Archaeology 6, 353-359. Pugliese Carratelli, G. (ed) (1996) The Western Greeks. Classical Civilization in the Western Mediterranean, London: Thames & Hudson. Redfield, J. M. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. (2003) The Locrian Maidens: Love and Death in Greek Italy, New Jersey: Princeton University Press. 122 Ridgway, D. (1973) ‘The First Western Greeks: Campanian Coasts and Southern Etruria’ in Hawkes, C. & S. (eds) Greeks, Celts and Romans: Studies in Venture and Resistance, New Jersey: Rowman and Littlefield, 5-38. Ridgway, D. (1990) ‘The First Western Greeks and their Neighbours, 1935-1985’ in Descoeudres, J. P. (ed) Greek Colonists and Native Populations, Oxford: Oxford University Press, 61-72. Ridgway, D. (1981) ‘The Foundation of Pithekoussai’ in Nouvelle contribution a l’etude de la societe et de la colonisation Eubeennes, Naples: Cahiers du centre Jean Berard, 45-60. Ridgway, D. (1992) The First Western Greeks, Cambridge: Cambridge University Press. Rihll, T. (1993) ‘War, Slavery, and Settlement in Early Greece’ in Rich, J. & Shipley, G. (eds) War and Society in the Greek World, London: Routledge, 77-107. Roebuck, C. (1959) Ionian Trade and Colonization, New York: Archaeological Institute of America. Rosenbaum, E. (1960) A Catalogue of Cyrenaican Portrait Sculpture, London: Oxford University Press. Rouge, J. (1970) ‘La Colonisation Grecque et les Femmes’, Cahiers d’histoire 15, 307-317. Rutter, K. (1986) ‘Sicily and South Italy: The Background to Thucydides Books 6 and 7’, Greece & Rome 33 (2), 142-155. Rutter, N. K. (2002) ‘Syracusan Democracy: Most like the Athenian?’ in Brock, R. & Hodkinson, S. (eds) Alternatives to Athens, Oxford: Oxford University Press, 137-151. Sallares, R. (1991) The Ecology of the Ancient Greek World, New York: Cornell University Press. Sampson, A. (1985) ΜΑΝΙΚΑ: ΜΙΑ ΠΡΩΤΟΕΛΛΑΔΙΚΗ ΠΟΛΗ ΕΤΗ ΧΑΛΚΙΔΑ, Athens: Society for Euboean Studies. Schaps, D. M. (1979) Economic Rights of Women in Ancient Greece, Edinburgh: Edinburgh University Press. Scheidel, W. (2003) ‘The Greek Demographic Expansion: Models and Comparisons’, Journal of Hellenic Studies 123, 122-40. 123 Schutkowski, H. (1993) ‘Sex Determination of Infant and Juvenile Skeletons: I. Morphognostic Features’, American Journal of Physical Anthropology 90, 199-205. Serrati, J. (2000) ‘Sicily from Pre-Greek Times to the Fourth Century’ in Smith, C. & Serrati, J. (eds) Sicily from Aeneas to Augustus: New Approaches in Archaeology and History, Edinburgh: Edinburgh University Press, 9-14. Seyrig, H. (1927) ‘Quatre Cultes de Thasos’, Bulletin de Correspondance Hellenique 51, 178-233. Shefton, B. B. (1994) ‘Massalia and Colonization in the North-Western Mediterranean’ in Tsetskhladze, G. R & de Angelis, F (eds) The Archaeology of Greek Colonisation, Oxford: Oxford University Committee for Archaeology, 61-86. Shepherd, G. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. (1995) ‘The Pride of Most Colonials: Burial and Religion in the Sicilian Colonies’, Acta Hyperborea 6, 51-82. Shepherd, G. (1999) ‘Fibulae and Females: Intermarriage in the Western Greek Colonies and the Evidence from Cemeteries’ in Tsetskhladze, G. R. (ed) Ancient Greeks West and East, Leiden: Brill, 267-300. Shepherd, G. (2000) ‘Greeks Bearing Gifts: Religious Relationships Between Sicily and Greece in the Archaic Period’, in Smith, C. & Serrati, J. (eds) Sicily from Aeneas to Augustus: New Approaches in Archaeology and History, Edinburgh: Edinburgh University Press, 55-70. Shepherd, G. (2005) ‘Dead Men Tell No Tales: Ethnic Diversity in Sicilian Colonies and the Evidence of the Cemeteries’, Oxford Journal of Archaeology 24 (2), 115-136. Shepherd, G. (2012) ‘Women in Magna Graecia’ in James, S. L. & Dillon, S. (eds) A Companion to Women in the Ancient World, West Sussex: Blackwell Publishing, 215-228. Shilik, K. K. (1997) ‘Oscillations of the Black Sea and Ancient Landscapes’ in Chapman, J. & Dolukhanov, P. (eds) (1997) Landscapes in Flux: Central and Eastern Europe in Antiquity, Oxford: Oxbow Books, 115-129. Sissa, G. (1990) Greek Virginity, Massachusetts: Harvard University Press. Sjoqvist, E. (1958) ‘Excavations at Serra Orlando (Morgantina): Preliminary Report II’, Americal Journal of Archaeology 62 (2), 155-164. Sjoqvist, E. (1973) Sicily and the Greeks: Studies in the Interrelationship Between the Indigenous Populations and the Greek Colonists, Ann Arbor: The University of Michigan Press. 124 Snodgrass, A. M. (1971) The Dark Age of Greece, Edinburgh: Edinburgh University Press. Snodgrass, A. M. (1971) The Dark Age of Greece, Edinburgh: Edinburgh University Press. Snodgrass, A. M. (1980) Archaic Greece: The Age of Experiment, California: University of California Press. Snodgrass, A. M. (1994) ‘The Nature and Standing of the Early Western Colonies’ in Tsetskhladze, G. & De Angelis, F. (eds) The Archaeology of Greek Colonisation, Oxford: Oxford University Committee for Archaeology. Snodgrass, A. M. (2005) ‘“Lesser Breeds”: The History of a False Analogy’ in Hurst, H. & Owen, S. (eds) Ancient Colonizations: Analogy, Similarity and Difference, London: Duckworth, 45-58. Solovyov, S. L. (1998) ‘Archaic Berezan: Historical – Archaeological Essay’ in Tsetskhladze, G. R. (ed) Greek Colonisation of the Black Sea Area, Stuttgart: Steiner, 205-225. Solovyov, S. L. (1999) Ancient Berezan: The Architecture, History and Culture of the First Greek Colony in the Northern Black Sea, Leiden: Brill. Solovyov, S. L. (2001) ‘The Archaeological Excavation of the Berezan Settlement (1987-1991)’ in Tsetskhladze, G. R. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. (ed) North Pontic Archaeology: Recent Discoveries and Studies, Leiden: Brill, 117-141. Stucchi, S. (1989) ‘Problems Concerning the Coming of the Greeks to Cyrenaica and the Relations with their Neighbours’, Meditarch 2, 73-84. Toms, J. (2000) ‘The Arch Fibula in Early Iron Age Italy’ in Ridgway, D., Ridgway, F. R. S., Pearce, M., Herring, E., Whitehouse, R. D. & Wilkins, J. B. (eds) Ancient Italy in its Mediterranean Setting, London: Accordia Research Centre, 91-116. Tsakirgis, B. (1995) ‘Morgantina: A Greek Town in Central Sicily’, Acta Hyperborea 6, 123-147. Tsetskhladze, G. R. (1994) ‘Greek Penetration of the Black Sea’ in Tsetskhladze, G. R & de Angelis, F (eds) The Archaeology of Greek Colonisation, Oxford: Oxford University Committee for Archaeology, 111-135. Tsetskhladze, G. R. (1995) ‘Did the Greeks go to Colchis for Metals?’, Oxford Journal of Archaeology 14 (3), 307-331. Tsetskhladze, G. R. (1997) ‘How Greek Colonists Adapted Their Way of Life to the Conditions in Kolkhis’ in Fossey, J. M. (ed) Proceedings of the First International Conference on the Archaeology and History of the Black Sea, Amsterdam: J. C. Gieben, 121-136. 125 Tsetskhladze, G. R. (ed) (1998a) Greek Colonisation of the Black Sea Area, Stuttgart: Steiner. Tsetskhladze, G. R. (1998b) ‘Greek Colonisation of the Black Sea Area: Stages, Models, and Native Population’ in Tsetskhladze, G. R. (ed) Greek Colonisation of the Black Sea Area, Stuttgart: Steiner, 9-68. Tsetskhladze, G. R. (ed) (1999) Ancient Greeks West and East, Leiden: Brill. Tsetskhladze, G. R. (ed) (2001) North Pontic Archaeology: Recent Discoveries and Studies, Leiden: Brill. Tsetskhladze, G. R. (2002) ‘Ionians Abroad’ in Tsetskhladze, G. R. & Snodgrass, A. M. (eds) Greek Settlements in the Eastern Mediterranean and the Black Sea. BAR International Series 1062, Oxford: Archaeopress, 81-96. Tsetskhladze, G. R. (ed) (2006a) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Volume One, Leiden: Brill. Tsetskhladze, G. R. (ed) (2006b) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Volume Two, Leiden: Brill. Tsetskhladze, G. R. (2006c) ‘Introduction: Revisiting Ancient Greek Colonisation’ in Tsetskhladze, G. R. (ed) Greek Colonisation: An Account of Greek Colonies and Other Settlements Overseas. Volume One, Leiden: Brill, xxiii-lxxxiii. Tsetskhladze, G. R. & De Angelis, F. (eds) (1994) The Archaeology of Greek Colonisation, Oxford: Oxford University Committee for Archaeology. Tsetskhladze, G. R. & Snodgrass, A. M. (eds) (2002) Greek Settlements in the Eastern Mediterranean and the Black Sea. BAR International Series 1062, Oxford: Archaeopress. Tsetskhladze, G. Tuplin, C. (ed) (2003) Pontus and the Outside World: Studies in Black Sea History, Historiography, and Archaeology, Leiden: Brill. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. R. & Treister, M. Y. (1995) ‘The Metallurgy and Production of Precious Metals in Colchis Before and After the Arrival of the Ionians (Towards the Problem of the Reason for Greek Colonisation)’, Bulletin of the Metals Museum 24, 1-32. Tuplin, C. (1999) ‘Greek Racism? Observations on the Character and Limits of Greek Ethnic Prejudice’ in Tsetskhladze, G. R. (ed) Ancient Greeks West and East, Leiden: Brill, 47-76. Tuplin, C. (ed) (2003) Pontus and the Outside World: Studies in Black Sea History, Historiography, and Archaeology, Leiden: Brill. 126 Van Compernolle, R. (1983) ‘Femmes indigenes et colonisateurs’ in Modes de Contacts et Processus de Transformation dans les Societes Anciennes, Rome: l’Ecole Francaise de Rome, 1033-1049. Van Dommelen, P. (1997) ‘Colonial Constructs: Colonialism and Archaeology in the Mediterranean’, World Archaeology 28 (3), 305-323. Van Dommelen, P. (2002) ‘Ambiguous Matters: Colonialism and Local Identities in Punic Sardinia’ in Lyons and Papadopoulos (eds) The Archaeology of Colonialism, Los Angeles: Getty Research Institute, 121-147. Van Gennep, A. (1960) The Rites of Passage, London: Routledge & Kegan Paul, trans Vizedom, M. B. & Caffee, G. L. Van Wees, H. (2004) Greek Warfare: Myths and Realities, London: Gerald Duckworth & Co Ltd. Waldbaum, J. C. (1997). ‘Greeks in the East or Greeks and the East?: Problems in the Definition and Recognition of Presence’, Bulletin of the American Schools of Oriental Research 305, 1–17. Walker, P. L. & Cook, D. C. (1998) ‘Gender and Sex: Vive La Difference!’, American Journal of Anthropology 106, 255-259. Wallinga, H. T. (1993) Ships and Sea-Power Before the Great Persian War, Leiden: Brill. Wanis, S. (1992) ‘A New Relief from Cyrene with a Libyan Scene’ Libyan Studies 23, 41-44. Waters, K. H. (1974) ‘The Rise and Decline of Some Greek Colonies in Sicily’, Ancient Society 5, 1-19. Watkins, C. (1976) ‘Observations on the “Nestor’s Cup” Inscription’, Harvard Studies in Classical Philology 80, 25-40. Weaver, D. S. (1980) ‘Sex Differences in the Ilia of a Known Sex and Age Sample of Fetal and Infant Skeletons’, American Journal of Physical Anthropology 52, 191-195. West, S. (2007) ‘Herodotus and Olbia’ in Braund, D. & Kryzhitskii, S. D. (eds) Classical Olbia and the Scythian World: From the Sixth Century BC to the Second Century AD, Oxford: Oxford University Press, 79-92. White, D. (1987) ‘Demeter Libyssa: Her Cult in Light of the Recent Excavations’, Quaderni di archeologia della Libia 12, 67-84. White, T. D. & Folkens, P. A. Koromila, M. (1991) The Greeks in the Black Sea, Athens: Panorama. (2005) The Human Bone Manual, Massachusetts: Elsevier Academic Press. 127 Whitehouse, R. & Wilkins, J. B. (1989) ‘Greeks and Natives in Southeast Italy: Approaches to the Archaeological Evidence’ in Champion, T. C. (ed) Centre and Periphery: Comparative Studies in Archaeology, London: Unwin Hyan, 102-126. Whitley, J. (2001) The Archaeology of Ancient Greece, Cambridge: Cambridge University Press. Wiedemann, T. E. J. (1983) ‘Thucydides, Women and the Limits of Rational Analysis’, Greece & Rome 30 (2), 163-170. Wilson, J. P. (1997) ‘The Nature of Greek Overseas Settlements in the Archaic Period: Emporion or Apoikia?’ in Mitchell, L. G. & Rhodes, P. J. (eds) The Development of the Polis in Archaic Greece, London: Routedge, 199-207 Wilson, J. P. (2006) ‘’Ideologies’ of Greek Colonization’ in Bradley, G. & Wilson, J. P. (eds) Greek and Roman Colonization: Origins, Ideologies, and Interactions, Swansea: The Classical Press of Wales, 25-57. Wilson, R. J. A. (1981-82) ‘Archaeology in Sicily, 1977-1981’, Archaeological Reports 28, 84-105. Wilson, R. J. A. (1987-88) ‘Archaeology in Sicily, 1982-87’, Archaeological Reports 34, 105-150. Woodhead, A. G. (1981) The Study of Greek Inscriptions, Cambridge: Cambridge University Press, 2nd ed. Wylie, A. (1991) ‘Gender Theory and the Archaeological Record: Why is There No Archaeology of Gender?’ in Gero, J. M. & Conkey, M. W. (eds) Engendering Archaeology: Women and Prehistory, Oxford: Basil Blackwell, 31-54. Zeitlin, F. (1986) ‘Configurations of Rape in Greek Myth’ in Tomaselli, S. & Porter, R. (eds) Rape, Oxford: Basil Blackwell, 122-151. 128
https://openalex.org/W3027918111	https://www.researchsquare.com/article/rs-7933/v2.pdf	English	null	Radial peripapillary capillary density in superior segmental optic hypoplasia measured with OCT angiography	BMC ophthalmology	2,020	cc-by	3,364	Radial peripapillary capillary density in superior segmental optic hypoplasia measured with OCT angiography. Maiko Abe Tohoku University Graduate School of Medicine Kazuko Omodaka Tohoku University Graduate School of Medicine Tsutomu Kikawa Topcon Corporation Toru Nakazawa (  ntoru@oph.med.tohoku.ac.jp ) https://orcid.org/0000-0002-5591-4155 Research article Keywords: optical coherence tomography angiography, diagnosis, radial peripapillary capillary density, superior segmental optic hypoplasia. Posted Date: January 6th, 2020 DOI: https://doi.org/10.21203/rs.2.17299/v2 DOI: https://doi.org/10.21203/rs.2.17299/v2 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 1/9 Page 1/9 Abstract Background: To investigate the diagnostic power of radial peripapillary capillary (RPC) density, measured with optical coherence tomography angiography (OCT-A), in patients with superior segmental optic hypoplasia (SSOH). Methods: Forty subjects with SSOH and 40 age- and axial length-matched control subjects were retrospectively registered for this study. SSOH was defined as intraocular pressure less than 21 mmHg with the presence of two of the following: superior rim thinning, superior entrance of the central retinal artery, scleral halo, and pale optic disc; as well as non-progressive visual field loss. RPC density was measured with swept-source OCT-A (Triton, Topcon) overall, in the quadrants, and in the 12 clock-wise sectors. Changes in RPC density were also compared in SSOH patients and age-matched patients with mild- or moderate-stage of glaucoma. RPC density was compared in pairs of groups with Welch’s t-test. Diagnostic power was assessed with the area under the receiver operating characteristics curve (AUC). Results: Overall cpRNFLT was significantly different in the normal (106.7 ± 9.5 μm) and SSOH (77.2 ± 13.7 μm, p < 0.001) subjects. RPC density overall and in the superior, nasal, and inferior quadrants was significantly lower in the SSOH group (all, p < 0.001), but not in the temporal (p = 0.756) quadrant. The diagnostic power of RPC density was highest in the superior quadrant (AUC = 0.928) and the 1 o’clock sector (0.896). Comparing the SSOH and glaucoma patients showed that there were no significant differences in RPC density either overall (p=0.391) or in the superior quadrant (p = 0.268), while RPC density was significantly higher in the inferior (p = 0.005) and temporal quadrants (p < 0.001) and lower in the nasal quadrant (p = 0.029). Conclusions: Low RPC density was found in the three non- temporal quadrants of the optic nerve head in SSOH patients, in comparison to normal subjects. Regionally, RPC density in SSOH was lower in the nasal quadrant and higher in the inferior and temporal quadrants in comparison to glaucoma patients. Measuring RPC density with OCT-A may help the diagnosis of SSOH and may improve the management of glaucoma. Background Superior segmental optic hypoplasia (SSOH) is a congenital anomaly of the optic nerve head (ONH) characterized by superior entry of the central retinal vessels, superior peripapillary scleral halo, and reduced circumpapillary retinal nerve fiber layer thickness (cpRNFLT) with non-progressive inferior-sector visual field defects. In Asia, normal-tension glaucoma (NTG) is the primary subtype of open-angle glaucoma (1) making it important to differentiate SSOH from NTG (2). Previously, it was reported that SSOH could be identified based on the location of cpRNFLT loss (3) and structural abnormalities in the extension of the retinal pigment epithelium (RPE) over the nasal disc margin (4), both measured with spectral-domain optical coherence tomography (OCT). Thus, both fundus photography and OCT may be useful in the clinical treatment of SSOH. We previously showed that tissue mean blur rate (MT), representing blood flow (BF) in the ONH, decreases in SSOH (5). MT is also correlated with deep-ONH BF, but not surface-ONH BF (6). Recently, OCT angiography (OCT-A), based on swept-source OCT, has become available clinically and is increasingly important for visualizing vessels both in the retina and ONH. The density of radial Page 2/9 Page 2/9 peripapillary capillaries (RPCs), a network of surface nerve fiber layers, has been shown to have diagnostic power for glaucoma (7). However, until now, there have been no reports on changes in RPC density or the diagnostic power of RPC density in SSOH. This study is thus the first to assess the microvasculature of the peripapillary retina and investigate whether vascular parameters are useful for diagnosing SSOH. peripapillary capillaries (RPCs), a network of surface nerve fiber layers, has been shown to have diagnostic power for glaucoma (7). However, until now, there have been no reports on changes in RPC density or the diagnostic power of RPC density in SSOH. This study is thus the first to assess the microvasculature of the peripapillary retina and investigate whether vascular parameters are useful for diagnosing SSOH. Methods Eighty eyes of 40 normal and 40 SSOH patients were included. SSOH was diagnosed by glaucoma specialists after a careful differential diagnosis for glaucoma (4). SSOH diagnosis was based on the presence of more than two of the following four symptoms: superior rim thinning, superior entrance of the central retinal artery, scleral halo, and pale optic disc; combined with non-progressive visual field loss (in an average of 6.3 ± 4.8 visual field tests), peripheral visual field defects in the V-4 target of the Goldmann visual field test, and intraocular pressure (IOP) less than 21 mmHg. Mean deviation measurements used reliable data from Humphrey field analyzer standard automated perimetry. We also recruited 50 separate patients, including 27 SSOH patients and 23 age-matched mild- or moderate-stage glaucoma patients, to compare regional differences in RPC density. RPC density was measured with OCT- A (Triton, Topcon), as previously described (7), overall, in the superior, temporal, inferior, and nasal quadrants, and in 12 clock-wise sectors. Swept-source OCT (SS-OCT) uses a tissue-penetrating laser system with a long central wavelength of 1,050 nm and allows patients to easily maintain good fixation during scanning due to its low glare. The device is also equipped with an eye-tracking system. Thus, OCT- A with SS-OCT has led to significant improvements in the observation of retinal capillaries, both at the surface and in the deep areas of the ONH, in glaucoma patients. cpRNFLT was measured with swept- source OCT. This study adhered to the tenets of the Declaration of Helsinki, and the protocols were approved by the institutional review board of the Tohoku University Graduate School of Medicine (study 2017-1-290). Clinical findings and other characteristics, including age, sex, axial length, and IOP, were compared in pairs of groups with Welch’s t-test. The area under the receiver operating characteristic curve (AUC) was calculated for differentiating normal subjects and SSOH patients based on OCT-A parameters. All statistical analyses were performed with JMP software (Pro version 13.1.0, SAS Institute Japan Inc., Tokyo, Japan). Discussion We found that the power of RPC density to differentiate SSOH patients and controls was strong, especially in the superior, nasal, and inferior areas. Previously, we used laser speckle flowgraphy to show that SSOH, a congenital anomaly, causes reduced microcirculation in the ONH (5), that the quadrant MT ratios in the ONH have a strong power to differentiate SSOH and NTG (indicating that SSOH has a fundamentally different pathogenesis from NTG) (5), and that the structure of the deep layers of the ONH, around the lamina cribrosa, are correlated to MT, but not RPC density (6). Here, we found that the density of the peripapillary retinal microvasculature decreased in SSOH, but that the pattern of RPC loss was distinctly different from that reported for cpRNFLT (3). RPC density in the 9 o’clock sector, outside the temporal area, was significantly lower, but RPC density in the 3 o’clock sector was not significantly different in the SSOH and normal subjects. On the other hand, cpRNFLT was not significantly different in the 9 o’clock sector (normal: 67.6 ± 9.9, SSOH: 70.3 ± 19.4, p = 0.431), but was in the 3 o’clock sector (normal: 63.6 ± 12.7, SSOH: 57.8 ± 26.9, p < 0.001). Thus, the location of changes in RPC density and cpRNFLT is on horizontally opposite sides of the ONH. This may be due to lower initial RPC density in the nasal ONH, higher RPC density in the temporal ONH, and the location of the major retinal central vessel in the superior and inferior ONH. These findings suggest that OCT-A might have the potential to diagnose SSOH. In this study, we also performed an investigation to compare SSOH patients and age-matched mild- and moderate-stage glaucoma patients. RPC density in the SSOH patients was not significantly different overall or in the superior quadrant but was significantly higher in the inferior and temporal quadrants and lower in the nasal quadrant. These differences in RPC density are understandable, because the area of the nerve fiber layer that is vulnerable to damage is different in SSOH and glaucoma. Our finding also shows that RPC density may be valuable for assessing damage in SSOH. Theoretically, cpRNFLT damage in SSOH is developmental, while damage in glaucoma is secondary. However, the diagnostic power of In this study, we also performed an investigation to compare SSOH patients and age-matched mild- and moderate-stage glaucoma patients. Results There were no differences between the groups in age (normal: 39.5 ± 8.2, SSOH: 41.2 ± 18.0, p = 0.326), axial length (24.6 ± 0.8, 25.1 ± 1.3, p = 0.088), or IOP (14.0 ± 2.2, 14.1 ± 2.5, p = 0.749). Figure 1 shows a representative SSOH patient without visual field progression over 5 years and a normal subject; RPC density in the superior to nasal quadrants is clearly lower in the SSOH subject. Table 1 shows differences in RPC density between the normal and SSOH groups. There were significant differences in overall RPC density and in the superior, nasal, and inferior (all: p < 0.001), but not temporal (p = 0.756) quadrants. Page 3/9 Page 3/9 RPC density was significantly lower in the 1, 2, 4, 5, 7, 9, 11, and 12 o’clock sectors. Table 2 shows the diagnostic power of RPC density. The highest AUC was in the superior quadrant (AUC = 0.928, Fig. 1E), with a cutoff value of 42.14% (83% sensitivity and 88% specificity). The 1 o’clock sector had the highest AUC (0.896, Fig. 1F), with a cutoff value of 34.8% (80% sensitivity and 83% specificity). We also compared changes in RPC density in 27 SSOH patients and 23 age-matched patients with mild- or moderate-stage of glaucoma. There were no differences between the groups in age (normal: 39.5 ± 8.2, SSOH: 41.2 ± 18.0, p = 0.326), IOP (14.0 ± 2.2, 14.1 ± 2.5, p = 0.749) or axial length (24.6 ± 0.8, 25.1 ± 1.3, p = 0.088), but there was a significant difference in mean deviation (24.6 ± 0.8, 25.1 ± 1.3, p = 0.088). We found that RPC density in the SSOH patients was significantly higher in the inferior (p = 0.005, 0.725) and temporal (p < 0.001, 0.781) quadrants and lower in the nasal quadrant (p = 0.029, 0.688). There was no difference in overall RPC density (p = 0.391, AUC = 0.568) or RPC density in the superior quadrant (p = 0.268, 0.593), as shown in Table 3. Discussion RPC density in the SSOH patients was not significantly different overall or in the superior quadrant but was significantly higher in the inferior and temporal quadrants and lower in the nasal quadrant. These differences in RPC density are understandable, because the area of the nerve fiber layer that is vulnerable to damage is different in SSOH and glaucoma. Our finding also shows that RPC density may be valuable for assessing damage in SSOH. Theoretically, cpRNFLT damage in SSOH is developmental, while damage in glaucoma is secondary. However, the diagnostic power of Page 4/9 RPC density to differentiate SSOH and glaucoma was not strong, and further study is needed to determine the relationship between cpRNFLT and vasculature. Limitations of this cross-sectional study include a small size. This study is the first to investigate changes in RPC density in SSOH. Generally, sample size calculations are hard to perform in such exploratory studies. Another limitation was the use of a specific manufacturer-dependent method for the evaluation of RPC density, even though there are currently several differing methods of calculating RPC density with devices from different companies. We excluded patients with high myopia and applied a method (8) to compensate for optical magnification of the eye, and excluded the major central retinal vessels by using image processing with a Laplacian of Gaussian filter with noise reduction. Thus, we are the first to demonstrate the potential of RPC density measurement for SSOH diagnosis. Conclusion We found that RPC density decreased in SSOH, and that changes in microcirculation were indicative of SSOH. However, it remains unclear whether these changes are congenital or secondary to retinal nerve fiber layer degeneration. In addition to cpRNFLT and the extension of the RPE over the nasal disc margin (4), OCT-A may be a new, non-invasive, objective instrument for SSOH diagnosis. Acknowledgements: The authors thank Mr. Tim Hilts for editing this manuscript. Co-author TK is employed by Topcon Corporation, a commercial company. The funders played no role in the design or conduct of the study, nor in the decision to submit the manuscript for publication. This retrospective study was approved by the institutional review board of Tohoku University Graduate School of Medicine (study 2017-1-290). Declarations Ethics approval and consent to participate: This retrospective study was approved by the institutional review board of Tohoku University Graduate School of Medicine (study 2019-1-165). Written patient’s informed consent was not required because the ethics committee allowed us to obtain data retrospectively from patients’ records under insured medical treatment and approved to carry out by opt-out method. No patient was individually identified in this study. Consent for publication: Not applicable. Availability of data and materials Not applicable. Competing interests: The authors declare that they have no competing interests except for co-author TK. Co-author TK is employed by Topcon Corporation, a commercial company but has no conflicts of interest associated with the content of this article. Page 5/9 Funding: This manuscript was supported in part by a JST grant from JSPS KAKENHI Grants-in-Aid for Scientific Research (B) (T.N. 26293372), by the JST Center for Revitalization Promotion and KAKENHI Grants-in-Aid for young scientists (B) (K.O. 17K16957) and by the Public Trust Suda Memorial Fund for Glaucoma Research. This manuscript was supported in part by a JST grant from JSPS KAKENHI Grants-in-Aid for Scientific Research (B) (T.N. 26293372), by the JST Center for Revitalization Promotion and KAKENHI Grants-in-Aid for young scientists (B) (K.O. 17K16957) and by the Public Trust Suda Memorial Fund for Glaucoma Research. Authors’ contributions: MA and KO contributed to data collection. KO and TN contributed to manuscript writing. TK contributed to technical support. References Comparison of RPC density in normal and SSOH subjects 　 RPC (%) 　 Normal (n = 40) SSOH (n = 40) p Average 46.1±3.3 39.6±4.7 < 0.001* Superior 50.3±7.1 36.6±7.1 < 0.001* Nasal 37.6±5.7 31.5±8.1 < 0.001* Inferior 50.7±5.8 44.0±10.2 < 0.001* Temporal 45.9±5.6 46.3±6.6 0.756 Superior 11 56.6±8.4 49.6±10.7 0.002 12 43.0±11.7 30.5±13.3 < 0.001* 1 51.5±11.2 30.3±12.9 < 0.001* Nasal 2 42.7±9.5 33.6±13.7 < 0.001* 3 31.6±8.1 29.0±10.9 0.238 4 38.7±7.8 32.0±11.7 0.004 Inferior 5 45.9±9.9 36.8±12.5 < 0.001* 6 44.9±8.5 41.3±13.6 0.158 7 61.3±9.0 54.9±11.3 0.007** Temporal 8 46.5±8.3 47.2±8.6 0.700 9 38.2±7.3 41.9±8.1 0.039* 10 52.7±8.9 50.0±10.7 0.226 References 1. Iwase A, Suzuki Y, Araie M, Yamamoto T, Abe H, Shirato S, et al. The prevalence of primary open- angle glaucoma in Japanese: the Tajimi Study. Ophthalmology. 2004;111(9):1641-8. 1. Iwase A, Suzuki Y, Araie M, Yamamoto T, Abe H, Shirato S, et al. The prevalence of primary open- angle glaucoma in Japanese: the Tajimi Study. Ophthalmology. 2004;111(9):1641-8. 2. Yamamoto T, Sato M, Iwase A. Superior segmental optic hypoplasia found in Tajimi Eye Health Care Project participants. Jpn J Ophthalmol. 2004;48(6):578-83. 2. Yamamoto T, Sato M, Iwase A. Superior segmental optic hypoplasia found in Tajimi Eye Health Care Project participants. Jpn J Ophthalmol. 2004;48(6):578-83. 3. Yagasaki A, Sawada A, Manabe Y, Yamamoto T. Clinical features of superior segmental optic hypoplasia: hospital-based study. Jpn J Ophthalmol. 2019;63(1):34-9. 3. Yagasaki A, Sawada A, Manabe Y, Yamamoto T. Clinical features of superior segmental optic hypoplasia: hospital-based study. Jpn J Ophthalmol. 2019;63(1):34-9. 4. Hayashi K, Tomidokoro A, Konno S, Mayama C, Aihara M, Araie M. Evaluation of optic nerve head configurations of superior segmental optic hypoplasia by spectral-domain optical coherence tomography. Br J Ophthalmol. 2010;94(6):768-72. 4. Hayashi K, Tomidokoro A, Konno S, Mayama C, Aihara M, Araie M. Evaluation of optic nerve head configurations of superior segmental optic hypoplasia by spectral-domain optical coherence tomography. Br J Ophthalmol. 2010;94(6):768-72. 5. Aizawa N, Kunikata H, Omodaka K, Nakazawa T. Optic disc microcirculation in superior segmental optic hypoplasia assessed with laser speckle flowgraphy. Clin Exp Ophthalmol. 2014;42(7):702-4. 5. Aizawa N, Kunikata H, Omodaka K, Nakazawa T. Optic disc microcirculation in superior segmental optic hypoplasia assessed with laser speckle flowgraphy. Clin Exp Ophthalmol. 2014;42(7):702-4. 6. Kiyota N, Kunikata H, Shiga Y, Omodaka K, Nakazawa T. Relationship between laser speckle flowgraphy and optical coherence tomography angiography measurements of ocular microcirculation. Graefes Arch Clin Exp Ophthalmol. 2017;255(8):1633-42. 6. Kiyota N, Kunikata H, Shiga Y, Omodaka K, Nakazawa T. Relationship between laser speckle flowgraphy and optical coherence tomography angiography measurements of ocular microcirculation. Graefes Arch Clin Exp Ophthalmol. 2017;255(8):1633-42. 7. Kiyota N, Kunikata H, Shiga Y, Omodaka K, Nakazawa T. Ocular microcirculation measurement with laser speckle flowgraphy and optical coherence tomography angiography in glaucoma. Acta Ophthalmol. 2018;96(4):e485-e92. 8. Iwase A, Sekine A, Suehiro J, Tanaka K, Kawasaki Y, Kawasaki R, et al. A New Method of Magnification Correction for Accurately Measuring Retinal Vessel Calibers From Fundus Photographs. Invest Ophthalmol Vis Sci. 2017;58(3):1858-64. Page 6/9 Page 6/9 Tables Table 1. Tables Table 1. Comparison of RPC density in normal and SSOH subjects RPC (%) 　 Normal (n = 40) SSOH (n = 40) p Average 46.1±3.3 39.6±4.7 < 0.001* Superior 50.3±7.1 36.6±7.1 < 0.001* Nasal 37.6±5.7 31.5±8.1 < 0.001* Inferior 50.7±5.8 44.0±10.2 < 0.001* Temporal 45.9±5.6 46.3±6.6 0.756 Superior 11 56.6±8.4 49.6±10.7 0.002 12 43.0±11.7 30.5±13.3 < 0.001* 1 51.5±11.2 30.3±12.9 < 0.001* Nasal 2 42.7±9.5 33.6±13.7 < 0.001* 3 31.6±8.1 29.0±10.9 0.238 4 38.7±7.8 32.0±11.7 0.004 Inferior 5 45.9±9.9 36.8±12.5 < 0.001* 6 44.9±8.5 41.3±13.6 0.158 7 61.3±9.0 54.9±11.3 0.007** Temporal 8 46.5±8.3 47.2±8.6 0.700 9 38.2±7.3 41.9±8.1 0.039* 10 52.7±8.9 50.0±10.7 0.226 Table 2. Diagnosis power of localized RPC density in normal and SSOH subjects. Table 2. Diagnosis power of localized RPC density in normal and SSOH subjects. Page 7/9 Page 7/9 AUC of RPC (%) Average 0.879 Superior 0.928 Nasal 0.744 Inferior 0.706 Temporal 0.520 Superior 11 0.692 12 0.769 1 0.896 Nasal 2 0.729 3 0.637 4 0.693 Inferior 5 0.709 6 0.614 7 0.669 Temporal 8 0.513 9 0.628 10 0.563 AUC of RPC (%) AUC of RPC (%) Table 3. Comparison of RPC density in SSOH and glaucoma subjects and the diagnostic power of regional RPC density for SSOH. RPC (%) SSOH (n = 27) Glaucoma (n = 23) P value AUC Average 57.3±11.1 54.7±10.4 0.391 0.568 Superior 56.8±15.7 62.3±18.7 0.268 0.592 Nasal 35.7±19.1 46.8±15.4 0.029* 0.688 Inferior 68.5±15.8 55.8±14.9 0.005 0.725 Temporal 68.7±11.0 55.0±13.0 < 0.001* 0.781 Table 3. Comparison of RPC density in SSOH and glaucoma subjects and the diagnostic power of regional RPC density for SSOH. Table 3. Comparison of RPC density in SSOH and glaucoma subjects and the diagnostic power of regional RPC density for SSOH. Table 3. Comparison of RPC density in SSOH and glaucoma subjects power of regional RPC density for SSOH. RPC (%) SSOH (n = 27) Glaucoma (n = 23) P value AUC Average 57.3±11.1 54.7±10.4 0.391 0.568 Superior 56.8±15.7 62.3±18.7 0.268 0.592 Nasal 35.7±19.1 46.8±15.4 0.029* 0.688 Inferior 68.5±15.8 55.8±14.9 0.005 0.725 Temporal 68.7±11.0 55.0±13.0 < 0.001* 0.781 Figures Page 8/9 Figure 1 Representative normal and SSOH subjects and the regional diagnostic power of RPC density for SSOH. A, B: Fundus photography. C, D: RPC density measured with OCT-A. A, C: Normal subject. B, D: SSOH subject. E, F: Grayscale map of the visual field obtained with the Humphrey field analyzer in the SSOH patient (E: baseline, F 3 years later). G Receiver operating characteristic (ROC) curve for superior-quadrant cpRNFLT. H: ROC curve for the 1 o’clock sector of cpRNFLT. Page 9/9
https://openalex.org/W4220855771	https://europepmc.org/articles/pmc8997088?pdf=render	English	null	ATRX-Deficient High-Grade Glioma Cells Exhibit Increased Sensitivity to RTK and PDGFR Inhibitors	Cancers	2,022	cc-by	10,589	cancers cancers Citation: Pladevall-Morera, D.; Castejón-Griñán, M.; Aguilera, P.; Gaardahl, K.; Ingham, A.; Brosnan-Cashman, J.A.; Meeker, A.K.; Lopez-Contreras, A.J. ATRX-Deﬁcient High-Grade Glioma Cells Exhibit Increased Sensitivity to RTK and PDGFR Inhibitors. Cancers 2022, 14, 1790. https://doi.org/ 10.3390/cancers14071790 Academic Editors: Ana María Zubiaga and Jone Mitxelena Received: 15 February 2022 Accepted: 28 March 2022 Published: 31 March 2022 Citation: Pladevall-Morera, D.; Castejón-Griñán, M.; Aguilera, P.; Gaardahl, K.; Ingham, A.; Brosnan-Cashman, J.A.; Meeker, A.K.; Lopez-Contreras, A.J. ATRX-Deﬁcient High-Grade Glioma Cells Exhibit Increased Sensitivity to RTK and PDGFR Inhibitors. Cancers 2022, 14, 1790. https://doi.org/ 10.3390/cancers14071790 Academic Editors: Ana María Zubiaga and Jone Mitxelena Received: 15 February 2022 Accepted: 28 March 2022 Published: 31 March 2022 Citation: Pladevall-Morera, D.; Castejón-Griñán, M.; Aguilera, P.; Gaardahl, K.; Ingham, A.; Brosnan-Cashman, J.A.; Meeker, A.K.; Lopez-Contreras, A.J. ATRX-Deﬁcient High-Grade Glioma Cells Exhibit Increased Sensitivity to RTK and PDGFR Inhibitors. Cancers 2022, 14, 1790. https://doi.org/ 10.3390/cancers14071790 Academic Editors: Ana María Zubiaga and Jone Mitxelena Received: 15 February 2022 Accepted: 28 March 2022 Published: 31 March 2022 Abstract: High-grade glioma, including anaplastic astrocytoma and glioblastoma (GBM) patients, have a poor prognosis due to the lack of effective treatments. Therefore, the development of new therapeutic strategies to treat these gliomas is urgently required. Given that high-grade gliomas frequently harbor mutations in the SNF2 family chromatin remodeler ATRX, we performed a screen to identify FDA-approved drugs that are toxic to ATRX-deﬁcient cells. Our ﬁndings reveal that multi-targeted receptor tyrosine kinase (RTK) and platelet-derived growth factor receptor (PDGFR) inhibitors cause higher cellular toxicity in high-grade glioma ATRX-deﬁcient cells. Furthermore, we demonstrate that a combinatorial treatment of RTKi with temozolomide (TMZ)–the current standard of care treatment for GBM patients–causes pronounced toxicity in ATRX-deﬁcient high-grade glioma cells. Our ﬁndings suggest that combinatorial treatments with TMZ and RTKi may increase the therapeutic window of opportunity in patients who suffer high-grade gliomas with ATRX mutations. Thus, we recommend incorporating the ATRX status into the analyses of clinical trials with RTKi and PDGFRi. 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. Keywords: glioblastoma; glioma; ATRX; RTKi; PDGFRi; drug screen 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/). David Pladevall-Morera 1,†, María Castejón-Griñán 1,2,†, Paula Aguilera 1,2, Karina Gaardahl 1, Andreas Ingham 1 , Jacqueline A. Brosnan-Cashman 3, Alan K. Meeker 3 and Andres J. Lopez-Contreras 1,2,* orera 1,†, María Castejón-Griñán 1,2,†, Paula Aguilera 1,2, Karina Gaardahl 1, , Jacqueline A. Brosnan-Cashman 3, Alan K. Meeker 3 and Andres J. Lopez-Contreras 1,2,* David Pladevall-Morera 1,†, María Castejón-Griñán 1,2,†, Paula Aguilera 1,2, Karina Gaardahl 1, Andreas Ingham 1 , Jacqueline A. Brosnan-Cashman 3, Alan K. Meeker 3 and Andres J. Lopez-Contreras 1,2,* 1 Department of Cellular and Molecular Medicine, DNRF Center for Chromosome Stability and Center for Healthy Aging, University of Copenhagen, 2200 Copenhagen, Denmark; davidpm@sund.ku.dk (D.P.-M.); maria.castejon@cabimer.es (M.C.-G.); paula.aguilera@cabimer.es (P.A.); kgaardahl@sund.ku.dk (K.G.); ingham@sund.ku.dk (A.I.) 1 Department of Cellular and Molecular Medicine, DNRF Center for Chromosome Stability and Center for Healthy Aging, University of Copenhagen, 2200 Copenhagen, Denmark; davidpm@sund.ku.dk (D.P.-M.); maria.castejon@cabimer.es (M.C.-G.); paula.aguilera@cabimer.es (P.A.); kgaardahl@sund.ku.dk (K.G.); ingham@sund.ku.dk (A.I.) 2 Centro Andaluz de Biología Molecular y Medicina Regenerativa (CABIMER), Consejo Superior de Investigaciones Cientíﬁcas (CSIC), Universidad de Sevilla, Universidad Pablo de Olavide, 41013 Seville, Spain 2 Centro Andaluz de Biología Molecular y Medicina Regenerativa (CABIMER), Consejo Superior de Investigaciones Cientíﬁcas (CSIC), Universidad de Sevilla, Universidad Pablo de Olavide, 41013 Seville, Spain 3 Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA; jackiebros@gmail.com (J.A.B.-C.); ameeker1@jhmi.edu (A.K.M.) g ( ) p 3 Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA; jackiebros@gmail.com (J.A.B.-C.); ameeker1@jhmi.edu (A.K.M.) j g * Correspondence: andres.lopez@cabimer.es † These authors contributed equally to this work. Simple Summary: High-grade glioma has a poor prognosis and new effective strategies to treat this aggressive form of cancer are highly needed. We have conducted a drug screen searching for compounds toxic to ATRX-deﬁcient cells, a frequent scenario in cancer, and particularly in high-grade gliomas. We have identiﬁed that ATRX-deﬁcient glioma cells are sensitive to several multi-targeted receptor tyrosine kinase and speciﬁc platelet-derived growth factor receptor inhibitors, some of which are currently under study in clinical trials. In view of our results, we believe that taking into consideration the presence/absence of ATRX mutations could provide valuable information to interpret the results of those clinical trials. cancers cancers 1. Introduction Inherited mutations in the SNF2 family chromatin remodeler ATRX cause the alpha thalassemia mental retardation X- linked syndrome, which this protein is named after [1]. ATRX is also a bona ﬁde tumor suppressor gene frequently mutated in several cancer types [2]. ATRX together with DAXX (Death domain-associated protein 6) acts as a histone chaperone to deposit the histone variant H3.3 in a replication-independent manner at heterochromatic regions of the genome [3,4]. ATRX is involved in the maintenance of https://www.mdpi.com/journal/cancers Cancers 2022, 14, 1790. https://doi.org/10.3390/cancers14071790 Cancers 2022, 14, 1790 2 of 17 2 of 17 genome stability in cells by several mechanisms. For instance, the deposition of histone H3.3 by ATRX/DAXX is important for timely and accurate double-strand break (DSB) repair by homologous recombination [5]. Furthermore, ATRX suppresses genome instability at telomeres and common fragile sites (CFS), and it has been shown to suppress R-loop formation and promote the resolution of G-quadruplexes [6–8]. Thus, loss of ATRX leads to increased genome instability manifested as elevated DSBs, increased CFS expression, and micronuclei formation [5,7,8], which can inﬂuence tumor progression and therapy response in cancer patients. ATRX mutations are also often associated with the Alternative Lengthening of Telomeres (ALT) in cancers [9], although the loss of ATRX alone is not sufﬁcient to trigger the ALT phenotype [9,10]. In addition, ATRX is a critical regulator of therapy-induced senescence, since ATRX-depleted cells are deﬁcient in triggering the cell senescence program [11]. p g ATRX mutations occur in a variety of human cancers such as hepatocellular carci- noma, pancreatic neuroendocrine tumors, and gliomas [12–14]. Often, these mutations are truncating mutations that lead to a loss of the functional protein [15]. ATRX mutations are associated with PDGFR ampliﬁcation [16,17] and with mutations in the TP53 and IDH1 genes in cancer cells. On the contrary, ATRX and DAXX mutations are mutually exclusive in glioma and other cancers [2,18]. Interestingly, in the rare pancreatic neuroendocrine cancer, ATRX/DAXX inactivating mutations are frequently associated with mutations of genes in the mTOR pathway [13,19]. Furthermore, it has been described that inactivating ATRX mutations are mutually exclusive with MYCN ampliﬁcation in neuroblastoma [20]. High-grade gliomas often harbor ATRX mutations frequently co-occurring with mutations in the TP53 and IDH1 genes in this type of malignancy (reviewed in [21]). Glioblastoma Multiforme (GBM) is the most prevalent malignant type of glioma in adults [21]. 1. Introduction The over- all survival rate of GBM is less than one year from the time of diagnosis. Currently, the ﬁrst line of treatment for GBM patients, if tumor resection is not possible, is the administration of the DNA alkylating agent temozolomide (TMZ) and radiotherapy, which increases patient survival by an average of 2.5 months [22]. Therefore, the low overall survival rate and the current lack of effective treatments highlight the need for new and more effective drugs for GBM treatment. In line with this, a recent study reports that one promising possibility for targeting ATRX-deﬁcient cancers is the use of WEE1 inhibitors [23]. However, WEE1 inhibitors are not clinically approved. In this study, we aimed to identify drugs that exhibit synthetic lethality with ATRX loss and, thus, exploit ATRX deﬁciency for the treatment of high-grade glioma patients har- boring ATRX mutations. Therefore, we compared the toxicity proﬁles of ATRX-proﬁcient versus ATRX-deﬁcient cells exposed to a compound library of 1496 FDA-approved drugs. We demonstrate that four multi-targeted receptor tyrosine kinase inhibitors (RTKi) and a speciﬁc inhibitor targeting the platelet-derived growth factor receptor (PDGFRi) exhibit higher toxicity in ATRX-deﬁcient high-grade glioma cells. Furthermore, we show that ATRX-deﬁcient cells are highly sensitive to combinatorial treatments of TMZ and RTKi or PDGFRi. The RTKi tested in this study (nintedanib, sunitinib, pazopanib, and sorafenib) are currently being assessed in multiple clinical trials for GBM treatment and other cancer types. Our ﬁndings suggest that the status of ATRX should be considered to stratify patients when evaluating the efﬁciency of RTKi in those clinical trials. 2.3. Western Blot Cells were lysed in RIPA buffer (Sigma) supplemented with complete protease in- hibitor cocktail tablet (Roche, Basel, Switzerland), 5 mM β-glycerolphosphate (Sigma), 5 mM sodium ﬂuoride (Sigma) and 1 mM sodium orthovanadate (Sigma). Then, lysis extracts were centrifuged at 13,000 rpm for 15 min at 4 ◦C and supernatants were collected for protein measurement. Protein quantiﬁcation of the cell lysates was performed with the DC Protein Assay kit (Bio-Rad, Richmond, CA, USA) according to supplier’s instructions. Cell lysates containing 30 µg protein were boiled (4:1 ratio) in NuPAGE™LDS Sample Buffer 4x (Invitrogen) with 10 mM DTT for 15 min at 70 ◦C. An equal amount of each sample was loaded into each well of 4–12% NuPage Bis–Tris gels (Invitrogen) and elec- trophoresis was performed using a dissociation running buffer MOPS (Thermo Fisher, Waltham, MA, USA), at 180 V for around 1 h. Next, proteins were transferred to a PVDF membrane by electrophoresis at 350 mAmp for 1 h 20 min at 4 ◦C, subsequently blocked with 5% milk or BSA in PBS supplemented with 0.1% Tween-20 (PBST) for 1 h at room temperature (RT) and thereafter incubated with primary antibodies overnight at 4 ◦C. The primary antibodies used in this study were mouse monoclonal ATRX (Santa Cruz, Dallas, TX, USA, sc-55584), mouse monoclonal p53 (Santa Cruz, sc-126), rabbit monoclonal p21 (Cell Signaling, Danvers, MA, USA, 2947), mouse vinculin (Sigma, V9131) and mouse monoclonal β-actin (Sigma, A5441). On the following day, membranes were washed in PBST, incubated with the appropriate HRP-conjugated secondary antibodies for 1 h at RT, then washed again and stained with the chemiluminescent substrate AmerSham™ECL™ Western Blotting Detection Reagents (GE Healthcare, Chicago, IL, USA). Images were acquired on an AmerSham™Imager 600 (GE Healthcare Life Sciences). Goat anti-rabbit immunoglobulin G (IgG) (Sigma, A6667) and goat anti-mouse IgG (Sigma, A4416) were used as secondary antibodies. 2.1. Cell Culture HeLa cells were grown in DMEM (Dulbecco’s modiﬁed Eagle’s medium) supple- mented with 10% of fetal bovine serum (FBS) (Life Technologies, Rocky Hill, NJ, USA) and 1% of penicillin/streptomycin (Life Technologies). The glioma cell lines MOG-G-UVW (grade III human adult anaplastic astrocytoma) and SF188 (grade IV human pediatric GBM) were grown in DMEM/F12 (DMEM nutrient mixture F-12) and U-251 (grade IV human adult GBM) cells in RPMI (Roswell Park Memorial Institute) medium. DMEM/F12 and RPMI mediums were supplemented with 10% FBS, 1% penicillin/streptomycin, 1% ampho- Cancers 2022, 14, 1790 3 of 17 3 of 17 tericin B (Sigma, St. Louis, MO, USA), 10 µg/mL gentamicin (Sigma), and 5 µg/mL Plas- mocin (Invitrogen, Carlsbad, CA, USA). Patient-derived glioma cells (all grade IV human adult GBM) were grown in Neurobasal and DMEM/F12 glutamax (ratio 1:1) supplemented with 1% of penicillin/streptomycin, N2, B27 (Life Technologies), EGF (10 ng/mL) and bFGF (10 ng/mL) (Peprotech, Rocky Hill, NJ, USA). These cells from the human glioblastoma cell culture (HGCC) collection were grown as adherent monolayers in laminin-coated dishes. All the cell lines were maintained at 37 ◦C in a 5% CO2 incubator. 2.2. Generation of ATRX KO Cells ATRX KO HeLa cells were generated by CRISPR/Cas9 following Ran et al., 2013 protocol [24]. The guide RNA (gRNA) sequences that were used to generate the plas- mid were as follows: 5′-CACCGCAGGATCGTCACGATCAAAG-3′ (forward) and 5′- AAACCTTTGATCGTGACGATCCTGC-3′ (reverse). The gRNA was cloned into pSpCas9(BB)-2A-GFP vector (#48138, addgene) for co-expression with Cas9. HeLa cells were seeded in 6-well dishes at a density of 70% a day prior transfection. The cells were transfected with 1.25 µg of the sequence-veriﬁed gRNA cloned into pSpCas9(BB)-2A-GFP plasmid using Fugene (Promega, Madison, WI, USA). The day after, GFP-positive single cells were sorted into a 96-well plate using a FACS Aria-II cell sorter. Clonal cell lines were expanded for 2 weeks and screened by Western blot analysis. 2.7. Lentivirus Synthesis HEK-293 T cells were used for the synthesis of third generation lentiviruses containing a tdTomato vector or a GFP vector. Then, 6 × 106 of cells were reverse transfected using lipofectamine 2000 with 10 µg of the plasmid of interest and the plasmids coding for the lentivirus packaging components (6.5 µg pMMDLRRE, 2.5 µg PRSVREV and 3.5 µg PMDGVSVG) (#12251, #12253 and #12259, respectively, addgene). Forty-eight hours post- transfection, the viruses were ﬁltered (0.25 µm) and collected. 2.6. Immunoﬂuorescence For all immunostaining, cells were seeded on µCLEAR 96-well plates. Then, cells that were treated as required were ﬁxed with 4% formaldehyde (VWR Chemicals, Radnor, PA, USA) for 15 min at RT and permeabilized with 0.5% Triton X-100 (v/v) in PBS for 10 min. Cells were washed twice with PBST and blocked with 3% BSA (Sigma) in PBST for 30 min and then labeled with primary antibody at 4 ◦C overnight. The primary antibodies used in this study were ATRX (Santa Cruz, sc-55584) and γH2AX (MilliporeSigma, St. Louis, MO, USA, 05-636). Next, wells were washed in PBST and ﬂuorescence-tagged secondary antibody (Alexa Fluor™Goat Anti-mouse IgG 488 (Invitrogen)) that was added for 2 h at RT in the dark. DAPI was used for nuclear staining. The DNA replication rate was determined by EdU incorporation using Click-iT technology following the manufacturer’s instructions (Life Technologies). Brieﬂy, EdU was added 30 min prior to ﬁxing the cells and click chemistry was performed prior to blocking. A dilution of 1 mM ascorbic acid was prepared fresh, and the click-it reaction mix was performed by mixing PBS, CuSO4, Azide 647 and ascorbic acid in the given order, which was then added to each well for 1 h at RT in the dark. Cellular viability = n◦of ATRX WT cells (drug)/n◦of ATRX WT cells (DMSO) n◦of ATRX KO cells (drug)/n◦of ATRX KO cells (DMSO) Cellular viability = n◦of ATRX WT cells (drug)/n◦of ATRX WT cells (DMSO) n◦of ATRX KO cells (drug)/n◦of ATRX KO cells (DMSO) Cellular viability = n◦of ATRX WT cells (drug)/n◦of ATRX WT cells (DMSO) n◦of ATRX KO cells (drug)/n◦of ATRX KO cells (DMSO) Drugs causing the same effect in the WT and KO clones scored a ratio of 1; drugs with a higher toxicity for the KO clones compared to the WT clones scored >1; and drugs with a higher toxicity for the WT clones compared to the KO clones scored <1. 2.4. High-Throughput Microscopy Images were acquired using the ScanR acquisition software 3.2.0 (r4066) ×64 (Olym- pus, Johann-Krane-Weg, Munich, Germany) controlling a motorized Olympus IX-81 wide- ﬁeld microscope. Olympus UPLSAPO 10x/0.4 NA objective was used. Single plane images corresponding to Z positions of maximal DAPI signal were acquired. At least nine images were acquired per well. Each ﬂuorophore emission was collected separately and both images were acquired exactly in the same space conditions. The number of cells and mean signal intensity were analyzed and quantiﬁed with the ScanR analysis software (Olympus). Cancers 2022, 14, 1790 4 of 17 2.5. Analysis of the Screen Data 2.5. Analysis of the Screen Data ATRX WT and ATRX KO HeLa clones were mixed. The mixed population was seeded on µCLEAR 96-well plates (Greiner Bio-One, Frickenhausen, Germany) at a density of 4000 cells/well. After 24 h, cells were treated with the compounds of the FDA-approved APExBIO drug library (L1021), with a ﬁnal concentration of 10 µM and for a period of 48 h by adding one compound per well. DMSO was used as control. The number of cells from each condition was analyzed with ScanR analysis software 3.2.0 (r4066) ×64 (Olympus). Brieﬂy, cell populations were selected according to their emitted ﬂuorescence corresponding to ATRX staining, as WT (ﬂuorescent, ATRX positive) and KO (non-ﬂuorescent, ATRX negative) from each condition for the primary screen, and according to their tdTomato (ATRX positive) or GFP signal (ATRX negative) for the secondary screen. DAPI was used to identify the total number of cells. Drug-treated wells with <190 WT alive cells were excluded from further analysis. The cell viability was assessed with the following formula: 2.11. Dose–Response Curves U-251 (EV and ATRX KO) high-grade glioma cells (4000 cells/well) or patient-derived GBM cells (8000 cells/well) were seeded in µCLEAR 96-well plates. On the following day, the medium was replaced with fresh medium containing increasing doses of the desired drug. After 48 h cells were ﬁxed with 4% formaldehyde (VWR Chemicals) for 15 min at RT and stained for DAPI. Images were acquired with ScanR acquisition software with a 10×/0.4 NA objective and quantiﬁcation of number of cells was performed with ScanR analysis software. The log of each known concentration in the dilution series (x-axis) was plotted against the number of cells (expressed as percentage) for that concentration (y-axis). The resulting dose–response curves were ﬁt using GraphPad Prism. Drug’s potency (IC50 value) was determined by non-linear regression analysis of the resulting dose–response curve. Each dose–response curve was performed in biological triplicates (with technical triplicates for each experiment) and a representative replicate is shown in the manuscript. 2.10. Cell Viability Analysis All the cell lines used for the cell viability analysis were seeded with a density of 4000 cells/well in µCLEAR 96-well plates. The following day, drugs were added in technical triplicates with the desired concentration and time. Subsequently, cells were ﬁxed with 4% formaldehyde (VWR Chemicals) for 15 min at RT and stained for DAPI. Images were acquired with ScanR acquisition software with a 10×/0.4 NA objective and quantiﬁcation of the number of cells was performed with ScanR analysis software. DMSO- treated cells were used as control. Cell viability was assessed by dividing the number of drug-treated cells by the number of DMSO-treated cells. 2.9. Drug Inhibitors Ibrutinib (A3001), niclosamide (B2283), pazopanib (A3022), nintedanib (A8252), suni- tinib (B1045), sorafenib (A3009), temozolomide (B1399) and CP-673451 (B2173) were pur- chased from APExBIO. 2.8. Cell Infection Cells were plated in 6-well dishes a day prior to infection. The cells were infected with different amounts of lentivirus containing the vector of interest and 10 µg/mL of polybrene to facilitate the infection. HeLa ATRX WT clones 1 and 2 were infected with a medium and high titer of lentivirus containing tdTomato plasmid, respectively; and HeLa ATRX KO clones 3, 4 and 5 were infected with a low, medium or high titer of lentivirus containing GFP plasmid, respectively. Furthermore, the U-251 (EV and ATRX KO) cells were infected with lentivirus containing short hairpin RNA (shRNA) targeting TP53 (kindly provided by Cancers 2022, 14, 1790 5 of 17 5 of 17 Prof. Oscar Fernandez-Capetillo) or the shRNA control of pLKO.1 (#8453, addgene). The following day, the infection efﬁciency was assessed by a ﬂuorescent microscope (Olympus). 2.9. Drug Inhibitors 2.12. Colony Formation Assay U-251 (EV and ATRX KO) high-grade glioma cells were seeded in 10 cm plates at a density of 2000 and 3500 cells, respectively. Four days after seeding, the medium was replaced with fresh-medium containing sunitinib at 1.25 µM, 2.5 µM, 5 µM or vehicle. After 6 days of treatment, when colonies were visible, cells were washed with PBS, ﬁxed and stained with a solution containing 20% absolute ethanol and 0.5% crystal violet in water for 1 h at RT. Cells were washed twice with water and after allowing to air dry, images were acquired and the number of colonies was quantiﬁed using the Analyze Particles plugin of ImageJ software. 2.13. Statistical Analysis Statistical analyses were performed using GraphPad Prism 8. The signiﬁcance was determined by either unpaired t-test or Fisher’s exact test. The p-values are indicated in each graph and in the ﬁgure legends. 3. Results 3.1. Identiﬁcation of FDA-Approved Compounds That Selectively Kill ATRX-Deﬁcient Cells To create isogenic cell lines that either possess or lack ATRX, we generated HeLa ATRX knockout (KO) cells using CRISPR/Cas9. We designed a gRNA targeting the fourth coding exon of the ATRX gene and obtained three independent ATRX KO clones. Two HeLa ATRX WT clones were obtained in parallel and used as a control for further experiments (Figures 1A and S1A). DNA replication was assessed by EdU (5-Ethynyl-2′-deoxyuridine) Cancers 2022, 14, 1790 6 of 17 6 of 17 incorporation. All the clones proliferated at similar rates, although the number of replicating cells slightly decreased in two ATRX KO clones as shown by the percentage of EdU-positive cells (Figure S1B). We also assessed the levels of H2AX phosphorylation (γH2AX) as a marker of DNA damage. The three ATRX KO clones exhibited higher levels of γH2AX as compared to the WT clones indicating that the loss of ATRX leads to increased DSBs formation and genome instability, as reported previously (Figure S1C) [8,25]. Recent studies have revealed that the use of CRISPR to generate KO cell lines often results in mutations in TP53 [26,27]. Therefore, we assessed the p53 status in all of the HeLa clones. We observed that all of the ATRX WT and ATRX KO cell lines used in this study were p53 proﬁcient and were able to activate p21 upon DNA damage (Figure S1D). Figure 1. FDA-Approved drug screen identiﬁes compounds synthetically lethal with ATRX-deﬁciency. (A) Immunoblotting of HeLa clones generated by CRISPR. (B) Drug screen ﬂowchart. (C) Primary drug screen. Cell viability (WT/KO) after 48 h of drug treatment. Each point represents one drug. Red dots indicate drugs with a 2-fold higher lethality effect in the ATRX KO clones compared to WT clones. Green dots indicate compounds that induce a 2.5-fold higher toxicity in ATRX WT clones compared to ATRX KO clones. (D) Secondary drug screen. Cell viability (WT/KO) after 48 h of drug treatment of the 29 top hits derived from the primary screen. Each point represents one drug. Data shown correspond to technical triplicates. Means and SDs are indicated. Figure 1. FDA-Approved drug screen identiﬁes compounds synthetically lethal with ATRX-deﬁciency. (A) Immunoblotting of HeLa clones generated by CRISPR. (B) Drug screen ﬂowchart. (C) Primary drug screen. Cell viability (WT/KO) after 48 h of drug treatment. Each point represents one drug. 3.2. ATRX-Deﬁcient Cells Show Increased Sensitivity to BTK, STAT3, and RTK Inhibitors To validate the effect of the 29 compounds identiﬁed in the initial screen, we performed a secondary screen whereby the two ATRX WT and three ATRX KO HeLa clones were labeled with different ﬂuorescent protein markers. The two WT clones (clones 1 and 2) were infected with different titers of lentivirus expressing ﬂuorescent tdTomato protein and the three ATRX KO clones (clones 3, 4 and 5) with different titers of lentivirus expressing GFP (Figure S1F). This strategy allowed us to distinguish the individual tdTomato-ATRX WT and GFP-ATRX KO clones within the mixed population (Figure S1G). Importantly, replication among the clones was not affected after the lentivirus infection and the GFP-ATRX KO clones showed increased levels of γH2AX as compared to their tdTomato-ATRX WT counterparts, as seen for the unlabeled clones used for the initial screen (Figure S1H,I). We employed a similar strategy for the secondary screen using a mixed population containing the ﬁve labelled clones; cells were treated with 10 µM of the 29 selected compounds for 48 h. Assessment of cell viability using high-content microscopy revealed that seven out of the twenty-nine analyzed compounds showed a toxicity of at least 1.5-fold higher in the ATRX KO clones as compared to the ATRX WT clones (Figure 1D). Among these seven validated hits, we further characterized the inhibitors targeting BTK (ibrutinib), STAT3 (niclosamide), and RTK (pazopanib), as previous studies have indicated the potential of these drugs to treat high-grade glioma [28–31]. To further validate the effect of ibrutinib, niclosamide, and pazopanib, we treated the tdTomato ATRX WT and GFP-ATRX KO HeLa clones individually at 10 µM and assessed cell viability at different time points (24 h, 48 h, and 72 h). The three ATRX- deﬁcient HeLa clones were more sensitive to ibrutinib, niclosamide, and pazopanib (Figures 2A,B and S2A). The higher toxicity of the drugs for ATRX-deﬁcient clones was observed at 48 h and 72 h of drug treatment at 10 µM (Figures 2A and S2A). Finally, we tested three additional FDA-approved RTKi (nintedanib, sunitinib, and sorafenib), all of which target PDGFR. Nintedanib and sunitinib caused higher cell toxicity in all the ATRX KO HeLa clones compared to the HeLa ATRX WT counterparts when used at 6 µM for 48 h (Figure S2B). 3.2. ATRX-Deﬁcient Cells Show Increased Sensitivity to BTK, STAT3, and RTK Inhibitors In addition, treatment with sorafenib in the same conditions induced higher toxicity in two out of the three ATRX KO HeLa clones compared to the HeLa WT cells, although not all the conditions were statistically signiﬁcant. Taken together, our data indicate that BTKi, STAT3i, and multitarget RTKi are toxic to ATRX-deﬁcient cells. 3. Results ATRX-Deﬁcient Cells Show Increased Sensitivity to BTK, STAT3, and RTK Inhibitors 3. Results Red dots indicate drugs with a 2-fold higher lethality effect in the ATRX KO clones compared to WT clones. Green dots indicate compounds that induce a 2.5-fold higher toxicity in ATRX WT clones compared to ATRX KO clones. (D) Secondary drug screen. Cell viability (WT/KO) after 48 h of drug treatment of the 29 top hits derived from the primary screen. Each point represents one drug. Data shown correspond to technical triplicates. Means and SDs are indicated. Red dots indicate drugs with a 2-fold higher lethality effect in the ATRX KO clones compared to WT clones. Green dots indicate compounds that induce a 2.5-fold higher toxicity in ATRX WT clones compared to ATRX KO clones. (D) Secondary drug screen. Cell viability (WT/KO) after 48 h of drug treatment of the 29 top hits derived from the primary screen. Each point represents one drug. Data shown correspond to technical triplicates. Means and SDs are indicated. To screen for drugs that are synthetic lethal with ATRX loss, we used the FDA- approved APExBIO drug library, which contains 1496 compounds. We performed the screen using a mixed population of cells (mixed in 1 to 1 ratio) containing the two ATRX WT and three ATRX KO clones that could be discriminated by ATRX immunoﬂuorescence (IF) (Figure S1E). Cells were treated with the compounds at 10 µM for 48 h and cell viability was assessed by DAPI staining combined with high-content microscopy (Figure 1B). We evaluated the effect of the drugs on the viability of WT and ATRX KO clones by calculating the ratio of the number of WT cells by the number of KO cells compared to their DMSO Cancers 2022, 14, 1790 7 of 17 7 of 17 control cells (Figure 1B,C). From this screen, we identiﬁed 29 compounds (red-colored drugs in Figure 1C and Table S1) that caused at least a 2-fold higher toxicity to the KO cells compared to the WT cells. In addition, we also identiﬁed 37 compounds that induced at least a 2.5-fold higher lethality to the ATRX WT cells compared to the ATRX KO cells (green-colored drugs in Figure 1C and Table S1). However, we chose to focus on the drugs that trigger death in ATRX-deﬁcient cells in this study, and further analyses should be performed to validate and characterize the compounds that appear to be less toxic for ATRX KO cells. 3.2. 3.3. ATRX KO High-Grade Glioma Cells Are Sensitive to RTK and PDGFR Inhibitors Given the relevance and frequency of ATRX mutations in high-grade glioma patients, we sought to validate the effect of BTK, STAT3, and RTK inhibitors in glioma cell lines derived from adult malignant glioma (MOG-G-UVW and U-251) and pediatric malignant glioma (SF188) patients. We used CRISPR/Cas9 engineered ATRX KO high-grade glioma cell lines and empty vector (EV) control cell lines (Figure S3A,B) that were generated in a previous study [10]. The ATRX KO malignant glioma cell lines incorporated EdU similar to the ATRX WT cell lines, indicating that ATRX KO cells replicate and proliferate at normal rates (Figure S3C), which is important to assess the toxicity caused by the drug treatments. However, unlike the HeLa ATRX KO clones, the ATRX KO glioma cell lines used in this study showed similar levels of γH2AX as compared to the EV cells (Figure S3D). Cancers 2022, 14, 1790 8 of 17 Figure 2. ATRX-deﬁcient HeLa cells are susceptible to BTK, STAT3 and RTK inhibitors. (A) Cell viability of tdTomato-ATRX WT and GFP-ATRX KO HeLa clones after 72 h of treatment with the indicated drugs and concentrations compared to DMSO controls. Data shown correspond to technical triplicates. Means and SEMs are shown. Statistics for signiﬁcant conditions are shown. Signiﬁcance was assessed by unpaired t-test. * p ≤0.05; p ≤0.01; * p ≤0.001; **** p ≤0.0001. (B) Representative images of clone 1 (ATRX WT) and clone 3 (ATRX KO) after 72 h of treatment with either DMSO or the indicated drugs at a concentration of 10 µM. Figure 2. ATRX-deﬁcient HeLa cells are susceptible to BTK, STAT3 and RTK inhibitors. (A) Cell viability of tdTomato-ATRX WT and GFP-ATRX KO HeLa clones after 72 h of treatment with the indicated drugs and concentrations compared to DMSO controls. Data shown correspond to technical triplicates. Means and SEMs are shown. Statistics for signiﬁcant conditions are shown. Signiﬁcance was assessed by unpaired t-test. * p ≤0.05; p ≤0.01; * p ≤0.001; ** p ≤0.0001. (B) Representative images of clone 1 (ATRX WT) and clone 3 (ATRX KO) after 72 h of treatment with either DMSO or the indicated drugs at a concentration of 10 µM. We treated MOG-G-UVW, U-251, and SF-188 (EV and ATRX KO) high-grade glioma cell lines with different concentrations (5 µM and 10 µM) of BTK, STAT3, and RTK inhibitors for 48 h in order to assess cell viability. 3.3. ATRX KO High-Grade Glioma Cells Are Sensitive to RTK and PDGFR Inhibitors Sunitinib treatment signiﬁcantly reduced clonogenic survival of U-251 ATRX KO cells compared to WT cells (Figure 4A,B),further validating the toxicity of RTKi for ATRX-deﬁcient high-grade glioma cells. Furthermore, we tested the effect of RTKi by performing clonogenic assays. Sunitinib treatment signiﬁcantly reduced clonogenic survival of U-251 ATRX KO cells compared to WT cells (Figure 4A,B),further validating the toxicity of RTKi for ATRX-deﬁcient high-grade glioma cells. Figure 4. ATRX-deﬁcient high-grade glioma cells are sensitive to combinatorial treatments of TMZ and sunitinib. (A,B) Clonogenic assays of U-251 (EV and ATRX KO) high-grade glioma cells exposed to the indicated doses of sunitinib. Colonies were quantiﬁed using Analyze Particles plugin of ImageJ software. Data shown correspond to biological triplicates. Means and SDs are indicated. Statistics for signiﬁcant conditions are shown. Signiﬁcance was assessed by unpaired t-test. p ≤0.01. (C) Cell viability of U-251 (EV and ATRX KO) high-grade glioma cells after 48 h of treatment with TMZ and/or sunitinib at the indicated concentrations compared to DMSO controls. Data shown correspond to four independent experiments. Mean and SEMs are indicated. Statistics for signiﬁcant conditions are shown. Signiﬁcance was assessed by unpaired t-test. * p ≤0.05; ** p ≤0.0001. (D) Representative images of U-251 (EV and ATRX KO) high-grade glioma cells after 48 h of treatment with either DMSO or the indicated drugs and concentrations. Figure 4. ATRX-deﬁcient high-grade glioma cells are sensitive to combinatorial treatments of TMZ and sunitinib. (A,B) Clonogenic assays of U-251 (EV and ATRX KO) high-grade glioma cells exposed to the indicated doses of sunitinib. Colonies were quantiﬁed using Analyze Particles plugin of ImageJ software. Data shown correspond to biological triplicates. Means and SDs are indicated. Statistics for signiﬁcant conditions are shown. Signiﬁcance was assessed by unpaired t-test. p ≤0.01. (C) Cell viability of U-251 (EV and ATRX KO) high-grade glioma cells after 48 h of treatment with TMZ and/or sunitinib at the indicated concentrations compared to DMSO controls. Data shown correspond to four independent experiments. Mean and SEMs are indicated. Statistics for signiﬁcant conditions are shown. Signiﬁcance was assessed by unpaired t-test. * p ≤0.05; **** p ≤0.0001. (D) Representative images of U-251 (EV and ATRX KO) high-grade glioma cells after 48 h of treatment with either DMSO or the indicated drugs and concentrations. 3.3. ATRX KO High-Grade Glioma Cells Are Sensitive to RTK and PDGFR Inhibitors Ibrutinib (BTKi) and niclosamide (STAT3i) did not show a consistent increased toxicity in the ATRX-deﬁcient glioma cells as compared to their ATRX-proﬁcient counterparts (Figure S3E,F), suggesting that the sensitivity of ATRX-deﬁcient cells to these inhibitors may be inﬂuenced by other factors. Interestingly, the four RTKi tested induced higher toxicity to a different extent in the three ATRX KO high-grade glioma cell lines when used at 5 µM and 10 µM (Figure S4A–D). To further conﬁrm this ﬁnding, we performed additional experiments with the U-251 cell lines. U-251 cells (EV and ATRX KO) were treated with increasing concentrations of the four RTKi to determine their IC50 values. In line with the previous results, the U-251 ATRX KO cells showed increased sensitivity to the four RTKi tested with an IC50 at least 1.5-fold lower than their ATRX WT counterpart cells (Figure 3A–D). Cancers 2022, 14, 1790 9 of 17 Figure 3. ATRX-deﬁcient high-grade glioma cells are sensitive to multi-targeted RTK and speciﬁ PDGFR inhibitors. (A–E) Relative cell viability of U-251 (EV and ATRX KO) high-grade glioma ce lines treated with increasing doses of the indicated drug. The IC50 values were calculated from th dose–response curve determined by GraphPad Prism. Data shown correspond to a representativ experiment (with technical triplicates) out of three biological replicates. SEMs from each dat point are indicated. Signiﬁcance was assessed by F-test and the p-value for each dataset is show Representative images of EV and ATRX KO cells after 48 h of treatment with DMSO or the indicate drug concentrations are shown. Figure 3. ATRX-deﬁcient high-grade glioma cells are sensitive to multi-targeted RTK and speciﬁc PDGFR inhibitors. (A–E) Relative cell viability of U-251 (EV and ATRX KO) high-grade glioma cell lines treated with increasing doses of the indicated drug. The IC50 values were calculated from the dose–response curve determined by GraphPad Prism. Data shown correspond to a representative experiment (with technical triplicates) out of three biological replicates. SEMs from each data point are indicated. Signiﬁcance was assessed by F-test and the p-value for each dataset is shown. Representative images of EV and ATRX KO cells after 48 h of treatment with DMSO or the indicated drug concentrations are shown. Cancers 2022, 14, 1790 10 of 17 Furthermore, we tested the effect of RTKi by performing clonogenic assays. 3.4. ATRX KO High-Grade Glioma Cells Are Sensitive to Combinatorial Treatments of TMZ and PDGFRi 3.4. ATRX KO High-Grade Glioma Cells Are Sensitive to Combinatorial Treatments of TMZ and PDGFRi The current standard therapy for GBM includes TMZ and radiotherapy. In addition, it has been previously shown that ATRX KO cells are more sensitive to TMZ [33]. Therefore, we explored the effect of combining TMZ and PDGFRi on ATRX-deﬁcient high-grade glioma cells. To this end, we assessed the toxicity of a combined treatment consisting of low doses of TMZ and sunitinib or CP-673451 in U-251 EV and ATRX KO cells. Treatments with low doses of either TMZ, sunitinib or CP-673451 resulted in a slightly de- creased cell viability in the ATRX KO cells as compared to WT cells (Figures 4C,D and S5A,B). However, the combinatorial treatments with TMZ and sunitinib or CP-673451 led to a more pronounced decreased viability in the ATRX KO cells, (Figures 4C,D and S5A,B). p y g These data suggest that the use of combinatorial treatments with TMZ and RTKi or PDGFRi may increase the therapeutic window of opportunity in GBM patients with ATRX mutations. 3.3. ATRX KO High-Grade Glioma Cells Are Sensitive to RTK and PDGFR Inhibitors The four RTKi used (nintedanib, sunitinib, pazopanib, and sorafenib) are very potent inhibitors of PDGFR but also inhibit other RTK, such as VEGFR or FGFR to a variable extent [32]. In order to decipher whether ATRX-deﬁcient cells are sensitive to PDGFRi, we assessed the sensitivity of MOG-G-UVW, U-251, and SF-188 (EV and ATRX KO) high-grade glioma cell lines to the potent and selective PDGFRi, CP-673451. The three isogenic (EV and ATRX KO) glioma cell lines were treated with different concentrations (0.3 µM and 0.6 µM) of CP-673451 for 48 h to assess cell viability. We found that all the ATRX KO high-grade glioma cell lines are more sensitive to CP-673451 at 0.6 µM compared to their Cancers 2022, 14, 1790 11 of 17 11 of 17 WT counterparts (Figure S4E). In addition, U-251 (EV and ATRX KO) were treated with increasing concentrations of CP-673451 to determine their IC50 values. Importantly, U-251 ATRX KO cells have an IC50 of 2-fold lower when treated with CP-673451 than the U-251 EV cells (Figure 3E), indicating that speciﬁc inhibition of PDGFR leads to higher cell toxicity in ATRX-deﬁcient cells. 3.5. Patient-Derived GBM Cells with ATRX Mutations Are Sensitive to RTK and PDGFR Inhibitors Next, we aimed to validate these ﬁndings in cells derived from adult malignant glioma patients harboring somatic ATRX mutations. We used patient-derived GBM cells from the human glioblastoma cell culture (HGCC) biobank (https://www.hgcc.se/, accessed on 21 March 2022). These cells were obtained from surgical samples of human adult GBM patients and cultured as described in Xie et al., 2015 [34]. We compared the effect of sunitinib and the more speciﬁc PDGFRi CP-673451 in two cell lines harboring mutations in ATRX (U3129 and U3034) and two without them (U3082 and U3024). The U3129 and U3034 patient-derived glioma cells used in this study harbor intronic and missense mutations in ATRX (https://www.hgcc.se/, accessed on 21 March 2022), which, besides potential alterations to protein functionality, lead to reduced ATRX levels (Figure S5C,D). Cells were grown and studied in pairs grouped according to their molecular subtype, proneural GBM (U-3082 and U-3129), and mesenchymal GBM (U-3024 and U-3034). These cells were treated with increasing concentrations of sunitinib or CP-67345 and IC50 s were calculated as described previously. Cells harboring ATRX mutations were more sensitive to sunitinib and CP-673451 than ATRX WT cells, showing an IC50 of at least 1.6-fold lower (Figure 5A–D). g ATRX mutations usually co-occur with mutations in the TP53 gene in high-grade glioma tumors, which could inﬂuence the response to these drugs [35–37]. The U-251 cell line used in this study harbors mutations in the TP53 gene [10]. However, it should be noted that we detected a functional p53 response upon DNA damage in terms of p53 accumulation and p21 activation in the three isogenic cell lines (Figure S6A). We tested whether p53 depletion could inﬂuence the sensitivity of U-251 (EV and ATRX KO) cells to the RTKi sunitinib and nintedanib. Of note, cells infected with the control shRNAs exhibited high p21 and p53 basal levels, probably due to the stress caused by the recent viral infection. Nevertheless, p53-depleted U-251 ATRX KO cells, generated by p53 shRNA transduction, were also more sensitive to RTKi than p53-depleted ATRX WT cells (Figure S6B,C). 12 of 17 Cancers 2022, 14, 1790 Figure 5. ATRX-mutant patient-derived high-grade glioma cells are sensitive to RTK and PDGFR inhibitors. (A–D) Relative cell viability of ATRX WT (U-3082 and U-3024) and ATRX Mutant (U-3129 and U-3034) high-grade glioma cell lines treated with increasing doses of the indicated drug. The IC50 values were calculated from the dose–response curve determined by GraphPad Prism. 4. Discussion Given the high prevalence of ATRX mutations in cancer, particularly in gliomas (reviewed in [15]), identifying drugs highly toxic for ATRX-deﬁcient cells may lead to clinical applications. Thus, we performed high-content microscopy-based drug screens using ATRX WT and ATRX KO isogenic HeLa cell lines generated by CRISPR/Cas9. Of note, we used an FDA-approved drug library, hence the biosafety of the compounds was already tested and the ﬁndings from this screen could be rapidly implemented in the clinic. We used two pipelines to perform drug screens, in both cases pooling ATRX WT and ATRX KO cells in the same wells, which facilitated the analysis of a high number of drugs and the direct comparison of the effect of the drugs. In the primary screen, the p96-well plates were stained with an ATRX antibody to discriminate between ATRX WT and KO cells after 48 h of drug treatment. The top hits validated in a secondary screen where the isogenic clones were GFP/tdTomato-labeled by lentiviral transduction prior to the screen. This second strategy allows to quantify the effect of the drugs on individual clones at different time points and doses in one single experiment. We focused on the validation of the drugs that caused higher toxicity to ATRX KO cells compared to WT cells. However, the drugs that caused lower toxicity could also provide interesting clinical information. For instance, patients with ATRX mutations could be refractory to treatments with those drugs. Since ATRX is a regulator of therapy-induced senescence [11], some of the drugs causing less toxicity in ATRX KO cells may act by inducing senescence in ATRX WT cells, and therefore, in that scenario, ATRX-deﬁcient cells might present a proliferative advantage compared to the ATRX-proﬁcient cells. g p p g p p Among the 29 top hits identiﬁed in the primary screen, seven compounds were validated in the secondary screen, also using HeLa cells. Next, we studied three of these drugs (ibrutinib, niclosamide, and pazopanib) in isogenic ATRX KO high-grade glioma cell lines previously generated by CRISPR/Cas9 [10]. Ibrutinib and niclosamide have shown positive results in the treatment of glioma cells, regardless of the ATRX status [28,29,31]. Furthermore, ibrutinib and pazopanib have been used in clinical trials for GBM patients (NCT03535350, NCT00459381). However, only the RTKi pazopanib led to increased toxicity in ATRX KO high-grade glioma cells compared to the WT counterparts. 3.5. Patient-Derived GBM Cells with ATRX Mutations Are Sensitive to RTK and PDGFR Inhibitors Data shown correspond to a representative Figure 5. ATRX-mutant patient-derived high-grade glioma cells are sensitive to RTK and P inhibitors. (A–D) Relative cell viability of ATRX WT (U-3082 and U-3024) and ATRX Mutant ( and U-3034) high-grade glioma cell lines treated with increasing doses of the indicated dru IC50 values were calculated from the dose–response curve determined by GraphPad Prism shown correspond to a representative experiment (with technical triplicates) out of three bio replicates. SEMs from each data point are indicated. Signiﬁcance was assessed using F-test a p-value for each dataset is shown. Representative images of ATRX WT and ATRX Mutant cel 48h of treatment with DMSO or the indicated drug concentrations are shown. Figure 5. ATRX-mutant patient-derived high-grade glioma cells are sensitive to RTK and PDGFR inhibitors. (A–D) Relative cell viability of ATRX WT (U-3082 and U-3024) and ATRX Mutant (U-3129 and U-3034) high-grade glioma cell lines treated with increasing doses of the indicated drug. The IC50 values were calculated from the dose–response curve determined by GraphPad Prism. Data shown correspond to a representative Figure 5. ATRX-mutant patient-derived high-grade glioma cells are sensitive to RTK and PDGFR inhibitors. (A–D) Relative cell viability of ATRX WT (U-3082 and U-3024) and ATRX Mutant (U-3129 and U-3034) high-grade glioma cell lines treated with increasing doses of the indicated drug. The IC50 values were calculated from the dose–response curve determined by GraphPad Prism. Data shown correspond to a representative experiment (with technical triplicates) out of three biological replicates. SEMs from each data point are indicated. Signiﬁcance was assessed using F-test and the p-value for each dataset is shown. Representative images of ATRX WT and ATRX Mutant cells after 48h of treatment with DMSO or the indicated drug concentrations are shown. Cancers 2022, 14, 1790 13 of 17 4. Discussion This suggests that the ATRX-related sensitivity to STATi and BTKi is inﬂuenced by other factors such as the cellular genetic background and the cell type from which the tumor originates. It would be worth analyzing the sensitivity to these drugs in other cancer types presenting ATRX mutations, such as hepatocellular carcinoma, pancreatic neuroendocrine tumors, and neuroblastomas, and in cells with different genetic backgrounds. Three different ATRX KO high-grade glioma cell lines showed sensitivity to pazopanib and three additional RTKi (sorafenib, nintedanib, and sunitinib), all of which target PDGFR and other RTKs. Interestingly, a speciﬁc PDGFRi (CP-673451) also produced higher toxicity in ATRX KO cells, indicating that PDGFR is the main target responsible for this ATRX- related sensitivity. Enhanced PDGFR signaling is associated with the progression of several pathologies, including tumorigenesis [38], hence many anticancer therapies are based on targeting PDGFR signaling. The overexpression of the PDGFR alpha subunit has been reported to correlate with ATRX mutations in cancer cells [16,17]. This may suggest a direct or indirect compensatory relationship between these two events, which may explain the higher toxicity of PDGFRi in ATRX-deﬁcient high-grade glioma cells. Furthermore, the increased sensitivity of ATRX-deﬁcient cells to RTKi (i.e., sunitinib) has been recently shown in other cancer cell types, such as neuroblastoma, reinforcing our ﬁndings [39]. In addition, our study shows that combinatorial treatments of RTKi or PDGFRi and TMZ (the current standard of care treatment for GBM patients) might be beneﬁcial for patients with tumors harboring ATRX mutations, which is in line with previous studies showing that TMZ inhibits glioma formation and increases the chemosensitivty of ATRX-deﬁcient gliomas in vivo [33]. Importantly, there are several clinical trials using similar combinatorial treatments for GBM (i.e., NCT00597493, NCT00544817), and it has been shown that combi- natorial treatments of RTKi plus TMZ can be safely administered to these patients [40,41]. Cancers 2022, 14, 1790 14 of 17 We found that patient-derived GBM cells harboring somatic ATRX mutations are also more sensitive to RTK and PDGFR inhibitors, supporting that high-grade glioma patients harboring ATRX mutations could respond better to these drugs. Most ATRX mutations found in glioma patients are truncating mutations leading to the complete loss of the functional ATRX protein. Thus, the isogenic ATRX KO cells used in this study resemble the ATRX status in most cancer patients with ATRX mutations. 4. Discussion However, some patients present ATRX missense mutations, which are predicted as pathogenic but are less characterized. Indeed, the mutations identiﬁed in the patient-derived glioma cells used in this study (U3129 and U3034) are intronic and missense mutations that lead to reduced ATRX levels. All ATRX KO high-grade glioma cell lines used in this study were more sensitive to RTKi and PDGFRi than ATRX WT cells, independently of the type of mutation in ATRX. Patients with ATRX-deﬁcient high grade glioma tumors often harbor mutations in other genes, such as TP53 or IDH1, which could inﬂuence the response to these drugs [35–37]. The high-grade glioma cell lines used in our study do not harbor mutations in the IDH1/2 genes [10,34,42]. Since secondary GBM frequently harbor mutations in both, IDH1 and ATRX genes [43], future studies should address the toxicity of RTKi and selective PDGFRi (with and without TMZ) in cells with this relevant mutational background. On the other hand, all the patient-derived GBM cells and two of the isogenic high-grade glioma cell lines used in our study, U-251 and SF188, harbor mutations in the TP53 gene [10,34], and all the ATRX-deﬁcient cell lines tested were sensitive to RTKi and PDGFRi irrespective of their TP53 status. Therefore, these inhibitors could also be effective in treating GBM patients harboring TP53 and ATRX co-occurrent mutations. Abbreviation ALT Alternative Lengthening of Telomeres ATRX alpha thalassemia mental retardation X-linked BTK Bruton’s tyrosine kinase BTKi inhibitors targeting BTK CFS common fragile sites DAXX death domain-associated protein 6 DMSO dimethyl sulfoxide DSB double-strand break EdU 5-Ethynyl-2′-deoxyuridine EV empty vector FDA food and drug administration GBM glioblastoma multiforme GFP green ﬂuorescent protein gRNA guide RNA H3.3 histone H3.3 HGCC human glioblastoma cell culture IDH1 isocitrate dehydrogenase 1 KO knockout PDGFR platelet-derived growth factor receptor PDGFRi inhibitors targeting PDGFR RTK receptor tyrosine kinase RTKi inhibitors targeting RTK shRNA short hairpin RNA STAT3 signal transducer and activator of transcription 3 STAT3i inhibitors targeting STAT3 TMZ temozolomide TP53 tumor protein 53 WT wild-type γH2AX histone H2AX phosphorylation 5. Conclusions Paula Aguilera was supported with a Juan de la Cierva formación fellowship from the MICINN and an Incorporación fellowship from the Junta de Andalucía. Toyota Fonden and Læge Sofus Carl Emil Friis og hustru Olga Doris Fonden funded the acquisition of the high-content microscope used in this study. Foundation (R218-2016-415) and funding from Dansk Kræftforskningsfond. María Castejón-Griñán holds an Incorporación fellowship from the Junta de Andalucía. Paula Aguilera was supported with a Juan de la Cierva formación fellowship from the MICINN and an Incorporación fellowship from the Junta de Andalucía. Toyota Fonden and Læge Sofus Carl Emil Friis og hustru Olga Doris Fonden funded the acquisition of the high-content microscope used in this study. a Juan de la Cierva formación fellowship from the MICINN and an Incorporación fellowship from the Junta de Andalucía. Toyota Fonden and Læge Sofus Carl Emil Friis og hustru Olga Doris Fonden funded the acquisition of the high-content microscope used in this study. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request to the corre- sponding author. Acknowledgments: We thank Hocine Mankouri for critical reading of this manuscript and our col- leagues at the Center for Chromosome Stability for valuable input and discussions about our results. Conﬂicts of Interest: The authors declare of no conﬂict of interest. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request to the corre- sponding author. Acknowledgments: We thank Hocine Mankouri for critical reading of this manuscript and our col- leagues at the Center for Chromosome Stability for valuable input and discussions about our results. Conﬂicts of Interest: The authors declare of no conﬂict of interest. 5. Conclusions In this study, we have demonstrated that ATRX deficient cells, including patient-derived high-grade glioma cell lines, are particularly sensitive to RTKi (pazopanib, nintedanib, suni- tinib, sorafenib) and to a specific PDGFRi (CP-673451). Besides, this sensitivity was enhanced when RTKi or PDGFRi were combined with TMZ (current standard of care in glioblastoma treatment). Importantly, there are several clinical trials using similar combinatorial treatments for glioblastoma (i.e., NCT02928575, NCT00597493, NCT00544817, NCT02331498). Therefore, based on our findings and the frequent presence of ATRX mutations in patients who suffer secondary GBM or anaplastic astrocytomas, we suggest that the analyses of these clinical trials data should consider the ATRX status. Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cancers14071790/s1, Figure S1: Characterization of HeLa ATRX- KO clones, Figure S2: Cellular viability of tdTomato ATRX WT and GFP ATRX KO HeLa clones upon treatment with BTK, STAT3 and RTK inhibitors Figure S3: CRISPR/Cas9 engineered MOG- G-UVW, U-251 and SF188 (EV and ATRX KO) high-grade glioma cells: ATRX expression, effect on EdU incorporation and γH2AX levels, and sensitivity to BTK and STAT3 inhibitors, Figure S4: Cell viability of MOG-G-UVW, U-251 and SF-188 (EV and ATRX KO) upon treatment with RTK and PDGFR inhibitors, Figure S5: ATRX-deﬁcient high-grade glioma cells are sensitive to combinatorial treatments of TMZ and CP-673451, Figure S6: TP53 loss does not inﬂuence the toxicity of RTK inhibitors in ATRX-KO U-251 high-grade glioma cells Table S1. List of compounds identiﬁed in the ﬁrst screen. Author Contributions: A.J.L.-C. conceived, designed and supervised the study. D.P.-M. and M.C.-G. performed and analyzed the drug screen. D.P.-M., M.C.-G. and P.A. designed, performed and analyzed most of the experiments. K.G. assisted with Western blot experiments. A.I. helped with the generation of HeLa ATRX KO cells. A.K.M. and J.A.B.-C. generated the MOG-G-UVW, U-251 and SF-188 (EV and ATRX KO) cells. A.J.L.-C., D.P.-M. and M.C.-G. wrote the manuscript. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported by grants from Danish National Research Foundation (DNRF115), Danish Cancer Society (KBVU-2017_R167-A11063), European Research Council (ERC-2015-STG- 679068), Nordea-fonden (02-2017-1749) and the Spanish Ministry of Science and Innovation (PID2020- 119329RB-I00). David Pladevall-Morera was supported with a PhD scholarship from the Lundbeck Cancers 2022, 14, 1790 15 of 17 Foundation (R218-2016-415) and funding from Dansk Kræftforskningsfond. María Castejón-Griñán holds an Incorporación fellowship from the Junta de Andalucía. 4. Lewis, P.W.; Elsaesser, S.J.; Noh, K.-M.; Stadler, S.C.; Allis, C.D. Daxx is an H3.3-speciﬁc histone chaperone and cooperates with ATRX in replication-independent chromatin assembly at telomeres. Proc. Natl. Acad. Sci. USA 2010, 107, 14075–14080. [CrossRef] References 1. Gibbons, R.; Picketts, D.; Villard, L.; Higgs, D. Mutations in a putative global transcriptional regulator cause X-linked mental retardation with alpha-thalassemia (ATR-X syndrome). Cell 1995, 80, 837–845. [CrossRef] 1. Gibbons, R.; Picketts, D.; Villard, L.; Higgs, D. Mutations in a putative global transcriptional regulator cause X-linked mental retardation with alpha-thalassemia (ATR-X syndrome). Cell 1995, 80, 837–845. [CrossRef] p g g 4. Lewis, P.W.; Elsaesser, S.J.; Noh, K.-M.; Stadler, S.C.; Allis, C.D. Daxx is an H3.3-speciﬁc histone chaperone and cooperates with ATRX in replication-independent chromatin assembly at telomeres. Proc. Natl. Acad. Sci. USA 2010, 107, 14075–14080. [CrossRef] 16 of 17 Cancers 2022, 14, 1790 5. Juhász, S.; Elbakry, A.; Mathes, A.; Löbrich, M. ATRX Promotes DNA Repair Synthesis and Sister Chromatid Exchange during Homologous Recombination. Mol. Cell 2018, 71, 11.e7–24.e7. [CrossRef] g 6. Nguyen, D.T.; Voon, H.P.J.; Xella, B.; Scott, C.; Clynes, D.; Babbs, C.; Ayyub, H.; Kerry, J.; Sharpe, J.A.; Sloane-Stanley, J.A.; et al. The chromatin remodelling factor ATRX suppresses R-loops in transcribed telomeric repeats. EMBO Rep. 2017, 18, 914–928. [CrossRef] [PubMed] [ ] [ ] 7. Wang, Y.; Yang, J.; Wild, A.T.; Wu, W.H.; Shah, R.; Danussi, C.; Riggins, G.J.; Kannan, K.; Sulman, E.P.; Chan, T.A.; et al. G-quadruplex DNA drives genomic instability and represents a targetable molecular abnormality in ATRX-deﬁcient malignant glioma. Nat. Commun. 2019, 10, 943. [CrossRef] g 8. Pladevall-Morera, D.; Munk, S.; Ingham, A.; Garribba, L.; Albers, E.; Liu, Y.; Olsen, J.V.; Lopez-Contreras, A.J. Proteomic characterization of chromosomal common fragile site (CFS)-associated proteins uncovers ATRX as a regulator of CFS stability. Nucleic Acids Res. 2019, 47, 8004–8018. [CrossRef] 9. Amorim, J.P.; Santos, G.; Vinagre, J.; Soares, P. The role of ATRX in the alternative lengthening of telomeres (ALT) phenotype. Genes. 2016, 7, 66. [CrossRef] 10. Brosnan-Cashman, J.A.; Yuan, M.; Graham, M.K.; Rizzo, A.J.; Myers, K.M.; Davis, C.; Zhang, R.; Esopi, D.M.; Raabe, E.H.; Eberhart, C.G.; et al. ATRX loss induces multiple hallmarks of the alternative lengthening of telomeres (ALT) phenotype in human glioma cell lines in a cell line-speciﬁc manner. PLoS ONE 2018, 13, e0204159. [CrossRef] g p 11. Kovatcheva, M.; Liao, W.; Klein, M.E.; Robine, N.; Geiger, H.; Crago, A.M.; Dickson, M.A.; Tap, W.D.; Singer, S.; Koff, A. ATRX is a regulator of therapy induced senescence in human cells. Nat. Commun. 2017, 8, 386. [CrossRef] [PubMed] 12. References [CrossRef] rico, S.; Chen, X.; Fan, Y.; Xu, B.; Stewart, E.; Zhou, X.; Jeon, J.; Grifﬁths, L.; Nguyen, R.; et al. MYCN TRX mutations are incompatible in neuroblastoma. Nat. Commun. 2020, 11, 913. [CrossRef] 20. Zeineldin, M.; Federico, S.; Chen, X.; Fan, Y.; Xu, B.; Stewart, E.; Zhou, X.; Jeon, J.; Grifﬁths, L.; N ampliﬁcation and ATRX mutations are incompatible in neuroblastoma. Nat. Commun. 2020, 11, 913. [C 21. Schneider, T.; Mawrin, C.; Scherlach, C.; Skalej, M.; Firsching, R. Gliomas in Adults. Dtsch. Arztebl. Int. 2010, 107, 799–807. [CrossRef] [PubMed] 22. Stupp, R.; Mason, W.P.; van den Bent, M.J.; Weller, M.; Fisher, B.; Taphoorn, M.J.B.; Belanger, K.; Brandes, A.A.; Marosi, C.; Bogdahn, U.; et al. Radiotherapy plus Concomitant and Adjuvant Temozolomide for Glioblastoma. N. Engl. J. Med. 2009, 352, 987–996. [CrossRef] [PubMed] 23. Liang, J.; Zhao, H.; Diplas, B.H.; Liu, S.; Liu, J.; Wang, D.; Lu, Y.; Zhu, Q.; Wu, J.; Wang, W.; et al. Genome-Wide CRISPR-Cas9 Screen Reveals Selective Vulnerability of ATRX-Mutant Cancers to WEE1 Inhibition. Cancer Res. 2020, 80, 510–523. [CrossRef] [PubMed] 24. Ran, F.A.; Hsu, P.D.; Wright, J.; Agarwala, V.; Scott, D.A.; Zhang, F. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 2013, 8, 2281–2308. [CrossRef] 24. Ran, F.A.; Hsu, P.D.; Wright, J.; Agarwala, V.; Scott, D.A.; Zhang, F. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 2013, 8, 2281–2308. [CrossRef] 25 Cl D J li k C X ll B A b H T l S Mit M B h ti C Z Hi D R Gibb R J ATRX D f ti C.; Xella, B.; Ayyub, H.; Taylor, S.; Mitson, M.; Bachrati, C.Z.; Higgs, D.R.; Gibbons, R.J. ATRX Dysfunction Defects in Primary Mouse Cells. PLoS ONE 2014, 9, e92915. [CrossRef] 25. Clynes, D.; Jelinska, C.; Xella, B.; Ayyub, H.; Taylor, S.; Mitson, M.; Bachrati, C.Z.; Higgs, D.R.; Gibbon Induces Replication Defects in Primary Mouse Cells. PLoS ONE 2014, 9, e92915. [CrossRef] p y 26. Haapaniemi, E.; Botla, S.; Persson, J.; Schmierer, B.; Taipale, J. CRISPR–Cas9 genome editing induces a p53-mediated DNA damage response. Nat. Med. 2018, 24, 927–930. [CrossRef] g p 27. Ihry, R.J.; Worringer, K.A.; Salick, M.R.; Frias, E.; Ho, D.; Theriault, K.; Kommineni, S.; Chen, J.; Sondey, M.; Ye, C.; et al. p53 inhibits CRISPR–Cas9 engineering in human pluripotent stem cells. Nat. Med. 2018, 24, 939–946. [CrossRef] 28. References Totoki, Y.; Tatsuno, K.; Covington, K.R.; Ueda, H.; Creighton, C.J.; Kato, M.; Tsuji, S.; Donehower, L.A.; Slagle, B.L.; Nakamura, H.; et al. Trans-ancestry mutational landscape of hepatocellular carcinoma genomes. Nat. Genet. 2014, 46, 1267–1273. [CrossRef] [PubMed] [ ] 13. Jiao, Y.; Shi, C.; Edil, B.H.; de Wilde, R.F.; Klimstra, D.S.; Maitra, A.; Schulick, R.D.; Tang, L.H.; Wolfgang, C.L.; Choti, M.A.; et al. DAXX/ATRX, MEN1, and mTOR pathway genes are frequently altered in pancreatic neuroendocrine tumors. Science 2011, 331, 1199–1203. [CrossRef] [ ] 14. Heaphy, C.M.; de Wilde, R.F.; Jiao, Y.; Klein, A.P.; Edil, B.H.; Shi, C.; Bettegowda, C.; Rodriguez, F.J.; Eberhart, C.G.; Hebbar, S.; et al. Altered telomeres in tumors with ATRX and DAXX mutations. Science 2011, 333, 425. [CrossRef] 15. Haase, S.; Garcia-Fabiani, M.B.; Carney, S.; Altshuler, D.; Núñez, F.J.; Méndez, F.M.; Núñez, F.; Lowenstein, P.R.; Castro, M.G. Mutant ATRX: Uncovering a new therapeutic target for glioma. Expert Opin. Ther. Targets 2018, 22, 599–613. [CrossRef] ; Garcia-Fabiani, M.B.; Carney, S.; Altshuler, D.; Núñez, F.J.; Méndez, F.M.; Núñez, F.; Lowenstein, P.R.; C aase, S.; Garcia-Fabiani, M.B.; Carney, S.; Altshuler, D.; Núñez, F.J.; Méndez, F.M.; Núñez, F.; Lowenstein, utant ATRX: Uncovering a new therapeutic target for glioma. Expert Opin. Ther. Targets 2018, 22, 599–613. [ ATRX: Uncovering a new therapeutic target for glioma. Expert Opin. Ther. Targets 2018, 22, 599–613. [CrossRe 16. Abedalthagaﬁ, M.; Phillips, J.; Kim, G.; Mueller, S.; Haas-Kogen, D.; Marshall, R.; Croul, S.; Santi, M.; Cheng, J.; Zhou, S.; et al. The alternative lengthening of telomere phenotype is signiﬁcantly associated with loss of ATRX expression in high-grade pediatric and adult astrocytomas: A multi-institutional study of 214 astrocytomas. Mod. Pathol. 2013, 26, 1425–1432. [CrossRef] [PubMed] y y y 17. Purkait, S.; Miller, C.A.; Kumar, A.; Sharma, V.; Pathak, P.; Jha, P.; Sharma, M.C.; Suri, V.; Suri, A.; Sharma, B.S.; et al. ATRX in Diffuse Gliomas With its Mosaic/Heterogeneous Expression in a Subset. Brain Pathol. 2017, 27, 138–145. [CrossRef] 18. Scarpa, A.; Chang, D.K.; Nones, K.; Corbo, V.; Patch, A.M.; Bailey, P.; Lawlor, R.T.; Johns, A.L.; Miller, D.K.; Mafﬁcini, A.; et al. Whole-genome landscape of pancreatic neuroendocrine tumours. Nature 2017, 543, 65–71. [CrossRef] [PubMed] 19. Shah, M.A.; Denton, E.L.; Arrowsmith, C.H.; Lupien, M.; Schapira, M. A global assessment of cancer genomic alterations in epigenetic mechanisms. Epigenetics Chromatin 2014, 7, 29. References Wang, J.; Liu, X.; Hong, Y.; Wang, S.; Chen, P.; Gu, A.; Guo, X.; Zhao, P. Ibrutinib, a Bruton’s tyrosine kinase inhibitor, exhibits antitumoral activity and induces autophagy in glioblastoma. J. Exp. Clin. Cancer Res. 2017, 36, 96. [CrossRef] y p gy g p 29. Wieland, A.; Trageser, D.; Gogolok, S.; Reinartz, R.; Höfer, H.; Keller, M.; Leinhaas, A.; Schelle, R.; Normann, S.; Klaas, L.; et al. Anticancer Effects of Niclosamide in Human Glioblastoma. Clin. Cancer Res. 2013, 19, 4124–4136. [CrossRef] 17 of 17 17 of 17 Cancers 2022, 14, 1790 30. Reardon, D.A.; Groves, M.D.; Wen, P.Y.; Nabors, L.; Mikkelsen, T.; Rosenfeld, S.; Raizer, J.; Barriuso, J.; McLendon, R.E.; Suttle, A.B.; et al. A Phase I/II Trial of Pazopanib in Combination with Lapatinib in Adult Patients with Relapsed Malignant Glioma. Clin. Cancer Res. 2013, 19, 900–908. [CrossRef] 31. Cheng, B.; Morales, L.D.; Zhang, Y.; Mito, S.; Tsin, A. Niclosamide induces protein ubiquitination and inhibits multiple pro- survival signaling pathways in the human glioblastoma U-87 MG cell line. PLoS ONE 2017, 12, e0184324. [CrossRef] [PubMed] 32. Hojjat-Farsangi, M. Small-Molecule Inhibitors of the Receptor Tyrosine Kinases: Promising Tools for Targeted Cancer Therapies. Int. J. Mol. Sci. 2014, 15, 13768–13801. [CrossRef] [PubMed] survival signaling pathways in the human glioblastoma U-87 MG cell line. PLoS ONE 2017, 12, e0184324. [CrossRef] [PubMed] 32. Hojjat-Farsangi, M. Small-Molecule Inhibitors of the Receptor Tyrosine Kinases: Promising Tools for Targeted Cancer Therapies. Int. J. Mol. Sci. 2014, 15, 13768–13801. [CrossRef] [PubMed] 32. Hojjat-Farsangi, M. Small-Molecule Inhibitors of the Receptor Tyrosine Kinases: Promising Tools fo Int. J. Mol. Sci. 2014, 15, 13768–13801. [CrossRef] [PubMed] J , , [ ] [ ] 33. Han, B.; Cai, J.; Gao, W.; Meng, X.; Gao, F.; Wu, P.; Duan, C.; Wang, R.; Dinislam, M.; Lin, L.; et al. Loss of ATRX suppresses ATM dependent DNA damage repair by modulating H3K9me3 to enhance temozolomide sensitivity in glioma. Cancer Lett. 2018, 419, 280–290. [CrossRef] 34. Xie, Y.; Bergström, T.; Jiang, Y.; Johansson, P.; Marinescu, V.D.; Lindberg, N.; Segerman, A.; Wicher, G.; Niklasson, M.; Baskaran, S.; et al. The Human Glioblastoma Cell Culture Resource: Validated Cell Models Representing All Molecular Subtypes. EBioMedicine 2015, 2, 1351–1363. [CrossRef] [PubMed] , , [ ] [ ] 35. Jiao, Y.; Killela, P.J.; Reitman, Z.J.; Rasheed, B.A.; Heaphy, C.M.; de Wilde, R.F.; Rodriguez, F.J.; Rosemberg, S.; Obashinjo, S.M.; Marie, S.K.N.; et al. References Frequent ATRX, CIC, FUBP1 and IDH1 mutations reﬁne the classiﬁcation of malignant gliomas. Oncotarget 2012, 3, 709–722. [CrossRef] [PubMed] 36. Ishii, N.; Maier, D.; Merlo, A.; Tada, M.; Sawamura, Y.; Diserens, A.-C.; Meir, E.G. Van Frequent Co-Alterations of TP53, p16/CDKN2A, p14ARF, PTEN Tumor Suppressor Genes in Human Glioma Cell Lines. Brain Pathol. 1999, 9, 469–479. [CrossRef] [PubMed] 37. O’Connor, P.M.; Jackman, J.; Bae, I.; Myers, T.G.; Fan, S.; Mutoh, M.; Scudiero, D.A.; Monks, A.; Sausville, E.A.; Weinstein, J.N.; et al. Characterization of the p53 Tumor Suppressor Pathway in Cell Lines of the National Cancer Institute Anticancer Drug Screen and Correlations with the Growth-Inhibitory Potency of 123 Anticancer Agents. Cancer Res. 1997, 57, 4285–4300. y y g 38. Levitzki, A. PDGF receptor kinase inhibitors for the treatment of PDGF driven diseases. Cytokine Growth Factor Rev. 2004, 15, 229–235. [CrossRef] 39. George, S.L.; Lorenzi, F.; King, D.; Hartlieb, S.; Campbell, J.; Pemberton, H.; Toprak, U.H.; Barker, K.; Tall, J.; da Costa, B.M.; et al. Therapeutic vulnerabilities in the DNA damage response for the treatment of ATRX mutant neuroblastoma. EBioMedicine 2020, 59, 102971. [CrossRef] 40. Reardon, D.A.; Vredenburgh, J.J.; Desjardins, A.; Peters, K.; Gururangan, S.; Sampson, J.H.; Marcello, J.; Herndon, J.E.; McLendon, R.E.; Janney, D.; et al. Effect of CYP3A-inducing anti-epileptics on sorafenib exposure: Results of a phase II study of sorafenib plus daily temozolomide in adults with recurrent glioblastoma. J. Neurooncol. 2011, 101, 57–66. [CrossRef] p y g 41. Hainsworth, J.D.; Ervin, T.; Friedman, E.; Priego, V.; Murphy, P.B.; Clark, B.L.; Lamar, R.E. Concurrent radiotherapy and temozolomide followed by temozolomide and sorafenib in the ﬁrst-line treatment of patients with glioblastoma multiforme. Cancer 2010, 116, 3663–3669. [CrossRef] [PubMed] 42. Johansson, P.; Krona, C.; Kundu, S.; Doroszko, M.; Baskaran, S.; Schmidt, L.; Vinel, C.; Almstedt, E.; Elgendy, R.; Elﬁneh, L.; et al. A Patient-Derived Cell Atlas Informs Precision Targeting of Glioblastoma. Cell Rep. 2020, 32, 107897. [CrossRef] [PubMed] 43. Liu, X.-Y.; Gerges, N.; Korshunov, A.; Sabha, N.; Khuong-Quang, D.-A.; Fontebasso, A.M.; Fleming, A.; Hadjadj, D.; Schwartzen- truber, J.; Majewski, J.; et al. Frequent ATRX mutations and loss of expression in adult diffuse astrocytic tumors carrying IDH1/IDH2 and TP53 mutations. Acta Neuropathol. 2012, 124, 615–625. [CrossRef] [PubMed]
https://openalex.org/W2508210137	https://europepmc.org/articles/pmc5013499?pdf=render	English	null	Impact of intensive treatment and remission on health-related quality of life in early and established rheumatoid arthritis	RMD open	2,016	cc-by	7,895	To cite: Scott IC, Ibrahim F, Lewis CM, et al. Impact of intensive treatment and remission on health-related quality of life in early and established rheumatoid arthritis. RMD Open 2016;2: e000270. doi:10.1136/ rmdopen-2016-000270 Key messages Objectives: To establish if using intensive treatment to reduce synovitis and attain remission in active rheumatoid arthritis (RA) improves all aspects of health-related quality of life (HRQoL). ▸It also demonstrates that in patients with previ- ously active RA, attaining remission substantially improves, but fails to normalise, HRQoL (espe- cially in established disease), and that the patient global assessment (PtGA) of disease activity has a strong association with HRQoL. 1Academic Department of Rheumatology, Centre for Molecular and Cellular Biology of Inflammation, King’s College London, London, UK 2Department of Medical and Molecular Genetics, King’s College London, Guy’s Hospital, London, UK 3Department of Rheumatology, Weston Education Centre, King’s College Hospital, London, UK 4Division of Immunology/ Rheumatology, Stanford University School of Medicine, Palo Alto, California, USA Correspondence to Dr I C Scott; ian scott@kcl ac uk How might this impact on clinical practice? ▸In order to optimise HRQoL in RA, it may be helpful to place an increased focus on improving PtGA scores; additional treatments based on reducing pain, fatigue and improving mood, which are likely to improve PtGA scores, may be of benefit in HRQoL. Rheumatoid arthritis Rheumatoid arthritis What is already known about this subject? What is already known about this subject? ▸An important treatment goal in rheumatoid arth- ritis (RA) is optimising health-related quality of life (HRQoL). Methods: A secondary analysis of two randomised clinical trials (CARDERA and TACIT) was undertaken. CARDERA randomised 467 patients with early active RA to different disease-modifying antirheumatic drug (DMARD) regimens, including high-dose tapering corticosteroids. TACIT randomised 205 established patients with active RA to combination DMARDs (cDMARDs) or tumour necrosis factor-α inhibitors (TNFis). Short-Form 36 (SF-36) measured HRQoL across eight domains, generating physical (PCS) and mental (MCS) component summary scores. Linear regression evaluated 6-month intensive treatment impacts. Mean SF-36 scores, stratified by end point disease activity category, were compared with age/ gender-matched population scores. ▸It is uncertain if intensive treatment and remis- sion improve all aspects of HRQoL, with previ- ous placebo-controlled trials of tumour necrosis factor-α inhibitors (TNFis) combined with methotrexate suggesting greater benefits on physical than mental health. ▸Prepublication history and additional material is available. To view please visit the journal (http://dx.doi.org/ 10.1136/rmdopen-2016- 000270). Impact of intensive treatment and remission on health-related quality of life in early and established rheumatoid arthritis I C Scott,1,2 F Ibrahim,3 C M Lewis,2 D L Scott,3 V Strand4 I C Scott,1,2 F Ibrahim,3 C M Lewis,2 D L Scott,3 V Strand4 I C Scott,1,2 F Ibrahim,3 C M Lewis,2 D L Scott,3 V Strand4 What does this study add? What does this study add? ▸This secondary analysis of two randomised clin- ical trials shows that in early active RA, intensive corticosteroid therapy improves physical but not mental HRQoL relative to placebo, and in estab- lished active RA, intensive treatment with com- bination disease-modifying antirheumatic drugs and TNFis has similar impacts on HRQoL. Received 25 February 2016 Revised 13 July 2016 Accepted 18 July 2016 Results: In CARDERA, intensive corticosteroid treatment gave significantly greater improvements in PCS but not MCS scores relative to placebo. In TACIT, all eight SF-36 domains had improvements from baseline exceeding minimal clinically important differences with cDMARDs and TNFis. Significantly greater improvements with TNFi relative to cDMARDs were reported in PCS only (p=0.034), after adjusting for covariates. Remission provided the best SF-36 profiles, but scores in physical functioning, role physical and general health in both trials remained below normative values. Patient global assessment of disease activity had a greater association with HRQoL than other disease activity score (DAS28) components. Conclusions: Intensive corticosteroid treatment in early RA improves physical but not mental health, relative to placebo. In established RA, cDMARDs and TNFi provide similar improvements in HRQoL. As remission optimises but fails to normalise HRQoL, a focus on treatment strategies targeting HRQoL is required. Health-related quality of life Short-Form 36 which mainly reﬂect reduced joint counts and acute phase markers. A key secondary goal is improving health-related quality of life (HRQoL). The treat-to-target (T2T) approach is based on the concept that HRQoL is maximised when synovitis is minimised through escalating drug treatment until patients achieve remission.3 which mainly reﬂect reduced joint counts and acute phase markers. A key secondary goal is improving health-related quality of life (HRQoL). The treat-to-target (T2T) approach is based on the concept that HRQoL is maximised when synovitis is minimised through escalating drug treatment until patients achieve remission.3 SF-36 evaluated HRQoL in both trials. It measures HRQoL across eight domains: physical functioning (PF) —ability to perform physical activities; role physical (RP) —interference with work and other daily activities; bodily pain (BP); general health (GH)—how individuals evaluate their health; vitality (VT)—fatigue, pep and energy; social functioning (SF)—interference with social activities; role emotional (RE)—interference with work/ daily activities due to emotional problems; general mental health (MH)—nervousness and depression.12 These domains are scored 0–100; higher scores indicate better health. Domain scores can be normalised, z-transformed and combined into physical and mental component summary scores (PCS and MCS) providing summary measures of physical and mental health, rela- tive to population means of 50 (SD 10).13 The PCS posi- tively weights the physical domains and VT and negatively weights the other mental domains; the MCS positively weights the four mental domains and nega- tively weights the four physical domains. Weights used are factor score coefﬁcients derived from the general population.13 Minimal clinically important differences (MCIDs) in SF-36 domain and component summary scores are 5.0 and 2.5 units, respectively.14 HRQoL is a broad, multidimensional concept spanning physical and mental health, function, social support and socioeconomic status.4 Patients with RA in remission have substantially improved HRQoL compared with patients in higher disease activity states.5 However, intensive treatment and remission may not improve all aspects of HRQoL equally. Placebo-controlled trials evaluating treatment with tumour necrosis factor-α inhibitors (TNFis) combined with methotrexate suggest greater beneﬁts for physical than mental health in DMARD-inadequate responders (DMARD-IRs),6 with no impact on mental health in methotrexate-naïve patients.7 8 Focusing on minimising synovitis using intensive drug treatment risks overlooking aspects of HRQoL affected by pain, mood, anxiety, joint damage and deconditioning. g g Our goal was to establish if using intensive treatment to reduce synovitis and attain remission improves all aspects of HRQoL. Short-Form 6D Short Form 6D SF-36 scores were converted into the Short-Form 6D (SF-6D), a health utility measure ranging from 0.290 (worst health) to 1.000 (perfect health), using the Ara and Brazier algorithm.15 16 Its minimum important dif- ference (MID) is 0.041.17 Analysis time point We evaluated treatment effects on HRQoL over the ﬁrst 6 months in CARDERA and TACIT for two reasons. First, the COBRA corticosteroid regimen in CARDERA (initially 60 mg/day, tapered to 7.5 mg/day by weeks 7–28 and stopped by week 36)11 had maximal effects at 6 months. Second, in TACIT after 6-month cDMARD treatment, 41% of patients switched to TNFis; 6-month results speciﬁcally compared TNFi with cDMARDs. To ensure 6 months was not too early a time point to observe treatment effects on HRQoL, we undertook a sensitivity analysis looking at treatment effects on HRQoL over 12 months (see online supplementary table 1). METHODS Patients We studied patients in the Combination Anti-Rheumatic Drugs in Early RA (CARDERA) and TNFi versus com- bination intensive therapy with conventional DMARDs in established RA (TACIT) trials; their primary results have been reported.9 10 CARDERA recruited 467 patients with active early RA (<2 years duration) from 42 English centres. Patients were randomised to receive methotrexate alongside high-dose rapidly reducing corticosteroids,11 ciclosporin, placebo or both active treatments in a factorial design. Patients were assessed 6-monthly for 24 months. The impact of remission and DAS28 components on HRQoL was assessed at study end points (24 months in CARDERA; 12 months in TACIT) when remission rates were greatest. y TACIT recruited 205 patients with active established RA (>1 year duration) from 24 English centres. Patients were randomised to receive (a) TNFi (adalimumab, eta- nercept or inﬂiximab), with a TNFi switch after 6 months if DAS28(ESR) reduction was <1.2; (b) inten- sive cDMARDs, with TNFi initiated after 6 months if DAS28(ESR) reduction was <1.2. Patients were assessed 6-monthly for 12 months. Health-related quality of life Short-Form 36 We undertook secondary analyses of two randomised clinical trials of intensive treatment in patients with early and established active RA. HRQoL was captured using the Short-Form 36 (SF-36). We had three aims: (1) establishing how intensive treatment with high-dose tapering corticosteroids, TNFi and combin- ation DMARDs (cDMARDs) affects SF-36 domains; (2) evaluating the impact of remission and other disease activity score (DAS28)-deﬁned disease activity categories on SF-36 domains and (3) evaluating which DAS28 com- ponents have the strongest associations with HRQoL. INTRODUCTION Treating rheumatoid arthritis (RA) with disease-modifying antirheumatic drugs (DMARDs), biologics and corticosteroids is primarily intended to reduce synovitis. Efﬁcacy is assessed by improving composite measures such as European League Against Rheumatism (EULAR) response criteria,1 2 Treating rheumatoid arthritis (RA) with disease-modifying antirheumatic drugs (DMARDs), biologics and corticosteroids is primarily intended to reduce synovitis. Efﬁcacy is assessed by improving composite measures such as European League Against Rheumatism (EULAR) response criteria,1 2 Trial registration numbers: CARDERA was registered as ISRCTN 32484878. TACIT was registered as ISRCTN 37438295; pre-results. Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 RMD Open Missing data imputation In CARDERA, missing data had been imputed at all time points using last observation carried forward (LOCF) analysis.10 Missing data were imputed in 19% of patients at 24 months; an observed case analysis had excluded a signiﬁcant impact of the LOCF assumption on the study end points, which included PCS and MCS scores. For consistency across studies, we imputed missing TACIT data using LOCF (undertaken at 6 months in 5 and 18 patients for SF-36 domain scores and DAS28(ESR) components, respectively, and at 12 months in 15 and 16 patients for SF-36 domain scores and DAS28(ESR) components, respectively). We under- took an additional analysis using non-imputed TACIT Associations between DAS28 components and SF-36 In both trials, most patients were female, with a mean age in the sixth decade (table 1). As expected from the inclusion criteria for each trial, mean disease duration in TACIT (8.2 years) was higher than CARDERA (0.3 years). TACIT patients also had higher baseline DAS28(ESR) (mean 6.26) compared with CARDERA (mean 5.78), attributable to higher TJC and PtGA scores. Baseline health assessment questionnaire (HAQ) levels were slightly higher in TACIT (mean 1.85) than CARDERA (mean 1.59). PCS scores were lower in TACIT (mean 26.0) than CARDERA (mean 29.6); MCS scores were similar in both trials (CARDERA mean 39.8; TACIT mean 39.2). To minimise type I error from multiple testing (4 DAS28 components; 8 health domains), associations between DAS28(ESR) components and PCS and MCS were tested. Linear regression models used ﬁnal time point PCS and MCS scores as response variables, and SJC, TJC, ESR and PtGA as explanatory variables, adjusted for covariates (treatment, age, sex and disease duration). Model 1 tested each DAS28(ESR) component separately. Model 2 included all DAS28(ESR) components as explanatory variables. To ensure multicollinearity between DAS28(ESR) components was not an issue in model 2, variance inﬂation factors (VIFs) were calcu- lated for each predictor; VIF was <2 for all explanatory variables.19 Standardised β values were calculated enab- ling direct comparison of effect sizes of each DAS28 (ESR) component on PCS and MCS. Within-trial comparisons of baseline patient character- istics between treatment arms showed patients were well matched (table 1). Signiﬁcant differences were observed between treatment arms for some SF-36 domains: in CARDERA, SF, RE and MCS scores were signiﬁcantly higher in the active steroid group; in TACIT, SF was sig- niﬁcantly lower in the TNFi group. Normative SF-36 profiles Age-matched and gender-matched US normative scores (A/G norms) were generated for CARDERA and TACIT protocol populations using data published in SF-36 manuals and updates.20 It was not possible to use UK A/ G norms as these data are not publicly available, although existing studies have highlighted similarities in mean SF-36 domain scores between UK and US populations.21 22 Impact of remission and disease activity categories on SF-36 Mean SF-36 domain scores at the ﬁnal time point in CARDERA and TACIT were plotted on spydergrams stratiﬁed by (a) DAS28(ESR) activity category: remission (DAS28(ESR) <2.6), low disease activity (LDA; ≥2.6 to <3.2), moderate disease activity (MDA; 3.2–5.1), high disease activity (HDA; >5.1) and (b) remission versus non-remission according to each DAS28 component: tender joint count (TJC) ≤1, swollen joint count (SJC) ≤1, patient global assessment (PtGA) of disease activity on a 100 mm visual analogue scale ≤10, erythrocyte sedi- mentation rate (ESR) ≤20 mm/hour. These component cut-offs represent the preliminary American College of Rheumatology (ACR)/EULAR Boolean-based deﬁnition of RA remission for clinical trials.18 As C reactive protein (CRP) data were not available, a normal ESR level was considered indicative of acute phase response remission. Statistical software Analyses were performed in R (V.3.1.3). Statistical software Analyses were performed in R (V.3.1.3). Analyses were performed in R (V.3.1.3). Ethics, consent and permissions CARDERA (ISRCTN 32484878 and Research Ethics Committee (REC) reference: MREC (1) 99/04) and TACIT (ISRCTN 37438295 and REC reference: MREC Ref 07/Q0505/57) were approved by research ethics commit- tees. All patients provided informed written consent. y Impact of treatment on SF-36 Treatment effects were evaluated using linear regression models including the 6-month changes in each SF-36 domain and summary score as the response variable. An 2 Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 Rheumatoid arthritis data to ensure our ﬁndings were not biased by LOCF imputation (see online supplementary tables 2 and 3). data to ensure our ﬁndings were not biased by LOCF imputation (see online supplementary tables 2 and 3). unadjusted model included treatment (active vs placebo corticosteroids in CARDERA; TNFi vs cDMARD therapy in TACIT) as the explanatory variable. An adjusted model included treatment, baseline SF-36 domain/ summary score, age, sex and disease duration as explanatory variables. unadjusted model included treatment (active vs placebo corticosteroids in CARDERA; TNFi vs cDMARD therapy in TACIT) as the explanatory variable. An adjusted model included treatment, baseline SF-36 domain/ summary score, age, sex and disease duration as explanatory variables. Normative SF-36 profiles Impact of corticosteroids versus placebo on SF-36 domains In CARDERA, 6-month increases in mean scores exceed- ing MCID were reported in all eight SF-36 domains with active and placebo corticosteroids, with the exception of GH in the placebo arm (ﬁgure 1). Active corticosteroids resulted in signiﬁcant improvements in all physical health domains, relative to placebo (table 2). Over 6-months increases in PF, RP, BP, GH and PCS scores were estimated to be 7.97 (p<0.001), 7.34 (p=0.047), 8.14 (p<0.001), 5.00 (p=0.003) and 3.78 (p<0.001) units Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 3 RMD Open Table 1 Patient baseline characteristics CARDERA (n=467) TACIT (n=205) Characteristic Placebo (n=236) Steroids (n=231) p Value cDMARDs (n=104) TNFi (n=101) p Value Demographic Number (%) female 157 (67) 166 (72) 0.251 73 (70) 79 (78) 0.249 Age (years) 53.8 (13.6) 54.5 (11.5) 0.567 58.0 (12.9) 56.7 (11.0) 0.459 RA specific Disease duration (years) 0.3 (0.4) 0.4 (0.5) 0.082 7.3 (8.4) 9.2 (9.2) 0.126 DAS28 5.84 (1.27) 5.71 (1.30) 0.266 6.21 (0.92) 6.30 (0.81) 0.461 SJC 10.0 (6.5) 9.7 (6.1) 0.577 10.5 (6.1) 10.8 (6.7) 0.753 ESR 40.9 (28.9) 40.8 (29.6) 0.979 33.1 (26.1) 30.1 (22.8) 0.379 TJC 12.3 (7.4) 11.3 (7.8) 0.130 16.4 (7.1) 17.5 (6.7) 0.259 PtGA 54.3 (26.7) 55.6 (26.0) 0.582 68.1 (19.7) 68.2 (21.3) 0.988 HAQ 1.59 (0.68) 1.59 (0.69) 0.984 1.80 (0.59) 1.90 (0.67) 0.251 HRQoL PF 33.6 (24.5) 34.3 (24.0) 0.754 30.1 (22.6) 24.6 (20.9) 0.067 RP 14.4 (29.5) 17.9 (32.8) 0.233 14.9 (30.1) 12.4 (26.1) 0.521 BP 33.2 (21.0) 34.4 (21.4) 0.541 28.1 (16.3) 26.3 (17.8) 0.435 GH 44.4 (21.1) 46.4 (19.9) 0.305 35.8 (18.2) 31.4 (16.8) 0.074 VT 31.2 (20.3) 34.7 (20.7) 0.063 30.3 (21.4) 26.6 (19.0) 0.185 SF 48.3 (31.2) 54.4 (29.7) 0.030 50.2 (25.2) 42.1 (25.3) 0.022 RE 38.3 (43.7) 46.5 (44.9) 0.046 43.9 (44.9) 35.3 (44.9) 0.172 MH 59.5 (20.0) 61.6 (19.0) 0.252 61.9 (20.2) 58.8 (23.1) 0.305 PCS 29.6 (8.5) 29.6 (9.5) 0.969 26.5 (7.5) 25.5 (7.8) 0.356 MCS 38.3 (15.2) 41.3 (14.3) 0.025 40.8 (14.8) 37.6 (14.6) 0.120 Data given as mean (SD) unless otherwise stated; p values for continuous and categorical variables derived from t-tests and χ2 tests, respectively. SF-36 domains by DAS28 category In both trials, increases in all SF-36 domains were observed with reducing DAS28(ESR) disease activity cat- egories (ﬁgures 2 and 3). Remission provided the highest mean SF-36 domain scores in CARDERA and TACIT. Impact of corticosteroids versus placebo on SF-36 domains BP, bodily pain; cDMARDs, combination disease-modifying antirheumatic drugs; DAS28, disease activity score; ESR, erythrocyte sedimentation rate; GH, general health; HRQoL, health-related quality of life; MCS, mental component summary; MH, mental health; PCS, physical component summary; PF, physical functioning; PtGA, patient global assessment of disease activity; RA, rheumatoid arthritis; RE, role emotional; RP, role physical; SF, social functioning; SJC, swollen joint count; TJC, tender joint count; TNFi, tumour necrosis factor-α inhibitor; VT, vitality. signiﬁcance after adjusting for their baseline scores and other covariates. As in CARDERA, despite TNFi and cDMARD therapy improving SF-36 domains, HRQoL remained substantially below that of A/G norms in both treatment arms (ﬁgure 1). A sensitivity analysis evaluating the impact of TNFi versus cDMARDs on changes in SF-36 domain scores over 12 months showed no effect of TNFi versus cDMARDs (see online supplementary table S1). greater with corticosteroids than placebo, respectively, when evaluated using the adjusted linear regression model. In contrast, corticosteroids had no effect on any mental health domains when evaluated using unadjusted and adjusted linear regression models. Despite improving physical health, 6-month SF-36 pro- ﬁles remained substantially lower than those of A/G norms (ﬁgure 1). A sensitivity analysis evaluating the impact of corticosteroids on changes in SF-36 domain scores over 12 months showed no treatment effect of corticosteroids compared with placebo, suggesting the beneﬁts of high-dose tapering corticosteroids on phys- ical HRQoL are not sustained over time (see online supplementary table S1). respectively. BP, bodily pain; cDMARDs, combination disease-modifying antirheumatic drugs; DAS28, disease activity score; ESR, erythrocyte sedimentation rate; GH, general health; HRQoL, health-related quality of life; MCS, mental component summary; MH, mental health; PCS, physical component summary; PF, physical functioning; PtGA, patient global assessment of disease activity; RA, rheumatoid arthritis; RE, role emotional; RP, role physical; SF, social functioning; SJC, swollen joint count; TJC, tender joint count; TNFi, tumour necrosis factor-α inhibitor; VT, vitality. In TACIT, SJC, TJC and PtGA, but not ESR had signiﬁ- cant associations with PCS and MCS scores in model 1 (table 3); as in CARDERA, the largest effect size was observed with PtGA (PCS and MCS model 1 standar- dised β values of −0.43 and −0.41, respectively). Only PtGA retained a signiﬁcant association with PCS (p<0.001) and MCS (p<0.001) scores in model 2. RE, respectively. Mean scores in MH were 2.7 units higher. RE, respectively. Mean scores in MH were 2.7 units higher. SF-6D utility scores SF-6D scores improved in CARDERA and TACIT with all treatments. In CARDERA, SF-6D scores increased from 0.594 to 0.669 (change 0.076) over 6 months with active corticosteroids, and from 0.577 to 0.633 (change 0.057) with placebo. In TACIT, SF-6D scores increased from 0.548 to 0.629 (change 0.081) over 6 months with TNFi and 0.574 to 0.626 (change 0.052) with cDMARDs. Improvements in both treatment arms in both trials exceeded MID. Figure 1 Mean Short-Form 36 (SF-36) domain scores at baseline and 6 months in patients receiving active or placebo corticosteroids, cDMARDs or TNFis. (A) Patients with early RA in the CARDERA trial randomised to receive either active or placebo high-dose tapering corticosteroids (baseline placebo and 6 months placebo=mean SF-36 domain scores at 0 and 6 months in patients receiving placebo corticosteroids; baseline active and 6 months active=mean SF-36 domain scores at 0 and 6 months in patients receiving active corticosteroids; A/G norms=mean SF-36 domain scores in an age-matched and gender-matched normative US population). (B) Patients with established RA in the TACIT trial randomised to receive either cDMARD or TNFi (baseline cDMARD and 6 months cDMARD=mean SF-36 domain scores at 0 and 6 months in patients receiving cDMARDs; baseline TNFi and 6 months TNFi=mean SF-36 domain scores at 0 and 6 months in patients receiving TNFi; A/G norms=mean SF-36 domain scores in an age-matched and gender-matched normative US population). BP, bodily pain; cDMARD, combination disease-modifying antirheumatic drug; GH, general health; MH, mental health; PF, physical functioning; RA, rheumatoid arthritis; RE, role emotional; RP, role physical; SF, social functioning; TNFi, tumour necrosis factor-α inhibitor; VT, vitality. RE, respectively. Mean scores in MH were 2.7 units higher. SF-36 domains by DAS28 component remission In both CARDERA and TACIT, PtGA-deﬁned remission resulted in the highest mean domain scores and ESR-deﬁned remission resulted in the lowest mean domain scores, compared with remission deﬁned by other DAS28(ESR) components (ﬁgures 2 and 3). SF-36 domains by DAS28 component remission In both CARDERA and TACIT, PtGA-deﬁned remission resulted in the highest mean domain scores and ESR-deﬁned remission resulted in the lowest mean domain scores, compared with remission deﬁned by other DAS28(ESR) components (ﬁgures 2 and 3). Association between DAS28 components and SF-36 summary scores In CARDERA, all four DAS28(ESR) components had signiﬁcant associations with PCS and MCS scores at 24 months in both models (model 1 testing each compo- nent separately; model 2 testing all components in same model) except TJC, which did not have a signiﬁcant association with PCS in model 2 (p=0.977; table 3). PtGA had the largest effect on PCS (model 1 standar- dised β=−0.57) and MCS (model 1 standardised β= −0.44) scores. Impact of TNFi versus cDMARDs on SF-36 domains pact of TNFi versus cDMARDs on SF-36 domains In TACIT, 6-month increases in mean scores exceeding MCID were reported in all eight SF-36 domains with cDMARDs and TNFi (ﬁgure 1). A total of 6 months of TNFi therapy provided a signiﬁcantly greater improve- ment in PCS compared with cDMARD therapy (table 2); the increase in PCS was estimated to be 3.04 units greater (p=0.034) with TNFi compared with cDMARDs when evaluated using the adjusted model. Signiﬁcantly greater improvements were reported in PF, GH and VT in the unadjusted model, although these domains lost In CARDERA (ﬁgure 2), patients in remission reported mean scores 9.6, 2.9, 3.9 and 5.0 units lower than A/G norms for PF, RP, GH and RE, respectively. Mean scores in BP, VT, SF and MH were 7.9, 1.6, 3.1 and 4.8 units higher, respectively. In TACIT (ﬁgure 3), patients in remission reported mean scores 19.6, 9.8, 3.3, 13.3, 5.7, 3.1 and 0.7 units lower than A/G norms for PF, RP, BP, GH, VT, SF and 4 Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 Rheumatoid arthritis DISCUSSION BP, bodily pain; cDMARD, combination disease-modifying antirheumatic drug; GH, general health; MCS, mental component summary; MH, mental health; PCS, physical component summary; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; SF-36, Short-Form 36; VT, vitality. treatment reference group in TACIT analysis. BP, bodily pain; cDMARD, combination disease-modifying antirheumatic drug; GH, general health; MCS, mental component summary; MH, mental health; PCS, physical component summary; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; SF-36 Short-Form 36; VT vitality Figure 2 Mean Short-Form 36 (SF-36) domain scores stratified by DAS28-defined disease activity category, and DAS28-component defined remission in CARDERA. (A) Mean SF-36 domain scores in patients in remission (DAS28<2.6), low disease activity (LDA; ≥2.6–<3.2), moderate disease activity (MDA; 3.2–5.1) and high disease activity (HDA; >5.1). (B) Mean scores in patients meeting/not meeting tender joint count (TJC)-defined remission (TJC≤1). (C) Mean scores in patients meeting/ not meeting swollen joint count (SJC)-defined remission (SJC≤1). (D) Mean scores in patients meeting/not meeting patient global assessment (PtGA) of disease activity-defined remission (100 mm PtGA≤10). (E) Mean scores in patients meeting/not meeting erythrocyte sedimentation rate (ESR)-defined remission (ESR≤20 mm/hour). These DAS28-component cut-offs represent the preliminary American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) Boolean-based definition of rheumatoid arthritis (RA) remission for clinical trials. A/G norms=mean SF-36 domain scores in age-matched and gender-matched individuals from the normative US population. Space between grid lines is 10 units, which represents twice minimal clinically important differences (MCIDs). BP, bodily pain; GH, general health; MH, mental health; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; VT, vitality. Figure 2 Mean Short-Form 36 (SF-36) domain scores stratified by DAS28-defined disease activity category, and DAS28-component defined remission in CARDERA. (A) Mean SF-36 domain scores in patients in remission (DAS28<2.6), low disease activity (LDA; ≥2.6–<3.2), moderate disease activity (MDA; 3.2–5.1) and high disease activity (HDA; >5.1). (B) Mean scores in patients meeting/not meeting tender joint count (TJC)-defined remission (TJC≤1). (C) Mean scores in patients meeting/ not meeting swollen joint count (SJC)-defined remission (SJC≤1). (D) Mean scores in patients meeting/not meeting patient global assessment (PtGA) of disease activity-defined remission (100 mm PtGA≤10). (E) Mean scores in patients meeting/not meeting erythrocyte sedimentation rate (ESR)-defined remission (ESR≤20 mm/hour). These DAS28-component cut-offs represent the preliminary American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) Boolean-based definition of rheumatoid arthritis (RA) remission for clinical trials. DISCUSSION We have undertaken secondary analyses of two com- pleted randomised clinical trials in early and established RA. Our ﬁrst aim was to evaluate the effect of intensive treatments—high-dose corticosteroids versus placebo in early RA, and cDMARDs versus TNFis in established RA —on SF-36 domains during the ﬁrst 6 months of treat- ment. With all active treatments, there were improve- ments in all eight SF-36 domains, which exceeded MCIDs (5 units). However, in early RA, corticosteroids only provided improvements beyond placebo in physical but not mental SF-36 domains. Furthermore, in estab- lished RA, there were similar improvements with cDMARDs and TNFis. These results show that in these two trials intensive treatments resulted in limited add- itional beneﬁts on HRQoL compared with less intensive treatments. Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 5 RMD Open Table 2 Effect of intensive treatment on 6-month changes in SF-36 domains SF-36 domain CARDERA TACIT Model 1: Unadjusted Model 2: Adjusted Model 1: Unadjusted Model 2: Adjusted β (SE) p Value β (SE) p Value β (SE) p Value β (SE) p Value PF 7.55 (2.27) <0.001 7.97 (2.18) <0.001 8.61 (3.83) 0.026 5.43 (3.62) 0.135 RP 5.52 (3.94) 0.162 7.34 (3.69) 0.047 5.05 (6.28) 0.423 1.55 (5.78) 0.789 BP 7.59 (2.34) 0.001 8.14 (2.20) <0.001 5.93 (3.29) 0.073 3.97 (2.89) 0.170 GH 4.52 (1.78) 0.011 5.00 (1.68) 0.003 7.38 (2.97) 0.014 4.16 (2.64) 0.117 VT 1.39 (1.86) 0.454 2.60 (1.76) 0.140 7.46 (3.28) 0.024 4.72 (2.91) 0.107 SF 1.86 (2.69) 0.488 4.57 (2.42) 0.060 6.01 (3.94) 0.129 0.11 (3.47) 0.975 RE −1.51 (4.52) 0.739 4.36 (3.82) 0.254 1.72 (7.72) 0.824 −6.93 (6.25) 0.269 MH −0.23 (1.77) 0.898 0.66 (1.61) 0.684 1.09 (3.45) 0.753 −2.25 (2.76) 0.415 PCS 3.84 (1.02) <0.001 3.78 (0.97) <0.001 3.70 (1.51) 0.015 3.04 (1.42) 0.034 MCS −1.33 (1.32) 0.316 0.36 (1.16) 0.757 0.85 (2.29) 0.710 −1.99 (1.91) 0.301 Adjusted model includes following covariates: age, gender, disease duration and baseline SF-36 domain/summary score; cDMARD used as treatment reference group in TACIT analysis. BP, bodily pain; cDMARD, combination disease-modifying antirheumatic drug; GH, general health; MCS, mental component summary; MH, mental health; PCS, physical component summary; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; SF-36, Short-Form 36; VT, vitality. Adjusted model includes following covariates: age, gender, disease duration and baseline SF-36 domain/summary score; cDMARD used as treatment reference group in TACIT analysis. DISCUSSION A/G norms=mean SF-36 domain scores in age-matched and gender-matched individuals from the normative US population. Space between grid lines is 10 units, which represents twice minimal clinically important differences (MCIDs). BP, bodily pain; GH, general health; MH, mental health; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; VT, vitality. 6 Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 Rheumatoid arthritis Figure 3 Mean Short-Form 36 (SF-36) domain scores stratified by DAS28-defined disease activity category, and DAS28-component defined remission in TACIT. (A) Mean SF-36 domain scores in patients in remission (DAS28<2.6), low disease activity (LDA; ≥2.6–<3.2), moderate disease activity (MDA; 3.2–5.1) and high disease activity (HDA; >5.1). (B) Mean scores in patients meeting/not meeting tender joint count (TJC)-defined remission (TJC≤1). (C) Mean scores in patients meeting/ not meeting swollen joint count (SJC)-defined remission (SJC≤1). (D) Mean scores in patients meeting/not meeting patient global assessment (PtGA) of disease activity-defined remission (100 mm PtGA≤10). (E) Mean scores in patients meeting/not meeting erythrocyte sedimentation rate (ESR)-defined remission (ESR≤20 mm/hour). These DAS28-component cut-offs represent the preliminary American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) Boolean-based definition of rheumatoid arthritis (RA) remission for clinical trials. A/G norms=mean SF-36 domain scores in an age-matched and gender-matched normative US population. Space between grid lines is 10 units, which represents twice minimal clinically important differences (MCIDs). BP, bodily pain; GH, general health; MH, mental health; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; VT, vitality. Rheumatoid arthritis Figure 3 Mean Short-Form 36 (SF-36) domain scores stratified by DAS28-defined disease activity category, and DAS28-component defined remission in TACIT. (A) Mean SF-36 domain scores in patients in remission (DAS28<2.6), low disease activity (LDA; ≥2.6–<3.2), moderate disease activity (MDA; 3.2–5.1) and high disease activity (HDA; >5.1). (B) Mean scores in patients meeting/not meeting tender joint count (TJC)-defined remission (TJC≤1). (C) Mean scores in patients meeting/ not meeting swollen joint count (SJC)-defined remission (SJC≤1). (D) Mean scores in patients meeting/not meeting patient global assessment (PtGA) of disease activity-defined remission (100 mm PtGA≤10). (E) Mean scores in patients meeting/not meeting erythrocyte sedimentation rate (ESR)-defined remission (ESR≤20 mm/hour). These DAS28-component cut-offs represent the preliminary American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) Boolean-based definition of rheumatoid arthritis (RA) remission for clinical trials. A/G norms=mean SF-36 domain scores in an age-matched and gender-matched normative US population. DISCUSSION Space between grid lines is 10 units, which represents twice minimal clinically important differences (MCIDs). BP, bodily pain; GH, general health; MH, mental health; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; VT, vitality. Our second aim was to evaluate the impact of remission and other DAS28-deﬁned disease activity categories on SF-36 domains at trial end points. Patients in remission had better HRQoL proﬁles than patients in higher disease activity states. However, scores in PF, RP and GH in both trials remained below normative values, with large deﬁcits observed in established RA patients enrolled to TACIT. The IMPROVED trial suggested that in patients with early inﬂammatory arthritis, attaining prompt remission normalises HRQoL.23 Overall, these ﬁndings suggest that in patients with established RA and in some patients with early RA, attaining remission does not normalise HRQoL, but attaining early remission in patients with early inﬂammatory arthritis may be sufﬁ- cient to achieve this goal. Our third aim was to assess which DAS28 components had the strongest associations with HRQoL at trial end points. In CARDERA and TACIT, PtGA had the strongest association with PCS and MCS scores; associations with SJC and ESR levels were weaker. Although different types of patient-reported outcomes such as PtGA and SF-36 component summary scores may be expected to be inter-related, our ﬁnding that PtGA was most strongly associated with physical and mental HRQoL compared with other DAS28 components is of clinical importance. It suggests that in order to optimise HRQoL in RA, it might be helpful to place an increased focus on improv- ing PtGA scores. In routine practice, the PtGA is not a key driver of rheumatologists’ treatment decisions, with Dutch24 and Canadian25 studies of RA patients demon- strating that SJC and physician global assessment of 7 Scott IC, et al. RMD Open 2016;2:e000270. DISCUSSION doi:10.1136/rmdopen-2016-000270 RMD Open Table 3 Associations between DAS28 components and SF-36 PCS and MCS at final trial time points DAS28 component CARDERA TACIT Model 1: DAS28 components tested individually Model 2: DAS28 components tested in same model Model 1: DAS28 components tested individually Model 2: DAS28 components tested in same model Standardised β (SE) p Value Standardised β (SE) p Value Standardised β (SE) p Value Standardised β (SE) p Value PCS SJC −0.45 (0.04) <0.001 −0.19 (0.05) <0.001 −0.24 (0.07) <0.001 −0.06 (0.08) 0.412 TJC −0.33 (0.04) <0.001 0.00 (0.05) 0.977 −0.33 (0.07) <0.001 −0.08 (0.09) 0.370 ESR −0.13 (0.05) 0.005 −0.08 (0.04) 0.036 −0.02 (0.07) 0.804 0.03 (0.07) 0.676 PtGA −0.57 (0.04) <0.001 −0.45 (0.05) <0.001 −0.43 (0.06) <0.001 −0.36 (0.08) <0.001 MCS SJC −0.29 (0.04) <0.001 −0.12 (0.06) 0.029 −0.16 (0.07) 0.024 0.05 (0.08) 0.527 TJC −0.13 (0.05) 0.004 0.15 (0.05) 0.003 −0.34 (0.07) <0.001 −0.16 (0.09) 0.080 ESR −0.14 (0.05) 0.004 −0.12 (0.04) 0.008 −0.04 (0.07) 0.578 0.00 (0.07) 0.974 PtGA −0.44 (0.04) <0.001 −0.43 (0.05) <0.001 −0.41 (0.06) <0.001 −0.33 (0.08) <0.001 All linear regression models include age, gender, disease duration and treatment as covariates; CARDERA model includes the 24-month PCS and MCS as the response variables; TACIT model includes the 12-month PCS and MCS as the response variables; the VIFs were <2 for all explanatory variables in model 2. DAS28, disease activity score; ESR, erythrocyte sedimentation rate; MCS, mental component summary; PCS, physical component summary; PtGA, patient global assessment of disease activity; SF-36, Short-Form 36; SJC, swollen joint count; TJC, tender joint count. All linear regression models include age, gender, disease duration and treatment as covariates; CARDERA model includes the 24-month PCS and MCS as the response variables; TACIT model includes the 12-month PCS and MCS as the response variables; the VIFs were <2 for all explanatory variables in model 2. DAS28, disease activity score; ESR, erythrocyte sedimentation rate; MCS, mental component summary; PCS, physical component summary; PtGA, patient global assessment of disease activity; SF-36, Short-Form 36; SJC, swollen joint count; TJC, tender joint count. to conﬁrm the relative beneﬁts of different intensive treatment regimens. disease activity to be central drivers of treatment escal- ation. DISCUSSION As PtGA does not purely reﬂect how active a patient considers their disease to be—a previous study by Studenic et al26 showed that 76% and 1.3% of the variability in PtGA was explained by pain and physical function, respectively—treatment intensity decisions should not be based on PtGA alone. However, there is likely to be a role for more holistic approaches to man- aging RA patients that move beyond purely targeting remission, and consider treating other disease impacts such as pain, fatigue and mood, which are likely to improve PtGA scores. In CARDERA and TACIT SF-36, health proﬁles were superior in patients attaining remission, compared with those attaining LDA. This replicates previous work by Radner et al,28 who reported that in 356 RA patients with longstanding RA from Austria, remission gave superior SF-36 proﬁles to LDA. Interestingly, as in our analysis, Radner et al also found most SF-36 domains in RA patients in remission were below general population levels. Their ﬁndings therefore support our view that additional management strat- egies are needed to normalise HRQoL in addition to attaining remission. p The failure of TNFi to improve individual SF-36 domains beyond cDMARDs in TACIT was of interest. Although limited by a modest sample size, TACIT pro- vided only little evidence that 6-months TNFi with DMARDs was superior to cDMARDs at improving HRQoL. Previous trials of TNFi with methotrexate versus methotrexate monotherapy in DMARD-IR patients show signiﬁcant improvements in PCS and MCS scores with biologics, although the effects are greater on physical than mental health.6 To our knowl- edge, only one other study has compared the effect of TNFi and cDMARDs on HRQoL. A secondary analysis of the BeST study showed 12-month EuroQol scores were similar with sequential DMARD monotherapy, step-up combination DMARDs, initial combination DMARDs with high-dose tapering prednisolone, and methotrexate with inﬂiximab, although prompter improvements were seen in the latter two groups.27 TACIT and BeST suggest TNFis have only limited bene- ﬁts on HRQoL compared with cDMARDs. There is insufﬁcient evidence, however, to know if this is correct in all clinical situations; further large trials are needed Our study has several strengths. As a secondary ana- lysis of two clinical trials, assessments were standardised. Its main ﬁndings were similar in early and established RA patients receiving different intensive treatments. Patients were recruited from many English centres, which followed consistent healthcare approaches. It also has several limitations. REFERENCES Drugs 2010;70:121–45. 7. Strand V, Rentz AM, Cifaldi MA, et al. Health-related quality of life outcomes of adalimumab for patients with early rheumatoid arthritis: results from a randomized multicenter study. J Rheumatol 2012;39:63–72. 8. Kekow J, Moots RJ, Emery P, et al. Patient-reported outcomes improve with etanercept plus methotrexate in active early rheumatoid arthritis and the improvement is strongly associated with remission: the COMET trial. Ann Rheum Dis 2010;69:222–5. 9. Scott DL, Ibrahim F, Farewell V, et al. Tumour necrosis factor inhibitors versus combination intensive therapy with conventional disease modifying anti-rheumatic drugs in established rheumatoid arthritis: TACIT non-inferiority randomised controlled trial. BMJ 2015;350:h1046. 10. Choy EHS, Smith CM, Farewell V, et al. Factorial randomised controlled trial of glucocorticoids and combination disease modifying drugs in early rheumatoid arthritis. Ann Rheum Dis 2008;67:656–63. 11. Boers M, Verhoeven AC, Markusse HM, et al. Randomised comparison of combined step-down prednisolone, methotrexate and sulphasalazine with sulphasalazine alone in early rheumatoid arthritis. Lancet 1997;350:309–18. 12. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30:473–83. Contributors ICS, DLS and FI conceived and designed the study. FI managed the trial data. ICS and VS performed the statistical analysis. ICS, VS, CML and DLS interpreted the data. ICS, VS and DLS drafted the manuscript. All authors revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript. Contributors ICS, DLS and FI conceived and designed the study. FI managed the trial data. ICS and VS performed the statistical analysis. ICS, VS, CML and DLS interpreted the data. ICS, VS and DLS drafted the manuscript. All authors revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript. 13. Ware JE Jr, Kosinski M, Dewey J. How to score version 2 of the SF-36 Health Survey (standard and acute forms). Lincoln, RI: QualityMetric, Inc., 2005. y 14. Lubeck DP. Patient-reported outcomes and their role in the assessment of rheumatoid arthritis. Pharmacoeconomics 2004;22:27–38. 15. Ara R, Brazier J. Deriving an algorithm to convert the eight mean SF-36 dimension scores into a mean EQ-5D preference-based score from published studies (where patient level data are not available). Value Health 2008;11:1131–43. REFERENCES Funding This work was supported by the National Institute for Health Research (NIHR) (Clinical Lectureship to ICS) and the NIHR Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London. It also presents research funded by the NIHR Programme Grants for Applied Research (http://www.ccf.nihr.ac.uk/PGfAR/ Pages/Home.aspx) on ‘Treatment Intensities and Targets in Rheumatoid Arthritis Therapy: Integrating Patients’ And Clinicians’ Views—The TITRATE Programme (RP-PG-0610-10066)’. ) ; 16. Ara R, Brazier J. Predicting the short form-6D preference-based index using the eight mean short form-36 health dimension scores: estimating preference-based health-related utilities when patient level data are not available. Value Health 2009;12:346–53. 17. Walters SJ, Brazier JE. Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Qual Life Res 2005;14:1523–32. 18. Felson DT, Smolen JS, Wells G, et al. American College of Rheumatology/European League Against Rheumatism provisional definition of remission in rheumatoid arthritis for clinical trials. Arthritis Rheum 2011;63:573–86. Disclaimer This article presents independent research funded by the National Institute for Health Research (NIHR). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. The funders had no role in the study design, data collection and analysis, data interpretation, the writing of the manuscript or the decision to submit the manuscript for publication. ; 19. Zuur AF, Ieno EN, Elphick CS. A protocol for data exploration to avoid common statistical problems. Methods Ecol Evol 2010;1:3–14. 20. Ware JE, Snow KK, Kosinski M, et al. SF-36 health survey manual and interpretation guide. Boston, MA: The Health Institute, 1993. 21. Jenkinson C, Layte R, Lawrence K. Development and testing of the Medical Outcomes Study 36-Item Short Form Health Survey summary scale scores in the United Kingdom. Results from a large-scale survey and a clinical trial. Med Care 1997;35:410–16. Competing interests None declared. Ethics approval Both trials had Research Ethical Committee approval (CARDERA reference MREC (1) 99/04; TACIT reference MREC Ref 07/Q0505/57). 22. Jenkinson C. Comparison of UK and US methods for weighting and scoring the SF-36 summary measures. J Public Health Med 1999;21:372–6. Provenance and peer review Not commissioned; externally peer reviewed. Data sharing statement No additional data are available. 23. Heimans L, Wevers-de Boer KV, Koudijs KK, et al. Health-related quality of life and functional ability in patients with early arthritis during remission steered treatment: results of the IMPROVED study. REFERENCES remission could reﬂect several underlying drivers. Loss of lean muscle mass, which is linked with RA disability,30 remission could reﬂect several underlying drivers. Loss of lean muscle mass, which is linked with RA disability,30 remission could reﬂect several underlying drivers. Loss of lean muscle mass, which is linked with RA disability,30 may occur rapidly in active RA and will not be reversed with drug treatment. Joint damage, though less common in contemporary RA cohorts, has detrimental effects on physical function31 and cannot be reversed by drug treatment. Persistent pain may also impair physical health. Finally, clinical remission may be insensitive at detecting low-level synovitis and inﬂammation, which could affect physical HRQoL. 1. Felson DT, Anderson JJ, Boers M, et al. American College of Rheumatology. Preliminary definition of improvement in rheumatoid arthritis. Arthritis Rheum 1995;38:727–35. ; 2. Fransen J, van Riel PL. The Disease Activity Score and the EULAR response criteria. Clin Exp Rheumatol 2005;23:S93–9. 3. Smolen JS, Aletaha D, Bijlsma JW, et al. Treating rheumatoid arthritis to target: recommendations of an international task force. Ann Rheum Dis 2010;69:631–7. ; 4. Centers for Disease Control and Prevention (CDC). Health-Related Quality of Life (HRQOL). 2011 (cited 16 July 2015). http://www.cdc. gov/hrqol/concept.htm g q p 5. Linde L, Sorensen J, Ostergaard M, et al. Does clinical remission lead to normalization of EQ-5D in patients with rheumatoid arthritis and is selection of remission criteria important? J Rheumatol 2010;37:285–90. p y Q Patient surveys suggest current RA management does not fully address their needs and expectations.32 33 The ‘RAISE patient needs’ survey found few patient–phys- ician consultations discussed HRQoL.32 The ‘Good Days Fast’ and ‘Getting to Your Destination Faster’ surveys found most patients rated ‘having a good day’ as their preferred target for RA management; being free of fatigue and pain often characterised ‘good days’.33 In these surveys, pain was a prevalent problem for patients with RA. Our ﬁndings suggest that attaining remission, though crucial to improving RA outcomes, is insufﬁcient by itself to entirely normalise HRQoL in active RA, par- ticularly in established disease. New ways are needed to identify and treat speciﬁcally impaired areas of HRQoL —including pain and fatigue—as an adjunct to treat- ments that reduce synovitis assessed by DAS28. Potential options include using psychological approaches, increas- ing exercise and improving pain management. 6. Strand V, Singh JA. Newer biological agents in rheumatoid arthritis: impact on health-related quality of life and productivity. DISCUSSION As a post-hoc analysis, it did not test a prespeciﬁed hypothesis. It was restricted to asses- sing treatment impacts in active RA, limiting its general- isability to all RA populations. It focused on 6-month and 12-month HRQoL changes; longer time frames could show greater beneﬁts. It used a generic HRQoL assessment (SF-36); disease-speciﬁc measures like RAQoL29 may better capture treatment effects. TACIT was a non-inferiority trial comparing cDMARDs with TNFi strategies; our analysis was underpowered to detect small improvements in SF-36.9 Finally, short-term high- dose steroids are not widely used in current practice. Our key ﬁnding that physical HRQoL remained impaired even when intensive treatment achieved 8 Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 Rheumatoid arthritis remission could reﬂect several underlying drivers. Loss of lean muscle mass, which is linked with RA disability,30 may occur rapidly in active RA and will not be reversed with drug treatment. Joint damage, though less common in contemporary RA cohorts, has detrimental effects on physical function31 and cannot be reversed by drug treatment. Persistent pain may also impair physical health. Finally, clinical remission may be insensitive at detecting low-level synovitis and inﬂammation, which could affect physical HRQoL. REFERENCES Arthritis Res Ther 2013;15:R173. Open Access This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http:// creativecommons.org/licenses/by/4.0/ 24. van Hulst LT, Kievit W, van Bommel R, et al. Rheumatoid arthritis patients and rheumatologists approach the decision to escalate care differently: results of a maximum difference scaling experiment. Arthritis Care Res 2011;63:1407–14. 9 Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270 RMD Open RMD Open 25. Pyne L, Bykerk VP, Boire G, et al. Increasing treatment in early rheumatoid arthritis is not determined by the disease activity score but by physician global assessment: results from the CATCH study. J Rheumatol 2012;39:2081–7. 26. Studenic P, Radner H, Smolen JS, et al. Discrepancies between patients and physicians in their perceptions of rheumatoid arthritis disease activity. Arthritis Rheum 2012;64:2814–23. 27. Allaart CF, Breedveld FC, Dijkmans BA. Treatment of recent-onset rheumatoid arthritis: lessons from the BeSt study. J Rheumatol 2007;80:25–33. 28. Radner H, Smolen JS, Aletaha D. Remission in rheumatoid arthritis: benefit over low disease activity in patient-reported outcomes and costs. Arthritis Res Ther 2014;16:R56. 29. Tijhuis GJ, de Jong Z, Zwinderman AH, et al. The validity of the Rheumatoid Arthritis Quality of Life (RAQoL) questionnaire. Rheumatology (Oxford) 2001;40:1112–19. 30. Giles JT, Bartlett SJ, Andersen RE, et al. Association of body composition with disability in rheumatoid arthritis: impact of appendicular fat and lean tissue mass. Arthritis Rheum 2008;59:1407–15. 25. Pyne L, Bykerk VP, Boire G, et al. Increasing treatment in early rheumatoid arthritis is not determined by the disease activity score but by physician global assessment: results from the CATCH study. J Rheumatol 2012;39:2081–7. ; 26. Studenic P, Radner H, Smolen JS, et al. Discrepancies between patients and physicians in their perceptions of rheumatoid arthritis disease activity. Arthritis Rheum 2012;64:2814–23. 31. Drossaers-Bakker KW, de Buck M, van Zeben D, et al. Long-term course and outcome of functional capacity in rheumatoid arthritis: the effect of disease activity and radiologic damage over time. Arthritis Rheum 1999;42:1854–60. y ; 27. Allaart CF, Breedveld FC, Dijkmans BA. Treatment of recent-onset rheumatoid arthritis: lessons from the BeSt study. J Rheumatol 2007;80:25–33. 32. McInnes IB, Combe B, Burmester G. Understanding the patient perspective—results of the Rheumatoid Arthritis: Insights, Strategies & Expectations (RAISE) patient needs survey. Clin Exp Rheumatol 2013;31:350–7. 28. Radner H, Smolen JS, Aletaha D. Remission in rheumatoid arthritis: benefit over low disease activity in patient-reported outcomes and costs. Arthritis Res Ther 2014;16:R56. 33. Strand V, Wright GC, Bergman MJ, et al. Patient expectations and perceptions of goal-setting strategies for disease management in rheumatoid arthritis. J Rheumatol 2015; 42:2046–54. ; 29. Tijhuis GJ, de Jong Z, Zwinderman AH, et al. The validity of the Rheumatoid Arthritis Quality of Life (RAQoL) questionnaire. Rheumatology (Oxford) 2001;40:1112–19. 10 Scott IC, et al. RMD Open 2016;2:e000270. doi:10.1136/rmdopen-2016-000270
https://openalex.org/W4391329615	https://jrssem.publikasiindonesia.id/index.php/jrssem/article/download/597/1315	English	null	Development of an Islamic-Based Independent Curriculum Based on the National Curriculum	Journal Research of Social Science, Economics, and Management/Journal Research of Social Science, Economics and Management	2,024	cc-by-sa	4,734	JRSSEM 2023, Vol. 03, No. 06, 1336 – 1344 E-ISSN: 2807 - 6311, P-ISSN: 2807 - 6494 JRSSEM 2023, Vol. 03, No. 06, 1336 – 1344 E-ISSN: 2807 - 6311, P-ISSN: 2807 - 6494 JRSSEM 2023, Vol. 03, No. 06, 1336 – 1344 E-ISSN: 2807 - 6311, P-ISSN: 2807 - 6494 DEVELOPMENT OF AN ISLAMIC-BASED INDEPENDENT CURRICULUM BASED ON THE NATIONAL CURRICULUM Daryatmo Rahman1 R. Madhakomala2 Universitas Negeri Jakarta1,2 Email: abuathif88412@gmail.com1, madhakomala@live.com2 Coresspondence: abuathif88412@gmail.com Daryatmo Rahman1 R. Madhakomala2 Universitas Negeri Jakarta1,2 Email: abuathif88412@gmail.com1, madhakomala@live.com2 Coresspondence: abuathif88412@gmail.com ABSTRACT: The background of this research is a study of how to develop an Islamic- based curriculum in informal education in order to provide options and solutions to create an Islamic-based curriculum to develop quality education. The purpose of this research is to develop a curriculum and meet the needs of the community in Islamic educational institutions. In this study, researchers used a qualitative approach while the type of research approach is descriptive with a case study method, which is intended to be used by researchers to describe and explain the development of an independent Islamic-based curriculum based on the national curriculum. The result of this study is that from the changes and developments in the existing curriculum, it turns out that it still does not reach the ideal level to be applied in all educational institutions, so this will have an impact that the curriculum that applies in this country cannot be fully applied in every educational institution, so the development of an independent Islamic-based curriculum will have an influence in improving Islamic education and affecting the fulfillment of community needs for Islamic knowledge. Keywords: Islamic-Based Curriculum, Education, National Curriculum 1337\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum models that were very varied, each of these institutions had their own advantages in carrying out the mandate of educating students. memorizing short letters in the Qur'an, learning the basics of understanding tawhid, adab and morals to deeper sciences. In this case, Islamic-based education is needed to meet the needs of people who are Muslim, because along with the development of the times there are also situations and conditions that exist in the community, educational institutions must compile and develop an ideal curriculum in accordance with what is desired by the community, including an Islam-based curriculum. Because this will have a positive impact and influence, especially on students, parents and society in general, if without the development of an ideal curriculum by what is needed, students will not get the right learning target, then the quality of education will decrease and far from quality. The mosque is the center of various kinds of activities both religious activities and social activities and mu'amalah. Since the beginning of the development of Islam in Indonesia, mosques have long been used as Islamic educational institutions even from the time of the Prophet Muhammad - Sallallahu alaihi wa sallam-, and companions, tabi'in, tabi‟ tabi'in even now, mosques and surau are used as Islamic educational institutions because this place is a very strategic place to study science both religious, social and general sciences. Pesantren is a complex educational institution, in which there are mosques, schools/madrassahs, santri, kiai, they carry out educational activities, guidance, and direction in pesantren almost approximately 24 hours, and the percentage of teaching is more Islamic religious education sciences including santrian activities, as for other general sciences such as: mathematics, physics, and others, studied also during teaching and learning activities in formal education in schools/madrasah in pesantren aforementioned. The curriculum cannot be separated from the pursuit of targets that will make students able to understand various learning materials easily. The curriculum also serves as an indicator of the success of learning activities and is used as a clear benchmark regarding the process of teaching activities and a benchmark for learning materials that must be given to students so that it becomes a reference for learning. With the curriculum, students can find out what material they must learn and understand then parents also understand more about their children's education, and help them determine their learning patterns. INTRODUCTION Based on this data, education has an important role in educating and fostering people in accordance with their religion and beliefs, especially people who are Muslim certainly have special attention in education based on sharia or Islamic provisions. Islamic educational institutions in Indonesia are very much starting from simple to more complete and modern stages such as; surau, mosques, then became pesantren, madrasas, Islamic schools and finally emerged many institutional Indonesia is the country with the largest Muslim population in the world. Based on a report by The Royal Islamic Strategic Studies Centre (RISSC) or MABDA entitled The Muslim 500 2022 edition, there are 231.06 million Indonesians who are Muslims, the rest are Christians (Catholics and Protestants), Hindus, Buddhists, and Confucians. All of them coexist and tolerate each other (Mayasari, 2023). https://jrssem.publikasiindonesia.id/index.php/jrssem/index DOI: 10.59141/jrssem.v3i06.301 1337\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum 1337\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum Daryatmo Rahman1 R. Madhakomala2 \| 1338 themselves to face challenges in the future. position of the curriculum, it must be kept in mind that it (curriculum) is a tool to achieve goals. Based on the background stated above, the purpose of writing literature that will be studied and discussed in this article is to see the extent of the importance of developing an Islamic- based independent curriculum in informal educational institutions that contribute to students, parents, and society. The term "Curriculum" has many definitions and is formulated by experts, each defining according to the core emphasis and views of the expert. The curriculum comes from Greek, namely "curir" which means runner, and "curere" which means a place to gallop. In the past, this term was used in the world of sports. The curriculum based on the term is defined as "The distance that must be traveled by a runner from start to finish to get a medal or award". This understanding is then adapted into the world of education and is interpreted as "A number of subjects that must be taken by a student from the beginning to the end of the program in order to obtain a diploma". In the book entitled "Curriculum Development and Development in Schools" by Dr. Nana Sudjana, it is stated that the definition of curriculum is a collection of intentions and expectations contained in the form of educational programs which are then implemented and implemented by teachers in the school concerned (Elyana & Das, 2022). In the Law on the National Education System No. 20 of 2003 article 1 point 19 stated, "curriculum is a set of arrangements and plans regarding objectives, content, and subject matter as well as ways used as guidelines for learning activities to achieve educational goals". Furthermore, in article 36 paragraph (3) it is stated that the curriculum is prepared by the level and type of education within the framework The term "Curriculum" has many definitions and is formulated by experts, each defining according to the core emphasis and views of the expert. The curriculum comes from Greek, namely "curir" which means runner, and "curere" which means a place to gallop. In the past, this term was used in the world of sports. The curriculum based on the term is defined as "The distance that must be traveled by a runner from start to finish to get a medal or award". RESEARCH METHODS In this study, researchers use a qualitative approach while this type of research approach is descriptive with a literature study method, which is intended to be used by researchers to describe and explain the development of an Islamic-based independent curriculum based on the national curriculum (Sugiyono, 2017). Daryatmo Rahman1 R. Madhakomala2 This understanding is then adapted into the world of education and is interpreted as "A number of subjects that must be taken by a student from the beginning to the end of the program in order to obtain a diploma". In the book entitled "Curriculum Development and Development in Schools" by Dr. Nana Sudjana, it is stated that the definition of curriculum is a collection of intentions and expectations contained in the form of educational programs which are then implemented and implemented by teachers in the school concerned (Elyana & Das, 2022). In the Law on the National Education System No. 20 of 2003 article 1 point 19 stated, "curriculum is a set of arrangements and plans regarding objectives, content, and subject matter as well as ways used as guidelines for learning activities to achieve educational goals". Furthermore, in article 36 paragraph (3) it is stated that the curriculum is prepared by the level and type of education within the framework 1337\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum That way, students can learn well and prepare From these Islamic educational institutions, the community has high hopes that their children and grandchildren can gain a lot of knowledge, especially Islamic sciences ranging from reading and writing iqra, Daryatmo Rahman1 R. Madhakomala2 RESULTS AND DISCUSSION This view rests on the assumption that science is the heart / center of attention of curriculum, education, and learning (Damri, 2021). Axiology is closely related to ethics that examines values, norms and morals and also aesthetics, namely the results obtained from experiences future of students. In the development of the curriculum the foundation lies the following factors including: 1) Philosophy; 2) Socio-cultural and religious prevailing in society; 3) Student development; 4) Development needs; 5) The development of science and technology. In the study of educational philosophy about the curriculum to be developed, it is grouped first in several relevant theories, namely: 1) ontology that discusses reality (the nature of reality); 2) epistemology that discusses the nature of knowledge; 3) Axiology which discusses the nature of value. Ontology in philosophy is grouped into three categories, the first is supernatural ontology, which views and places reality in the supernatural realm (the divine spiritual realm or the realm of ideas derived from Plato). The second ontology is the ontology of the earth, which is to view and place reality on the assumption that the source is on earth. The third ontology is a human ontology that places reality on human experience. Epistemology related to curriculum always seeks to uncover the truth and error of thinking and acting in learning. This view rests on the assumption that science is the heart / center of attention of curriculum, education, and learning (Damri, 2021). Axiology is closely related to ethics that examines values, norms and morals and also aesthetics, namely the results obtained from experiences In the development of the curriculum the foundation lies the following factors including: 1) Philosophy; 2) Socio-cultural and religious prevailing in society; 3) Student development; 4) Development needs; 5) The development of science and technology. In the study of educational philosophy about the curriculum to be developed, it is grouped first in several relevant theories, namely: 1) ontology that discusses reality (the nature of reality); 2) epistemology that discusses the nature of knowledge; 3) Axiology which discusses the nature of value. Ontology in philosophy is grouped into three categories, the first is supernatural ontology, which views and places reality in the supernatural realm (the divine spiritual realm or the realm of ideas derived from Plato). The second ontology is the ontology of the earth, which is to view and place reality on the assumption that the source is on earth. RESULTS AND DISCUSSION The curriculum is one of the most important parts in educational institutions, if without the appropriate curriculum, students will not get the right learning targets. The curriculum contains a set of plans, objectives, and learning materials. Including teaching methods that will be a guideline for each teacher in order to achieve learning targets and objectives well. The curriculum is one of the main components in education, it is a compass to guide where students want to be taken. Therefore, the position of the curriculum in educational practice is very important, but how important the 1339\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum 339\| Development of An Islamic-Based Independent Curriculum Based on The National urriculum of the Unitary State of the Republic of Indonesia by taking into account: a) Increased faith and piety; b) Improvement of noble morals; c) Increased potential, intelligence, and interest of learners; d) Diversity of regional and environmental potential; e) Regional and national development demands; f) Claims of the world of work; g) The development of science, technology, and art; h) Religion; i) Global development dynamics; And j) National associations and national values the development of educational curricula by the needs of society and the future of students. curricula by the needs of society and the future of students. In the development of the curriculum the foundation lies the following factors including: 1) Philosophy; 2) Socio-cultural and religious prevailing in society; 3) Student development; 4) Development needs; 5) The development of science and technology. In the study of educational philosophy about the curriculum to be developed, it is grouped first in several relevant theories, namely: 1) ontology that discusses reality (the nature of reality); 2) epistemology that discusses the nature of knowledge; 3) Axiology which discusses the nature of value. Ontology in philosophy is grouped into three categories, the first is supernatural ontology, which views and places reality in the supernatural realm (the divine spiritual realm or the realm of ideas derived from Plato). The second ontology is the ontology of the earth, which is to view and place reality on the assumption that the source is on earth. The third ontology is a human ontology that places reality on human experience. Epistemology related to curriculum always seeks to uncover the truth and error of thinking and acting in learning. RESULTS AND DISCUSSION From the review of the above understanding, it can be concluded that the curriculum is an education program that contains various kinds of teaching materials and learning experiences that are planned and then carefully and systemically designed to increase potential, intelligence, interests and talents and the development of science and technology as well as art and national values based on applicable norms and used as guidelines in the teaching and learning process for education staff and learners to achieve planned educational goals. The study of the curriculum cannot be separated from what must be delivered and mastered by students? How to convey it? And what kind of measuring instrument is used to determine what is desired? This question requires answers as an effort to face the challenges of a changing era that demands changes and The third ontology is a human ontology that places reality on human experience. Daryatmo Rahman1 R. Madhakomala2 \| 1340 obtained through sight, smell, touch, hearing etc. curriculum development designed for the needs of the Muslim community, especially with the aim of achieving the vision and mission of the educational institution and facilitating education personnel in the development and process of educational activities and evaluations that can be applied in a short time so that obstacles that occur in the middle of the learning process can be overcome properly, Fast and efficient. In the plan to develop an Islamic-based independent curriculum, first determine the goals and competencies to be achieved, for example in determining the goals in elementary schools (SD) is so that graduates have the basics of character in accordance with Islamic law, skills, skills, and knowledge that are adequate to develop their potential optimally. Then determine learning materials that can solve problems in everyday life, namely valid materials that have been tested for truth and validity and see the importance, feasibility, and interest to what extent the material is needed by students and provides feasibility benefits so as to give rise to encouragement to develop their abilities, and from all the material set is directed to achieve educational goals according to national standards. Then apply methods and strategies, namely the way used to deliver subject matter in an effort to achieve curriculum goals, this can be through three approaches, namely; 1) A subject-focused approach; 2) A learner-focused approach; 3) An approach oriented to people's lives; Socio-cultural and religious factors that prevail in society in an ecosystem both between human, cultural, biological, and geographical relationships are also related to the development of students to be the basis for developing curriculum in education because education is an effort to prepare students to plunge into the community environment. We certainly do not want the birth of students to become alienated and excluded in the community, but what we hope is the birth of students who better understand, adapt, and are able to build their community lives. Therefore, the preparation of plans, objectives, content, and processes, the educational curriculum must adjust to the social, cultural, characteristic, and development conditions that exist in society. Development factors, science and technology in order to accelerate the achievement of national strength and excellence. The third ontology is a human ontology that places reality on human experience. The great influence on human civilization is caused by advances in the fields of information, technology, telecommunications, so a knowledgeable society is needed through learning with superior educational quality standards through an ideal curriculum arrangement according to the times for the future development of students, one of which is the development of an independent Islamic-based curriculum. The development of an Islamic- The development of an Islamic- based independent curriculum is a 1341\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum 1341\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum Then creating a curriculum organization, Oemar Hamalik in his book entitled "Curriculum and Learning" he stated that there are six characteristics in curriculum organization; 1) Isolated subjects; 2) Correlated subjects; 3) Field of study (broadfield); 4) Child centered program; 5) Core programs; 6) Eclectic program;. With this curriculum organization, it will map the material according to the subject matter so that it is arranged systematically to make it easier for students to learn it. history of the companions, characters and Islamic scholars. 2) General subjects, consisting of (Indonesian, Mathematics, Science, Social Studies, Civics, PJOK, Art, Crafts and Skills, Life Skills of the 21st century) all general subjects are targeted based on the provisions of national standards, but the content of the material content has changes and developments so as to provide detailed understanding and deep understanding to students. 3) Extracurricular, consisting of (Qur'an house, Islamic studies, field trip). The third ontology is a human ontology that places reality on human experience. Madhakomala2 ال يف هدجن امك ةيلمعلا لئاسو يف و ةيرظن ال مث ،ا لو أ ةيسي ئر ةف صب ةنسل ا و نآ رق ال متها يتلا ىرخألا ةيركفلا دوهجلاب ةناعتسا به نوثدحملا و ءاهقفلا نم ريبكلا ليعرلا كلذ ا ومالسإلا يركف نم مهريغو ةفسالفلا ال يف هدجن امك ةيلمعلا لئاسو يف و ةيرظن ال مث ،ا لو أ ةيسي ئر ةف صب ةنسل ا و نآ رق ال متها يتلا ىرخألا ةيركفلا دوهجلاب ةناعتسا به نوثدحملا و ءاهقفلا نم ريبكلا ليعرلا كلذ ا ومالسإلا يركف نم مهريغو ةفسالفلا methods, means, concepts, to develop Islam-based education for individuals, communities, as a whole in all aspects for the life of the world and the hereafter. The development of Islamic-based educational curriculum must be guided by and refer to the Qur'an and hadith as its normative foundation. Al Syaibani, as quoted by Umar et al., explained the basic framework of the Islamic curriculum, including: An intellectual framework that addresses various educational issues and educational concepts in their theoretical and practical foundations, as we find in the Qur'an as well as the Hadith of the Prophet -Shallallahualaihi wa salam- mainly, and then using other intellectual endeavors that are of interest to other Islamic jurists, philosophers and thinkers. (Markaz ilmiyah in Islamic education, 1983, in a book entitled: Taushiyat Al- mu'tamiraat At-ta'liimiyah Al-Islaamiyah) 1. The basis of religion as the spirit and the highest target in the curriculum by referring to the main sources of Islamic teachings, namely the Qur'an and hadith. 2. The philosophical basis that provides philosophical guidance on the objectives of Islamic education so that the purpose, content, and organization of the curriculum contain values that are believed to be truth both in terms of ontological, epistemological, and also axiological. 3. إ عيمج نم الماك ادادعإ ملسملا دادع ال ةرخآلا و ايندلل هومن لحارم عيمج يف يحاون في ءاج يتلا ةيبرتلا قرطو ميقلاو ئدابملا ءوض بمالسإلا اه 3. A psychological basis that provides a foundation in the formulation of the curriculum to be in line with the psychological development of learners. Prepare a fully Muslim in all aspects at all stages of his development for the world and the Hereafter based on the principles, values, and methods of education that Islam has developed. (Yaljon Miqdad, 1989. In the book entitled: Ahdaf At- Tarbiyyah Al-Islamiyyah Wa Ghayaatuha). 4. The third ontology is a human ontology that places reality on human experience. In the independent curriculum based on Islam is very focused on the concept of Islamic education, Islamic education in Arabic is called -At- tarbiyyah Al-Islamiyyah- there are several experts defining -At-tarbiyyah Al- Islamiyyah- as follows: The preparation of an independent Islamic-based curriculum is grouped based on the national curriculum with several changes and developments tailored to student needs including, diniyyah (religious) subjects, general subjects and extracurriculars; 1) The subjects of diniyyah (religion) or Islamic Religious Education, consisting of subjects (Al-Qur'an, Aqidah, Hadith, Fiqh, Siroh, Arabic), each has a target goal of achievement at each level, here are examples that can be applied at the elementary level, for example in Arabic maple the target of graduates is to be able to communicate in Arabic and memorize a thousand vocabulary, and in the Qur'an maple the target of graduates is to be able to memorize three to six juz, and in the Aqidah maple the target of graduates is able to know their god, prophet, and religion, apply the Shar'i adab-adab, and in the Hadith maple the target of graduates is able to memorize the hadith arbaeen an- nawawiyah, and in the Siroh maple the target of graduates is able to know the 1. مج و ةيلقنلا لئاسولا و قئارطلا ةعوم ا يتلا ةيبرجتلا و ةيملعلا و ةيعامتجالا و ةيلقعل يس و بيدأت لل نوبرم لاو ءاملع لا اهمدخ ت ا ةيرشبلا و عمتجملا و درفلل ةيمنتلا و ،بيذهتل بق ةيشخلا و بولقلا يف هللا ىوقت قيقحت دص مسوفنلا يف هن A collection of methods and means, mental, social, scientific and experimental that scientists and educators use to discipline, and develop individuals, society and humanity to achieve piety to Allah and fear Him in him. (Ahmad Farhan Ishak, 1983. In his book entitled At-tarbiyyatul Islamiyah Bainal Asholati Wal Mu'ashirah) A collection of methods and means, mental, social, scientific and experimental that scientists and educators use to discipline, and develop individuals, society and humanity to achieve piety to Allah and fear Him in him. (Ahmad Farhan Ishak, 1983. In his book entitled At-tarbiyyatul Islamiyah Bainal Asholati Wal Mu'ashirah) 2. ال فلتخم لوانتي يذلا يركفلا راطإ قض اهسس أ يف ةيبرتلا ميها فم و ميل عتل ا ايا 2. ال فلتخم لوانتي يذلا يركفلا راطإ قض اهسس أ يف ةيبرتلا ميها فم و ميل عتل ا ايا \| 1342 Daryatmo Rahman1 R. The third ontology is a human ontology that places reality on human experience. Social basis that provides an illustration so that Islamic education is rooted in the life and culture of the community. With the basics of this development, it can be used as a guide in compiling educational curricula, especially in informal education units based on Islam which refers and is guided by two main sources of Islamic From this definition it can be concluded that Islamic education the main source is from the Qur'an and the Hadith of the Prophet -Sallallahu alaihi wasallam- then experts collect 1343\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum 1343\| Development of An Islamic-Based Independent Curriculum Based on The National Curriculum make a plan, targets in improving and developing learning, then a curriculum is needed, curriculum formation based on the needs of students in order to achieve the desired target, it is necessary to develop a curriculum, if what is needed is students who are Muslim, then the development is Islamic-based based based on two main sources, namely the Qur'an and Al-Hadith in accordance with the method, The concept that has been set in Islamic law, then combined with general sciences based on national standards, so that in running the wheels of education runs on the lines, foundations and targets that have been set. In developing an independent Islamic-based curriculum, it is necessary to compile good articulation by building good cooperation between teachers, principals and the community in order to realize the continuity of learning experiences from every level of education. teachings, namely the Qur'an and Al- Hadith. However, it is necessary to pay attention to articulation in the development of Islamic-based curriculum, researching thoroughly, removing things that are not needed, eliminating duplication. If the articulation is done well, there will be a continuous Islamic education in the learning experience from kindergarten to college. To compile this articulation, good cooperation between school principals, teachers from every level of education, parents of students and experts and community leaders will arise, because in the future there will inevitably be obstacles in the development of this curriculum, including obstacles that lie in teachers, meaning that in this case teachers do not participate in curriculum development due to several things. First, lack of time. Second, it is not in accordance with the opinions or ideas expressed between fellow teachers and with the principal. REFERENCE Bahri, S. (2018). Pengembangan kurikulum berbasis multikulturalisme di indonesia (landasan filosofis dan psikologis pengembangan kurikulum berbasis multikulturalisme). Jurnal Ilmiah DIDAKTIKA VOL. 19, NO. 1, 69-88 Damri, D. (2021). Pengambangan Kurikulum dan Pembelajaran di Sekolah yang Beragam Peserta Didik. The third ontology is a human ontology that places reality on human experience. Third, because of the ability and knowledge of the teachers themselves, and there are also other obstacles coming from the community, it is necessary to support the community in the development of Islam-based curriculum both in financing and in providing feedback on the ongoing curriculum. CONCLUSION The curriculum is one of the main components in education, it is a compass to guide where students want to be taken. An institution is required to Daryatmo Rahman1 R. Madhakomala2 \| 1344 Dhomiri, A. (2023). Konsep Dasar dan Peranan serta Fungsi Kurikulum dalam Pendidikan. Jurnal Pendidikan dan Sosial Humaniora Vol.3, No.1. 118-128 Martin, R., & Marianus Simanjorang, M. (2022). Pentingnya Peranan Kurikulum yang Sesuai dalam Pendidikan di Indonesia. Jurnal prosiding seminar nasional pendidikan dasar. Volume 1 Nomor 1, https://journal.mahesacenter.org/i ndex.php/ppd/index. Martin, R., & Marianus Simanjorang, M. (2022). Pentingnya Peranan Kurikulum yang Sesuai dalam Pendidikan di Indonesia. Jurnal prosiding seminar nasional pendidikan dasar. Volume 1 Nomor 1, Dwi Nur Khalim, A. (2019). Landasan sosiologis pengembangan kurikulum sebagai persiapan generasi yang berbudaya islam. Jurnal Kajian Kritis Pendidikan Islam dan Manajemen Pendidikan Dasar https://journal.mahesacenter.org/i ndex.php/ppd/index. Mayasari, R. (2023). Sosialisasi Sertifikasi Halal Pada Usaha Mikro Dan Kecil Melalui Kategori Self Declare: Studi Pada Umkm Desa Mojogeneng Kecamatan Jatirejo Kabupaten Mojokerto. Musyawarah: Jurnal Pengabdian Masyarakat, 1(4), 125– 132. Elyana, L., & Das, R. K. (2022). Management of Islamic Education on Construction of Early Childhood Curriculum. Journal of Nonformal Education, 8(2), 286–294. Sugiyono, P. D. (2017). Metode penelitian bisnis: pendekatan kuantitatif, kualitatif, kombinasi, dan R&D. Penerbit CV. Alfabeta: Bandung, 225. Hamalik, O. (2008). Kurikulum dan pembelajaran. Sinar Grafika Hidayat, S., & Asroi. (2013). Manajemen pendidikan substansi dan implementasi dalam praktik pendidikan di Indonesia. Pustaka Mandiri. Hidayat, S., & Asroi. (2013). Manajemen pendidikan substansi dan implementasi dalam praktik pendidikan di Indonesia. Pustaka Mandiri. Zahra Qurrata Ainy, F., & Effane, A. (2023). Peran kurikulum Dan Fungsi kurikulum. Jurnal Karimah Tauhid, Volume 2 Nomor 1 Zahra Qurrata Ainy, F., & Effane, A. (2023). Peran kurikulum Dan Fungsi kurikulum. Jurnal Karimah Tauhid, Volume 2 Nomor 1 Zahra Qurrata Ainy, F., & Effane, A. (2023). Peran kurikulum Dan Fungsi kurikulum. Jurnal Karimah Tauhid, Volume 2 Nomor 1 Mansur, R. (2016). Pengembangan kurikulum pendidikan agama islam multikultural. Jurnal Ilmiah Vicratina, Volume 10, No. 2 Kependidikan Dan Keislaman FAI Unisma © 2023 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/
https://openalex.org/W2464075771	https://europepmc.org/articles/pmc6159285?pdf=render	English	null	Incidental radiological diagnosis of asymptomatic Abernethy malformations—two case reports	BJR case reports	2,017	cc-by	3,420	Received: 09 December 2015 Revised: 07 June 2016 Accepted: 28 June 2016 Cite this article as: Shah A, Aziz A, Awwad A, Ramjas G, Higashi Y. Incidental radiological diagnosis of asymptomatic Abernethy malformations——two case reports. BJR Case Rep 2017; 2: 20150496. BJR\|case reports https://doi.org/10.1259/bjrcr.20150496 Received: 09 December 2015 Revised: 07 June 2016 Accepted: 28 June 2016 Cite this article as: Shah A, Aziz A, Awwad A, Ramjas G, Higashi Y. Incidental radiological diagnosis of asymptomatic Abernethy malformations——two case reports. BJR Case Rep 2017; 2: 20150496. BJR\|case reports https://doi.org/10.1259/bjrcr.20150496 Cite this article as: Shah A, Aziz A, Awwad A, Ramjas G, Higashi Y. Incidental radiological diagnosis of asymptomatic Abernethy malformations——two case reports. BJR Case Rep 2017; 2: 20150496. SUMMARY while the second case is under observation post recovery from his traumatic injuries and will be subsequently referred to the hepatology team in the future. The diagnosis of congenital extrahepatic portosystemic shunts (CEPS) is of clinical significance because of the risk of hepatic encephalopathy; liver dysfunction; and associ- ated cardiac, gastrointestinal, vascular, skeletal and genito- urinary anomalies. While the actual incidence and clinical importance of CEPS in asymptomatic patients is not known, they are extremely rare in healthy individuals. ALI SHAH, MBBS, MRCS, 1ABDUL AZIZ, MRCS, 2,3AMIR AWWAD, MRCS, 4GREG RAMJAS, FRCR and YUTARO HIGASHI, FRCR Trauma and Orthopaedics, Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK Si Pete M n field Im ging Cent e (SPMIC) Q een’ Medic l Cent e Nottingh m Uni e it Ho pit l N 1Trauma and Orthopaedics, Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK 2Sir Peter Mansfield Imaging Centre (SPMIC), Queen’s Medical Centre, Nottingham University Hospitals, N 3Radiology Department, Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK 4Interventional Radiology Queen’s Medical Centre Nottingham University Hospitals Nottingham UK 1Trauma and Orthopaedics, Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK 2Sir Peter Mansfield Imaging Centre (SPMIC), Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK 3Radiology Department, Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK 4Interventional Radiology, Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK gy p , , g y p , g , 4Interventional Radiology, Queen’s Medical Centre, Nottingham University Hospitals, Nottingham, UK Address correspondence to: Mr Ali Shah il li h h1@ h li h h@d k E-mail: ali.shah1@nhs.net;ali.shah@doctors.net.uk Conclusion Although uncommon, extrahepatic portosystemic shunts can cause significant morbidity and mortality and all new cases diagnosed radiologically should be further investi- gated by referring them to a hepatologist. ABSTRACT The diagnosis of the rare congenital extrahepatic portosystemic shunts is of clinical significance because of the risk of hepatic encephalopathy; liver dysfunction; and associated cardiac, gastrointestinal, vascular, skeletal and genitourinary anomalies. This article describes two varying cases showing the same type of the extrahepatic congenital shunts (Type II). Both the patients were clinically asymptomatic. The first patient initially presented with unprovoked deep venous thrombosis and a staging CT scan was performed to identify any potential underlying malignancy. The second was a polytrauma patient in whom a congenital extrahepatic portosystemic shunt was identified on the CT scan performed to investigate the trauma-related injuries. The first case underwent hepatological investigations, including a fibroscan to rule out liver fibrosis, and was diagnosed as having a Type II congenital malformation, while the second case is under observation post recovery from his traumatic injuries and will be subsequently referred to the hepatology team in the future. Although uncommon, extrahepatic portosystemic shunts can cause significant morbidity and mortality, and all new cases diagnosed radiologically should be further investigated by referring them to a hepatologist. CASE REPORT 1 This article describes two varying cases showing the same type of extrahepatic congenital shunts (Type II). Both the patients were healthy and clinically asymptom- atic. The first patient initially presented with unprovoked deep venous thrombosis (DVT) and a staging CT scan was performed to identify any potential underlying malignancy. The second was a polytrauma patient in whom CEPS was identified on the CT scan performed to evaluate the traumas. © 2017 The Authors. Published by the British Institute of Radiology. This is an open access article under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution and reproduction in any medium, provided the original author and source are credited Incidental radiological diagnosis of asymptomatic Abernethy malformations——two case reports 1ALI SHAH, MBBS, MRCS, 1ABDUL AZIZ, MRCS, 2,3AMIR AWWAD, MRCS, 4GREG RAMJAS, FRCR and 3YUTARO HIGASHI, FRCR Diagnosis and management The radiologist reporting the CT scan recommended a hepatol- ogy clinic referral, which was subsequently organized. During the clinic visit, a full assessment, including physical examination, was conducted. Contrary to any plausible clinical expectations, no central or peripheral signs of cardiac or chronic hepatic dis- ease were identified. There were neither audible murmurs on auscultation nor any evidence of hyperdynamic circulation. The liver was not palpable, and there was no free fluid in the abdo- men. There was, however, a mildly enlarged spleen, about 2 cm below the left costal margin; this was also confirmed on the CT images. until 4 years ago, when her daily intake increased by an addi- tional 10 units. She often indulged in binge drinking, mainly for the pain associated with her musculoskeletal and rheumatologi- cal comorbidities. There was no clinical history to suggest any liver abnormality (e.g. jaundice, pedal oedema, ascites, encephalopathy, gastrointestinal bleeding) or symptoms sugges- tive of cardiac overload such as shortness of breath or history of cardiac ischaemic disease. The patient, however, was still asymptomatic with a Type II Abernethy malformation, and her previous surgical history did not seem to qualify as an iatrogenic cause for this shunt. There- fore, further investigations were conducted. These included a non-invasive liver screen with immunoglobulins, autoimmune and viral hepatitis (hepatitis B and C) markers; a fibroscan with a view to proceed to liver biopsy; and an endoscopy to look for portal hypertension and varices. until 4 years ago, when her daily intake increased by an addi- tional 10 units. She often indulged in binge drinking, mainly for the pain associated with her musculoskeletal and rheumatologi- cal comorbidities. There was no clinical history to suggest any liver abnormality (e.g. jaundice, pedal oedema, ascites, encephalopathy, gastrointestinal bleeding) or symptoms sugges- tive of cardiac overload such as shortness of breath or history of cardiac ischaemic disease. On her subsequent review by the hepatology team, the fibro- scan showed a median liver stiffness measurement value of 5.5 kPa (normal healthy adult < 7.0 kPa, median 5.3 kPa), which was within the normal range to exclude liver fibrosis. Additionally, a subsequent non-invasive liver screen was also negative. The platelet count on the most recent pathology test was normal (platelet count 154). Diagnosis and management Given these results, chronic liver disease and portal hypertension were deemed unlikely and the most likely cause of her shunt was believed to be a long-standing congenital anomaly, hence a liver biopsy was not indicated. Clinical presentation A 68-year-old female was referred to the haematology clinic with left-sided above-knee DVT, which was essen- tially unprovoked as per the obtained clinical history. Her medical history included osteoporosis, osteoarthritis and sciatica. She had undergone a subtotal colectomy with ileorectal anastomosis for large bowel obstruction due to a histologically proven benign stricture secondary to colonic diverticular disease 16 years ago. The first case underwent hepatological investigations, including a fibroscan to rule out liver fibrosis, and was diagnosed as having a Type 1 congenital malformation, Although she was a non-smoker, she had been consuming 20–30 units of alcohol per week for the past many years BJR\|case reports Shah et al Figure 1. (a) Axial CT image (1 mm, maximum intensity projection) showing an engorged left renal vein posteriorly, rela- tive to the portal vein, due to the congenital extrahepatic porto- systemic shunt in a 68-year-old asymptomatic female patient. (b) Coronal reconstructed CT image (5mm, maximum intensity projection) showing a tortuous and dilated left gastric (shunt) and splenic vein within the left upper abdomen and communicat- ing superior mesenteric vein and portal vein with the left renal vein and inferior vena cava (Supplementary Video 1a,b). mesenteric and splenic veins. This was seen to anastomose with an engorged left adrenal vein and ultimately drain into the left renal vein. The hepatic portal vein was evidently patent. The appearance of the liver was consistent with fatty infiltration but was otherwise unremarkable (Figure 1a,b). Investigations Haematological and liver function tests (LFTs) revealed asymp- tomatic mild thrombocytopenia that had been ongoing since 2006. Mild derangement of the LFTs was also noted (Table 1). A portovenous phase CT scan was performed by the haematolo- gist to look for any possible underlying malignancy as the cause of the DVT. The scan showed bulky enlargement of the left thy- roid lobe with multiple nodules. There was neither any supracla- vicular, thoracic or axillary lymphadenopathy nor any focal lung lesions. The gallbladder, pancreas, spleen, adrenal glands and kidneys were also unremarkable. Case report: Asymptomatic Abernethy malformations BJR\|case reports CASE REPORT 2 Clinical presentation Figure 2. (a) Axial CT image of the upper abdomen (1 mm maximum intensity projection) showing a distended left gas- tric vein acting as a shunt (asterisk) anterior to the portal vein (arrow), which is seen anterior to the normal size inferior vena cava. (b) Coronal reconstructed CT image (1 mm, maxi- mum intensity projection) showing the shunt joining at the confluence of the superior mesenteric and splenic veins (side- to-side anastomosis) within the upper abdomen, surrounding the head of the pancreas (Supplementary Video 2a,b). A 56-year-old male was brought in by the regional ambulance team to our trauma centre (level 1) after being involved in a high speed road traffic accident. His past medical history included schizophrenia. Otherwise, he was fairly fit and well, with no significant comorbidities. He was seen and assessed by the trauma team and had a series of investigations and imaging studies, which included performing a CT scan to evaluate the traumas. He was found to have acute multiple traumatic injuries, all right-sided, with several fractures of the right upper and lower limbs, and the right hemipelvis. vasculature before the division of the portal vein. Our case series only deals with extrahepatic portosystemic shunts. Investigations They may also influence the systemic concentrations of compounds metabolized by the liver and potentially become a risk factor in patients with gastrointestinal tumours for earlier pulmonary metastasis.1 Outcome There were no findings on history taking or physical examina- tion to suggest any relevant symptomatology, and thus a full hepatological screen (viral screen, immunoglobulins, neutrophil cytoplasmic antibody level and antinuclear antibody level) was not carried out owing to the circumstances. Investigations The CT scan revealed an abnormal enhancing and distended extrahepatic portosystemic communication between the left renal vein/inferior vena cava and the splenic vein. A markedly hypoplastic portovenous system was also noted, likely due to considerable flow diversion into the systemic veins without any intrahepatic shunts identified. The findings suggested an incidental Type II CEPS, draining into the left renal vein (Figure 2a,b). vasculature before the division of the portal vein. Our case series only deals with extrahepatic portosystemic shunts. His LFTs on admission were normal, as detailed in Table 1. Fol- lowing his management, he spent a significant amount of time in high dependency care and was repatriated to his base hospital for further management. CEPS were first described by John Abernethy in 1793. They are extremely uncommon, especially in an asymptomatic individual. A systematic review spanning over 29 years (1982–2011) of publi- cations yielded a total of 80 studies with only 49 cases of CEPS.1 Two patients had no symptoms at the time of diagnosis.1 There- fore, while the exact incidence of CEPS is difficult to quantify, they are a rarity.1 They are, however, of significance because of their associated anomalies involving the cardiac and the hepatic systems along with the risk of hepatic encephalopathy.2 These are detailed in Table 2.3 Furthermore, shunts may affect the hepatic immune surveillance owing to the toxins and pathogens bypass- ing the liver. They may also influence the systemic concentrations of compounds metabolized by the liver and potentially become a risk factor in patients with gastrointestinal tumours for earlier pulmonary metastasis.1 CEPS were first described by John Abernethy in 1793. They are extremely uncommon, especially in an asymptomatic individual. A systematic review spanning over 29 years (1982–2011) of publi- cations yielded a total of 80 studies with only 49 cases of CEPS.1 Two patients had no symptoms at the time of diagnosis.1 There- fore, while the exact incidence of CEPS is difficult to quantify, they are a rarity.1 They are, however, of significance because of their associated anomalies involving the cardiac and the hepatic systems along with the risk of hepatic encephalopathy.2 These are detailed in Table 2.3 Furthermore, shunts may affect the hepatic immune surveillance owing to the toxins and pathogens bypass- ing the liver. Outcome Given the above findings in the absence of any local and/or sys- temic complications, she was discharged back to the care of her general practitioner without any further follow-up planned. However, there was an incidental finding of an extrahepatic por- tosystemic connection, with an enlarged vein arising from the portal vein just superior to the confluence of the superior Table 1. Patients’ test results (haematology and biochemistry) Tests Case 1 Case 2 Normal range Haemoglobin 140 100 115–165 g l–1 Platelets 142 117 150–450 108 l–1 Bilirubin 24 25 0–21 mmol l–1 Alkaline phosphatase 60 70 40–130 U l–1 Alanine aminotransferase 32 34 0–45 U l–1 Aspartate aminotransferase 51 74 0–35 U l–1 2 of 4 birpublications.org/bjrcr BJR Case Rep;2:20150496 Table 1. Patients’ test results (haematology and biochemistry) Table 1. Patients’ test results (haematology and biochemistry) 2 of 4 2 of 4 birpublications.org/bjrcr BJR Case Rep;2:20150496 DISCUSSION This case series has limitations, as the second patient is still recovering from injuries caused by the trauma and therefore has not had a formal hepatology review. There are two types of CEPS. Type I, in which there is absence of intrahepatic portal venous supply, is due to the complete diversion of portal blood into systemic circulation (end-to-side shunt).3 It is more common in females. 2 In Type II, the intrahe- patic supply via the portal vein is preserved but some of the por- tal flow is diverted into a systemic vein (side-to-side shunt). The There are two types of congenital portosystemic shunts, intrahe- patic and extrahepatic. Intrahepatic shunts include persistent patent ductus venosus and congenital hepatic vascular lesions, where the branches of the portal vein, after its division, join the hepatic veins or the inferior vena cava directly. In extrahe- patic shunts, a systemic vein joins the portomesenteric Table 2. Associated anomalies Circulatory Gastrointestinal Genitourinary ▪ Atrial septal defect ▪ Congenital aortic valve stenosis ▪ Dextrocardia ▪ Mesocardia ▪ Patent ductus arteriosus ▪ Patent foramen ovale ▪ Tetralogy of Fallot ▪ Ventricular septal defect ▪ Interruption of the inferior vena cava ▪ Double inferior vena cava ▪ Biliary atresia ▪ Choledochal cyst ▪ Intrahepatic gallbladder ▪ Polysplenia ▪ Bilateral ureteropelvic stenosis ▪ Crossed fused renal ectopia ▪ Hypospadias ▪ Multicystic dysplastic kidney ▪ Vesicoureteral reflux 3 of 4 birpublications.org/bjrcr BJR Case Rep;2:20150496 3 of 4 birpublications.org/bjrcr 3 of 4 BJR Case Rep;2:20150496 BJR\|case reports Shah et al age of diagnosis varies from 31 weeks of intrauterine life to 76 years.2 age of diagnosis varies from 31 weeks of intrauterine life to 76 years.2 carried to the liver. If a shunt is present, the isotope will be detected on the images of both the liver and lungs, otherwise only on the images of the liver. CONSENT Appropriate consent was obtained to publish the report from the first patient; however, informed consent could not be obtained from the second patient or the next of kin despite exhaustive efforts. Both cases have been sufficiently anonymized to protect patient identity. Other investigations include portal scintigraphy performed with rectal administration of iodine 123 iodoamphetamine, the iso- tope of which are absorbed by the inferior mesenteric vein and DISCUSSION Thus nuclear studies can calcu- late shunt ratios in Type II CEPS.3 Portosystemic shunts arising without concomitant liver disease or portal hypertension4 can also occur owing to abdominal trauma, prior surgery or postnatal massive necrosis secondary to either viral or hepatotoxic injury.5 Most extrahepatic systemic shunts involve a mesenteric vein and the vena cava.5 Liver biopsy, although not necessary, may help to differentiate between Type I and II CEPS by showing small portal venules within the portal triads in Type I, a finding that cannot be seen with imaging tests, and hence can influence the type of management. There are different modalities for the diagnosis of CEPS. Ultra- sound scan is the investigation of choice in symptomatic infants owing to its non-invasive nature; it requires no sedation and does not expose the patient to radiation. Often CEPS is an inci- dental finding when an ultrasound scan is performed for another purpose. Doppler ultrasound scan can detect the flow direction. The limitations of ultrasound scan include failure to fully visualize all associated shunts.6 A recently published sys- tematic review showed that 31 out of 112 patients (27.6%) required two or more different modalities for diagnosis.1 Treatment of CEPS is dependent on patient factors and presen- tation. Children are usually asymptomatic owing to their central nervous system being less sensitive to the effects of possible hyperammonaemia. Therefore, they may need to be closely monitored, as medical therapy and dietary changes, for example, a low protein diet, can be used to manage most mild metabolic abnormalities.3 For symptomatic patients, it is prudent to determine the type of shunt. While Type II shunts can be occluded either surgically or by embolization (percutaneous transcatheter coil placement), Type I shunts are the only drainage route for splenic and mesen- teric blood, and hence liver transplantation is the choice of treat- ment in symptomatic patients.3 For this purpose, CT and MR angiography scans are recom- mended. Radiological Society of North America recommends MR angiography to be considered first owing to its reliability in assessing hepatic vascular anatomy without exposing the patient to radiation.3 With the development of multidetector CT scanners, it has been reported that its spatial resolution in the detection of small vascular branches is superior to that of MR angiography.7 Currently, Doppler and CT angiography are the most common radiological techniques used for making a diagnosis.1 LEARNING POINTS 1. CEPS represent a rare but important condition with many associated anomalies. 2. Identifying CEPS on scans early can assist with prompt diagnosis and management. Conventional angiography, on the other hand, has risks of radia- tion, vascular injury and anaesthesia, and is not routinely required owing to MRI or CT scan findings leading to a diagnosis in most cases. Techniques such as indirect mesenteric portovenography can clarify the portal system anatomy, whereas percutaneous transhepatic portography can assist with selective embolization in Type II shunts as well as visualization of features of an extrahepatic shunt.3 3. The diagnosis is made radiologically and the approach is multidisciplinary, with the radiologist, hepatologist and specialist for each of the systems involved. REFERENCES 1. Matthews TJ, Trochsler MI, Bridgewater FH, Maddern GJ. Systematic review of congenital and acquired portal-systemic shunts in otherwise normal livers. Br J Surg 2014; 101: 1509–17. doi: https://doi.org/10.1002/ bjs.9619 2. Murray CP, Yoo SJ, Babyn PS. Congenital extrahepatic portosystemic Shunts. Pediatr Radiol 2002; 33: 614–20. 3. Alonso-Gamarra E, Parrón M, Pérez A, Prieto C, Hierro L, López-Santamaría M. Clinical and radiologic manifestations of congenital extrahepatic portosystemic shunts: a comprehensive review. congenital extrahepatic portosystemic shunts: a comprehensive review. Radiographics 2011; 31: 707–22. doi: https:// doi.org/10.1148/rg.313105070 4. Nishimoto Y, Hoshino H, Sato S, Oguri A, Yamada M, Nishimura D, et al. Extrahepatic portosystemic venous shunt without portal hypertension. Intern Med 1997; 36: 886–9. doi: https://doi.org/10.2169/ internalmedicine.36.886 5. Ali S, Stolpen AH, Schmidt WN. Portosystemic encephalopathy due to mesoiliac shunt in a patient without cirrhosis. J Clin Gastroenterol 2010; 44: 381–3. doi: https://doi.org/10.1097/MCG. 0b013e3181aae51b 6. Hu GH, Shen LG, Yang J, Mei JH, Zhu YF. Insight into congenital absence of the portal vein: is it rare? World J Gastroenterol 2008; 14: 5969–79. doi: https://doi.org/10.3748/ wjg.14.5969 7. Prokop M. Multislice CT angiography. Eur J Radiol 2000; 36: 86–96. doi: https://doi.org/ 10.1016/S0720-048X(00)00271-0 4. bjs.9619 congenital extrahepatic portosystemic shunts: a comprehensive review. Radiographics 2011; 31: 707–22. doi: https:// doi.org/10.1148/rg.313105070 4. Nishimoto Y, Hoshino H, Sato S, Oguri A, Yamada M, Nishimura D, et al. Extrahepatic portosystemic venous shunt without portal hypertension. Intern Med 1997; 36: 886–9. doi: https://doi.org/10.2169/ internalmedicine.36.886 5. Ali S, Stolpen AH, Schmidt WN. Portosystemic encephalopathy due to wjg.14.5969 4 of 4 4 of 4 birpublications.org/bjrcr BJR Case Rep;2:20150496
https://openalex.org/W4362486715	https://figshare.com/articles/journal_contribution/Figure_S5_from_Diffuse_Intrinsic_Pontine_Glioma_Cells_Are_Vulnerable_to_Mitotic_Abnormalities_Associated_with_BMI-1_Modulation/22515660/1/files/39978174.pdf	English	null	Figure S1, A from Diffuse Intrinsic Pontine Glioma Cells Are Vulnerable to Mitotic Abnormalities Associated with BMI-1 Modulation	null	2,023	cc-by	53	Kumar, Fig.S5 Kumar, Fig.S5 Kumar, Fig.S5 p value : 0.024 PTC596 in vivo – CCHMC-DIPG-1 Animal Survival (n=3) Drug Holiday start to end p value : 0.024 PTC596 in vivo – CCHMC-DIPG-1 Animal Survival (n=3) Drug Holiday start to end PTC596 in vivo – CCHMC-DIPG-1 Animal Survival (n=3) Drug Holiday start to end
https://openalex.org/W3216575875	https://figshare.com/articles/journal_contribution/Modelling_of_Hepatitis_E_virus_RNA-dependent_RNA_polymerase_genotype_3_from_a_chronic_patient_and_i_in_silico_i_interaction_analysis_by_molecular_docking_with_Ribavirin/17124950/1/files/31660208.pdf	English	null	Modelling of Hepatitis E virus RNA-dependent RNA polymerase genotype 3 from a chronic patient and<i>in silico</i>interaction analysis by molecular docking with Ribavirin	Journal of biomolecular structure and dynamics/Journal of biomolecular structure & dynamics	2,021	cc-by	327	Table S1. HEV 1a-8a subtype reference sequences and different RNA viral sequences included in the phylogenetic reconstruction Strain Accession Number Genotype/Subtype HEV (Burma) M73218 1a HEV (Mex-14) KX578717 2a HEV (Meng) AF082843 3a HEV (US1) AF060668 3a HEV (US2) AF060669 3a HEV (JRA1) AP003430 3b HEV (wbGER27) FJ705359 3c HEV (swJ8-5) AB248521 3e HEV (E116-YKH98C) AB369687 3f HEV (Osh 205) AF455784 3g HEV (TR19) JQ013794 3h HEV (BB02) FJ998008 3i HEV (Arkell) AY115488 3j HEV (E088-STM04C) AB369689 3k HEV (FR-SHEV3c-like) JQ953664 3l HEV (IC2011) KU513561 3m HEV (JKO-ChiSai98C) AB197673 4a HEV (JBOAR135-Shiz09) AB573435 5a HEV (wbJOY_06) AB602441 6a HEV (178C) KJ496143 7a HEV (12XJ) KX387865 8a Hepatitis C virus P26664 1a Hepatitis A virus P08617 IB Coxsackievirus P08291 B1 Norwalk virus Q83883 GI Rubella virus (Therien) P13889 - Beet necrotic yellow vein virus A0A0H4LTT4 - Porcine Aichi virus (S1-HUN) 6rliA - Norwalk virus (Ast6139/01/Sp) 1sh0A II Coxsackie virus (Nancy) 3cduA B3 Enterovirus 71 3n6l - Table S1. HEV 1a-8a subtype reference sequences and different RNA viral sequences included in the phylogenetic reconstruction Strain Accession Number Genotype/Subtype HEV (Burma) M73218 1a HEV (Mex-14) KX578717 2a HEV (Meng) AF082843 3a HEV (US1) AF060668 3a HEV (US2) AF060669 3a HEV (JRA1) AP003430 3b HEV (wbGER27) FJ705359 3c HEV (swJ8-5) AB248521 3e HEV (E116-YKH98C) AB369687 3f HEV (Osh 205) AF455784 3g HEV (TR19) JQ013794 3h HEV (BB02) FJ998008 3i HEV (Arkell) AY115488 3j HEV (E088-STM04C) AB369689 3k HEV (FR-SHEV3c-like) JQ953664 3l HEV (IC2011) KU513561 3m HEV (JKO-ChiSai98C) AB197673 4a HEV (JBOAR135-Shiz09) AB573435 5a HEV (wbJOY_06) AB602441 6a HEV (178C) KJ496143 7a HEV (12XJ) KX387865 8a Hepatitis C virus P26664 1a Hepatitis A virus P08617 IB Coxsackievirus P08291 B1 Norwalk virus Q83883 GI Rubella virus (Therien) P13889 - Beet necrotic yellow vein virus A0A0H4LTT4 - Porcine Aichi virus (S1-HUN) 6rliA - Norwalk virus (Ast6139/01/Sp) 1sh0A II Coxsackie virus (Nancy) 3cduA B3 Enterovirus 71 3n6l - Table S1. HEV 1a-8a subtype reference sequences and different RNA viral sequences included in the phylogenetic reconstruction
https://openalex.org/W3134633341	https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1008714&type=printable	English	null	Using deep neural networks to evaluate object vision tasks in rats	PLOS computational biology/PLoS computational biology	2,021	cc-by	15,664	a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Using deep neural networks to evaluate object vision tasks in rats Kasper VinkenID1,2, Hans Op de BeeckID3 Kasper VinkenI 1 Department of Ophthalmology, Children’s Hospital, Harvard Medical School, Boston, Massachusetts, United States of America, 2 Laboratory for Neuro- and Psychophysiology, KU Leuven, Leuven, Belgium, 3 Department of Brain and Cognition & Leuven Brain Institute, KU Leuven, Leuven, Belgium kasper.vinken@childrens.harvard.edu * kasper.vinken@childrens.harvard.edu * kasper.vinken@childrens.harvard.edu a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Editor: Ronald van den Berg, Stockholm University, SWEDEN Received: June 17, 2020 Accepted: January 17, 2021 Published: March 2, 2021 Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pcbi.1008714 Abstract In the last two decades rodents have been on the rise as a dominant model for visual neuro- science. This is particularly true for earlier levels of information processing, but a number of studies have suggested that also higher levels of processing such as invariant object recog- nition occur in rodents. Here we provide a quantitative and comprehensive assessment of this claim by comparing a wide range of rodent behavioral and neural data with convolutional deep neural networks. These networks have been shown to capture hallmark properties of information processing in primates through a succession of convolutional and fully con- nected layers. We find that performance on rodent object vision tasks can be captured using low to mid-level convolutional layers only, without any convincing evidence for the need of higher layers known to simulate complex object recognition in primates. Our approach also reveals surprising insights on assumptions made before, for example, that the best perform- ing animals would be the ones using the most abstract representations–which we show to likely be incorrect. Our findings suggest a road ahead for further studies aiming at quantify- ing and establishing the richness of representations underlying information processing in animal models at large. OPEN ACCESS Citation: Vinken K, Op de Beeck H (2021) Using deep neural networks to evaluate object vision tasks in rats. PLoS Comput Biol 17(3): e1008714. https://doi.org/10.1371/journal.pcbi.1008714 * kasper.vinken@childrens.harvard.edu Introduction 2018.02.037, https://doi.org/10.1523/JNEUROSCI. 3663-13.2014, and https://doi.org/10.1093/cercor/ bhw111). All relevant data are either contained within the original manuscripts or available on the following OSF repository: http://doi.org/10.17605/ OSF.IO/4W39D. Two decades ago, macaque monkeys were uncontested as the primary animal model for study- ing visual processing at the cortical level. Since then, rodents have become increasingly popu- lar, in particular because developments in neurotechnology such as cell-level imaging and optogenetics capitalized upon genetic rodent models. A major question emerging from this evolution is how far we can take rodents as a model for the more complex aspects of visual information processing. Funding: This work was supported by Research Foundation Flanders, Belgium, www.fwo.be (postdoctoral fellowship of K.V.); by KU Leuven Research Council, www.kuleuven.be (C14/16/031 of H.O.d.B.); by Excellence of Science (EOS), www. eosprogramme.be (grant HUMVISCAT of H.O.d. B.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The most promising evidence comes from rats. Several high-profile studies have docu- mented seemingly complex visual object recognition capabilities in rats, using classification tasks with computer-rendered objects [1–5] or natural videos [6]. In these experiments, rats show the ability to recognize objects despite various transformations in viewing conditions such as position, size, and viewpoint. Neurophysiological recordings have revealed neural responses in lateral extrastriate cortex that might underlie these behavioral abilities [7,8]. To cite [9], “The picture emerging from this survey is very encouraging with regard to the possi- bility of using rats as complementary models to monkeys in the study of higher-level vision.” Competing interests: The authors have declared that no competing interests exist. However, the extent to which previous object recognition experiments in rats probed higher-level vision has never been tested empirically. That is, to what extent did these classifi- cation tasks actually require an abstract, invariant representation of the visual stimuli? From previous work we know that rats are experts at finding and using the lowest-level feature that is predictive of a correct response in a discrimination task [10]. Thus, in order to appreciate the capabilities of the rodent visual system, it is critical to understand the minimum level of abstraction that is required to solve the tasks that these animals are able to perform. Already from the first landmark study by Zoccolan et al. Introduction [1], it has been argued that the behavior is However, the extent to which previous object recognition experiments in rats probed higher-level vision has never been tested empirically. That is, to what extent did these classifi- cation tasks actually require an abstract, invariant representation of the visual stimuli? From previous work we know that rats are experts at finding and using the lowest-level feature that is predictive of a correct response in a discrimination task [10]. Thus, in order to appreciate the capabilities of the rodent visual system, it is critical to understand the minimum level of abstraction that is required to solve the tasks that these animals are able to perform. Already from the first landmark study by Zoccolan et al. [1], it has been argued that the behavior is unlikely to be based upon representations found in the primary visual cortex (V1). However, the level of invariance found in the later stages of the primate ventral stream may not be neces- sary to explain generalization in rodent object vision experiments, as even similarly small sized Marmoset monkeys far outperformed rats on the same task [11]. In primates there is a multi- step progression of representations with increasing levels of abstraction beyond V1. The first steps still retain properties such as local receptive fields and a retinotopic organization. In the latest stages, invariant representations emerge where objects and categories are easily separable (Fig 1A) [12]. If not based on a V1-like representation, then what level of abstraction would support rodent behavior in object vision tasks? The difficulty in answering this question is that there is no clear cut definition of what con- stitutes different levels of intermediate representations between V1 and a high-level represen- tation specialized in object recognition. Here we demonstrate the value of a computational framework to address this issue. Advances in deep learning have brought about convolutional deep neural networks (DNNs) that allow to simulate this hierarchical information processing. It turns out that DNNs trained on object recognition learn a cascade of representations similar to the ventral stream in monkeys and humans [14–22] and capture important aspects of object perception [15,18,23]. The architecture of these models consists of a series of layers that trans- form pixel inputs into increasingly abstract representations where objects categories are increasingly separable (Fig 1B and 1C). PLOS COMPUTATIONAL BIOLOGY PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats 2018.02.037, https://doi.org/10.1523/JNEUROSCI. 3663-13.2014, and https://doi.org/10.1093/cercor/ bhw111). All relevant data are either contained within the original manuscripts or available on the following OSF repository: http://doi.org/10.17605/ OSF.IO/4W39D. Author summary Despite years of investigating object recognition in rodents, it remains unclear to what extent their visual system supports a capacity for high-level, abstract representations. Here, we used computational deep neural network models to assess which level of abstrac- tion is required to reproduce rodent behavior in several studies, and which level matches representations in higher rodent visual areas. We found that both behavioral and neural data support mid-level representations at best, and show that certain evidence available in the literature may not provide as strong support for invariant visual object recognition as previously thought. Going forward, our findings suggest that computational models could serve as a principled benchmark for evaluating the richness of information processing across species and for designing experiments to push the boundaries of animal models. Copyright: © 2021 Vinken, Op de Beeck. 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: The authors confirm that all data underlying the findings are fully available without restriction. The experimental data analyzed in the current manuscript have previously been published elsewhere (https://doi.org/10.1073/ pnas.0811583106, https://doi.org/10.1016/j.cub. 1 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 Introduction Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. (d) For yes-no tasks, correct or incorrect classification of a test stimulus (square marker) was evaluated by comparing the output label of the classifier with the true object label. (e) For 2AFC tasks, both the position of the target (yellow marker) and the distractor (blue marker) relative to the decision boundary were taken into account: for a correct response, the target stimulus needed to be positioned more towards the direction of the target side of the decision boundary (see yellow arrow in iii) than the distractor stimulus. https://doi.org/10.1371/journal.pcbi.1008714.g001 focusing upon several studies suggesting a high level of invariance and abstraction. We evalu- ated for each DNN layer’s representational space whether a linear classifier could generalize from the stimuli used in the training phase of experiments in rats, to the novel stimuli used in the test phase. Thus, rather than asking whether the stimuli of a given experiment are linearly separable (which is likely true for a small amount of stimuli in a very high dimensional space), we are asking whether a linear classifier easily finds a solution for the training set that general- izes well to the test set. That is, we are asking how well each DNN layer’s representational space lends itself to the task the rats were probed with. We reasoned that if generalization as observed in rodents is successful in early DNN layer representations, the task does not require high-level, invariant object representations, and thus one should be careful in interpreting gen- eralization performance of rodents as evidence for invariant object recognition. The findings are very consistent across datasets: data from earlier studies can be explained by low to mid- level convolutional representations that fall short of the representations that underlie object recognition in primates. To connect these results back to the rodent visual system, we com- pared the representational geometry between DNN layers and several visual areas along the putative rodent ventral stream. Consistent with the assessment of behavioral studies, mid-level convolutional layers mapped best onto the higher visual areas in rodents. focusing upon several studies suggesting a high level of invariance and abstraction. Introduction It is important to note that the current DNN models do not capture all aspects of object vision in primates (for example see [24,25]), yet they cap- ture ventral stream processing better than any currently available model type [26]. A frame- work based on pre-trained DNNs has been used to model visual object recognition behavior in primates, accurately capturing object- but not image-level error patterns [27]. Here we used the same DNN models as a principled framework to assess the level of abstraction required to explain the data obtained from rats in object recognition tasks, PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 2 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Fig 1. Evaluating visual object recognition tasks using DNN-based models. (a) The primate ventral stream is thought to transform visual representations where objects are entangled, into more abstract visual representations where objects are increasingly separable [12]. (b) A feedforward DNN trained in object categorization of natural images, separated in a sequence of models with increasingly higher-level representations, by training separate fully connected decoder layers (i.e., linear classifiers) on top of each DNN layer’s feature representation. (c) Higher-level DNN representations (AlexNet) show increased separability of visual stimuli according to object identity [for the stimuli of [1] (left) and of [5] (middle)] and category [for the stimuli of [6] (right; orange: natural distractors; grey: scrambled distractors)]. Separability was quantified as the sign-reversed Generalized Discrimination Value (GDV [13]), which is 0 for non-separable classes and 1 for perfectly separable classes. Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. (d) For yes-no tasks, correct or incorrect classification of a test stimulus (square marker) was evaluated by comparing the output label of the classifier with the true object label. (e) For 2AFC tasks, both the position of the target (yellow marker) and the distractor (blue marker) relative to the decision boundary were taken into account: for a correct response, the target stimulus needed to be positioned more towards the direction of the target side of the decision boundary (see yellow arrow in iii) than the distractor stimulus. https //doi org/10 1371/journal pcbi 1008714 g001 Fig 1. Evaluating visual object recognition tasks using DNN-based models. Introduction (a) The primate ventral stream is thought to transform visual representations where objects are entangled, into more abstract visual representations where objects are increasingly separable [12]. (b) A feedforward DNN trained in object categorization of natural images, separated in a sequence of models with increasingly higher-level representations, by training separate fully connected decoder layers (i.e., linear classifiers) on top of each DNN layer’s feature representation. (c) Higher-level DNN representations (AlexNet) show increased separability of visual stimuli according to object identity [for the stimuli of [1] (left) and of [5] (middle)] and category [for the stimuli of [6] (right; orange: natural distractors; grey: scrambled distractors)]. Separability was quantified as the sign-reversed Generalized Discrimination Value (GDV [13]), which is 0 for non-separable classes and 1 for perfectly separable classes. Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. (d) For yes-no tasks, correct or incorrect classification of a test stimulus (square marker) was evaluated by comparing the output label of the classifier with the true object label. (e) For 2AFC tasks, both the position of the target (yellow marker) and the distractor (blue marker) relative to the decision boundary were taken into account: for a correct response, the target stimulus needed to be positioned more towards the direction of the target side of the decision boundary (see yellow arrow in iii) than the distractor stimulus. Fig 1. Evaluating visual object recognition tasks using DNN-based models. (a) The primate ventral stream is thought to transform visual representations where objects are entangled, into more abstract visual representations where objects are increasingly separable [12]. (b) A feedforward DNN trained in object categorization of natural images, separated in a sequence of models with increasingly higher-level representations, by training separate fully connected decoder layers (i.e., linear classifiers) on top of each DNN layer’s feature representation. (c) Higher-level DNN representations (AlexNet) show increased separability of visual stimuli according to object identity [for the stimuli of [1] (left) and of [5] (middle)] and category [for the stimuli of [6] (right; orange: natural distractors; grey: scrambled distractors)]. Separability was quantified as the sign-reversed Generalized Discrimination Value (GDV [13]), which is 0 for non-separable classes and 1 for perfectly separable classes. PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 Results We focus on three studies: the first showed that rats can learn to discriminate two objects invariant to changes in size and azimuth-rotation [1], the second found that rats employ differ- ent object recognition strategies that seem to vary in complexity [5], the third showed that rats are capable of ordinate-level categorization of natural videos [6]. We evaluated these experi- ments by modeling the behavioral tasks using pre-trained DNNs. Computational models based on pre-trained DNNs have been used before to model primate object recognition experi- ments, by replacing the original decoder of the DNN with a new linear decoder trained on the experimental task [27,28]. The linear decoder layer simulates what a readout neuron/area could decode if it had access to the same visual representation as the penultimate DNN layer. The decoder itself does not add any further non-linear processing and thus requires object rep- resentations to already be disentangled in order to be able to successfully generalize in an object recognition task. Such disentangled object representations have been demonstrated in inferotemporal cortex [29]. Linear decoders trained on neural representations in inferotem- poral cortex are sufficient to predict core human object recognition performance [30], proving that the framework of a linear decoder trained on object representations is sufficient to capture object recognition behavior. Because the new decoder in models of primate object recognition received its input from the penultimate DNN layer [27,28], these models used the full depth of the original DNN and had access to a very high-level representation. However, if more of the top layers of the original DNN are removed, the new decoder will be trained on a layer that is lower in the hierarchy of the original DNN, resulting in a different model, which has access to a relatively lower-level representation. Thus, instead of training only one model that uses the full depth of the DNN, we constructed several models, each removing a different number of top layers of the pre- trained DNN before adding the new decoder (Fig 1B). The simplest model retained only the first layer of the DNN, a second model retained the first two layers, and so on. Results As the sequence of layers contains increasingly more abstract representations that are increasingly more useful for invariant visual object recognition, each successive model simulates a visual system which has access to increasingly higher level representations that each map best onto successive stages of the primate ventral stream [15,16,31]. Indeed, for the stimulus sets considered in the present study we found that higher DNN layers showed increasingly better separation for object iden- tity and category (Fig 1C), even though these models were trained on ImageNet [32]. Testing and comparing the performance of these models on behavioral experiments of visual object recognition in rats, provides information about the task difficulty in terms of the level of DNN processing steps that would be required to obtain a representational space that lends itself well to the task. For example, if a linear decoder model based on a certain DNN lay- er’s outputs successfully generalizes in a particular task, this suggests that the DNN layer’s representation is sufficiently abstract to capture the level of invariance probed by the experi- ment. Alternatively, a model that fails to generalize suggests that additional non-linear process- ing may be required to obtain a higher level of invariance and solve the task. Introduction We evalu- ated for each DNN layer’s representational space whether a linear classifier could generalize from the stimuli used in the training phase of experiments in rats, to the novel stimuli used in the test phase. Thus, rather than asking whether the stimuli of a given experiment are linearly separable (which is likely true for a small amount of stimuli in a very high dimensional space), we are asking whether a linear classifier easily finds a solution for the training set that general- izes well to the test set. That is, we are asking how well each DNN layer’s representational space lends itself to the task the rats were probed with. We reasoned that if generalization as observed in rodents is successful in early DNN layer representations, the task does not require high-level, invariant object representations, and thus one should be careful in interpreting gen- eralization performance of rodents as evidence for invariant object recognition. The findings are very consistent across datasets: data from earlier studies can be explained by low to mid- level convolutional representations that fall short of the representations that underlie object recognition in primates. To connect these results back to the rodent visual system, we com- pared the representational geometry between DNN layers and several visual areas along the putative rodent ventral stream. Consistent with the assessment of behavioral studies, mid-level convolutional layers mapped best onto the higher visual areas in rodents. 3 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Any DNN layer can account for size and rotation tolerance in a task with two rendered objects The first DNN layers account for generalization performance across a range of object transformations, but higher layers explain more variance in the transformation-level behavioral performance signature in Zoccolan et al. [1]. (a,b) The full sets of size and azimuth-rotation combinations of the two objects used in the behavioral task. For copyright reasons we do not show the stimuli from the original paper, but images of similar objects created in Blender 2.91 for illustrative purposes only. For the model we did use the originals from [1]. Rats were first trained on a subset of 14 of these transformations (purple) and subsequently asked to generalize to the remaining 40 novel combinations (green). (c,d) Average percentage correct discrimination of object 1 and object 2 by models incorporating increasingly more DNN layers of AlexNet (c) and VGG16 (d) (the X-axis indicates the highest DNN layer). Each model’s decoder layer was first trained on the same object transformations as the rats, after which performance was evaluated on these trained transformations (purple) as well as the untrained transformations (green). Horizontal lines indicate the average behavioral performance across rats reported by Zoccolan et al. [1]. Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. (e,f) Percentage variance in performance across stimuli explained by each DNN layer (estimated by a linear mapping fit on the data and stimuli of the training set only) from AlexNet (e) and VGG16 (f). Same conventions as in (c,d). All error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling size and azimuth-rotation combinations). https://doi.org/10.1371/journal.pcbi.1008714.g002 presented and the rat had to indicate the object identity by licking either a left or right feeding tube. After training, the rats were tested on the full stimulus set, which included novel combi- nations of size and azimuth-rotation (Fig 2A and 2B). In the original experiment, the rats generalized remarkably well to the object transforma- tions they had never seen before. We trained DNN-based models on the same task (Materials and Methods, Computational modeling, Modeling behavioral tasks) and found that this level of generalization turned out to require surprisingly little processing: the models incorporating only the first convolutional layer of both AlexNet and VGG16 already achieved near perfect generalization performance on the test set (Fig 2C and 2D). Any DNN layer can account for size and rotation tolerance in a task with two rendered objects We started by assessing the first landmark paper that reported evidence for invariant object recognition in rats [1]. In this study, rats were first trained to discriminate between two differ- ent objects and then to tolerate variations in size (15 to 40 degrees of visual angle) and azi- muth-rotation (60˚ left to 60˚ right), using a yes-no task in an operant box with liquid rewards (see Materials and Methods, Behavior, Task paradigms). At each trial, one object was 4 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Fig 2. The first DNN layers account for generalization performance across a range of object transformations, but higher layers explain more variance in the transformation-level behavioral performance signature in Zoccolan et al. [1]. (a,b) The full sets of size and azimuth-rotation combinations of the two objects used in the behavioral task. For copyright reasons we do not show the stimuli from the original paper, but images of similar objects created in Blender 2.91 for illustrative purposes only. For the model we did use the originals from [1]. Rats were first trained on a subset of 14 of these transformations (purple) and subsequently asked to generalize to the remaining 40 novel combinations (green). (c,d) Average percentage correct discrimination of object 1 and object 2 by models incorporating increasingly more DNN layers of AlexNet (c) and VGG16 (d) (the X-axis indicates the highest DNN layer). Each model’s decoder layer was first trained on the same object transformations as the rats, after which performance was evaluated on these trained transformations (purple) as well as the untrained transformations (green). Horizontal lines indicate the average behavioral performance across rats reported by Zoccolan et al. [1]. Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. (e,f) Percentage variance in performance across stimuli explained by each DNN layer (estimated by a linear mapping fit on the data and stimuli of the training set only) from AlexNet (e) and VGG16 (f). Same conventions as in (c,d). All error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling size and azimuth-rotation combinations). https://doi org/10 1371/journal pcbi 1008714 g002 Fig 2. https://doi.org/10.1371/journal.pcbi.1008714.g002 Any DNN layer can account for size and rotation tolerance in a task with two rendered objects Even a model trained on a pixel representation achieved 93% correct on the training set and 89% on the test set. Thus, no non- linear processing is required to explain a high level of generalization performance from the trained to untrained object transformations. The performance of the rats was not the same for all combinations of size and azimuth- rotation: performance was highest for the combination of the most common viewpoint and size in the training set (i.e. the center object in the purple “cross” in Fig 2A and 2B), and decreased for objects that deviated from that combination. We call this pattern of perfor- mances across object transformations (or–more generally–across images) the behavioral per- formance signature. To assess whether DNN representations could capture the behavioral results beyond overall generalization accuracy, we tested whether a linear combination of the 5 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats activation patterns in a DNN layer could also accurately predict this behavioral performance signature (Materials and Methods, Computational modeling, Predicting behavioral perfor- mance signatures). This mapping from DNN representations to behavioral performance per combination of object size and rotation was fit using the training stimuli and data only. Next, we used this behavioral mapping to predict generalization performances for each of the novel (green) object transformations. We calculated the variance in behavioral performance across object transformations that was explained by the DNN representations. These results show that the first convolutional layer only explained a limited amount of transformation-level vari- ance of 35–50%, which then sharply increased and slowly reached a maximum of >80% in higher convolutional layers (Fig 2E and 2F). Thus, whereas representations in the first layer of AlexNet and VGG16 can already explain successful generalization to untrained object trans- formations, higher convolutional layers get increasingly better at explaining transformation- level differences in behavioral performance of rats. None of the networks showed any improve- ment from incorporating the fully connected layers. Earliest DNN layers can account for the stimulus-level performance signature across rendered objects and sizes, also for the best performing rats To further investigate how well the DNN models can explain object and transformation-level differences in the behavioral performance of rats, we turned to a recent study by Djurdjevic et al. [5]. In this study, rats were trained to discriminate a reference object from 11 distractor objects at different sizes ranging from 15 to 35 degrees of visual angle (Fig 3A), using a yes-no task in an operant box with liquid rewards (see Materials and Methods, Behavior, Task para- digms). The goal of the study was to use the variable discrimination performance observed across object conditions to infer the complexity of the rats’ perceptual strategy. The authors found that there were "good performers", which performed above chance for all distractors at a size of 30˚, and "poorer performers", which performed below chance for more challenging dis- tractors (e.g., the T-shape highlighted in blue in Fig 3A). Can the behavioral fingerprints of good and poorer performing rats be explained by a dif- ference in readout from the same DNN representation, or is a higher-level representation required to explain the behavior of good performing rats? To investigate this, we calculated how much of the object and size-level variance in behavioral performance could be explained, using the same DNN layer activations for good and poorer performing rats, but fitting a separate linear mapping for the two groups as defined before (using the data that was displayed in Figs 1 and 2 of Djurdjevic et al. [5]; Materials and Methods, Computa- tional modeling, Predicting behavioral performance signatures). For both good and poorer performing rats, the percentage explained variance was with 85% already near maximum for the first convolutional layer, with an absolute maximum of 91% at pool2 of AlexNet (Fig 3B and 3C). Surprisingly, no higher-level representation was required to capture the behavioral performance signature of good performers, suggesting that in principle the dif- ference in behavioral performance signature between good and poorer performing rats can be explained by a different weighting of features of the same representation from the same DNN, in contrast with the idea that good performers relied on more advanced processing of shape information [5]. In sum, object and size-level differences in behavioral perfor- mance of both good and poorer performers could be captured well by the earliest layers with the least amount of processing, suggesting that the rats’ perceptual strategies could have relied on relatively low-level visual representations. Earliest DNN layers can account for the stimulus-level performance signature across rendered objects and sizes, also for the best performing rats For copyright reasons we do not show the stimuli from the original paper, but similar silhouette images were redrawn by hand or adapted from close-matching clip-art (https://openclipart.org/) for illustrative purposes only. For the model we did use the originals from [5]. The 3 example distractors of Fig 1 in Djurdjevic et al. [5] are highlighted in yellow, green, and light blue. Bottom: in a later phase of the experiment, the rats were trained to tolerate size changes in all objects from 15˚ to 35˚ of visual angle (here only shown for the subset of 4 objects indicated in color on the top). (b,c) Percentage of variance in average discrimination performance of good performing (blue) and poorer performing (red) rats, explained by each DNN layer (estimated by a linear mapping with leave-one-out cross-validation) from AlexNet (b) and VGG16 (c). The linear mapping was estimated using the data available in Figs 1 and 2 of Djurdjevic et al. [5]: all sizes of the 4 example objects indicated by purple, yellow, green, and light blue in (a), and the 30˚ size for the remaining eight objects. Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling stimuli). (d) Top: average discrimination performances of good performing rats, as a function of object size, for the reference object and the 3 example distractors. Middle: average predicted discrimination performances of good performing rats, based on pool2 of AlexNet, which explained the most out-of-sample variance in behavioral performance. Bottom: same as middle, but for conv1 of AlexNet. (e) Same as (d), but for poorer performing rats. https://doi.org/10.1371/journal.pcbi.1008714.g003 Earliest DNN layers can account for the stimulus-level performance signature across rendered objects and sizes, also for the best performing rats Stimulus-level discrimination performance signatures for objects from Djurdjevic et al. [5] map onto earliest DNN layers. (a) Top: the reference object (purple) and 11 distractor objects (rest) that rats were trained to discriminate in the behavioral task. For copyright reasons we do not show the stimuli from the original paper, but similar silhouette images were redrawn by hand or adapted from close-matching clip-art (https://openclipart.org/) for illustrative purposes only. For the model we did use the originals from [5]. The 3 example distractors of Fig 1 in Djurdjevic et al. [5] are highlighted in yellow, green, and light blue. Bottom: in a later phase of the experiment, the rats were trained to tolerate size changes in all objects from 15˚ to 35˚ of visual angle (here only shown for the subset of 4 objects indicated in color on the top). (b,c) Percentage of variance in average discrimination performance of good performing (blue) and poorer performing (red) rats, explained by each DNN layer (estimated by a linear mapping with leave-one-out cross-validation) from AlexNet (b) and VGG16 (c). The linear mapping was estimated using the data available in Figs 1 and 2 of Djurdjevic et al. [5]: all sizes of the 4 example objects indicated by purple, yellow, green, and light blue in (a), and the 30˚ size for the remaining eight objects. Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling stimuli). (d) Top: average discrimination performances of good performing rats, as a function of object size, for the reference object and the 3 example distractors. Middle: average predicted discrimination performances of good performing rats, based on pool2 of AlexNet, which explained the most out-of-sample variance in behavioral performance. Bottom: same as middle, but for conv1 of AlexNet. (e) Same as (d), but for poorer performing rats. Fig 3. Stimulus-level discrimination performance signatures for objects from Djurdjevic et al. [5] map onto earliest DNN layers. (a) Top: the reference object (purple) and 11 distractor objects (rest) that rats were trained to discriminate in the behavioral task. Earliest DNN layers can account for the stimulus-level performance signature across rendered objects and sizes, also for the best performing rats PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 6 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Fig 3. Stimulus-level discrimination performance signatures for objects from Djurdjevic et al. [5] map onto earliest DNN layers. (a) Top: the reference object (purple) and 11 distractor objects (rest) that rats were trained to discriminate in the behavioral task. For copyright reasons we do not show the stimuli from the original paper, but similar silhouette images were redrawn by hand or adapted from close-matching clip-art (https://openclipart.org/) for illustrative purposes only. For the model we did use the originals from [5]. The 3 example distractors of Fig 1 in Djurdjevic et al. [5] are highlighted in yellow, green, and light blue. Bottom: in a later phase of the experiment, the rats were trained to tolerate size changes in all objects from 15˚ to 35˚ of visual angle (here only shown for the subset of 4 objects indicated in color on the top). (b,c) Percentage of variance in average discrimination performance of good performing (blue) and poorer performing (red) rats, explained by each DNN layer (estimated by a linear mapping with leave-one-out cross-validation) from AlexNet (b) and VGG16 (c). The linear mapping was estimated using the data available in Figs 1 and 2 of Djurdjevic et al. [5]: all sizes of the 4 example objects indicated by purple, yellow, green, and light blue in (a), and the 30˚ size for the remaining eight objects. Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling stimuli). (d) Top: average discrimination performances of good performing rats, as a function of object size, for the reference object and the 3 example distractors. Middle: average predicted discrimination performances of good performing rats, based on pool2 of AlexNet, which explained the most out-of-sample variance in behavioral performance. Bottom: same as middle, but for conv1 of AlexNet. (e) Same as (d), but for poorer performing rats. Fig 3. Stimulus-level discrimination performance signatures for objects from Djurdjevic et al. [5] map onto earliest D ion performance signatures for objects from Djurdjevic et al. [5] map onto earliest DNN layers. (a) Top: the reference Fig 3. PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 Mid-level DNN layers can account for natural video categorization behavior Up to this point, we have only discussed studies that used a small number of computer-render- ings of abstract and more naturalistic objects. However, rats have also been shown to be able to learn category rules from more complex natural videos that generalize to novel category exem- plars [6]. In this study, rats were trained on a two-alternative forced choice task in a visual water maze (see Materials and Methods, Behavior, Task paradigms) to classify five-second vid- eos featuring a rat (target category), from phase scrambled versions of the target videos and target-matched natural distractor videos featuring various moving objects. The videos were 24 degrees of visual angle as seen from the choice point where the maze splits in two arms [6]. In each trial the target and distractor were presented simultaneously, with one video on the left and the other on the right. The rats were first trained on a training set of 15 videos (Fig 4A), initially with a fixed target-distractor pairing, followed by a phase where all possible target-dis- tractor combinations were presented. In the subsequent test phase the rats were probed with 40 novel videos with a fixed target-distractor pairing (Fig 4B), without negative feedback for incorrect trials. 7 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats The rats were able to generalize well to the novel videos, independently of temporal information or local luminance cues (although local luminance did explain some of the stimulus-level response variance [6]). To test whether this generalization could be explained by other low-level visual information, we trained DNN-based models incorpo- rating increasingly more layers from AlexNet, VGG16, and VGG11-C3D on the same two alternative forced choice task. VGG11-C3D is a convolutional neural network with 3D spa- tio-temporal filters (16-frame temporal bins) and thus able to also encode temporal fea- tures [33]. The features encoded by VGG11-C3D range from moving edges/blobs and changes in orientation or color, to more complex motion patterns such as moving circular objects or biking-like motions [33]. For training the DNN-based models, we took the first 144 frames of the total of 150 frames (4.8 out of 5s) of each video and split those into nine 16-frame bins. The inputs to each classifier were the time-averaged activations of each 16-frame video clip (Materials and Methods, Computational modeling, Feature extrac- tion). Mid-level DNN layers can account for natural video categorization behavior All three networks performed very similarly, suggesting there was no real advantage of the motion features encoded by VGG11-C3D for this task. For scrambled distractors the models based on the first layers already performed almost at the behavioral level of rats (Fig 4D), but for natural distractors they were at chance and did not exceed rat-level per- formance until the model included conv4 of AlexNet, conv4b of VGG16, or conv3b of VGG11-C3D (Fig 4C). A model trained on a pixel representation achieved only 60% for scrambled distractors and 66% for natural distractors on the test set. We again tested whether a linear combination of DNN activation patterns could accurately predict differences in behavioral performance across target and distractor frame bins, by fitting a regression model on the average rat performances for each target-distractor combination of the training set and then using this linear mapping to predict test set generalization perfor- mances (Materials and Methods, Computational modeling, Predicting behavioral performance signatures). Around 50–65% of variance in behavioral performance for scrambled distractors (Fig 4F) and around 40–50% for natural distractors (Fig 4E) of the test set was explained by middle convolutional DNN layers. The results show that stimulus representations of higher DNN layers are required to achieve generalization performance levels on the rat/non-rat classification task that are comparable to rats. Could this layer effect merely be a consequence of the hierarchical architecture of the DNNs, which have properties that systematically change across layers? Even without training, DNNs can provide a powerful set of features, and trends across lay- ers in the ability to capture experimental data can be explained by the number of linear- nonlinear stages rather than the task-optimization of the network [34]. To test this, we repeated our simulations of the behavioral experiments using randomly initialized DNN models (S1 Fig). Random networks achieved high generalization in the object classification task of Fig 2 and the video classification task of Fig 4 with scrambled distractors but not with natural distractors, where performance levels stayed below those of rats even for the highest layers. Thus, the layer-wise increase in generalization performance in Fig 4C depends on the increasingly more abstract features that were learned when the network was trained on image classification. Mid-level DNN layers can account for natural video categorization behavior Mid-level convolutional layers are required for generalization in a rat versus non-rat categorization task of natural v l layers are required for generalization in a rat versus non-rat categorization task of natural videos in Vinken et al. [6]. (a Fig 4. Mid-level convolutional layers are required for generalization in a rat versus non-rat categorization task of natural videos in Vinken et al. [6]. (a,b) Single frames of all the training (a) and test (b) videos that rats were asked to classify in the behavioral task. Each rat was trained with a subset of 15 videos (purple), and tested for generalization with 40 novel videos (green). Ten test videos (the 5 pairs with natural distractor in the left-most green rectangle) were modified to further probe the rats, for example by reducing playback speed to 25% or equalizing average pixel values in the lower-half of the videos [see 6]. (c,d) Average percentage correct classification of rat versus non-rat frame bin pairs by models incorporating increasingly more DNN layers from AlexNet, VGG16, and VGG11-C3D (the X-axis indicates the highest DNN layer) and for natural (c) and scrambled (d) distractors separately. Performance is evaluated on the training set (purple; all 50 target-distractor combinations) as well as the test set (green; the 25 tested target-distractor pairs, with 25% playback speed and pixel value modifications for 5 pairs [see (b)]). Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. Horizontal lines indicate the average behavioral performance across rats. (e,f) Percentage variance in performance across target-distractor pairs explained by each DNN layer (estimated by a linear mapping fit on the training set data and stimuli only) from AlexNet, VGG16, and VGG11-C3D (left to right) and for natural (e) and scrambled (f) distractors. Same conventions as in as in c,d. All error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling target-distractor pairs). https://doi.org/10.1371/journal.pcbi.1008714.g004 Mid-level DNN layers can account for natural video categorization behavior Together, these results suggest that, in terms of DNNs trained on object recognition, a learned, mid-level representation based on several convolutional and pooling operations is required to explain the observed generalization to novel category exemplars in a classification task of natural videos, while a modest amount of target-distractor-level variance is explained by early to mid-level layers. Again there is no evidence of an added benefit of fully connected layers. 8 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Fig 4. Mid-level convolutional layers are required for generalization in a rat versus non-rat categorization task of natural videos in Vinken et al. [6]. (a,b) Single frames of all the training (a) and test (b) videos that rats were asked to classify in the behavioral task. Each rat was trained with a subset of 15 videos (purple), and tested for generalization with 40 novel videos (green). Ten test videos (the 5 pairs with natural distractor in the left-most green rectangle) were modified to further probe the rats, for example by reducing playback speed to 25% or equalizing average pixel values in the lower-half of the videos [see 6]. (c,d) Average percentage correct classification of rat versus non-rat frame bin pairs by models incorporating increasingly more DNN layers from AlexNet, VGG16, and VGG11-C3D (the X-axis indicates the highest DNN layer) and for natural (c) and scrambled (d) distractors separately. Performance is evaluated on the training set (purple; all 50 target-distractor combinations) as well as the test set (green; the 25 tested target-distractor pairs, with 25% playback speed and pixel value modifications for 5 pairs [see (b)]). Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. Horizontal lines indicate the average behavioral performance across rats. (e,f) Percentage variance in performance across target-distractor pairs explained by each DNN layer (estimated by a linear mapping fit on the training set data and stimuli only) from AlexNet, VGG16, and VGG11-C3D (left to right) and for natural (e) and scrambled (f) distractors. Same conventions as in as in c,d. All error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling target-distractor pairs). https://doi.org/10.1371/journal.pcbi.1008714.g004 g 4. Representations of natural and scrambled videos in the rat lateral extrastriate cortex change in parallel with representations in DNNs DNN-based models suggest that several early and intermediate layers of hierarchical process- ing in an object recognition model are required for a representation that can support natural video categorization. The rodent cortex houses a complex network of specialized higher-order visual areas [35], but is there any evidence for such an intermediate representation in the rat brain? A likely candidate pathway is found in the lateral extrastriate cortex, which anatomically resembles the primate ventral visual stream [36] and shows several functional properties 9 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats thought to be typical of an object recognition pathway, such as increased tolerance for changes in position [7], size, rotation, and illumination [8]. Previously, we investigated neural representations of natural and scrambled videos along the putative rodent ventral stream [37]. In this experiment, we presented the 10 natural videos of the training set in Fig 4A and their 10 scrambled counterparts in randomized order to awake, passively watching rats which were never trained with these videos. The videos were shown at sizes ranging from 50 to 74 degrees of visual angle (as the eye-to-stimulus distance varied according to the position on the screen, which was optimized for each recording site’s receptive field location) separated by a 2 s blank screen, with 10 repetitions per video. We recorded single and multi-unit spiking activity in primary visual cortex (V1), a middle latero- intermediate area (LI), and the most lateral temporal occipital cortex (TO). In the original paper, we reported evidence for an increased dissociation of natural and scrambled videos from V1 to TO, but not for a categorical representation [37]. Here, we investigated the similarity between neural representations of these videos in the putative rat ventral stream and the DNN representations in each layer of each model. We cal- culated representational dissimilarity matrices (RDMs) based on the neural responses per 16-frame time bin (Fig 5A–5C, 180 time bins in total), and compared these with RDMs based on DNN layer activity (Fig 5D–5F). Visual inspection reveals that, similar to the neural RDMs, the DNN layer RDMs show a progression towards an increased dissociation between natural and scrambled videos. However, unlike the neural data, the RDMs of the last fc8 layers suggest a category-like representation differentiating the rat from the non-rat videos (see checkered pattern in the left-upper quadrant). Representations of natural and scrambled videos in the rat lateral extrastriate cortex change in parallel with representations in DNNs A further quantitative comparison between each neural and DNN layer RDM showed that the similarity between neural and DNN layer representations increases for LI and TO in later layers, well beyond the similarity for V1 data (non overlapping 95% confidence intervals, Fig 5G–5I). When we visualized all between-RDM similarities using multidimensional scaling, the plots suggested a progression from V1 to TO parallel to the progression across successive DNN layers (Fig 5J–5L). On average, the distance between the neural and DNN layer RDMs was 2.27 times the average distance between DNN layer RDMs from the three different DNN models. The representational similarity was consistently higher for TO, peaking between pool1 and fc6 for AlexNet (Fig 5G), pool2 and conv5a for VGG16 (Fig 5H), and pool3 and conv5a for VGG11-C3D (Fig 5I). The strong interaction between DNN layer and cortical area could not be explained by randomly initialized DNNs (S3A–S3C Fig), suggesting a significant role of the specific features that were learned when the network was trained on image classification. Overall, these results suggest that the neural representations of the videos in the most lateral visual area TO correspond best to the mid-level representations in DNNs trained on object recognition. Discussion We examined the object recognition capabilities of the rodent visual system, by focusing on several studies and formally assessing the level of abstraction required to explain behavioral performance. Using convolutional neural networks, we assessed at which stage of processing each network can solve the task, to shed light on the extent to which successful generalization performance of rats can be taken as evidence for invariant visual object recognition. This computational approach provides a generally consistent picture of the representations that underlie the behavior of rats: generalization in all tasks could be captured by convolutional lay- ers only, even the most challenging task which required later convolutional layers (Fig 4). This PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 10 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Fig 5. Neural representational dissimilarity matrices for natural and scrambled videos per putative ventral stream area change in parallel with DNN layer representations. (a-c) Neural RDMs for V1, LI, and TO data. Rows and columns correspond to 16-frame bins: nine per video, with first the five rat videos, then the five non-rat videos, and finally the scrambled versions of the natural videos in the same order. The color scale corresponds to percentiles of each RDM’s Pearson correlation distances (excluding the diagonal values). (d-f) Artificial neural network RDMs (Pearson correlation distance) for 4 layers of AlexNet, VGG16, and VGG11-C3D: the first layer, the earliest layer for which the RDM corresponds better to extra-striate (LI/TO) data, the layer with the best normalized correlation with neural data (TO in all three cases), and the last layer (fc8). (g-i) Spearman correlations between each artificial neural network layer RDM and each neural RDM (calculated using above diagonal elements only), normalized by each area’s noise ceiling (V1 in blue, LI in yellow, TO in red). Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. Grey text labels indicate the DNN RDMs shown in (d-f). Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling neural units). (j-l) Two-dimensional representation of similarities between neural and artificial neural network RDMs, derived from applying non-metric multidimensional scaling on Spearman correlation distances between RDMs. Each marker corresponds to an RDM and similar RDMs are plotted closer together. Discussion Despite the ability of the first convolu- tional layer to generalize across different sizes and rotations in Zoccolan et al. [1], later convo- lutional layers could explain more transformation-level variance in performance (Fig 2). On the other hand, performance differences between objects and sizes in Djurdjevic et al. [5] were equally well explained by the earliest convolutional layers, also for the best performing rats (Fig 3). For the natural movies of Vinken et al. [6] the stimulus-pair-level explained variance was generally low, which is more consistent with the finding that the same DNNs fail to account for the image-level behavioral signatures of primates [27]. Later convolutional layers matched the representational geometry increasingly better for extrastriate areas in the rat visual cortex (Fig 5), which is consistent with our finding in Fig 4 that only models that also include later layers generalized to novel exemplars from Vinken et al. [6]. These findings might seem to contradict Cadena et al. [34], who did not find evidence for a hierarchical correspondence between mouse visual areas and layers in a similar DNN. However, Cadena et al. [34] targeted three extrastriate areas that all border V1 and are moder- ately (lateromedial, LM) to strongly (anterolateral, AL; rostrolateral, RL) associated with the dorsal stream. Both AL and RL show conclusive anatomical and functional dorsal stream properties [36,38]. LM is a gateway to the putative ventral stream, but its role is considered divided because of dense connections to both ventral and dorsal areas [36,39] and inconsistent evidence for ventral-like computations [38,40]. Therefore, it is no surprise that these areas do not map onto the hierarchy of a ventral-like DNN. In contrast, we targeted extrastriate areas LI and TO, which are both positioned laterally to LM and span together with V1 the extent of the putative rodent ventral stream [7]. Being one step downstream from LM, LI lies at the cen- ter of this pathway and is both functionally and anatomically considered ventral-like [36,38,39]. Thus, our findings are consistent with functional and anatomical evidence and do not contradict the results of [34]. While our results show that generalization in rodent object vision experiments can be explained by representations in convolutional layers, in primates fully connected layers of the same or similar DNNs capture perceived shape similarity better [18,23,41]. This suggests that the behavioral studies in rats provide evidence for object representations which are markedly less abstract than primates. Discussion Text labels indicate the neural RDMs and the DNN RDMs shown in (d-f). https://doi.org/10.1371/journal.pcbi.1008714.g005 Fig 5. Neural representational dissimilarity matrices for natural and scrambled videos per putative ventral stream area change in parallel with DNN layer representations. (a-c) Neural RDMs for V1, LI, and TO data. Rows and columns correspond to 16-frame bins: nine per video, with first the five rat videos, then the five non-rat videos, and finally the scrambled versions of the natural videos in the same order. The color scale corresponds to percentiles of each RDM’s Pearson correlation distances (excluding the diagonal values). (d-f) Artificial neural network RDMs (Pearson correlation distance) for 4 layers of AlexNet, VGG16, and VGG11-C3D: the first layer, the earliest layer for which the RDM corresponds better to extra-striate (LI/TO) data, the layer with the best normalized correlation with neural data (TO in all three cases), and the last layer (fc8). (g-i) Spearman correlations between each artificial neural network layer RDM and each neural RDM (calculated using above diagonal elements only), normalized by each area’s noise ceiling (V1 in blue, LI in yellow, TO in red). Black and grey bars on the X-axis indicate layer blocks and markers indicate layer types (see legend insert); the division between convolutional and fully connected layer blocks is indicated by a dashed line. Grey text labels indicate the DNN RDMs shown in (d-f). Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling neural units). (j-l) Two-dimensional representation of similarities between neural and artificial neural network RDMs, derived from applying non-metric multidimensional scaling on Spearman correlation distances between RDMs. Each marker corresponds to an RDM and similar RDMs are plotted closer together. Text labels indicate the neural RDMs and the DNN RDMs shown in (d-f). https://doi.org/10.1371/journal.pcbi.1008714.g005 picture is much more precise than in earlier papers, which generally relied on the assumption that rats could not have used low-level representations to generalize across variations in appearance and identity preserving transformations; an assumption that up until now had not been tested scientifically. As such, this computational approach provides a deeper and more principled understanding of the combined behavioral and neural data available in the literature. PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 11 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Besides the general performance, there were interesting commonalities between behavior and DNN representations at a more fine-grained level. Discussion First, in Zoccolan et al. [1], the overall performance level is surprisingly easy to explain based on activations in the first convolutional layer. A more in-depth look at the behavioral performance signature across object transformations turned out to be more rel- evant for showing at least mid-level processing. Second, in Djurdjevic et al. [5], the behavioral signature for good performers is actually consistent with early convolutional representations, contrary to the intuitive assumption of the authors that these animals must rely on a complex strategy. Third, a paradigm where only a single stimulus is presented on every trial [as in 1,5] has been suggested [9] to probe more complex strategies compared to a two-alternative forced choice task where the target and distractor can be directly compared [6]. However, here we show that, if anything, the latter paradigm can provide evidence for representations that are at least as high level as the paradigm used by Zoccolan et al. [1]. These considerations are of critical importance for future research on rodent vision. For example, one of the most large-scale initiatives to investigate the visual system in rodents is headed by the Allen Institute, where the most recent efforts yielded nearly 100,000 recorded neurons [47,48]. The interpretation of the neural data is greatly facilitated when behavioral read-outs are available. The current study points to two challenges in this respect. First, we need a computational approach to validate assumptions about the difficulty of the visual tasks and the strategies used for task performance. Second, prior to starting data collection, a computational approach would be highly valuable at the stage of deciding which stimuli to use. Even for the large scale neural recordings [48], the bottleneck for comprehensively characteriz- ing higher-level visual processing could be in the design of the stimulus set rather than the number of recorded neurons. Finally, it is important to emphasize that the fact that we have not found any evidence for truly high-level visual object recognition behavior does not imply that it is not there or cannot be there. For example, a high generalization performance in the video categorization experi- ment required higher layer representations than the other experiments, underlining that the evidence we have is not only limited by the capabilities of the rat visual system, but also by the nature of the task. Discussion For the similarity between neural representations in primate infer- otemporal cortex and DNN layers the story is less consistent: in some cases it does peak at the fully connected layers [22,42], whereas in other studies it peaks at the last convolutional layer [16,18,43,44], suggesting that it might depend on the particular stimulus set, how anterior in the ventral stream the neural data were recorded [22], and/or the network architecture [45]. This shows the road ahead for future studies on comparing behavior and neurophysiology between animal species: preferably the exact same stimuli and tasks would be used, designed to be able to differentiate among representations and strategies of varying complexity. Currently no such data are available. It is important to note that the DNNs used in this study are supervised, feedforward neural networks trained on object recognition, which may constrain the features that are encoded by each layer. These models were not designed to model rodent vision [although the overall con- clusions did hold when the stimuli were resized so that receptive field sizes in early DNN layers match V1 receptive field sizes (S2 and S3D–S3F Figs)], because they are neither optimized for the same goals, nor are they subject to the same biological constraints of either the rodent or primate visual systems. Therefore, we do not claim that feedforward DNNs trained on object recognition are optimal models of the putative rodent ventral stream, which potentially encodes different information compared to the primate ventral stream. On the other hand, rodents do rely on vision for navigation, which demands to some extent properties that are PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 12 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats fundamental to object recognition, such as tolerance for changes in view-point and illumina- tion [46]. Regardless, DNNs are mechanistic models that can capture the steps of information processing required to solve visual recognition of the main object in natural images. While these steps of information processing in DNNs do not capture all aspects of primate vision [24,25], on a macroscale they do map onto successive stages of primate ventral stream process- ing [15,16,31]. Thus, as a general modeling framework for object vision, DNNs allowed us to computationally evaluate previous assumptions about the level of abstraction required for gen- eralization in behavioral experiments. Our results provide important qualifications for researcher assumptions about the stimulus set or task complexity. PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 Behavior Task paradigms. In the studies of Zoccolan et al. [1] and Djurdjevic et al. [5], rats were trained on a yes-no task in an operant box with liquid rewards. Briefly, a rat could initiate a trial by licking a central sensor, starting the presentation of a single stimulus with a default presenta- tion time of 3s. Only one object was presented per trial, and the rats had to associate each object identity with a left or right feeding tube. A correct response was given by licking the feeding tube associated with the presented stimulus, which prompted the delivery of the liquid reward. In the study of Vinken et al. [6], rats were trained on a two-alternative forced choice task (2AFC) in a V-shaped visual water maze [50], with a target video presented at the end of one arm of the maze, and a distractor video at the end of the other arm. A trial was started by plac- ing the rat in the water at a start position where both stimuli were visible. During training, a platform was located only under the target video, which the rat had to reach in order to escape the water maze and end the trial. Both videos were played repeatedly until the trial ended. In the subsequent test phase, a platform was placed under both the target and distractor for trials with novel stimuli, to exclude additional learning from negative feedback for incorrect trials. For further methodological details about the experimental setups and the tasks we refer to For further methodological details about the experimental setups and the tasks, we refer to the original papers. Data and stimulus extraction. We extracted the data and stimuli from the manuscripts of Zoccolan et al. [1] and Djurdjevic et al. [5]. The object images were directly copied from Fig 2A and Fig 1A, respectively. The behavioral data were copied from the values displayed in Fig 2B of Zoccolan et al. [1] (percentage correct based on 70–90 trials for each size and azimuth- rotation and per animal, averaged across N = 6 rats), and extracted from Fig 1C and 1D of Djurdjevic et al. Behavior [5] using WebPlotDigitizer 4.2 (https://apps.automeris.io/wpd/; percentage correct based on an unspecified number of trials for each size of the target object and of three distractor objects, as well as for the 30˚ size of all remaining distractor objects, for each of N = 6 rats). For the analyses of the natural video categorization experiment, we used the origi- nal movies from Vinken et al. [6], and had access to the original data which are available at http://doi.org/10.17605/OSF.IO/4W39D (percentage correct based on 14–49 [M = 30, SD = 10] trials for each target-distractor combination pooled across N = 5 rats). Discussion It is possible that none of the studies discussed here really pushed the limits of rodent object vision. A road ahead for addressing this question is to use DNNs to construct stimulus sets and design paradigms that explore the boundaries of rodent vision by getting the best out of them. This will likely also include considerations about the ecological validity of the tasks from the perspective of rodents [34,46,49]. Another interesting avenue is the use of paradigms that investigate visual strategies by means of experimental manipulations that also made the task more challenging and exclude some of the simplest pixel-based strategies. For example, Alemi-Neissi et al. [3] presented sti- muli covered by masks that randomly occluded parts of the objects, which effectively excludes the possibility of the animals using one highly specific local-luminance based strategy. In addition, in a second part of their study, Djurdjevic et al. [5] presented random varia- tions of the reference image to infer perceptual templates, showing that a template-matching model using only one fixed perceptual template could not account for the animal’s PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 13 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats performance across image manipulations such as changes in object size. Given the difficulty to relate these strategies to the level of abstraction of the underlying representations, it will be interesting to combine a more computational DNN approach with these template paradigms. In summary, we used convolutional deep neural networks for a comprehensive and quanti- tative assessment of the level of abstraction required to explain rodent visual object recogni- tion. A combination of behavioral and neural results reveals a level of invariance comparable to mid-level, classification-trained DNN representations, consistent with the idea of a visual system that is reasonably advanced but not the primary modality. The main conclusion can be phrased in different ways, depending on whether one takes a glass half-full or half-empty per- spective. On the one hand, our findings confirm that rodent visual task performance is non- trivial, displays a certain degree of invariance, and requires multi-layer networks to be simu- lated. On the other hand, the behavioral performance as well as neural responses point to rep- resentations of a limited level of abstraction relative to primate vision. conv: convolutional layer, suffixes a, b, or c are used if a block contains multiple conv layers; norm: local response normalization; pool: max pooling operation; fc: fully connected layer. Neurophysiology Data. The neurophysiological data are from a study published previously in Vinken et al. [37] (N = 7 rats) and consist of single cell and multi-unit responses to natural videos recorded Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 14 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats from primary visual cortex (N = 50 single cells, N = 25 multi-unit sites), latero-intermediate visual area (N = 53 single cells, N = 33 multi-unit sites), and temporal occipital cortex (N = 52 single cells, N = 26 multi-unit sites). For methodological details about the experimental setup and recordings, we refer to the original paper. Representational dissimilarity matrices. For comparison with DNNs (see Computa- tional modeling, Comparing neural and DNN stimulus representations), we calculated neural representational dissimilarity matrices [RDMs; 51]. In short, for each of the 20 five-second vid- eos we considered the first nine 16-frame bins (533ms each–together covering 144 out of the full length of 150 video frames, or 4.8 out of 5 seconds) to match the temporal bins explained under Computational modeling. This yielded a total of 180 16-frame bins. Next, for each of these 16-frame video bins, we calculated a neural response vector with the average standard- ized firing rate of each single and multi-unit response, taking into account a temporal shift cor- responding to the response latency which was estimated separately for each neuron or site [see 37]. This resulted in 180 response vectors (one per 16-frame bin of each stimulus), which were correlated pairs-wise (Pearson r) in order to obtain RDMs with distances 1−r. Stimulus pairs that elicit a similar neural response pattern result in a lower dissimilarity. https://doi.org/10.1371/journal.pcbi.1008714.t001 https://doi.org/10.1371/journal.pcbi.1008714.t001 Computational modeling We calculated these activations for every convolutional, normalization, max pooling, and fully connected layer, standardized the values across inputs and reduced the dimensionality using principal component analysis (always retaining the full set of principle components). In the case of videos, activations were averaged over time within each 16-frame bin before dimensionality reduction using principal component analysis. For VGG11-C3D this was done by taking the mean activation of each spatiotemporal convolutional kernel across the temporal dimension. For AlexNet and VGG16 (which were pre-trained on static images) we averaged outputs across frames for each 16-frame bin, by taking the mean activation of each unit across frames. Note that the parameters for the standardization and the transformation to principal com- ponent space were always calculated only based on the stimulus set that the animals were trained on in the behavioral experiments, thus excluding the stimuli of the test set (except for the Djurdjevic et al. [5] experiments, where there was no separate test set for generalization). The resulting feature vectors were then used to model the behavioral task for comparison with neural representations. Modeling behavioral tasks. Visual object recognition experiments were modeled using pre-trained DNNs, by placing a new linear decoder layer on top of an original network layer and training the new decoder on the specific task and stimuli of the experiment. To simulate visual systems with lower levels of ventral stream-like processing, we not only trained a decoder on top of the penultimate DNN layer [as in 27,28], but trained separate decoders for each DNN layer, each constituting a different model (Fig 1B). For each task, the decoder was a linear support vector machine (SVM) classifier trained on the same binary tasks that the rats were trained on (object 1 vs 2 for [1]; reference object vs distractors for [5]; rat movies versus distractor movies for [6]), using standardized DNN layer activations in principle component space as inputs (note that this is equivalent to training a fully connected neural network layer with binary output). As a baseline control, we also trained a classifier on a pixel representation obtained by transforming pixel values to a principal component space calculated from the stimulus set that the animals were trained on (i.e. identical to what we did with DNN activa- tions). Computational modeling DNNs. We used three trained DNN architectures as computational models of visual pro- cessing in the primate ventral stream: AlexNet, VGG16, and VGG11-C3D. AlexNet [52] and VGG16 [53] were both taken from the MATLAB 2017b Deep Learning Toolbox and had been pre-trained on the ImageNet dataset [32] to classify images into 1000 object categories. Both architectures have been extensively used to model ventral stream processing and object per- ception [14,16–20,22,23]. For the experiments involving videos we also used VGG11-C3D [called C3D in 33], an architecture which is similar to VGG11 [53], but performs convolution and pooling across the two spatial dimensions and a temporal dimension (operating on 16-frame bins). This network had been pre-trained on the Sports-1M dataset [54] to classify videos into 487 sports categories and has previously been used to model human brain responses to natural videos [55]. All three networks consist of a sequence of convolutional and max pooling layers followed by three fully connected layers. Only AlexNet also includes local response normalization layers. Each convolutional and all except the last fully connected layers were followed by a rectifying linear activation function (ReLU). We always extracted activa- tions of individual units from each layer before the ReLU. For each model the full set of layers is typically divided in eight layer blocks (see Table 1). Table 1. Nomenclature used to refer to each architecture’s layers and the division across layer blocks. Table 1. Nomenclature used to refer to each architecture’s layers and the division across layer blocks. Layer block AlexNet VGG16 VGG11-C3D 1 conv1 norm1 pool1 conv1a, b pool1 conv1a pool1 2 conv2 norm2 pool2 conv2a, b pool2 conv2a pool2 3 conv3 conv3a, b, c pool3 conv3a, b pool3 4 conv4 conv4a, b, c pool4 conv4a, b pool4 5 conv5 pool5 conv5a, b, c pool5 conv5a, b pool5 6 fc6 fc6 fc6 7 fc7 fc7 fc7 8 fc8 fc8 fc8 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 15 / 15 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats Feature extraction. Each network has learned a rich series of feature representations which can be accessed from every layer by obtaining unit activations for an input stimulus. Computational modeling Each classifier was trained on the experiment’s training set only, and tested for generali- zation on the test set to assess whether its feature representation can support the task (i.e., whether the object/category representations are such that a linear combination of features optimized for the training set generalizes to the test set). Each classifier was trained using the MATLAB 2017b function fitclinear, with the limited-memory BFGS solver and default regular- ization. For yes-no tasks with one stimulus per trial [1,5], task performance was evaluated for each stimulus individually: classification was considered correct if the stimulus fell on the cor- rect side of the SVM decision boundary (Fig 1D). For the two-alternative forced choice task, which had both a target and distractor per trial [6], a model response was considered correct if the target stimulus fell on the correct side of the distractor, relative to the decision boundary. That is, when (a) the target and distractor were both on their respective correct side of the SVM boundary (Fig 1E i), (b) both were on the target side, but the target was further away from the decision boundary (Fig 1E ii), or (c) both were on the distractor side, but the target was closer to the decision boundary (Fig 1E iii). Code for this model is available at http://doi. org/10.17605/OSF.IO/4W39D. In the behavioral experiments, rats were not head fixed or fixating, so the actual retinal pro- jection of an object image could vary from trial to trial. Importantly, rats ware trained with this variability and the distributions of retinal projections during training trials should cover those during test trials. We found that explicitly modeling such variability by varying object Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 16 / 21 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats positions during training and testing for the Zoccolan et al. [1] experiment did not lead to qualitatively different results, thus we proceeded without such variability. Predicting behavioral performance signatures. The probability of correct response by the rats was not the same for every trial, but varied depending on the presented stimulus (i.e. size and rotation in [1], object and size in [5], and category exemplar in [6]). We call this pat- tern of percentages correct across training or test images in the task the behavioral perfor- mance signature [27]. Computational modeling We used a linear regression approach to assess whether a linear mapping could predict these behavioral performance signatures from activation patterns in DNN layers. Specifically, for each DNN layer, a partial least squares (PLS) regression of the (logit-transformed) average rat performance onto the DNN layer features was fit using stimuli and data from the training phase of the experiment only. For the experiment with a 2AFC task [6] we used the DNN features of the target stimulus minus the DNN features of the distractor stimulus as predictor variables for each stimulus pair. We used three PLS components for the experiment with two objects [1], and ten components otherwise. The accuracy of each model at predicting behavioral performance signatures was assessed by calculating the percentage explained variance as the squared Pearson correlation between predicted performance values ypred and observed (logit-transformed) performance values yobs: corr(ypred, yobs)2. For experiments with an independent generalization phase [1,6], the PLS regression was fit using all stimuli and data from the training phase of the experiment, and thus out-of-sample prediction accuracy was assessed on the independent test set consisting of all stimuli and data from the generalization phase. Leave-one-out cross-validation was used instead when there was no independent generalization phase in the experiment (i.e., Fig 3). Comparing neural and DNN stimulus representations. In order to estimate how closely the representational geometry of each DNN layer matched that of visual areas along the puta- tive rodent ventral stream, we calculated RDMs based on DNN features. As for the neural data, the stimulus feature vectors (one vector for each 16-frame video bin) were correlated pairs-wise in order to obtain an RDM for each DNN layer. As for neural RDMs, stimulus pairs that share a similar representation across features in a layer result in a lower dissimilarity. We then quantified the correspondence between neural and DNN RDMs by calculating the Spear- man correlation between off-diagonal upper halves of the matrices. We normalized the corre- lations between neural and DNN RDMs by dividing by each area’s noise ceiling. To estimate the noise ceiling we split the trials per movie in two halves and computed the Spearman corre- lation between the two resulting neural RDMs (one from each split half). The noise ceiling was the Spearman-Brown-corrected average (across 1000 random splits) split-half correlation. Supporting information Main results of Fig 5, for randomly initialized AlexNet and for stimuli scaled to match RF sizes in early DNN layers to V1 RF sizes. (a-c) We repeated the analyses of Fig 5G and 5J with 10 randomly initialized AlexNet architectures (see S1 Fig). Error bounds ar 95% confidence intervals calculated using Jackknife standard error estimates (resampling t 10 random initializations). (d-f) The analyses of Fig 5D, 5G and 5J, after downsizing the vid to 137x137 pixels (60%) to match 62 degrees of visual angle (see explanation S2 Fig). Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling neural units). All other conventions match those of the corresponding figures the main text. (TIF) conv1 (11x11 pixels), pool1 (19x19 pixels), and conv2 (51x51 pixels) corresponded to 5, 8.6, and 23.2 degrees of visual angle, respectively, and repeated the analyses of the main figures in AlexNet. All stimuli were downsized from a default size of 227x227 pixels to match the reported presentation size in degrees of visual angle (followed by padding to the DNN input size of 227x227 pixels). (a,b) the analyses of Fig 2C and 2E, after downsizing the images to 129x129 pixels (57%) to match the largest object size of 40 degrees of visual angle. (c) the analy- sis of Fig 3B, after downsizing the images to 99x99 pixels (44%) to match the largest object size of 35 degrees of visual angle. (d,e) the analyses of Fig 4C–4F, after downsizing the videos to 53x53 pixels (23%) to match 24 degrees of visual angle. All conventions match those of the cor- responding figures in the main text. The effects of downsizing the stimuli are most notable in (d) and (e), where the stimuli were reduced to a much lower resolution, leading to the model requiring a higher DNN layer to reach rat-level accuracies and explain most stimulus-level var- iance. S3 Fig. Main results of Fig 5, for randomly initialized AlexNet and for stimuli scaled to match RF sizes in early DNN layers to V1 RF sizes. (a-c) We repeated the analyses of Fig 5D, 5G and 5J with 10 randomly initialized AlexNet architectures (see S1 Fig). Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling the 10 random initializations). Supporting information (d-f) The analyses of Fig 5D, 5G and 5J, after downsizing the videos to 137x137 pixels (60%) to match 62 degrees of visual angle (see explanation S2 Fig). Error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling neural units). All other conventions match those of the corresponding figures in the main text. (TIF) Acknowledgments We thank Davide Zoccolan and Thomas P. O’Connell for helpful comments on this work. We thank Davide Zoccolan and Thomas P. O’Connell for helpful comments on this work. Author Contributions Conceptualization: Kasper Vinken, Hans Op de Beeck. Formal analysis: Kasper Vinken. Funding acquisition: Kasper Vinken, Hans Op de Beeck. Supervision: Hans Op de Beeck. Visualization: Kasper Vinken. Writing – original draft: Kasper Vinken. Writing – review & editing: Kasper Vinken, Hans Op de Beeck. Conceptualization: Kasper Vinken, Hans Op de Beeck. Formal analysis: Kasper Vinken. Funding acquisition: Kasper Vinken, Hans Op de Beeck. Funding acquisition: Kasper Vinken, Hans Op de Beeck. Supervision: Hans Op de Beeck. Supervision: Hans Op de Beeck. Visualization: Kasper Vinken. Writing – original draft: Kasper Vinken. Writing – original draft: Kasper Vinken. Writing – review & editing: Kasper Vinken, Hans Op de Beeck. Supporting information S1 Fig. Randomly initialized AlexNet can account for the main results of Figs 2 and 3, but not Fig 4. We repeated the analyses of the main figures with 10 randomly initialized AlexNet architectures (using the same scheme as the trained version, i.e., uniform Glorot initialization for the weights [56] and zero bias). (a,b) the analyses of Fig 2C and 2E. (c) the analysis of Fig 3D. (d,e) the analyses of Fig 4C–4F. All error bounds are 95% confidence intervals calculated using Jackknife standard error estimates (resampling the 10 random initializations). All other conventions match those of the corresponding figures in the main text. (TIF) S2 Fig. Main results of Figs 2–4, for stimuli scaled to match receptive field (RF) sizes in early DNN layers to V1 RF sizes. The V1 RF sizes reported in pigmented rats cover a broad range between 3 and 20+ degrees of visual angle [8,57,58]. To match DNN RF sizes in early layers with the range observed in rat V1, we downsized the stimuli so that the RF sizes in 17 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats conv1 (11x11 pixels), pool1 (19x19 pixels), and conv2 (51x51 pixels) corresponded to 5, 8.6 and 23.2 degrees of visual angle, respectively, and repeated the analyses of the main figures AlexNet. All stimuli were downsized from a default size of 227x227 pixels to match the reported presentation size in degrees of visual angle (followed by padding to the DNN inpu size of 227x227 pixels). (a,b) the analyses of Fig 2C and 2E, after downsizing the images to 129x129 pixels (57%) to match the largest object size of 40 degrees of visual angle. (c) the an sis of Fig 3B, after downsizing the images to 99x99 pixels (44%) to match the largest object of 35 degrees of visual angle. (d,e) the analyses of Fig 4C–4F, after downsizing the videos to 53x53 pixels (23%) to match 24 degrees of visual angle. All conventions match those of the responding figures in the main text. The effects of downsizing the stimuli are most notable (d) and (e), where the stimuli were reduced to a much lower resolution, leading to the mod requiring a higher DNN layer to reach rat-level accuracies and explain most stimulus-level iance. (TIF) S3 Fig. References 1. Zoccolan D, Oertelt N, DiCarlo JJ, Cox DD. A rodent model for the study of invariant visual object recog- nition. Proc Natl Acad Sci U S A. 2009; 106: 8748–8753. https://doi.org/10.1073/pnas.0811583106 PMID: 19429704 2. Tafazoli S, Di Filippo A, Zoccolan D. Transformation-tolerant object recognition in rats revealed by visual priming. J Neurosci. 2012; 32: 21–34. https://doi.org/10.1523/JNEUROSCI.3932-11.2012 PMID: 22219267 2. Tafazoli S, Di Filippo A, Zoccolan D. Transformation-tolerant object recognition in rats revealed by visual priming. J Neurosci. 2012; 32: 21–34. https://doi.org/10.1523/JNEUROSCI.3932-11.2012 PMID: 22219267 18 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats 3. Alemi-Neissi A, Rosselli FB, Zoccolan D. Multifeatural shape processing in rats engaged in invariant visual object recognition. J Neurosci. 2013; 33: 5939–5956. https://doi.org/10.1523/JNEUROSCI.3629- 12.2013 PMID: 23554476 4. Rosselli FB, Alemi A, Ansuini A, Zoccolan D. Object similarity affects the perceptual strategy underlying invariant visual object recognition in rats. Front Neural Circuits. 2015; 9: 1–22. https://doi.org/10.3389/ fncir.2015.00001 PMID: 25713515 5. Djurdjevic V, Ansuini A, Bertolini D, Macke JH, Zoccolan D. Accuracy of Rats in Discriminating Visual Objects Is Explained by the Complexity of Their Perceptual Strategy. Curr Biol. 2018; 28: 1005–1015. e5. https://doi.org/10.1016/j.cub.2018.02.037 PMID: 29551414 6. Vinken K, Vermaercke B, Op de Beeck HPHP. Visual categorization of natural movies by rats. J Neu- rosci. 2014; 34: 10645–10658. https://doi.org/10.1523/JNEUROSCI.3663-13.2014 PMID: 25100598 7. Vermaercke B, Gerich FJ, Ytebrouck E, Arckens L, Op de Beeck HP, Van den Bergh G. Functional spe- cialization in rat occipital and temporal visual cortex. J Neurophysiol. 2014; 112: 1963–1983. https://doi. org/10.1152/jn.00737.2013 PMID: 24990566 8. Tafazoli S, Safaai H, De Franceschi G, Rosselli FB, Vanzella W, Riggi M, et al. Emergence of transfor- mation-tolerant representations of visual objects in rat lateral extrastriate cortex. Elife. 2017; 6: 1–39. https://doi.org/10.7554/eLife.22794 PMID: 28395730 9. Zoccolan D. Invariant visual object recognition and shape processing in rats. Behav Brain Res. 2015; 285: 10–33. https://doi.org/10.1016/j.bbr.2014.12.053 PMID: 25561421 10. Minini L, Jeffery KJ. Do rats use shape to solve “shape discriminations”? Learn Mem. 2006; 13: 287– 297. https://doi.org/10.1101/lm.84406 PMID: 16705141 11. Kell AJE, Bokor SL, Jeon Y-N, Toosi T, Issa EB. Conserved core visual object recognition across simian primates: Marmoset image-by-image behavior mirrors that of humans and macaques. bioRxiv. 2020; 1–32. https://doi.org/10.1101/2020.10.19.345561 12. Dicarlo JJ, Zoccolan D, Rust NC. How does the brain solve visual object recognition? Neuron. 2012; 73: 415–34. https://doi.org/10.1016/j.neuron.2012.01.010 PMID: 22325196 13. References Deep Learning: The Good, the Bad, and the Ugly. Annu Rev Vis Sci. 2019; 5: 399–426. https:// doi.org/10.1146/annurev-vision-091718-014951 PMID: 31394043 26. Kietzmann TC, McClure P, Kriegeskorte N. Deep Neural Networks in Computational Neuroscience. Oxford Research Encyclopedia of Neuroscience. Oxford University Press; 2019. pp. 1–28. https://doi. org/10.1093/acrefore/9780190264086.013.46 27. Rajalingham R, Issa EB, Bashivan P, Kar K, Schmidt K, DiCarlo JJ. Large-scale, high-resolution com- parison of the core visual object recognition behavior of humans, monkeys, and state-of-the-art deep artificial neural networks. J Neurosci. 2018; 38: 7255–7269. https://doi.org/10.1523/JNEUROSCI.0388- 18.2018 PMID: 30006365 28. Kar K, Kubilius J, Schmidt K, Issa EB, DiCarlo JJ. Evidence that recurrent circuits are critical to the ven- tral stream’s execution of core object recognition behavior. Nat Neurosci. 2019; 354753. https://doi.org/ 10.1038/s41593-019-0392-5 PMID: 31036945 29. Hung CP, Kreiman G, Poggio T, DiCarlo JJ. Fast readout of object identity from macaque inferior tem- poral cortex. Science. 2005; 310: 863–866. https://doi.org/10.1126/science.1117593 PMID: 16272124 30. Majaj NJ, Hong H, Solomon EA, DiCarlo JJ. Simple learned weighted sums of inferior temporal neuronal firing rates accurately predict human core object recognition performance. J Neurosci. 2015; 35: 13402–13418. https://doi.org/10.1523/JNEUROSCI.5181-14.2015 PMID: 26424887 31. Nonaka S, Majima K, Aoki SC, Kamitani Y. Brain Hierarchy Score: Which Deep Neural Networks are Hierarchically Brain-Like? SSRN Electron J. 2020. https://doi.org/10.2139/ssrn.3664362 32. Russakovsky O, Deng J, Su H, Krause J, Satheesh S, Ma S, et al. ImageNet Large Scale Visual Recog- nition Challenge. Int J Comput Vis. 2015; 115: 211–252. https://doi.org/10.1007/s11263-015-0816-y 33. Tran D, Bourdev L, Fergus R, Torresani L, Paluri M. Learning Spatiotemporal Features with 3D Convo- lutional Networks. Int J Comput Vis. 2014; 101: 6–21. https://doi.org/10.1007/s11263-012-0542-7 34. Cadena SA, Sinz FH, Muhammad T, Froudarakis E, Cobos E, Walker EY, et al. How well do deep neu- ral networks trained on object recognition characterize the mouse visual system? NeurIPS 2019 Work Neuro AI Blind Submiss. 2019; 1–5. Available from: https://openreview.net/forum?id=rkxcXmtUUS. 35. Glickfeld LL, Olsen SR. Higher-Order Areas of the Mouse Visual Cortex. Annu Rev Vis Sci. 2017; 3: 251–273. https://doi.org/10.1146/annurev-vision-102016-061331 PMID: 28746815 36. Wang Q, Sporns O, Burkhalter A. Network Analysis of Corticocortical Connections Reveals Ventral and Dorsal Processing Streams in Mouse Visual Cortex. J Neurosci. 2012; 32: 4386–4399. https://doi.org/ 10.1523/JNEUROSCI.6063-11.2012 PMID: 22457489 37. Vinken K, Van den Bergh G, Vermaercke B, Op de Beeck HP. Neural Representations of Natural and Scrambled Movies Progressively Change from Rat Striate to Temporal Cortex. Cereb Cortex. 2016; 26: 3310–3322. https://doi.org/10.1093/cercor/bhw111 PMID: 27146315 38. References Schilling A, Metzner C, Rietsch J, Gerum R, Schulze H, Krauss P. How deep is deep enough?—Quanti- fying class separability in the hidden layers of deep neural networks. arXiv. 2018. Available from: http:// arxiv.org/abs/1811.01753. 14. Cadieu CF, Hong H, Yamins DLK, Pinto N, Ardila D, Solomon EA, et al. Deep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognition. PLoS Comput Biol. 2014; 10. https://doi.org/10.1371/journal.pcbi.1003963 PMID: 25521294 15. Yamins DLK, Hong H, Cadieu CF, Solomon EA, Seibert D, DiCarlo JJ. Performance-optimized hierar- chical models predict neural responses in higher visual cortex. Proc Natl Acad Sci. 2014; 111: 8619– 8624. https://doi.org/10.1073/pnas.1403112111 PMID: 24812127 16. Gu¨c¸lu¨ U, van Gerven MAJ. Deep Neural Networks Reveal a Gradient in the Complexity of Neural Rep- resentations across the Ventral Stream. J Neurosci. 2015; 35: 10005–10014. https://doi.org/10.1523/ JNEUROSCI.5023-14.2015 PMID: 26157000 17. Kalfas I, Kumar S, Vogels R. Shape Selectivity of Middle Superior Temporal Sulcus Body Patch Neu- rons. eneuro. 2017. https://doi.org/10.1523/ENEURO.0113-17.2017 PMID: 28660250 18. Kalfas I, Vinken K, Vogels R. Representations of regular and irregular shapes by deep Convolutional Neural Networks, monkey inferotemporal neurons and human judgments. PLOS Comput Biol. 2018; 14: e1006557. https://doi.org/10.1371/journal.pcbi.1006557 PMID: 30365485 19. Pospisil DA, Pasupathy A, Bair W. ‘Artiphysiology’ reveals V4-like shape tuning in a deep network trained for image classification. Elife. 2018; 7: 1–31. https://doi.org/10.7554/eLife.38242 PMID: 30570484 20. Bashivan P, Kar K, DiCarlo JJ. Neural population control via deep image synthesis. Science. 2019;364. https://doi.org/10.1126/science.aay2204 PMID: 31624212 21. Cadena SA, Denfield GH, Walker EY, Gatys LA, Tolias AS, Bethge M, et al. Deep convolutional models improve predictions of macaque V1 responses to natural images. PLoS Comput Biol. 2019; 15: 1–27. https://doi.org/10.1371/journal.pcbi.1006897 PMID: 31013278 22. Zeman AA, Ritchie JB, Bracci S, Op de Beeck H. Orthogonal Representations of Object Shape and Cat- egory in Deep Convolutional Neural Networks and Human Visual Cortex. Sci Rep. 2020; 10: 1–12. https://doi.org/10.1038/s41598-019-56847-4 PMID: 31913322 23. Kubilius J, Bracci S, Op de Beeck HP. Deep Neural Networks as a Computational Model for Human Shape Sensitivity. PLoS Comput Biol. 2016; 12: 1–26. https://doi.org/10.1371/journal.pcbi.1004896 PMID: 27124699 19 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats 24. Geirhos R, Michaelis C, Wichmann FA, Rubisch P, Bethge M, Brendel W. Imagenet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness. 7th Int Conf Learn Represent ICLR 2019. 2019; 1–22. 25. Serre T. References Marshel JH, Garrett ME, Nauhaus I, Callaway EM. Functional specialization of seven mouse visual cor- tical areas. Neuron. 2011; 72: 1040–1054. https://doi.org/10.1016/j.neuron.2011.12.004 PMID: 22196338 39. Wang Q, Gao E, Burkhalter A. Gateways of ventral and dorsal streams in mouse visual cortex. J Neu- rosci. 2011; 31: 1905–18. https://doi.org/10.1523/JNEUROSCI.3488-10.2011 PMID: 21289200 40. Murakami T, Matsui T, Ohki K. Functional segregation and development of mouse higher visual areas. J Neurosci. 2017; 37: 9424–9437. https://doi.org/10.1523/JNEUROSCI.0731-17.2017 PMID: 28847805 41. Bracci S, Ritchie JB, Kalfas I, Op de Beeck HP. The Ventral Visual Pathway Represents Animal Appear- ance over Animacy, Unlike Human Behavior and Deep Neural Networks. J Neurosci. 2019; 39: 6513– 6525. https://doi.org/10.1523/JNEUROSCI.1714-18.2019 PMID: 31196934 42. Khaligh-Razavi SM, Kriegeskorte N. Deep Supervised, but Not Unsupervised, Models May Explain IT Cortical Representation. PLoS Comput Biol. 2014; 10. https://doi.org/10.1371/journal.pcbi.1003915 PMID: 25375136 43. O’Connell TP, Chun MM, Kreiman G. Zero-shot neural decoding of visual categories without prior exemplars. bioRxiv. 2019; 1–18. https://doi.org/10.1101/700344 44. Grossman S, Gaziv G, Yeagle EM, Harel M, Me´gevand P, Groppe DM, et al. Convergent evolution of face spaces across human face-selective neuronal groups and deep convolutional networks. Nat Com- mun. 2019; 10: 4934. https://doi.org/10.1038/s41467-019-12623-6 PMID: 31666525 45. Xu Y, Vaziri-Pashkam M. Limited correspondence in visual representation between the human brain and convolutional neural networks. bioRxiv. 2020. https://doi.org/10.1101/2020.03.12.989376 20 / 21 PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 PLOS COMPUTATIONAL BIOLOGY Evaluating object vision tasks in rats 46. Cox DD. Do we understand high-level vision? Curr Opin Neurobiol. 2014; 25: 187–193. https://doi.org/ 10.1016/j.conb.2014.01.016 PMID: 24552691 47. Siegle JH, Jia X, Durand S, Gale S, Bennett C. A survey of spiking activity reveals a functional hierarchy of mouse corticothalamic visual areas. bioRxiv. 2019; 1–53. https://doi.org/10.1101/805010 48. De Vries SEJ, Lecoq J, Buice MA, Peter A, Ocker GK, Oliver M, et al. A large-scale, standardized physi- ological survey reveals higher order coding throughout the mouse visual cortex. bioRxiv. 2018. https:// doi.org/10.1101/359513 49. Vinken K, Vogels R, Op de Beeck H. Recent Visual Experience Shapes Visual Processing in Rats through Stimulus-Specific Adaptation and Response Enhancement. Curr Biol. 2017. https://doi.org/10. 1016/j.cub.2017.02.024 PMID: 28262485 50. Prusky GT, West PW, Douglas RM. Behavioral assessment of visual acuity in mice and rats. Vision Res. 2000; 40: 2201–2209. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10878281. https://doi. org/10.1016/s0042-6989(00)00081-x PMID: 10878281 51. Kriegeskorte N, Mur M, Bandettini P. Representational similarity analysis—connecting the branches of systems neuroscience. Front Syst Neurosci. 2008; 2: 1–28. https://doi.org/10.3389/neuro.06.001.2008 PMID: 18958245 52. PLOS Computational Biology \| https://doi.org/10.1371/journal.pcbi.1008714 March 2, 2021 References Krizhevsky A, Sutskever I, Hinton GE. ImageNet Classification with Deep Convolutional Neural Net- works. Adv Neural Inf Process Syst. 2012; 1–9. http://dx.doi.org/10.1016/j.protcy.2014.09.007 53. Simonyan K, Zisserman A. Very Deep Convolutional Networks for Large-Scale Image Recognition. Inf Softw Technol. 2014; 51: 769–784. https://doi.org/10.1016/j.infsof.2008.09.005 54. Karpathy A, Toderici G, Shetty S, Leung T, Sukthankar R, Fei-Fei L. Large-Scale Video Classification with Convolutional Neural Networks. 2014 IEEE Conference on Computer Vision and Pattern Recogni- tion. IEEE; 2014. pp. 1725–1732. https://doi.org/10.1109/CVPR.2014.223 55. Gu¨c¸lu¨ U, van Gerven MAJ. Increasingly complex representations of natural movies across the dorsal stream are shared between subjects. Neuroimage. 2017; 145: 329–336. https://doi.org/10.1016/j. neuroimage.2015.12.036 PMID: 26724778 56. Glorot X, Bengio Y. Understanding the difficulty of training deep feedforward neural networks. Proc 13th Int Conf Artif Intell Stat. 2010; 9: 249–256. 57. Fagiolini M, Pizzorusso T, Berardi N, Domenici L, Maffei L. Functional postnatal development of the rat primary visual cortex and the role of visual experience: Dark rearing and monocular deprivation. Vision Res. 1994; 34: 709–720. https://doi.org/10.1016/0042-6989(94)90210-0 PMID: 8160387 58. Girman SV, Sauve´ Y, Lund RD. Receptive field properties of single neurons in rat primary visual cortex. J Neurophysiol. 1999; 82: 301–311. Available from: http://jn.physiology.org/content/82/1/301.short https://doi.org/10.1152/jn.1999.82.1.301 PMID: 10400959 21 / 21
https://openalex.org/W2147490810	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0009319&type=printable	English	null	Spatial Distribution of Dominant Arboreal Ants in a Malagasy Coastal Rainforest: Gaps and Presence of an Invasive Species	PloS one	2,010	cc-by	7,614	Funding: The authors have no support or funding to report. Competing Interests: The authors have declared that no competing interests exist. Competing Interests: The authors have declared that no competing interests exist. Competing Interests: The authors have declared that no competing interests exist. * E-mail: alain.dejean@wanadoo.fr Alain Dejean1, Brian L. Fisher2, Bruno Corbara3, Raymond Rarevohitra4, Richard Randrianaivo5, Balsama Rajemison5, Maurice Leponce6 1 E´cologie des Foreˆts de Guyane, Centre National de la Recherche Scientifique, Unite´ Mixte de Recherche 8172, Campus agronomique, BP 709, Kourou, France, 2 Department of Entomology, California Academy of Sciences, San Francisco, California, United States of America, 3 Laboratoire Microorganismes Ge´nome et Environnement, Centre National de la Recherche Scientifique, Unite´ Mixte de Recherche 6023, Universite´ Blaise Pascal, Aubie`re, France, 4 De´partement de Recherches Forestie`res et Piscicoles, BP 904, Antananarivo, Madagascar, 5 Parc Botanique et Zoologique de Tsimbazaza, BP 4096, Antananarivo, Madagascar, 6 Biological Evaluation Section, Royal Belgian Institute of Natural Sciences, Brussels, Belgium PLoS ONE \| www.plosone.org Introduction dominant’’ species with small colonies of hundreds of workers. Under favorable conditions, some of these non-dominant species are able to develop larger colonies that behave like territorially- dominant ants. Such colonies are known as ‘‘sub-dominant’’. Two colonies of territorially-dominant arboreal species have been known to share the same territory; these ‘‘co-dominant’’ species generally have complementary rhythms of activity: one is often diurnal, while the other is nocturnal [4,7]. Ants dominate the animal communities of tropical rainforest canopies in terms of biomass and number of individuals and have become adapted to the climatic conditions of this environment. Their ecological success is made possible by the fact that most arboreal ants are at least partially herbivorous – feeding on extra- floral nectaries, food bodies, pollen, epiphylls, and sap – and they are also ‘‘cryptic herbivores’’ that feed on hemipteran honeydew [1–3]. In these canopies, a few ‘‘territorially-dominant’’ arboreal ant species are characterized by very populous colonies of up to several million workers, large and/or polydomous nests, and an absolute intra- and interspecific territoriality. The territories of these species are distributed in a mosaic pattern, creating what have become known as ‘‘arboreal ant mosaics’’ [4–8]. Territorially-dominant arboreal ants tolerate within their territories the presence of ‘‘non- Ants dominate the animal communities of tropical rainforest canopies in terms of biomass and number of individuals and have become adapted to the climatic conditions of this environment. Their ecological success is made possible by the fact that most arboreal ants are at least partially herbivorous – feeding on extra- floral nectaries, food bodies, pollen, epiphylls, and sap – and they are also ‘‘cryptic herbivores’’ that feed on hemipteran honeydew [1–3]. Types of arboreal ant nesting sites include: pre-existing cavities (typically branches bored out by xylophagous insects and used by opportunistic species); galleries bored by carpenter ants; silk nests built by weaver ants; and carton nests [7–9]. Certain myrmeco- phytes (plants sheltering ant colonies in hollow structures such as leaf pouches, hollow branches and thorns [9]) grow large enough to reach the canopy, enabling their associated plant-ants to become a part of the ant mosaic [7,8]. Also, along with In these canopies, a few ‘‘territorially-dominant’’ arboreal ant species are characterized by very populous colonies of up to several million workers, large and/or polydomous nests, and an absolute intra- and interspecific territoriality. Citation: Dejean A, Fisher BL, Corbara B, Rarevohitra R, Randrianaivo R, et al. (2010) Spatial Distribution of Dominant Arboreal Ants in a Malagasy Coastal Rainforest: Gaps and Presence of an Invasive Species. PLoS ONE 5(2): e9319. doi:10.1371/journal.pone.0009319 Abstract This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. Funding: The authors have no support or funding to report. Competing Interests: The authors have declared that no competing interests exist. E-mail: alain.dejean@wanadoo.fr Citation: Dejean A, Fisher BL, Corbara B, Rarevohitra R, Randrianaivo R, et al. (2010) Spatial Distribution of Dominant Arboreal Ants in a Malagasy Coastal Rainforest: Gaps and Presence of an Invasive Species. PLoS ONE 5(2): e9319. doi:10.1371/journal.pone.0009319 Editor: Corrie S. Moreau, Field Museum of Natural History, United States of America Received September 16, 2009; Accepted November 20, 2009; Published February 19, 2010 Copyright: 2010 Dejean et al. This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. Funding: The authors have no support or funding to report. Funding: The authors have no support or funding to report. Editor: Corrie S. Moreau, Field Museum of Natural History, United States of America Abstract We conducted a survey along three belt transects located at increasing distances from the coast to determine whether a non-random arboreal ant assemblage, such as an ant mosaic, exists in the rainforest on the Masoala Peninsula, Madagascar. In most tropical rainforests, very populous colonies of territorially dominant arboreal ant species defend absolute territories distributed in a mosaic pattern. Among the 29 ant species recorded, only nine had colonies large enough to be considered potentially territorially dominant; the remaining species had smaller colonies and were considered non-dominant. Nevertheless, the null-model analyses used to examine the spatial structure of their assemblages did not reveal the existence of an ant mosaic. Inland, up to 44% of the trees were devoid of dominant arboreal ants, something not reported in other studies. While two Crematogaster species were not associated with one another, Brachymyrmex cordemoyi was positively associated with Technomyrmex albipes, which is considered an invasive species—a non-indigenous species that has an adverse ecological effect on the habitats it invades. The latter two species and Crematogaster ranavalonae were mutually exclusive. On the other hand, all of the trees in the coastal transect and at least 4 km of coast were occupied by T. albipes, and were interconnected by columns of workers. Technomyrmex albipes workers collected from different trees did not attack each other during confrontation tests, indicating that this species has formed a supercolony along the coast. Yet interspecific aggressiveness did occur between T. albipes and Crematogaster ranavalonae, a native species which is likely territorially dominant based on our intraspecific confrontation tests. These results suggest that the Masoala rainforest is threatened by a potential invasion by T. albipes, and that the penetration of this species further inland might be facilitated by the low density of native, territorially dominant arboreal ants normally able to limit its progression. Citation: Dejean A, Fisher BL, Corbara B, Rarevohitra R, Randrianaivo R, et al. (2010) Spatial Distribution of Dominant Arboreal Ants in a Malagasy Coastal Rainforest: Gaps and Presence of an Invasive Species. PLoS ONE 5(2): e9319. doi:10.1371/journal.pone.0009319 Editor: Corrie S. Moreau, Field Museum of Natural History, United States of America Received September 16, 2009; Accepted November 20, 2009; Published February 19, 2010 Copyright: 2010 Dejean et al. Tree and Ant Species Composition along the Three Transects Tree and Ant Species Composition along the Three Transects Twelve tree species (seven families) were found along the ‘‘coastal transect’’ (situated along the shore), but three species represented 81.3% of the cases (N= 150): Barringtonia butonica (Lecythidaceae), Bruguiera gymnorhiza, and B. sexangula (Rhizophoraceae) (N= 65, 36 and 21 trees, respectively). In addition, 64 clusters of Medinilla sp., an epiphytic Melastomataceae, were noted on 46 trees (30.7%; N = 150). Large numbers of T. albipes workers patrolled the branches, foliage, and trunks of all of the trees, and columns of workers following trails traversed the ground between trees whose crowns were not interconnected. We also noted the presence of several ant species with small colonies (with the exception of introduced Brachymyrmex cordemoyi) that typically nested under Medinilla clusters; these colonies were tolerated by T. albipes (Data set S1). In the two ‘‘inland transects’’, where we recorded 73 tree and eight liana species, the above-cited three tree species were absent, while Medinilla sp. was noted only once. There was little similarity in tree species between these two inland transects (Chao-Jaccard abundance-based index: 0.2060.27 [mean 6 SE]; Fig. 1). Among the 29 total ant species gathered, 18 species were found in both inland transects, and nine were shared species (Chao-Jaccard incidence-based similarity index = 0.5860.18; Fig. 1). Crematogaster ranavalonae was the most common species in both inland transects (Fig. 1) and was found on 35 tree species belonging to 20 different families (Data set S1). Yet, the area studied was characterized by the absence of arboreal ants in the crown of 44% (54/120) and 26% (23/89) of the trees in the inland 1 and inland 2 transects, respectively (Data set S1; Fig. S1), but the difference is not significant (Fisher’s exact-test: P = 0. 17). Together, the absence of natural enemies for introduced species (enemy release hypothesis) and a reduction in the costs associated with intraspecific territoriality for supercolonies permit more energy to be allocated to the production of workers [11]. This results in high worker densities that can monopolize habitat by excluding other ant species through exploitative and interference competition [11,19,20]. Consequently, invasive ants are among the most harmful bioinvaders known. They penetrate ecosystems by disassembling the native ant community, and occasionally even eliminate other species. Tree and Ant Species Composition along the Three Transects By lowering native ant abundance and diversity, they directly or indirectly affect all other organisms that depend on those species, and modify large geographical regions by disrupting native communities [11,21]. Among the potential invasive ant species reported in Madagascar is the dolichoderine ant Technomyrmex albipes which is particularly abundant along the coast [22]. Likely native to the Pacific Islands, T. albipes is an extremely successful tramp species that nests both terrestrially and arboreally with workers that attend a wide range of hemipterans [22]. Its success is also facilitated by the mode of reproduction it shares with other species from the albipes group: the reproductive castes include ergatoid females and males in addition to alates of both sexes, which facilitates the formation of supercolonies [22]. In the past, other species from the albipes group were often misidentified as T. albipes (see p. 70 in Bolton [22]); for example, the ants identified in references made to T. albipes in Terayama [23] and Tsuji & Yamauchi [24] are, in fact, T. brunneus [22]. By climbing trees or using the canopy sledge (a device carried by a blimp), we noted that the colonies of only nine ant species were large enough to occupy at least one tree crown; they occupied adjacent trees in certain cases, as we observed large numbers of workers passing along the branches from one tree to another (see Fig.S1). We verified this by cutting off relatively large branches, which permitted us to sample workers. Here we also noted the presence of other ant species with much smaller colonies (hundreds of workers) corresponding to the status of non-dominant species (see Data set S1). Some were represented by a few foraging workers, while in other cases we found entire colonies represented by at most several hundred individuals after we cut the branches into smaller pieces with a machete and pruning scissors. Of the nine ant species with large colonies, only B. cordemoyi and T. albipes (two introduced, opportunistic nesters that frequently share trees and even nest areas) were recorded along the coast. In the inland transects, five native arboreal species had colonies large enough to possibly be territorially dominant. Crematogaster rasoher- inae, Cr. madagascariensis and Cr. kelleri nested in hollow branches several centimeters wide. Very populous colonies of Crematogaster sp.1 nested in the naturally hollow branches of Vitex beraviensis (Lamiaceae; trees A7 and A12; Data set S1 and Fig. Introduction The territories of these species are distributed in a mosaic pattern, creating what have become known as ‘‘arboreal ant mosaics’’ [4–8]. Territorially-dominant arboreal ants tolerate within their territories the presence of ‘‘non- PLoS ONE \| www.plosone.org 1 February 2010 \| Volume 5 \| Issue 2 \| e9319 February 2010 \| Volume 5 \| Issue 2 \| e9319 Gaps in an Ant Mosaic tested T. albipes worker aggression against a native arboreal species thought to be territorially dominant. territorially-dominant species, invasive ants may form a new component in some ant mosaics, as Technomyrmex albipes (Dolicho- derinae) does in Southeast Asia [10]. Of the approximately 14,000 ant species known, about 150 ‘‘tramp species’’ have been transported and introduced into many parts of the world through human activity, but only some of them have become invasive [11]. The aptitude for invasiveness primarily stems from an intrinsic ability to shift their colony structure. Invasive ants that form a multi-colonial social structure in their native range can switch to supercoloniality (the formation of colonies extending over extremely large areas) in their introduced, and, among certain species, also in their native range. Unicoloniality refers to a population forming a single supercolony over several hundred or even thousand kilometers [11–13]. All of these huge colonies are based on extreme polydomy (multiple nests) and polygyny (multiple queens), so that the workers’ relatedness is very low in most supercolonies [12,13]. Ants are often highly territorial because they can recognize kin through differences or similarities in their cuticular hydrocarbons. Those living within each supercolony are tolerant of one another, freely mixing between different nests even if they are tens of kilometers apart [11–15]. The most plausible hypothesis explaining the absence of aggressiveness between workers from the same supercolony is that recognition and aggressiveness may be genetically based. For this reason, species that form supercolonies are believed to share similar or identical heritable recognition cues that surpass the influence of the environment on the composition of their cuticular hydrocarbons [16–18]. PLoS ONE \| www.plosone.org Tree and Ant Species Composition along the Three Transects Colonies of the last two species, Tapinoma subtile and Camponotus sp.1, occupied the entire crown of a tree on occasion, but were most frequently associated with other ant species (co-dominance). interconnected all trees through their foliage or along trails on the ground. When confronted with Cr. ranavalonae individuals, T. albipes workers initiated 40% of the combats (N = 60) and fought and bit, which corresponds to strongly agonistic behavior. By contrast, T. albipes workers showed a complete lack of intraspecific aggressive- ness (Table 1), even when individuals were gathered from areas separated by up to 8.4 km. This was not the case between Cr. ranavalonae workers belonging to different areas, and which were frequently involved in reciprocal full attacks. Tree and Ant Species Composition along the Three Transects S1), while numerous foraging workers invaded lianas and adjacent trees, demonstrating that their territory was not limited to their host plant. The relatively large Cr. ranavalonae workers built ovoid carton nests 35 to 60 cm in height and 25 to 50 cm in diameter. These nests, reminiscent of those belonging to territorially- dominant arboreal African Crematogaster, were distributed over the main branches of several adjacent trees and interconnected by Initially, the aim of this study was to verify during a snapshot field survey whether or not an ant mosaic existed in a Malagasy rainforest. To do this, we had planned to study three transects, one along the coast, and two others inland. Because we observed T. albipes workers on numerous trees over several kilometers along the coast, we decided to investigate the possible presence of a supercolony. We thus verified whether all of the trees were occupied by this species, and if columns of workers interconnected these trees over large distances along the coast. We conducted standard behavioral assays to establish whether workers gathered from distant areas showed aggression toward one another. We also PLoS ONE \| www.plosone.org February 2010 \| Volume 5 \| Issue 2 \| e9319 February 2010 \| Volume 5 \| Issue 2 \| e9319 2 Gaps in an Ant Mosaic Figure 1. Relative frequency for the 29 ant species recorded in the coastal transect and the two inland transects. Because one tree can shelter several ant species, the total percentages per transect can surpass 100%.* introduced species. doi:10.1371/journal.pone.0009319.g001 trails (polydomous colonies). Colonies of the last two species, interconnected all trees through their foliage or along trails on Figure 1. Relative frequency for the 29 ant species recorded in the coastal transect and the two inland transects. Because one tree can shelter several ant species, the total percentages per transect can surpass 100%.* introduced species. doi:10.1371/journal.pone.0009319.g001 Figure 1. Relative frequency for the 29 ant species recorded in the coastal transect and the two inland transects. Because one tree can shelter several ant species, the total percentages per transect can surpass 100%.* introduced species. doi:10.1371/journal.pone.0009319.g001 trails (polydomous colonies). Colonies of the last two species, Tapinoma subtile and Camponotus sp.1, occupied the entire crown of a tree on occasion, but were most frequently associated with other ant species (co-dominance). trails (polydomous colonies). Discussion Workers from the same area/ colony were not aggressive with each other, while this was not the case for those from different areas. Statistical comparisons (Kruskal–Wallis tests): H5 60 = 49.46; P,0.0001; Dunn’s multiple comparison tests, workers from the same area: P.0.05 in all cases; from two different areas: P,0.05; workers from same area vs. from two different areas (control vs. experimental lots): P,0.01. doi:10.1371/journal.pone.0009319.t001 (I) Interspecific confrontations between Technomyrmex albipes (colonies from areas A–F) and Cr. ranavalonae workers (from areas F, G). Areas A–E correspond to a total of ca. 8.4 km of shoreline, while area F is situated 1.7 km inland. Tests of species co-occurrences revealed no ant mosaic structure since dominant species were generally independently distributed or positively associated. Nevertheless, our survey yielded several arboreal ant species likely to be territorially dominant, in particular the carton-building Cr. ranavalonae. This species has polydomous colonies spread over several trees and shows strong intraspecific and interspecific aggressiveness towards T. albipes. Its habits are reminiscent of the carton-building Crematogaster species that participate in African rainforest ant mosaics [7,8]; the smaller ovoid nests are likely adapted to the harsh climatic conditions of the Masoala Peninsula. Crematogaster rasoherinae and Cr. kelleri workers, which colonize several adjacent trees of different species, are likely to bore galleries into branches like the African Myrmicinae Atopomyrmex mocquerisii [29]. That Crematogaster sp.1 nests in the naturally hollow V. beraviensis branches indicates that this tree species could be a myrmecophyte like several others of the genus Vitex [30]. We also recorded a nocturnal species, Camponotus sp.1, which shared trees with different dominant species (co-dominance). Finally, B. cordemoyi, which - like other species of the genus - has very tiny workers, was associated with T. albipes. To study whether group size influences aggressiveness, we also transported groups of ca. 200 foraging T. albipes workers and installed them among conspecifics in a different zone of the same tree (control) or on a tree situated several hundreds meters away (experiment). The results confirmed those recorded during individual bouts, as the transferred individuals were integrated into nearby worker columns in all cases. The same experiments conducted with Cr. ranavalonae workers resulted in the absence of aggressiveness between workers originating from different parts of the same tree crown, but all of those transferred to distant trees were spread-eagled and then killed by resident ants, illustrating the existence of intercolonial aggressiveness. Intra- and Interspecific Aggressiveness between Technomyrmex albipes and Crematogaster ranavalonae Colonies ranavalonae 2.9 2.7 2.9 3.7 3.0 3.1 II Sites A B C D E F A 1.1 1.1 1.2 1.1 1.0 1.1 B 1.2 1.4 1.2 1.2 1.1 C 1.2 1.3 1.4 1.2 D 1.3 1.1 1.0 E 1.1 1.1 F 1.1 III Sites F G H F 1.1 3.7 3.3 G 1.0 3.2 H 1.0 albipes and B. cordemoyi, but a negative association between Cr. ranavalonae and T. albipes and between Cr. ranavalonae and B. cordemoyi (Table 2). In the coastal transect, we found a significant positive association between B. cordemoyi and Camponotus sp.4. Note that numerically-dominant native species present in both inland transects were absent here. Discussion Both inland transects were characterized by a high proportion of trees devoid of dominant arboreal ants (areas where no arboreal ants were recorded, but discrete species might have been present), something never before reported to the best of our knowledge. Indeed, all other studies conducted so far in the humid tropics have found that most, if not all, canopy trees sheltered ants, regardless of whether or not these ants were dominant. The presence of numerous trees devoid of native dominant arboreal ants cannot be ascribed to an ability to repel arboreal ants, as all of these tree species sheltered ant colonies elsewhere in the transects (see Data set S1). Rather, several non-exclusive factors might be involved, such as canopy structure, the size of the colonies’ territories, and climatic impacts such as particularly strong storms that destroy exposed arboreal nests on this peninsula. The latter Discussion In addition to the lack of intra-specific aggressiveness between workers gathered from places separated from each other by up to 8.4 km, the presence of trails between trees along the shore supports our argument that a T. albipes supercolony exists in Madagascar [see 25] and probably covers an area larger than we were able to explore. Genetic data would be necessary for definitive proof of a supercolony’s existence [13,17,26]. As reported for T. brunneus [23], environmental conditions along the coast may favor the formation of a T. albipes supercolony (see also the 12 out of 14 Malagasy locales cited by Bolton [22]). The present study also suggests that T. albipes can spread inland through patches where resident workers will not fight with those from the shore. The extension of these patches can lead to interconnections and the formation of a huge colony. Note that the populations in supercolonies do not necessarily span a contiguous area [27], so that T. albipes from the inland transects could belong to the same supercolony as those from the coast, and these entities might even be interconnected from time to time. Indeed, T. albipes, described as invasive at least once [28], has been reported inland in Madagascar [22], Borneo and Malaysia [10]. (I) Interspecific confrontations between Technomyrmex albipes (colonies from areas A–F) and Cr. ranavalonae workers (from areas F, G). Areas A–E correspond to a total of ca. 8.4 km of shoreline, while area F is situated 1.7 km inland. Aggressiveness values were recorded only for T. albipes workers and calculated for cases where they initiated the encounters. Statistical comparisons (Kruskal– Wallis tests); T. albipes vs. Cr. ranavalonae: H5 60 = 9.5; P = 0.092. (II) Intraspecific aggressiveness between T. albipes workers; control lot or individuals from the same tree: H5 60 = 3.4; P = 0.80; experimental lot or individuals from two different areas: H14 150 = 13.2; P = 0.51. Comparison taking into account inter- and intraspecific confrontations: H2 150 = 1573.1; P,0.0001; Dunn’s multiple comparison tests, T. albipes workers confronted with Cr. ranavalonae vs. T. albipes workers from the control or the experimental group: P,0.001 in both cases; confrontations between T. albipes workers (control vs. experimental lot): P.0.05. (III) Intraspecific aggressiveness between Cr. ranavalonae workers gathered from three different areas (F, G, H). February 2010 \| Volume 5 \| Issue 2 \| e9319 Intra- and Interspecific Aggressiveness between Technomyrmex albipes and Crematogaster ranavalonae Colonies While searching for evidence of the presence of T. albipes along 4 km of coast, we noticed that columns of workers PLoS ONE \| www.plosone.org February 2010 \| Volume 5 \| Issue 2 \| e9319 February 2010 \| Volume 5 \| Issue 2 \| e9319 3 Gaps in an Ant Mosaic Gaps in an Ant Mosaic Table 1. Values for aggressiveness during one-on-one confrontations between workers. I Sites A B C D E F F, H for Cr. ranavalonae 2.9 2.7 2.9 3.7 3.0 3.1 II Sites A B C D E F A 1.1 1.1 1.2 1.1 1.0 1.1 B 1.2 1.4 1.2 1.2 1.1 C 1.2 1.3 1.4 1.2 D 1.3 1.1 1.0 E 1.1 1.1 F 1.1 III Sites F G H F 1.1 3.7 3.3 G 1.0 3.2 H 1.0 (I) Interspecific confrontations between Technomyrmex albipes (colonies from areas A–F) and Cr. ranavalonae workers (from areas F, G). Areas A–E correspond to a total of ca. 8.4 km of shoreline, while area F is situated 1.7 km inland. Aggressiveness values were recorded only for T. albipes workers and calculated for cases where they initiated the encounters. Statistical comparisons (Kruskal– Wallis tests); T. albipes vs. Cr. ranavalonae: H5 60 = 9.5; P = 0.092. (II) Intraspecific aggressiveness between T. albipes workers; control lot or individuals from the same tree: H5 60 = 3.4; P = 0.80; experimental lot or individuals from two different areas: H14 150 = 13.2; P = 0.51. Comparison taking into account inter- and intraspecific confrontations: H2 150 = 1573.1; P,0.0001; Dunn’s multiple comparison tests, T. albipes workers confronted with Cr. ranavalonae vs. T. albipes workers from the control or the experimental group: P,0.001 in both cases; confrontations between T. albipes workers (control vs. experimental lot): P.0.05. (III) Intraspecific aggressiveness between Cr. ranavalonae workers gathered from three different areas (F, G, H). Workers from the same area/ colony were not aggressive with each other, while this was not the case for those from different areas. Statistical comparisons (Kruskal–Wallis tests): H5 60 = 49.46; P,0.0001; Dunn’s multiple comparison tests, workers from the same area: P.0.05 in all cases; from two different areas: P,0.05; workers from same area vs. from two different areas (control vs. experimental lots): P,0.01. doi:10.1371/journal.pone.0009319.t001 Table 1. Values for aggressiveness during one-on-one confrontations between workers. I Sites A B C D E F F, H for Cr. Looking for an Ant Mosaic Inland doi:10.1371/journal.pone.0009319.t002 Symbols indicate the nature of the association.: +: positive, [2] negative, 0: not significant. X indicates that the test is meaningless since Technomyrmex albipes is present in every sample. doi:10 1371/journal pone 0009319 t002 may explain why we noted only one species, Cr. ranavalonae, with external nests; all other species nest in hollow branches, the cavities formed by rough bark or the root area of the epiphyte Medinilla. Ethics Statement This work was conducted according to relevant national and international guidelines. Thanks to their number and aggressiveness, T. albipes workers might have excluded native dominant arboreal ants from the coast (present inland, they were absent from the coast). An invasive process likely had begun, as numerical dominance often favors invasive species through the rapid recruitment of relatively aggressive nestmates that eliminate native species through exploitative and interference competition [11,19,20]. The fact that numerous inland trees were apparently devoid of ants could favor the explanation that invasive species had penetrated the area (see [11] about unsaturated island ecosystems). Indeed, T. albipes was noted in both inland transect 1 and ca. 2 km inland. The success of invasive ants is associated with the absence of intraspecific aggressiveness due to the formation of supercolonies accompanied by high interspecific aggressiveness [11]. This is true for T. albipes, as (1) workers gathered inland did not fight with those from the shoreline; and (2) all were aggressive toward Cr. ranavalonae individuals. In Southeast Asia, where territorially- dominant species such as Oecophylla smaragdina are present, T. albipes colonies are involved in the formation of an ant mosaic [10]. In this case, although T. albipes successfully penetrated inland, its invasive action seems limited by native, territorially-dominant arboreal ants. PLoS ONE \| www.plosone.org February 2010 \| Volume 5 \| Issue 2 \| e9319 Looking for an Ant Mosaic Inland For the nine species with large colonies, which were also the most frequently encountered, the null model analysis indicated that species co-occurrences were more frequent than might be expected in the inland transects (P,0.05 and P = 0.07, respec- tively) and in the coastal transect (P,0.001) due to chance. More specifically, we found a significant positive association between Cr. madagascariensis and Cr. kelleri in the inland 1 transect. In the inland 2 transect, we found a significant positive association between T. PLoS ONE \| www.plosone.org February 2010 \| Volume 5 \| Issue 2 \| e9319 4 Gaps in an Ant Mosaic Table 2. Associations between the most frequent species (relative frequency .5%) from the three transects sorted by decreasing rank of occurrence and tested using Chi-square tests (1 d.f., Yates’ correction). Relative frequency Species 1 2 3 4 5 Coastal 1 70% Technomyrmex albipes transect 2 15% Brachymyrmex cordemoyi X 3 9% Tetraponera longula X 0 4 6% Camponotus sp.4 X + 0 Inland 1 1 21% Crematogaster ranavalonae transect 2 19% Crematogaster kelleri 0 3 12% Cremato. madagascariensis 0 + 4 6% Crematogaster rasoherinae 0 0 0 5 6% Camponotus sp.1 0 0 0 0 6 5% Crematogaster sp.1 0 0 0 0 0 Inland 2 1 20% Crematogaster ranavalonae transect 2 16% Technomyrmex albipes [2] 3 15% Brachymyrmex cordemoyi [2] + 4 10% Cremato. madagascariensis 0 0 0 5 5% Tapinoma subtile 0 0 0 0 Symbols indicate the nature of the association.: +: positive, [2] negative, 0: not significant. X indicates that the test is meaningless since Technomyrmex albipes is present in every sample. doi:10.1371/journal.pone.0009319.t002 Table 2. Associations between the most frequent species (relative frequency .5%) from the three transects sorted by decreasing rank of occurrence and tested using Chi-square tests (1 d.f., Yates’ correction). Table 2. Associations between the most frequent species (relative frequency .5%) from the three transects sorted by decreasing rank of occurrence and tested using Chi-square tests (1 d.f., Yates’ correction). Table 2. Associations between the most frequent species (relative frequency .5%) from the three transects sorted by decreasing rank of occurrence and tested using Chi-square tests (1 d.f., Yates’ correction). Symbols indicate the nature of the association.: +: positive, [2] negative, 0: not significant. X indicates that the test is meaningless since Technomyrmex albipes is present in every sample. The Transects ranavalonae workers from sites F, G and H. living in each tree would, however, require a much greater effort incompatible with the conditions of a snapshot study. The first, or ‘‘coastal,’’ transect (20 m wide at ground level; 175 m long; altitude 1–3 m; all tree crowns inspected), included 150 trees, 6–20 m tall, and was situated along the shore, beginning on the right bank of the estuary of the Tampolo River. The second, or ‘‘inland 1’’ transect (20 m wide; 175 m long and with an additional area, Fig. S1; altitude 15 m), included 120 trees about 30 m tall and was parallel to the coast 100 m away, beginning 400 m from the right bank of the estuary of the Tampolo River. The third, or ‘‘inland 2’’ transect, included 89 adjacent trees about 30 m tall and was located 2 km inland (10– 15 m wide; ca.100 m long; altitude 35 m). The Transects For this study, conducted between 12 October and 10 November 2001 in a rainforest on the western coast of the Masoala National Park around the estuary of the Tampolo River (15u 439 450S, 49u 579 380E), Madagascar, we surveyed three belt transects. For all of the surveys, we cut off two to four relatively large sections (diameter.10 cm) of branches from each tree. Because arboreal ants mark these branches as part of their territories, they remained on the branches for more than one hour after these sections were removed. Any ants found on or under tree bark or in hollow twigs were collected with aspirators. The trees were typically identified based on the flowers and/or fruit attached to the branches. We used the single rope technique to reach the canopy in the first two transects (see [33]) and the canopy sledge in the third transect. The canopy sledge (‘‘Luge des cimes’’) is an inflatable device carried by a blimp that can transport two persons from treetop to treetop [34]. Branches harvested via the single rope technique were collected between 8:00 and 12:00, and at night (between 21:30 and 23:30) on trees apparently devoid of ants during the day. Branches harvested via the canopy sledge were collected between 5:30–7:30 in low wind conditions. Beachhead invasions by ants have already occurred on the Galapagos, Hawaii, Mauritius, New Caledonia and Christmas Island, and pose a significant conservation concern [11,31,32]. Although we noted the presence of arboreal ant species likely to be territorially dominant, the distribution of their territories was very loose, a factor that could favor the penetration inland of T. albipes. The forest canopy of the Masoala National Park could be under threat from T. albipes, which has already established itself along the coast and could easily spread inland. Our aim was to rapidly assess the distribution of the dominant arboreal ants over a wide area, not to conduct an exhaustive inventory of the arboreal ant assemblage. Preliminary tests showed that clipping two to four large branches from each tree was sufficient to capture opportunistic nesting, diurnal, dominant arboreal ants (the nests made by weaver or carton-building ants are easily detectable). A complete survey of all of the ant colonies PLoS ONE \| www.plosone.org February 2010 \| Volume 5 \| Issue 2 \| e9319 5 Gaps in an Ant Mosaic transported in each case. The same experiment was conducted with Cr. Acknowledgments This study was conducted during the 5th Ope´ration canope´e as part of a project set up by Pro-Natura International and Oce´an Vert sponsored by Solvay S.A. through the SolVin-Bretzel programme. We would also like to thank Francis Halle´ and Olivier Pascal for inviting us to participate in this operation. We are grateful to Nicolas Lochu and Jean-Yves Serein who gathered samples from the canopy using the single rope technique and to Andrea Dejean for proofreading the first version of the manuscript. For comparison, the behavior of T. albipes from each site was recorded during confrontations with Cr. ranavalonae, the most frequent species collected in the inland transects. We also set up intraspecific confrontations between Cr. ranavalonae workers belong- ing to three colonies (sites F, G, H) separated by more than 1 km. Levels of aggressiveness between sample pairs were compared using the Kruskal-Wallis test (GraphPad Prism 4.0 Software). Intra- and Interspecific Aggressiveness between Technomyrmex albipes and Crematogaster ranavalonae Colonies Specific associations between the most common species were tested using Chi-square tests (Yates’ correction). During sampling, we noted when ant colonies were gathered from the same tree crown, when host trees were adjacent, and whether an ant species was represented by a small or a large colony. When the colonies of two dominant species occupied different areas of the same tree crown, two different territories were distinguished (i.e., some of the large branches sampled were occupied by one dominant species, and others by a second dominant species), and the species were not considered co-occurring for the species association analysis [6,7]. After surveying the coastal transect, we checked for the presence of a T. albipes supercolony along 4 km of coast. We observed whether or not the columns of T. albipes workers went from one tree to another (ca. 3400 trees). In addition, we employed the standard behavioral assay created by Suarez et al. [26, see also 16, 25] to test the aggressiveness between T. albipes individuals collected from different sites. We paired workers originating from six sites (one tree per site) or from the same tree at each site (control). Five of these sites (sites A–E) were situated along 8.4 km of coast. Three sites (A–C) were located to the north of the Tampolo River estuary between Andronabe´ and ca. 0.7 km to the south of Camp Tampolo; two others (D and E) were situated to the south of the estuary (see map in Radeau des cimes 2000 [35]). The distance between sites A and B was ca. 6.3 km and the other sites were spaced at intervals of 0.7 km. The sixth site (F) was situated 1.7 km inland across from the fifth site. Supporting Information Figure S1 Schematic representation of the distribution of ant species recorded on trees from the inland 1 transect. Found at: doi:10.1371/journal.pone.0009319.s001 (0.07 MB DOC) Figure S1 Schematic representation of the distribution of ant species recorded on trees from the inland 1 transect. Found at: doi:10.1371/journal.pone.0009319.s001 (0.07 MB DOC) Data set S1 Series and tree species monitored in the inland 1 transect (A1-A120) and the inland 2 transect (L1-L89), and ant species recorded nesting in tree crowns. // : cases when ant species were recorded on different branches of the same tree (two different territories on the same tree). For the inland transects we provide the trees’ code in the first column (see also Figure S1 for inland transect 1); for the coastal transect we provide only the number of trees. For the tests, two individual workers were placed in a neutral arena (Ø: 6 cm; height: 7 cm) whose walls were coated with fluonH to prevent the ants from climbing out. We scored interactions between the workers over a 5 minute period on a scale from 1 to 4: 1 = physical contact, but no aggressive response (may include antennation or trophallaxis), 2 = avoidance (the ants touch, and one or both recoils and runs in the opposite direction), 3 = aggressiveness (biting legs or antennae), and 4 = fighting (prolonged biting, pulling, or gaster bending by one or both ants). We repeated the confrontations 10 times, retaining the highest value noted each time, and used each worker only once. p y Found at: doi:10.1371/journal.pone.0009319.s002 (0.24 MB DOC) PLoS ONE \| www.plosone.org Comparing the Two Inland Transects and Testing the Existence of an Ant Mosaic The similarity between samples was calculated using the Chao- Jaccard abundance-based similarity index for trees, and the Chao- Jaccard incidence-based similarity index for ants. These indices are appropriate for the comparison of incompletely sampled species-rich communities [37]. Standard errors for the Chao- Jaccard estimators were computed through 200 bootstrap procedures using EstimateS 7.5 software. Global trends in species associations were investigated using a fixed-equiprobable null model and the C-score co-occurrence index available in the EcoSim software [38]. The fixed- equiprobable algorithm, appropriate for data-matrices with unoccupied sites, maintains the species occurrence frequencies and considers all sites (trees) equiprobable [39]. Tests not shown here confirmed that the outcome of the null model analysis was not altered by including trees not occupied by ants. The C-score index used in combination with the fixed-equiprobable algorithm has generally good statistical properties and is not prone to false positives [39]. All plants were morphotyped at least to family save for five dead or unrecognizable trees in the inland 1 transect. Ants were preserved in 70% ethanol for later identification to species or morphospecies. Voucher specimens were deposited at the Bibikely Biodiversity Center, Parc Botanique et Zoologique de Tsimbazaza, BP 4096, Antananarivo, Madagascar, and the Department of Entomology, California Academy of Sciences, San Francisco, USA. Intra- and Interspecific Aggressiveness between Technomyrmex albipes and Crematogaster ranavalonae Colonies References 21. Sanders NJ, Gotelli NJ, Heller NE, Gordon DM (2003) Community disassembly by an invasive species. Proc Natl Acad Sci USA 100: 2474–2477. 1. Davidson DW, Cook SC, Snelling RR, Chua TH (2003) Explaining the abundance of ants in lowland tropical rainforest canopies. Science 300: 969–972. p p 2. Hunt JH (2003) Cryptic herbivores of the rainforest canopy. Science 300: 916. 22. Bolton B (2007) Taxonomy of the dolichoderine ant genus Technomyrmex Mayr (Hymenoptera: Formicidae) based on the worker cast. Contr Amer Entomol Inst 35: 1–150. 3. Blu¨thgen N, Gebauer G, Fiedler K (2003) Disentangling a rainforest food web using stable isotopes: dietary diversity in a species-rich ant community. Oecologia 137: 426–435. 23. Terayama M (1989) Ant community of the mangrove swamps of Japan. Bull Toho Gakuen 3: 1–11. g 4. Leston D (1973) The ant mosaic, tropical tree crops and the limiting of pests and diseases. Pest Articles and News Summaries 19: 311–341. 24. Tsuji K, Yamauchi K (1994) Colony level sex allocation in a polygynous and polydomous ant. Behav Ecol Sociobiol 34: 157–167. 5. Majer JD (1993) Comparison of the arboreal ant mosaic in Ghana, Brazil, Papua New Guinea and Australia: its structure and influence of ant diversity. In: LaSalle J, Gauld ID, eds. Hymenoptera and Biodiversity. Wallingford: CAB International. pp 115–141. 25. Fadamiro HY, Xiaofa H, Chen L (2009) Aggression in imported fire ants: an explanation for shifts in their spatial distributions in Southern United States? Ecol Entomol 34: 427–436. pp 6. Blu¨thgen N, Stork NE (2007) Ant mosaics in a tropical rainforest in Australia and elsewhere: A critical review. Aust Ecol 32: 93–104. 26. Suarez AV, Tsutsui ND, Holway DA, Case TJ (1999) Behavioral and genetic differentiation between native and introduced populations of the Argentine ant. Biol Invas 1: 43–53. 7. Dejean A, Corbara B, Orivel J, Leponce M (2007) Rainforest canopy ants: the implications of territoriality and predatory behavior. Funct Ecosyst Commun 1: 105–120. 27. Giraud T, Pedersen JS, Keller L (2002) Evolution of supercolonies: the Argentine ants of southern Europe. Proc Natl Acad Sci, USA 99: 6075–6079. 8. Dejean A, Djie´to-Lordon C, Ce´re´ghino R, Leponce M (2008) Ontogenetic succession and the ant mosaic: an empirical approach using pioneer trees. Basic Appl Ecol 9: 316–323. 28. Sulaiman SFM (1997) Impact of weed management on ant density and fruit yield in the control of pineapple wilt disease. Acta Horticult 425: 475–484. 29. References Kenne M, Djie´to-Lordon C, Orivel J, Mony R, Fabre A, et al. (2003) Influence of insecticide treatments on ant-hemiptera associations in tropical plantations. J Econ Entomol 96: 251–258. pp 9. Rico-Gray V, Oliveira P (2007) The ecology and evolution of ant-plant interactions. Chicago: The University of Chicago press. 10. Pfeiffer M, Tuck HC, Lay TC (2008) Exploring arboreal ant community composition and co-occurrence patterns in plantations of oil palm (Elaeis guineensis) in Borneo and Peninsular Malaysia. Ecography 31: 21–32. 30. Djie´to-Lordon C, Dejean A, Ring RA, Lauga J, Nkongmeneck BA, et al. (2005) Ecology of an improbable association: the pseudomyrmecine plant-ant Tetraponera tessmanni and the myrmecophytic vine Vitex thyrsiflora (Lamiaceae). Biotropica 37: 420–429. g ) y g p y 11. Holway D, Lach L, Suarez AV, Tsutui ND, Case TJ (2002) The causes and consequences of ant invasions. Ann Rev Ecol Syst 33: 181–233. consequences of ant invasions. Ann Rev Ecol Syst 33: 181–23 31. Le Breton J, Jourdan H, Chazeau J, Orivel J, Dejean A (2005) Niche opportunity and ant invasion: the case of Wasmannia auropunctata in a New Caledonian rain forest. J Trop Ecol 21: 93–98. 12. Boosma JJ (2003) Social mutualism and social parasitism: conflict and cooperation at the family and species level. In: Kikuchi T, Azuma N, Higashi S, eds. Genes, Behavior and Evolution in Social Insects. Hokkaido PP: Hokkaido university Press. pp 131–170. 32. Abbott KL (2006) Spatial dynamics of supercolonies of the invasive yellow crazy ant, Anoplolepis gracilipes, on Christmas Island, Indian Ocean. Divers Distrib 12: 101–110. y pp 13. Helantera H, Strassmann JE, Carrillo J, Queller DC (2009) Unicolonial ants: where do they come from, what are they and where are they going? Trends Ecol Evol 24: 9341–349. 33. Steege H (1998) Single rope techniques in tropical rain forest trees: going down safe and sound. Biotropica 30: 496–497. 14. Abbott KL, Greaves SNJ, Ritchie PA, Lester PJ (2007) Behaviourally and genetically distinct populations of an invasive ant provide insight into invasion history and impacts on a tropical ant community. Biol Invas 9: 453–463. 34. Basset Y, Novotny V, Miller SE, Kitching RL (2003) Arthropods of tropical forests. Spatio-temporal dynamics and resource use in the canopy. Cambridge: Cambridge University Press. pp 474. 15. Foucaud J, Fournier D, Orivel J, Delabie JHC, Loiseau A, et al. (2007) Sex and clonality in the little fire ant. Mol Biol Evol 24: 2465–2473. 35. Author Contributions Because there was a demonstrated effect of group size on aggressiveness [25,36], in a complementary experiment, we transported foraging workers from two zones of the same tree from sites C, E and F (control), and between trees situated on the three sites; 10 groups of approximately 200 workers were Conceived and designed the experiments: AD BLF BC. Performed the experiments: AD BLF BC. Analyzed the data: AD ML. Contributed reagents/materials/analysis tools: AD BLF BC BR ML. Wrote the paper: AD BC ML. Identified the ants: BLF BR. Identified the trees: RR RR. February 2010 \| Volume 5 \| Issue 2 \| e9319 6 Gaps in an Ant Mosaic Gaps in an Ant Mosaic References Radeau des cimes (2000) Available: http://www.radeau-des-cimes.org/radeau/ expe%20mada_fr.pdf. clonality in the little fire ant. Mol Biol Evol 24: 2465–2473. 16. Tsutsui ND, Suarez AV, Holway DA, Case TJ (2000) Reduced genetic variation and the success of an invasive species. Proc Natl Acad Sci, USA 97: 5948–5953. 36. Tanner CJ (2006) Numerical assessment affects aggression and competitive ability: a team-fighting strategy for the ant Formica xerophila. Proc Roy Soc London B 273: 2737–2742. 17. Tsutsui ND, Suarez AV, Grosberg RK (2003) Genetic diversity, asymmetrical aggression, and cooperation in a widespread invasive species. Proc Natl Acad Sci, USA 100: 1078–1083. 37. Chao A, Chazdon RL, Colwell RK, Shen T-J (2005) A new statistical approach for assessing compositional similarity based on incidence and abundance data. Ecol Let 8: 148–159. 18. Pedersen JS, Krieger MJB, Vogel V, Giraud T, Keller L (2006) Native supercolonies of unrelated individuals in the invasive argentine ant. Evolution 60: 782–791. 38. Gotelli NJ, Entsminger GL (2000) EcoSim: null models software for ecology. Version 7.20. Acquired Intelligence Inc. & Kesey-Bear. Available: http:// homepages.together.net/,gentsmin/ecosim.htm. 19. Davidson DW (1998) Resource discovery versus resource domination in ants: a functional mechanism for breaking the trade-off. Ecol Entomol 23: 484–490. 39. Gotelli NJ (2000) Null model analysis of species co-occurrence patterns. Ecology 81: 2606–2621. 20. Holway DA, Suarez AV (2004) Colony-structure variation and interspecific competitive ability in the invasive Argentine ant. Oecologia 138: 216–222. PLoS ONE \| www.plosone.org February 2010 \| Volume 5 \| Issue 2 \| e9319 7
https://openalex.org/W4206774330	http://old.scielo.br/pdf/rbpv/v30n4/1984-2961-rbpv-30-4-e017721.pdf	English	null	Detection of Trypanosoma vivax in tissues of experimentally infected goats: what is the role of adipose tissue in the life cycle of this protozoon?	Revista Brasileira de Parasitologia Veterinária/Brazilian Journal of Veterinary Parasitology	2,021	cc-by	5,987	1/10 Braz J Vet Parasitol 2021; 30(4): e017721 \| https://doi.org/10.1590/S1984-29612021092 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. Received October 04, 2021. Accepted November 16, 2021. Corresponding author: Rosangela Zacarias Machado. E-mail: rzacariasmachado@gmail.com Detecção de Trypanosoma vivax em tecidos de caprinos experimentalmente infectados: qual é o papel do tecido adiposo no ciclo de vida desse protozoário? Rosangela Zacarias Machado1 ; Kayo José Garcia de Almeida Castilho Neto1; Luiz Ricardo Gonçalves1; Gisele Fabrino Machado2; Maria Cecília Rui Luvizotto2; Rosemeri de Oliveira Vasconcelos1; Giulia Jussiani2; Thiago Merighi Vieira da Silva1; Lindsay Unno Gimenes1; Andrew Jackson3; Gavin Wright4; i ld i i 5 1Departamento de Patologia, Reprodução e Saúde Única, Faculdade de Ciências Agrárias e Veterinárias – FCAV, Universidade Estadual Paulista – UNESP, Jaboticabal, SP, Brasil J 2Departamento de Clínica, Cirurgia e Reprodução Animal, Faculdade de Medicina Veterinária – FMVA, Universidade Est UNESP, Araçatuba, SP, Brasil ç 3Department of Infection Biology, University of Liverpool, Liverpool, United Kingdom 4Department of Biology Hull York Medical School University of York Wentworth Way York United Kingdom 3Department of Infection Biology, University of Liverpool, Liverpool, United Kingdom 4 f i l ll k di l h l i i f k h k i d i d 4Department of Biology, Hull York Medical School, University of York, Wentworth Way, York, United Kingdom 5Departamento de Parasitologia, Instituto de Ciências Biomédicas – ICB, Universidade de São Paulo – USP, São Paulo, SP, Bra How to cite: Machado RZ, Castilho Neto KJGA, Gonçalves LR, Machado GF, Luvizotto MCR, Vasconcelos RO et al. Detection of Trypanosoma vivax in tissues of experimentally infected goats: what is the role of adipose tissue in the life cycle of this protozoon? Braz J Vet Parasitol 2021; 30(4): e017721. https://doi.org/10.1590/S1984-29612021092 Abstract Trypanosomiasis, caused by Trypanosoma vivax, is responsible for great economic losses among livestock in Africa and South America. During the life cycle of these parasites, they may present different morphological, metabolic and physiological characteristics depending on the interactions that are encountered at each point of their life cycle. Although T. vivax is frequently reported in the circulation of its mammalian hosts, it has the ability to migrate to the tissues of these individuals. However, this characteristic is poorly understood. In this context, we aimed to investigate the presence of T. vivax and the changes caused in different tissues of experimentally infected goats. Despite the animals were not perfused before tissues collection, using different approaches, we demonstrated its presence in different samples, including in the adipose tissue and skin of infected animals. In addition, a mononuclear inflammatory reaction, mostly characterized by an infiltrate of lymphocytes, plasma cells and macrophages were observed. The results highlight the possibility that, like other trypanosomatids, T. vivax may use these tissues during its life cycle. Future studies aiming to elucidate the length of time for which T. vivax remains active in these sites, and whether it uses these sites as a refuge from trypanocidal drugs, and whether it is capable of recolonizing the blood circulation, are much needed. Keywords: LAMP, livestock, refuge, trypanosomiasis, Trypanosoma vivax. ISSN 1984-2961 (Electronic) www.cbpv.org.br/rbpv Original Article Introduction Trypanosomiasis is a cosmopolitan disease that affects humans, domestic animals and wild animals. Trypanosoma vivax, T. evansi, T. equiperdum, T. cruzi and T. theileri are found in animals in South America (Hoare, 1972; Gardiner et al., 1989), and T. vivax is responsible for major economic losses in cattle-rearing in Africa, Asia, Central America and South America (Dávila & Silva, 2000; Osório et al., 2008). In the African continent, in areas where Glossina spp. (tsetse) is present, T. vivax is transmitted cyclically, developing in the digestive tract of the fly. In the Americas, where this fly is not present, transmission is performed mechanically by tabanids (horseflies), Stomoxys calcitrans (stable fly) and Haematobia irritans (Serra-Freire & Rezende, 1988; Cadioli et al., 2012), or iatrogenically by means of fomites (Silva et al., 1996; Cadioli et al., 2012). Transplacental infections have also been described (Silva et al., 2013). In Brazil, T. vivax infections in ruminant herds have been occurring with increasing frequency. Over the last two decades, outbreaks have been observed in different Brazilian states (Bastos et al., 2020), causing a major economic impact on Brazilian cattle-rearing (Seidl et al., 1999; Jones & Dávila, 2001). The clinical signs triggered by the parasite are not very specific, which makes diagnosis difficult. During the course of trypanosomiasis, there are fluctuations in parasitemia or even intervals of undetectable parasitaemia (Almeida et al., 2012; Fidelis et al., 2016). These fluctuations are related to the host’s immune response and the antigenic variation of surface variant glycoproteins of trypanosomes (Nantulya, 1990; Stijlemans et al., 2010). In low- parasitemia phases, detection of T. vivax becomes even more challenging (Rebeski et al., 1999; Waal, 2012). Thus, diagnoses based on immunological response (ELISA), molecular tests (PCR and LAMP) and histopathological analyses, contribute greatly to higher-precision diagnosis, thus avoiding maintenance of false-negative animals in the herd and contributing to greater disease control. In animals that are experimentally and/or naturally infected by T. vivax, different anatomopathological findings and histological alterations have been reported in very distinct tissue types (Batista et al., 2006, 2007; Almeida et al., 2010). So far, no studies have demonstrated any presence of T. vivax in the adipose tissue or skin of its vertebrate hosts. However, recent research on T. brucei (which causes human African trypanosomiasis, i.e. sleeping sickness) showed, using a mouse model, that adipose tissue constituted a third major reservoir for this parasite (Trindade et al., 2016). Resumo A tripanossomíase, causada por Trypanosoma vivax, é responsável por grandes perdas econômicas na bovinocultura da África e da América do Sul. Durante seu ciclo de vida, o parasita pode apresentar diferentes características morfológicas, metabólicas e fisiológicas em função das interações que ele encontra em cada ponto do seu ciclo. Embora o T. vivax seja reportado, frequentemente, na circulação dos seus hospedeiros mamíferos, o protozoário tem a capacidade de migrar para os tecidos desses indivíduos. Entretanto, essa característica é pobremente 1/10 Goats experimentally infected with T. vivax conhecida. Neste contexto, o objetivo foi verificar a presença, assim como as alterações causadas pelo T. vivax nos diferentes tecidos de caprinos experimentalmente infectados. Apesar dos animais não terem sido perfundidos antes da coleta dos tecidos, utilizando-se diferentes abordagens, foi evidenciada a presença do T. vivax em diferentes amostras teciduais, incluindo no tecido adiposo e pele dos animais infectados. Além disso, foi observada reação inflamatória mononuclear, caracterizada majoritariamente por infiltrado de linfócitos, plasmócitos e macrófagos. Os resultados evidenciam a possibilidade de que, assim como outros tripanossomatídeos, T. vivax pode usar esses tecidos durante o seu ciclo de vida. São necessários futuros estudos, objetivando elucidar o período em que o T. vivax permanece ativo nesses sítios, se ele utiliza esses locais como refúgio das drogas tripanocidas, e se ele é capaz de recolonizar a circulação sanguínea. Palavras-chave: LAMP, pecuária, refúgio, tripanossomíase, Trypanosoma vivax. Braz J Vet Parasitol 2021; 30(4): e017721 Histopathology Organ/tissue fragments were collected for light microscopy analysis. These were fixed in 10% formalin solution and buffered with phosphates (0.15 molar; pH 7.2). After 24 hours of fixing, the samples were removed from the formalin solution and placed in 70% alcohol. They were then dehydrated in solutions of decreasing alcohol concentration, diaphanized in xylol and embedded in paraffin, in accordance with the routine histological technique. Sections of thickness 5 μm were cut and stained with hematoxylin and eosin (HE), and these were then used to identify the main morphological alterations. Unstained sections were used for the Immunofluorescence technique. Introduction Moreover, Capewell et al. (2016) showed conclusive evidence from a mouse model that T. b. brucei (animal trypanosomiasis) could invade the extracellular tissue of the skin, including but not restricted to the adipose tissue. The same authors also showed the presence of trypanosomes within the skin of undiagnosed humans (Capewell et al., 2016). Although T. vivax (like T. congolense) is considered to be a species confined to the host’s vascular system, some strains in late infections may reach extravascular sites (e.g. lymph nodes, eyes and cerebrospinal fluid). This can cause tissue lesions that are less accessible for drug treatment (Whitelaw et al., 1988; Osório et al., 2008; D’Archivio et al., 2013). Furthermore, it is extremely important to have knowledge of how trypanocidal drugs used for therapeutic and preventive purposes remain in the circulation (Giordani et al., 2016) and whether or not they enter different tissue types. These characteristics may determine whether or not trypanocidal drugs are effective for controlling trypanosomiasis. Thus, detecting positive animals in a herd, and finding out whether, at some point during the course of infection, T. vivax lodges in tissues, is vital for controlling this disease. 2/10 Braz J Vet Parasitol 2021; 30(4): e017721 Goats experimentally infected with T. vivax The aim of the present study was to investigate the presence of T. vivax and evaluate the histopathological alterations in different tissues of experimentally infected goats. The aim of the present study was to investigate the presence of T. vivax and evaluate the histopathological alterations in different tissues of experimentally infected goats. Animals, Experimental Period and Collection of Material Six male goats (Saanen breed) aged four to six months were used. Thirty days before the experimental period, all the animals were wormed (albendazole 7.5 mg/kg, orally). Also, before starting the experiment, the following tests were used to ensure that the animals were not naturally infected: blood smears to investigate hemoparasites; PCR and LAMP to detect T. vivax DNA (Cortez et al., 2009; Njiru et al., 2011); and IFAT (Sampaio et al., 2015) and ELISA (Machado et al., 1997) to detect anti-T. vivax IgG. Initially, the Miranda isolate of T. vivax cryopreserved in liquid nitrogen was thawed and used to infect a goat. The goat was monitored and during the parasitemia peak, the blood sample was collected in tubes containing EDTA. Thereafter, the number of trypomastigote forms was assessed (Brener, 1961) and the animals inoculated. Five animals, after a period of acclimatization (30 days), were inoculated with 1.0 X 103 trypomastigote forms of T. vivax (Miranda isolate) intravenously, through the external jugular vein. Similarly, the control animal was inoculated with physiological saline solution. Before (in the acclimatization period) and during the experimental period, the goats were subjected to daily physical examinations, with evaluations of heart and respiratory rates, mucosal staining, skin turgor, lymph node changes and body condition. In addition, blood samples were obtained for laboratory tests (Brener, 1961) in order to ascertain the presence of the parasite. The animals were euthanized 40 days after infection (DAI), using a combination of xylazine (0.2 mg/kg) and ketamine (2 mg/kg), followed by intrathecal injection of lidocaine (40 ml). During the necropsies on the animals, fragments of the following organs/tissues were collected: testicle, skin, liver, spleen, lymph nodes, peritesticular adipose tissue and perirenal adipose tissue. This research project was approved by the Ethics Committee for Animal Use (CEUA) of the School of Agrarian and Veterinary Sciences, UNESP, Jaboticabal campus, under protocol number 001494/18. The animals were euthanized in accordance with CFMV resolution no. 1000 of May 11, 2012, and this procedure was also approved by the CEUA of FCAV/UNESP. Indirect Immunofluorescence Reaction on Histological Sections The sections of the fragments from the spleen, liver, lymph node, testicle, skin and fat were subjected to the immunofluorescence test. Initially, antigen recovery was performed in a steaming pan (at 100° C), in Envision Target Retrieval Solution High pH buffer (Dako, K8004) for 20 minutes. Nonspecific blockade was implemented using a protein block (Abcam, ab64226) for 30 minutes. Immune serum samples obtained from cattle (40th DAI) that had been experimentally infected by T. vivax were used as the primary antibody. Serum from uninfected cattle was used as the negative control. The serum samples were diluted in PBS in the proportions of 1:40 and were applied to the sections and incubated at 37° C for 30 minutes. Subsequently, the sections were incubated with bovine anti-IgG conjugated with fluorescein isocyanate (Sigma, St. Louis, USA) diluted in PBS and Evans Blue, at 37 ºC for 30 minutes. Lastly, coverslips were mounted on the slides, with addition of buffered glycerin, for viewing under a fluorescence microscope (Olympus, BX-FLA). 3/10 Braz J Vet Parasitol 2021; 30(4): e017721 Goats experimentally infected with T. vivax DNA Extraction and Quality DNA was extracted from 25 mg of each goat tissue sample (except for the spleen tissue, for which only 10 mg was used), using the DNeasy® Blood & Tissue kit (Qiagen®, Valencia, California, USA), following the manufacturer’s recommendations. To avoid false-negative results caused by the presence of inhibitors, and to check for amplifiable DNA, the DNA samples were subjected to a conventional polymerase chain reaction (cPCR) to amplify the endogenous glyceraldehyde-3-phosphate dehydrogenase gene (gapdh) of mammals, following the established protocol (Birkenheuer et al., 2003). Internal control-PCR positive samples were subjected to T. vivax screening by means of loop-mediated isothermal amplification (LAMP) targeting the satellite repeat DNA. The LAMP assay was carried out as described elsewhere (Njiru et al., 2011) using the following set of primers: outer primers: F3 (TGTTCTGGTGGCCTGTTGC) and B3 (GGCCGGAGCGAGAGGTGC); inner primers: FIP (GTGGAGCGTGCCAACGTGGCACCCGCTCCCAGACCATA) and BIP (TGTCTAGCGTGACGCGATGGAAGAGGGAGTGGGGAAGG); and loop primers: LF (CACATGGAGCATCAGGAC) and LB (CCGTGCACTGTCCCGCAC). The LAMP reactions were performed in a reaction volume of 25 µL, consisting of the following: 5 pmol of the outer primers, 20 pmol of the loop primers, 40 pmol of the inner primers, 4 mM of extra MgSO4, 1 M betaine (Sigma- Aldrich, St. Louis, MO, USA) and 2.5 mM of deoxynucleotide triphosphates mix (dNTP). The 1X ThermoPol reaction buffer (New England BioLabs, MA, USA) was used, containing 20 mM of Tris-HCl (pH 8.8), 10 mM of KCl, 10 mM of (NH4)2SO4, 2 mM of MgSO4 and 0.1% Triton X-100. The Bst DNA polymerase volume (large fragment; New England Biolabs) was 1 µL (8 units). SYTO-9 fluorescence dye at 1.5 M (Molecular Probes, OR, USA) and 2.5 µL of each DNA sample were used as a template for each LAMP reaction (Njiru et al., 2011). The reactions were conducted at 63 ºC for 60 minutes, using a QuantStudio 3 Thermal Cycler (Applied Biosystems). The reaction was stopped by increasing the temperature to 80 ºC for 5 mins. Melt curves were acquired using 0.5 ºC steps, with holds of 5 s, from 63 to 96 ºC. The results were assessed through observation of amplification curves using the QuantStudio 3 software. All the Cq (cycle thresholds) of each sample were annotated. Each LAMP assay was performed including duplicates of each goat tissue DNA sample. DNA of T. vivax (Cadioli et al., 2012) and ultrapure water were used as positive and non-template controls, respectively, in all LAMP assays. Molecular Detection of T. vivax The LAMP assay was carried out as described elsewhere (Njiru et al., 2011) using the following set of primers: outer primers: F3 (TGTTCTGGTGGCCTGTTGC) and B3 (GGCCGGAGCGAGAGGTGC); inner primers: FIP (GTGGAGCGTGCCAACGTGGCACCCGCTCCCAGACCATA) and BIP (TGTCTAGCGTGACGCGATGGAAGAGGGAGTGGGGAAGG); and loop primers: LF (CACATGGAGCATCAGGAC) and LB (CCGTGCACTGTCCCGCAC). Results In general, throughout the experimental period, none of the animals showed any severe changes in physical examinations. However, some animals had apathy, pale mucous membranes and enlarged lymph nodes, but without pain on palpation. Parasites were detected using the technique of Brener (1961) at six days after inoculation in all animals (Figure 1). Fluctuations in T. vivax parasitemia were observed during the trial period, with the highest peak observed at eight DAI for the animal C10 with 1.1 X 107 parasites/mL of blood (Figure 1). Braz J Vet Parasitol 2021; 30(4): e017721 4/10 Figure 1. Mean T. vivax parasitemia recorded during the trial period. C9–C13: Infected goats. C-: Negative control. DAI: Days after inoculation. The animals were inoculated with 1.0 X 103 trypomastigote forms of T. vivax (Miranda isolate) intravenously. The parasite quantification was carried out using the thick-drop counting method as previously described (Brener, 1961). Figure 1. Mean T. vivax parasitemia recorded during the trial period. C9–C13: Infected goats. C-: Negative control. DAI: Days after inoculation. The animals were inoculated with 1.0 X 103 trypomastigote forms of T. vivax (Miranda isolate) intravenously. The parasite quantification was carried out using the thick-drop counting method as previously described (Brener, 1961). Braz J Vet Parasitol 2021; 30(4): e017721 4/10 Goats experimentally infected with T. vivax Histologically, a mononuclear inflammatory reaction characterized mainly by infiltration of lymphocytes, plasmocytes and perivascular and/or interstitial macrophages was observed in all the experimentally infected goats (Figure 2). Additionally, a hyperplastic reaction of the lymphoid tissues was observed, particularly in the spleen and peripheral lymph nodes, thus demonstrating a systemic reaction to persistent antigenic stimulation (Figure 2). In general, the endothelial cells of the tissues analyzed showed exacerbated reactivity, especially in the endothelium of venules and lymphatics system. pathological study of goat infected with T. vivax. Skin – Perivascular mononuclear infiltrate (arrow filtrate (inset) in the superficial dermis (A). Skin – Perivascular mononuclear infiltrate () an ves (arrow) in the deep dermis (B). Lymph node – cortical and paracortical follicular () hyper splenic lymphoid follicle (D). Liver – Discrete multifocal perivascular mononuclear infiltrate s (arrow) and vacuolization of hepatocytes (F). Hypodermis Perivascular mononuclear infiltrat mononuclear infiltrate in the interstitium of the seminiferous tubules (H). The tissue sample aining. Figure 2. Histopathological study of goat infected with T. vivax. Skin – Perivascular mononuclear infiltrate (arrow) and eosinophilic perivascular infiltrate (inset) in the superficial dermis (A). Results Skin – Perivascular mononuclear infiltrate () and infiltrate around peripheral nerves (arrow) in the deep dermis (B). Lymph node – cortical and paracortical follicular () hyperplasia (C). Spleen – hyperplasia of splenic lymphoid follicle (D). Liver – Discrete multifocal perivascular mononuclear infiltrate (E). Hemosiderosis in Kupffer cells (arrow) and vacuolization of hepatocytes (F). Hypodermis Perivascular mononuclear infiltrate in adipose tissue (G). Testicle – mononuclear infiltrate in the interstitium of the seminiferous tubules (H). The tissue samples were collected in 40th DAI. HE staining. Figure 2. Histopathological study of goat infected with T. vivax. Skin – Perivascular mononuclear infiltrate (arrow) and eosinophilic perivascular infiltrate (inset) in the superficial dermis (A). Skin – Perivascular mononuclear infiltrate () and infiltrate around peripheral nerves (arrow) in the deep dermis (B). Lymph node – cortical and paracortical follicular () hyperplasia (C). Spleen – hyperplasia of splenic lymphoid follicle (D). Liver – Discrete multifocal perivascular mononuclear infiltrate (E). Hemosiderosis in Kupffer cells (arrow) and vacuolization of hepatocytes (F). Hypodermis Perivascular mononuclear infiltrate in adipose tissue (G). Testicle – mononuclear infiltrate in the interstitium of the seminiferous tubules (H). The tissue samples were collected in 40th DAI. HE staining. 5/10 Braz J Vet Parasitol 2021; 30(4): e017721 Goats experimentally infected with T. vivax Remarkably, all the samples obtained from the infected animals were positive for tissue-based IFA (Figure 3). None of the tissue samples from the negative control goat were positive for tissue-based IFA at 1:40 dilution (Figure 3). Remarkably, all the samples obtained from the infected animals were positive for tissue-based IFA (Figure 3). None of the tissue samples from the negative control goat were positive for tissue-based IFA at 1:40 dilution (Figure 3). ably, all the samples obtained from the infected animals were positive for tissue based IFA (Fig e samples from the negative control goat were positive for tissue-based IFA at 1:40 dilutio sue-based immunofluorescent assay for detection of T. vivax. Spleen – Positive (A) and negative (B) sam tive (C) and negative (D) samples. Skin – Positive (E) and negative (F) samples. Testicle – Positive (G) Serum from T. vivax experimentally infected cattle diluted in PBS (1:40) was used as the primary a FITC conjugate) was used as the secondary antibody. Fluorescence (green) reveals the binding betwee ntibody and the antigen-antibody complex. Both primary and secondary antibodies were incubated a 30 minutes, respectively. Figure 3. Tissue-based immunofluorescent assay for detection of T. vivax. Discussion In the present study, we report histopathological changes and the presence of T. vivax in different tissues, including in the adipose tissue of experimentally infected goats. To date, this is the first report of the occurrence of T. vivax in adipose tissue. Our results, together with the previously reported findings (Trindade et al., 2016), suggest the possibility that this site can be used by this protozoon during its life cycle. Interestingly, although parasitemia was reported in all the experimentally infected goats, the clinical signs presented by the animals were mild. This finding was different from what had previously been reported (Batista et al., 2011). In that study, it was reported that goats infected with T. vivax (1.25 × 105 trypomastigotes) isolated from a cow during an outbreak in the state of Paraíba, northeastern Brazil) presented marked mucosal pallor, apathy, fever and neurological disorders. These findings may have reflected the differences in virulence of the isolates used (Osório et al., 2008). In addition, the number of parasites used in the inoculums may have influenced the disparity of the results observed. Generally, T. vivax is found in the circulatory system. However, during the course of infection, it has the ability to migrate to the tissues of its vertebrate hosts (Gardiner et al., 1989). In these sites, it plays an important role in the pathogenesis of inflammatory and degenerative lesions (Batista et al., 2011). The histopathological lesions seen in the present study were also observed in a previous study on experimentally T. vivax-infected goats (Batista et al., 2011). In addition, those authors described severe lesions in the central nervous system of the animals, including meningitis and meningoencephalitis. Through using two different approaches (LAMP and IFA), we reported the presence of T. vivax in the adipose tissue of infected animals. Recently, Trindade et al. (2016) reported the presence of T. brucei in the adipose tissue of mice, which they suggested would be an important niche for the parasite. Furthermore, those authors demonstrated that this protozoon was metabolically active and that, from the fat, heart and brain, it was able to reestablish infection through colonizing the circulation (Trindade et al., 2016). Lastly, those same authors revealed metabolic differences between T. brucei obtained from the circulation (BSF) and those from adipose tissue (ATF). The ATF transcriptome included upregulated genes, including putative fatty acid β-oxidation enzymes. Furthermore, they reported that ATF could use fatty acids, i.e. Results Spleen – Positive (A) and negative (B) samples. Adipose tissue – Positive (C) and negative (D) samples. Skin – Positive (E) and negative (F) samples. Testicle – Positive (G) and negative (H) samples. Serum from T. vivax experimentally infected cattle diluted in PBS (1:40) was used as the primary antibody. Anti- bovine IgG (FITC conjugate) was used as the secondary antibody. Fluorescence (green) reveals the binding between FITC labeled secondary antibody and the antigen-antibody complex. Both primary and secondary antibodies were incubated at 37 ºC for 45 minutes and 30 minutes, respectively. 6/10 Braz J Vet Parasitol 2021; 30(4): e017721 Goats experimentally infected with T. vivax All of the DNA samples extracted were positive in gapdh internal control PCR assays. All of the 35 goat tissue DNA samples from infected animals that were subjected to the LAMP assay were positive. The average melting temperature for the positive samples was 88.2 ºC (SD = 0.2). In addition, the average quantification cycle was 30.4 (SD = 4.2). Likewise, analyzing only the adipose tissue DNA samples, the average melting temperature and quantification cycle were 88.0 ºC (SD = 0.09) and 30.7 (SD = 3.6), respectively. All tissue DNA samples from the negative control animal were negative. Discussion myristate, and catabolize them via beta-oxidation, which could lead to production of ATP via the cycle of tricarboxylic acid and oxidative phosphorylation (Trindade et al., 2016). Likewise, T. cruzi has been recorded in adipose tissue from mice and humans (Ferreira et al., 2011). In that study, the authors also reported that T. cruzi parasitized primary adipocytes in vitro and that both in humans and in mice, T. cruzi may persist in adipose tissue for a long time and become a reservoir of infection (Ferreira et al., 2011). These findings raise some interesting points about the role of adipose tissue in the biology of T. vivax. Could the adipose tissue be a niche for T. vivax during its mammalian life cycle? Can T. vivax parasites use fatty acids as an external carbon source? Is it possible that the persistence of T. vivax in mammalian fat may contribute to treatment failures in animals? Relapses of T. vivax infection after treatment with trypanocidal drugs such as diminazene diaceturate (DA) and isometamidium chloride (ISM) have been reported (Schönefeld et al., 1987; Cadioli et al., 2012; Bastos et al., 2017; Castilho et al., 2021). Despite some authors had suggested that T. vivax resistance to the drug is caused due to indiscriminate use, underdosing and formulations with inadequate chemotherapy concentrations (Schönefeld et al., 1987; Dagnachew & Bezie, 2015; Tekle et al., 2018), it is also possible that the parasite is able to escape from trypanocide drugs by invading host tissues in which the drug does not reach levels high enough to eliminate it (Batista et al., 2011; Bastos et al., 2020; Castilho et al., 2021). In the latter scenario, after the half-life of the trypanocide drugs has expired, the parasites may leave their “refuge sites” and returns to the host’s bloodstream. In Brazil, DA and ISM are the only two trypanocide drugs licensed by the Ministry of Agriculture, Livestock and Supply (MAPA). They are generally used against T. vivax infection (Castilho et al., 2021). These drugs have half-lives of approximately 4.5 and 12 days, respectively (Eisler, 1996; Kaur et al., 2000). Braz J Vet Parasitol 2021; 30(4): e017721 7/10 Goats experimentally infected with T. vivax DA is widely used as a trypanocide among cattle, goats and sheep due to its activity against T. congolense and T. vivax, and it only has low toxic effects. Discussion It is a curative drug (Peregrine & Mamman, 1993) and has higher concentrations in the liver and kidney (FAO, 1990). On the other hand, ISM accumulates in the liver, kidneys and spleen and at the inoculation site. From these sites, the drug is slowly released into the plasma and exerts prophylactic action (Kinabo & Bogan, 1988). Thus, because of the short half-lives of these drugs, especially that of DA, the likelihood that these drugs might reach refuge sites for this parasite, e.g. skin and fat, at ideal concentrations for clearing parasites from these sites should be taken into account in cases of relapse of T. vivax infection. Although we have presented robust results, given that blood of the animals was not removed before tissues collection, we can not rule out the possibility that tissue-based IFAs may reflect the parasites in the vasculature of the tissues rather than parasites that have crossed the endothelium and established in the extravascular spaces of tissues. In this way, future studies using perfused animals are much necessary to verify if T. vivax is able to cross the endothelium or if they are restricted to vascular systems. In the current study, using serological and molecular approaches, we showed the presence of T. vivax in tissue samples from experimentally infected goats. Our findings, coupled with those of previous studies, show that adipose tissue and skin may be used as a refuge site for the parasite. These results shed some light on relapsed T. vivax cases after treatment with trypanocide drugs. Future studies aiming to assess whether these parasites from tissues are metabolically distinct from bloodstream forms, and how long the parasites can remain active in these sites and thereafter colonize the bloodstream, are necessary. References Almeida KS, Costa AF, Silva PC, Fagliari JJ, Machado RZ, Nascimento AAD. Acute phase proteins: a potential approach for diagnosing chronic infection by Trypanosoma vivax. Rev Bras Parasitol Vet 2012; 21(2): 97-100. http://dx.doi.org/10.1590/S1984- 29612012000200005. PMid:22832747. Almeida KS, Freitas FLC, Tebaldi JH, Alessi AC, Machado RZ, Nascimento AA. Alterações clínicas, histopatológicas e enzimáticas em ovinos infectados experimentalmente por Trypanosoma vivax. Cienc Anim Bras 2010; 11(3): 669-676. http://dx.doi.org/10.5216/ cab.v11i3.5971. Bastos TSA, Faria AM, Cavalcante ASA, Madrid DMC, Zapa DMB, Nicaretta JE, et al. Comparison of therapeutic efficacy of different drugs against Trypanosoma vivax on experimentally infected cattle. Prev Vet Med 2020; 181: 105040. http://dx.doi.org/10.1016/j. prevetmed.2020.105040. PMid:32516747. Bastos TSA, Faria AM, Madrid DMC, Bessa LCD, Linhares GFC, Fidelis OL Jr, et al. First outbreak and subsequent cases of Trypanosoma vivax in the state of Goiás, Brazil. Rev Bras Parasitol Vet 2017; 26(3): 366-371. http://dx.doi.org/10.1590/s1984- 29612017019. PMid:28678894. Bastos TSA, Faria AM, Madrid DMC, Bessa LCD, Linhares GFC, Fidelis OL Jr, et al. First outbreak and subsequent cases of Trypanosoma vivax in the state of Goiás, Brazil. Rev Bras Parasitol Vet 2017; 26(3): 366-371. http://dx.doi.org/10.1590/s1984- 29612017019. PMid:28678894. Batista JS, Riet-Correa F, Barbosa RC, Guerra JL. Infecção experimental por Trypanosoma vivax em ovinos. Pesq Vet Bras 2006; 26(1): 31-37. http://dx.doi.org/10.1590/S0100-736X2006000100007. Batista JS, Riet-Correa F, Teixeira MMG, Madruga CR, Simões SVD, Maia TF. Trypanosomiasis by Trypanosoma vivax in cattle in the Brazilian semiarid: description of an outbreak and lesions in the nervous system. Vet Parasitol 2007; 143(2): 174-181. http:// dx.doi.org/10.1016/j.vetpar.2006.08.017. PMid:16965857. Batista JS, Rodrigues CM, García HA, Bezerra FS, Olinda RG, Teixeira MMG, et al. Association of Trypanosoma vivax in extracellular sites with central nervous system lesions and changes in cerebrospinal fluid in experimentally infected goats. Vet Res 2011; 42(1): 63. http://dx.doi.org/10.1186/1297-9716-42-63. PMid:21569364. Birkenheuer AJ, Levy MG, Breitschwerdt EB. Development and evaluation of a seminested PCR for detection and differentiation of Babesia gibsoni (Asian genotype) and B. canis DNA in canine blood samples. J Clin Microbiol 2003; 41(9): 4172-4177. http:// dx.doi.org/10.1128/JCM.41.9.4172-4177.2003. PMid:12958243. Brener Z. Contribuição ao estudo da terapêutica experimental da doença de Chagas [Tese Livre-Docência]. Belo Horizonte: Universidade Federal de Minas Gerais; 1961. Cadioli FA, Barnabé PA, Machado RZ, Teixeira MCA, André MR, Sampaio PH, et al. First report of Trypanosoma vivax outbreak in dairy cattle in São Paulo state, Brazil. Rev Bras Parasitol Vet 2012; 21(2): 118-124. http://dx.doi.org/10.1590/S1984-29612012000200009. PMid:22832751. Capewell P, Cren-Travaillé C, Marchesi F, Johnston P, Clucas C, Benson RA, et al. References The skin is a significant but overlooked anatomical reservoir for vector-borne African trypanosomes. eLife 2016; 5: e17716. http://dx.doi.org/10.7554/eLife.17716. PMid:27653219. 8/10 Braz J Vet Parasitol 2021; 30(4): e017721 Goats experimentally infected with T. vivax Castilho KJGA No, Garcia ABCF, Fidelis OL Jr, Nagata WB, André MR, Teixeira MMG, et al. Follow-up of dairy cattle naturally infected by Trypanosoma vivax after treatment with isometamidium chloride. Rev Bras Parasitol Vet 2021; 30(1): e020220. http://dx.doi. org/10.1590/s1984-29612021019. PMid:33909833. Cortez AP, Rodrigues AC, Garcia HA, Neves L, Batista JS, Bengaly Z, et al. Cathepsin L-like genes of Trypanosoma vivax from Africa and South America - characterization, relationships and diagnostic implications. Mol Cell Probes 2009; 23(1): 44-51. http://dx.doi. org/10.1016/j.mcp.2008.11.003. PMid:19063960. D’Archivio S, Cosson A, Medina M, Lang T, Minoprio P, Goyard S. Non-invasive in vivo study of the Trypanosoma vivax infectious process consolidates the brain commitment in late infections. PLoS Negl Trop Dis 2013; 7(1): e1976. http://dx.doi.org/10.1371/ journal.pntd.0001976. PMid:23301112. Dagnachew S, Bezie M. Review on Trypanosoma vivax. Afr J Basic Appl Sci 2015; 7(1): 41-64. Dávila AMR, Silva RAMS. Animal trypanosomiasis in South America: current status, partnership, and information technology. Ann N Y Acad Sci 2000; 916(1): 199-212. http://dx.doi.org/10.1111/j.1749-6632.2000.tb05291.x. PMid:11193622. Eisler MC. Pharmacokinetics of the chemoprophylactic and chemotherapeutic trypanocidal drug isometamidium chloride (Samorin) in cattle. Drug Metab Dispos 1996; 24(12): 1355-1361. PMid:8971142. FAO - Food and Agriculture Organization of the United Nations. Residues of some veterinary drugs in animals and foods. Rome: Food and Agriculture Organization of the United Nations - FAO; 1990 Ferreira AVM, Segatto M, Menezes Z, Macedo AM, Gelape C, Andrade LO, et al. Evidence for Trypanosoma cruzi in adipose tissue in human chronic Chagas disease. Microbes Infect 2011; 13(12-13): 1002-1005. http://dx.doi.org/10.1016/j.micinf.2011.06.002. PMid:21726660. Fidelis OL Jr, Sampaio PH, Machado RZ, André MR, Marques LC, Cadioli FA. Evaluation of clinical signs, parasitemia, hematologic and biochemical changes in cattle experimentally infected with Trypanosoma vivax. Rev Bras Parasitol Vet 2016; 25(1): 69-81. http://dx.doi.org/10.1590/S1984-29612016013. PMid:27007249. Gardiner PR, Assoku RKG, Whitelaw DD, Murray M. Haemorrhagic lesions resulting from Trypanosoma vivax infection in ayrshire cattle. Vet Parasitol 1989; 31(3-4): 187-197. http://dx.doi.org/10.1016/0304-4017(89)90069-1. PMid:2763442. Giordani F, Morrison LJ, Rowan TG, Koning HP, Barrett MP. The animal trypanosomiases and their chemotherapy: a review. Parasitology 2016; 143(14): 1862-1889. http://dx.doi.org/10.1017/S0031182016001268. PMid:27719692. omes of mammals. Oxford: Blackwell Scientific Publications; 1972. Hoare CA. The trypanosomes of mammals. Oxford: Blackwell Scientific Publications; 1972. Jones TW, Dávila AMR. Trypanosoma vivax - out of Africa. Trends Parasitol 2001; 17(2): 99-101. References http://dx.doi.org/10.1016/S1471- 4922(00)01777-3. PMid:11228017. Kaur G, Chaudhary RK, Srivastava AK. Pharmacokinetics, urinary excretion and dosage regimen of diminazene in crossbred calves. Acta Vet Hung 2000; 48(2): 187-192. PMid:11402701. Kinabo LDB, Bogan TLJ. The pharmacology of isometamidium. J Vet Pharmacol Ther 1988; 11(3): 233-245. http://dx.doi. org/10.1111/j.1365-2885.1988.tb00148.x. PMid:3054150. Machado RZ, Montassier HJ, Pinto AA, Lemos EG, Machado MRF, Valadão IFF, et al. An enzyme-linked immunosorbent assay (ELISA) for the detection of antibodies against Babesia bovis in cattle. Vet Parasitol 1997; 71(1): 17-26. http://dx.doi.org/10.1016/ S0304-4017(97)00003-4. PMid:9231985. Nantulya VM. Trypanosomiasis in domestic animals: the problems of diagnosis. Rev Sci Tech 1990; 9(2): 357-367. http://dx.doi. org/10.20506/rst.9.2.507. PMid:2132685. Njiru ZK, Ouma JO, Bateta R, Njeru SE, Ndungu K, Gitonga PK, et al. Loop-mediated isothermal amplification test for Trypanosoma vivax based on satellite repeat DNA. Vet Parasitol 2011; 180(3-4): 358-362. http://dx.doi.org/10.1016/j.vetpar.2011.03.021. PMid:21482026. Osório ALAR, Madruga CR, Desquesnes M, Soares CO, Ribeiro LRR, Costa SCG. Trypanosoma (Duttonella) vivax: its biology, epidemiology, pathogenesis, and introduction in the New World - a review. Mem Inst Oswaldo Cruz 2008; 103(1): 1-13. http:// dx.doi.org/10.1590/S0074-02762008000100001. PMid:18368231. Peregrine AS, Mamman M. Pharmacology of diminazene: a review. Acta Trop 1993; 54(3-4): 185-203. http://dx.doi.org/10.1016/0001- 706X(93)90092-P. PMid:7902657. Rebeski DE, Winger EM, Van Rooij EMA, Schöchl R, Schuller W, Dwinger RH, et al. Pitfalls in the application of enzyme-linked immunoassays for the detection of circulating trypanosomal antigens in serum samples. Parasitol Res 1999; 85: 550-556. http:// dx.doi.org/10.1007/s004360050594. PMid:10382604. 9/10 Braz J Vet Parasitol 2021; 30(4): e017721 Goats experimentally infected with T. vivax Sampaio PH, Fidelis OL Jr, Marques LC, Machado RZ, Barnabé PA, André MR, et al. Acute-phase protein behavior in dairy cattle herd naturally infected with Trypanosoma vivax. Vet Parasitol 2015; 211(3-4): 141-145. http://dx.doi.org/10.1016/j.vetpar.2015.05.014. PMid:26012859. Schönefeld A, Röttcher D, Moloo SK. The sensitivity to trypanocidal drugs of Trypanosoma vivax isolated in Kenya and Somalia. Trop Med Parasitol 1987; 38(3): 177-180. PMid:3432953. Seidl A, Dávila AMR, Silva RAMS. Estimated financial impact of Trypanosoma vivax on the Brazilian Pantanal and Bolivian lowlands. Mem Inst Oswaldo Cruz 1999; 94(2): 269-272. http://dx.doi.org/10.1590/S0074-02761999000200027. PMid:10224541. Serra-Freire NM, Rezende AML. Stomoxys calcitrans, um vetor mecânico do Trypanosoma vivax no Brasil e notas sobre o comportamento do parasito no vetor. Arq Univ Fed Rural Rio de Janeiro 1988; 11: 77-82. Silva RAMS, Silva JA, Schneider RC, Freitas J, Mesquita DP, Mesquita TC, et al. Outbreak of trypanosomiasis due to Trypanosoma vivax (Ziemann, 1905) in bovines of the Pantanal, Brazil. Mem Inst Oswaldo Cruz 1996; 91(5): 561-562. Braz J Vet Parasitol 2021; 30(4): e017721 References http://dx.doi.org/10.1590/ S0074-02761996000500005. PMid:9137742. Silva TMF, Olinda RG, Rodrigues CMF, Câmara ACL, Lopes FC, Coelho WAC, et al. Pathogenesis of reproductive failure induced by Trypanosoma vivax in experimentally infected pregnant ewes. Vet Res 2013; 44(1): 1. http://dx.doi.org/10.1186/1297-9716- 44-1. PMid:23289625. Stijlemans B, Vankrunkelsven A, Caljon G, Bockstal V, Guilliams M, Bosschaerts T, et al. The central role of macrophages in trypanosomiasis-associated anemia: rationale for therapeutical approaches. Endocr Metab Immune Disord Drug Targets 2010; 10(1): 71-82. http://dx.doi.org/10.2174/187153010790827966. PMid:20158497. Tekle T, Terefe G, Cherenet T, Ashenafi H, Akoda KG, Teko-Agbo A, et al. Aberrant use and poor quality of trypanocides: a risk for drug resistance in south western Ethiopia. BMC Vet Res 2018; 14(1): 4. http://dx.doi.org/10.1186/s12917-017-1327-6. PMid:29304792. Trindade S, Rijo-Ferreira F, Carvalho T, Pinto-Neves D, Guegan F, Aresta-Branco F, et al. Trypanosoma brucei parasites occupy and functionally adapt to the adipose tissue in mice. Cell Host Microbe 2016; 19(6): 837-848. http://dx.doi.org/10.1016/j. chom.2016.05.002. PMid:27237364. Waal T. Advances in diagnosis of protozoan diseases. Vet Parasitol 2012; 189(1): 65-74. http://dx.doi.org/10.1016/j. vetpar.2012.03.033. PMid:22503386. Whitelaw DD, Gardiner PR, Murray M. Extravascular foci of Trypanosoma vivax in goats: the central nervous system and aqueous humor of the eye as potential sources of relapse infections after chemotherapy. Parasitology 1988; 97(1): 51-61. http://dx.doi. org/10.1017/S0031182000066737. PMid:3174238. 10/10
https://openalex.org/W2895289060	https://www.scielo.br/j/rbca/a/fF5NtQk3t7vZzFXnqPXJnKF/?lang=en&format=pdf	English	null	Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach	Brazilian Journal of Poultry Science	2,018	cc-by	6,741	Brazilian Journal of Poultry Science Revista Brasileira de Ciência Avícola Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Introduction Brazil is the second largest broiler chicken producer in the world. In 2015, about 5.2 billion broiler chickens were slaughtered in establishments under the Federal Inspection Service (SIF) of the Ministry of Agriculture, Livestock and Food Supply (MAPA), and there is a projection of a 46.4% increase in chicken meat production by 2023. Simultaneously, there is an increasing demand for information on ethical aspects of animal production. Based on this, governmental actions are increasing worldwide. In the European Union (EU), Directive 2007/43/ CE (European Commission, 2007), on the protection of chickens kept for meat production, sets out compliance inputs for poultry farms, such as maximum stocking density, minimum lighting intensity, and air quality parameters. Additionally, outputs such as mortality and meat inspection data are considered with the purpose of establishing maximum stocking density values. Dermatitis, parasitic infections and systemic illness are also measured by the official veterinarian at the slaughterhouse to identify signs of poor welfare. Keywords Animal-based indicators, carcass downgrading, risk analysis, welfare assessment, welfare surveillance. Submitted: 05/December/2017 Approved: 08/April/2018 Corresponding author e-mail address Carla Forte Maiolino Molento Universidade Federal do Paraná - Rua dos Funcionários, 1540, Curitiba/PR - 80035050 - Brazil. Phone: +55 41 3350-5788 Email: carlamolento@ufpr.br Abstract We aimed to study the potential use of carcass condemnation data of broiler chicken slaughterhouses in Brazil as indicators in an animal welfare monitoring program, and to identify points to be addressed to increase data reliability. Data from 2010 to 2015 in the states of Paraná (PR), Santa Catarina (SC) and Rio Grande do Sul (RS) were used. Fractures and bruising were recorded together, representing the most prevalent welfare problem, followed by skin lesion or inflammation. In PR, progressive increases on injury, arthritis, ineffective bleeding, and air sacculitis condemnation may reveal important welfare aspects. High correlation between AWI within PR was more commonly observed than in RS and SC, perhaps as a result of earlier implementation of local meat inspection standardization. Principal component analysis showed changes on condemnation data pattern in PR after standardization, pointing injury and Escherichia coli problems as the main causes for condemnation related to animal welfare. There is considerable potential to improve animal health and welfare surveillance using meat inspection structure that is already in place for food safety purposes, provided that the competent authority harmonizes the procedure of meat inspection among the States, sets specific animal welfare outcomes to be monitored, and integrates condemnation, transport and flock data. It seems crucial to update data collection to establish a routine that allows risk analysis regarding both food safety and animal welfare. In this regard, cooperative work between Federal Inspection and companies seems an interesting approach to promote transparency of the production processes, which would benefit society and animals. Mail Address Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach ISSN 1516-635X Jul - Sept 2018 / v.20 / n.3 / 547-554 http://dx.doi.org/10.1590/1806-9061-2017-0706 Author(s) Souza APOI Taconeli CAII Plugge NFIII Molento CFMI I Animal Welfare Laboratory, Federal Univer- sity of Paraná, Rua dos Funcionários 1540, 80035-050, Curitiba, Paraná, Brazil. II Department of Statistics, Federal University of Paraná, Centro Politécnico, 81531-990, Curitiba, Paraná, Brazil. III Ministry of Agriculture, Livestock and Food Supply, Rua Romário Martins 625, 84165- 010, Castro, Paraná, Brazil. III Ministry of Agriculture, Livestock and Food Supply, Rua Romário Martins 625, 84165- 010, Castro, Paraná, Brazil. Brazilian Journal of Poultry Science Revista Brasileira de Ciência Avícola Animal-based indicators, carcass downgrading, risk analysis, welfare assessment, welfare surveillance. Material and Methods Publicly available official slaughter and carcass condemnation data from January 2010 to December 2015 were obtained from the SIGSIF platform, MAPA website (www.agricultura.gov.br). Reports were ge- nerated in portable document format (PDF) and we transformed them into Excel® files to be analyzed. We then selected the three main producer States, Paraná (PR), Santa Catarina (SC) and Rio Grande do Sul (RS), all located in Southern Brazil. We analyzed general data regarding total and partial carcass condemnation in these three States. Additionally, we identified animal welfare target indicators (AWI) to be further assessed: abscess (ABS), airsacculitis/respiratory disease (AIR), arthritis (ART), ascites (ASC), bruises, contact dermatitis, dead on arrival (DOA), emaciation (EMA), dehydration, fracture, hepatitis (HEP), inadequate bleeding (INB), pericarditis (PER) and septicaemia (SEP)(EFSA, 2013, 2012; European Commission, 2007). Temperature (oC) and humidity (%) were collected from the National Institute of Meteorology (http://www.inmet.gov.br) for the same period to study correlations. In Brazil, the SIF is responsible for sanitary inspection at slaughterhouses under federal control and it is linked to the Department of Inspection of Products of Animal Origin (DIPOA) of the MAPA. Inspection is performed by a permanent team composed by official veterinarians and auxiliary staff. Activities performed by the SIF are regulated by the Decree 30,691, known as RIISPOA, which establishes the procedures of sanitary inspection of animal origin products (Brasil, 2017). Additionally, there is a specific regulation for the inspection of broiler chicken meat (MAPA, 1998). All carcass condemnation data obtained by SIF is recorded at the information management system (SIGSIF) and reports are publicly available. Meat inspection data was transformed in broiler chicken carcass condemnation per 100,000 birds. Descriptive statistics was used to verify the frequency of condemnations. Spearman rank correlation test was used to analyze the correlation between carcass condemnation data and climate variables, correlation of each condemnation cause between and within States. Correlations where R > 0.6 were considered high, and 0.6 < R < 0.3 were considered moderate. Nonparametric change point analysis (James & Matteson, 2013) was used to detect possible change points observed in carcass condemnation data from PR. Biplots based on Principal Component Analysis (PCA) were used to explore variance and covariance structure of data. PCA was based on the correlation matrix, using standardized data, to eliminate scale effects. The biplot was used to assess condemnation data and time simultaneously in a two-dimensional representation (Rencher, 2003). Keywords Animal-based indicators, carcass downgrading, risk analysis, welfare assessment, welfare surveillance. 547 Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Souza APO, Taconeli CA, Plugge NF, Molento CFM and to identify points to be addressed to increase data reliability. and to identify points to be addressed to increase data reliability. Outcomes assessed at the slaughterhouse have the potential to improve animal welfare (Grandin, 2017). Outcomes assessed at the slaughterhouse have the potential to improve animal welfare (Grandin, 2017). The use of carcass condemnation data as an official monitoring program of animal welfare (AW) is expected to promote practical consequences to animals, since feedback from slaughterhouse may gradually improve practices on farm (European Commission, 2017). However, based on the EU example, there are challenges to effectively implement such control, mainly regarding the variability of procedures among Member States (Butterworth et al., 2016). Thus, EU Members States organized a network for exchanging technical information to improve implementation of the Directive 2007/43/CE. Additionally, meat inspection data have been considered useful to investigate animal welfare (Correia-Gomes et al., 2017, 2016; Huneau- Salaün et al., 2015; Knage-Rasmussen et al., 2015). Thus, creation and use of a meat inspection database seems to constitute a potential tool to improve public policies related to the welfare of farm animals. This seems also a practical approach, since there is a structure already in place with the primary purpose of controlling food safety, which may benefit AW actions. Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Results The total number of broiler chickens slaughtered in Brazil under SIF, from January 2010 to December 2015, is observed in Figure 1. In Southern Brazil, almost 19 billion broiler chickens were slaughtered between 2010 and 2015, representing 62.2% of national broiler chicken production. Considering the 27 States in Brazil, there are 18 that produce broiler chicken meat, of which, Paraná accounts for one third of total national production. Figure 2 – Number of broiler chickens slaughtered, and number of carcasses conde- mned in the States of Southern Brazil, from 2010 to 2015. Columns refer to broiler chickens slaughtered, lines and percentages refer to carcass condemnation. Animal welfare indicators and the main causes for carcass condemnation in PR, SC and RS are presented in Figure 3. Causes for condemnations that were below 80/100,000 carcasses were presented as ‘others’, such as coligranuloma, over scalding, delayed evisceration, myositis, tumor, salpingitis and hemorrhagic syndrome. Paraná reported additional data for aspergillosis, hypertrophy and omphalitis; SC for colibacillosis and omphalitis; and RS for nephritis, myocarditis and enteritis. There was no record for the prevalence of contact dermatitis, DOA and dehydration. p Figure 1 – Broiler chicken slaughter in Brazil, by state and region, from 2010 to 2015. Percentages refer to the proportion of the total of 30.4 billion broiler chickens slaughte- red under Federal Inspection Service; PR, Paraná; SC, Santa Catarina; RS, Rio Grande do Sul; SP, São Paulo; MG, Minas Gerais; GO, Goiás; MT, Mato Grosso; MS, Mato Grosso do Sul; DF, Federal District; BA, Bahia; PA, Pará; PE, Pernambuco; ES, Espírito Santo; PB, Paraíba; TO, Tocantis; RO, Rondônia; PI, Piaui; SE, Sergipe. Considering AWI, bruises and fractures were both registered as injury, with no discrimination between them. Injury was the main cause of condemnation in PR, SC and RS, representing, in 2015, 22.1% of items condemned in PR, 19.4% in SC and 23.7% in RS. Dermatosis was the second most common cause in PR (14.8%) and in RS (9.1%). Increasing occurrence of a type of myopathy, named as dorsal cranial myopathy (MYO), was observed in the three States (Figure 3). From 2010 to 2015, MYO increased from 0.01% to 4.4% of total of carcasses downgraded in PR, being the sixth cause of condemnation in this State in 2015. Material and Methods This technique was applied to data from PR to further understand the effect of the standardization of the data collection within the State, to allow comparison of the data in two periods, before and after the standardization procedure. Robustness of biplot was verified by identifying outlier values (López-de-Lacalle, 2016) and repeating data analysis by replacing outlier with values derived from the According to Vannier et al. (2014), a set of harmonized welfare outcome indicators may be used by competent authorities in the framework of inspection and by private sectors to improve transparency in the market of animal products. Food production chain provides valuable data collection that can be used to improve disease control, animal health, public health and animal welfare. However, carcass condemnation data are not used for animal welfare purposes in Brazil. Our hypotheses were that broiler meat inspection data in Brazil comprises important AW indicators, and that adjustments are required to improve data collection. Thus, we aimed to study the potential use of carcass condemnation data of broiler chicken slaughterhouses in Brazil as indicators in an AW monitoring program 548 Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Figure 2 – Number of broiler chickens slaughtered, and number of carcasses conde- mned in the States of Southern Brazil, from 2010 to 2015. Columns refer to broiler chickens slaughtered, lines and percentages refer to carcass condemnation. Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Souza APO, Taconeli CA, Plugge NF, Molento CFM Souza APO, Taconeli CA, Plugge NF, Molento CFM statistical average of previous and subsequent months. Analysis were performed using R Statistical Computing Environment version 3.3.1 (R Core Team, 2016). Figure 2 – Number of broiler chickens slaughtered, and number of carcasses conde- mned in the States of Southern Brazil, from 2010 to 2015. Columns refer to broiler chickens slaughtered, lines and percentages refer to carcass condemnation. Results In SC, MYO was the second cause of condemnation in 2015, moving from 0.1% to 10.1% of total of carcasses downgraded; and it was the third cause of condemnation in RS, moving from 0.7% to 8.1%. Figure 1 – Broiler chicken slaughter in Brazil, by state and region, from 2010 to 2015. Percentages refer to the proportion of the total of 30.4 billion broiler chickens slaughte- red under Federal Inspection Service; PR, Paraná; SC, Santa Catarina; RS, Rio Grande do Sul; SP, São Paulo; MG, Minas Gerais; GO, Goiás; MT, Mato Grosso; MS, Mato Grosso do Sul; DF, Federal District; BA, Bahia; PA, Pará; PE, Pernambuco; ES, Espírito Santo; PB, Paraíba; TO, Tocantis; RO, Rondônia; PI, Piaui; SE, Sergipe. General slaughter and carcass condemnation data for the States of Paraná (PR), Santa Catarina (SC) and Rio Grande do Sul (RS) are presented in Figure 2. In PR broiler chicken slaughter increased by 27.9% from 2010 to 2015, with increased carcass condemnation rates; the moment for the changepoint was statistically estimated to be March 2013. In SC and RS, broiler chicken slaughter was stable from 2010 to 2015, and a tendency toward increasing carcass condemnation reports through these years was not observed. Higher condemnation in 2015 in SC occurred due to an unusual peak of condemnation for dermatosis in February (19,120/100,000 birds slaughtered); this was not representative of the situation in SC. Excluding data from February/2015, condemnation rate in 2015 was slightly high in SC, reaching 5.6%. In PR, high correlations between condemnation causes were more commonly observed than in RS and SC (Figure 4). Principal component analysis in PR showed changes in the condemnation data pattern after a standardization procedure was introduced in 2012 (Figure 5). For example, one group of indicators was strongly related with the component 1, representing 39.4% of total data variability. The group was composed of the indicators INJ, ABS, AIR, CEL, COL and DER. Notification of indicators was strengthened in 2014 and 2015 in PR, observed by the distribution of dates in Figure 5. Outliers did not cause significant changes on the original data. 549 neli CA, to CFM Broiler Chicken Meat Inspection Data in B A First Glimpse into an Animal Welfare Ap Figure 3 – Broiler chicken carcass condemnation per 100,000 birds in the States of Paraná, Santa Catarina and Rio Grande do Sul, Southern Brazil, from 2010 to 2015. Results CON, contamination; DER, dermatosis; INJ, injury; ABS, abscess; AIR, airsacculitis; ART, arthritis; ASC, ascites; CEL, cellulitis; COL, colibacillosis; STE, steatosis; MYO, dorsal cranial myopathy; RAS, abnormal aspect; EMA, emaciation; INB, inadequate bleeding; SEP, septicaemia; HEP, hepatitis; PER, pericarditis; others include all reasons for condemnations below 80/100,000 carcasses, except target animal welfare indicators of interest in this study. Souza APO, Taconeli CA, Plugge NF, Molento CFM Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Figure 3 – Broiler chicken carcass condemnation per 100,000 birds in the States of Paraná, Santa Catarina and Rio Grande do Sul, Southern Brazil, from 2010 to 2015. CON, contamination; DER, dermatosis; INJ, injury; ABS, abscess; AIR, airsacculitis; ART, arthritis; ASC, ascites; CEL, cellulitis; COL, colibacillosis; STE, steatosis; MYO, dorsal cranial myopathy; RAS, abnormal aspect; EMA, emaciation; INB, inadequate bleeding; SEP, septicaemia; HEP, hepatitis; PER, pericarditis; others include all reasons for condemnations below 80/100,000 carcasses, except target animal welfare indicators of interest in this study. Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Souza APO, Taconeli CA, Plugge NF, Molento CFM Figure 3 – Broiler chicken carcass condemnation per 100,000 birds in the States of Paraná, Santa Catarina and Rio Grande do Sul, Southern Brazil, from 2010 to 2015. CON, contamination; DER, dermatosis; INJ, injury; ABS, abscess; AIR, airsacculitis; ART, arthritis; ASC, ascites; CEL, cellulitis; COL, colibacillosis; STE, steatosis; MYO, dorsal cranial myopathy; RAS, abnormal aspect; EMA, emaciation; INB, inadequate bleeding; SEP, septicaemia; HEP, hepatitis; PER, pericarditis; others include all reasons for condemnations below 80/100,000 carcasses, except target animal welfare indicators of interest in this study. We observed disparate values among the three States. As an example, condemnation data for abscess in RS and PR were, respectively, 21.0 and 8.8 times greater than the reported value in SC; and there was almost double the rate of condemnation for dermatosis in PR as compared to SC and RS. There was poor correlation in respect to condemnation rates among the three States. High correlation was observed for arthritis between PR and RS (p<0.001; R = 0.86), injury between PR and SC (p<0.001; R = 0.62), and for ascites between PR and SC (p<0.001; R = 0.66), PR and RS (p<0.001; R = 0.81) and between SC and RS (p<0.001; R = 0.83). Results Ascites was the only AWI that presented high correlation with a climate variable Figure 4 – Correlation of broiler chicken carcass condemnation indicators in the States of Paraná, Santa Catarina and Rio Grande do Sul, Southern Brazil, from 2010 to 2015 (broiler chicken carcass condemnation per 100,000 birds). ABS, abscess; AIR, airsaccu- litis; ART, arthritis; ASC, ascites; RAS, abnormal aspect; EMA, emaciation; CEL, cellulitis; COL, colibacillosis; DER, dermatosis; MYO, dorsal cranial myopathy; SAL, salpingitis; SEP, septicaemia; HEP, hepatitis; PER, pericarditis; ellipse shape is directly proportional to correlation strength, higher correlations appear close to an ellipse format; the orienta- tion of the ellipse indicate positive (upwards to the right) or negative (upwards to the left) correlations. 550 Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Souza APO, Taconeli CA, Plugge NF, Molento CFM (Temperature; SC, p<0.001, R = -0.71; RS, p<0.001, R = -0.76; PR, p<0.001, R = -0.61). Other correlations between condemnation data and climate variables were moderate and low. According to MAPA, dermatosis is a generic term used to record any skin or meat lesion without inflammation; and inflammatory processes, such as cellulitis and dermatitis, must be recorded as specific indicators (MAPA, 1998). There was no record of condemnation for contact dermatitis. In the case of footpad dermatitis, the absence of records occurred because broiler chicken feet with contact dermatitis are exported as lower grade product authorized by DIPOA to China and Hong Kong. In other cases, feet that were not marketed were discarded before inspection by the competent authority. Thus, since feet were not condemned, there was no official record about the incidence of footpad dermatitis. Since dermatosis includes a wide range of occurrences, it may contribute to high variability between SIF records, and is a potential item to be improved on data collection. In addition, considering that contact dermatitis is relevant to broiler chicken welfare (EFSA, 2012; European Commission, 2017), the implementation of an official monitoring program covering this issue seems crucial in Brazil. Figure 5 – Principal Component Analysis of broiler chicken carcass condemnation indicators in the state of Paraná, Southern Brazil, from 2010 to 2011, and from 2012 to 2015. Discussion Dorsal cranial myopathy (MYO) has been observed in Brazil since 2006, and it was reported as a lesion of the anterior latissimus dorsi (ALD) muscle (Zimmermann et al., 2012). Zimmermann et al. (2012) suggested that MYO is related to fast growing breeds, whose body is unbalanced and may cause intermittent interruption of blood flow of ALD when wings move over the large pectoral muscle of birds. Information about causes of MYO and its impact on animal welfare is scarce. Other studies about similar myopathies in broiler chickens have discussed the influence of genetics (Petracci et al., 2015) and both genetics and environment (Bailey et al., 2015). Due to the possible correlation of MYO with items that impact broiler chicken welfare and its intrinsic welfare impact, it seems an interesting indicator to be recorded and further studied. Results DER, dermatosis; ABS, abscess; AIR, airsacculitis; ART, arthritis; ASC, ascites; CEL, cellulitis; COL, colibacillosis; MYO, dorsal cranial myopathy; RAS, abnormal aspect; EMA, emaciation; SAL, salpingitis; SEP, septicaemia; numbers inside the biplot represent month/year of the data derived from the database. Figure 5 – Principal Component Analysis of broiler chicken carcass condemnation indicators in the state of Paraná, Southern Brazil, from 2010 to 2011, and from 2012 to 2015. DER, dermatosis; ABS, abscess; AIR, airsacculitis; ART, arthritis; ASC, ascites; CEL, cellulitis; COL, colibacillosis; MYO, dorsal cranial myopathy; RAS, abnormal aspect; EMA, emaciation; SAL, salpingitis; SEP, septicaemia; numbers inside the biplot represent month/year of the data derived from the database. Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Broiler chicken condemnation data in Brazil Condemnation for ART and AIR suggest that changes promoted in the broiler chicken industry in PR in the last six years, like migration from natural lit poultry houses to those working exclusively with artificial lighting (Souza et al., 2015), had negative impacts on animal welfare and should be scrutinized from this perspective. Even though genetics has been pointed as the main cause for skeletal disorders in fast growing breeds, the lack of bird activity aggravates the problem (Bradshaw et al., 2002; EFSA, 2010). In the case of PR, broiler activity may have been reduced by both increased stocking density and low lighting, contributing to higher levels of skeletal disorders. High stocking density is also correlated to reduced air quality, increased heat stress and increased transmission of infectious diseases. Thus, higher condemnation for airsacculitis may indicate worse managing practices on farm since factors associated to the etiology of air sac disease are poor air quality, mainly high levels of dust and ammonia, associated with Mycoplasma gallisepticum or Escherichia coli infection (EFSA, 2012; Gross, 1961). Additionally, thermal conditions have the potential to cause stress and, thus, to decrease the immune response in poultry (Lara & Rostagno, 2013), predisposing birds to disease. The specific Brazilian regulation about broiler chicken slaughter, Ordinance no. 210/1998, includes a list of condemnation causes to be reported by the SIF of each plant (MAPA, 1998). It is not an exhaustive list; however, it covers most of the selected AWI. In general, States in Southern Brazil recorded items demanded by national ordinance; however, each local SIF personnel may provide additional information and each State may standardize which items will be informed in the SIGSIF. This has created variations across States. Since 2009, MAPA has demanded States to set guidelines for the management of inspection service, including the standardization of post-mortem procedures. As an example, in PR, a group of official veterinarians met in 2012 to discuss about criteria for carcass evaluation and destination, and outputs from this meeting oriented SIF personnel in each slaughterhouse within the State. We observed that the standardization of procedures in PR took about one year to be fully visible, clearly dividing general condemnation data into two levels, before and after 2013 (Figure 2) and changing condemnation data pattern (Figure 5). Broiler chicken condemnation data in Brazil Additionally, thermal conditions have the potential to cause stress and, thus, to decrease the immune response in poultry (Lara & Rostagno, 2013), predisposing birds to disease. Dead on arrival is controlled by SIF for each batch slaughtered and data may be recorded at SIGSIF. In addition DOA higher than 1% must be reported to The extent of blood loss is affected by stunning, type of neck cut, time between stunning and bleeding and time for bleeding (Bilgili, 1988). All causes mentioned are controlled by SIF, by the national Ordinance 210/1998 and the Normative Instruction 3/2000. In addition, modification of the processing line speed must be approved by SIF regarding food safety concerns and proper post-mortem inspection. However, higher line speed combined with expansion of Halal meat exportation in PR may have affected bleeding efficiency. In PR, exportation of broiler chicken products to Middle East countries increased 70.0% from 2010 to 2015. Faster line speed requires more staff to perform neck cut during religious slaughter, thus space on the slaughter line and/or number of employees for neck cutting may potentially be insufficient. An indicator of bleeding efficiency based on the ratio between line speed and number of employees for manual slaughter may be an interesting approach to be studied in Brazil. Correlations of ascites and temperature, as well as the correlation of ascites data among the three States in Southern Brazil, suggest it is a well-established health indicator at SIF. Disparate results in other indicators may be caused by several factors. Specific characteristics of each company, such as orientation to broiler chicken farmers, infrastructure, management policies, export market and labor will directly affect carcass condemnation data. Thus, it is possible that weak correlations between condemnation categories in certain states could be a consequence of combining data from companies with heterogeneous management practices and health problems, which may be further explored in future studies. However, lack of standardization is one weakness of meat inspection as health and animal welfare surveillance system (Huneau-Salaün et al., 2015). Based on our data, difference on carcass evaluation among SIF seemed to be the core point to improve quality of meat inspection data. Broiler chicken condemnation data in Brazil High prevalence of injury, skin problems and arthritis observed in Southern Brazil had already been observed in carcass condemnation data from 2006 to 2011(Oliveira et al., 2016); thus, these items have been important animal welfare issues in Brazil for a decade. Discrimination between bruises and fractures is described in scientific literature (Grandin, 2010) and is an important point to be improved with regard to meat inspection data in Brazil. Bruising and fractures used to be controlled separately as part of the MAPA Circular 294/2006, which established that companies had to implement self-monitoring programs, including animal welfare, and determined SIF as responsible for verifying those programs. Circular 294/2006 was repealed and current regulation about self-monitoring, Normative 01/2017, does not include the requirement for assessing injuries. In addition, data provided by self-monitoring programs were not recorded in the national database, and remain under used or even unused. Moreover, recognition of the relevance of injuries as a food safety problem may be variable amongst official veterinarians. Consequently, low values may be reported due to the acceptance of injured meat by less demanding markets or to be used as raw material in processed products. In this case, the inclusion of an AW concept to carcass condemnation data is encouraged to provide a standard procedure within all Brazilian States. In PR, where both percent condemnation and number of carcasses slaughtered increased, progressive increase in condemnation rates for injury, inadequate bleeding, arthritis and airsacculitis may indicate important welfare aspects to be considered. For example, injury may increase if employees are not adequately trained to handle live birds and if structure to transport live birds or staff responsible for catching and shackling birds is undersized (Grandin, 2010). In the case of PR, the competent authority at each slaughterhouse may accept carcasses or parts of carcass with small bruises. Nevertheless, condemnation for injury increased. Thus, we consider it important to assess whether the whole production chain structure, including activities where live birds are handled, 551 Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Souza APO, Taconeli CA, Plugge NF, Molento CFM Souza APO, Taconeli CA, Plugge NF, Molento CFM presented a proportional increase as that observed in the number of broiler chickens slaughtered. presented a proportional increase as that observed in the number of broiler chickens slaughtered. Broiler chicken condemnation data in Brazil were not presented on SIGSIF condemnation reports, which prevented us to further study this indicator. In Brazil, both staff and a database to register this information are already in place. Thus, it seems feasible to standardize the procedure of registering and analyzing DOA, as well as making it publicly available, representing a structural advancement for the meat chain and public policies. g The extent of blood loss is affected by stunning, type of neck cut, time between stunning and bleeding and time for bleeding (Bilgili, 1988). All causes mentioned are controlled by SIF, by the national Ordinance 210/1998 and the Normative Instruction 3/2000. In addition, modification of the processing line speed must be approved by SIF regarding food safety concerns and proper post-mortem inspection. However, higher line speed combined with expansion of Halal meat exportation in PR may have affected bleeding efficiency. In PR, exportation of broiler chicken products to Middle East countries increased 70.0% from 2010 to 2015. Faster line speed requires more staff to perform neck cut during religious slaughter, thus space on the slaughter line and/or number of employees for neck cutting may potentially be insufficient. An indicator of bleeding efficiency based on the ratio between line speed and number of employees for manual slaughter may be an interesting approach to be studied in Brazil. Condemnation for ART and AIR suggest that changes promoted in the broiler chicken industry in PR in the last six years, like migration from natural lit poultry houses to those working exclusively with artificial lighting (Souza et al., 2015), had negative impacts on animal welfare and should be scrutinized from this perspective. Even though genetics has been pointed as the main cause for skeletal disorders in fast growing breeds, the lack of bird activity aggravates the problem (Bradshaw et al., 2002; EFSA, 2010). In the case of PR, broiler activity may have been reduced by both increased stocking density and low lighting, contributing to higher levels of skeletal disorders. High stocking density is also correlated to reduced air quality, increased heat stress and increased transmission of infectious diseases. Thus, higher condemnation for airsacculitis may indicate worse managing practices on farm since factors associated to the etiology of air sac disease are poor air quality, mainly high levels of dust and ammonia, associated with Mycoplasma gallisepticum or Escherichia coli infection (EFSA, 2012; Gross, 1961). Broiler chicken condemnation data in Brazil It may be the result of strengthened training performed with SIF staff, in addition to possible problems related to Dead on arrival is controlled by SIF for each batch slaughtered and data may be recorded at SIGSIF. In addition, DOA higher than 1% must be reported to the Animal Health Service of each State, according to the MAPA Normative Instruction 17/2006. However, these data were not available for consultation and 552 Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Souza APO, Taconeli CA, Plugge NF, Molento CFM Souza APO, Taconeli CA, Plugge NF, Molento CFM companies may be an interesting approach to increase AW data collection, in addition to condemnation data collected by the competent authority. As stated by Short and Toffel (Short & Toffel, 2008), success of self- monitoring depends on the continued involvement of regulators with coercive powers. Thus, SIF supervision on food safety and animal welfare issues to support activities on animal production seems essential. broiler chicken chain. In RS, the memorandum 048/ SICAO/014 and a manual were published in 2014 and 2015, respectively, to guide official veterinarians within the State. As result, since 2014 data from RS seem better organized regarding terms used to describe cause and type of condemnation; however, information about downgrading of parts and giblets, including condemnation for hepatitis and pericarditis, was suppressed. Similarly, in PR condemnation of liver for hepatitis and heart for pericarditis have been recorded as carcass partial condemnation for colibacillosis since 2012, with the loss of valuable information. Thus, the standardization procedure adopted in Southern Brazil may lead to unreported indicators, reducing power of condemnation data as a surveillance system for animal health and welfare. Brazilian government was moving to pass responsibility of meat inspection to industry. Reducing SIF operation will also affect surveillance in AW at the slaughterhouse, because it is part of the official veterinary activities. Similarly, in 2016 the Department of Environment, Food and Rural Affairs tried to put the welfare code on chicken farming under the control of the poultry industry in the United Kingdom. This process was interrupted mainly due to public opinion and pressure from non-governmental organizations. Recent disclosure of Brazilian Federal Police investigation related to meat inspection revealed high public concern about meat quality. Conclusion The inclusion of an AW view on meat inspection data is a new concept that seems applicable to Brazil, since the data present information with potential use as AW indicators. Our results indicate a need to harmonize SIF procedures among States, to set specific AW outcomes to be monitored and to integrate condemnation, transport and flock data. Points to be improved include differentiation of bruising from fracture when recording these lesions; refined assessment of skeletal disorders and contact dermatitis; and monitoring the ratio of line speed and number of employees for neck cutting. Results suggest a need to update data collection to keep pace with modern animal production, as well to establish a routine of data analysis as part of risk analysis for both food safety and animal welfare. Overall, there is considerable potential to improve animal health and welfare surveillance using the structure of meat inspection that is already in place for food safety purposes, provided that MAPA addresses issues related to the weakness of data collection process. In this regard, cooperative work between Federal Inspection and companies seems to be an interesting approach to Broiler chicken condemnation data in Brazil Thus, it is advisable that the MAPA take public opinion into consideration, since consumers represents a powerful stakeholder in the food chain. In the case of Brazil, society was not inquired about the changes on food inspection system proposed by MAPA, which may be detrimental to the relation between society, government and industry. Higher coherence of indicators presenting high correlation in PR, as compared to SC and RS (Figure 4) may be a result of the standardization of meat inspection procedure. In contrast to PR, lack of high correlation among condemnation data in SC (Figure 4) may point variance on carcass judgment. The development and maintenance of a robust system of meat inspection data collection at national level is challenging. Regional organization of SIF proposed in Brazil may be more dynamic and improve activities within each State as compared to a national guidance; however, it may result in increased variation among States, creating uncertainty about the efficiency of Competent Authority in delivering reliable data. It seems advisable that standardization be either centralized at federal level by MAPA or that the regional organization be indirectly guided by MAPA. References Knage-Rasmussen KM, RousingT, Sørensen JT, Houe H. Assessing animal welfare in sow herds using data on meat inspection, medication and mortality. Animal 2015;9:509–515. Algers B, Anil H, Blokhuis H, Fuchs K, Hultgren J, Lambooij B, et al. Project to develop animal welfare risk assessment guidelines on stunning and killing. EFSA Supporting Publication 2009;6:1–88. Lara LJ, Rostagno MH. Impact of heat stress on poultry production. Animals 2013;3:356–369. Bailey RA, Watson KA, Bilgili SF, Avendano S. The genetic basis of pectoralis major myopathies in modern broiler chicken lines. Poultry Science 2015;94:2870–2879. Available from: https://doi:10.3382/ps/pev304. López-de-Lacalle, J. Detection of Outliers in Time Series. In: López-de- Lacalle, J. Tsoutliers T package. Vizcaya: University of the Basque Country; 2016. Bilgili SF. Electrical stunning of broilers - basic concepts and carcass quality implications:a review. Journal of Applied Poultry Research 1988;135– 146. MAPA – Ministério da Agricultura, Agropecuária e Abastecimento. Portaria 210 de 10 de novembro de 1998.Aprova o regulamento técnico da inspeção tecnológica e higiênico-sanitaria de carne de aves. Brasília, DF; 1998. Bradshaw RH, Kirkden RD, Broom DM. A review of the aetiology and pathology of leg weakness in broilers in relation to welfare. Avian and Poultry Biology Reviews 2002;13:45–103. Oliveira AA, Andrade MA, Armendaris PM, Bueno PHS. Principais causas de condenação ao abate de aves em matadouros frigoríficos registrados no serviço brasileiro de inspeção federal entre 2006 e 2011. Ciência Animal Brasileira 2016;17:79–89. Brasil. Decreto 9013 de 29 de março de 2017. Regulamenta a Lei no 1.283, de 18 de dezembro de 1950, e a Lei no 7.889, de 23 de novembro de 1989, que dispõem sobre a inspeção industrial e sanitária de produtos de origem animal. Brasília, DF; 2017. Petracci M, Mudalal S, Soglia F, Cavani C. Meat quality in fast-growing broiler chickens. Worlds Poultry Science Journal 2015;71:363–374. Butterworth A, Jong IC, Keppler C, Knierim U, Stadig L, Lambton S. What is being measured, and by whom? Facilitation of communication on technical measures amongst competent authorities in the implementation of the European Union Broiler Directive (2007/43/EC). Animal 2016;10:302–308. R Core Team. A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2016. Rencher AC. Methods of multivariate analysis. New York: John Wiley & Sons; 2003. Correia-Gomes C, Eze JI, Borobia-Belsué J, Tucker AW, Sparrow D, Strachan D, et al. Federal Inspection Service potential to improve animal welfare in Brazil Brazilian Federal Inspection Service has been working to push companies to higher sanitary status to prevent foodborne diseases. This has been achieved through the implementation of hazard analysis and critical control points program (HACCP). In Brazil, since 1998 animal product processors are demanded to implement HACCP, but in broiler chicken slaughterhouses implementation has been strengthened since 2006, with the publication of Circular 668/2006. The same principle of hazard analysis for food safety concerns may be applied to animal welfare issues (Algers et al., 2009; Smulders, 2009). This is a new research area, and meat inspection data may facilitate its development. Potential AWI may be chosen as critical control points to be monitored, including the proposal of setting critical limits for AWI. Self-monitoring programs by 553 Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Broiler Chicken Meat Inspection Data in Brazil: A First Glimpse into an Animal Welfare Approach Souza APO, Taconeli CA, Plugge NF, Molento CFM Souza APO, Taconeli CA, Plugge NF, Molento CFM increase public information about animal welfare and to promote transparency of production process, which would benefit society and animals. European Commission. Study on the application of the broilers directive (DIR 2007/43/EC) and development of welfare indicators. Brussels: European Union; 2017. pp 261. European Commission. Study on the application of the broilers directive (DIR 2007/43/EC) and development of welfare indicators. Brussels: European Union; 2017. pp 261. Grandin T. On-farm conditions that compromise animal welfare that can be monitored at the slaughter plant. Meat Science 2017;132:52–58. Acknowledgements Grandin T. Welfare during transport of livestock and poultry.In: Grandin T, editor. Improving animal welfare: a practical approach. Oxfordshire: CABI; 2010. p.115–138. The authors wish to thank Liziè Pereira Buss who encouraged this study, the undergraduates Rita de Cássia G. Silva and Mariana Tiepo for data organization and two reviewers who offered valuable recommendations to this manuscript. We also acknowledge that Ana Paula de Oliveira Souza is the recipient of a CAPES (Ministry of Education, Brazil) doctorate scholarship. GrossWB. The development of “air sac disease.” Avian Diseases 1961;5:431–439. Huneau-Salaün A, Stärk KDC, Mateus A, Lupo C, Lindberg A, Le Bouquin- Leneveu S. Contribution of meat inspection to the surveillance of poultry health and welfare in the European Union. Epidemiology & Infection 2015;143:2459–2472. James NA, Matteson DS. ECP: an R package for nonparametric multiple change point analysis of multivariate data. Journal of Statistical Software 2014;62(7):1-25. References Voluntary monitoring systems for pig health and welfare in the UK:Comparative analysis of prevalence and temporal patterns of selected non-respiratory post mortem conditions. Preventive Veterinary Medicine 2017;146:1–9. Short JL, Toffel MW. Coerced confessions: self-policing in the shadow of the regulator. Journal of Law, Economics & Organization 2008;24:45–71. Smulders FJM. A praticable approach to assessing risks for animal welfare - methodological considerations. In: Smulders FJM, Algers B., editors. Welfare of production animals:assessment and management of risks. Wageningen: Wageningen Academic Publishers; 2009. p.239–274. Correia-Gomes C, Smith RP, Eze JI, Henry MK, Gunn GJ, Williamson S, et al. Pig abattoir inspection data: can it be used for surveillance purposes? PLoS One 2016;11:e0161990. EFSA. Preparation of a data collection system of welfare indicators in EU broilers’ slaughterhouses. EFSA Journal 2013;11:19–28. Souza APO, Sans ECO, Müller BR, Molento CFM. Broiler chicken welfare assessment in GLOBALGAP certified and non- certified farms in Brazil. Animal Welfare 2015;24:45–54. EFSA. Scientific report updating the EFSA opinions on the welfare of broilers and broiler breeders. EFSA Supporting Publication 2012;9:116. Vannier P, Michel V, Keeling LJ. Science-based management of livestock welfare in intensive systems :looking to the future. Revue Scientifique et Technique 2014;33:153–160. EFSA. Scientific Opinion on the influence of genetic parameters on the welfare and the resistance to stress of commercial broilers. EFSA Journal 2010;8:1–82. Zimmermann FC, Fallavena LCB, Salle CTP, Moraes HLS, Soncini RA, Barreta MH, et al. Downgrading of heavy broiler chicken carcasses due to myodegeneration of the anterior latissimus dorsi: pathologic and epidemiologic studies. Avian Pathology 2012;56:418–421. European Commission. Laying down minimun rules for the protection of chickens kept for meat production [Directive 43/2007/EC]. Brussels; 2007. 554
https://openalex.org/W1832484820	https://www.mongoliajol.info/index.php/MJAS/article/download/506/486	English	null	A HISTOLOGICAL STRUCTURE AND SPECIFIC PROTEIN OF BROWN ADIPOSE TISSUE OF MONGOLIAN LAMBS	Mongolian journal of agricultural sciences/Hôdôô až ahujn šinžlèh uhaan	2,015	cc-by	2,302	KhorolmaaCh, Demberel Sh, Mongolian Journal of Agricultural Sciences №13 (02), 2014:8-12 KhorolmaaCh, Demberel Sh, Mongolian Journal of Agricultural Sciences №13 (02), 2014:8-12 8 ABSTRACT Brown fat tissue of Mongolian lamb is divided histologically into a number of lobules with connective tissue and connective-muscular tissue septa,consists of fat cells, cytoplasm of which contains small vacuoles showing the presence of multiple lipid droplets (multilocular), and blood vessels are abundant there. However, lipid droplets in cytoplasm of brown adipose tissue /BAT/ cells of 7-day old lambs enlarge and cells are generally unilocular type. Immunohistochemical analysis of this fat tissue by using specific antiserum of sheep UCP1 protein (rabbit anti-serum againstUCP1) detected an antigen of UCP1protein and approximately 350 bp length DNA fragment encoding specific protein UCP1 was found in lamb genome by PCR. KEY WORDS: Brown fat, thermogenin, UCP1, multilocular adipocyte, unilocular adipocyte KEY WORDS: Brown fat, thermogenin, UCP1, multilocular adipocyte, unilocular adipocyte INTRODUCTION (hypothermia)(S.John Martin, 2010) and then the body temperature becomes constant due to activation of specific mechanism of thermal regulation for a shorter period. Thus the essence of such stabilization of the body temperature for a shorter period is associated with that animal body heat production process is different from mature animals. A factor for essential and specific processes brown adipose tissue (BAT) is brown fat (Brown fat tissue). (B.Cannon, J.Nedergaard, 2003, J.C.Rauch, 1969, G.Purevdorj 2005, M.E.Symons 1991, S.Àndrei 1991, G.Àmgàlànbààtàr 2004) Mongolian livestock has reproductive feature that they directly produce their offspring on the pasture during the spring season, when average air temperature is minimal and wind speed is maximal. Therefore, the body temperature of neonates, as well as keeping its homeostasis at the normal level requires thermal exchange and specific factor of thermal regulation, when the fetus shifts from constantly warm environment of maternal body to severely cool external environment. Not only offspring of various mammals, but also human infants face the necessity of withstanding above mentioned natural challenges and during first moments of the neonatal period the body temperature drops Mongolian livestock has reproductive feature that they directly produce their offspring on the pasture during the spring season, when average air temperature is minimal and wind speed is maximal. Therefore, the body temperature of neonates, as well as keeping its homeostasis at the normal level requires thermal exchange and specific factor of thermal regulation, when the fetus shifts from constantly warm environment of maternal body to severely cool external environment. Not only offspring of various mammals, but also human infants face the necessity of withstanding above mentioned natural challenges and during first moments of the neonatal period the body temperature drops Localization of brown fat differs with various animals as reported by many authors. For KhorolmaaCh, Demberel Sh, Mongolian Journal of Agricultural Sciences №13 (02), 2014:8-12 9 thyroid and adrenal glands are also of greater significance (Toyoshi Endo and Tetsuro Kobayashi, 2008, Liu X, Li Q, Lin Q, Sun R., 2001, Ìostovoi À.V. , IvànîvS.L.,2004, Alison Sharpe Avram, 2005). MATERIALSAND METHODS methods, the antigen of UCP1 protein was detected by the immunohistochemical method using an antiserum to UCP1, and the gene of cell mitochondrial thermogenin (UCP1), was determined by PCR using specific primers of UCP1 (forward- AGATCTCAGCGGGCCTAAC,reverse- GGCCTCCTTCATTAGGTCAT). The thoracic, pericardial, renal and flank BAT for our study were prepared by sacrificing and dissecting 12 fetuses obtained from 4-5 months pregnant ewes via Cesarean section, and 0, 3, 7 and 14-day old 6 healthy lambs. The thoracic, pericardial, renal and flank BAT for our study were prepared by sacrificing and dissecting 12 fetuses obtained from 4-5 months pregnant ewes via Cesarean section, and 0, 3, 7 and 14-day old 6 healthy lambs. A histological structure of BATwas determined according to conventional INTRODUCTION example, normally grown calf brown fat is deposited in the perirenal region and it accounts for 27% of its total brown fat and 17.5% of subscapular brown fat, as well as significantly greater depositions are found in thoracic and abdominal cavities, along aorta, pericardium and axillary regions (PaivaiSoppela, 1986), and according to the study performed on mouse Peromyscusmanicularis, it is broadly scattered on interscapular, subscapular, axillary, perirenal, pericardial and intercostals regions (J.C.Rauch, 1969). A specific protein called thermogenin (UCP1) was found in the inner membrane of cellular mitochondria plays a role of absorbing energy due to oxidation into blood in the form of heat energy (B.Cannon, J.Nedergaard, Abraham L, Alison Sharpe Avram). A specific protein called thermogenin (UCP1) was found in the inner membrane of cellular mitochondria plays a role of absorbing energy due to oxidation into blood in the form of heat energy (B.Cannon, J.Nedergaard, Abraham L, Alison Sharpe Avram). In other words majority of chemical energy produced in BAT can be converted into thermal energy (B.Cannon, 2004, Cousin B., 1992, D.Enebish, 2003). Thus it is most important quality of BAT function and therefore this topic has been attracting greater interests in the field of both human health and veterinary sciences. Once the parturition ends, a thermoregulatory center in the central nervous system is triggered due to effect of lower temperature of external environment, main physiological processes leading to heat production in animal body is stimulated, functions of BAT is intensified for a shorter period, and specific process of heat production takes place there. Maintenance of the body temperature at the normal level via stimulation of BAT functions results from such morphological characteristics as its innervations, blood circulations and localizations, as well as physiological factors including noradrenaline released at the sympathetic nerve ending, STH, TTH and ACTH of pituitary gland and hormones of Within the framework of our research on structure and development of brown fat in the body of Mongolian lamb being adapted to such severe changes of temperature, the present study aimed to detect a specific histological structure of BAT, the gene of a specific protein (UCP1) in brown fat in the genome, and the antigen to UCP1 in brown fat during fetal development and early neonatal periods. RESULTS connective-muscular tissue septa, cell sizes were not uniform (Picture 1), multilocular cells with smaller vacuoles evidencing cytoplasmic lipid droplets accounts for major part of adipocytes, whereas unilocular cells with larger vacuoles for a minor part. As well, presence of abundant blood vessels and Hematoxilin and eosin (HE) staining and microscopy of micro preparations of both thoracic superficial and deep, thoracic cavity, renal and flank brown fats obtained from fetuses and 0 to 3-day old neonatal lambs demonstrate that there are numerous lobules separated by connective tissue, in some cases KhorolmaaCh, Demberel Sh, Mongolian Journal of Agricultural Sciences №13 (02), 2014:8-12 10 great number of intercellular capillaries makes brown fat very different from white fat. in its cytoplasm (Picture 1). In other words, there is a principle that brown fat is becoming typical to be white fat cells since the lamb is 7-day old. During the microstructural study, it is also observed that color of brown fat is changing into bright at this age. Size of brown fat cells on above localizations in 7-day old lambs was slightly larger than those in 0 to 3-day old lambs, as well as fat cells as a whole are unilocular a type of white blood cells containing only large lipid droplet Picture 1.A histological structure of fetal BAT A.Fetus thoracic, B.Fetus flank, C.0-day old lamb pericardium, D.0-day old lamb perirenal, E.7-day old thoracic, F.7-day old perirenalbrown fat Picture 1.A histological structure of fetal BAT A.Fetus thoracic, B.Fetus flank, C.0-day old lamb pericardium, D.0-day old lamb perirenal, E.7-day old thoracic, F.7-day old perirenalbrown fat The BAT cell size is relatively smaller from fetal development period to 3-day old age and cellular diameter is approximately 18 to 25 µm in average. According to our measurements, BAT cells in thoracic and flank regions are slightly smaller than brown fat tissue cells in both thoracic and abdominal cavities or approximately 18 to 22 µm. From 7-day old age, the cell diameter becomes 25 to 30 µm, there was an observed principle that cells were slightly enlarged with typical changes (Table 1). RESULTS Table 1 Table 1 Micro-measurements of brown fat tissue ¹ Localization of brown fat Cell size with ages(µm) Size of lipid droplet (µm) Fetus and 0 to 3-day old lambs 7-day old lambs Fetus and 0 to 3-day old lambs 7-day old 1 Thoracic brown fat 18-22 20-35 3-10 20-25 2 Flank brown fat 22 3 Pericardium brown fat 20-25 4 Perirenal brown fat 20-25 Micro-measurements of brown fat tissue Micro-measurements of brown fat tissue KhorolmaaCh, Demberel Sh, Mongolian Journal of Agricultural Sciences №13 (02), 2014:8-12 11 Picture 2. Result of PCR A - 1.Ìàrker /100bp/, 2,3. Fetal thoracic brown fat, 5,6.Thoracic BAT of neonate /350bp/ B - 1. Ìàrker /500bp/, 2.Perirenal BAT of lamb, 5.Periscrotum BAT of lamb /350bp/ Picture 2. Result of PCR A - 1.Ìàrker /100bp/, 2,3. Fetal thoracic brown fat, 5,6.Thoracic BAT of neonate /350bp/ B - 1. Ìàrker /500bp/, 2.Perirenal BAT of lamb, 5.Periscrotum BAT of lamb /350bp/ DNA was extracted from brown fat of Mongolian lambs by using a phenol- chloroform method, then polymerase chain reaction was performed using specific primers of sheep UCP1 protein (forward- AGATCTCAGCGGGCCTAAC, reverse encoding specific protein UCP1 was detected in the genome.(Picture2) Immunohistochemical analysis by using arabbit anti-sheep UCP1serum in brown fat of lambs thorax, scrotum, pericardium and kidneys revealed an antigen to UCP1 was detected. (Picture 3) GGCCTCCTTCATTAGGTCAT), and a DNA fragment with approximately 350 bp length Picture 3.Results of immunohistochemical analysis A.Thoracic brown fat, B.Scrotum brown fat, C.Perirenal brown fat, D.Pericardium brown fat, E.Positive control, F.Negative control Picture 3.Results of immunohistochemical analysis A.Thoracic brown fat, B.Scrotum brown fat, C.Perirenal brown fat, D.Pericardium brown fat, E.Positive control, F.Negative control CONCLUSION was detected in the genome of Mongolian sheep lambs. 1. For histological structures, brown fat tissues are divided into lobules separated with connective tissue septa, it consists of 18 to 30 µm multilocular type cells with multiple small lipid droplets in their own cytoplasm, and capillary blood vessels are abundant between them. 3. 3. An antigen to UCP1 was detected by immunohistochemical analysis using rabbit anti-sheep UCP1 antiserum in brown fat of lambs thorax, scrotum, pericardium and kidneys. 2. DNA fragment with approximately 350 bplength encoding specific protein UCP1 2. DNA fragment with approximately 350 bplength encoding specific protein UCP1 DISCUSSIONS for the minor part. As well, presence of abundant blood vessels and great number of intercellular capillaries makes brown fat very different from white fat(B.Cannon and J.Nederhaard, 2004, Abraham L, 2002, À.Ham, D.Êîrmàê, 1983).Typical changes of fat cells since 7-day old age of lambs and enlargement of the cell size is also consistent with results of studies by other researchers.(Andrei S 1991, G.Purevdorj 2005) Simple staining of micropreparation of brown fat of both fetuses and newborn lambs revealed the tissue was divided with connective tissue, in some cases with connective-muscular tissue septa into multiple lobules (Pictures 1, 2 and 3) and our results are in agreement with that multilocular cells with smaller vacuoles evidencing cytoplasmic lipid droplets accounts for the major part of adipocytes, whereas unilocular cells with large vacuoles Simple staining of micropreparation of brown fat of both fetuses and newborn lambs revealed the tissue was divided with connective tissue, in some cases with connective-muscular tissue septa into multiple lobules (Pictures 1, 2 and 3) and our results are in agreement with that multilocular cells with smaller vacuoles evidencing cytoplasmic lipid droplets accounts for the major part of adipocytes, whereas unilocular cells with large vacuoles KhorolmaaCh, Demberel Sh, Mongolian Journal of Agricultural Sciences №13 (02), 2014:8-12 12 In the present study, it was found that the genome of Mongolian lambs contains information on UCP1 protein and detection of the antigen to USP1 protein in brown fat in the localizations described by us results in creation of theoretical background that development of the brown fat, synthesis of its specific protein and their functions can be greater than the same species of animals under different ecological conditions and other non-hibernating animal species depending on our country’s climate and specific ecological factors. REFERENCES 10. Thyroid-stimulating hormone receptor in brown adipose tissue is involved in the regulation of thermogenesis, Toyoshi Endo and Tetsuro Kobayashi, Am J PhysiolEndocrinolMetab 295: E514– E518, 2008. 1. À.Ham, D.Êîrmàê, Histology, 1983 1. À.Ham, D.Êîrmàê, Histology, 1983 2. “Brown Adipose Tissue: Function and Physiological Significance”, B.Cannon and J.Nederhaard, Physiology Rev., 2004 3. Histology and Cell Biology an Introduction to Pathology, Abraham L, 2002 11. Uncoupling protein1 mRNA, mitochondrial GTP-binding, and T4 5'- deiodinase of brown adipose tissue in euthermicDaurian ground squirrel during cold exposure. Liu X, Li Q, Lin Q, Sun R., Comp BiochemPhysiol A MolIntegr Physiol. 2001 Apr;128(4):827-35 11. Uncoupling protein1 mRNA, mitochondrial GTP-binding, and T4 5'- deiodinase of brown adipose tissue in euthermicDaurian ground squirrel during cold exposure. Liu X, Li Q, Lin Q, Sun R., Comp BiochemPhysiol A MolIntegr Physiol. 2001 Apr;128(4):827-35 4. S.John Martin, Hyphotermia in new born lambs, 2010 5. Topography and vascularization of brown fat in small non-hibernator, J.C.Rauch and J.S.Hayaward, Can. J. Zool., 1969, 1301-1314 12. Ìîstîvîi À.V., IvànîvS.L., Features of thermoregulations in newborn animals,Minsitry of healthcare, RF. Department of anaestesiology and reanimation, Sanct-Peterburg State Academy of medicine, 2004 6. G.Purevdorj, Structure and chemical compositions of brown fat of both humans and animals, 2005 7. Andrei S., Brown adipose tissue of animals, J. Agriculture, 1991, ¹2, p 26 g 8. Thyroid-stimulating hormone receptor in brown adipose tissue is involved in the regulation of thermogenesis, Toyoshi Endo and Tetsuro Kobayashi, Am J PhysiolEndocrinolMetab295: E514– E518, 2008. 13. Anatomy and physiology of white and brown adipose tissue, Alison Sharpe Avram, MD,a Mathew M. Avram, MD, JD,b and William D. James, MDc, J. Aì.Acad.Dermatol. volume 53, No 4, 11/2005 9. Anatomy of brown adipose tissue in neonate reindeer, Paivi Soppela, Raija Sormunen and Mauri Nieminen, Rangifer, 1986, No. 1, 116 14. D. Enebish, About mechanism of thermal production in brown fat, J.Diagnosis, 2003
https://openalex.org/W2940578773	https://europepmc.org/articles/pmc6357388?pdf=render	English	null	The association of tumor necrosis factor superfamily 13 with recurrence of immunoglobulin A nephropathy in living related kidney transplantation	BMC nephrology	2,019	cc-by	5,155	© The Author(s). 2019 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. Open Access The association of tumor necrosis factor superfamily 13 with recurrence of immunoglobulin A nephropathy in living related kidney transplantation Hyung Ah Jo1,2, Seung Seok Han1,3, Sunhwa Lee1,3, Joo Young Kim3, Seung Hee Yang3, Hajeong Lee1,3, Jae Seok Yang4,5, Jung Pyo Lee1,3,6, Kwon Wook Joo1,3, Chun Soo Lim1,6, Yon Su Kim1,3, Curie Ahn1,4, Jin Suk Han1,3 and Dong Ki Kim1,3* Abstract Background: An increasing amount of evidence has demonstrated an association between an increase in the level of tumor necrosis factor superfamily 13 (TNFSF13) and immunoglobulin A nephropathy (IgAN) progression. We aimed to evaluate if the level of pre-transplant serum TNFSF13 is predictive of IgAN recurrence after kidney transplantation. Methods: This analysis was based on the clinical and laboratory data of 69 patients with IgAN who underwent first kidney transplantation with no evidence of mesangial IgA deposits in zero-time transplantation biopsy. We measured pre-transplant serum TNFSF13, total IgA, and galactose-deficient IgA1 levels. Results: The recurrence rate of IgAN over a median follow-up duration of 5.1 years was 15.9% (11/69 patients), with a mean time to the first recurrence of 1.7 years. The high pre-transplant TNFSF13 level was associated with IgAN recurrence after kidney transplantation among patients who received a graft from a living related donor. Conclusions: This study highlights association of TNFSF13 levels in recurrent IgAN patients who undergo living related donor transplantation Further research is needed to clarify mechanisms by which TNFSF13 Results: The recurrence rate of IgAN over a median follow-up duration of 5.1 years was 15.9% (11/69 patients), with a mean time to the first recurrence of 1.7 years. The high pre-transplant TNFSF13 level was associated with IgAN recurrence after kidney transplantation among patients who received a graft from a living related donor. Conclusions: This study highlights association of TNFSF13 levels in recurrent IgAN patients who undergo living related donor transplantation. Further research is needed to clarify mechanisms by which TNFSF13 affects the recurrence of IgA nephropathy. Keywords: IgA nephropathy, Tumor necrosis factor superfamily 13, Recurrence RESEARCH ARTICLE Open Access The association of tumor necrosis factor superfamily 13 with recurrence of immunoglobulin A nephropathy in living related kidney transplantation Hyung Ah Jo1,2, Seung Seok Han1,3, Sunhwa Lee1,3, Joo Young Kim3, Seung Hee Yang3, Hajeong Lee1,3, Jae Seok Yang4,5, Jung Pyo Lee1,3,6, Kwon Wook Joo1,3, Chun Soo Lim1,6, Yon Su Kim1,3, Curie Ahn1,4, Jin Suk Han1,3 and Dong Ki Kim1,3* Jo et al. BMC Nephrology (2019) 20:33 https://doi.org/10.1186/s12882-019-1222-4 RESEARCH ARTICLE Open Access The association of tumor necrosis factor superfamily 13 with recurrence of immunoglobulin A nephropathy in living related kidney transplantation Hyung Ah Jo1,2, Seung Seok Han1,3, Sunhwa Lee1,3, Joo Young Kim3, Seung Hee Yang3, Hajeong Lee1,3, Jae Seok Yang4,5, Jung Pyo Lee1,3,6, Kwon Wook Joo1,3, Chun Soo Lim1,6, Yon Su Kim1,3, Curie Ahn1,4, Jin Suk Han1,3 and Dong Ki Kim1,3* Jo et al. BMC Nephrology (2019) 20:33 https://doi.org/10.1186/s12882-019-1222-4 Jo et al. BMC Nephrology (2019) 20:33 https://doi.org/10.1186/s12882-019-1222-4 Background follow-up duration and biopsy policies across institutions [4, 5]. While the impact of the recurrence of IgA deposits on graft function has been debated, recurrent IgAN has been associated with unfavorable outcomes in long-term follow-up studies [2, 6–8]. Unfortunately, however, there is no biomarker that can predict IgAN recurrence in these patients. Immunoglobulin A nephropathy (IgAN) is the most com- mon type of primary glomerulonephritis [1, 2], and a sig- nificant number of patients progress to end-stage renal disease (ESRD) requiring dialysis or kidney transplantation [3]. In cases of transplant recipients, recurrence of primary glomerulonephritis commonly affects graft and patient outcomes [2]. The recurrence rate of IgAN after kidney transplantation has been estimated to be 9 to 61%, with this wide variability likely to reflect differences in Tumor necrosis factor superfamily 13 (TNFSF13), also known as a proliferation inducing ligand (APRIL), is se- creted from antigen-presenting cells and is engaged in adaptive and innate immune responses by binding to re- ceptors on B and T cells [9, 10]. An in vivo analysis using TNFSF13 gene knockout mice showed impaired IgA anti- body responses to mucosal immunization [11]. Con- versely, TNFSF13 transgenic mice showed enhanced T-cell independent humoral responses [12]. The TNFSF13 * Correspondence: dkkim73@gmail.com 1Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea 3Kidney Research Institute, Seoul National University, Seoul, Republic of Korea Full list of author information is available at the end of the article Statistical analysis Comparisons were performed between patients with and without recurrent IgAN after kidney transplantation. The data are presented as the mean ± standard deviation (S.D) for continuous variables and as proportions for categorical variables. Differences between recurrent and non-recurrent group were evaluated using the t-test for normally distrib- uted continuous variables and Mann-Whitney U test for non-normally distributed continuous variables. The chi-squared or Fisher’s exact test was used for categorical variables. A Cox regression model was used to calculate the unadjusted and adjusted hazard ratios (HRs) and the 95% confidence intervals (CIs) for the factors that were as- sociated with IgAN recurrence after kidney transplant- ation. The interaction terms between TNFSF13 and adjusted variables were confirmed by Cox analysis. A p-value < 0.05 was considered statistically significant. All statistical analyses were performed using IBM®SPSS® soft- ware, version 21.0 (IBM Corporation, Armonk, NY, USA). Page 2 of 7 Jo et al. BMC Nephrology (2019) 20:33 Jo et al. BMC Nephrology (2019) 20:33 Page 2 of 7 Jo et al. BMC Nephrology (2019) 20:33 CA, USA), based on previously published methods [14]. Serum Gd-IgA1 levels were quantified using a lectin ELISA with Helix promatia agglutinin. Serum Gd-IgA1 levels were expressed in U/mL, where 1 U of Gd-IgA1 was defined as 10.0 μg of standard, and 1 U/mL was expressed as 1 U/ng IgA after normalization to total IgA levels. We purified Gd-IgA1 from an IgAN patient’s plasma using immobilized Jacalin (Thermo Scientific, Rockford, IL, USA) for the Gd-IgA1 method, as previously reported [14]. CA, USA), based on previously published methods [14]. Serum Gd-IgA1 levels were quantified using a lectin ELISA with Helix promatia agglutinin. Serum Gd-IgA1 levels were expressed in U/mL, where 1 U of Gd-IgA1 was defined as 10.0 μg of standard, and 1 U/mL was expressed as 1 U/ng IgA after normalization to total IgA levels. We purified Gd-IgA1 from an IgAN patient’s plasma using immobilized Jacalin (Thermo Scientific, Rockford, IL, USA) for the Gd-IgA1 method, as previously reported [14]. locus has recently been identified as a susceptibility gene in a Han Chinese genome-wide association study of IgAN patients [13]. Furthermore, high serum TNFSF13 levels in IgAN patients could predict the progression of renal dis- ease based on B cell stimulation [14]. The high recurrence rates after kidney transplantation in IgAN patients sug- gests an impaired host IgA immune system. Therefore, TNFSF13, which is known to be involved in B cell and im- munoglobulin A immune system function, could affect re- current IgAN. The objective of this study was to find an association between pre-transplant serum TNFSF13 levels and recurrent IgAN. Patients and data collection Between January 2011 and October 2015, 80 patients with biopsy-proven IgAN underwent kidney transplant- ation at our institution. Among these patients, we en- rolled 69 who underwent first kidney transplantation and did not show mesangial IgA deposits on biopsy at the time of kidney transplantation and have allograft bi- opsy data within one year after transplantation. The fol- lowing information was extracted from medical records for analysis: age, sex, renal replacement therapy, the donor age and sex, number of human leukocyte antigen (HLA) mismatches, ABO mismatch, donor type, desensitization treatment, induction treatments adminis- tered, withdrawal of steroid treatment and acute rejec- tion events and trough level of tacrolimus within one year after kidney transplantation. Highly sensitized pa- tients defined as having panel reactive antibody higher than 50% [15]. Allograft failure was defined by the need to re-initiate dialysis or when the estimated glomerular filtration rates (eGFR) fell below 15 mL/min/1.73 m2. Clinicopathologic recurrence of IgAN after kidney trans- plantation was defined as a clinically evident disease with deteriorating renal function and identification of mesangial IgA deposits on immunofluorescence staining with mesangial hypercellularity in allograft biopsy tissue [16]. Renal function decline was defined as eGFR-time slope, differences in eGFR between the peak value in the post-transplant period within 2 months after kidney transplantation and the nadir value at the time of last follow up time divided by follow up duration. Baseline characteristics and clinical outcomes Baseline characteristics and clinical outcomes This study included 40 men (58.0%) and 29 women who underwent first kidney transplantation for ESRD due to biopsy-proven IgAN. At the time of transplantation, their mean age was 43.7 years (Table 1). Over a median follow up of 1865 days (mean, 1835 days), there were no mortalities, and one case of allograft failure caused by the patient discontinuing the immunosuppressive ther- apy was observed. Recurrent IgAN were identified in 11 patients (15.9%). Among these cases, 3 received de- ceased donor grafts, 1 received an unrelated living donor graft and the other 7 received related living donor grafts. There was no difference between the recurrent and non-recurrent groups in the proportions of patients treated with tacrolimus, cyclosporine or mycophenolate mofetil. Also, the proportion of patients who received a desensitization treatment and highly sensitized patients who had panel reactive activity over 50% were not differ- ent between the groups. The rate of acute rejection was marginally higher among patients with than without IgAN recurrence (p = 0.056). Pre-transplant serum TNFSF13 levels in the study population were signifi- cantly higher than in the 382 patients with non-ESRD Laboratory tests BMC Nephrology (2019) 20:33 Page 3 of 7 Table 1 Baseline characteristics of the study subjects All Recurrence Non-recurrence p-value Patients, n 69 11 58 – Age, years 43.7 ± 13.4 43.4 ± 13.5 43.8 ± 13.5 0.923 Sex (male/female) 40/29 10/1 28/30 0.019 Dialysis/Preemptive kidney transplantation 53/16 9/2 44/14 1.000 Living related donor 36 (52.2) 7 (63.6) 29 (50.0) 0.406 Panel reactive activity > 50% 10 (14.5) 0 (0) 10 (17.2) 0.345 Desensitization treatment 13 (18.8) 1 (9.1) 12 (20.7) 0.676 Number of HLA mismatches 2.88 ± 1.84 2.45 ± 1.97 2.97 ± 1.82 0.354 ABO incompatible transplantation 10 (14.5) 1 (9.1) 9 (15.5) 1.000 Induction treatment (ATG/anti-IL-2) 0/68 0/10 0/58 0.159 Mean follow up duration after transplantation (days) 1835 ± 440 1782 ± 542 1845 ± 423 0.663 Time to transplantation from dialysis (days, IQR) 1387 ± 1611 (84–2109) 1514 ± 1477 (219–2516) 1360 ± 1652 (83–2133) 0.568 Acute rejection 32 (46.4) 8 (72.7) 24 (41.4) 0.056 Mean time of recurrence of IgA on allograft (days, IQR) – 614 ± 443 (390–726) – – Serum creatinine (at time of recurrence, mg/dL) 2.96 ± 1.52 eGFR-time slope (delta eGFR/year) −3.64 ± 4.88 −4.50 ± 3.99 −3.47 ± 5.04 0.302 HLA-A2 30 (43.5) 6 (54.5) 24 (41.4) 0.514 HLA-B35 12 (17.4) 2 (18.2) 10 (17.2) 1.000 HLA-DR4 40 (58.0) 9 (81.8) 31 (53.4) 0.104 Tacrolimus 66 (95.7) 10 (90.9) 56 (96.6) 0.411 Cyclosporine 7 (10.1) 2 (18.2) 5 (8.6) 0.309 Mycophenolate mofetil 53 (76.8) 7 (63.6) 46 (79.3) 0.265 Trough level of tacrolimus within one year from transplantation 8.30 ± 1.70 8.92 ± 1.38 8.19 ± 1.74 0.218 TNFSF13 (ng/mL) 16.03 ± 12.95 17.75 ± 14.14 15.66 ± 12.79 0.525 Gd-IgA1 (units/ng IgA) 5.77 ± 6.79 3.28 ± 1.43 6.35 ± 7.40 0.208 HLA human leucocyte antigen, ATG anti-thymocyte globulin, IL-2 interleukin-2, IQR interquartile range, eGFR estimated glomerular filtration rate, TNFSF13 Tumor necrosis factor superfamily 13, Gd-IgA1 Galactose deficient-IgA1 Table 1 Baseline characteristics of the study subjects HLA human leucocyte antigen, ATG anti-thymocyte globulin, IL-2 interleukin-2, IQR interquartile range, eGFR estimated glomerular filtration rate, TNFSF13 Tumor necrosis factor superfamily 13, Gd-IgA1 Galactose deficient-IgA1 IgAN (mean 1.75 ± 11.94 ng/mL) (Fig. 1). However, the levels were not different between the recurrent (mean 17.75 ± 14.14 ng/mL) and non-recurrent (mean 15.66 ± 12.79 ng/mL) groups (Fig. 1). Laboratory tests Specific HLA types (A2, B35, and DR4) and pre-transplant serum levels of TNFSF13 and galactose deficient IgA1 (Gd-IgA1) were equivalents between groups. The eGFR-time slope was steeper in the recurrent than in the non-recurrent group, although this difference was not statistically significant. was a significant interaction between pre-transplant serum TNFSF13 levels with living related transplantation for re- current IgAN (p = 0.018). Pre-transplant TNFSF13 levels were associated with an increased risk of IgAN recurrence among patients who received a graft from a living related donor after adjusting the recipient’s age and follow up duration (Table 3, HR, 1.685; p = 0.025). Patients who re- ceived a graft from a living related donor were dichoto- mized based on the median pre-transplant serum TNFSF13 level (9.97 ng/mL). Recurrence-free survival was significantly lower among patients with a pre-transplant serum TNFSF13 level greater than median values (Fig. 2). Laboratory tests Of the 69 IgAN patients initially enrolled in the study, serum samples at the time of kidney transplantation to quantify pre-transplant TNFSF13 levels were obtained from 62 patients. We also recruited the serum samples from non-ESRD IgAN patients who had eGFR > 30 mL/min/1.73 m2. The samples were frozen at −80 °C until analysis. Serum levels TNFSF13 were quantified using an enzyme-linked im- munosorbent assay (ELISA) kit (eBioscience, San Diego, Jo et al. Discussion We did identify a specific association between pre-trans- plant serum TNFSF13 levels with IgAN recurrence among patients who received living related donor grafts in an interaction analysis (Table 2), though not significant all transplant patients (adjusted HR, 1.111; p = 0.620). There Several factors, including genetic and environmental fac- tors, contribute to the pathogenesis of IgAN. Among the several factors involved in the pathogenesis of IgAN, ab- normal mucosal immunity is an important factor for Jo et al. BMC Nephrology (2019) 20:33 Page 4 of 7 Fig. 1 Comparison of serum TNFSF13 levels in patients with recurrent, non-recurrent immunoglobulin A nephropathy and patients with immunoglobulin A nephropathy not on dialysis. p = 0.000, n.s: non-significant Fig. 1 Comparison of serum TNFSF13 levels in patients with recurrent, non-recurrent immunoglobulin A nephropathy and patients with immunoglobulin A nephropathy not on dialysis. p = 0.000, n.s: non-significant triggering IgAN. As previously noted, TNFSF13 is known to be involved in abnormal mucosal immunity via T-cell independent production of IgA- producing B cells and IgA1 to IgA2 class switching [9, 17–20]. TNFSF13 levels can affect the etiology of recurrent IgAN, based on findings of various studies about the ef- fect of TNFSF13 on the IgA mucosal immune system [11, 14]. It is well known that a longer follow-up period after transplantation and younger age of a recipient at transplantation increase the risk of IgAN recurrence [2]. A previous study reported a high recurrence rate of IgAN after kidney transplantation among recipients with zero-HLA mismatches [21]. In our study, we identified that a higher pre-transplant serum level of TNFSF13 was associated with IgAN recurrence on multivariate analysis among patients who received a graft from a liv- ing related donor after adjusting for the recipient’s age at transplantation, the length of follow up duration. We could not elucidate why there was an association between the levels of TNFSF13 on the recurrent IgAN only in living related donor graft but not in entire sub- jects. Although the mechanism of increased recurrent IgAN in living related grafts is unclear, several explana- tions could be possible. First, TNFSF13 shown to be in- creased the overall IgA secretion, although not specific to Gd-IgA1 [14]. However, unfortunately, we could not confirm that TNFSF13 would affect recurrent IgAN via increasing the relative amount of Gd-IgA1 because there were no differences in the level of Gd-IgA1 between the groups. Discussion Second, the pathogenesis of IgAN requires Table 2 An interaction analysis between pre-transplant serum TNFSF13 levels and adjusted variables for recurrent IgA nephropathy Variables Adjusted HR 95% CI p-value for interaction Recipient age ≥50 years 0.981 0.915–1.052 0.597 Donor age ≥50 years 1.017 0.976–1.060 0.427 HLA full matches 1.054 0.960–1.158 0.266 ABO mismatch 0.985 0.830–1.170 0.864 Follow up duration ≥2000 days 1.030 0.993–1.069 0.115 Gd-IgA1 (units/ng IgA) 0.995 0.985–1.006 0.370 Living related donor 1.046 1.008–1.085 0.018 HR: hazard ratio, 95% CI: confidence interval. The hazard ratio was expressed as 1 units of Gd-IgA1/ng IgA increase Jo et al. BMC Nephrology (2019) 20:33 Page 5 of 7 Table 3 Hazard ratios for recurrent immunoglobulin A nephropathy in patients who underwent kidney transplantation with living- related donor grafts (n = 36) Variables Unadjusted HR 95% CI p-value Adjusted HR 95% CI p-value Recipient age 1.003 0.952–1.057 0.902 1.006 0.947–1.068 0.846 Donor age 0.988 0.924–1.056 0.718 HLA full matches 2.037 0.394–10.541 0.396 ABO mismatch 0.554 0.066–4.625 0.586 Follow up duration 0.999 0.997–1.001 0.332 0.999 0.997–1.001 0.198 Gd-IgA1 (units/ng IgA) 0.881 0.644–1.205 0.427 TNFSF 13 (10 ng/mL) 1.585 1.030–2.439 0.036 1.685 1.068–2.658 0.025 HR: hazard ratio, 95% CI: confidence interval. The hazard ratio was expressed as 1 units of Gd-IgA1/ng IgA increase, 10 ng/mL TNFSF13 increase Table 3 Hazard ratios for recurrent immunoglobulin A nephropathy in patients who underwent kidney transplantation with living- related donor grafts (n = 36) several hits, and we carefully assume there was the role of TNFSF13 in the production of anti-glycan antibody. It is essential for the generation of anti-glycan IgG due to a defect of allorecognition to Gd-IgA1 in the pathogenesis of IgAN [22]. A recent study has shown that TNFSF13 also involved in adaptive immunity through influencing on memory T cells [23]. We could not measure anti-glycan antibodies in the present study; however, we carefully sug- gest that TNFSF13 would affect recurrent IgAN by en- gaging in alloantibody response. Third, the similar genetic background in the living related grafts with recipients would have affected the recurrent IgAN [24]. It can be in- ferred that the mesangial immune complex deposition of living related grafts is more likely to occur [25, 26]. TNFSF13 may have upregulated the production of sys- temic IgA or pathognomonic IgA1-immune complex, and these factors may be profound in living related donor grafts having the genetic susceptibility. Discussion This issue would be re- solved by the further experimental investigation. study, we observed only one event of graft failure due to the self-discontinuing steroid. Therefore, we did not find an association between serum TNFSF13 levels and graft outcome. Additionally, we did not identify an HLA-specific effect, as reported in previous studies, including a negative effect for HLA-B35 and DR4 and a protective effect for HLA-A2 [6, 29]. As the majority of the patients in our study group received anti-interleukin 2 for induction ther- apy, we were unable to demonstrate the protective role of antithymocyte globulin induction therapy on IgAN recur- rence, as reported in a previous study [30]. According to Australia and New Zealand Dialysis and Transplant (ANZDATA) registry, which has accumu- lated data over 30 years, showed that recurrence of IgAN was 5.1, 15% at 5, 15 years after kidney transplantation [2]. The rate of IgAN recurrence in our study was 15.9% over a median follow up duration of 5.1 years, a rate which was higher than the rate based on the ANZDATA registry data but lower than recurrence rate of 29–61% reported in other studies [4, 29]. Differences in reported rates of IgAN recurrence might result from differences in biopsy policy between institutions and differences in the length of follow up. Further prospective studies with Although IgAN was traditionally considered as a benign disease, more recent evidence has demonstrated that re- currence of IgAN in renal allografts is associated with un- favorable outcomes in graft function [2, 8, 27, 28]. In this Fig. 2 Kaplan-Meier survival curves of recurrence after kidney transplantation in immunoglobulin A nephropathy patients who received living related donor grafts according to the median levels of pre-transplant serum TNFSF13 Fig. 2 Kaplan-Meier survival curves of recurrence after kidney transplantation in immunoglobulin A nephropathy patients who received living related donor grafts according to the median levels of pre-transplant serum TNFSF13 Jo et al. BMC Nephrology (2019) 20:33 Page 6 of 7 Page 6 of 7 Page 6 of 7 sufficient follow-up duration are needed to clarify the as- sociation between pre-transplant serum TNFSF13 levels and the recurrence rate and graft functionality in IgAN patients. Hospital (No. H 1605–090-762). Serum samples were provided by the kidney transplantation biorepository (H-1102-082-353) and Seoul National University Human Biobank (H-1004-037-315). We obtained informed written consent to use their samples from all participants. Competing interests Competing interests Competing interests The authors declare that they have no competing interests. Author details 1 In summary, we identified an association between pre-transplant serum TNFSF13 levels and IgAN recurrence among patients who received a graft from a living related donor. Our findings suggest that there may be a possible role of TNFSF13 in IgAN recurrence among patients who undergo living related donor graft transplantation. 1Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea. 2Department of Internal Medicine, Inje University Ilsan Paik Hospital, Ilsan, Republic of Korea. 3Kidney Research Institute, Seoul National University, Seoul, Republic of Korea. 4Transplantation Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea. 5Department of Surgery, Transplantation Center, Seoul National University Hospital, Seoul, Republic of Korea. 6Department of Internal Medicine, Seoul National University Boramae Medical Center, Seoul, Republic of Korea. 1Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea. 2Department of Internal Medicine, Inje University Ilsan Paik Hospital, Ilsan, Republic of Korea. 3Kidney Research Institute, Seoul National University, Seoul, Republic of Korea. 4Transplantation Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea. 5Department of Surgery, Transplantation Center, Seoul l l S l R bl f 6D f Availability of data and materials The dataset used during the current study is available from the corresponding author upon request. 10. He B, Xu W, Santini PA, Polydorides AD, Chiu A, Estrella J, et al. Intestinal bacteria trigger T cell-independent immunoglobulin a 2 class switching by inducing epithelial-cell secretion of the cytokine APRIL. Immunity. 2007; 26(6):812–26. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Funding 7. Nijim S, Vujjini V, Alasfar S, Luo X, Orandi B, Delp C, et al. Recurrent IgA nephropathy after kidney transplantation. Transplant Proc. 2016;48(8):2689–94. 7. Nijim S, Vujjini V, Alasfar S, Luo X, Orandi B, Delp C, et al. Recurrent IgA nephropathy after kidney transplantation. Transplant Proc. 2016;48(8):2689–94. g This study was supported by a grant from the Seoul National University Hospital Research Fund (3020130050). The funder had no role in study design, data collection and analysis, and interpretation of data and in writing the manuscript. 8. Cosio FG, Cattran DC. Recent advances in our understanding of recurrent primary glomerulonephritis after kidney transplantation. Kidney Int. 2016; 91(2):304–14. 8. Cosio FG, Cattran DC. Recent advances in our understanding of recurrent primary glomerulonephritis after kidney transplantation. Kidney Int. 2016; 91(2):304–14. 9. Litinskiy MB, Nardelli B, Hilbert DM, He B, Schaffer A, Casali P, et al. DCs induce CD40-independent immunoglobulin class switching through BLyS and APRIL. Nat Immunol. 2002;3(9):822–9. 9. Litinskiy MB, Nardelli B, Hilbert DM, He B, Schaffer A, Casali P, et al. DCs induce CD40-independent immunoglobulin class switching through BLyS and APRIL. Nat Immunol. 2002;3(9):822–9. Abbreviations anti-IL-2: anti-interleukin 2; APRIL: A proliferation-inducing ligand; ATG: Antithymocyte globulin; eGFR: Estimated glomerular filtration rates; ESRD: End-stage renal disease; Gd-IgA1: Galactose deficient IgA1; HLA: Human leucocyte antigen; HR: Hazard ratio; IgAN: Immunoglobulin A nephropathy; TNFSF13: Tumor necrosis factor superfamily 3. Schena FP. A retrospective analysis of the natural history of primary nephropathy worldwide. Am J Med. 1990;89(2):209–15. 3. Schena FP. A retrospective analysis of the natural history of primary IgA nephropathy worldwide. Am J Med. 1990;89(2):209–15. 4. Ponticelli C, Glassock RJ. Posttransplant recurrence of primary glomerulonephritis. Clin J Am Soc Nephrol. 2010;5(12):2363–72. l k ll k l k 4. Ponticelli C, Glassock RJ. Posttransplant recurrence of primary glomerulonephritis. Clin J Am Soc Nephrol. 2010;5(12):2363–72. 5. Ortiz F, Gelpi R, Koskinen P, Manonelles A, Räisänen-Sokolowski A, Carrera M, et al. IgA nephropathy recurs early in the graft when assessed by protocol biopsy. Nephrol Dial Transplant. 2012;27(6):2553–8. References l k 1. Kiryluk K, Li Y, Sanna-Cherchi S, Rohanizadegan M, Suzuki H, Eitner F, et al. Geographic differences in genetic susceptibility to IgA nephropathy: GWAS replication study and geospatial risk analysis. PLoS Genet. 2012;8(6): e1002765. 2. Allen PJ, Chadban SJ, Craig JC, Lim WH, Allen RD, Clayton PA, et al. Recurrent glomerulonephritis after kidney transplantation: risk factors and allograft outcomes. Kidney Int. 2017;92(2):461–9. Authors’ contributions HAJ, SSH, and DKK participated in the conception of the paper, to the analysis of the data, and the writing of the manuscript. SHL, JYK, SHY, and JSY participated in the data collection and to the writing of the manuscript. HJL, JPL, KWJ, CSL, YSK, CA and JSH participated in the interpretation of the data and in the writing of the manuscript. All the authors have revised the article and approve the final version. 11. Castigli E, Scott S, Dedeoglu F, Bryce P, Jabara H, Bhan AK, et al. Impaired IgA class switching in APRIL-deficient mice. Proc Natl Acad Sci U S A. 2004; 101(11):3903–8. 12. Stein JV, López-Fraga M, Elustondo FA, Carvalho-Pinto CE, Rodríguez D, Gómez-Caro R, et al. APRIL modulates B and T cell immunity. J Clin Invest. 2002;109(12):1587–98. Received: 17 January 2018 Accepted: 18 January 2019 This study provides evidence to suggest that pre-transplant serum TNFSF13 levels is associated with recurrent IgAN who had living related transplantation. Regarding the con- cern of recurrence of IgAN on graft function, these results require further studies to validate the role of TNFSF13 in the recurrence of IgAN. Acknowledgments l Serum samples were provided by the Seoul National University Hospital Human Biobank, a member of the National Biobank of Korea, which is supported by the Ministry of Healthy and Welfare, Republic of Korea. 6. Choy BY, Chan TM, Lo SK, Lo WK, Lai KN. Renal transplantation in patients with primary immunoglobulin a nephropathy. Nephrol Dial Transplant. 2003;18(11):2399–404. 6. Choy BY, Chan TM, Lo SK, Lo WK, Lai KN. Renal transplantation in patients with primary immunoglobulin a nephropathy. Nephrol Dial Transplant. 2003;18(11):2399–404. Discussion We have been approved by IRB to use the stored serum samples with written consent from participants. The present study has several limitations. The number of total subjects and events in the present study was too small to obtain the predictive value of TNFSF13 for re- current IgAN. Relatively shorter follow up duration was also a limitation of the present study. TNFSF13 was only associated with living related transplantation, and the predictive value was not shown. Further studies in the sufficient number of cohorts are needed to confirm these findings and to show cut off values for the predict- ive value of TNFSF13. Consent for publication Not applicable Consent for publication Not applicable 12. Stein JV, López-Fraga M, Elustondo FA, Carvalho-Pinto CE, Rodríguez D, Gómez-Caro R, et al. APRIL modulates B and T cell immunity. J Clin Invest. 2002;109(12):1587–98. Conclusions Received: 17 January 2018 Accepted: 18 January 2019 Ethics approval and consent to participate 13. Yu X-Q, Li M, Zhang H, Low H-Q, Wei X, Wang J-Q, et al. A genome-wide association study in Han Chinese identifies multiple susceptibility loci for IgA nephropathy. Nat Genet. 2012;44(2):178–82. The study protocol complied with the Declaration of Helsinki and was approved by our institutional review board at Seoul National University The study protocol complied with the Declaration of Helsinki and was approved by our institutional review board at Seoul National University Page 7 of 7 Jo et al. BMC Nephrology (2019) 20:33 Jo et al. BMC Nephrology (2019) 20:33 Jo et al. BMC Nephrology (2019) 20:33 14. Han SS, Yang SH, Choi M, Kim H-R, Kim K, Lee S, et al. The role of TNF superfamily member 13 in the progression of IgA nephropathy. J Am Soc Nephrol. 2016;27(11):3430–9. 15. Jordan SC, Tyan D, Stablein D, McIntosh M, Rose S, Vo A, et al. Evaluation of intravenous immunoglobulin as an agent to lower allosensitization and improve transplantation in highly sensitized adult patients with end-stage renal disease: report of the NIH IG02 trial. J Am Soc Nephrol. 2004;15(12): 3256–62. 16. Berthoux F, Suzuki H, Mohey H, Maillard N, Mariat C, Novak J, et al. Prognostic value of serum biomarkers of autoimmunity for recurrence of IgA nephropathy after kidney transplantation. J Am Soc Nephrol. 2017;28(6): 1943–50. 17. Kiryluk K, Novak J. The genetics and immunobiology of IgA nephropathy. J Clin Invest. 2014;124(6):2325–32. 18. Tezuka H, Abe Y, Iwata M, Takeuchi H, Ishikawa H, Matsushita M, et al. Regulation of IgA production by naturally occurring TNF/iNOS-producing dendritic cells. Nature. 2007;448(7156):929–33. 19. Schneider P, Takatsuka H, Wilson A, Mackay F, Tardivel A, Lens S, et al. Maturation of marginal zone and follicular B cells requires B cell activating factor of the tumor necrosis factor family and is independent of B cell maturation antigen. J Exp Med. 2001;194(11):1691–8. 20. Mackay F, Schneider P, Rennert P, Browning J. BAFF AND APRIL: a tutorial on B cell survival. Annu Rev Immunol. 2003;21(1):231–64. 21. McDonald SP, Russ GR. Recurrence of IgA nephropathy among renal allograft recipients from living donors is greater among those with zero HLA mismatches. Transplantation. 2006;82(6):759–62. 22. Kiryluk K, Novak J, Gharavi AG. Pathogenesis of immunoglobulin a nephropathy: recent insight from genetic studies. Annu Rev Med. 2013;64: 339. 23. Gorbacheva V, Ayasoufi K, Fan R, Baldwin WM III, Valujskikh A. Ethics approval and consent to participate B cell activating factor (BAFF) and a proliferation inducing ligand (APRIL) mediate CD40-independent help by memory CD4 T cells. Am J Transplant. 2015; 15(2):346–57. 24. Freese P, Svalander C, Nordén G, Nyberg G. Clinical risk factors for recurrence of IgA nephropathy. Clin Transpl. 1999;13(4):313–7. 25. Brensilver JM, Mallat S, Scholes J, McCabe R. Recurrent IgA nephropathy in living-related donor transplantation: recurrence or transmission of familial disease? Am J Kidney Dis. 1988;12(2):147–51. 26. Hardy MA. Disappearance of glomerular mesangial IgA deposits after renal allograft transplantation. Transplantation. 1982;33(2):214–5. 27. Briganti EM, Russ GR, McNeil JJ, Atkins RC, Chadban SJ. Risk of renal allograft loss from recurrent glomerulonephritis. N Engl J Med. 2002;347(2):103–9. 28. Moroni G, Longhi S, Quaglini S, Gallelli B, Banfi G, Montagnino G, et al. The long-term outcome of renal transplantation of IgA nephropathy and the impact of recurrence on graft survival. Nephrol Dial Transplant. 2013;28(5): 1305–14. 29. Wang AYM, Lai FM, Yu AW-Y, Lam PKW, Chow KM, Choi PCL, et al. Recurrent IgA nephropathy in renal transplant allografts. Am J Kidney Dis. 2001;38(3):588–96. 30. Berthoux F, El Deeb S, Mariat C, Diconne E, Laurent B, Thibaudin L. Antithymocyte globulin (ATG) induction therapy and disease recurrence in renal transplant recipients with primary IgA nephropathy. Transplantation. 2008;85(10):1505–7.
https://openalex.org/W2000760864	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0123523&type=printable	English	null	Breast Cancer Biology and Ethnic Disparities in Breast Cancer Mortality in New Zealand: A Cohort Study	PloS one	2,015	cc-by	7,389	Materials and Methods Data on 2849 women with primary invasive breast cancers diagnosed between 1999 and 2012 were extracted from the Waikato Breast Cancer Register. Differences in distribution of cancer biological characteristics between Māori and NZ European women were explored adjusting for age and socioeconomic deprivation in logistic regression models. Impacts of socioeconomic deprivation, stage and cancer biological characteristics on breast cancer mortality disparity between Māori and NZ European women were explored in Cox regression models. Data Availability Statement: Ethics committee approval for this study restricts the public sharing of data as some of these data contain patient identifiable information, and hence can only be made available to a third party upon request. Please direct all data requests to corresponding author. Email: sanjeewa_sa@yahoo.com; Address: Breast Cancer Research Office, Ground Floor, Hockin Building, Waikato Hospital, Hamilton 3240, New Zealand. Introduction Citation: Seneviratne S, Lawrenson R, Scott N, Kim B, Shirley R, Campbell I (2015) Breast Cancer Biology and Ethnic Disparities in Breast Cancer Mortality in New Zealand: A Cohort Study. PLoS ONE 10(4): e0123523. doi:10.1371/journal.pone.0123523 Indigenous Māori women have a 60% higher breast cancer mortality rate compared with European women in New Zealand. We investigated differences in cancer biological charac- teristics and their impact on breast cancer mortality disparity between Māori and NZ European women. Academic Editor: Robert M Lafrenie, Sudbury Regional Hospital, CANADA Received: September 5, 2014 Accepted: February 20, 2015 Published: April 7, 2015 Copyright: © 2015 Seneviratne 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. RESEARCH ARTICLE Breast Cancer Biology and Ethnic Disparities in Breast Cancer Mortality in New Zealand: A Cohort Study Sanjeewa Seneviratne1, Ross Lawrenson1, Nina Scott2, Boa Kim3, Rachel Shirley3, Ian Campbell1 Sanjeewa Seneviratne1, Ross Lawrenson1, Nina Scott2, Boa Kim3, Rachel Shirley3, Ian Campbell1 1 Waikato Clinical School, University of Auckland, Hamilton, New Zealand, 2 Māori Health Services, Waikato District Health Board, Hamilton, New Zealand, 3 Waikato Breast Cancer Trust, Waikato Hospital, Hamilton, New Zealand a1111 * Sanjeewa.Seneviratne@waikatodhb.health.nz Introduction New Zealand has the seventh highest age standardized mortality rate from breast cancer in the world, a figure which is 20% higher compared with Australia [1, 2]. Indigenous Māori women in New Zealand have one of the highest known population incidences of breast cancer in the world and this incidence is 28% higher compared with NZ European women. Furthermore, mortality from breast cancer for Māori women is 60% higher compared to NZ European women [3, 4]. Despite gradual improvement in breast cancer survival observed for Māori women over the last decade, a significant survival gap persists [5]. Advanced cancer stage at diagnosis in Māori mostly due to lack of healthcare access has been shown to be the major contributor for lower breast cancer survival in Māori compared with NZ European women [3, 6, 7]. However, significant ethnic differences in breast cancer survival remain after adjustment for stage at diagnosis [3]. Hence, factors other than stage in- cluding differences in timeliness and quality of treatment [8–10] and/or differences in cancer biology are likely to be important contributors to the mortality disparity. Data on biological differences in breast cancer between Māori and NZ European women have so far been limited [11–13]. The largest study to date was published by McKenzie et al based on a cohort of women diagnosed during 1994–2004 from the New Zealand Cancer Registry [13]. The authors of this paper have reported significant differences in biological characteristics, including higher rates of poorly differentiated and human epidermal growth factor receptor type 2 (HER-2) positive cancers, and lower rates oestrogen (ER) and proges- terone receptor (PR) negative cancers in Māori compared with non-Māori/non-Pacific (i.e. NZ European) women, which appeared to be independent of socioeconomic deprivation. Two other groups from Auckland and Christchurch have also investigated biological differ- ences using smaller regional cohorts, but have reported on ethnic differences that significant- ly differ from McKenzie at al report, including for tumour grade and hormone receptor status [11, 12]. Although all three studies have contributed significantly to the knowledgebase on ethnic differences in biological characteristics, exact nature of these differences and their impact on breast cancer survival inequity between Māori and NZ European women remain unclear at present. We conducted this study to further investigate differences in breast cancer biological char- acteristics between Māori and NZ European women, and to compare with previously reported figures. Ethnic Differences in Breast Cancer Biology in New Zealand Conclusions Energy Trust towards the WBCR and for additional data collection included for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. More advanced cancer stage at diagnosis has the greatest impact while differences in bio- logical characteristics appear to be a minor contributor for inequities in breast cancer mortal- ity between Māori and NZ European women. Strategies aimed at reducing breast cancer mortality in Māori should focus on earlier diagnosis, which will likely have a greater impact on reducing breast cancer mortality inequity between Māori and NZ European women. Competing Interests: The authors have declared that no competing interests exist. Results Compared with NZ European women (n=2304), Māori women (n=429) had significantly higher rates of advanced and higher grade cancers. Māori women also had non-significantly higher rates of ER/PR negative and HER-2 positive breast cancers. Higher odds of ad- vanced stage and higher grade remained significant for Māori after adjusting for age and dep- rivation. Māori women had almost a 100% higher age and deprivation adjusted breast cancer mortality hazard compared with NZ European women (HR=1.98, 1.55-2.54). Advanced stage and lower proportion of screen detected cancer in Māori explained a greater portion of the excess breast cancer mortality (HR reduction from 1.98 to 1.38), while the additional con- tribution through biological differences were minimal (HR reduction from 1.38 to 1.35). Funding: The authors acknowledge funding support received from the Waikato Breast Cancer Trust, Cancer Society of New Zealand, New Zealand Breast Cancer Foundation, Lion Foundation and the WEL 1 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Introduction We used data from a cohort of women diagnosed over a 14-year period from a compre- hensive regional breast cancer registry to investigate these differences. We also attempted to identify the impact of biological differences on ethnic disparities in breast cancer mortality in New Zealand. 2 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Ethnic Differences in Breast Cancer Biology in New Zealand Study covariates Patient ethnicity was identified from the WBCR, which records self-identified ethnicity collect- ed as a part of the WBCR consent process, as per the Ministry of Health ethnicity data proto- cols [16]. Ethnicity was categorized into Māori, Pacific, NZ European and Other. Socioeconomic deprivation was classified according to the New Zealand Deprivation Index 2006 (NZDep2006) [17]. The NZDep2006 assigns small areas of residence (mesh-blocks with a median population of approximately 100) a deprivation decile on a scale of 1 to 10 based on Patient ethnicity was identified from the WBCR, which records self-identified ethnicity collect- ed as a part of the WBCR consent process, as per the Ministry of Health ethnicity data proto- cols [16]. Ethnicity was categorized into Māori, Pacific, NZ European and Other. ed as a part of the WBCR consent process, as per the Ministry of Health ethnicity data proto- cols [16]. Ethnicity was categorized into Māori, Pacific, NZ European and Other. Socioeconomic deprivation was classified according to the New Zealand Deprivation Index 2006 (NZDep2006) [17]. The NZDep2006 assigns small areas of residence (mesh-blocks with a median population of approximately 100) a deprivation decile on a scale of 1 to 10 based on nine socio-economic variables measured during the 2006 population census; decile 1-least de- prived, decile 10-most deprived. Cancer stage at diagnosis was defined according to the Tumour, Node, and Metastasis (TNM) staging system [18]. Invasive tumour grade was defined according to the Elston and Ellis modified Scarff-Bloom-Richardson breast cancer grading system [19]. Oestrogen (ER) and progesterone (PR) receptor status was determined based on the results of immunohis- tochemistry tests and classified as positive or negative. HER-2 status was based on Fluorescent In-Situ Hybridization (FISH) test or when this was not available, on immunohistochemistry [20]. Comorbidity was measured using the Charlson Comorbidity Index based on documented comorbidities at diagnosis identified from the WBCR [21]. Receipt of chemotherapy, and hor- monal therapy were considered under systemic breast cancer treatment. Outcome variables Date and cause of death for all deceased women (censored at 31/12/2013) were identified from the WBCR and the Mortality Collection of the Ministry of Health. Follow up duration was cal- culated from the date of diagnosis to date of death, or to the date of the last follow up when the patient was known to be alive (censored at 31/12/2013). Materials and Methods Study population All women with newly diagnosed invasive primary breast cancers from 01/01/1999 to 31/12/ 2012 were identified from the Waikato Breast Cancer Register (WBCR). The WBCR is a pro- spectively maintained database that includes over 98% of all breast cancers in women who were resident in the Waikato District Health Board area at the time of diagnosis. The WBCR includes more comprehensive and complete breast cancer data for the Waikato population, in- cluding cancer biological characteristics compared with the New Zealand Cancer Registry. The completeness and accuracy of the WBCR data have been validated previously [14]. Of the total New Zealand population of 4.5 million, Waikato District Health Board covers a population of approximately 380,000. This includes a Māori population of over 75,000 which is the second largest regional Māori population in New Zealand [15]. Ethnic Differences in Breast Cancer Biology in New Zealand and socioeconomic deprivation. The initial base model calculated hazard ratios for breast can- cer specific mortality controlling for age and socioeconomic deprivation. Additional variables were introduced sequentially, starting with breast cancer screening followed by stage at diagno- sis, biological characteristics, treatment and comorbidities. Due to small numbers Pacific and Other ethnic group women were excluded from analyses and ethnic comparisons were per- formed for Māori and NZ European women. Breast cancer-specific survival curves for Maori and NZ European women were estimated using the Kaplan-Meier method and compared by log-rank test. Deaths due to causes other than breast cancer were considered as censored events. As some of the variables included high numbers of missing data, survival analysis was repeated using women diagnosed from 2006 onwards, where rates of missing data were signifi- cantly lower. A further analysis was performed using only cases with complete data for all vari- ables. Results of this analysis were almost similar to those obtained from the full Cox proportional hazards regression model, and these data are not presented in this report. Imputa- tion of missing values was not undertaken due to the similarity of these results. Statistical anal- yses were performed in SPSS (Version 22). proportional hazards regression model, and these data are not presented in this report. Imputa- tion of missing values was not undertaken due to the similarity of these results. Statistical anal- yses were performed in SPSS (Version 22). Ethics statement Ethical approval for this study was obtained from the New Zealand Northern ‘A’ Ethics Com- mittee (Ref. No. 12/NTA/42). Statistical analysis Categorical measures were summarized as numbers with percentages and continuous variables were summarized as means with standard deviation. Chi squared (χ2) test for trend was used to test for univariate differences in age adjusted rates of cancer biological characteristics be- tween Māori and NZ European women. Logistic regression models were used to explore associ- ations of tumour biology with socioeconomic deprivation and ethnicity, adjusting for age. Multivariable Cox proportional hazard models were used to calculate hazard ratios with 95% confidence intervals to identify the association of ethnicity, cancer stage and different cancer biological factors with breast cancer specific mortality independently, and adjusting for age PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 3 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Results A total of 2856 women with new primary invasive breast cancer diagnosed in the Waikato area over the study period were identified. Of these, Pacific (n = 53) and Other (n = 63) ethnic women and seven women in whom a diagnosis of breast cancer was made post-mortem were excluded, leaving 2733 for analysis. There were a total of 688 (25.2%) deaths, out of which 407 (59.2%) were due to breast cancer; 317 (77.9%) in NZ European and 90 (22.1%) in Maori women. The study cohort was followed up for a median of 58 months (mean 66 months) and 67% women were followed up for a minimum of five years or until death. Majority of the study women were of NZ European ethnicity (n = 2304, 80.9%) and 15.1% (n = 429) were Māori. Distribution of tumour biological characteristics by ethnicity is shown in Table 1. Māori women were significantly younger with a mean age difference of approximately six years (61.5 vs. 55.6 years, p<0.001) keeping in with relatively younger Māori population compared with NZ Europeans. A significantly higher age adjusted rate of invasive ductal cancer was observed in Māori compared with NZ European women (85.0% vs. 80.5%, p = 0.032). A corresponding reduction in the rate of invasive lobular carcinoma was seen in Māori compared with NZ European (8.7% vs. 11.7%, p = 0.072), although this difference was not statistically sig- nificant. Māori women had higher likelihoods of larger breast tumours (p<0.001), positive lymphadenopathy (p<0.001), metastatic cancer (p<0.001) and overall more advanced stage cancer (p<0.001) compared with NZ European women. Breast cancers among Māori were of higher grade (p = 0.008) compared with NZ European women, with less grade I and more grade II cancers after adjusting for age. Age adjusted rates of ER+ and PR+ cancers tended to be lower (61.9% vs. 64.2%, p = 0.373), and ER- and PR- cancers tended to be higher in Māori (17.9% vs. 14.4%, p = 0.071) compared with NZ European women. When ER status is consid- ered alone, age adjusted rate of ER positive cancers was significantly lower in Māori compared with NZ European women (80.6% vs. 84.5%, p = 0.011) (data not shown). Māori women had a statistically non-significant, higher age adjusted rate of HER-2 amplified tumours (20.2% vs. 16.3%, p = 0.068) compared to NZ European women (Table 1). 4 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Ethnic Differences in Breast Cancer Biology in New Zealand Table 1. Age and breast cancer biological characteristics at diagnosis compared between NZ European and Māori women. Characteristic NZ European (N = 2304) Māori (N = 429) p n (crude %) Age adjusted % n (crude %) Age adjusted % Age Mean ± SD 61.5 ± 13.9 55.6 ± 12.2 <0.001 Median T Stage 1 1249 (54.4) 53.2 178 (41.7) 42.5 <0.001 2 821 (35.8) 36.6 172 (40.4) 40.8 3 100 (4.4) 4.5 27 (6.3) 5.8 4 121 (5.3) 5.8 49 (11.5) 11.0 Unknown 13 3 N stage 0 1404 (61.4) 62.9 220 (51.9) 53.3 <0.001 1 582 (25.5) 24.7 130 (30.7) 29.7 2 188 (8.2) 7.8 40 (9.4) 9.3 3 112 (4.9) 4.6 34 (8.0) 7.7 Unknown 18 5 M Stage 0 2197 (95.4) 94.9 379 (88.3) 88.6 <0.001 1 107 (4.6) 5.1 50 (11.7) 11.4 Stage category I 972 (42.2) 41.6 142 (33.1) 34.2 <0.001 II 887 (38.5) 39.1 163 (38.0) 37.5 III 338 (14.7) 14.2 74 (17.2) 16.9 IV 107 (4.6) 5.1 50 (11.7) 11.4 Histology Ductal 1831 (81.2) 80.5 352 (85.2) 85.0 0.032a Lobular 258 (11.4) 11.7 35 (8.5) 8.7 0.072b Mixed 42 (1.9) 1.8 9 (2.2) 2.3 Other 125 (5.5) 5.9 17 (4.1) 4.0 Unknown 48 16 Grade Grade I 543 (25.3) 25.6 69 (17.6) 18.5 0.008 Grade II 1118 (52.1) 52.7 229 (58.4) 59.7 Grade III 485 (22.6) 21.7 93 (23.7) 21.8 Unknown 158 38 ER/PR ER+/PR+ 1414 (64.1) 64.2 252 (60.9) 61.9 0.204 ER+/PR- 430 (19.5) 20.2 75 (18.1) 18.6 0.071c ER-/PR+ 28 (1.3) 1.1 8 (1.9) 1.6 ER-/PR- 333 (15.1) 14.4 79 (19.1) 17.9 ER or PR Unknown 99 15 HER-2 Negative 1239 (74.8) 75.2 257 (72.4) 73.8 0.091 Equivocal 135 (8.1) 8.5 21 (5.9) 6.0 0.069d Positive 283 (17.1) 16.3 77 (21.7) 20.2 Unknown 647 74 (Continued) breast cancer biological characteristics at diagnosis compared between NZ European and Māori women. Table 1. Age and breast cancer biological characteristics at diagnosis compared between NZ European PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 5 / 13 Ethnic Differences in Breast Cancer Biology in New Zealand Table 1. (Continued) Characteristic NZ European (N = 2304) Māori (N = 429) p n (crude %) Age adjusted % n (crude %) Age adjusted % TNBC e No 1455 (91.5) 91.7 316 (92.9) 93.0 0.446 Yes 135 (8.5) 8.3 24 (7.1) 7.0 Unknown 714 89 Detection method Non-screen 1443 (62.6) 64.1 300 (69.9) 68.9 0.056 Screen 861 (37.4) 35.9 129 (30.1) 31.1 Comorbidity score 0 1903 (82.6) 79.6 304 (70.9) 67.6 <0.001 1–2 366 (15.9) 18.4 110 (25.6) 28.2 3+ 35 (1.5) 2.0 15 (3.5) 4.1 Loco-regional therapy BCS with RT 1082 (41.0) 40.3 153 (27.5) 27.6 <0.001 BCS without RT 196 (14.4) 13.6 30 (15.2) 14.5 Mastectomy 852 (37.0) 36.8 188 (43.8) 43.3 No primary surgery 174 (7.6) 9.2 58 (13.5) 14.7 Chemotherapy No 1580 (68.6) 73.5 270 (62.9) 67.3 0.015 Yes 724 (31.4) 26.5 159 (37.1) 32.7 Endocrine therapy No 678 (29.4) 30.3 146 (34.0) 33.8 0.041 Yes 1626 (70.6) 69.7 283 (66.0) 66.2 Post 2005 breast cancers NZ European (N = 1247) Māori (N = 278) p n (crude %) Age adjusted % n (crude %) Age adjusted % HER-2 Negative 946 (79.5) 79.9 202 (74.8) 75.9 0.072 Equivocal 65 (5.5) 5.8 11 (4.1) 4.4 0.012d Positive 179 (15.0) 14.3 57 (21.1) 19.7 Unknown 57 8 TNBC e No 1054 (92.6) 92.6 238 (93.3) 93.5 0.536 Yes 84 (7.4) 7.4 17 (6.7) 6.5 Unknown 109 23 a ductal vs. other histology types, As the rate of missing HER-2 data was relatively high (26.2%), an analysis was performed only including breast cancers diagnosed from 2006, where the rate of missing HER-2 status was only 4.3%. This showed figures similar to complete NZ European and Māori cohorts, with a higher age adjusted rate of HER-2 positivity in Māori compared with NZ European women (19.7% vs. 14.3%, p = 0.076) (Table 1). As the rate of missing HER-2 data was relatively high (26.2%), an analysis was performed only including breast cancers diagnosed from 2006, where the rate of missing HER-2 status was only 4.3%. This showed figures similar to complete NZ European and Māori cohorts, with a higher age adjusted rate of HER-2 positivity in Māori compared with NZ European women (19.7% vs. 14.3%, p = 0.076) (Table 1). Ethnic Differences in Breast Cancer Biology in New Zealand Table 2. Age adjusted odds ratios (OR) with 95% confidence intervals (95% CI) for tumour biological characteristics by socio-economic depriva- tion category (NZDep 2006) Table 2. Age adjusted odds ratios (OR) with 95% confidence intervals (95% CI) for tumour biological characteristics by socio-economic depriva- tion category (NZDep 2006). Deprivation quintile Stage a n = 2733 Grade b n = 2537 ER/PR c n = 2673 HER-2 d n = 2012 TNBC e n = 1930 OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p Dep 1–2 Ref 0.006 Ref 0.095 Ref 0.011 Ref 0.939 Ref 0.270 Dep 3–4 0.85 (0.55– 1.30) 0.75 (0.51– 1.11) 0.98 (0.59– 1.64) 0.97 (0.58– 1.63) 0.67 (0.30– 1.52) Dep 5–6 0.88 (0.65– 1.27) 1.05 (0.74– 1.47) 1.31 (0.86– 2.00) 1.05 (0.69– 1.60) 0.90 (0.49– 1.68) Dep 7–8 1.27 (0.90– 1.78) 0.98 (0.71– 1.36) 1.41 (0.94– 2.13) 0.98 (0.65– 1.47) 1.15 (0.64– 2.08) Dep 9–10 1.33 (0.94– 1.87) 1.18 (0.84– 1.67) 1.77 (1.18– 2.67) 1.11 (0.73– 1.67) 1.31 (0.73– 2.38) with 95% confidence intervals (95% CI) for tumour biological characteristics by socio-economic depriva- Table 2. Age adjusted odds ratios (OR) with 95% confidence intervals (95% CI) for tumour biological characteristics by socio-economic depriva- tion category (NZDep 2006). Table 2. Age adjusted odds ratios (OR) with 95% confidence intervals (95% CI) for tumour biological charac tion category (NZDep 2006). a—Stage III & IV compared with stage I & II, c—ER and PR negative compared with ER and/or PR positive, —HER-2 positive compared with HER-2 equivocal and negative, e—Triple negative breast cancer (TNBC) compared with non-TNBC. Triple receptor status (ER, PR and HER-2) was determined for a total 1930 (70.7%) women with invasive breast cancer. Of this group, 8.3% of cancers were negative for all three receptor types; i.e. triple negative breast cancer (TNBC). NZ European women had a higher age adjusted rate of TNBC compared with Māori women, although this was statistically not significant (7.0% vs. 8.3%, p = 0.446). Of women diagnosed from 2006 onwards, TNBC status was avail- able for 1393 (91.3%) women. Age-adjusted rate of TNBC was higher in NZ European than in Māori, but this was still statistically non-significant (7.4% vs. 6.5%, p = 0.532). As the rate of missing HER-2 data was relatively high (26.2%), an analysis was performed only including breast cancers diagnosed from 2006, where the rate of missing HER-2 status was only 4.3%. This showed figures similar to complete NZ European and Māori cohorts, with a higher age adjusted rate of HER-2 positivity in Māori compared with NZ European women (19.7% vs. 14.3%, p = 0.076) (Table 1). 6 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 doi:10.1371/journal.pone.0123523.t002 doi:10.1371/journal.pone.0123523.t003 006) adjusted odds ratios (OR) with 95% confidence intervals (95% CI) for breast cancer biological charac- ropean women. Table 3. Age and deprivation (NZDep 2006) adjusted odds ratios (OR) with 95% confidence in teristics for Māori compared with NZ European women. chemotherapy and endocrine therapy) and comorbidity resulted in a final hazard ratio of 1.25 (0.97–1.61), which was no longer statistically significantly (p = 0.088) (Tables 4 & 5). Multivari- ate Cox regression model was repeated only including women diagnosed from 2006 onwards, where rates of missing data were significantly smaller (Table 4). Overall results of this model were much similar to the model that included all women (final HR 1.25 vs. 1.28). Kaplan- Meier survival curves for crude breast cancer specific survival by ethnicity is shown in Fig 1. Ten year breast cancer specific survival in NZ European women was significantly higher com- pared with Maori women (p<0.001) with 10-year survival rates of 79.9% (95% CI 79.88–79.92) and 66.4% (95% CI 66.33–66.47), respectively. Increasing social deprivation significantly increased the risk of (age adjusted) advanced stage (p = 0.001) and ER/PR negative (p = 0.011) invasive cancers. No significant associations were observed between deprivation and age adjusted rates of high tumour grade (p = 0.095), HER-2 positivity (p = 0.939) or TNBC status (p = 0.270) (Table 2). Higher socioeconomic dep- rivation status was significantly higher in Maori compared with NZ European women (Dep. 7– 10 72.2% in Maori vs. 51.7% in NZ European, p<0.001, data not shown). Compared to NZ Eu- ropean women, age and socioeconomic deprivation adjusted risk of advanced stage, higher grade, ER and PR negativity and HER-2 positivity were higher while the rate of TNBC was lower(8.3% vs. 7.0) in Māori women. Differences in stage and grade were statistically signifi- cant, while differences in ER /PR, HER-2 and TNBC were not (Table 3). Results from the survival analysis with Cox regression model are shown in Table 4, and change in hazard ratios with sequential introduction of variables of interest into the Cox regres- sion model is shown in Table 5. Māori women had a significantly higher age adjusted breast cancer mortality compared with NZ European women (HR 2.07, p<0.001). Adjusting for so- cioeconomic deprivation marginally reduced the age-adjusted hazard of mortality for Māori compared with NZ European from 2.07 (1.64–2.61) to 1.98 (1.55–2.54) (Table 5). As the pro- portion of screen detected cancer was significantly higher in NZ European compared with Maori (37.4% vs. 30.1% p = 0.002, data not shown), detection method was included as a covari- ate in the survival model. Adjusting for screening status and tumour stage (TNM stage) re- duced the HR for mortality to 1.38 (1.06–1.78) (Table 5). Adjusting for tumour biological factors (i.e., grade, hormone receptor status, HER-2 status and histology type) further attenuat- ed this estimate (HR 1.35, 1.04–1.75). Further adjustments for treatment characteristics (i.e., 7 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Ethnic Differences in Breast Cancer Biology in New Zealand Table 3. Age and deprivation (NZDep 2006) adjusted odds ratios (OR) with 95% confidence intervals (95% CI) for breast cancer biological charac- teristics for Māori compared with NZ European women. Ethnicity Stage a n = 2733 Grade b n = 2537 ER/PR c n = 2673 HER-2 d n = 2012 TNBC e n = 1930 OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p NZ European Ref <0.001 Ref 0.007 Ref 0.539 Ref 0.214 Ref 0.159 Māori 1.58 (1.24–2.01) 1.48 (1.13–1.96) 1.09 (0.83–1.44) 1.21 (0.89–1.61) 0.72 (0.45–1.14) a—Stage III & IV compared with stage I & II, b—Grade II & III compared with grade I, c—ER and PR negative compared with ER and/or PR positive, d—HER-2 positive compared with HER-2 equivocal and negative, e—Triple negative breast cancer (TNBC) compared with non-TNBC. chemotherapy and endocrine therapy) and comorbidity resulted in a final hazard ratio of 1.25 (0.97–1.61), which was no longer statistically significantly (p = 0.088) (Tables 4 & 5). Multivari- ate Cox regression model was repeated only including women diagnosed from 2006 onwards, where rates of missing data were significantly smaller (Table 4). Overall results of this model were much similar to the model that included all women (final HR 1.25 vs. 1.28). Kaplan- Meier survival curves for crude breast cancer specific survival by ethnicity is shown in Fig 1. b f l f l h h Table 3. Age and deprivation (NZDep 2006) adjusted odds ratios (OR) with 95% confidence intervals (95% CI) for breast cancer biological charac- teristics for Māori compared with NZ European women. Ethnicity Stage a n = 2733 Grade b n = 2537 ER/PR c n = 2673 HER-2 d n = 2012 TNBC e n = 1930 OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p NZ European Ref <0.001 Ref 0.007 Ref 0.539 Ref 0.214 Ref 0.159 Māori 1.58 (1.24–2.01) 1.48 (1.13–1.96) 1.09 (0.83–1.44) 1.21 (0.89–1.61) 0.72 (0.45–1.14) a—Stage III & IV compared with stage I & II, b—Grade II & III compared with grade I, c—ER and PR negative compared with ER and/or PR positive, d—HER-2 positive compared with HER-2 equivocal and negative, e—Triple negative breast cancer (TNBC) compared with non-TNBC. doi:10.1371/journal.pone.0123523.t003 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Discussion From this study we have observed some differences in breast cancer biological characteristics between Māori and NZ European women. Although Māori women had higher likelihoods of exhibiting certain biological characteristics associated with worse breast cancer outcomes, this appears to be only a minor contributor while advanced stage at diagnosis in Māori had the greatest impact towards the breast cancer survival inequity between Māori and NZ European women. Overall, Māori women had higher rates of advanced stage and higher grade, and possi- bly a higher rate of HER-2 positive cancers. No significant differences were observed in rates of ER/PR negative or triple negative breast cancers (TNBC). We have observed several key differences in our findings compared with previous studies [11–13]. For example, McKenzie study based on the New Zealand Cancer Registry, reported that Māori women have higher rates of ER/PR positive and poorly differentiated (i.e. grade III) cancers compared with NZ European women [13]. A higher rate of grade III cancers in Māori was reported from another study based on the Auckland Breast Cancer Register [11], while a third study from Christchurch reported that Māori women have a significantly lower rate of grade III cancers compared with NZ European women [12]. Differences in sample selection, rates of missing data, statistical methods used for analysis and possible regional variations in breast cancer biology could explain some of these differences. For instance, McKenzie study based on the New Zealand Cancer Registry included very high rates of missing data; 65.1% for tumour grade and 59.8% for ER status. Further, as the authors of the Christchurch study have proposed [12], regional variations in breast cancer biological characteristics, may also have contributed, especially for differences between Auckland and Christchurch datasets. Such re- gional differences have been observed in other countries [22, 23], and could be related to differ- ences in distribution of risk factors associated with tumour biological expressions. 8 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Ethnic Differences in Breast Cancer Biology in New Zealand Table 4. Cox regression model for factors associated with breast cancer specific mortality in the Wai- kato, New Zealand 1999–2012. PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Discussion Characteristic Univariate Multivariate Multivariate (Post 2005 cancers only) HR 95% CI p HR 95% CI p HR 95% CI p Ethnicity a NZ European Ref <0.001 Ref 0.088 Ref 0.226 Māori 1.98 1.55–2.54 1.25 0.97–1.61 1.28 0.86–1.91 Year of diagnosis 1999–2002 Ref 0.556 Ref 0.84 - 2003–2006 1.13 1.19–1.43 1.11 0.85–1.45 Ref 0.793 2007–2009 0.95 0.70–1.27 0.80 0.58–1.12 0.95 0.62–1.46 2010–2012 0.97 0.66–1.41 0.77 0.51–1.16 0.84 0.50–1.41 Mode of detection Non-screen Ref <0.001 Ref 0.031 Ref 0.029 Screen 0.26 0.20–0.35 0.71 0.52–0.96 0.49 0.26–0.93 T stage T1 Ref <0.001 Ref <0.001 Ref 0.001 T2 3.33 2.56–4.33 1.87 1.41–2.48 3.39 1.81–6.35 T3 8.68 6.01–12.5 3.13 2.10–4.66 4.51 2.04–9.95 T4 18.6 13.8–25.1 3.10 2.15–4.48 4.55 2.19–9.49 N stage N0 Ref <0.001 Ref <0.001 Ref 0.002 N1 2.03 1.58–2.60 1.63 1.25–2.13 1.49 0.94–2.36 N2+ 4.04 3.01–5.42 2.67 1.92–3.71 1.88 1.04–3.42 M Stage M0 Ref <0.001 Ref <0.001 Ref <0.001 M1 15.4 12.2–19.4 3.60 2.61–4.96 4.21 2.63–6.73 Grade I Ref <0.001 Ref <0.001 Ref <0.001 II 4.93 2.90–8.38 3.01 1.76–5.17 3.94 1.39–11.2 III 13.4 7.85–22.7 5.22 2.98–9.14 6.47 2.25–18.6 ER/PR ER &/or PR + Ref <0.001 Ref <0.001 Ref 0.008 ER & PR - 2.47 1.98–3.07 1.57 1.22–1.88 1.71 1.14–2.53 HER-2 Negative Ref <0.001 Ref 0.157 Ref 0.386 Equivocal 0.62 0.38–1.03 0.93 0.56–1.55 1.13 0.52–2.42 Positive 1.80 1.39–2.33 0.93 0.71–1.22 0.76 0.49–1.16 Histology Ductal Ref 0.293 Ref 0.139 Ref 0.458 Lobular 0.87 0.62–1.22 0.83 0.57–1.21 0.85 0.46–1.56 Mixed 0.91 0.45–1.83 1.29 0.62–2.65 0.84 0.32–2.17 Other 0.58 0.32–1.05 0.73 0.39–1.36 0.47 0.21–1.04 Comorbidity score 0 Ref <0.001 Ref 0.003 Ref 0.005 1–2 2.05 1.64–2.56 1.49 1.17–1.90 1.90 1.28–2.82 3+ 2.48 1.32–4.67 1.24 0.64–2.37 1.65 0.67–4.07 (Continued) Table 4. Cox regression model for factors associated with breast cancer specific mortality in the Wai- kato, New Zealand 1999–2012. PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 9 / 13 Ethnic Differences in Breast Cancer Biology in New Zealand Table 4. (Continued) Characteristic Univariate Multivariate Multivariate (Post 2005 cancers only) HR 95% CI p HR 95% CI p HR 95% CI p Chemotherapy No Ref 0.387 Ref 0.103 Ref 0.154 Yes 1.09 0.89–1.34 0.81 0.62–1.64 0.72 0.46–1.13 Endocrine therapy No Ref <0.001 Ref 0.019 Ref 0.057 Yes 0.24 0.20–0.30 0.71 0.53–0.94 0.64 0.40–1.01 HR—hazard ratios, 95% CI—95% conﬁdence intervals, a—adjusted for age and socio-economic deprivation. Discussion doi:10.1371/journal.pone.0123523.t004 African American women with breast cancer in both the USA and the UK are known to harbour more high grade, ER/PR negative and TNBC than their European American counter- parts [22, 24–26]. Further, these differences are known to be major contributors for excess breast cancer mortality in African American women [22, 25]. Although, compared to NZ Euro- pean women, Māori women had a higher rates of ER/PR negative cancers (crude rate 15.1% vs. 19.1%) and grade III cancers (crude rate 22.6% vs. 23.7%), these rates were much lower than rates of respective characteristics observed in African American women, in whom the rates of ER/PR negative or grade III cancers were approximately 30–35% [23]. Further, in contrast to African American women, the rate of TNBC tended to be lower in Māori compared with NZ European women, although this difference was not significant in either unadjusted or adjusted analyses. Although it appears that differences tumour biology in Māori women may be a con- tributor to higher mortality; it certainly is not as significant a contributor as it is for African American women. Studies from the USA and the UK have demonstrated that women of lower socioeconomic groups to have significantly higher rates of advanced stage, ER/PR negative, high grade and in- vasive ductal cancers compared with women living in affluent socioeconomic circumstances Table 5. Hazard ratios for breast cancer-specific mortality risk in Māori compared with NZ European women with stepwise adjustment for screening status, cancer stage, biological characteristics, co- morbidity and treatment factors. Characteristics HR (95% CI) Baseline—Age adjusted 2.07 (1.64–2.61) Model A (Adjusted for socioeconomic deprivation) 1.98 (1.55–2.54) Model B (Model A + Screening status) 1.82 (1.42–2.32) Model C (Model B + Cancer stage at diagnosis) Tumour, Lymph nodes & Metastasis 1.38 (1.06–1.78) Model D (Model C + Cancer biological factors) ER/PR, Grade, HER-2 & histology 1.35 (1.05–1.75) Model E (Model D + Treatment) Use of chemo and endocrine therapy 1.33 (1.03–1.73) Model F (Model E + Patient factors) Comorbidity index score 1.25 (0.97–1.61) doi:10.1371/journal.pone.0123523.t005 Table 5. Hazard ratios for breast cancer-specific mortality risk in Māori compared with NZ European women with stepwise adjustment for screening status, cancer stage, biological characteristics, co- morbidity and treatment factors. PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Discussion Characteristics HR (95% CI) Baseline—Age adjusted 2.07 (1.64–2.61) Model A (Adjusted for socioeconomic deprivation) 1.98 (1.55–2.54) Model B (Model A + Screening status) 1.82 (1.42–2.32) Model C (Model B + Cancer stage at diagnosis) Tumour, Lymph nodes & Metastasis 1.38 (1.06–1.78) Model D (Model C + Cancer biological factors) ER/PR, Grade, HER-2 & histology 1.35 (1.05–1.75) Model E (Model D + Treatment) Use of chemo and endocrine therapy 1.33 (1.03–1.73) Model F (Model E + Patient factors) Comorbidity index score 1.25 (0.97–1.61) doi:10.1371/journal.pone.0123523.t005 Table 5. Hazard ratios for breast cancer-specific mortality risk in Māori compared with NZ European women with stepwise adjustment for screening status, cancer stage, biological characteristics, co- morbidity and treatment factors. PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 10 / 13 Ethnic Differences in Breast Cancer Biology in New Zealand Fig 1. Kaplan-Meier survival curves by ethnicity for crude breast cancer specific survival for Māori and NZ European women with invasive breast cancer in the Waikato 1999–2012. doi:10.1371/journal.pone.0123523.g001 Fig 1. Kaplan-Meier survival curves by ethnicity for crude breast cancer specific survival for Māori and NZ European women with invasive breast cancer in the Waikato 1999–2012. doi:10.1371/journal.pone.0123523.g001 doi:10.1371/journal.pone.0123523.g001 [27, 28]. Although many New Zealand studies have reported on the influence of socioeconomic status on cancer stage at diagnosis for many cancers including breast [7, 29], only the McKen- zie study to date has reported on biological differences in breast cancer by socioeconomic status [13]. This study did not observe significant differences in tumour grade, ER/PR and HER-2 sta- tus among different socioeconomic groups. In contrast, we observed a higher age-adjusted rate of ER/PR negative cancers in women of low socioeconomic groups, which was marked for women from the most deprived socioeconomic quintile. Further, in our study, adjusting for age and socioeconomic status resulted only in a marginal attenuation of higher grade cancers observed in Māori compared with NZ Europeans. Despite the differences in the nature of bio- logical differences between the two studies, both indicate that Māori women may have differ- ences in breast cancer biology compared with NZ European women, which are likely to be independent of age at diagnosis and socioeconomic deprivation. The main strengths of our study include the comprehensive nature of our data which in- cluded approximately 98% of cancers diagnosed in the Waikato region over the 14-year period of this study. PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Ethnic Differences in Breast Cancer Biology in New Zealand although it may not accurately represent the socioeconomic status of each individual. Third, we acknowledge the possible differences in analysis and reporting of biological characteristics by different laboratories [31]. More than 95% of the pathology tests for cancers included in our study were performed by two laboratories; one public and one private. These two laboratories have used similar equipment, tests and reporting protocols over the time period and are ex- pected to have had minimal analysis and reporting variations. In conclusion, we have observed Māori ethnicity and lower socioeconomic status to be sig- nificantly associated with some breast cancer biological characteristics associated with worse cancer outcomes. However, differences in tumour biological factors appear to be contributing minimally, while delay in diagnosis in Māori appears to have a major impact on the breast can- cer mortality inequity between Māori and NZ European women. Strategies aimed at reducing breast cancer mortality in Māori should focus on earlier diagnosis through increasing screening coverage and other methods, which will likely have a greater impact on minimizing the breast cancer mortality inequity between Māori and NZ European women. Author Contributions Conceived and designed the experiments: SS RL IC. Performed the experiments: SS BK RS. An- alyzed the data: SS BK RS. Contributed reagents/materials/analysis tools: SS RS BK. Wrote the paper: SS BK NS RS RL IC. References 1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010; 127(12):2893–917. doi: 10.1002/ijc.25516 PMID: 21351269 2. Althuis MD, Dozier JM, Anderson WF, Devesa SS, Brinton LA. Global trends in breast cancer incidence and mortality 1973–1997. Int J Epidemiol. 2005; 34(2):405–12. PMID: 15737977 3. Robson B, Purdie G, Cormack D. Unequal Impact II: Maori and Non-Maori Cancer Statistics by Depri- vation and Rural-Urban Status 2002–2006. Wellington: Ministry of Health; 2010. 4. Ajwani S, Blakely T, Robson B, Tobias M, Bonne M. Decades of disparity: Ethnic mortality trends in New Zealand 1980–1999, Ministry of Health and University of Otago. Wellington, New Zealand 2003. PMID: 15475978 5. Cunningham R, Shaw C, Blakely T, Atkinson J, Sarfati D. Ethnic and socioeconomic trends in breast cancer incidence in New Zealand. BMC Cancer. 2010; 10:674. doi: 10.1186/1471-2407-10-674 PMID: 21138590 6. Ajwani S, Blakely T, Robson B, Tobias M, Bonne M. Decades of disparity: ethnic mortality trends in New Zealand 1980–1999. Ministry of Health and University of Otago, New Zealand; 2003. PMID: 15475978 7. Jeffreys M, Stevanovic V, Tobias M, Lewis C, Ellison-Loschmann L, Pearce N, et al. Ethnic inequalities in cancer survival in New Zealand: linkage study. Am J Public Health. 2005; 95(5):834–7. PMID: 15855462 8. Seneviratne S, Campbell I, Scott N, Coles C, Lawrenson R. Treatment delay for Maori women with breast cancer in New Zealand. Ethn Health. 2015; 20(2):178–93. doi: 10.1080/13557858.2014.895976 PMID: 24635721 9. Seneviratne S, Campbell I, Scott N, Kuper-Hommel M, Kim B, Pillai A, et al. Adherence to adjuvant en- docrine therapy: Is it a factor for ethnic differences in breast cancer outcomes in New Zealand? The Breast. 2015; 24(1):62–7. doi: 10.1016/j.breast.2014.11.011 PMID: 25486877 10. Seneviratne S, Campbell I, Scott N, Kuper-Hommel M, Round G, Lawrenson R. Ethnic differences in timely adjuvant chemotherapy and radiation therapy for breast cancer in New Zealand: a cohort study. BMC Cancer. 2014; 14(1):839. doi: 10.1186/1471-2407-14-1 PMID: 24383403 11. Weston MK, Moss DP, Stewart J, Hill AG. Differences in breast cancer biological characteristics be- tween ethnic groups in New Zealand. Breast Cancer Res Treat. 2008; 111(3):555–8. PMID: 18026873 12. Dachs GU, Kano M, Volkova E, Morrin HR, Davey VC, Harris GC, et al. A profile of prognostic and mo- lecular factors in European and Maori breast cancer patients. BMC Cancer. 2010; 10:543. doi: 10. PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 Discussion All breast cancer data were directly extracted from prospectively collected data forms, patient clinical records and histopathology reports, and this has helped to maintain a high rate of data accuracy and completeness in our database, which far exceeds other New Zea- land cancer databases such as the national cancer registry. We acknowledge some limitations in our analysis. First, some biological characteristics had significant proportions of missing data. For example HER-2 data were missing for approximately 25% of women, most of who were diagnosed prior to 2006 when HER-2 testing was not routine in New Zealand. Further, even among women diagnosed post-2006, in addition to missing HER-2 rate of 4.3%, a further 5% had an equivocal result, which may have been a source of misclassification bias. Second, we have used NZDep2006 as a measure of socioeconomic status, which is based on area level dep- rivation [17]. Area level deprivation measured with NZDep2006 has been validated to be an ac- curate proxy measure of socioeconomic deprivation for epidemiological research [30], 11 / 13 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 References 1186/1471-2407-10-543 PMID: 20932344 PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 12 / 13 Ethnic Differences in Breast Cancer Biology in New Zealand 13. McKenzie F, Jeffreys M, Mannetje A, Pearce N. Prognostic factors in women with breast cancer: in- equalities by ethnicity and socioeconomic position in New Zealand. Cancer Causes Control. 2008; 19 (4):403–11. PMID: 18066672 14. Seneviratne S, Campbell I, Scott N, Shirley R, Peni T, Lawrenson R. Accuracy and completeness of the New Zealand Cancer Registry for staging of invasive breast cancer. Cancer Epidemiol. 2014; 38 (5):638–44. doi: 10.1016/j.canep.2014.06.008 PMID: 25037979 15. Statistics New Zealand. 2006 Census Data 2007. Available: http://www.stats.govt.nz 16. Ministry of Health. Ethnicity Data Protocols for the Health and Disability Sector Wellington2004 [15/05/ 2014]. Available: http://www.health.govt.nz/publication/ethnicity-data-protocols-health-and-disability- sector. 17. Salmond C, Crampton P, Atkinson J. NZDep2006: Index of Deprivation. Wellington: Department of Public Health, University of Otago, Wellington., 2007. 18. American Joint Committee on Cancer. AJCC Cancer Staging Manual. New York: Springer-Verlag, 2010. 19. Elston CW, Ellis IO. Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. Histopathology. 1991; 19 (5):403–10. PMID: 1757079 20. Pauletti G, Godolphin W, Press MF, Slamon DJ. Detection and quantitation of HER-2/neu gene amplifi- cation in human breast cancer archival material using fluorescence in situ hybridization. Oncogene. 1996; 13(1):63–72. E PMID: 8700555 21. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987; 40(5):373–83. PMID: 3558716 22. O'Brien KM, Cole SR, Tse CK, Perou CM, Carey LA, Foulkes WD, et al. Intrinsic breast tumor subtypes, race, and long-term survival in the Carolina Breast Cancer Study. Clin Cancer Res. 2010; 16(24):6100– 10. doi: 10.1158/1078-0432.CCR-10-1533 PMID: 21169259 23. Li CI, Malone KE, Daling JR. Differences in breast cancer hormone receptor status and histology by race and ethnicity among women 50 years of age and older. Cancer Epidemiol Biomarkers Prev. 2002; 11(7):601–7. PMID: 12101106 24. Carey LA, Perou CM, Livasy CA, Dressler LG, Cowan D, Conway K, et al. Race, breast cancer sub- types, and survival in the Carolina Breast Cancer Study. JAMA. 2006; 295(21):2492–502. PMID: 16757721 25. Rosenberg J, Chia YL, Plevritis S. The effect of age, race, tumor size, tumor grade, and disease stage on invasive ductal breast cancer survival in the U.S. SEER database. Breast Cancer Res Treat. 2005; 89(1):47–54. PLOS ONE \| DOI:10.1371/journal.pone.0123523 April 7, 2015 References PMID: 15666196 26. Jack RH, Davies EA, Renshaw C, Tutt A, Grocock MJ, Coupland VH, et al. Differences in breast cancer hormone receptor status in ethnic groups: a London population. Eur J Cancer. 2013; 49(3):696–702. doi: 10.1016/j.ejca.2012.09.012 PMID: 23058788 27. Vona-Davis L, Rose DP. The influence of socioeconomic disparities on breast cancer tumor biology and prognosis: a review. J Womens Health (Larchmt). 2009; 18(6):883–93. doi: 10.1089/jwh.2008. 1127 PMID: 19514831 28. Taylor A, Cheng KK. Social deprivation and breast cancer. J Public Health Med. 2003; 25(3):228–33. PMID: 14575198 29. Haynes R, Pearce J, Barnett R. Cancer survival in New Zealand: ethnic, social and geographical in- equalities. Soc Sci Med. 2008; 67(6):928–37. doi: 10.1016/j.socscimed.2008.05.005 PMID: 18573580 30. Salmond CE, Crampton P. Development of New Zealand's deprivation index (NZDep) and its uptake as a national policy tool. Can J Public Health. 2012; 103(8 Suppl 2):S7–11. PMID: 23618071 31. Allred DC. Commentary: hormone receptor testing in breast cancer: a distress signal from Canada. On- cologist. 2008; 13(11):1134–6. doi: 10.1634/theoncologist.2008-0184 PMID: 18987048 13 / 13
https://openalex.org/W2516504335	https://pure.rug.nl/ws/files/36643962/art_3A10.1186_2Fs12983_016_0168_7.pdf	English	null	Unique wing scale photonics of male Rajah Brooke’s birdwing butterflies	Frontiers in zoology	2,016	cc-by	10,400	University of Groningen University of Groningen Unique wing scale photonics of male Rajah Brooke's birdwing butterflies Wilts, Bodo D.; Giraldo, Marco A.; Stavenga, Doekele G. DOI: 10.1186/s12983-016-0168-7 IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publication date: 2016 Publication date: 2016 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Wilts, B. D., Giraldo, M. A., & Stavenga, D. G. (2016). Unique wing scale photonics of male Rajah Brooke's birdwing butterflies. Frontiers in Zoology, 13, Article 36. https://doi.org/10.1186/s12983-016-0168-7 Citation for published version (APA): Wilts, B. D., Giraldo, M. A., & Stavenga, D. G. (2016). Unique wing scale photonics of male Rajah Brooke's birdwing butterflies. Frontiers in Zoology, 13, Article 36. https://doi.org/10.1186/s12983-016-0168-7 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). © 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. Unique wing scale photonics of male Rajah Brooke’s birdwing butterflies Bodo D. Wilts1, Marco A. Giraldo2 and Doekele G. Stavenga3 Abstract Background: Ultrastructures in butterfly wing scales can take many shapes, resulting in the often striking coloration of many butterflies due to interference of light. The plethora of coloration mechanisms is dazzling, but often only single mechanisms are described for specific animals. Results: We have here investigated the male Rajah Brooke’s birdwing, Trogonoptera brookiana, a large butterfly from Malaysia, which is marked by striking, colorful wing patterns. The dorsal side is decorated with large, iridescent green patterning, while the ventral side of the wings is primarily brown-black with small white, blue and green patches on the hindwings. Dense arrays of red hairs, creating a distinct collar as well as contrasting areas ventrally around the thorax, enhance the butterfly’s beauty. The remarkable coloration is realized by a diverse number of intricate and complicated nanostructures in the hairs as well as the wing scales. The red collar hairs contain a broad-band absorbing pigment as well as UV-reflecting multilayers resembling the photonic structures of Morpho butterflies; the white wing patches consist of scales with prominent thin film reflectors; the blue patches have scales with ridge multilayers and these scales also have centrally concentrated melanin. The green wing areas consist of strongly curved scales, which possess a uniquely arranged photonic structure consisting of multilayers and melanin baffles that produces highly directional reflections. Conclusion: Rajah Brooke’s birdwing employs a variety of structural and pigmentary coloration mechanisms to achieve its stunning optical appearance. The intriguing usage of order and disorder in related photonic structures in the butterfly wing scales may inspire novel optical materials as well as investigations into the development of these nanostructures in vivo. Keywords: Animal coloration, Polarization, Grating, Iridescence, Diffraction, FDTD, Photonics, Melani Copyright 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. down policy believe that this document breaches copyright please contact us providing details, and we will remove access to the wo vestigate 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. 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 Wilts et al. Frontiers in Zoology (2016) 13:36 DOI 10.1186/s12983-016-0168-7 Correspondence: bodo.wilts@unifr.ch 1Adolphe Merkle Institute, University of Fribourg, Chemin des Verdiers 4, CH-1700 Fribourg, Switzerland Full list of author information is available at the end of the article Background B fli The pigments encountered in butterflies are rather fam- ily specific. For instance, pterins are prominently en- countered in the Pieridae [3], ommochromes are the pigments of the Nymphalidae [4], while papiliochrome pigments have been only found in Papilionidae [5–8]. The pigments are dispersed in scales that cover butterfly wings like tiles on a roof. A butterfly wing scale basically consists of a lower lamina, which is essentially a thin plate with thickness 100-200 nm, connected by pillar- like trabeculae to the upper lamina, which is made up of an array of parallel ridges, with interdistance 1-2 μm, and connecting cross-ribs [9]. Butterflies are a hallmark of biodiversity and multifunc- tionality in nature. Of particular beauty are the birdwing butterflies of Australasia, which are noted by their ex- ceptional size and a birdlike flight [1, 2]. An especially attractive butterfly is Rajah Brooke’s birdwing, Trogonop- tera brookiana, the national butterfly of Malaysia. How- ever, a detailed explanation of the striking coloration has yet to be made. Recent studies on butterfly coloration have revealed a multitude of optical mechanisms that strongly alter the composition of incident light and reflect strong colors. Pigments are generally the major means to create color. The fine structure of the scales often creates structural coloration. For instance, the lower lamina of the scales universally acts as a thin film reflector [10, 11]. Further- more, the ridges of the upper lamina are built of lamel- lae, which in most Morpho species have extensive * Correspondence: bodo.wilts@unifr.ch 1Adolphe Merkle Institute, University of Fribourg, Chemin des Verdiers 4, CH-1700 Fribourg, Switzerland Full list of author information is available at the end of the article Wilts et al. Frontiers in Zoology (2016) 13:36 Page 2 of 12 Page 2 of 12 Page 2 of 12 overlap [12, 13]. The stacked lamellae then act as optical multilayers, which create the butterflies’ striking blue- metallic reflections. Multilayers with perforations exist in the scale lumen of many lycaenids [14, 15]. Intricate structures, acting as three-dimensional photonic crystals, have been demonstrated in several lycaenid as well as papilionid species [16–19]. Results Overall appearance of Trogonoptera brookiana The upperside of the wings of the male Rajah Brooke’s birdwing, T. brookiana, is marked by mainly jet-black margins, which on the dorsal forewings surround a row of seven tooth-shaped green areas and on the hindwings border a large green patch. A bright red collar exists in the border area between head and thorax (Fig. 1a). p p p We recently reported the coloration mechanisms of the Ornithoptera, a genus of birdwings closely related to Trogonoptera sp. [1], with colorful scales consisting of a large membrane stack that acts as a chirped multilayer [7], where the reflected light is filtered by papiliochrome pig- ments. A previous study also reported the intriguingly complex ultrastructure of the strongly green reflecting wing scales of the birdwing butterfly T. brookiana [20]. The anatomy of the birdwing scales has inspired the pro- duction of advanced materials via replication into inor- ganic materials [21–24]. However, we found that the latter papers contain confusing data and erroneous identifica- tions of the scales of T. brookiana. Here we specifically focus at the optics of T. brookiana’s green reflecting scales, but we also analyze the interplay of pigmentary and struc- tural coloration realized in other wing scale types, using spectrophotometry, imaging scatterometry, and transmis- sion and scanning electron microscopy. We demonstrate that the lower lamina of the white and black scales acts as a thin film blue reflector and that the ridges of the blue scales and red hairs function as multilayer reflectors. The uniquely structured green-iridescent scales contain com- plex three-dimensional photonic crystals, the optical sig- nature of which could be understood by applying finite- difference time-domain modelling. The underside of the wings is mainly dark-brown, with a few radiating, vivid-blue lines. A row of yellow-green chevrons with blue borders is present in the middle of the forewings. The hindwings are mainly black with at the outer rim a row of white spots. Red patches sur- round the thorax (Fig. 1b). We measured the reflectance spectra of the various colored areas with a bifurcated fiber-probe spectropho- tometer, which yielded intriguing spectral differences, es- pecially for the green patches (Fig. 1c-f). Below we describe the very different pigmentary and structural as- pects that contribute to the different colors, and we ex- plain the resulting optical effects. Thin films in the white and black scales Light microscopy of a white area (Fig. 1b, #1) shows that the color resides in a lattice of whitish cover scales over- lapping ground scales that are either white or black (Fig. 2a, b). In the latter case the cover scales have a bluish hue. Interestingly, the scales of the brown-black areas (Fig. 1a, b, #2, 3; Fig. 2b) also display a bluish hue, although much weaker. Reflectance spectra measured with a microspectrophotometer revealed the pigmentary and/or structural origin of the different colors. The Fig. 1 Rajah Brooke’s birdwing, Trogonoptera brookiana, and reflectance spectra. a, b Photographs of the upper (dorsal) and lower (ventral) side of a male. c-f Reflectance spectra; the number of the spectra corresponds to the wing and body location indicated in (a, b). Scale bar: a, b 5 cm Fig. 1 Rajah Brooke’s birdwing, Trogonoptera brookiana, and reflectance spectra. a, b Photographs of the upper (dorsal) and lower (ventral) side of a male. c-f Reflectance spectra; the number of the spectra corresponds to the wing and body location indicated in (a, b). Scale bar: a, b 5 cm Page 3 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 Fig. 2 White and black scales. a, b Epi-illumination light micrographs of wing areas with white (area 2 of Fig. 1b) and black cover scales (area 3 of Fig. 1b). c, d Scanning electron micrographs of a white and a black scale, respectively. e Cross-sectional SEM image of a white scale showing the basic scale architecture with the lower lamina of thickness ~200 nm (arrowhead). f Reflectance spectra; spectrum 1 is from a white scale overlapping a white scale, and spectrum 2 is from a central area of a white scale above a black ground scale, as indicated in (a), while spectrum 3 is from a central area of a black scale above a black ground scale, as indicated in (b). g Scatterogram of a white scale. The red circles indicate directional angles of 5°, 30°, 60°, and 90°. Scale bars: a, b 100 μm, c-e 2 μm Fig. 2 White and black scales. a, b Epi-illumination light micrographs of wing areas with white (area 2 of Fig. 1b) and black cover scales (area 3 of Fig. 1b). c, d Scanning electron micrographs of a white and a black scale, respectively. Thin films in the white and black scales e Cross-sectional SEM image of a white scale showing the basic scale architecture with the lower lamina of thickness ~200 nm (arrowhead). f Reflectance spectra; spectrum 1 is from a white scale overlapping a white scale, and spectrum 2 is from a central area of a white scale above a black ground scale, as indicated in (a), while spectrum 3 is from a central area of a black scale above a black ground scale, as indicated in (b). g Scatterogram of a white scale. The red circles indicate directional angles of 5°, 30°, 60°, and 90°. Scale bars: a, b 100 μm, c-e 2 μm which strongly reduces the light flux reaching the lower lamina as well as filters the reflected light flux. The lower lamina’s thickness of ~200 nm, suggested by the reflectance spectrum, was confirmed by scanning elec- tron microscopy (Fig. 2e). spectrum measured from a white cover scale overlapping a white ground scale showed a slightly undulating, broad-band spectrum (Fig. 2f, #1) similar as that mea- sured with the bifurcated fiber probe (Fig. 1c, #1). How- ever, a white cover scale overlapping a black ground scale yielded a very different, blue-peaking spectrum (Fig. 2f, #2), which resembles that of a chitinous thin film reflector with thickness ~200 nm [10, 25]. As demonstrated for several butterfly wing scales, the can- didate thin film reflector is undoubtedly the scale’s lower lamina [10, 11]. Thin film reflectors, like those encountered in soap bubbles, usually reflect incident light very directionally, but this is not the case for the scales when illuminated from the side of the upper lamina, because the disor- dered cross-rib lattice acts as a diffuser. This was dir- ectly demonstrated by imaging scatterometry, using a narrow aperture illumination, which yielded a broad, dif- fuse pattern with a faint horizontal line caused by light diffraction at the ridge grating (Fig. 2g). To ascertain the structure of the scales of the whitish wing areas of T. brookiana, we performed scanning elec- tron microscopy. The white cover and ground scales as well as the black ground scales showed the common pa- pilionid wing scale structure, with a disordered cross-rib lattice connecting regularly spaced ridges (Fig. 2c-e; see further [5, 26, 27]). Pigmentary and structurally colored red hairs The surface of the hairs featured prominent longitudinal ridges, very similar to the ridges of the white and black wing scales (e.g., Fig. 2c-e). How- ever, in the cylindrical hairs the ridges are present along the complete circumference, while in the wing scales the ridges are only present in the upper lamina; the lower lamina is an approximately flat plate. As in the wing scales, the ridges of the hairs consist of overlapping la- mellae. Notably however, the lamellae of the collar hairs overlap each other much more than those of the ventral hairs. This reminds us of the case of the yellow and or- ange wing scales of various pierid butterflies where over- lapping ridge lamellae create UV-reflecting multilayers [29–32]. The UV-reflectance peak of the collar hairs has clearly the same structural basis. As noted above, imaging scatterometry of wing scales with multilayered ridges generally yield characteristic, line-shaped scatterograms. However, scatterograms of the blue scales of T. brookiana had a rather different ap- pearance. Illumination with a narrow aperture beam yielded a reflected light beam with a restricted spatial distribution (Fig. 4c), indicating a special scale structure that reflects directionally. We performed electron microscopy to unravel the structural basis of the blue scales’ colors. Somewhat sur- prisingly, scanning electron microscopy showed that the scale ridges consist of elaborate stacks of lamellae, simi- lar to the well-known Morpho ridge multilayers (Fig. 4d). Transmission electron microscopy confirmed this, but revealed a material organization of the ridges more com- plex than that of the Morpho scales. At the ridge interior a highly electron dense medium exists, suggesting local deposition of melanin (Fig. 4e). Light microscopic obser- vations as well as microspectrophotometry on blue scales immersed in refractive index matching fluid in- deed revealed a substantial amount of melanin. y Imaging scatterometry on wing scales with ridge mul- tilayers demonstrated that the parallel ridges can act as a diffraction grating [32–34] (see also Fig. 2g). To in- vestigate whether the hairs behave similarly, we per- formed scatterometry on single, isolated hairs (Fig. 3d). The scatterograms obtained from the two hair types both showed a colorful, line-shaped pattern perpen- dicular to the hair axis, clearly due to diffraction by the parallel ridges. The additional vague, diffuse red back- ground is due to randomly scattered light filtered by the short-wavelength absorbing pigment in the con- necting cross-ribs and additional underlying structures (cf. Pigmentary and structurally colored red hairs Light microscopic inspection of the red-colored ventral areas around the abdomen (Fig. 1b, #4) and the red col- lar area (Fig. 1a, #5) revealed very similar, densely- packed hairs (Fig. 3a). The reflectance spectra of both areas (Fig. 1d) showed a high reflectance at long wave- lengths, consistent with papiliochrome pigment absorb- ing up to the red wavelength range [5, 26]; the spectral The reflectance spectra measured of black scales (Fig. 2f, #3) resembled that of a white cover scale when on a black ground scale (Fig. 2f, #2) except for a much lower amplitude. The latter is evidently due to a high concentration of melanin in the scales’ upper lamina, Page 4 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 Fig. 3 Red hairs. a Epi-illumination light micrograph of the red hairs near the abdomen (area 4 of Fig. 1b); scale bar 100 μm. b, c Scanning electron micrograph of red hairs of the area near the abdomen and collar, respectively; scale bar: 5 μm. d Scatterogram of a red hair Fig. 3 Red hairs. a Epi-illumination light micrograph of the red hairs near the abdomen (area 4 of Fig. 1b); scale bar 100 μm. b, c Scanning electron micrograph of red hairs of the area near the abdomen and collar, respectively; scale bar: 5 μm. d Scatterogram of a red hair Multilayered ridges in the blue scales separation of spectra # 4 and #5 in Fig. 1d is most probably due to a higher amount of pigment in the hairs in the collar region. However, this explanation is at variance with the high UV-reflectance of the collar hairs (Fig. 1d, #5), which therefore suggests a struc- tural origin. Epi-illumination light microscopy of the blue patches (Fig. 1b, #6, 7) showed that the scale lattice consists of blue-colored cover scales overlapping black ground scales (Fig. 4a). The color of the individual blue scales rather varies, which is of course reflected in the mea- sured reflectance spectra (Fig. 4b). The narrow bands and fine structure of the spectra closely resemble the reflectance spectra of the blue wing scales of Morpho and pierid butterflies that have multilayered ridges [12, 31, 35]. To further investigate this, we performed scanning electron microscopy (Fig. 3b, c). As found in other cases, the ventral and collar hairs have a more or less cylin- drical shape [9, 28]. Pigmentary and structurally colored red hairs the scatterogram of Hebomoia glaucippe; Fig. 3c of ref. [32]). The melanin deposition in the center of the ridges will act as an optical isolation mechanism, which thus ex- plains the restricted spatial spread of reflected light in the scatterogram (Fig. 4c). Below we will recognize a similar and even much stronger isolation mechanism in the green scales. Page 5 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 Fig. 4 Optics and ultrastructure of the blue wing scales. a Epi-illumination light micrograph of an area with variously blue colored scales (area 6 in Fig. 1b). b Reflectance spectra of scales with numbers corresponding to those in panel (a). c Scatterogram (the red circles indicate scattering angles of 5, 30, 60, and 90°. d Scanning electron micrograph of a blue scale, showing high ridges. e Transmission electron micrograph of a blue scale, with an electron-dense core of the ridges. Scale bars: a 100 μm, d, e 1 μm Fig. 4 Optics and ultrastructure of the blue wing scales. a Epi-illumination light micrograph of an area with variously blue colored scales (area 6 in Fig. 1b). b Reflectance spectra of scales with numbers corresponding to those in panel (a). c Scatterogram (the red circles indicate scattering angles of 5, 30, 60, and 90°. d Scanning electron micrograph of a blue scale, showing high ridges. e Transmission electron micrograph of a blue scale, with an electron-dense core of the ridges. Scale bars: a 100 μm, d, e 1 μm Complex photonics of the green scales cross-section observed with scanning electron micros- copy (Fig. 5e). Co p e p oto cs o t e g ee sca es The strongly variable reflectance spectra of the green patches (Fig. 1f, #8–10) betrayed complex structural properties of the wing scales. Epi-illumination light mi- croscopy showed strongly curved scales with locally very bright reflections (Fig. 5a). Scanning electron microscopy of single scales demonstrated pronounced ridges, but the space in between the ridges was filled with an almost continuous layer, except for a narrow striped burrow (Fig. 5b, c). Interestingly, the ridge centers of the green scales con- tain a highly electron-dense material (Fig. 5d), like the blue scales (Fig. 4e). The suggested presence of melanin was demonstrated by embedding a scale in immersion oil and observing it with a light microscope using trans- mitted light (Fig. 5f). Pigmentary and structurally colored red hairs a, b Local epi-illumination of a green scale with normally-incident light linearly-polarized perpendicular (TE) and parallel (TM) to the ridges. c Reflectance spectra corresponding to panel (a) and (b). d, e Photographs of a dorsal right wing under about perpendicular and oblique (~45°) illumination. Scale bars: a, b 5 μm, d, e 1 cm Fig. 5 Structure of the green-colored wing scales. a Epi-illumination light micrograph of a green area (#10 of Fig. 1a). b Scanning elecron micrograph of a single green scale. c Close-up on-view SEM image showing ridges and minute gaps between them. d TEM cross- sectional image of a green cover scale and a black ground scale. e Cross-sectional SEM image of a green scale showing the extreme order of multiple ridges. d An isolated green wing scale immersed in refractive index matching fluid observed in transmitted light. Scale bars: a 100 μm. b 50 μm, c-f: 2 μm Fig. 6 Polarizing optics of the green scales. a, b Local epi-illumination of a green scale with normally-incident light linearly-polarized perpendicular (TE) and parallel (TM) to the ridges. c Reflectance spectra corresponding to panel (a) and (b). d, e Photographs of a dorsal right wing under about perpendicular and oblique (~45°) illumination. Scale bars: a, b 5 μm, d, e 1 cm Fig. 6 Polarizing optics of the green scales. a, b Local epi-illumination of a green scale with normally-incident light linearly-polarized perpendicular (TE) and parallel (TM) to the ridges. c Reflectance spectra corresponding to panel (a) and (b). d, e Photographs of a dorsal right wing under about perpendicular and oblique (~45°) illumination. Scale bars: a, b 5 μm, d, e 1 cm illumination on the central, flat part of the scale (diam- eter of illumination spot ~10 μm; Fig. 7a, arrow b) cre- ated a remarkably clear diffraction pattern (Fig. 7b). The diffraction pattern’s first order for 500 nm light was ~45°, which corresponds to a grating parameter of 0.77 μm, in excellent agreement with the measured ridge distance. Illumination with larger spots (diameter 30 and 60 μm; Fig. 7a, arrows c and d) yielded fuzzier observed by illuminating a wing from different angular directions; huge shifts in hue can be elicited that way (Fig. 6d, e). The distinct angle dependence of the scale reflectance might be fully due to the scale’s fine structure but could also be prominently affected by the scale’s strong curva- ture. Pigmentary and structurally colored red hairs a, b Local epi-illumination of a green scale with normally-incident light linearly-polarized perpendicular (TE) and parallel (TM) to the ridges. c Reflectance spectra corresponding to panel (a) and (b). d, e Photographs of a dorsal right wing under about perpendicular and oblique (~45°) illumination. Scale bars: a, b 5 μm, d, e 1 cm Fig. 5 Structure of the green-colored wing scales. a Epi-illumination light micrograph of a green area (#10 of Fig. 1a). b Scanning elecron micrograph of a single green scale. c Close-up on-view SEM image showing ridges and minute gaps between them. d TEM cross- sectional image of a green cover scale and a black ground scale. e Cross-sectional SEM image of a green scale showing the extreme order of multiple ridges. d An isolated green wing scale immersed in refractive index matching fluid observed in transmitted light. Scale bars: a 100 μm. b 50 μm, c-f: 2 μm Fig. 6 Polarizing optics of the green scales. a, b Local epi-illumination of a green scale with normally-incident light linearly-polarized perpendicular (TE) and parallel (TM) to the ridges. c Reflectance spectra corresponding to panel (a) and (b). d, e Photographs of a dorsal right wing under about perpendicular and oblique (~45°) illumination. Scale bars: a, b 5 μm, d, e 1 cm Fig. 5 Structure of the green-colored wing scales. a Epi-illumination light micrograph of a green area (#10 of Fig. 1a). b Scanning elecron micrograph of a single green scale. c Close-up on-view SEM image showing ridges and minute gaps between them. d TEM cross- sectional image of a green cover scale and a black ground scale. e Cross-sectional SEM image of a green scale showing the extreme order of multiple ridges. d An isolated green wing scale immersed in refractive index matching fluid observed in transmitted light. Scale bars: a 100 μm. b 50 μm, c-f: 2 μm Fig. 6 Polarizing optics of the green scales. a, b Local epi-illumination of a green scale with normally-incident light linearly-polarized perpendicular (TE) and parallel (TM) to the ridges. c Reflectance spectra corresponding to panel (a) and (b). d, e Photographs of a dorsal right wing under about perpendicular and oblique (~45°) illumination. Scale bars: a, b 5 μm, d, e 1 cm Fig. 6 Polarizing optics of the green scales. Pigmentary and structurally colored red hairs This revealed dark, brown-black stripes with distance of 0.8 μm, fully corresponding with the ridge distance (Fig. 5c-e). Absorbance measurements unequivocally confirmed the local concentration of mel- anin in the ridge centers. Transmission electron microscopy revealed that the upper lamina of the green scales is indeed organized in a unique way. The lamellae fully fill the space in be- tween neighboring ridges and even touch each other, so creating an almost continuous multilayer. This organization strongly deviates from that of ordinary papilionid scales, which normally have irregular ar- ranged cross-ribs (e.g., Fig. 2c, d). Usually the laterally extending lamellae of the ridges are narrow and skewed with respect to the scale surface and have only limited overlap (compare the cross-sections of the green cover scale with the underlying black ground scale in Fig. 5d and the blue cover scale of Fig. 4d). That the elaborate ridges of the green scales have an extra-ordinary regular organization is also clear from a The strongly anisotropic, longitudinal organization of the green scales suggested a strong polarization depend- ence of the scales’ optical properties. We investigated this by first applying from about a normal direction epi- illumination of a scale with TE- and TM-polarized light, i.e., light polarized perpendicularly and parallel to the ridges, respectively. That yielded distinctly different colors (Fig. 6a, b), as was documented also in the corre- sponding reflectance spectra measured with the micro- spectrophotometer (Fig. 6c). The high reflectivity of the wing scales and the polarization dependence suggested that the wing reflection characteristics could be strongly angle dependent. This could indeed be strikingly Page 6 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 Fig. 5 Structure of the green-colored wing scales. a Epi-illumination light micrograph of a green area (#10 of Fig. 1a). b Scanning elecron micrograph of a single green scale. c Close-up on-view SEM image showing ridges and minute gaps between them. d TEM cross- sectional image of a green cover scale and a black ground scale. e Cross-sectional SEM image of a green scale showing the extreme order of multiple ridges. d An isolated green wing scale immersed in refractive index matching fluid observed in transmitted light. Scale bars: a 100 μm. b 50 μm, c-f: 2 μm Fig. 6 Polarizing optics of the green scales. Pigmentary and structurally colored red hairs Here, the shape of the scale severely affected the scattering pattern, as the strong curvature of the scale stretched the scattering pattern across the hemisphere (see also [32]). time-domain simulations on various, idealized photonic structures with varying grades of complexity (Fig. 8a). Sample parameters were extracted from electron micro- graphs, yielding a chitin layer thickness of dc = 150 nm with an air gap of da = 80 nm. The thickness and spacing of the melanized ridges was 200 nm and 800 nm, re- spectively. The regular spaced air holes (also with a mean spacing of 800 nm) were estimated to have a diameter of 105 nm and a lateral spacing of 230 nm (i.e., dc + da). We furthermore investigated the diffraction pattern as a function of the angle of illumination and the polarization. To allow this, we inserted a linear polarizer and a pinhole into the secondary illumination beam of the imaging scat- terometer, allowing illumination with polarized light from angles of incidence between 0° and 90°. Figure 7e, f shows the angle-dependency for TE- and TM-polarized light, re- spectively. As expected, when the angle of incidence in- creases, the diffraction pattern also shifts to larger angles. More importantly, the color of the 0th order reflection strongly varies with polarization and angle of incidence. Unexpectedly, the color change of the 0th order reflection deviates from the typical iridescence of a multilayer, i.e., a blue shift with increasing angle of incidence). Somewhat erratically, the reflection first shifts to the red (with a max- imum in-between 40 and 50°) before shifting to the blue. The angle-dependency of the diffraction pattern, i.e., the 0th, 1st and -1st diffraction order, can be well understood as to be from a grating with a ridge distance of ~0.77 μm (shaded bands overlaid in Fig. 7e). Normal illumination of an ideal chitin-air multilayer structure (Fig. 8a, i) causes a polarization-independent reflectance spectrum with a maximum of ~0.99 at 630 nm (i.e., peak wavelength λmax = 2 (dcnc + da) with a chitin refractive index nc ≈1.55 [25]). Regular-spaced, or- thogonal air slabs in the chitin-air multilayer (Fig. 8a, ii) invoke a minor form birefringence, but still a saturating reflectance at ~620 nm, with small oscillations in the re- flectance spectrum (Fig. 8c). Putting melanin-containing baffles in the chitin-air multilayer (Fig. 8a, iii) has a se- vere optical effect, however, causing a pronounced polarization-dependency. Pigmentary and structurally colored red hairs To clarify this, we performed imaging scatterome- try on a single, isolated green scale (Fig. 7). Local Page 7 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 Fig. 7 Imaging scatterometry of a green scale. a Side-view of a single scale glued to the end of a pulled glass micropipette. The width of the rectangles indicate the size of the illumination area used in panels b-d. b-d Scatterograms resulting from illumination of a small (b), medium- sized (e) and large (f) area. e, f Angle dependence of the diffraction patterns obtained with a narrow-aperture illumination beam (as in panel b) for TE- and TM-polarized light with the angle of light incidence increasing in ~7.5° steps. g, h Calculated angle dependence of the full photonic structure (see Fig. 8a, iv) for TE- and TM-polarization with the angle of light incidence varying from 0 to 80° in 10° steps Fig. 7 Imaging scatterometry of a green scale. a Side-view of a single scale glued to the end of a pulled glass micropipette. The width of the rectangles indicate the size of the illumination area used in panels b-d. b-d Scatterograms resulting from illumination of a small (b), medium- sized (e) and large (f) area. e, f Angle dependence of the diffraction patterns obtained with a narrow-aperture illumination beam (as in panel b) for TE- and TM-polarized light with the angle of light incidence increasing in ~7.5° steps. g, h Calculated angle dependence of the full photonic structure (see Fig. 8a, iv) for TE- and TM-polarization with the angle of light incidence varying from 0 to 80° in 10° steps Fig. 7 Imaging scatterometry of a green scale. a Side-view of a single scale glued to the end of a pulled glass micropipe rectangles indicate the size of the illumination area used in panels b-d. b-d Scatterograms resulting from illumination of sized (e) and large (f) area. e, f Angle dependence of the diffraction patterns obtained with a narrow-aperture illuminatio for TE- and TM-polarized light with the angle of light incidence increasing in ~7.5° steps. g, h Calculated angle depende structure (see Fig. 8a, iv) for TE- and TM-polarization with the angle of light incidence varying from 0 to 80° in 10° steps scatterograms (Fig. 7c, d), although the principal diffrac- tion pattern remained well-recognizable. Pigmentary and structurally colored red hairs The TE-reflectance spectrum shows a distinct peak in the green, at ~540 nm, with sidelobes at larger wavelengths, whereas the multi-lobed TM-reflectance spectrum peaks in the red, at ~720 nm (Fig. 8d). The multilayer with both the regular spaced air holes and the melanized baffles (Fig. 8a, iv) is a photonic structure as that found in the butterfly scales (Fig. 5). The spectra obtained by FDTD modelling (Fig. 8e) are FDTD modelling of the photonic response The TE-polarized reflectance is maximal in the green, at ~530 nm, with minor sidelobes in the yellow- red wavelength range, whereas the TM-polarized reflect- ance spectrum has multiple peaks with very similar reflectance values of ~0.3 at 520, 610 and 700 nm, re- spectively. The angle-dependent scattering calculated for this optical structure (Fig. 7g, h) corresponds closely to the experimentally observed scattering, showing strong grating-like diffraction (Fig. 7e, f). indeed very similar to those measured experimentally (Fig. 6c). The TE-polarized reflectance is maximal in the green, at ~530 nm, with minor sidelobes in the yellow- red wavelength range, whereas the TM-polarized reflect- ance spectrum has multiple peaks with very similar reflectance values of ~0.3 at 520, 610 and 700 nm, re- spectively. The angle-dependent scattering calculated for this optical structure (Fig. 7g, h) corresponds closely to the experimentally observed scattering, showing strong grating-like diffraction (Fig. 7e, f). FDTD modelling of the photonic response To understand the role of the various optical compo- nents in the green scales, we performed finite-difference Page 8 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 Fig. 8 FDTD modelling of multilayered scale structures. a Sketches of four investigated topologies: (i) ideal multilayer of chitin (cyan) and air spaces, (ii) multilayers with air holes (white), (iii) multilayers with orthognal melanin baffles (brown), and (iv) the full structure. b-e Reflectance spectra of the four cases for TE- and TM-polarized light Fig. 8 FDTD modelling of multilayered scale structures. a Sketches of four investigated topologies: (i) ideal multilayer of chitin (cyan) and air spaces, (ii) multilayers with air holes (white), (iii) multilayers with orthognal melanin baffles (brown), and (iv) the full structure. b-e Reflectance spectra of the four cases for TE- and TM-polarized light Thin film reflectors in black and white scales It is commonly assumed that the color of pigmented butterfly wing scales is due to the scattering of light by the wing scale’s irregular structures and that the wavelength-dependent absorbing pigment determines the scale’s color, because it acts as a high-pass spectral filter on the scattered light. We found that the white and black wing scales of T. brookiana have widely open win- dows and irregular cross-ribs (Fig. 2c, d). The lower lam- ina, acting as a thin film reflector, determines the reflectance spectrum of the white scales. The peak re- flectance is at ~450 nm, indicating a lower lamina thick- ness of ~200 nm (Fig. 2e). This falls well in line with recent observations across all major butterfly families of the presence of thin film reflectors in pigmented wing scales [4, 8, 10, 11, 13]. For instance, the blue eye spots in the wings of the Peacock butterfly, Inachis io (Nym- phalidae), were found to be created by pigmentless cover scales overlapping black ground scales, where the cover scales’ lower lamina acts as a blue-reflecting thin film. On the other hand, wing areas of the Peacock and other nymphalines where both cover and ground scales are pigmentless have a distinctly white color as a result from cumulative reflections of cover and ground scales as well as the wing substrate. The reflectance spectra of Fig. 1c indeed very similar to those measured experimentally (Fig. 6c). Complex photonic structures in Trogonoptera: ordered Morpho-type photonic structures The photonics of the blue and green wing scales is re- markably complex and seems to have uniquely evolved in the animal kingdom. Compared to the well-known Morpho butterflies, two main differences are obvious: the order of the ridges is extremely high, and baffles of melanin within the ridges affect the optical characteris- tics (Figs. 4 and 5). Morpho butterflies show extreme blue iridescence due to a disordered multilayer, with se- vere height and position disorder of the ridge layers [36–38], causing the iconic Christmas tree structure of transmission electron micrographs [12]. Light diffraction by the ridges causes the line-like pattern in scattero- grams [34]. g g p Whether or not the wing scale colors are tuned by evolutionary selection for intraspecific recognition is a challenging question. Sexual dichromatism is present in birdwing butterflies where females generally possess a smaller number of color markings of smaller wing area compared to the males. In Trogonoptera brookiana the sexual dichromatism is far less distinctive than that of other sympatric birdwing species (e.g., Troides or Ornithoptera [7]). Female T. brookiana carry green wing patches on the dorsal wings, quite similar to the male, however in females these green areas are associated with an extensive white pattern. Furthermore, the discodial area of the hindwing is completely blue in the hindwing, but in the male these blue lines are restricted to the basal areas of the wing. When in motion, these different patterns will most likely emit quite a different signal. What is the optical advantage of the extreme order in the green wing scales of Trogonoptera brookiana? First of all, the ordered ridges invoke form birefringence resulting in a strong polarization-dependent reflectance (Figs. 6, 7 and 8), which is much less in butterflies with similar, disordered architectures. Secondly, the order in- vokes a grating-like, spatially restricted reflection pat- tern, resulting in extreme color changes with minute changes in viewing angle (Fig. 6d, e). Similar optical ef- fects have been noticed in other organisms, like birds- of-paradise, but the underlying structures have very different topologies [39, 40]. Thirdly, the presence of melanin in the ridges adds a novel optical mechanism that severely changes the spectral response and shapes the spatial reflections. Sexual dichromatism functions in mate recognition, enabled by a rich set of spectral photoreceptors [46]. Biological significance g g The coloration of Rajah Brooke’s birdwings appears to be developed in different ways. As noted above, each scale type is optimized for a specific signaling purpose using various photonic structures. The prominent tooth- shaped green markings that are contrasted by the jet- black wing borders will likely serve as a potential aposematic signal to predators, similar to other strongly colored markings of related papilionids [2, 7, 42–44]. Es- pecially so, since larval stages of T. brookiana have poi- sonous Aristolochiaceae as their main foodplants [1]. Furthermore, the Trogonoptera brookiana phenotype is unique amongst the birdwing butterflies and the striking pattern is probably associated to its unique communal be- haviour where males often assemble in large groups [45]. Discussion The coloration toolkit of the male Rajah Brooke’s Bird- wing, Trogonoptera brookiana, relies on a diversity of nanostructures, which, in combination with various ab- sorbing pigments, create the strikingly vivid coloration of the butterfly. The optics of the pigmentary colored scales can be well understood as a basic combination of a thin film reflector and a strongly absorbing pigment. Papiliochrome pigment and multilayers determine the reflection properties of the red hairs; complex structured ridge multilayers and melanin-containing, absorbing baf- fles play a central role in the optics of the blue and green wing scales. Wilts et al. Frontiers in Zoology (2016) 13:36 Page 9 of 12 and Fig. 2f are very similar to those measured in the blue and white wing areas of the nymphalines as well as in another papilionid, Papilio xuthus [26]. Yet, whereas the lower lamina of the blue scales of the nymphaline butterflies was a single, chitinous layer, the lower lamina of the bluish scales of P. xuthus was a multilayer consist- ing of two membranes with an air gap in between the membranes. In all cases, the scale’s reflectance spectrum, which depends on the thin film thickness (~150– 250 nm), appears to be well-tuned to the absorbance spectrum of the scale’s pigment [4, 8, 10, 13, 26]. We note here that the anatomy of the black scales of T. brookiana is almost identical to that of the white scales, and their lower lamina also acts as a blue thin film re- flector. However, a high concentration of melanin se- verely reduces the reflectance, thus causing the scales’ blackness. shifting and spreading the viewing direction, as was demonstrated in some pierid [32] or riodinid [41] butter- fly species having photonic structures similar to those encountered in Trogonoptera brookiana. Complex photonic structures in Trogonoptera: ordered Morpho-type photonic structures Most likely Troides and Trogonoptera, which are in the same tribe (Troidini) in the family of Papilionidae [1, 2], evolved quite similar sets of spectral receptors. For the Golden Birdwing butterfly, Troides aeacus formosanus, Chen et al. [47] determined by intracellular recordings seven different photoreceptor types, with spectral sensi- tivities ranging from the UV to red. A comparison of the reflectance spectra of the colored wing areas in T. brookiana with the spectral sensitivities of the different photoreceptors indicates a stark spectral contrast of the various wing areas with the surrounding black framing of melanized scales (Fig. 1a, b). This indeed suggests that the wing colors are tuned to the butterfly’s visual system; similar to what has been previously observed in related butterflies [7]. The rather extreme scale curvature of the green scales, with their very convex tip, spreads the light reflected by the grating structure into the hemisphere along the long axis of the scale (Fig. 7a-d). Scale curvature is not un- usual in structured butterfly wing scales and is usually attributed to maximizing the viewing angle and/or Page 10 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 Wilts et al. Frontiers in Zoology (2016) 13:36 Wing scale development controlled structure may well provide inspiration for biomimetic applications. g p The degree of order in the green and blue scales, but also the multilayering of the red bristles, is quite remark- able. All butterfly wing scales are hypothesized to de- velop via a similar pathway by the (out-)folding of cell membranes and organelles (see Ghiradella’s seminal studies [9, 29, 48]) as well as by the preferential align- ment of intracellular F-actin fiber networks [49]. For the folding of multilayered ridges, Ghiradella’s observations of developing wing scales indicate that the multilayers are formed by elastic buckling of the cell membrane and subsequent backfilling with nascent cuticle [9, 29, 48]. Drying of this chitinous cuticle after cell apoptosis re- sults in the final, highly anisotropic cell shape [49], as those observed in the SEM images of Figs. 2, 3, 4 and 5. The deposition of the melanin in the ridge centers will most likely happen subsequent to the buckling process. It will be of extreme interest to further investigate which cellular parameters control or drive the cellular pro- cesses on the nanoscale given the unusual degree of order observed in these wing scales, especially because the order must be the result of evolutionary selection. Transmission electron microscopy For transmission electron microscopy (TEM) of the scales, wing parts were prefixed in 2 % paraformaldehyde and 2.5 % glutaraldehyde in 0.1 mol l−1 sodium cacodylate buf- fer (CB, pH 7.3) for ∼45 min. After dehydrating with a graded series of ethanol and infiltration with propylene oxide, the tissues were embedded in Spurr's resin. The tis- sues were cut into 50 nm ultrathin sections, double- stained with uranyl acetate and lead citrated and observed using a Hitachi H7650 (Tokyo, Japan) transmission elec- tron microscope (as outlined in [34]). Bio-inspired replicas Recently extensive nanotechnological attempts have been undertaken to mimic the photonic structures of butterfly scales in materials with novel optical effects and/or ad- vanced functionalities. The exact fabrication of complex three-dimensional topologies is still beyond current nano- fabrication capabilities, however [50–52]. The wing scales of T. brookiana have served as templates for creating light trapping structures based on the black wing scales via an inverse SiO2 replica [21, 23, 24]. Although improperly de- scribed at some points, the latter authors show that repli- cation of the black scales results in black optical structures with similar morphology. The green wing scales could inspire the production of highly efficient gas sen- sors, due to the large surface area, similar to the sensors based on the well-structured, multilayered wing scales of Morpho butterflies [53–55]. Spectroscopy fl Reflectance spectra of the wings of intact butterflies were measured (in air) with a bifurcated fiber-optic probe. The probe comprised six light guides, delivering light from a halogen-deuterium source (AvaLight-D(H)-S-bal; Avantes, Eerbeek, the Netherlands), which surrounded a central fiber that acts as a light collector of reflected light (of a spot with diameter ~1 mm) and which delivered it to a fiber optic spectrometer (Maya2000Pro; Ocean Optics, Duiven, the Netherlands). A white diffusing reflectance standard (Ocean Optics WS-1) served as the reference. Absorbance spectra of single wing scales immersed in refractive index matching fluid as well as polarization- dependent reflectance spectra of single scales were mea- sured with a custom-built microspectrophotometer. The light beam of a xenon light source was coupled with a quartz lens into the microscope, equipped with an Olympus 20x/0.45 objective. The spectral range of the microspectrophotometer was limited to wavelengths ≥ 360 nm. Scanning electron microscopy The ultrastructure of the wing scales was investigated with a Tescan MIRA 3 LMH field-emission scanning electron microscope (Tescan, Brno, Czech Republic). To prevent charging, the samples were sputtered with a thin layer of palladium or gold prior to imaging. Methods Specimen p The investigated specimens of Rajah Brooke’s Birdwing, Trogonoptera brookiana albescens (Rothschild, 1895) (Lepidoptera: Papilionidae: Troidini), were obtained from Worldwide Butterflies (Dorset, UK; www.wwb.co.uk). p The investigated specimens of Rajah Brooke’s Birdwing, Trogonoptera brookiana albescens (Rothschild, 1895) (Lepidoptera: Papilionidae: Troidini), were obtained from Worldwide Butterflies (Dorset, UK; www.wwb.co.uk). Conclusions In conclusion, the coloration mechanisms of Rajah Brooke’s Birdwing consist of thin films, multilayers and higher dimensional photonic structures that are ex- tremely ordered at the nanoscale, resulting in the birdw- ing’s stunning coloration. We for the first time observed structural color in butterfly hairs and observed that the wing scales of the showy green patches provide yet an- other example of uniquely arranged structures that cre- ate reflection patterns with strong polarization contrast, thus expanding our insight into biophotonic coloration, especially in insects. The birdwings’ novel, nanoscale- Imaging scatterometry The hemispherical far-field light scattering pattern of single scales was visualized with an imaging scatterom- eter [14, 34, 56]. The scatterometer is built around an Wilts et al. Frontiers in Zoology (2016) 13:36 Page 11 of 12 Page 11 of 12 Page 11 of 12 ellipsoidal mirror, which collects light from a full hemi- sphere around its first focal point, where the sample is positioned. Illumination was with a white light source (a xenon lamp), which delivered a narrow aperture (~5°) beam. For polarisation-dependent measurements, a lin- ear sheet polarizer was added to the light path of the secondary beam. A small piece of magnesium oxide served as a white diffuse reference object. Images were acquired with an Olympus DP-70 camera and were sub- sequently corrected for geometrical distortions using a MATLAB routine. 4. Stavenga DG, Leertouwer HL, Wilts BD. Coloration principles of nymphaline butterflies - thin films, melanin, ommochromes and wing scale stacking. J Exp Biol. 2014;217:2171–80. 4. Stavenga DG, Leertouwer HL, Wilts BD. Coloration principles of nymphaline butterflies - thin films, melanin, ommochromes and wing scale stacking. J Exp Biol. 2014;217:2171–80. 5. Wilts BD, Trzeciak TM, Vukusic P, Stavenga DG. Papiliochrome II pigment reduces the angle-dependency of structural wing colouration in nireus group papilionids. J Exp Biol. 2012;215:796–805. 5. Wilts BD, Trzeciak TM, Vukusic P, Stavenga DG. Papiliochrome II pigment reduces the angle-dependency of structural wing colouration in nireus group papilionids. J Exp Biol. 2012;215:796–805. 6. Umebachi Y. Papiliochrome, a new pigment group of butterfly. Zool Sci. 1985;2:163–74. 6. Umebachi Y. Papiliochrome, a new pigment group of butterfly. Zool Sci. 1985;2:163–74. 7. Wilts BD, Matsushita A, Arikawa K, Stavenga DG. Spectrally tuned structural and pigmentary coloration of birdwing butterfly wing scales. J R Soc Interface. 2015;12:20150717. 7. Wilts BD, Matsushita A, Arikawa K, Stavenga DG. Spectrally tuned structural and pigmentary coloration of birdwing butterfly wing scales. J R Soc Interface. 2015;12:20150717. 8. Stavenga DG, Leertouwer HL, Wilts BD. The colouration toolkit of the Pipevine Swallowtail butterfly, Battus philenor: thin films, papiliochromes, and melanin. J Comp Physiol A. 2014;200:547–61. y 9. Ghiradella H. Insect cuticular surface modifications: Scales and other structural formations. Adv Insect Physiol. 2010;38:135–80. 9. Ghiradella H. Insect cuticular surface modifications: Scales and other structural formations. Adv Insect Physiol. 2010;38:135–80. Funding Thi This research was partly supported through the National Centre of Competence in Research “Bio-Inspired Materials” and the Adolphe Merkle Foundation (to BDW) and the Air Force Office of Scientific Research/ European Office of Aerospace Research and Development AFOSR/EOARD (grant FA9550-15-1-0068 to DGS). 18. Saranathan V, Osuji CO, Mochrie SG, Noh H, Narayanan S, Sandy A, Dufresne ER, Prum RO. Structure, function, and self-assembly of single network gyroid (I4132) photonic crystals in butterfly wing scales. Proc Natl Acad Sci U S A. 2010;107:11676–81. 19. Schröder-Turk GE, Wickham S, Averdunk H, Brink F, Fitz Gerald JD, Poladian L, Large MC, Hyde ST. The chiral structure of porous chitin within the wing- scales of Callophrys rubi. J Struct Biol. 2011;174:290–5. Acknowledgments W h k We thank two reviewers for constructive comments, Dr. Nicholas Roberts for reading an early version of the manuscript and Hein L. Leertouwer for on-going support and technical assistance. 16. Michielsen K, Stavenga DG. Gyroid cuticular structures in butterfly wing scales: biological photonic crystals. J R Soc Interface. 2008;5:85–94. 17. Poladian L, Wickham S, Lee K, Large MC. Iridescence from photonic crystals and its suppression in butterfly scales. J R Soc Interface. 2009;6:S233–42. Received: 9 June 2016 Accepted: 4 August 2016 26. Stavenga DG, Matsushita A, Arikawa K. Combined pigmentary and structural effects tune wing scale coloration to color vision in the swallowtail butterfly Papilio xuthus. Zool Lett. 2015;1:14. Finite-difference time-domain simulations 10. Stavenga DG. Thin film and multilayer optics cause structural colors of many insects and birds. Mater Today Proc. 2014;1S:109–21. 10. Stavenga DG. Thin film and multilayer optics cause structural colors of many insects and birds. Mater Today Proc. 2014;1S:109–21. The polarization-dependent light scattering of various models was simulated with three-dimensional finite- difference time-domain (FDTD) calculations. We used Lumerical FDTD Solutions 8.15, a commercial-grade Maxwell equation solver. Each model was placed in a three-dimensional simulation volume of 2x10x6 μm3. The light source covered a wavelength range of 350 to 800 nm. For the refractive index dispersion of melanin and chitin we used previously published data [25, 57]. 11. Wasik BR, Liew SF, Lilien DA, Dinwiddie AJ, Noh H, Cao H, Monteiro A. Artificial selection for structural color on butterfly wings and comparison with natural evolution. Proc Natl Acad Sci U S A. 2014;111:12109–14. 11. Wasik BR, Liew SF, Lilien DA, Dinwiddie AJ, Noh H, Cao H, Monteiro A. Artificial selection for structural color on butterfly wings and comparison with natural evolution. Proc Natl Acad Sci U S A. 2014;111:12109–14. 12. Vukusic P, Sambles JR, Lawrence CR, Wootton RJ. Quantified interference and diffraction in single Morpho butterfly scales. Proc R Soc B. 1999;266: 1403–11. 12. Vukusic P, Sambles JR, Lawrence CR, Wootton RJ. Quantified interference and diffraction in single Morpho butterfly scales. Proc R Soc B. 1999;266: 1403–11. 13. Giraldo MA, Stavenga DG. Brilliant iridescence of Morpho butterfly wing scales is due to both a thin film lower lamina and a multilayered upper lamina. J Comp Physiol A. 2016;202:381–8. 13. Giraldo MA, Stavenga DG. Brilliant iridescence of Morpho butterfly wing scales is due to both a thin film lower lamina and a multilayered upper lamina. J Comp Physiol A. 2016;202:381–8. 14. Wilts BD, Leertouwer HL, Stavenga DG. Imaging scatterometry and microspectrophotometry of lycaenid butterfly wing scales with perforated multilayers. J R Soc Interface. 2009;6:S185–92. 14. Wilts BD, Leertouwer HL, Stavenga DG. Imaging scatterometry and microspectrophotometry of lycaenid butterfly wing scales with perforated multilayers. J R Soc Interface. 2009;6:S185–92. 15. Biro LP, Kertesz K, Vertesy Z, Mark GI, Balint Z, Lousse V, Vigneron J. Living photonic crystals: butterfly scales - Nanostructure and optical properties. Mater Sci Eng C. 2007;27:941–6. Authors’ contributions 20. Prum RO, Quinn T, Torres RH. Anatomically diverse butterfly scales all produce structural colours by coherent scattering. J Exp Biol. 2006;209:748–65. All authors designed research, performed experiments and analyzed experiments. BDW and DGS wrote the paper. All authors read and approved the final manuscript. 21. Han Z, Niu S, Yang M, Zhang J, Yin W, Ren L. An ingenious replica templated from the light trapping structure in butterfly wing scales. Nanoscale. 2013;5:8500–6. Competing interests 22. Han Z, Niu S, Shang C, Liu Z, Ren L. Light trapping structures in wing scales of butterfly Trogonoptera brookiana. Nanoscale. 2012;4:2879–83. The authors declare that they have no competing interests. 23. Han Z, Li B, Mu Z, Yang M, Niu S, Zhang J, Ren L. An ingenious super light trapping surface templated from butterfly wing scales. Nanoscale. 2015;10: 1–8. Author details 1 d l h kl 1Adolphe Merkle Institute, University of Fribourg, Chemin des Verdiers 4, CH-1700 Fribourg, Switzerland. 2Biophysics Group, Institute of Physics, University of Antioquia, Calle 70 No. 52-21, A.A.1226, Medellín, Colombia. 3Computational Physics, Zernike Institute for Advanced Materials, University of Groningen, Nijenborgh 4, NL-9747AG Groningen, The Netherlands. 24. Han Z, Li B, Mu Z, Yang M, Niu S, Zhang J, Ren L. Fabrication of the replica templated from butterfly wing scales with complex light trapping structures. Appl Surf Sci. 2015;355:290–7. 25. Leertouwer HL, Wilts BD, Stavenga DG. Refractive index and dispersion of butterfly chitin and bird keratin measured by polarizing interference microscopy. Opt Express. 2011;19:24061–6. Received: 9 June 2016 Accepted: 4 August 2016 Received: 9 June 2016 Accepted: 4 August 2016 56. Vukusic P, Stavenga DG. Physical methods for investigating structural colours in biological systems. J R Soc Interface. 2009;6:S133–48. 57. Stavenga DG, Leertouwer HL, Hariyama T, De Raedt H, Wilts BD. Sexual dichromatism of the damselfly Calopteryx japonica caused by a melanin-chitin multilayer in the male wing veins. PLoS One. 2012;7:e49743. References 27. Ghiradella H. Structure and development of iridescent lepidopteran scales - the papilionidae as a showcase family. Ann Entomol Soc Am. 1985;78: 252–64. 1. Condamine FL, Sperling FAH, Wahlberg N, Rasplus J, Kergoat GJ. What causes latitudinal gradients in species diversity? Evolutionary processes and ecological constraints on swallowtail biodiversity. Ecol Lett. 2012;15:267–77. y 2. Condamine FL, Toussaint EF, Clamens A, Genson G, Sperling FA, Kergoat GJ. Deciphering the evolution of birdwing butterflies 150 years after Alfred Russel Wallace. Sci Rep. 2015;5:11860. 28. Ghiradella H. Hairs, bristles, and scales. In: Locke M, editor. Microscopic anatomy of invertebrates, vol. 11A. New York: Wiley-Liss; 1998. p. 257–87. 28. Ghiradella H. Hairs, bristles, and scales. In: Locke M, editor. Microscopic anatomy of invertebrates, vol. 11A. New York: Wiley-Liss; 1998. p. 257–87. 29. Ghiradella H, Aneshansley D, Eisner T, Silberglied RE, Hinton HE. Ultraviolet reflection of a male butterfly: interference color caused by thin-layer elaboration of wing scales. Science. 1972;178:1214–7. 29. Ghiradella H, Aneshansley D, Eisner T, Silberglied RE, Hinton HE. Ultraviolet reflection of a male butterfly: interference color caused by thin-layer elaboration of wing scales. Science. 1972;178:1214–7. 3. Wijnen B, Leertouwer HL, Stavenga DG. Colors and pterin pigmentation of pierid butterfly wings. J Insect Physiol. 2007;53:1206–17. 3. Wijnen B, Leertouwer HL, Stavenga DG. Colors and pterin pigmentation of pierid butterfly wings. J Insect Physiol. 2007;53:1206–17. Page 12 of 12 Wilts et al. Frontiers in Zoology (2016) 13:36 30. Stavenga DG, Giraldo MA, Hoenders BJ. Reflectance and transmittance of light scattering scales stacked on the wings of pierid butterflies. Opt Express. 2006;14:4880–90. 31. Wilts BD, Pirih P, Stavenga DG. Spectral reflectance properties of iridescent pierid butterfly wings. J Comp Physiol A. 2011;197:693–702. 32. Pirih P, Wilts BD, Stavenga DG. Spatial reflection patterns of iridescent wings of male pierid butterflies: curved scales reflect at a wider angle than flat scales. J Comp Physiol A. 2011;197:987–97. 33. Kinoshita S. Structural colors in the realm of nature. Singapore: World Scientific; 2008. 34. Stavenga DG, Leertouwer HL, Pirih P, Wehling MF. Imaging scatterometry of butterfly wing scales. Opt Express. 2009;17:193–202. 35. Kinoshita S, Yoshioka S, Kawagoe K. Mechanisms of structural colour in the Morpho butterfly: cooperation of regularity and irregularity in an iridescent scale. Proc R Soc B. 2002;269:1417–21. 36. Zhu D, Kinoshita S, Cai D, Cole JB. Investigation of structural colors in Morpho butterflies using the nonstandard-finite-difference time-domain method: effects of alternately stacked shelves and ridge density. References Phys Rev E. 2009;80:051924. 37. Kinoshita S, Yoshioka S, Fujii Y, Osanai M. Photophysics of structural color in the Morpho butterflies. Forma. 2002;17:103–21. 38. Kinoshita S, Yoshioka S. Structural colors in nature: the role of regularity and irregularity in the structure. ChemPhysChem. 2005;6:1443–59. 39. Stavenga DG, Leertouwer HL, Marshall NJ, Osorio D. Dramatic colour changes in a bird of paradise caused by uniquely structured breast feather barbules. Proc R Soc B. 2011;278:2098–104. 40. Wilts BD, Michielsen K, De Raedt H, Stavenga DG. Sparkling feather reflections of a bird-of-paradise explained by finite-difference time-domain modeling. Proc Natl Acad Sci U S A. 2014;111:4363–8. 41. Vukusic P, Sambles JR, Lawrence CR, Wootton RJ. Limited-view iridescence in the butterfly Ancyluris meliboeus. Proc R Soc B. 2002;269:7–14. 41. Vukusic P, Sambles JR, Lawrence CR, Wootton RJ. Limited-view iridescence in the butterfly Ancyluris meliboeus. Proc R Soc B. 2002;269:7–14. in the butterfly Ancyluris meliboeus. Proc R Soc B. 2002;269:7–14. 42. Stevens M, Merilaita S. Animal camouflage: mechanisms and function. Cambridge: Cambridge University Press; 2011. 43. Wilts BD, IJbema N, Stavenga DG. Pigmentary and photonic coloration mechanisms reveal taxonomic relationships of the Cattlehearts (Lepidoptera: Papilionidae: Parides). BMC Evol Biol. 2014;14:160. 44. Wilts BD, Michielsen K, De Raedt H, Stavenga DG. Iridescence and spectral filtering of the gyroid-type photonic crystals in Parides sesostris wing scales. Interface Focus. 2012;2:681–7. 44. Wilts BD, Michielsen K, De Raedt H, Stavenga DG. Iridescence and spectral filtering of the gyroid-type photonic crystals in Parides sesostris wing scales. Interface Focus. 2012;2:681–7. 45. Panchen A. Notes on the behaviour of Rajah Brooke’s birdwing butterfly Trogonoptera brookiana brookiana (Wallace) in Sarawak. Entomol Record. 1981;55:98–102. 45. Panchen A. Notes on the behaviour of Rajah Brooke’s birdwing butterfly Trogonoptera brookiana brookiana (Wallace) in Sarawak. Entomol Record. 1981;55:98–102. 46. Doucet SM, Meadows MG. Iridescence: a functional perspective. J R Soc Interface. 2009;6:S115–32. 46. Doucet SM, Meadows MG. Iridescence: a functional perspective. J R Soc Interface. 2009;6:S115–32. 47. Chen P, Arikawa K, Yang E. Diversity of the photoreceptors and spectral opponency in the compound eye of the Golden Birdwing, Troides aeacus formosanus. PLoS One. 2013;8:e62240. 47. Chen P, Arikawa K, Yang E. Diversity of the photoreceptors and spectral opponency in the compound eye of the Golden Birdwing, Troides aeacus formosanus. PLoS One. 2013;8:e62240. 48. Ghiradella H. Structure and development of iridescent butterfly scales: Lattices and laminae. J Morphol. 1989;202:69–88. 48. Ghiradella H. Wilts et al. Frontiers in Zoology (2016) 13:36 References Structure and development of iridescent butterfly scales: Lattices and laminae. J Morphol. 1989;202:69–88. 49. Dinwiddie A, Null R, Pizzano M, Chuong L, Krup AL, Tan HE, Patel NH. Dynamics of F-actin prefigure the structure of butterfly wing scales. Dev Biol. 2014;392:404–18. 49. Dinwiddie A, Null R, Pizzano M, Chuong L, Krup AL, Tan HE, Patel NH. Dynamics of F-actin prefigure the structure of butterfly wing scales. Dev Biol. 2014;392:404–18. 50. Dolan JA, Wilts BD, Vignolini S, Baumberg JJ, Steiner U, Wilkinson TD. Optical properties of gyroid structured materials: from photonic crystals to metamaterials. Adv Opt Mater. 2015;3:12–32. 50. Dolan JA, Wilts BD, Vignolini S, Baumberg JJ, Steiner U, Wilkinson TD. Optical properties of gyroid structured materials: from photonic crystals to metamaterials. Adv Opt Mater. 2015;3:12–32. 51. Barthelat F. Biomimetics for next generation materials. Phil Trans R Soc A. 2007;365(1861):2907–19. 51. Barthelat F. Biomimetics for next generation materials. Phil Trans R Soc A. 2007;365(1861):2907–19. 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: 52. Parker AR, Townley HE. Biomimetics of photonic nanostructures. Nat Nanotechnol. 2007;2:347–53. 52. Parker AR, Townley HE. Biomimetics of photonic nanostructures. Nat Nanotechnol. 2007;2:347–53. • 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: 53. Potyrailo RA, Ghiradella H, Vertiatchikh A, Dovidenko K, Cournoyer JR, Olson E. Morpho butterfly wing scales demonstrate highly selective vapour response. Nat Photon. 2007;1:123–8. 53. Potyrailo RA, Ghiradella H, Vertiatchikh A, Dovidenko K, Cournoyer JR, Olson E. Morpho butterfly wing scales demonstrate highly selective vapour response. Nat Photon. 2007;1:123–8. 54. Potyrailo RA, Starkey TA, Vukusic P, Ghiradella H, Vasudev M, Bunning T, Naik RR, Tang Z, Larsen M, Deng T, Zhong S, Palacios M, Grande JC, Zorn G, Goddard G, Zalubovsky S. Discovery of the surface polarity gradient on iridescent Morpho butterfly scales reveals a mechanism of their selective vapor response. Proc Natl Acad Sci U S A. 2013;110:15567–72. 55. Pris AD, Utturkar Y, Surman C, Morris WG, Vert A, Zalyubovskiy S, Deng T, Ghiradella HT, Potyrailo RA. Towards high-speed imaging of infrared photons with bio-inspired nanoarchitectures. Nat Photon. 2012;6:195–200. 55. Pris AD, Utturkar Y, Surman C, Morris WG, Vert A, Zalyubovskiy S, Deng T, Ghiradella HT, Potyrailo RA. Towards high-speed imaging of infrared photons with bio-inspired nanoarchitectures. Nat Photon. 2012;6:195–200.
https://openalex.org/W4353109676	https://digital.csic.es/bitstream/10261/305846/1/1-s2.0-S1364032123001260-main.pdf	English	null	Chemo-hydro-mechanical effects of CO2 injection on reservoir and seal rocks: A review on laboratory experiments	Renewable & sustainable energy reviews	2,023	cc-by	24,972	Atefeh Vafaie a Institute of Environmental Assessment and Water Research (IDAEA), CSIC, Barcelona, Catalonia, Spain b Department of Earth Sciences, University of Barcelona, Barcelona, Catalonia, Spain c Global Change Research Group (GCRG), IMEDEA, CSIC-UIB, Esporles, Spain Renewable and Sustainable Energy Reviews 178 (2023) 113270 Renewable and Sustainable Energy Reviews 178 (2023) 113270 Contents lists available at ScienceDirect * Corresponding author. Institute of Environmental Assessment and Water Research (IDAEA), CSIC, Barcelona, Catalonia, Spain. E-mail address: atefeh.vafaie@idaea.csic.es (A. Vafaie). 1. Introduction At geological storage sites, the injected CO2 into deep reservoirs (>800 m) reaches supercritical conditions (pressure >7.3 MPa and temperature >31.1 ◦C), leading to a liquid-like density that makes the storage volumetrically efficient. Yet, supercritical CO2 density is lower than that of resident brine, resulting in buoyancy that tends to drive CO2 upward back to the surface and, thus, low-permeability caprocks are essential to maintain CO2 deep underground permanently [5,6]. Further, CCS at the gigatonne scale will imply the interference between multiple CO2 injection wells targeting the same formation, giving rise to basin-wide pressurization of several megapascals [7,8]. This over pressure may constrain the maximum allowable injection rates and the storage capacities [9,10] to limit CO2 and brine leakage through the overlying caprock(s) [11–13] and induced earthquakes [14–16]. These phenomena could negatively impact the environment and harm the A R T I C L E I N F O Keywords: Geological carbon storage Coupled chemo-hydro-mechanical effects Time-dependent deformation Reservoir integrity Caprock sealing capacity Laboratory experiments Geological carbon storage is one of the technologies required to arrive at the ambitious yet realistic net-zero emission target and climate change neutrality. Injected CO2 in geological formations acidifies the resident brine, inducing chemical reactions with the minerals. Reactions may alter the hydraulic and mechanical prop erties of the rock and have an impact on reservoir and wellbore integrity, reservoir injectivity, long-term compaction and caprock sealing capacity. We provide a comprehensive review of chemo-hydro-mechanical (CHM) effects of CO2 on the reservoir and sealing rocks, either intact or fractured, which have been studied through laboratory experiments under no-flow and open-flow conditions, i.e., batch and flow-through experi ments. The hydraulic and mechanical rock properties affected by geochemical processes comprise (a) physical properties influencing the flow and transport of solutes (pore size distribution, porosity, permeability and fracture aperture and roughness), (b) multi-phase flow properties (capillary entry pressure and relative perme ability), (c) mechanical characteristics, including stiffness (i.e., elastic moduli for intact rocks and normal stiffness for fractures), strength (cohesion, friction angle and fracture toughness), and poroelastic response (Biot’s coefficient), and (d) time-dependent behavior of the rock (compaction creep). The experiments alone cannot capture the complex dynamics of CO2 flow underground, but they provide critical insights into short-term alteration mechanisms of rock with implications to geologic CO2 storage. Studying naturally altered rocks, extending the experiments to tens-of-meters-scale underground rock laboratories, and bringing together exper imental observations and numerical simulations are valuable for the advance in the understanding of CHM processes and upscaling them across space and time. Available online 22 March 2023 1364-0321/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1016/j.rser.2023.113270 Received 19 June 2022; Received in revised form 3 March 2023; Accepted 16 March 2023 Chemo-hydro-mechanical effects of CO2 injection on reservoir and seal rocks: A review on laboratory experiments Atefeh Vafaie a,b,*, Jordi Cama a, Josep M. Soler a, Iman R. Kivi a,c, Victor Vilarrasa 1.2. CO2-brine-rock interactions Large volumes of stored supercritical CO2 disturb the existing equi librium between formation brine and reservoir/caprocks (Fig. 1), trig gering several chemical, mechanical, thermal, and transport processes (e.g., mineral dissolution and precipitation, porosity and permeability variations, injection-induced overpressure and thermal and poroelastic stresses) over different spatial and temporal scales [17–19]. These coupled processes give rise to uncertainties concerning the reservoir pressurization response [20,21] and the sealing capacity and integrity of caprocks [22–26]. Continuous injection of dry supercritical CO2 with low chemical reactivity sweeps out and dries the rock from the resident brine in the near-wellbore zone, which may result in salt precipitation and possible changes in the hydraulic properties of zone A1 (Fig. 1) [17,28,34]. In contrast to dry CO2, multiple lines of evidence from laboratory experi ments suggest the reactive nature of wet CO2 towards the surface of some minerals, in particular those comprising divalent metal cations [35]. Separated water from wet CO2 may form a thin, reactive water film on the surface of silicate minerals that promotes a series of mineral transformation reactions, including silicate carbonation [36,37]. These reactions are of paramount importance to the mineral trapping of CO2 in basaltic and ultramafic rocks owing to their high mineral carbonation potential [38–40]. The type and extent of chemical interactions vary across different zones around the wellbore, divided by fluid phase saturations, i.e., volumetric proportions of CO2 and aqueous phases (or brine) in pores, and the concentration of dissolved CO2 and water molecules in the two phases [27–30] (Fig. 1a). A1 is the nearest-wellbore zone, where the rock pore network is fully occupied by dry supercritical CO2. A2 represents a zone of water-bearing supercritical CO2 (wet CO2). A3 illustrates the transition zone where the two fluid phases coexist at varying pro portions, and A4 is the zone of CO2-rich brine surrounding the CO2 2 Fig. 1. Schematic of a geological CO2 storage site with zones formed around the injection well (a) and potential reactions in these zones (b). A1 to A4 represent zones formed in the reservoir rock. A1, or the near-wellbore zone, is fully occupied by dry su percritical CO2. A3 is a two-phase flow zone and is surrounded by zones A2 and A4 comprising water- bearing CO2 and CO2-bearing brine, respectively. A5 is the far-field or uninvaded zone. 1.1. Geological carbon storage 1.1. Geological carbon storage 1.1. Geological carbon storage Carbon Capture and Storage (CCS) in deep geological formations is a reliable strategy to mitigate climate change [1–3]. CCS involves capturing CO2 from large point sources or the atmosphere and injecting it into suitable deep formations, such as saline aquifers, depleted oil and gas reservoirs, coal beds, and salt domes. Most mitigation pathways consistent with the Paris climate goal rely on large-scale deployment of CCS. The annual storage rates are expected to increase around the globe and reach up to 10 Gt and 25 Gt per year in 2050 and 2100, respectively, indicating cumulative storage of 800–3000 Gt CO2 in the current cen tury [4]. A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 plume. The zones A2 and A4 are driven by the mutual solubilities of CO2 and water, which are pressure and temperature dependent [31]. On the one hand, the CO2 phase in zones A2 and A3 may host variable amounts of dissolved water, i.e., hardly exceeding 1 mol% at temperatures <60 ◦C, but reaching 2–3 mol% at an elevated temperature of 100 ◦C and pressures >10 MPa [32]. On the other hand, CO2 dissolves into the aqueous phase in zones A3 and A4, where CO2 solubility has a strong pressure dependence and generally ranges between 2 and 3 mol% at pressures >10 MPa and temperatures <60 ◦C [31,33]. Finally, A5 is the uninvaded zone, which is furthest from the injection well. Unlike zone A5, which remains unaffected by CO2, the other zones that are in direct contact with CO2 may experience different physicochemical phenomena explained below. public perception on CCS, which may jeopardize its widespread imple mentation. Consequently, the hydrogeological response of the subsur face to CO2 injection becomes a matter of paramount importance when assessing potential geological storage sites and the long-term fate of the stored CO2. Fig. 1. Schematic of a geological CO2 storage site with zones formed around the injection well (a) and potential reactions in these zones (b). A1 to A4 represent zones formed in the reservoir rock. A1, or the near-wellbore zone, is fully occupied by dry su percritical CO2. A3 is a two-phase flow zone and is surrounded by zones A2 and A4 comprising water- bearing CO2 and CO2-bearing brine, respectively. A5 is the far-field or uninvaded zone. The caprock is divided into 3 zones: two-phase CO2-brine zones (dry CO2 in C1 and wet CO2 in C2) and a CO2-rich brine zone (C3). Fig. 1. Schematic of a geological CO2 storage site with zones formed around the injection well (a) and potential reactions in these zones (b). A1 to A4 represent zones formed in the reservoir rock. A1, or the near-wellbore zone, is fully occupied by dry su percritical CO2. A3 is a two-phase flow zone and is surrounded by zones A2 and A4 comprising water- bearing CO2 and CO2-bearing brine, respectively. A5 is the far-field or uninvaded zone. The caprock is divided into 3 zones: two-phase CO2-brine zones (dry CO2 in C1 and wet CO2 in C2) and a CO2-rich brine zone (C3). A. Vafaie et al. A. Vafaie et al. Dissolution of CO2 into water (zones A3 and A4) forms carbonic acid (H2CO3), which dissociates into protons (H+), bicarbonate (HCO− 3 ) and carbonate (CO2− 3 ) ions as follows [17] and mechanical characteristics (stiffness, strength, poroelastic response, and time-dependent behavior of the rock) [74–76]. These alterations may occur at varying distances from the injection well and result in substantial rock deformation and variations in flow [14]. CO2(g) + H2O ↔H2CO3(aq) (1) H2CO3(aq) ↔H+(aq) + HCO− 3 (aq) (2) HCO− 3 (aq) ↔H+(aq) + CO2− 3 (aq) (3) (1) l Mineral dissolution is perceived to be disadvantageous when enhancing caprock porosity and permeability, known as self-enhancing [77,78], or causing strength weakening [79,80], compromising the caprock integrity and sealing capacity. Furthermore, the chemically-induced stiffness degradation and reservoir creep may speed up compaction and, consequently, surface subsidence [80–83]. On the other hand, precipitation-induced improvement in caprock sealing ca pacity, known as self-sealing [23], and dissolution-induced growth of reservoir injectivity [84–87] are considered beneficial. Therefore, assessing the subsurface perturbations brought on by coupled CHM processes at different temporal and spatial scales is essential in large-scale deployment of CCS. Direct reliance on mineralogy and pore structure renders these processes more complicated and site-specific, requiring adequate methodologies and techniques to accurately address them. (2) (3) These H+-producing reactions lead to a pH drop in the resident brine from approximately 7.0 to an acidic range of 3.5–5.0 [17,33]. The acidic fluid induces chemical reactions with the reservoir and caprock, mainly mineral dissolution and precipitation, with potential pivotal impacts on their hydraulic and mechanical properties [28,30,41]. These irreversible interactions within the storage strata are constrained by rock miner alogy, in-situ pressure and temperature, and reservoir fluid salinity, making their assessment the main priority of the baseline geochemical evaluations of CCS sites [17]. Research carried out to date on CHM processes in the context of CCS has relied on (1) laboratory experiments on centimeter-long rock spec imens accompanied by accurate observations on the short-term effects (from several hours up to 2–3 years in exceptional cases), (2) charac terization of analogue samples from natural CO2 reservoirs exposed to CO2 over geological time scales, (3) field observations, both at pilot- and commercial-scale, and (4) numerical modeling employed to better interpret laboratory and field observations [68,88–90]. 1.3. Chemically-induced alterations of rocks CO2 storage reservoirs in sedimentary basins are mainly chalk, limestones and dolostones (carbonate rocks), and sandstones [38,42]. Thus, chemical reactions following CO2 injection are subject to the host rock minerals, commonly carbonates (e.g., calcite and dolomite) and silicates (e.g., quartz, feldspars, and clays) [17,21]. The mineral dis solution/precipitation reactions with different kinetic rates are active over distinct time scales. Calcite is the most prominent carbonate min eral in most host rocks, acting as a building block in limestones or cement in sandstones. It dissolves rapidly in acidic brines, releasing Ca2+ and bicarbonate (HCO3 −) and consuming protons [43,44]. The changes in ion concentrations increase the salinity and buffer the pH, resulting in less acidic conditions that favor reactions with silicates and result in precipitation of secondary minerals (e.g., dawsonite) over larger time scales (i.e., up to thousands of years), leading to permanent trapping of CO2 [45–49]. A. Vafaie et al. Laboratory ex periments, commonly conducted at constant stress and flow conditions, are unable in essence to capture the fully coupled nature of CHM pro cesses. They rather provide key insights into these processes individu ally, i.e., chemically induced changes in hydromechanical properties of the rock, on the one hand, and flow and mechanical controls on chem ical reactions, on the other hand (we hereinafter use the term coupled CHM processes without referring to its exact meaning). 1.2. CO2-brine-rock interactions The caprock is divided into 3 zones: two-phase CO2-brine zones (dry CO2 in C1 and wet CO2 in C2) and a CO2-rich brine zone (C3). Fig. 1. Schematic of a geological CO2 storage site with zones formed around the injection well (a) and potential reactions in these zones (b). A1 to A4 represent zones formed in the reservoir rock. A1, or the near-wellbore zone, is fully occupied by dry su percritical CO2. A3 is a two-phase flow zone and is surrounded by zones A2 and A4 comprising water- bearing CO2 and CO2-bearing brine, respectively. A5 is the far-field or uninvaded zone. The caprock is divided into 3 zones: two-phase CO2-brine zones (dry CO2 in C1 and wet CO2 in C2) and a CO2-rich brine zone (C3). Fig. 1. Schematic of a geological CO2 storage site with zones formed around the injection well (a) and potential reactions in these zones (b). A1 to A4 represent zones formed in the reservoir rock. A1, or the near-wellbore zone, is fully occupied by dry su percritical CO2. A3 is a two-phase flow zone and is surrounded by zones A2 and A4 comprising water- bearing CO2 and CO2-bearing brine, respectively. A5 is the far-field or uninvaded zone. The caprock is divided into 3 zones: two-phase CO2-brine zones (dry CO2 in C1 and wet CO2 in C2) and a CO2-rich brine zone (C3). A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 1.4. Paper overview Caprocks are primarily shales, commonly known for their tight na ture and low intrinsic permeability (typically, k < 10−18 m2) [50–53]. They possess ultrafine pore size in the micro-to the nanometer range, creating high capillary forces across the caprock. These features empower shales to function as efficient sealing barriers, preventing massive upward migration of CO2 [23,54]. Thus, CO2 infiltration into shales that lack permeable faults or fractures occurs through sluggish molecular diffusion unless their high capillary entry pressure is excee ded [55–57], where the non-wetting CO2 phase expels the wetting aqueous phase out of the largest pore throats of the caprock [58,59]. Chemical interactions between the injected CO2 and the caprock take place at the reservoir-caprock interface, where a small fraction of CO2 penetrates the caprock in free phase (C1 and C2) and, upwards, in an area that contains CO2-rich brine (C3) (Fig. 1a). As a result of the high capillarity of shales, the two-phase flow of CO2 and brine occurs only in a narrow region adjacent to the reservoir rock and close to the wellbore (C1 and C2), where injection-induced overpressure can overcome shale capillary entry pressure. Moreover, the bulk CO2 front spreads into the caprock at rates several orders of magnitude lower than in the reservoir, affecting the rate and extent of geochemical interactions [60]. This paper provides a comprehensive review on the present knowl edge of chemical, hydraulic and mechanical processes involved in geological carbon storage that has been acquired from laboratory ex periments using both intact and fractured rock samples. The review focuses on processes related to interactions of dry CO2 (zone A1 and C1) and aqueous solutions containing dissolved CO2, i.e., aqueous-phase dominated reactions (zones A3, A4, and C3), with rocks as they have been investigated in more detail so far. Although mineral reactivity with the counterpart water-bearing CO2 phase, i.e., non-aqueous dominated system (zones A2 and C2), is also important for CO2 transport and storage underground and the literature on relevant processes has become rich [37,91,92], it is not covered here. Furthermore, the potential impacts of CO2 impurities, such as SO2, NOX, H2S, NH3 and O2, on interactions with rocks have been assessed in several experimental studies [93–96], but are not reviewed here. This review targets studies on sedimentary rocks, including seventy- six studies on intact reservoir and shaly caprock specimens and sixteen on fractured rocks, conducted mainly since 2005. 1.4. Paper overview Of the former group, 30 studies used carbonate rocks (mainly chalks and limestones), 41 sandstones, and 13 shales. Our paper complements some recent reviews on experimental studies dealing with chemo-mechanical aspects of CO2 injection into rocks [90,97,98]. We provide detailed information on geochemical controls on mechanical, transport and multi-phase flow properties, which received relatively little attention in earlier studies. We first describe the experimental approaches employed to assess the coupled CHM processes in rocks induced by injected CO2 (Section 2). Then, we discuss geochemical controls on transport, flow and mechan ical properties of reservoir rocks (Section 3), caprocks (Section 4), and fractured rocks (Section 5). Finally, we conclude with recent advances in This review targets studies on sedimentary rocks, including seventy- six studies on intact reservoir and shaly caprock specimens and sixteen on fractured rocks, conducted mainly since 2005. Of the former group, 30 studies used carbonate rocks (mainly chalks and limestones), 41 sandstones, and 13 shales. Our paper complements some recent reviews on experimental studies dealing with chemo-mechanical aspects of CO2 injection into rocks [90,97,98]. We provide detailed information on geochemical controls on mechanical, transport and multi-phase flow properties, which received relatively little attention in earlier studies. i Fractures are ubiquitous in the subsurface and encountered at different length scales [61], resulting in strong heterogeneities in deformation, transport, and flow fields in the subsurface, which, in turn, may affect the geochemical processes of CO2 injection [62–64]. The CO2-induced chemical reactions may likely change the microstructure and mineral distribution in intact or fractured reservoir and sealing rocks. The resulting modifications give rise to changes in transport (pore size distribution, porosity, and permeability) [65–69], multi-phase flow properties (capillary pressure and relative permeability curves) [70–73], We first describe the experimental approaches employed to assess the coupled CHM processes in rocks induced by injected CO2 (Section 2). Then, we discuss geochemical controls on transport, flow and mechan ical properties of reservoir rocks (Section 3), caprocks (Section 4), and fractured rocks (Section 5). Finally, we conclude with recent advances in 3 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. the understanding of coupled CHM processes and their implications for field CO2 storage practices. We also highlight some perspectives for future research to fill the existing knowledge gaps (Section 6). intact rocks and fractured tight rocks (Fig. A1b). 1.4. Paper overview The resulting residence time of the solution is longer in batch experiments than in flow-through tests, in which several hundreds of pore volumes are usually circulated. CO2 injection experiments on fractured rocks are mainly conducted on evaporite and carbonate samples, occasionally acting as caprocks for CO2 storage in some geological settings [118]. The high reactivity of these rocks raises concerns about chemically assisted CO2 leakage through fractures [108,119]. The tested specimens contain either nat ural [120] or artificial fractures created by saw cutting [121] or appli cation of tensile stress (e.g., using the Brazilian method or chisel-type bits) [108,122]. Note that these artificial fractures may differ from shear fractures created under high normal stress due to the lack of fine-grained gouge (i.e., fracture or fault wear products upon slip). Table 1 Summary of common laboratory techniques used to evaluate the alterations in rock properties upon interaction with CO2. The references point to some ex amples of using the techniques in the context of CO2-water-rock interactions. g Measured parameters Measurement technique References Chemical analysis Influent/effluent solution chemical composition Inductively Coupled Plasma- Atomic Emission Spectroscopy (ICP-AES) Inductively Coupled Plasma- Mass Spectroscopy (ICP-MS) Spectrophotometry Ion chromatography [99,100] [101] [102] [103] Influent/effluent solution pH pH measurement [104] Microstructural-hydraulic analysis Porosity Helium/Nitrogen porosimetry [67] Fracture roughness and aperture High-resolution optical profilometer [105] Porosity, pore size distribution, skeletal and grain density, capillary entry pressure Mercury Intrusion Capillary Pressure (MICP) [26] Porosity and surface area Brunauer Emmett Teller (BET) [78] Permeability Permeability measurement (steady-state or pulse decay methods) [60,106] Porosity, microstructural evolution Scanning Electron Microscopy (SEM) [69] porosity, pore size distribution, microstructural evolution, fracture aperture X-ray Micro Computed Tomography (XMCT) Synchrotron X-ray Micro Computed Tomography (SXMCT) [107] [108,109] Mineralogical analysis Mineral composition Scanning Electron Microscopy (SEM) [110] Mineral composition X-Ray Diffraction (XRD) [111] Mechanical analysis Friction behavior of fault gouge Direct shear experiment [112] Fracture toughness Double torsion fracture mechanic test [113] Scratch toughness and scratch hardness Scratch test [114] Elastic, poroelastic, strength, and creep properties of intact rocks Normal stiffness of fractures Triaxial compression test [115] [116] Elastic, unconfined strength, and creep properties Uniaxial compression test [117] Dynamic elastic properties Ultrasonic velocity measurement [74] Most of the samples used in the laboratory experiments were extracted from representative reservoir and sealing formations that constitute CO2 storage sites. The experiments have also been conducted on synthetic cores [85,136,137] and natural analogue samples extracted from either deep CO2 reservoirs [25] or outcrops exposed to CO2 leakage over geological time scales [115]. Samples can be centimeter-size cy lindrical cores, which is commonly the case in flow-through experi ments, or of other shapes like cubic specimens [138,139], rock fragments [78,103], powder [140], or disk [141], which are prevalent in batch experiments. The studied sedimentary rock samples are carbon ates (including limestones, dolostones and chalks), sandstone, and shales with a large range of porosities (3%–42% in carbonates, 7%–29% in sandstones, and 2%–22% in shales), and permeabilities (10−18 < k < 10−12 m2 in carbonates, 10−17 < k < 10−12 m2 in sandstones, and 10−21 < k < 10−19 m2 in shales) (Table S2). 2. Experimental methodology Many experimental studies have been carried out in the last fifteen years to assess the impacts of CO2 injection on the hydromechanical properties of intact or fractured rocks. CO2 injection experiments have been combined with analytical techniques to evaluate changes in microstructural, mineralogical, chemical, hydraulic, and mechanical properties of the rock before, during, and after exposure to CO2. Table 1 summarizes these techniques and the corresponding evaluated parameters. l As for the phase state of the injected CO2 in the experiments, pure dry gaseous-phase [123–125], liquid-phase [126], supercritical-phase CO2 [127–131], and dissolved CO2 in aqueous phase commonly equilibrated with subcritical to supercritical CO2 [132–134] (hereinafter referred to as CO2-rich water) are used. The aqueous phase in the injected solution and initial pore fluid of the rock may be deionized, fresh, brackish or saline water besides brine. We use the general term water to refer to all these compositions throughout the paper unless differently indicated. The fluid compositions for all experiments are however listed in Table S1. We also note that the experimental pressures and temperatures vary largely (i.e., P = 0.1–60 MPa and T = 22–200 ◦C; summarized in Table S2) to replicate field storage conditions, considering surface temperature of 20 ◦C and gradients of hydrostatic pressure (10 MPa/km) and temperature (30–45 ◦C/km) [135]. CO2 injection experiments mainly differ in the flow conditions and the phase state of the injected CO2 (Table S1). Batch experiments emulate static or no flow conditions (i.e., diffusive transport conditions) to evaluate the short-term (usually for several hours and up to a few years) effects on rock exposure to CO2 (Fig. A1a). Flow-through exper iments represent open-flow conditions to assess the short-term effects of CO2 injection on the properties of both more porous and permeable 2.2. Flow-through experiments alterations, making carbonate rocks the focus of a large number of studies. Chemical interactions between dissolved CO2 and carbonates involve both dissolution and precipitation reactions, whose extents are sensitive to the pH and CO2 solubility in water which are functions of CO2 pressure, temperature and salinity [151,152]. To explore rock al terations at representative reservoir conditions, batch and flow-through experiments have been carried out under a wide range of pressure and temperature conditions (P = 0.1–60 MPa and T = 25–120 ◦C). Flow-through experiments (a.k.a. Percolation experiments) are suitable for assessing the combined effects of advective and diffusive transport of reactive solutes on rock properties. In these experiments, CO2 either as (1) a free phase in a gaseous, liquid, or supercritical state, (2) dissolved in water, or (3) a two-phase mixture with water, is forced to flow through permeable limestones and sandstones [130]. Moreover, flow-through experiments by injecting CO2-rich water into pre-fractured rock specimens have been usually carried out at elevated temperatures, pressures, and stresses representative of reservoir conditions. The applicability of flow-through experiments to shale samples of nano-darcy permeability is subject to discrimination between the viscous bulk flow of CO2 and its diffusive transport through high-precision measurements together with appropriate analytical or numerical interpretations [55,145]. Furthermore, fluid permeation through the entire body of the shale specimen could be very time-consuming and has to be assured. l In batch experiments, i.e., a closed system, exposure of carbonates to CO2 causes only slight alterations in the pore structures owing to the limited water-to-solid ratio. Under such conditions, the fast dissolution of calcite increases the aqueous concentration of carbonate species and consumes protons (pH buffering), leading to the stability of the solid calcite phase (equilibrium) [66,84,101,147,153]. An increase in porosity may occur, but barely reaches a small percentage [66,101,154]. Correspondingly, insignificant permeability changes are expected [27, 147,153,155,156]. Dissolution traits are more pronounced in samples exposed to CO2-rich water than those treated with dry supercritical CO2, as water is the key agent for the acid-producing reaction (Eq. (1)) [66]. The amount of CO2 infiltrating the rock is low under static conditions owing to the slow pace of the diffusion process, preventing pH from reaching values lower than about four [101]. Rock specimens used in the flow-through experiments may be dry or saturated with water before CO2 injection. The injected fluids are designed to reproduce fluid-rock interaction conditions at distinct zones around wellbores (Fig. 2.2. Flow-through experiments 1). For instance, the injection of dry CO2 into a water-saturated rock specimen mimics the near-wellbore zone (zone A1 in Fig. 1a), where CO2 displaces the resident fluid and dries out the rock [67,146]. In such cases, the injection rate plays a significant role in replicating underground injection conditions [60,146]. It is worth mentioning that some flow-through experiments were conducted under closed downstream conditions where the fluid flow out of the specimen was impeded, and molecular diffusion is the primary CO2 transport process through the rock in the long term [126,147]. This testing approach helps overcome the limitation of neglecting stress state in conventional autoclave batch reactors although suffering from limiting reactions to low water-to-rock ratio. Assuming that the reservoir as a whole emulates a closed hydrody namic system with some brine circulation but not renewal, results from the batch experiments would suggest negligible changes in the hydraulic properties of a carbonate reservoir [147]. However, this assumption may not be fulfilled locally (regions A1 to A4 in Fig. 1). Indeed, disso lution and/or precipitation reactions could persist over a period long enough to induce significant alterations in rock properties [28,97]. Percolation experiments are supposed to simulate conditions under which continuous injection of dissolved CO2 into carbonate rock sam ples results in a progressive dissolution of calcite and increase in porosity (Fig. 2). Absolute porosity enhancements of up to 5–6% (20% relative enhancement) occurred in limestone samples of the L´erouville and Savonniere Formations under open-flow conditions [65,74,107]. Moreover, calcite dissolution and pore widening may cause large enhancement in intrinsic permeability (documented by up to 3 orders [65,67,102]). The positive feedback between the rapid dissolution of calcite and acid renewal by fluid flow promotes chemical reactions. The higher the injection rate, the higher the extent of interactions with rock [157]. However, under specific flow and reaction conditions, the un stable evolution of flow pathways in carbonates can create highly conductive channels known as wormholes (explained in Section 3.1.2). 2.1. Batch experiments In the CCS framework, batch experiments are primarily used to characterize slow evolution of rock samples through sluggish reactions under no-flow conditions at constant temperature and pressure (e.g. Refs. [73,140,142]). Samples are usually immersed in CO2-rich water by supplying supercritical CO2 at the top and can react over a wide range of pressures and temperatures (ambient to 60 MPa and 200 ◦C, Table S1) to reproduce subsurface conditions [87,103,143], where CO2 dissolves in the pore fluid at varying degrees [31], determining the solution pH. Aqueous CO2 diffuses through the rock sample at a rate that inversely correlates with the pore network tortuosity [55]. Long-term batch ex periments (from weeks to months, even up to 2–3 years) are well suited to study alterations in clay-rich shale samples as their low intrinsic permeability renders flow-through experiments very time-consuming and cumbersome [104,111,144]. The main limitations of batch experi ments are the lack of the effects from advective transport (importantly when concerning fluid-rock interactions during and shortly after injec tion into the reservoir) and confining stress on chemical reactions, and restriction of sample characterization to ex-situ measurements before and after exposure to CO2 (Table 1). 4 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 2.3. Natural analogue samples Some studies used samples from natural emplacements where CO2 from mantle degassing, metamorphic processes, or organic matter breakdown was accumulated in subsurface formations and stored over geological time scales (104–106 years). Rock samples retrieved from these reservoirs provide critical insights into the long-term effects of CO2-rock exposures after a long residence time that cannot be repro duced using laboratory timescales [148,149]. Numerous natural CO2 accumulations are found worldwide, but experimental research on the potential long-term CO2-rock interactions is scarce [25,115]. In sandstones, the mineral composition of the rock cement, i.e., silicate-cemented or calcite-cemented sandstone, plays a major role in the evolution of the rock properties. Sandstones with high content of carbonates are prone to dissolution-induced porosity and permeability enhancements in a range comparable to carbonate rocks [69,115,117, 141,158–162], particularly under open flow conditions, hindering calcite buffering effects [60,163]. Likewise, in natural-analogue speci mens with a marked carbonate content (14 wt% and 23–38 wt% in the Entrada Sandstone and Summerville Siltstone, respectively) exposed to CO2 over 400 k years, a porosity increase of 3–6% was observed due to the dissolution of grain-coating hematite and calcite cement. 3.1. Changes in hydraulic properties 3.1. Changes in hydraulic properties 3. CHM effects of CO2 on reservoir rocks Mineral dissolution and precipitation induced by acidified water can cause changes in pore structure, affecting the rock porosity and permeability (i.e., hydraulic properties) and, eventually, the mechanical structure and strength of the targeted reservoirs. The distribution and content of the reactive minerals constituting the rocks and flow hydro dynamics determine the extent to which rock properties may change and affect the capacity/efficiency of geological carbon storage. Most oper ating storage sites comprise carbonates (e.g., chalks, limestones and dolostones) and sandstones as reservoir rocks, overlain by tight, thick shales as caprocks (Tables S2 and S3). An exception to CO2 injection in sedimentary rocks is storage in basalt and ultramafic rocks, e.g., the CARBFIX project [150]. The dissolution rate of silicates (feldspars, clays, and quartz) is slower than that of calcite by up to nine orders of magnitude [164,165]. Hence, trivial alterations in pore structure, porosity, and permeability are conceivable for carbonate-free sandstone treated under open-flow conditions (Fig. 3b [115,156]). Yet, low- or no-carbonate sandstones that contain clays and feldspars, which are more reactive than quartz, react with dissolved CO2 over long times, implemented commonly in batch experiments (months to a few years, Fig. 3a). More complex re actions, including mineral dissolution, precipitation and transformation, may occur simultaneously or sequentially, leading to positive or nega tive contributions to hydraulic properties. The longer the exposure time, the greater the extent of alterations. For instance, 8% and 3-fold 3.1.1. Role of mineralogy and flow conditions 3.1.1. Role of mineralogy and flow conditions l Carbonate rocks are mainly composed of highly reactive calcite, the content of which is decisive in determining the extent of rock 5 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. gy ( ) Fig. 2. Porosity changes in carbonate rocks during CO2 injection: (a) cross-plot of porosity for treated Ekofisk and Tor Chalks subjected to supercritical CO2 injection versus intact porosities (Modified from Alam et al., 2014 [84]) and (b) porosity variations in a limestone specimen with injected pore volumes of CO2-rich water (Modified from Vialle and Vanorio, 2011 [74]). f Fig. 2. Porosity changes in carbonate rocks during CO2 injection: (a) cross-plot of porosity for treated Ekofisk and Tor Chalks subjected to supercritical CO2 injection versus intact porosities (Modified from Alam et al., 2014 [84]) and (b) porosity variations in a limestone specimen with injected pore volumes of CO2-rich water (Modified from Vialle and Vanorio, 2011 [74]). riation in porosity in sandstones with different mineral contents reacted with CO2-rich water under no-flow [69,87,104,138,139,141,153,166,167] (a) and conditions [60,66,117,124,137,156,158,160,163] (b). The duration of the experiments, if reported, is also illustrated. Variation in porosity in sandstones with different mineral contents reacted with CO2-rich water under no-flow [69 87 104 138 139 141 153 166 167] (a) and Fig. 3. Variation in porosity in sandstones with different mineral contents reacted with CO2-rich water under no-flow [69,87,104,138,139,141,153,166,167] (a) and open-flow conditions [60,66,117,124,137,156,158,160,163] (b). The duration of the experiments, if reported, is also illustrated. Fig. 3. Variation in porosity in sandstones with different mineral contents reacted with CO2-rich water under no-flow [69,87,104,138,139,141,153,166,167] (a) and open-flow conditions [60,66,117,124,137,156,158,160,163] (b). The duration of the experiments, if reported, is also illustrated. 6 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. increases in porosity and permeability, respectively, were observed in specimens from the Hutton [104], Stuttgart [166], and Mt. Simon [138] Formations exposed to CO2-rich water for up to 4 months. Dissolution of clays and feldspars, widening the pore space and creating micro frac tures along the bedding planes, and corrosion of quartz were responsible for the enhancement of the hydraulic properties. Conversely, the com bined effects of K-feldspar dissolution, kaolinite, barite, and celestine precipitation reduced the porosity and permeability of a sandstone specimen from the Tuscaloosa Formation by 2% and 13%, respectively, after 6-month exposure to dissolved CO2 [143]. Fig. 5. 3.1.1. Role of mineralogy and flow conditions Neutron radiographs of different dissolution patterns caused by reactive fluids in limestone: (a) uniform dissolution, (b) compact dissolution, (c) conical wormhole, (d) ramified wormhole, and (e) dominant wormhole formation (Adapted from Fredd and Fogler [182]). Ranges of Damkohler (Da) and Peclect (Pe) numbers are taken from Menke [179]. Mineral dissolution and precipitation processes dominate the tran sition zones A3 and A4 (Fig. 1). As the aqueous phase saturation in the region adjacent to the injection well, i.e., zone A1, decreases toward the residual value, its mobility decreases (due merely to reduction in the relative permeability to this phase), raising a trade-off between bulk phase (water and dissolved minerals as a whole) displacement by CO2 and water evaporation to the CO2 phase [67]. If the latter prevails, which is particularly the scenario in small pores due to the lower fluid velocity, the concentration of dissolved minerals progressively in creases, resulting in dehydration-induced salt precipitation if exceeding the salt solubility limits [168,169]. This phenomenon is prevalent not only in the CCS context but also during geothermal energy exploitation from depleted high-temperature gas reservoirs where CO2 serves as a circulating fluid [170–172]. Dramatic injectivity reductions due to pore clogging by salt crystals are anticipated even in the short-term for high-salinity systems [168,173]. Laboratory experiments of dry CO2 injection into brine-saturated sandstones give credence to dehydration occurrence [63,146,174]. The lower the rock porosity, the higher the extent of changes in hydraulic properties, with up to 10% and 20% relative decreases in porosity and permeability, respectively (Fig. 4 [67]). Fig. 5. Neutron radiographs of different dissolution patterns caused by reactive fluids in limestone: (a) uniform dissolution, (b) compact dissolution, (c) conical wormhole, (d) ramified wormhole, and (e) dominant wormhole formation (Adapted from Fredd and Fogler [182]). Ranges of Damkohler (Da) and Peclect (Pe) numbers are taken from Menke [179]. dissolution patterns may develop in carbonate rocks: (1) Compact dissolution under high Da and low Pe conditions (diffusion-controlled system), where the reaction front advances along the flow direction, and dissolution progresses as a thin front (Fig. 5b) [180]; (2) Conical wormholes form under similar reactive conditions but with an inter mediate Pe number where the reaction front becomes unstable, espe cially in heterogeneous porous media (Fig. 5c); (3) dominant wormhole formation occurs under high Pe conditions (i.e., advection dominance) where dissolution concentrates along preferential flow paths (Fig. 5e). 3.1.1. Role of mineralogy and flow conditions The high velocity under such conditions leads to a minimal residence time with the result that wormholes do not have a substantial width; (4) Ramified wormholes form in advection-dominated systems, in which reaction is not very rapid, resulting in longer residence times under which wormhole spreading and branching occur (Fig. 5d); (5) Uniform dissolution is typically observed when the reaction rates are slow, and the acid fluid has time to access the entire pore space in the sample (Fig. 5a) [179,181]. Nevertheless, the boundaries between these disso lution regimes are primary functions of pore space heterogeneity, varying largely from one carbonate rock to another and, thus, are not definite. Besides, precise determination and definition of the Da number are still challenging because of complexities in determining the reactive surface area as a major heterogeneity factor controlling the reaction rate. 3.1.2. Wormhole formation and rock heterogeneity The feedback between dissolved CO2 transport, rapid chemical re actions, and the highly diverse textural attributes (i.e., heterogeneity in pore size and shape, grain size and sorting, among others, encountered at different scales) in carbonate rocks leads to instabilities in the advancing reactive fronts and localized dissolution regimes (e.g., wormhole formation, Fig. 5) [67,74,175–178]. Experimental and theo retical studies have demonstrated that dimensionless P´eclet (Pe) and Damnk¨ohler (Da) numbers can be used as practical criteria to determine the resulting dissolution patterns. While P´eclet number measures the relative magnitude of pressure-driven advective flow to diffusive transport, Damk¨ohler number compares the reaction rate to the advec tive mean fluid velocity [177,179]. i l Depending on the range of Pe and Da numbers, five distinct Fig. 4. Decrease in porosity and permeability for quartz-rich Fontainebleau sandstone specimens after injection of gaseous CO2 (modified after Vanorio et al. [67]). The formation of wormholes is of particular importance for reservoir conductivity and its mechanical integrity as will be discussed later. Laboratory observations unravel two-stage permeability enhancements during wormhole formation (Fig. 6) [65,97,102]. In the initial stage, a slow to moderate increase in permeability results from dissolution-enhanced pore connectivity, removal of pore-clogging par ticles, the initiation of wormhole formation, and flow localization [97, 182]. Subsequently, permeability sharply increases as the wormhole breaks through the sample [65,102]. Power-law permeability-porosity relationships (k ∝ φn) with high values of n (as high as 56) are required to capture the localization of flow and reaction in wormholes [65,102]. Thus, the conventional choice of n = 3 in the Kozeny-Carman Fig. 4. Decrease in porosity and permeability for quartz-rich Fontainebleau sandstone specimens after injection of gaseous CO2 (modified after Vanorio et al. [67]). 7 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. Fig. 6. Permeability changes as a function of porosity in an Estaillades Lime stone specimen flooded with a CO2-rich brine/distilled water sequence for 115 days (Modified from Vialle et al. [102]). Conclusively, alterations in relative permeability curves are pri marily correlated to how the pore size distribution evolves upon inter action with CO2. However, the application of temperature-retarded acids should be further verified as they statically dissolve minerals uniformly at the pore scale that may overlook actual heterogeneous velocity fields of pore-scale CO2 invasion, probably dissolving the larger pores (i.e., where fluid flow is focused) more easily. 3.1.3. Geochemical controls on multiphase flow In light of the limited effects of CO2 on the pore structure of car bonates or carbonate-rich sandstones under no-flow conditions due to the buffering effect of dissolved calcite, slight changes in elastic prop erties, either static or dynamic, are expected and accredited by labora tory measurements [66,139,147,153]. In contrast to no-flow conditions, flooding carbonate and carbonate-bearing sandstone samples with dis solved CO2, where the injected acidic solution is continuously renewed, induces significant changes in rock porosity and, thus, in elastic con stants, i.e., Young’s, bulk, and shear moduli, and Poisson’s ratio [27,29, 84,147], and in seismic behavior, i.e., compressional and shear ultra sonic velocities [74,86,195]. Geochemical interactions, rock weakening, and ultrasonic velocity attenuation continue as long as a CO2-rich so lution is injected into the specimen [147,155,196]. Conversely, rock stiffening and velocity amplification is expected during the injection of dry gaseous/supercritical CO2 into water-saturated rock samples, as salt precipitation is likely to occur [67]. The behavior of Poisson’s ratio, however, does not show any clear trend with exposure to CO2, not correlating with the rock type and experimental conditions, decreasing in some cases [197] and increasing in others [167,196]. i Assessment of the potential impacts of chemical reactions on the two- phase flow behavior of water and CO2 is carried out by studying relative permeability, defined as the ratio of the permeability of a given fluid in the presence of other fluids to the intrinsic rock permeability. In essence, relative permeability quantifies the extent to which CO2 and water interfere as both migrate through rocks, affecting all critical storage processes, from residual CO2 trapping mechanism [185,186] to CO2 leakage through the caprock [145]. However, relative permeability has not been adequately characterized in CO2-altered rocks. Appraisal of two-phase flow properties in treated rocks requires a continuous pore network, a condition hardly met in carbonate rocks in the laboratory, where flooding with CO2 likely leads to large wormhole formation [177, 182]. Temperature-controlled acids have been proposed to overcome characterization challenges posed by wormhole formation [187–189]. These acids are water-soluble organic compounds activated above a certain temperature, resulting in uniform rock dissolution, not only at the core scale but also at the pore scales [26]. This treatment approach enabled Niu and Krevor [26] to obtain the relative permeability curves of two altered specimens from the Ketton and Estaillades Limestones. 3.1.2. Wormhole formation and rock heterogeneity A combination of high-resolution in-situ imaging techniques, e.g., synchrotron XMCT imaging [109], and modeling approaches [190] to capture the evolution of pore structure upon interactions with CO2 prior to wormhole for mation should enable better assessment of changes in two-phase flow properties. 3.2. Geochemical effects on mechanical behavior This section presents recent advances in understanding potential changes in rock deformation, stiffness, strength, and time-dependent behavior, i.e., compaction creep, originating from interactions with CO2. The key concept of these studies is that the evolution of the rock microstructure, i.e., pore structure and grains as the fundamental con stituents of the load-bearing framework, would potentially cause changes in the rock mechanical behavior [191,192]. Given the short duration of laboratory experiments, the studies have mostly focused on carbonate dissolution and the subsequent effects on mechanical prop erties. Alterations of the mechanical properties of reservoir rocks are broadly relevant to assess the subsurface CO2 flow and storage perfor mance, given the tightly coupled behavior of flow, mechanical and geochemical processes [30,90,193,194]. Fig. 6. Permeability changes as a function of porosity in an Estaillades Lime stone specimen flooded with a CO2-rich brine/distilled water sequence for 115 days (Modified from Vialle et al. [102]). permeability-porosity relationship and other adapted values for porous rocks [183,184] are only valid for a finite range of pore structure evo lution where the continuity of the porous medium is not questioned. 3.1.3. Geochemical controls on multiphase flow Relative permeability curves at intermediate water saturations and pore size distribution for limestone samples fr Estaillades Formations (c and d, respectively) before and after treatment with CO2 (Adapted from Niu and Krevor [26]). Fig. 7. Relative permeability curves at intermediate water saturations and pore size distribution for limestone samples from the Ketton (a and b, respectively) and Estaillades Formations (c and d, respectively) before and after treatment with CO2 (Adapted from Niu and Krevor [26]). stems from the compaction-induced reduction in rock porosity and consequently in the available reactive surface area. The stiffer the initial pore structure of the rock, as in the case of a micrite-rich limestone [85], the lower the stress dependence of fluid-rock interactions. that intensify the deformation of the solid part of the rock (equivalent to a lower solid bulk modulus). As a result, even if the bulk modulus of a rock does not change, which is plausible in a closed geochemical system [147,153,200], a slight decrease in the Biot coefficient can still be ex pected, e.g., from 0.85 to 0.81 in Apulian Limestone [156]. 3.1.3. Geochemical controls on multiphase flow The results showed increased and decreased relative permeabilities to CO2 and water, respectively, for the Ketton Limestone specimen in response to dissolution-enhanced porosity (Fig. 7a). This selective alteration is due to the enlargement of small pores previously occupied by the wetting phase (i.e., water), but now provide pathways for the non-wetting phase (i.e., CO2) to flow (Fig. 7b). Interestingly, the altered specimen from the Estaillades Limestone underwent a decrease in rela tive permeability for both water and CO2 in spite of dissolution-induced porosity and intrinsic permeability enhancement (Fig. 7c). The reason is a rise in the volume fraction of middle-range pores while both large and small pore fractions decrease (Fig. 7d). This indicates a competition between wetting and non-wetting phases to pass through the middle-range pore network. A two-phase flow interference decreases relative permeability for both phases as increasing fractions of both CO2 and water pass through the same-size pore cluster. Changes in the Biot effective stress coefficient (α) that delineates the poroelastic behavior of the rock can be evaluated as α = 1 – (bulk modulus of the rock)/(bulk modulus of the solid part of the rock), herein referred to as solid bulk modulus [198]. Assuming an ideal porous me dium, which is occasionally the case for homogeneous rocks with a low content of compressible clay minerals and an interconnected pore network, the solid bulk modulus is well correlated to the mean bulk modulus of the grains constituting the rock [199]. Considering this parameter as constant, a decrease in the rock bulk modulus translates into an increase in the Biot coefficient. Nevertheless, Kim and Makh nenko [156] provided evidence for alterations in the solid bulk moduli of the Apulian and Indiana Limestone samples, as well as the bulk modulus of pure calcite crystals in contact with liquid CO2 under no-flow conditions. These moduli decreased by up to 20% in these samples, indicating local dissolution of the calcite crystal surface. The formation of non-interconnected pores due to re-precipitation of locally dissolved calcite also gives rise to micro-scale inhomogeneities in compressibility 8 8 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. Fig. 7. Relative permeability curves at intermediate water saturations and pore size distribution for limestone samples from the Ketton (a and b, respectively) and Estaillades Formations (c and d, respectively) before and after treatment with CO2 (Adapted from Niu and Krevor [26]). Fig. 7. 3.2.2. Failure behavior Circles, triangles, and solid lines indicate pore collapse onset (i.e., the point at which stress-strain curves deviate from linear elastic behavior), shear strength (i.e., the point where rock failure occurs under elevated deviatoric stresses), and failure envelopes (laboratory data fitting). Characteristics of intact and altered rocks are discriminated by blue and red colors, respectively (adapted from Alam et al. [84]). Fig. 8. Deviatoric versus mean effective stress q-p′ plots of failure properties of intact and CO2-injected specimens from the (a) Ekofisk Chalk and (b) Tor Chalk. Circles, triangles, and solid lines indicate pore collapse onset (i.e., the point at which stress-strain curves deviate from linear elastic behavior), shear strength (i.e., the point where rock failure occurs under elevated deviatoric stresses), and failure envelopes (laboratory data fitting). Characteristics of intact and altered rocks are discriminated by blue and red colors, respectively (adapted from Alam et al. [84]). broad range, from negligible to reduction by more than 50%, depending on their mineralogical content and the rock texture [25,66,115,197]. Dissolution of locally-concentrated reactive minerals acting as load-bearing cement or grains, even if the content is small, can give rise to significant strength degradation [25,134]. By contrast, the dissolution of dispersed pore-filling reactive cement may not affect the mechanical behavior, despite its potential effect on hydraulic conductivity enhancement [25,197]. broad range, from negligible to reduction by more than 50%, depending on their mineralogical content and the rock texture [25,66,115,197]. Dissolution of locally-concentrated reactive minerals acting as load-bearing cement or grains, even if the content is small, can give rise to significant strength degradation [25,134]. By contrast, the dissolution of dispersed pore-filling reactive cement may not affect the mechanical behavior, despite its potential effect on hydraulic conductivity enhancement [25,197]. dissolution, diffusive transport of reaction products through the grain boundary fluid film/coating, and potential precipitation of secondary minerals on the free-face grain surface in pores [212]. The interface reaction kinetics depends on the fluid chemistry, rock composition, porosity, grain size, temperature and stress state and is fast in granular chalk and limestone due to the high solubility and reaction rate of calcite [147,217–220]. CO2 injection strongly accelerates pressure solution and, thus, compaction creep of porous granular carbonate rocks by up to two orders of magnitude [24,89,147,219,221]. 3.2.2. Failure behavior This creep-accelerating effect can be explained based on the notion that CO2 injection in creases the calcite dissolution rate (both in pores and grain contacts), which in turn increases the rock porosity, decreases the grain contact area and enhances stress-induced contact dissolution in parallel with grain sliding, rearrangement and crushing [24,89]. Naturally altered sandstones offer an opportunity to explore the strength degradation associated with long-term reaction mechanisms, e. g., those of quartz dissolution. The measured compressive strength of an outcrop Summerville specimen exposed to CO2-charged brine leaking through an adjacent fault for over hundreds of thousand years was up to 90% lower than their apparently intact twin samples [115]. The failure regime shifted from brittle strain-softening in intact specimens to ductile slightly strain-softening in altered ones, indicating a significant decrease in rock brittleness (Fig. 9d) [210]. Similar evolution trends, although less pronounced, were observed for the marginally more porous sample retrieved from the Entrada Sandstone at the same site (Fig. 9b). This discrepancy in the extent of alterations suggests that drastic strength deterioration of Summerville samples should not be viewed as purely long-term chemical effects. Indeed, altered Summerville samples could endure a variety of loading paths linked to their proximity to the fault core [211], leading to severe layering and partially mineralized frac tures (Fig. 9c). Therefore, rock mechanical alterations are driven by the combination of geochemical interactions and the presence of heteroge neities, acting as planes of weakness. Subcritical crack growth is simply understood as quasi-static, stable crack (or fracture) propagation at stress levels well below the tensile failure strength, i.e., opening-mode fracture toughness according to linear elastic fracture mechanics [214]. Adsorption of aqueous fluid species, i.e., H+, OHˉ, and H2O, onto mineral surfaces and breakage of their bonds, e.g., Si–O bonds in quartz, by hydrolysis reduces the surface energy of the mineral, i.e., the energy required to create a new surface, and promotes the crack formation and subcritical growth, a process called stress-corrosion cracking [222,223]. This chemically-activated cracking mechanism is primarily controlled by the solution pH [224]. It is well established that the surface energy of quartz (and probably feldspar) increases with decreasing solution pH [81,225]. Accordingly, CO2 injection into quartz- and feldspar-bearing sandstone, acidifying the pore fluid, progressively inhibits stress corrosion and, thus, creeping. 3.2.2. Failure behavior The observed evolution of rock stiffness and wave velocities is of critical importance to seismic monitoring of the CO2 plume dynamics. Common geophysical monitoring approaches rely on the well-known Gassmann model [201], which benefits from the difference in physical properties (density and compressibility) of CO2 and water to address the effect of fluid substitution on elastic wave propagation [202,203]. However, the Gassmann model does not account for the geochemically-induced time-dependent rock’s stiffness and seismic response. Developing alternative empirical relationships considering these interactions is a hot research topic in rock physics [75,76]. Dissolution of inter-granular cementing minerals and grain surfaces in the pore space deteriorates contact surfaces and individual grains, causing pore collapse and microcracking at lower stress levels [84,89, 196,205,206]. Such microstructural degradation yields lower tensile and compressive strengths in most carbonates and sandstones exposed to dissolved CO2 (Table S3). Expectedly, carbonates and primarily chalks with high porosity and surface area undergo the largest chemically-driven weakening of the load-bearing capacity [84,132,147, 196,207]. Pore collapse is markedly susceptible to chemical degradation (Fig. 8), whereas chemical controls on the friction yield surface are still unresolved. The friction coefficient of altered rocks may diminish (Fig. 8b) [84,115,159,163]) or remain almost unaffected by CO2 injec tion (Fig. 8a) [25,84,206]. Depending on the prevalence of compactive pore collapse and dilatant microcracking mechanisms, the rock perme ability will probably decrease or increase, respectively [208,209], notably affecting reservoir injectivity. Changes in the rock stiffness alter the reservoir propensity to expansion and compaction during and after CO2 injection, affecting surface uplift and subsidence, respectively [204]. Furthermore, rock deformation could also considerably impact CO2 flow and the resulting geochemical processes in rock, as elaborated by Vanorio et al. [67] through laboratory experiments. The bulk modulus of carbonate rock samples during continuous CO2-rich water injection diminished by up to 70% under no lateral confinement, significantly higher than stiffness degradations of less than 20% at 15 MPa of confinement. This difference Changes in the compressive strength of sandstone samples reacting with CO2 over the exposure times in laboratory experiments span a 9 9 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. Fig. 8. Deviatoric versus mean effective stress q-p′ plots of failure properties of intact and CO2-injected specimens from the (a) Ekofisk Chalk and (b) Tor Chalk. 3.2.2. Failure behavior In line with theory, experiments on samples from the Adamswiller Sand stone [89], Berea Sandstone [226] and Captain Sandstone [81,197] have revealed almost similar or even lower long-term compaction rates when converting the pore fluid from water to CO2-rich water. Yet, sili ciclastic rocks with even small contents of reactive cementing materials, i.e., carbonates and clays, could experience non-trivial time-dependent deformation, by up to a factor of 4 [117,159]. 3.2.3. Time-dependent behavior In addition to pure chemical effects of CO2 in pores, i.e., dissolution and precipitation of minerals, a number of other fluid-controlled phys icochemical processes may operate at grain-to-grain contacts. Inter granular pressure solution [212,213] and subcritical cracking of grains [214,215] are commonly invoked to elucidate time-dependent rock deformation, i.e., rock deformation with time under elevated but con stant external stress, referred to as compaction creep [30,81,216]. These laboratory observations have implications for reservoir compaction and surface subsidence in the context of CO2 storage, more importantly, in depleted reservoirs that already underwent significant Pressure solution is a deformation mechanism where the concen tration of normal stress at grain contacts results in localized mineral 10 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. Fig. 9. Deformation and failure characteristics of Entrada Sandstones (a and b) and Summerville Siltstone (c and d) (modified from Espinoza et al. [115]). (a) X-ray CT section of the altered specimen with an evident single plane shear failure. The magnified CT image shows enhanced porosity (black area) by cement dissolution (red arrows); (b) stress-strain curves show brittle failure for both intact and altered Entrada Sandstone; (c) X-ray CT image illustrating initial heterogeneities (red arrows) which likely affect strain localization and failure pattern in the altered sample; and (d) stress-strain curves depicting brittle failure for intact Summerville Siltstone and ductile slightly strain-softening for altered ones. Fig. 9. Deformation and failure characteristics of Entrada Sandstones (a and b) and Summerville Siltstone (c and d) (modified from Espinoza et al. [115]). (a) X-ray CT section of the altered specimen with an evident single plane shear failure. The magnified CT image shows enhanced porosity (black area) by cement dissolution (red arrows); (b) stress-strain curves show brittle failure for both intact and altered Entrada Sandstone; (c) X-ray CT image illustrating initial heterogeneities (red arrows) which likely affect strain localization and failure pattern in the altered sample; and (d) stress-strain curves depicting brittle failure for intact Summerville Siltstone and ductile slightly strain-softening for altered ones. scarcity. It should also be pointed out that a significant body of literature explores the effect of CO2 as a working fluid for hydraulic fracturing and enhanced gas recovery from shales [77,142,228]. As a result, the focus is primarily on shale failure behavior while ignoring possible changes in transport properties that are crucial for risk assessment of CO2 leakage in geological storage. 4. CHM effects of CO2 on shales Geochemical processes of CO2 infiltration into water-saturated shales are fundamentally similar to those in sandstones and carbon ates [18,30,73]. However, as shales host ultrafine pores, chemical re actions between minerals and CO2 are controlled by intrinsically slow diffusive transport [145]. Batch experiments of long durations are thus suitable to study these reactions and their subsequent hydromechanical effects (Table S2). However, as the rate of diffusive penetration of the dissolved CO2 front (the reaction front) into a shale specimen is a few millimeters per month [145], batch experiments are very time-consuming, costly and technically challenging, justifying their 3.2.3. Time-dependent behavior Moreover, candidate shales for fracturing are often calcite- and/or quartz-rich with low clay contents, which makes them relatively brittle and frackable [50,210,229]. Hence these shales may not be representative of ductile clay-rich caprocks suitable for geological CO2 storage. time-dependent deformation in response to pore pressure decline [24, 197]. CO2 injection could exacerbate compaction creeping by promoting pressure solution in porous carbonate-rich formations while deceler ating it in quartz- and feldspar-bearing sandstones by attenuating grain-scale microcracking. Furthermore, CO2 injection is likely to pro gressively hamper such time-dependent behaviors in the near-wellbore region by pervasively drying out the reservoir [227]. 4.2. Geochemical effects on mechanical behavior: effect on caprock integrity The mechanical response of shales to geochemical processes induced by CO2 injection is critically important in terms of caprock integrity. The extent of reactions primarily depends on how CO2 infiltrates the shale, whether by molecular diffusion dissolved in water (i.e., zone C3 in Fig. 1) or by advection as a free phase (i.e., zone C1 in Fig. 1). In the former case, mineral dissolution dominates the mechanical behavior of altered shales, resulting in less stiff, less brittle and weaker rock frameworks [77,80,131,142,228]. The compressive and tensile strengths of shales exposed to dissolved CO2 degraded by up to 66% [130] and 27% [73], respectively. The shear failure of shaly caprocks is unlikely to signifi cantly affect their sealing potential because of their ductile nature, hindering the formation of conductive flow paths upon rupture [90]. In contrast, tensile fracturing has the potential to dramatically enhance the shale permeability, jeopardizing the permanent containment of CO2 underground [12]. Yet, stiffening and strengthening of shales due to the precipitation of secondary minerals, although underreported, can not be ruled out [115]. i Despite invaluable insights gained from laboratory observations into the mechanisms of dry CO2 interactions with shales, it remains unknown under which conditions and to what extent these mechanisms control the mechanical response of the caprock to CO2 intrusion. These mech anisms compete to potentially reinforce or compromise caprock integ rity. Although the high capillarity of shales likely limits the pervasive invasion of CO2 to the lowermost portion of the caprock (zone C1 [145]), shale (or clay) interactions with CO2 may become significant under certain circumstances, e.g., (1) the presence of annulus between cement and the shaly caprock, (2) formation of (micro)fractures, and (3) dissolution-enhanced porosity and permeability, altogether extending the CO2 rise and exposure to the caprock [106,111], (4) CO2 injection directly below an intraformational baffle (e.g., the Illinois Basin–Decatur project [246]) and (5) a reservoir containing non-negligible clay contents [236]. If the capillary entry pressure of the pore network, which commonly is in the order of a few megapascals, is exceeded, CO2 would enter the caprock in free phase [56,58]. In such situation, a series of other physico-chemical processes may gain prominence. CO2 intrusion volu metrically sweeps the pore fluid or evaporates the clay interlayer water at the CO2/aqueous phase interface, leading to rock dehydration [98]. 4.1. Changes in hydraulic properties: concerns about caprock sealing capacity In long-duration batch experiments for shales, calcite dissolution is accompanied by dissolution of slowly-reacting minerals, including feldspar, chlorite, kaolinite, and even quartz, along with clay mineral transformations (more importantly, illitization of smectite and mixed layer illite-smectite) [77,78,230,231]. Mineral transformations may contribute to changing hydraulic properties of the rock if the density of 11 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. expand and the shale to swell [235]. Likewise, physical or chemical adsorption of CO2 on clay platelets may lead to expansion, the extent of which is correlated with the amount of adsorbed CO2 – a function of temperature, pressure, interlayer water content and the surface area and charge of clay sheets [236,237] – and the stiffness of the clay minerals, among others [238]. The stiffer the shale, the lower the volumetric deformation [98]. Under no deformation constraints, montmorillonite (a subclass of the smectite clay group) has reportedly undergone expansion as high as 9% perpendicular to the layering direction [237]. In the subsurface, where the rock deformation is restricted, swelling of clay minerals may yield swelling stresses, which deteriorate the shales’s stiffness and strength if cementation bonds are overcome. While the mechanical weakening of water-sensitive shales resulting from hydra tion swelling has been broadly acknowledged [239–241], experimental evidence for similar weakening behaviors following CO2 adsorption-induced swelling is rare [242]. The synergy between direc tional swelling of clay-rich shales and deviatoric stresses originating from injection overpressure and cooling effects of the injected CO2 may cause (micro)fracturing and damage the caprock. the secondary minerals and, thus, the volume they occupy differ from the reactants [55]. The extent of these reactions increases with the exposure time and temperature [17,103,140,232]. Non-negligible dissolution of quartz can only be expected at notably high tempera tures (>150 ◦C [17]). The consequent changes in shale microstructure contribute to noticeable porosity enhancement of up to 10% [78,140, 230], which may even exceed those in reservoir rock samples [78,140]. ], y p [ , ] The risk of CO2 leakage through altered shales may substantially increase with small perturbations in porosity because they lead to orders of magnitude enhancement in shale permeability [77]. Indeed, the porosity-permeability relationships in low-permeability shales (in the nano-Darcy range) are found to follow power-laws with exponents as high as 16 (i.e., k = k0(φ/φ0)16, where k0 and φo are the initial permeability and porosity, respectively) [233]. 4.1. Changes in hydraulic properties: concerns about caprock sealing capacity Such a relationship im plies that a small porosity increase of 1% will lead to a permeability enhancement by a factor of roughly 3. In addition, dissolution-induced enlargement of pore throats can reduce the capillary entry pressure of shales to a large extent (by a factor of up to 10 following 5% porosity increase [78,140]), suggesting degradation of the capillary sealing ca pacity [55,56,58,228]. Nevertheless, batch experiments do not account for the potential coupling between multi-phase flow and chemical re actions. CO2 bubbles flow preferentially through large pores due to their lower capillarity, which may direct supersaturated water into low-velocity tiny pores, enhancing mineral precipitation. Although dry CO2 invasion into shales has been shown to cause mineral precipitation, resulting in slight permeability loss [73,131], the governing mecha nisms remain unresolved. Dehydration-induced shrinkage can also alter the microstructure of clay-rich shales. This mechanism has recently been shown to contribute to the stiffening and strengthening of saturated shales (up to 70% in crease in tensile and compressive strengths and 44% increase in Young’s modulus [111,243]). However, shrinkage remains a critical issue con cerning crack development [111]. At the pore scale, given that clays are water-wet in the presence of CO2 [244], capillary forces normal to the CO2-water interface invading the pores or clay platelets push the sur rounding particles away from the invasion area, resulting in crack for mation and propagation, a process referred to as desiccation cracking (see Ref. [245] for a detailed explanation). The formed cracks degrade the rock strength and provide conductive pathways for CO2 to migrate deeper into the caprock. Yet, desiccation cracking, a phenomenon observed in unconfined clays, is likely to be inhibited due to the high confining pressure at the depths of CO2 storage. 4.2. Geochemical effects on mechanical behavior: effect on caprock integrity Water progressively evaporates into the CO2 phase as far as the solubi lity limits and the equilibrium state are not achieved [31]. Dehydration increases the salt concentration, leading to salt precipitation when the water becomes oversaturated, as in the case of sandstones near the in jection well (see Section 3.1.1). The precipitated salt may clog the nanometer-sized shale pore throats, possibly decreasing its intrinsic permeability and strengthening the rock body, although these processes need to be experimentally verified. The described processes also have important implications for devising future laboratory plans to improve the understanding of CO2- shale interactions. Particularly, they explicitly outline the challenging essence of evaluating the mechanical properties of shales and their subsequent effects of exposure to CO2 under in-situ conditions. Even a small gain or loss in the water content or change in its activity can thoroughly change the shale behavior [106,111,247]. Consequently, the unsolicited wetting/drying effects have to be isolated from the physico-chemical process of CO2 injection, although not addressed by the vast majority of existing studies. Outcrop samples, which have tolerated varying degrees of weathering, can not represent deep, intact shale behaviors. Future laboratory experiments should be conducted on samples retrieved from the subsurface at relevant depths to CO2 storage, well preserved and handled to retain their innate water and original structure. Furthermore, variations in temperature and applied stresses may cause microstructural damage in shales and alter their porosity, permeability, and swelling tendency [248,249]. Accordingly, initial i Shales containing water-sensitive clay minerals, especially smectite, may shrink or swell in response to dehydration and hydration, respec tively [230,234]. Following hydration, the influx of water into the nanometer-spaced clay sheets causes the layered structure of the clay to 12 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 equilibration of the shale specimen with the desired in-situ conditions and maintaining them throughout CO2 injection has to be assured to avoid experimental artifacts. pore scale phenomena to occur concurrently or sequentially to control fracture evolution. Dissolution of the more reactive mineral leaves a trail of uncemented, less reactive minerals filling the fracture. The less reactive mineral particles may become contacting asperities [272] and/or progressively reorganize as a diffusive barrier on the reactant grains (e.g., microporous clay coating on calcite grains in an argillaceous limestone), dramatically decreasing the fracture dissolution rate [105, 108]. Therefore, the spatial pattern of minerals is the key determinant of flow trajectories. 4.2. Geochemical effects on mechanical behavior: effect on caprock integrity An extreme flow behavior happens as the less reacting minerals spread out over the fracture surface or form repetitive layering orthogonal to the flow direction, similar to shale beddings. In this sce nario, normal fracture compaction hinders channeling and leaves only flow bottlenecks, drastically decreasing fracture permeability, report edly by up to two orders of magnitude [272]. This self-sealing phe nomenon is also plausible if the less-reactive minerals, particularly clays in shales, are removed from the fracture surface. The massive displacement of clay particles may clog the fracture [105]. On the contrary, for a nodular distribution pattern of less reacting minerals, warranting continuous bands of reactive minerals along the flow di rection, channeling takes place around the nodular areas, acting as persistent asperities that sustain mechanical loads [108]. These labo ratory observations can be numerically reproduced by coupling reactive transport and geomechanical processes [119]. 5. Geochemical processes for CO2 injection into fractured rocks Fractures can create rapid flow pathways through the reservoir and seal rocks. Fracture permeability may be several orders of magnitude greater than that of the surrounding rock matrix, depending primarily on the roughness and degree of mismatch of fracture surfaces and the effective stress state [250–252]. A sufficiently high pressure gradient drives a long-lasting localized flow of fluids, likely out of chemical equilibrium with minerals on the fracture surfaces, rendering them prone to chemical reactions [253]. Mineral precipitation and/or disso lution have been widely acknowledged from a diagenetic perspective, under prolonged deep burial and elevated temperature conditions, to extensively alter the fracture geometry and mineralogy and, thus, their hydraulic and mechanical properties [211,254]. Multiple lines of evi dence from along-fault leakage of CO2-charged fluid from natural ac cumulations over geological time scales at the Colorado Plateau region in Utah, US, including severe carbonate mineralization in the fracture network [255,256], also reproduced by numerical simulations [257], lend support to potential reactive CO2 flow in fractures in the course of geologic storage. Despite observations of geochemically-modulated al terations of fracture mineralogy, geochemical controls of CO2 injection on hydromechanical properties of fractures are still poorly constrained [252]. Basin-wide pressurization in gigatonne-scale injection may pose a risk to safe CO2 storage by triggering perceivable seismicity, compro mising the wellbore integrity, and creating leaking pathways through the caprock [193,273,274]. Self-healing of fractures in the reservoir boosts overpressurization, whereas flow channeling leads to injectivity enhancement. Similarly, these mechanisms can be of utmost importance to the leakage risk through caprocks, as conductive fractures become primary threats to the caprock sealing capacity [253]. Although the geochemical effects on the flow field are observed in the short laboratory experiments, their occurrence in the subsurface may be delayed or accelerated by variations in the fracture length, rock composition, pore pressure, and stress state [119,122,272]. It is also questionable whether the spatial patterns of less-reactive minerals observed in the laboratory can be uniformly distributed across scales. In fact, vertically heteroge neous fractures/faults crossing multiple stratigraphic layers hinder up ward CO2 migration [275]. If the stratigraphic layers are of different chemical reactivity, the caprock sealing capacity would be even increased. 5.1. Evolution of reaction and flow trajectories Insights from recent experiments on fractured rock imply the pri mary control of the injection rate on geochemical reactions along the fracture [120,122,258]. Generally, at high injection rates, favoring short residence compared to the reaction time (low Da number), minerals dissolve almost uniformly throughout the fracture [120]. In contrast, low injection rates, reproducing high Da numbers, result in channeled dissolution and fracture enlargement [122,259]. Flow channeling es tablishes positive feedback between reaction progress and enlargement of fracture aperture, which may be accompanied by channel break through along the rock and, thus, runaway permeability growth [119, 120,122]. Caution should be taken, however, when upscaling these re action patterns to reservoir conditions where fracture length can vary across many orders of magnitude [260]. The residence time increases with the fracture length, causing reaction front instabilities and channel formation to occur at significantly higher injection rates than the ones registered in the laboratory [261]. In addition, precipitation may also take place if the fluid flowing through the fracture becomes supersatu rated [262]. Mineral precipitation may take place in several conditions, e.g., (1) when CO2-rich fluid rises along a fracture and CO2 exsolution following pressure reduction results in pH increase and carbonate mineral precipitation [254], (2) locally at smaller aperture portions of a single fracture, where the water-to-mineral ratio is lowest [263,264] or (3) at fracture intersections where fluid mixing or splitting assist in attaining the supersaturation conditions [265]. l 5.2. Localized deformation and fracture propagation Reactive transport of CO2 in fractures may alter their mechanical characteristics because the non-uniform erosion of the fracture surface increases the fracture roughness and local stress distribution is mainly transformed from intact asperities to new uncemented contacts. Nu merical simulations show that the described geochemical processes divide the fracture surface into two regions: a flattened contact area, which rapidly stiffens even at low stress levels and does not contribute to further normal deformation afterward, and a rough surface that bounds the dissolution-induced channel and is susceptible to compaction at high stresses [276]. Thus, the lower the contact surface area and poorly developed surface mating (higher roughness), the lower the fracture stiffness but the more sensitive the stiffness and flow to stress changes [270]. Laboratory experiments of CO2 injection, albeit rare, give credence to insights obtained from simulations and conceptual models. Skurtveit et al. [116] highlight markedly stress-sensitive flow in conductive fractures naturally exposed to dissolved CO2, featuring normal stiffness values lower than half of those measured for low-permeability unaltered fractures. These observations point to the tightly coupled nature of fluid flow, deformation and geochemical re actions in fractures. Accordingly, Pyrak-Nolte and Nolte [277] suggested considering geochemical alterations of fractures in a universal flow-stiffness scaling relationship. Furthermore, the localized The evolution of flow and mechanical deformation that may change the dissolution patterns is profoundly affected by mineralogical het erogeneities, not only in abundance but also in spatial distribution [110, 266,267], and applied stresses [119,268]. In homogeneous mono-mineral fractured rocks, e.g., carbonate rocks, uniform mineral dissolution gives rise to stress concentration at contacting asperities of fracture walls, promoting pressure solution [269] or brittle failure once the stress surpasses the compressive strength of asperities [268]. The consequence is a competition between dissolution-enhanced flow and permeability loss caused by the well-documented mechanical compac tion [270,271]. In heterogeneous multi-mineral rocks, the different solubility and reaction kinetics of the minerals enable a number of other 13 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 time-dependent deformation of altered fractures may dominate the volumetric response of the reservoir and become relevant to risks associated with reservoir compaction or induced seismicity [278]. Our understanding of the frictional behavior of fractured rocks is limited to a number of CO2 injection experiments on fault gouge that better represent a mature, gouge-bearing fault. 5.2. Localized deformation and fracture propagation Friction coefficient values for treated fault gouges are systematically inconsistent, yielding negligible changes in some claystone and sandstone [112] or reduction of up to 15% in an anhydrite sample [281]. At experimental slip and time scales, CO2 injection does not impose any clear, strong influence on the slip velocity dependence. The slip behavior remains predominantly velocity-strengthening, meaning that the friction strength increases as the slip accelerates and seismic activities are unlikely [282]. These ob servations imply that chemical weakening of the friction coefficient may drive slip on fractures that could otherwise remain stable, given the excess pore pressure and shear stress buildup or release by poroelastic and thermoelastic effects [14]. CO2 injection does not change the pro pensity of reservoir and caprock faults to slip aseismically, not to negate induced seismicity risk in CCS projects. Yet, aseismic slip may enhance Similar to fracture stiffness, experimental measurements to study the influence of CO2 injection on fracture strength are scarce. Opening- mode fracture toughness as an indicator of the fracture resistance against propagation under tensile load has been found to systematically increase and decrease by up to two folds with carbonate cement pre cipitation and dissolution, respectively, in a suite of siliciclastic samples diagenetically altered by CO2-rich water [113]. The resulting weakening effects may threaten caprock integrity far beyond the CO2 injection and monitoring period. Coupled chemical-mechanical numerical models have been developed to address the weakening (strengthening) impacts of calcite dissolution (precipitation) on fracture initiation and/or tensile propagation in calcite-rich or -cemented rocks across spatial and tem poral scales [279,280]. ig. 10. An overview of the geochemical effects of CO2 injection on hydromechanical properties of the reservoir and sealing rocks and the associated ke ncluding reservoir injectivity, deformation and integrity, caprock integrity and sealing capacity and induced seismicity. Note that zones A1 to A5 in the rese 1 to C3 in the caprock are defined in the same way as in Fig. 1. Fig. 10. An overview of the geochemical effects of CO2 injection on hydromechanical properties of the reservoir and sealing rocks and the associated key issues, including reservoir injectivity, deformation and integrity, caprock integrity and sealing capacity and induced seismicity. Note that zones A1 to A5 in the reservoir and C1 to C3 in the caprock are defined in the same way as in Fig. 1. 14 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. 5.2. Localized deformation and fracture propagation fracture permeability, reaping the benefits of enhanced injectivity in the reservoir [274], but raising significant concerns around the loss of caprock integrity [12]. variables (pressure, temperature, fluid chemistry, applied stresses), intrinsic rock properties (mineralogy, microstructure, porosity and permeability), and the hydrodynamics of CO2 flow and transport in rock. l This review provides a perspective on the geochemical interactions of supercritical CO2 with reservoir (carbonates, sandstones) and sealing rocks (shales) crucial to ensuring secure geological storage of CO2. The identified gaps in our fundamental understanding of coupled CHM processes, their monitoring, prediction, and upscaling to industrial-scale CO2 injection warrant further research. Fig. 10 summarizes the insights gained from this review into the potential impacts of CO2-water-rock interactions on hydromechanical rock properties and their implications for geological CO2 storage, including the influences on reservoir injec tivity and integrity, reservoir compaction and surface subsidence, wellbore integrity and caprock sealing capacity and integrity. l 6.1. Fundamental understanding of the CHM processes of CO2 injection 6.1. Fundamental understanding of the CHM processes of CO2 injection Laboratory experiments with reservoir and caprock specimens reacting at the P-T ranges relevant to geological CO2 storage provide fundamental insights into the processes that control CO2-water-rock interactions. The CO2 fraction that dissolves in the pore fluid forms acidic CO2-rich water (pH ≈3–5) that induce the dissolution of primary minerals and the precipitation of secondary minerals. These reactions lead to major changes in sedimentary rock properties, including mineralogy, pore structure, porosity, permeability, stiffness, strength and time-dependent deformation. The non-dissolved supercritical CO2 (dry supercritical CO2) barely reacts with primary minerals but dries out the formation fluids and causes salt precipitation, leading to alterations in hydromechanical properties. It is inferred that mineral reactions primarily depend on the content of highly soluble carbonate minerals. Moreover, the extent of reactions is a complex function of environmental Carbonates: In laboratory experiments, flow conditions affect the geochemical processes. Exposure of carbonate rocks to CO2-rich water under no-flow conditions results in minor changes in rock composition because the CO2-rich solution rapidly reaches equilibrium with calcite, hindering further interactions [101,143,147,283]. The most significant changes occur in cores flooded with CO2-rich water under open-flow conditions with a continuous renewal of the acidic solution at the inlet, where minerals dissolve and porosity and permeability increase, 11. Box-and-whisker plots for changes in porosity (a), permeability (b), Young’s modulus (c), and Poisson’s ratio (d). The CO2 exposure types, batch, flow- gh, or natural, are highlighted in different colors. The dividing lines in the boxes indicate the median values, and the bottom and top edges illustrate the nd third quartiles, respectively. The whiskers extend to the most extreme data points not statistically considered outliers. The permeability enhancement of 125 measured by Noiriel et al. [65] for L´erouville limestone is not illustrated in the figure for the sake of more clear observation of other experimental data. Data in Table S3 used to plot this figure. Fig. 11. Box-and-whisker plots for changes in porosity (a), permeability (b), Young’s modulus (c), and Poisson’s ratio (d). The CO2 exposure types, batch, flow- through, or natural, are highlighted in different colors. The dividing lines in the boxes indicate the median values, and the bottom and top edges illustrate the first and third quartiles, respectively. The whiskers extend to the most extreme data points not statistically considered outliers. The permeability enhancement of 125 folds measured by Noiriel et al. 6.1. Fundamental understanding of the CHM processes of CO2 injection [65] for L´erouville limestone is not illustrated in the figure for the sake of more clear observation of other experimental data. Data listed in Table S3 used to plot this figure. 15 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 resulting in mechanical weakening (Fig. 11). resulting in mechanical weakening (Fig. 11). observations. Nevertheless, free-phase CO2 intrusion into shales is un likely owing to their high capillarity. The knowledge of CHM coupling in shales remains undeveloped and should be subject to complementary experimental, observational and modeling research. Additional studies are warranted to elucidate the interplay between physical and chemical processes of CO2 interactions with clay minerals. The structure and heterogeneity of limestones affect the chemically- induced evolution of the hydromechanical properties. Given the high porosity, smaller grain-to-grain contacts and large reactive surface area, grain-supported carbonate rocks (chalks) commonly endure the most considerable CHM coupling effects both in the short and long terms. The rapid dissolution of carbonate minerals (e.g., calcite) when subjected to CO2-rich water circulation depends on the inherent heterogeneity of the rock and leads to the formation of highly conductive flow channels, i.e., wormholes. Wormholes are susceptible to collapse under the elevated reservoir confinement pressure, potentially speeding up reservoir compaction and reducing reservoir injectivity. Fractured rocks: Chemical reactions of CO2 with fractured caprocks or reservoir rocks and the ensuing changes in flow and mechanical behavior are tightly coupled and primarily controlled by the injection rate, length scale, fracture roughness, applied stresses and the content and spatial distribution of reactive minerals. Low injection rates (or at several tens to hundreds of meters from the injection well as the flux decreases with the logarithm of distance from the well) and mineral heterogeneity, featuring a continuous path of reactive minerals along the flow direction, favor fracture channeling. This dissolution regime results in runaway permeability growth, smaller stiffness, and heavily stress-dependent flow. In contrast, CO2 injection at high rates into ho mogeneous fractures or certain distribution of less-reactive minerals on the fracture surface may impede channeled flow and reactions. The consequence would be rapid compaction, stiffening, and self-sealing. The use of dimensionless P´eclet and Damk¨ohler numbers and the char acterization of rock fabric evolutions across scales may be helpful to upscale these dissolution patterns to the field. Chemically-assisted flow channeling may increase reservoir injectivity, while it may jeopardize the caprock sealing capacity. 6.1. Fundamental understanding of the CHM processes of CO2 injection To move the field forward, we suggest focusing future research on the following directions. • Experimental and modeling studies on reactive CO2 transport in fractured rocks should accurately account for stress states represen tative of reservoir conditions. Neglecting the confining stress, using artificial fractures with unrealistic roughness and aperture, and holding the fracture open, e.g., by using proppant, may overlook geomechanical processes, leading to overestimation of the dissolution-induced permeability enhancement. i • At the field scale, fractures have presumably undergone prolonged shearing or may experience slip due to physical and chemical per turbations caused by CO2 injection. Hence, fracture roughness is modulated, microfractures may branch, fine-grained gouge material forms and similar lithologies already juxtaposed may be offset along the flow direction [286,287]. Laboratory experiments on artificial shear-mode fractures offer an opportunity to study the evolution of flow and shear deformation in such morphologically heterogeneous fractures. Furthermore, CO2 or CO2-rich water can be used as the working fluid for fracture slip experiments to observe the combined effects of fracture reactivation and chemistry on flow and deformation. i Shales: Technical challenges to characterize low-permeability shales have significantly limited the experimental assessment of CO2 exposure effects on shaly caprocks. Usually, shales are treated under no-flow conditions to replicate diffusive CO2 leakage through the caprock. Under these conditions, an increase in dissolution-driven mechanical weakening and an increase in permeability and attenuation of capillarity occur. These changes in shale properties can be even more striking than those of altered reservoir rocks (Fig. 11), with the potential to endanger the caprock sealing capacity. • Laboratory experiments have mainly focused on single fractures, and a knowledge gap in the presence of fracture intersections is evident. Fluid mixing where fractures of different origins converge or split ting where multiple fractures branch may locally alter the fluid ve locity field and chemistry [265]. Laboratory studies of fracture sets or networks can benefit from breakthroughs in 3D X-ray tomography at representative underground conditions [288] and from a combi nation of etching, machining, and fast-evolving 3D printing tech niques to produce fracture networks in a repeatable manner [289, 290]. 6.1. Fundamental understanding of the CHM processes of CO2 injection Furthermore, geochemical processes may degrade both tensile and frictional strengths of fractures, raising con cerns about loss of caprock integrity during and beyond the injection period. At the moment, a better characterization of the processes leading to localized flow and deformation in wormholes is necessary (1) to eval uate the induced CHM effects as the continuity of the porous network is questioned and (2) to upscale the results from the laboratory to the field as the rock sample is no longer a representative elementary volume. Attempts to find appropriate injection rates to approach uniform dissolution patterns have been made by adjusting the Pe and Da numbers. Further research using real-time and high-resolution imaging of reacting rocks combined with reactive transport modeling would be highly instructive [109,284]. g y Sandstones: In sandstones, the content of carbonate minerals and their load-bearing role in the pore structure determine the fate of the mechanical properties. Quartz-rich sandstones undergo trivial in teractions with CO2 because of the slow dissolution kinetics of silicates, resulting in slight alterations in hydromechanical properties [17]. Measurable amounts of dissolved quartz over the time-scale of labora tory experiments are only expected under high-temperature conditions (e.g., T > 150 ◦C) can be anticipated [44,69,141]. In contrast, sand stones with intergranular carbonate cement have the most variable structures. Dissolution of a small amount of load-bearing carbonate cement can provoke drastic declines in stiffness and strength and cause time-dependent deformation by promoting pressure solution at grain contacts. Clay and/or feldspar (alumino-silicate)-cemented sandstones with minor carbonate content can also undergo significant alterations through dissolution and transformation of the cement material. There fore, the extent of alterations in hydromechanical properties depends on the mineralogical composition of sandstones. Considerable dissolution-enhanced porosity and permeability, and mechanical weakening greater than those of carbonates, can occur (Fig. 11). Although alterations in sandstone properties under no-flow and open-flow conditions are comparable (Fig. 11a and b), batch and flow-through experiments with the same sandstone samples and similar experimental conditions should be performed to more accurately isolate the effect of rock-to-CO2 exposure hydrodynamics. In addition, given the scarcity of knowledge on relative permeability hysteresis (two-phase flow) in chemically-altered rocks (all rock types) [185,285], an improved understanding of this property is essential in order to better assess the efficiency of CO2 residual trapping. We conclude that despite recent advances, chemo-mechanical coupling in fractured rocks, particularly in the context of CO2 injec tion, is underappreciated. 6.1. Fundamental understanding of the CHM processes of CO2 injection l If dry CO2 reaches the caprock (note that the injected dry CO2 takes water as it advances through the reservoir, i.e., zone C2), the water within the pores or clay interlayers could be dehydrated, giving rise to self-sealing mechanisms and mechanical strengthening or self- enhancing and mechanical weakening, the influence of which is controversial [98,111,244]. Recent studies on the sealing capacity of caprock analogue specimens invaded by CO2 also yield inconsistent • Microfluidics and transparent analogue systems are promising techniques to visualize and elucidate coupled (multi-phase) flow, transport, deformation and geochemical processes in fractures under in-situ pressure, temperature and stress conditions. Three notable • Microfluidics and transparent analogue systems are promising techniques to visualize and elucidate coupled (multi-phase) flow, transport, deformation and geochemical processes in fractures under in-situ pressure, temperature and stress conditions. Three notable 16 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. examples of using these techniques to study chemically-eroded fractures during CO2 injection are Detwiler [268], Fazeli et al. [291] and Jimenez-Martinez et al. [263]. examples of using these techniques to study chemically-eroded fractures during CO2 injection are Detwiler [268], Fazeli et al. [291] and Jimenez-Martinez et al. [263]. (CO2LPIE) at the Mont Terri rock laboratory, Switzerland, which aims at understanding coupled flow, geochemical and geomechanical processes for CO2 injection into Opalynus Clay, a representative caprock [297]. • Multiple physical processes may simultaneously operate during CO2 injection into fractured rocks. Physics-based models and numerical simulations are well-positioned to diagnose the relative importance and contribution of each process in the experiments. In addition, machine-learning approaches will improve the characterization and analysis of flow and fracture evolution by extracting information from massive laboratory signals and images [292]. The progress of CO2-water-rock interactions in subsurface systems critically depends on the hydrodynamics of the CO2 plume propagation in the reservoir and caprocks. Reproduction of these scenarios is com plex and demands laboratory experiments with different CO2 exposures and physical conditions. Coupled flow, reactive transport and geo mechanical modeling is a powerful tool for quantitatively linking laboratory-scale observations, intermediate-scale rock laboratory ex periments and industrial-scale CO2 injection [43,298,299]. At the moment, a fully coupled CHM modeling package is scarce owing to its complexities and multidisciplinary nature (see Ilgen et al. [194] and Viswanathan et al. [300] for a review of these models). 6.2. Upscaling in time and space Laboratory experiments to understand CO2-rich water-rock in teractions have provided insights into short-term geochemical processes (up to 2–3 years). However, long-term reaction mechanisms, particu larly those of quartz reaction, are less well understood and need further research [17,18]. It has been demonstrated that evaluating naturally altered intact and fractured rocks exposed to CO2 over geological time scales is essential to understand long-term CHM processes. Overall, the observed changes in hydraulic and mechanical properties of these samples are within the wide ranges recognized for laboratory-treated specimens (Fig. 11). However, samples taken from outcrops may differ from those retrieved from representative depths for CCS because they were commonly exposed to weathering and underwent smaller stresses. Thus, geological intuitions of CO2-rock interactions sourced from outcrop observations should be treated with caution because reactive transport regimes that cause the hydraulic and mechanical properties of (fractured) rocks to evolve are strongly stress-sensitive. Moreover, outcrop samples probably endured various loading paths (see Ref. [211] for a review of relevant mechanisms), e.g., mechanical alterations due to their proximity to faults as the main conduits for CO2 leakage [149,253]. Thus, differentiating between the coupled CHM effects of CO2 and other processes operating over geological time scales is vital, although quite challenging [115]. Integrating reactive transport modeling and thermo-hydromechanical simulators to reproduce these natural features may unlock new insights into their origin [293]. Numerical simulations help draw more accurate constraints on the kinetic rates of slow-reacting minerals (e.g., quartz) whose uncertainties in predicting the fate of CO2 underground could grow nonlinearly over the geologic time of interest to CO2 storage [18,294]. i Acknowledgments This work has received funding from European Joint Programme on Radioactive Waste Management, EJP-EURAD of the the European Union’s Horizon 2020 Research and Innovation Programme 2014–2018 under agreement No. 847593, specifically from the WP MAGIC. I.R.K. and V.V. acknowledge funding from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Inno vation Program through the Starting Grant GEoREST (www.georest.eu) (Grant agreement No. 801809). I.R.K. also acknowledges support by the PCI2021-122077-2 B project (www.easygeocarbon.com) funded by MCIN/AEI/10.13039/501100011033 and the European Union Next GenerationEU/PRTR. J.C. and J.M.S. also acknowledge funding from the Catalan Government through project 2017SGR1733. IDAEA-CSIC is a Centre of Excellence Severo Ochoa (Spanish Ministry of Science and Innovation, Project CEX2018-000794-S funded by MCIN/AEI/ 10.13039/501100011033). This research has been carried out within the framework of the activities of the Spanish Government through the “Maria de Maeztu Centre of Excellence” accreditation to IMEDEA (CSIC- UIB) (CEX2021-001198). Open access funding is provided by the Spanish National Research Council (CSIC). The authors acknowledge Dr. Nicolas Espinoza for providing source figures to prepare Fig. 9. Extending the research to field-scale injections, with all accompa nying geologic complexities and heterogeneities, could see significant developments using purpose-designed, densely monitored experiments in highly characterized, in-situ, tens-of-meters-scale underground research laboratories [295]. These experiments are particularly valuable to advance our fundamental understanding of relevant physical pro cesses gained from core-scale experiments and develop and test con ceptual models, numerical approaches, and cutting-edge real-time monitoring technologies under more realistic conditions [296]. A pertinent example is the CO2 Long-term Periodic Injection Experiment Declaration of competing interest 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. 6.1. Fundamental understanding of the CHM processes of CO2 injection Further efforts should be dedicated to developing CHM constitutive models that work for different rock types. By reproducing laboratory experiments, these numerical models would be validated to ensure upscaling of the key observed phenomena. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.rser.2023.113270 Appendix A. Experimental apparatus Appendix A. Experimental apparatus Fig. A1 illustrates schematics of the two types of experimental apparatus used to inject CO2 into rock samples 17 17 Renewable and Sustainable Energy Reviews 178 (2023) 113270 A. Vafaie et al. Schematic experimental setups used to expose rock samples to CO2: (a) static (no flow) batch and (b) flow-through experiments. In the flow-through setup, 3 represent axial and radial (confining) stresses, respectively, and V1 and V2 are three-sided valves allowing free-phase CO2 or CO2-rich water to be injected. nces chu S, Adams J. Sequestration of CO2 in geological media in response to forward. Energy Environ Sci 2018;11:1062–176. https://doi.org/10.1039/ C7EE02342A. [5] Benson S, Cole DR. CO2 Sequestration in deep sedimentary formations. Elements 2008;4:325 31 https://doi org/10 2113/gselements 4 5 325 A1 Schematic experimental setups used to expose rock samples to CO : (a) static (no flow) batch and (b) flow through experiments In the flow throug Fig. A1. Schematic experimental setups used to expose rock samples to CO2: (a) static (no flow) batch and (b) flow-through experiments. In the flow-through setup, σ1 and σ3 represent axial and radial (confining) stresses, respectively, and V1 and V2 are three-sided valves allowing free-phase CO2 or CO2-rich water to be injected. Fig. A1. Schematic experimental setups used to expose rock samples to CO2: (a) static (no flow) batch and (b) flow-through experiments. In the flow-through setup, σ1 and σ3 represent axial and radial (confining) stresses, respectively, and V1 and V2 are three-sided valves allowing free-phase CO2 or CO2-rich water to be injected. forward. Energy Environ Sci 2018;11:1062–176. https://doi.org/10.1039/ C7EE02342A. forward. Energy Environ Sci 2018;11:1062–176. h C7EE02342A. forward. Energy Environ Sci 2018;11:1062–176. https://doi.org/10.1039/ C7EE02342A. [5] Benson S, Cole DR. CO2 Sequestration in deep sedimentary formations. Elements 2008;4:325–31. https://doi.org/10.2113/gselements.4.5.325. [6] Oelkers EH, Cole DR. Carbon dioxide sequestration a solution to a global problem. Elements 2008;4:305–10. https://doi.org/10.2113/gselements.4.5.305. [7] Zhou C, Birkholzer JT, Mehnert E, Lin YF, Zhang K. Modeling Basin- and plume- scale processes of CO2 storage for full-scale Deployment. Ground Water 2009;48: 494–514. https://doi.org/10.1111/j.1745-6584.2009.00657.x. [8] Bandilla KW, Celia MA. Active pressure management through brine production for basin-wide deployment of geologic carbon sequestration. Int J Greenhouse Gas Control 2017;61:155–67. https://doi.org/10.1016/j.ijggc.2017.03.030. [9] Szulczewski ML, MacMinn CW, Herzog HJ, Juanes R. Lifetime of carbon capture and storage as a climate-change mitigation technology. Proc Natl Acad Sci USA 2012;109:5185–9. https://doi.org/10.1073/pnas.1115347109. [6] Oelkers EH, Cole DR. Carbon dioxide sequestration a solution to a global problem. Elements 2008;4:305–10. https://doi.org/10.2113/gselements.4.5.305. [8] Bandilla KW, Celia MA. Active pressure management through brine production for basin-wide deployment of geologic carbon sequestration. Int J Greenhouse Gas Control 2017;61:155–67. https://doi.org/10.1016/j.ijggc.2017.03.030. [7] Zhou C, Birkholzer JT, Mehnert E, Lin YF, Zhang K. Modeling Basin- and plume- scale processes of CO2 storage for full-scale Deployment. Ground Water 2009;48: 494–514. https://doi.org/10.1111/j.1745-6584.2009.00657.x. [5] Benson S, Cole DR. CO2 Sequestration in deep sedimentary formations. Elements 2008;4:325–31. https://doi.org/10.2113/gselements.4.5.325. References [37] Loring JS, Thompson CJ, Wang Z, Joly AG, Sklarew DS, Schaef HT, Ilton ES, Rosso KM, Felmy AR. In situ infrared spectroscopic study of forsterite carbonation in wet supercritical CO2. Environ Sci Technol 2011;45(14):6204–10. https://doi. org/10.1021/es201284e. [13] Kim S, Hosseini SA. Above-zone pressure monitoring and geomechanical analyses for a field-scale CO2 injection project in Cranfield, MS. Greenhouse Gases Sci Technol 2013;4:81–98. https://doi.org/10.1002/ghg.1388. [38] Oelkers EH, Gislason SR, Matter J. Mineral carbonation of CO2. Elements 2008;4: 333–7. https://doi.org/10.2113/gselements.4.5.333. [14] Rutqvist J. The geomechanics of CO2 storage in deep sedimentary formations. Geotech Geol Eng 2012;30:525–51. https://doi.org/10.1007/s10706-011-9491- 0. [39] Rosenbauer RJ, Thomas B, Bischoff JL, Palandri J. Carbon sequestration via reaction with basaltic rocks: geochemical modeling and experimental results. Geochem Cosmochim Acta 2012;89:116–33. https://doi.org/10.1016/j. gca.2012.04.042. [15] Zoback MD, Gorelick SM. Earthquake triggering and large-scale geologic storage of carbon dioxide. Proc Natl Acad Sci USA 2012;109:164–8. https://doi.org/ 10.1073/pnas.1202473109. g [40] Assayag N, Matter J, Ader M, Goldberg D, Agrinier P. Water–rock interactions during a CO2 injection field-test: implications on host rock dissolution and alteration effects. Chem Geol 2009;265(1–2):227–35. https://doi.org/10.1016/j. chemgeo.2009.02.007. p [16] Vilarrasa V, Carrera J. Geologic carbon storage is unlikely to trigger large earthquakes and reactivate faults through which CO2 could leak. Proc Natl Acad Sci USA 2015;112:5938–43. https://doi.org/10.1073/pnas.1413284112. [41] Zhang Y, Zhang Z, Arif M, Lebedev M, Busch A, Sarmadivaleh M, Iglauer M. Carbonate rock mechanical response to CO2 flooding evaluated by a combined X- ray computed tomography – DEM method. J Nat Gas Sci Eng 2020;84:103675A. https://doi.org/10.1016/j.jngse.2020.103675. g [17] Gaus I. Role and impact of CO2–rock interactions during CO2 storage in sedimentary rocks. Int J Greenh Gas Control 2010;4:73–89. https://doi.org/ 10.1016/j.ijggc.2009.09.015. ll j jgg [18] Kampman N, Bickle MJ, Wigley M, Dubacq B. Fluid flow and CO2-fluid-mineral interactions during CO2 storage in sedimentary basins. Chem Geol 2013;369: 22–50. https://doi.org/10.1016/j.chemgeo.2013.11.012. g j j g [42] Rochelle CA, Czernichowski-Lauriol I, Milodowski AE. The impact of chemical reactions on CO2 storage in geological formations: a brief review. Geological Society London Special Publications 2004;233:87–106. https://doi.org/10.1144/ GSL.SP.2004.233.01.07. [19] Pawar R, Guthrie G. Hydraulic-Chemical coupling associated with the injection and storage of CO2 into the subsurface. Science of Carbon Storage in Deep Saline Formations 2019:291–309. https://doi.org/10.1016/B978-0-12-812752- 0.00013-7. [43] Dai Z, Viswanathan H, Xiao T, Hakala A, Lopano C, Guthrie G, McPherson B. Reactive transport modeling of geological carbon storage associated with CO2 and brine leakage. Science of Carbon Storage in Deep Saline Formations 2019:67–88. https://doi.org/10.1016/B978-0-12-812752-0.00005-8. References [1] Bachu S, Adams J. Sequestration of CO2 in geological media in response to climate change: capacity of deep saline aquifers to sequester CO2 in solution. Energy Convers Manag 2003;44:3151–75. https://doi.org/10.1016/S0196-8904 (03)00101-8. ( ) [2] Celia MA. Geological storage of captured carbon dioxide as a large-scale carbon mitigation option. Water Resour Res 2017;53:3527–33. https://doi.org/10.1002/ 2017WR020841. [8] Bandilla KW, Celia MA. Active pressure management through brine production for basin-wide deployment of geologic carbon sequestration. Int J Greenhouse Gas Control 2017;61:155–67. https://doi.org/10.1016/j.ijggc.2017.03.030. [3] Ringrose PS, Meckel TA. Maturing global CO2 storage resources on offshore continental margins to achieve 2DS emissions reductions. Sci Rep 2019;9:17944. https://doi.org/10.1038/s41598-019-54363-z. p g j jgg [9] Szulczewski ML, MacMinn CW, Herzog HJ, Juanes R. Lifetime of carbon capture and storage as a climate-change mitigation technology. Proc Natl Acad Sci USA 2012;109:5185–9. https://doi.org/10.1073/pnas.1115347109. [9] Szulczewski ML, MacMinn CW, Herzog HJ, Juanes R. Lifetime of carbon capture and storage as a climate-change mitigation technology. Proc Natl Acad Sci USA 2012;109:5185–9. https://doi.org/10.1073/pnas.1115347109. p g [4] Bui M, Adjiman CS, Bardow A, Anthony EJ, Boston A, Brown S, Fennell PS, Fuss S, Galindo A, Hackett LA. Carbon capture and storage (CCS): the way 18 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 [10] De Simone S, Jackson SJ, Krevor S. The error in using superposition to estimate Pressure during multi-site subsurface CO2 storage. Geophys Res Lett 2019;46: 6525–33. https://doi.org/10.1029/2019GL082738. [35] Miller QRS, Thompson CJ, Loring JS, Windisch CF, Bowden ME, Hoyt DW, Hu JZ, Arey BW, Rosso KM, Schaef HT. Insights into silicate carbonation processes in water-bearing supercritical CO2 fluids. Int J Greenh Gas Control 2013;15:104–18. https://doi.org/10.1016/j.ijggc.2013.02.005. g [11] Birkholzer JT, Zhou Q, Tsang CF. Large-scale impact of CO2 storage in deep saline aquifers: a sensitivity study on pressure response in stratified systems. Int J Greenh Gas Control 2009;3:181–94. https://doi.org/10.1016/j. ijggc.2008.08.002. p g j jgg [36] Kwak JH, Hu JZ, Turcu RV, Rosso KM, Ilton ES, Wang C, Sears JA, Engelhard MH, Felmy AR, Hoyt DW. The role of H2O in the carbonation of forsterite in supercritical CO2. Int J Greenh Gas Control 2011;5(4):1081–92. https://doi.org/ 10.1016/j.ijggc.2011.05.013. [12] Guglielmi Y, Nussbaum C, Cappa F, De Barros L, Rutqvist J, Birkholzer J. Field- scale fault reactivation experiments by fluid injection highlight aseismic leakage in caprock analogues: Implications for CO2 sequestration. Int J Greenhouse Gas Control 2021;111:103471. https://doi.org/10.1016/j.ijggc.2021.103471. References [20] Nguyen BN, Hou Z, Bacon DH, Murray CJ, White MD. Three-dimensional modeling of the reactive transport of CO2 and its impact on geomechanical properties of reservoir rocks and seals. Int J Greenhouse Gas Control 2016;46: 100–15. https://doi.org/10.1016/j.ijggc.2016.01.004. p g [44] Rosenbauer RJ, Koksalan T, Palandri JL. Experimental investigation of CO2–brine-–rock interactions at elevated temperature and pressure: implications for CO2 sequestration in deep saline aquifers. Fuel Process Technol 2005:1581. https://doi.org/10.1016/j.fuproc.2005.01.011. –1597. p g j jgg [21] Zhang R, Wu YS. Hydrologic, mechanical, thermal, and chemical process coupling triggered by the injection of CO2. Science of Carbon Storage in Deep Saline Formations 2019:291–309. https://doi.org/10.1016/B978-0-12-812752- 0.00016-2. p g j p [45] White SP, Allis RG, Moore J, Chidsey T, Morgan C, Gwynn W. Simulation of reactive transport of injected CO2 on the Colorado Plateau. Chem Geol 2005;217: 387–405. https://doi.org/10.1016/j.chemgeo.2004.12.020. [22] Noiriel C, Mad´e B, Gouze P. Impact of coating development on the hydraulic and transport properties in argillaceous limestone fracture. Water Resour Res 2007; 43. https://doi.org/10.1029/2006WR005379. [46] Samper J, Dai Z, Molinero J, Garcia-Gutierrez M, Missana T, Mingarro M. Interpretation of solute transport experiments in compacted Ca-bentonites using inverse modeling. Phys Chem Earth 2006;31:640–8. https://doi.org/10.1016/j. pce.2006.04.013. [23] Gherardi F, Xu T, Pruess K. Numerical modeling of self-limiting and self- enhancing caprock alteration induced by CO2 storage in a depleted gas reservoir. Chem Geol 2007;244:103–29. https://doi.org/10.1016/j.chemgeo.2007.06.009. [47] Johnson JW, Nitao JJ, Knauss KG. Reactive transport modeling of CO2 storage in saline aquifers to elucidate fundamental processes, trapping mechanisms, and sequestration partitioning. In: The first national conference on carbon sequestration Washington DC; 2001. [24] Liteanu E, Niemeijer A, Spiers CJ, Peach CJ, De Bresser JHP. The effect of CO2 on creep of wet calcite aggregates. J Geophys Res 2012;117:B03211. https://doi. org/10.1029/2011JB008789. [48] Johnson JW, Nitao JJ, Morris JP, Blair SC. Reactive transport modeling of cap rock integrity during natural and engineered CO2 storage. Lawrence Livermore National Laboratory Livermore CA; 2004 [ ]. g [25] Hangx S, Bakker E, Bertier P, Nover G, Busch A. Chemical–mechanical coupling observed for depleted oil reservoirs subjected to long-term CO2-exposure — a case study of the Werkendam natural CO2 analogueue field. Earth Planet Sci Lett 2015;428:230–42. https://doi.org/10.1016/j.epsl.2015.07.044. [49] Gunter WD, Wiwehar B, Perkins EH. Aquifer disposal of CO2-rich greenhouse gases: extension of the time scale of experiment for CO2-sequestering reactions by geochemical modeling. Mineral Petrol 1997;59(1–2):121. [26] Niu B, Krevor S. References The impact of mineral dissolution on drainage relative permeability and residual trapping in two carbonate rocks. Transport Porous Media 2020;131:363–80. https://doi.org/10.1007/s11242-019-01345-4. [50] Vafaie A, Habibnia B, Moallemi SA. Experimental investigation of the pore structure characteristics of the Garau gas shale formation in the Lurestan Basin, Iran. J Nat Gas Sci Eng 2015;27:432–42. https://doi.org/10.1016/j. jngse.2015.06.029. [27] Bemer E, Lombard JM. From injectivity to integrity studies of CO2 geological storage. Oil Gas Sci Technol 2010;65:445–59. https://doi.org/10.2516/ogst/ 2009028. [51] Vafaie A, Kivi IR. An investigation on the effect of thermal maturity and rock composition on the mechanical behavior of carbonaceous shale formations. Mar Petrol Geol 2020;116:104315. https://doi.org/10.1016/j. marpetgeo.2020.104315. [28] Kim S, Santamarina JC. CO2 geological storage: hydro-chemo-mechanical analyses and implications. Greenhouse Gases Sci Technol 2014;4:528–43. https://doi.org/10.1002/ghg.1421. i [29] Rathnaweera TD, Ranjith PG, Perera MSA, Haque A, Lashin A, Al Arifi N, Chandrasekharam D, Yang S, Xu T, Wang S, Yasar E. CO2-induced mechanical behavior of Hawkesbury sandstone in the Gosford basin: an experimental study. Mater Sci Eng, A 2015;641:123–37. https://doi.org/10.1016/j. msea.2015.05.029. [52] Vafaie A, Kivi IR, Moallemi SA, Habibnia B. Permeability prediction in tight gas reservoirs based on pore structure characteristics: a case study from South Western Iran. Unconv Resour 2021;1:9–17. https://doi.org/10.1016/j. uncres.2021.08.001. [53] Neuzil CE. Permeability of clays and shales. Annu Rev Earth Planet Sci 201;47: 247–273. https://doi.org/10.1146/annurev-earth-053018-060437. [30] Rohmer J, Pluymakers A, Renard F. Mechano-chemical interactions in sedimentary rocks in the context of CO2 storage: weak acid, weak effects? Earth Sci Rev 2016;157:86–110. https://doi.org/10.1016/j.earscirev.2016.03.009. [54] Ilgen AG, Aman M, Espinoza DN, Rodriguez MA, Griego JM, Dewers TA, Feldman JD, Stewart TA, Choens RC, Wislon J. Shale-brine-CO2 interactions and the long-term stability of carbonate-rich shale caprock. Int J Greenh Gas Control 2018;78:244–53. https://doi.org/10.1016/j.ijggc.2018.07.002. [31] Spycher N, Pruess K, Ennis-king J. CO2–H2O mixtures in the geological sequestration of CO2. I. Assessment and calculation of mutual solubilities from 12 to 100 ◦C and up to 600 bar. Geochimica Cosmochimica Acta 2003;67:3015–31. https://doi.org/10.1016/S0016-7037(03)00273-4. [55] Busch A, Alles S, Gensterblum Y, Prinz D, Dewhurst DN, Raven MD, Stanjek H, Krooss BM. Carbon dioxide storage potential of shales. Int J Greenh Gas Control 2008;2:297–308. https://doi.org/10.1016/j.ijggc.2008.03.003. [32] Greenwood HJ, Barnes HL. Binary mixtures of volatile components. In: Clark SP, editor. Handbook of physical constants, vol. 97. The Geological Society of America Memoir; 1966. [56] Makhnenko RY, Vilarrasa V, Mylnikov D, Laloui L. Hydromechanical aspects of CO2 breakthrough into clay-rich caprock. Energy Proc 2017;114:3219–28. References g [101] Sterpenich J, Jobard E, El Hajj H, Pironon J, Randi A, Caumon MC. Experimental study of CO2 injection in a simulated injection well: the MIRAGES experiment. Greenhouse Gas Sci Technol 2014;4:210–24. https://doi.org/10.1002/ghg.1389. [102] Vialle S, Contraires S, Zinzsner B, Clavaud JB, Mahiouz K, Zuddas P, Zamora M. Percolation of CO2-rich fluids in a limestone sample: evolution of hydraulic, electrical, chemical, and structural properties. J Geophys Res Solid Earth 2014; 119:2828–47. https://doi.org/10.1002/2013JB010656. l [78] Hadian P, Rezaee R. The effect of supercritical CO2 on shaly caprocks. Energies 2019;13:149. https://doi.org/10.3390/en13010149. [79] Al-Ameri WA, Abdulraheem A, Mahmoud M. Long-term effects of CO2 sequestration on rock mechanical properties. J Energy Resour Technol 2016;138. https://doi.org/10.1115/1.4032011. 012201-1. [103] Kaszuba JP, Janecky DR, Snow MG. Experimental evaluation of mixed fluid reactions between supercritical carbon dioxide and NaCl brine: relevance to the integrity of a geologic carbon repository. Chem Geol 2005;217:277–93. https:// doi.org/10.1016/j.chemgeo.2004.12.014. [80] Jayasekara DW, Ranjth PG. Effect of geochemical characteristics on caprock performance in deep saline aquifers. Energy Fuel 2021;35(3):2443–55. https:// doi.org/10.1021/acs.energyfuels.0c03726. g j g [104] Farquhar SM, Pearce JK, Dawson GKW, Golab A, Sommacal S, Kirste D, Biddle D, Golding SD. A fresh approach to investigating CO2 storage: experimental CO2–water-rock interactions in a low-salinity reservoir system. Chem Geol 2015; 399:98–122. https://doi.org/10.1016/j.chemgeo.2014.10.006. g gy [81] Hangx SJT, Spiers CJ, Peach CJ. The effect of deformation on permeability development in anhydrite and implications for caprock integrity during geological storage of CO2. Geofluids 2010;10:369–87. https://doi.org/10.1111/ j.1468-8123.2010.00299.x. j [82] Espinoza DN, Kim SH, Santamarina JC. CO2 geological storage−geotechnical implications. KSCE J Civ Eng 2011;15:707–19. https://doi.org/10.1007/s12205- 011-0011-9. p g j g [105] Noiriel C, Mad´e B, Gouze P. Impact of coating development on the hydraulic and transport properties in argillaceous limestone fracture. Water Resour Res 2007;43 (9). https://doi.org/10.1029/2006WR005379. [83] Nguyen MT, Bemer E, Dormieux L. Micromechanical modeling of carbonate geomechanical properties evolution during acid gas injection. ARMA 2011: 11–207 [ ]. [106] Busch A, Bertier P, Gensterblum Y, Rother G, Spiers CJ, Zhang M, Wentinck HM. On sorption and swelling of CO2 in clays. Geomech Geophys Geo-Energy Geo- Resour (GGGG) 2016;2:111–30. https://doi.org/10.1007/s40948-016-0024-4. [107] Lebedev M, Zhang Y, Sarmadivaleh M, Barifcani A, Al-Khdheeawi E, Iglauer S. Carbon geosequestration in limestone: pore-scale dissolution and geomechanical weakening. Int J Greenhouse Gas Control 2017;66:106–19. https://doi.org/ 10.1016/j.ijggc.2017.09.016. [84] Alam MM, Hjuler ML, Foged H, Lykke I, Christensen HF, Fabricius IL. References l j g [96] Turner LG, Dawson GK, Golding SD, Pearce JK. CO2 and NOx reactions with CO2 storage reservoir core: NOx dissolution products and mineral reactions. Int J Greenh Gas Control 2022;120:103750. https://doi.org/10.1016/j. ijggc.2022.103750. p g j jgg [71] Vialle S, Druhan JL, Maher K. Multi-phase flow simulation of CO2 leakage through a fractured caprock in response to mitigation strategies. Int J Greenh Gas Control 2016;44:11–25. https://doi.org/10.1016/j.ijggc.2015.10.007. l [72] Pini R, Krevor S. Laboratory studies to understand the controls on flow and transport for CO2 storage. Science of Carbon Storage in Deep Saline Formations 2019:109–37. https://doi.org/10.1016/B978-0-12-812752-0.00007-1. [97] Akono AT, Davila G, Druhan J, Shi Z, Jessen K, Tsotsis T, Fuchs S, Crandall D, Dalton L, Tkach MK, Goodman AL, Frailey S, Werth CJ. A review of geochemical–mechanical impacts in geological carbon storage reservoirs. Greenhouse Gas Sci Technol 2019;9:474–504. https://doi.org/10.1002/ ghg.1870. p g [73] Minardi A, Stavropoulou E, Kim T, Ferrari A, Laloui L. Experimental assessment of the hydro-mechanical behavior of a shale caprock during CO2 injection. Int J Greenh Gas Control 2021;106:103225. https://doi.org/10.1016/j. ijggc.2020.103225. [98] Bhuiyan MH, Agofack N, Gawel KM, Cerasi PR. Micro- and macroscale consequences of interactions between CO2 and shale rocks. Energies 2020;13: 1167. https://doi.org/10.3390/en13051167. jgg [74] Vialle S, Vanorio T. Laboratory measurements of elastic properties of carbonate rocks during injection of reactive CO2-saturated water. Geophys Res Lett 2011;38: 1–5. https://doi.org/10.1029/2010GL045606. [99] Garcia-Rios M, Cama J, Luquot L, Soler JM. Interaction between CO2-rich sulfate solutions and carbonate reservoir rocks from atmospheric to supercritical CO2 conditions: Experiments and modeling. Chem Geol 2015;414:95–108. https:// doi.org/10.1016/j.chemgeo.2014.06.004. p g [75] Vanorio T, Mavko G. Laboratory measurements of the acoustic and transport properties of carbonate rocks and their link with the amount of microcrystalline matrix. Geophysics 2011;76:E105–15. https://doi.org/10.1190/1.3580632. g j g [100] Davila G, Dalton L, Crandall DM, Garing C, Werth CJ, Druhan JL. Reactive alteration of an Mt. Simon Sandstone due to CO2-rich brine displacement. Geochem Cosmochim Acta 2020;271:227–47. https://doi.org/10.1016/j. gca.2019.12.015. p y p g [76] Miller K, Vanorio T, Yang S, Xiao X. A scale-consistent method for imaging porosity and micrite in dual-porosity carbonate rocks. Geophysics 2019;84: MR115–27. https://doi.org/10.1190/geo2017-0812.1. p g g [77] Zou Y, Li S, Ma X, Zhang S, Li N, Chen M. Effects of CO2–brine–rock interaction on porosity/permeability and mechanical properties during supercritical-CO2 fracturing in shale reservoirs. J Nat Gas Sci Eng 2018;49:157–68. https://doi.org/ 10.1016/j.jngse.2017.11.004. References h d i j [33] Kaszuba J, Yardley B, Andreani M. Experimental perspectives of mineral dissolution and precipitation due to carbon dioxide-water-rock interactions. Rev Mineral Geochem 2013;77(1):153–88. https://doi.org/10.2138/rmg.2013.77.5. l [57] Myers M, Roberts JJ, White C, Stalker L. The impact of water on CO2 leak rate measurements for CCS projects. In: 14th Greenhouse Gas Control Technologies Conference Melbourne Melbourne; 2018. p. 21–6. https://doi.org/10.2139/ ssrn.3366400. i [34] Andre L, Audigane M, Azaroual AM. Numerical modeling of fluid-rock chemical interactions at the supercritical CO2–liquid interface during CO2 injection into a carbonate reservoir, the Dogger aquifer (Paris Basin, France). Energy Convers Manag 2017;48:1782–97. https://doi.org/10.1016/j.enconman.2007.01.006. [58] Hildenbrand A, Schlomer S, Krooss BM. Gas breakthrough experiments on fine- grained sedimentary rocks. Geofluids 2002;2:3–23. https://doi.org/10.1046/ j.1468-8123.2002.00031.x. 19 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 [59] Espinoza DN, Santamarina JC. Water-CO2-mineral systems: interfacial tension, contact angle, and diffusion—implications to CO2 geological storage. Water Resour Res 2010;46. https://doi.org/10.1029/2009WR008634. [86] Grombacher D, Vanorio T, Ebert Y. Time-lapse acoustic, transport, and NMR measurements to characterize microstructural changes of carbonate rocks during injection of CO2-rich water. Geophysics 2015;77:WA169–79. https://doi.org/ 10.1190/geo2011-0281.1. p g [60] Luquot L, Gouze P, Niemi A, Bensabat J, Carrera J. CO2-rich brine percolation experiments through Heletz reservoir rock samples (Israel): role of the flow rate and brine composition. Int J Greenh Gas Control 2016;48:44–58. https://doi.org/ 10.1016/j.ijggc.2015.10.023. g [87] Nover G, Von Der Gonna J, Heikamp S, K¨oster J. Changes of petrophysical properties of sandstones due to interaction with supercritical carbon dioxide; a laboratory study. Eur J Mineral 2013:70176. https://doi.org/10.1127/0935- 1221/2013/0025-2295. [61] Aguilera R. Naturally fractured reservoirs. United States: N. 1980. [88] Gaus I, Azaroual M, Czernichowski-Lauriol I. Reactive transport modelling of the impact of CO2 injection on the clayey cap rock at Sleipner (North Sea). Chem Geol 2005;3–4:319–37. https://doi.org/10.1016/j.chemgeo.2004.12.016. [62] Tsang CF. Coupled hydromechanical-thermochemical processes in rock fractures. Rev Geophys 1991;4:537–51. https://doi.org/10.1029/91RG01832. [63] Zimmerman RW, Bodvarsson GS. Hydraulic conductivity of rock fractures. Transport Porous Media 1996;23:1–30. https://doi.org/10.1007/BF00145263. [89] Le Guen Y, Renard F, Hellmann R, Brosse E, Collombet M, Tisserand D, Gratier J. Enhanced deformation of limestone and sandstone in the presence of high P CO2. Fluid 2007;112:1–21. https://doi.org/10.1029/2006JB004637. [64] Hu M, Rutqvist J. Multi-scale coupled processes modeling of fractures as porous, interfacial and granular systems from rock images with the numerical manifold method. Rock Mech Rock Eng 2022;55:3041–59. https://doi.org/10.1007/ s00603-021-02455-6. [90] Vilarrasa V, Makhnenko RY, Rutqvist J. References Field and laboratory studies of geomechanical response to the injection of CO2. Science of Carbon Storage in Deep Saline Formations 2019:209–36. https://doi.org/10.1016/B978-0-12- 812752-0.00009-5. [65] Noiriel C, Bernard D, Gouze P, Thibault X. Hydraulic properties and microgeometry evolution accompanying limestone dissolution by acidic water. Oil Gas Sci Technol Rev IFP 2005;60:177–92. https://doi.org/10.2516/ogst: 2005011. [91] Schaef HT, Windisch Jr CF, McGrail BP, Martin PF, Rosso KM. Brucite [Mg (OH2)] carbonation in wet supercritical CO2: an in situ high pressure X-ray diffraction study. Geochem Cosmochim Acta 2011;75(23):7458–71. https://doi.org/ 10.1016/j.gca.2011.09.029. [66] Rimmel´e G, Barlet-Gou´edard V, Renard F. Evolution of the petrophysical and mineralogical properties of two reservoir rocks under thermodynamic conditions relevant for CO2 geological storage at 3 km depth. Oil Gas Sci Technol 2010;65: 565–80. https://doi.org/10.2516/ogst/2009071. j g [92] Pearce JK, Dawson GK, Law AC, Biddle D, Golding SD. Reactivity of micas and cap-rock in wet supercritical CO2 with SO2 and O2 at CO2 storage conditions. Appl Geochem 2016;72:59–76. https://doi.org/10.1016/j.apgeochem.2016.06.010. [67] Vanorio T, Nur A, Ebert Y. Rock physics analysis and time-lapse rock imaging of geochemical effects due to the injection of CO2 into reservoir rocks. Geophysics 2011;76. https://doi.org/10.1190/geo2010-0390.1. O23–O33. [93] Thaysen EM, Soler JM, Boone M, Cnudde V, Cama J. Effect of dissolved H2SO4 on the interaction between CO2-rich brine solutions and limestone, sandstone, and marl. Chem Geol 2017;450:31–43. https://doi.org/10.1016/j. chemgeo.2016.11.037. [68] Davila MG, Cama J, Luquot L, Soler JM, Ayora C. Reactivity of a marl caprock in contact with acid solutions under different PCO2 conditions (atmospheric, 10, and 37 bar). Procedia Earth and Planetary Science 2017;17:528–31. https://doi.org/ 10.1016/j.proeps.2016.12.133. l [94] Pearce JK, Dawson GK, Blach TP, Bahadur J, Melnichenko YB, Golding SD. Impure CO2 reaction of feldspar, clay, and organic matter-rich cap rocks: decreases in the fraction of accessible mesopores measured by SANS. Int J Coal Geol 2018;185:79–90. https://doi.org/10.1016/j.coal.2017.11.011. [69] Akono AT, Davila G, Druhan G, Shi Z, Jessen K, Tsotsis T. Influence of geochemical reactions on the creep behavior of Mt.Simon sandstone. Int J Greenh Gas Control 2020;103:103183. https://doi.org/10.1016/j.ijggc.2020.103183. [95] Dawson GK, Pearce JK, Biddle D, Golding SD. Experimental mineral dissolution in Berea Sandstone reacted with CO2 or SO2–CO2 in NaCl brine under CO2 sequestration conditions. Chem Geol 2015;2(399):87–97. https://doi.org/ 10.1016/j.chemgeo.2014.10.005. [70] Wollenweber J, Alles S, Busch A, Krooss BM, Stanjek H, Littke R. Experimental investigation of the CO2 sealing efficiency of caprocks. Int J Greenh Gas Control 2010;4:231–41. https://doi.org/10.1016/j.ijggc.2010.01.003. References l [142] Lyu Q, Ranjith PG, Long X, Ji B. Experimental investigation of mechanical properties of black shales after CO2-water-rock interaction. Materials 2016;9:663. https://doi.org/10.3390/ma9080663. [143] Soong Y, Howard BH, Dilmore RM, Haljasmaa I, Crandall DM, Zhang L, Zhang W, Lin R, Irdi GA, Romanov VN, Mclendon TR. CO2/brine/rock interactions in Lower Tuscaloosa formation. Greenhouse Gas Sci Technol 2016:1–14. https://doi.org/ 10.1002/ghg.1611. 00. i g j jgg [119] Spokas K, Peters CA, Pyrak-Nolte L. Influence of rock mineralogy on reactive fracture evolution in carbonate-rich caprocks. Environ Sci Technol 2018;52(17): 10144–52. https://doi.org/10.1021/acs.est.8b01021. p g [120] Elkhoury JE, Ameli P, Detwiler RL. Dissolution and deformation in fractured carbonates caused by flow of CO2-rich brine under reservoir conditions. Int J Greenh Gas Control 2013;16:S203–15. https://doi.org/10.1016/j. ijggc 2013 02 023 [144] Bertier P, Swennen R, Laenen B, Lagrou D, Dreesen R. Experimental identification of CO2–water-rock interactions caused by sequestration of CO2 in Westphalian and Buntsandstein sandstones of the Campine Basin (NE-Belgium). J Geochem Explor 2006;89:10–4. https://doi.org/10.1016/j.gexplo.2005.11.005. l g j g [145] Kivi IR, Makhnenko RY, Vilarrasa V. Two-phase flow mechanisms controlling CO2 intrusion into shaly caprock. Transport Porous Media 2022;141:771–98. https:// doi.org/10.1007/s11242-022-01748-w. [121] Foroutan M, Ghazanfari E, Amirlatifi A, Moradian O. Evolution of fracture permeability and aperture during CO2 storage in varyingly cemented sedimentary rocks. Geomech Energy Environ 2021;27:100289. https://doi.org/10.1016/j. gete.2021.100289. g [146] Ott H, Kloe KD, Bakel MV, Vos F, Pelt AV, Legerstee P, Bauer A, Eide K, Linden AVD, Berg S, Makurat A. Core-flood experiment for transport of reactive fluids in rocks. Rev Sci Instrum 2012;83:084501. https://doi.org/10.1063/ 1.4746997. l g [122] Deng H, Fitts JP, Crandall D, McIntyre D, Peters CA. Alterations of fractures in carbonate rocks by CO2-acidified brines. Environ Sci Technol 2015;49(16): 10226–34. https://doi.org/10.1021/acs.est.5b01980. [147] Grgic D. Influence of CO2 on the long-term chemomechanical behavior of an oolitic limestone. J Geophys Res Solid Earth 2011;116:1–22. https://doi.org/ 10.1029/2010JB008176. [123] Adam L, MacFarlane J, Wijk KV, Shragge J, Higgs K. Monitoring how carbonate cement dissolution affects rock frame properties due to CO2 injection. In: 3rd international Conference on rock physics perth Australia; 2015. [148] Dai J, Yang S, Chen H, Shen X. Geochemistry and occurrence of inorganic gas accumulations in Chinese sedimentary basins. Org Geochem 2005;36:1664–88. https://doi.org/10.1016/j.orggeochem.2005.08.007. i [124] Zhou H, Hu D, Zhang F, Shao J, Feng X. Laboratory investigations of the hydro- mechanical–chemical coupling behavior of sandstone in CO2 storage in aquifers. Rock Mech Rock Eng 2016;49:417–26. https://doi.org/10.1007/s00603-015- 0752-8. References Petrophysical and rock-mechanics effects of CO2 injection for Enhanced oil recovery: experimental study on chalk from south Arne field, north sea. J Pet Sci Eng 2014;122:468–87. https://doi.org/10.1016/j.petrol.2014.08.008. [85] El Husseiny A, Vanorio T. The effect of micrite content on the acoustic velocity of carbonate rocks. Geophysics 2015;80:l45. https://doi.org/10.1190/geo2014- 0599.1. –l55. [108] Ellis B, Peters C, Fitts J, Bromhal G, McIntyre D, Warzinski R, Rosenbaum E. Deterioration of a fractured carbonate caprock exposed to CO2-acidified brine flow. Greenhouse Gases Sci Technol 2011;1(3):248–60. https://doi.org/10.1002/ ghg.25. 20 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 [109] Voltolini M, Ajo-Franklin J. The effect of CO2-induced dissolution on flow properties in Indiana limestone: an in situ synchrotron X-ray micro-tomography study. Int J Greenh Gas Control 2019;82:38–47. https://doi.org/10.1016/j. ijggc.2018.12.013. [132] Rinehart AJ, Dewers TA, Broome ST, Eichhubl P. Effects of CO2 on mechanical variability and constitutive behavior of the Lower Tuscaloosa Formation, Cranfield injection site. USA. Int J Greenhouse Gas Control 2016;53:305–18. https://doi.org/10.1016/j.ijggc.2016.08.013. p g j jgg [133] Seyyedi M, Mahmud HKB, Verrall M, Giwelli A, Esteban L, Ghasemiziarani M, Clennell B. Pore structure changes occur during CO2 injection into carbonate reservoirs. Sci Rep 2020;10:3624. https://doi.org/10.1038/s41598-020-60247-4. jgg [110] Garcia-Rios M, Luquot L, Soler JM, Cama J. The role of mineral heterogeneity on the hydrogeochemical response of two fractured reservoir rocks in contact with dissolved CO2. Appl Geochem 2017;84:202–17. https://doi.org/10.1016/j. apgeochem 2017 06 008 [134] Foroutan M, Ghazanfari E. CO2-enriched brine injection’s impact on mechanical properties of a sandstone specimen. In: E3S web of conferences; 2020. https://doi. org/10.1051/e3sconf/202020502005. pg [111] Dewhurst DN, Raven MD, Shah SSBM, Ali SSBM, Giwelli M, Firns S, Josh M, White C. Interaction of super-critical CO2 with mudrocks: impact on composition and mechanical properties. Int J Greenh Gas Control 2020;102:103163. https:// doi.org/10.1016/j.ijggc.2020.103163. g [135] Bachu S. Screening and ranking of sedimentary basins for sequestration of CO2 in geological media in response to climate change. Environ Geol 2003;44:277–89. https://doi.org/10.1007/s00254-003-0762-9. [112] Samuelson J, Spiers CJ. Fault friction and slip stability not affected by CO2 storage: evidence from short-term laboratory experiments on North Sea reservoir sandstones and caprocks. Int J Greenh Gas Control 2012;11:S78–90. https://doi. org/10.1016/j.ijggc.2012.09.018. [136] Falcon-Suarez I, North L, Amalokwu K, Best A. Integrated geophysical and hydromechanical assessment for CO2 storage: shallow low permeable reservoir sandstones. Geophysical Prospecting 2016;64:828–47. https://doi.org/10.1111/ 1365-2478.12396. [113] Major JR, Eichhubl P, Dewers TA, Olson JE. References Effect of CO2–brine–rock interaction on fracture mechanical properties of CO2 reservoirs and seals. Earth Planet Sci Lett 2018;499:37–47. https://doi.org/10.1016/j.epsl.2018.07.013. [137] Yu Z, Yang S, Liu K, Zhou Q, Yang L. An experimental and numerical study of CO2–brine-synthetic sandstone interactions under high-pressure temperature reservoir conditions. Appl Sci 2019;9:3354. https://doi.org/10.3390/ app9163354. p g j p [114] Aman M, Espinoza DN, Ilgen AG, Major JR, Eichhubl P, Dewers TA. CO2-induced chemo-mechanical alteration in reservoir rocks assessed via batch reaction experiments and scratch testing. Greenhouse Gases Sci Technol 2018;8(1): 133–49. https://doi.org/10.1002/ghg.1726. [138] Fuchs SJ, Espinoza DN, Lopano CL, Akono AT, Werth CJ. Geochemical and geomechanical alteration of siliciclastic reservoir rock by supercritical CO2- saturated brine formed during geological carbon sequestration. Int J Greenh Gas Control 2019;88:251–60. https://doi.org/10.1016/j.ijggc.2019.06.014. p g g g [115] Espinoza DN, Jung H, Major JR, Sun Z, Ramos MR, Eichhubl P, Balhoff MT, Choens RC, Dewers TA. CO2 charged brines changed rock strength and stiffness at Crystal Geyser, Utah: implications for leaking subsurface CO2 storage reservoirs. Int J Greenh Gas Control 2018;73:16–28. https://doi.org/10.1016/j. ijggc 2018 03 017 [139] Shi Z, Sun L, Haljasmaa I, Harbert W, Sanguinito S, Tkach M, Goodman A, Tsotsis TT, Jessen K. Impact of brine/CO2 exposure on the transport and mechanical properties of the Mt Simon sandstone. J Pet Sci Eng 2019;177: 295–305. https://doi.org/10.1016/j.petrol.2019.01.112. jgg [116] Skurtveit E, Sundal A, Bjørnarå TI, Soldal M, Sauvin G, Zuchuat V, Midtkandal I, Braathen A. Experimental investigation of natural fracture stiffness and flow properties in a faulted CO2 bypass system (Utah, USA). J Geophys Res Solid Earth 2020;125(7):e2019JB018917. https://doi.org/10.1029/2019JB018917. [140] Rezaee R, Saeedi A, Iglauer S, Evans B. Shale alteration after exposure to supercritical CO2. Int J Greenh Gas Control 2017;62:91–9. https://doi.org/ 10.1016/j.ijggc.2017.04.004. j jgg [141] Li S, Zhang S, Zou Y, Ma X, Ding Y, Li N, Zhang X, Kasperczyk D. Pore structure alteration induced by CO2–brine–rock interaction during CO2 energetic fracturing in tight oil reservoirs. J Pet Sci Eng 2020;191:107147. https://doi.org/10.1016/j. petrol.2020.107147. [117] Tarokh A, Makhnenko RY, Kim K, Zhu X, Popovics JS, Segvic B, Sweet DE. Influence of CO2 injection on the poromechanical response of Berea sandstone. Int J Greenhouse Gas Control 2020;95:102959. https://doi.org/10.1016/j. ijggc.2020.102959. jgg [118] Michael K, Golab A, Shulakova V, Ennis-King J, Allinson G, Sharma S, Aiken T. Geological storage of CO2 in saline aquifers—a review of the experience from existing storage operations. Int J Greenh Gas Control 2010;4(4):659–67. https:// doi.org/10.1016/j.ijggc.2009.12.011. References [149] Miocic J, Gilfillan SMV, McDermott C, Haszeldine RS. Controls on CO2 storage security in natural reservoirs and implications for CO2 storage site selection. Int J Greenhouse Gas Control 2016;51:118–25. https://doi.org/10.1016/j. ijggc.2016.05.019. [125] Zekri AY, Shedid A, Almehaideb RA. An experimental investigation of interactions between supercritical CO2, aspheltenic crude oil, and reservoir brine in carbonate cores. In: International Symposium on Oilfield Chemistry Houston Texas USA OnePetro SPE104750; 2007. https://doi.org/10.2118/104750-MS. [150] Gislason SR, Wolff-Boenisch D, Stefansson A, Oelkers EH, Gunnlaugsson E, Sigurdardottir H, Sigfusson B, Broecker WS, Matter JM, Stute M, Axelsson G, Fridriksson T. Mineral sequestration of carbon dioxide in basalt: a pre-injection overview of the CarbFix project. Int J Greenh Gas Control 2010;4:537–45. https://doi.org/10.1016/j.ijggc.2009.11.013. [126] Kim K, Vilarrasa V, Makhnenko RY. CO2 injection effect on geomechanical and flow properties of calcite-rich reservoirs. Fluid 2018;3:66. https://doi.org/ 10.3390/fluids3030066. l [127] Marbler H, Erickson KP, Schmidt M, Lempp C, P¨ollmann H. Geomechanical and geochemical effects on sandstones caused by the reaction with supercritical CO2: an experimental approach to in situ conditions in deep geological reservoirs. Environ Earth Sci 2013;69:1981. https://doi.org/10.1007/s12665-012-2033-0. –1998. p g j jgg [151] Pokrovsky OS, Schott J. Surface Chemistry and dissolution kinetics of divalent metal carbonates. Environ Sci Technol 2002;36(3):426–32. https://doi.org/ 10.1021/es010925u. [152] Duan Z, Sun R. An improved model calculating CO2 solubility in pure water and aqueous NaCl solutions from 273 to 533 K and from 0 to 2000 bar. Chem Geol 2003;193(3–4):257–71. https://doi.org/10.1016/S0009-2541(02)00263-2. [128] Mikhaltsevitch V, Lebedev M, Gurevich B. Measurements of the elastic and anelastic properties of sandstone flooded with supercritical CO2. Geophys Prospect 2014;62:1266–77. https://doi.org/10.1111/1365-2478.12181. [153] Tariq Z, Abdulraheem A, Elkatatny S, Mahmud M, Muqtadir A, Murtaza M. Geomechanical studies on CO2 sequestrated rocks in an aqueous saline environment. In: SPE-122242-MS; 2018 [ ]. [129] Raza A, Gholami R, Sarmadivaleh M, Tarom N, Rezaee R, Bin CH, Nagarajan R, Hamid MA, Elochukwu H. Integrity analysis of CO2 storage sites concerning geochemical-geomechanical interactions in saline aquifers. J Nat Gas Sci Eng 2016;36PA:224–40. https://doi.org/10.1016/j.jngse.2016.10.016. l [154] Wang H, Alvarado V, Bagdonas DA, McLaughlin JF, Kaszuba JP, Grana D, Campbell E, Ng K. Effect of CO2-brine-rock reactions on pore architecture and permeability in dolostone: implications for CO2 storage and EOR. Int J Greenh Gas Control 2021;107:103283. https://doi.org/10.1016/j.ijggc.2021.103283. [130] Huang YH, Yang SQ, Li WP, Hall MR. Influence of supercritical CO2 on the strength and fracture behavior of brine-saturated sandstone specimens. Rock Mech Rock Eng 2019;53:653–70. https://doi.org/10.1007/s00603-019-01933-2. References Environ Earth Sci 2013;70:3687–708. https:// doi.org/10.1007/s12665-013-2669-4. g j [193] Newell P, Ilgen AG. Overview of geological carbon storage (GCS). In: Science of carbon storage in deep saline formations. Elsevier; 2019. p. 1–13. https://doi.org/ 10.1016/B978-0-12-812752-0.00001-0. g [167] Harbert W, Goodman A, Spaulding R, Haljasmaa I, Crandall D, Sanguinito S, Kutchko B, Tkach M, Fuchs S, Werth CJ, Tsotsis T, Dalton L, Jessen K, Shi Z, Frailey S. CO2 induced changes in Mount Simon sandstone: understanding links to post CO2 injection monitoring, seismicity, and reservoir integrity. Int J Greenhouse Gas Control 2020;100:103109. https://doi.org/10.1016/j. ijggc.2020.103109. [194] Ilgen AG, Newell P, Hueckel T, Espinoza DN, Hu M. Coupled chemical-mechanical processes associated with the injection of CO2 into subsurface. In: Science of carbon storage in deep saline formations. Elsevier; 2019. p. 337–59. https://doi. org/10.1016/B978-0-12-812752-0.00015-0. [195] Raza A, Gholami R, Sarmadivaleh M. Feasibility of limestone reservoirs as a carbon dioxide storage site: an experimental study. AAPG (Am Assoc Pet Geol) Bull 2020;104:83–96. https://doi.org/10.1306/04241918124. [168] Baumann G, Henninges J, De Lucia M. Monitoring of saturation changes and salt precipitation during CO2 injection using pulsed neutron-gamma logging at the Ketzin pilot site. Int J Greenh Gas Control 2014;28:134–46. https://doi.org/ 10.1016/j.ijggc.2014.06.023. [196] Perera MSA, Rathnaweera TD, Ranjith PG, Wanniarachchi WAM, Nasvi MCA, Abdulagatov LM, Haque A. Laboratory measurement of deformation-induced hydro-mechanical properties of reservoir rock in deep saline aquifers: an experimental study of Hawkesbury formation. Mar Petrol Geol 2016;77:640–52. https://doi.org/10.1016/j.marpetgeo.2016.07.012. [169] Miri R, Hellevang H. Salt precipitation during CO2 storage-A review. Int J Greenh Gas Control 2016;51:136–47. https://doi.org/10.1016/j.ijggc.2016.05.015. [170] Cui G, Zhang L, Ren B, Enechukwu E, Liu Y, Ren S. Geothermal exploitation from depleted high temperature gas reservoirs via recycling supercritical CO2: heat mining rate and salt precipitation effects. Appl Energy 2016;183:837–52. https:// doi.org/10.1016/j.apenergy.2016.09.029. [197] Hangx S, van der Linden A, Marcelis F, Bauer A. The effect of CO2 on the mechanical properties of the Captain sandstone: geological storage of CO2 at the Goldeneye field (UK). Int J Greenh Gas Control 2013;19:609–19. https://doi.org/ 10.1016/j.ijggc.2012.12.016. [171] Cui G, Ren S, Dou B, Ning F. Geothermal energy exploitation from depleted high- temperature gas reservoirs by recycling CO2: the superiority and existing problems. Geosci Front 2021;12(6):101078. https://doi.org/10.1016/j. gsf.2020.08.014. [198] Ahmadinejad A, Kivi IR. An experimental investigation on the poroelastic response of a water-saturated limestone to hydrostatic compression. Bull Eng Geol Environ 2021;80:3817–32. https://doi.org/10.1007/s10064-021-02157-x. [172] Norouzi AM, Babaei M, Han WS, Kim Ky, Niasar V. References [155] Clark AC, Vanorio T. The rock physics and geochemistry of carbonates exposed to reactive brines. J Geophys Res Solid Earth 2016;121:1497–513. https://doi.org/ 10.1002/2015JB012445. [131] Choi CS, Kim J, Song JJ. Analysis of shale property changes after geochemical interaction under CO2 sequestration conditions. Energy 2021;214:118933. https://doi.org/10.1016/j.energy.2020.118933. 21 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 [156] Kim K, Makhnenko RY. Changes in rock matrix compressibility during deep CO2 storage. Greenhouse Gases Sci Technol 2021;11(5):954–73. https://doi.org/ 10.1002/ghg.2106. [179] Menke H. Reservoir condition pore-Scale imaging of reaction. Ph.D. thesis Department of Earth Science and Engineering Imperial College London; 2016. i [180] Golfier f, Zarcone C, Bazin B, Lenormand R, Lasseux D, Quintard M. On the ability of a Darcy-scale model to capture wormhole formation during the dissolution of a porous medium. J Fluid Mech 2002;457:213–54. https://doi:10.1017/S0022112 002007735. g g [157] Luquot L, Gouze P. Experimental determination of porosity and permeability changes induced by injection of CO2 into carbonate rocks. Chem Geol 2009;265 (1–2):148–59. https://doi.org/10.1016/j.chemgeo.2009.03.028. [181] Cohen CE, Ding D, Quintard M, Bazin B. From pore scale to wellbore scale: impact of geometry on wormhole growth in carbonate acidization. Chem Eng Sci 2008; 63:3088–99. https://doi.org/10.1016/j.ces.2008.03.021. l [158] Canal J, Delgado J, Falcon I, Yang Q, Juncosa R, Barrientos V. Injection of CO2- saturated water through a siliceous sandstone plug from the Hontomin test site (Spain): experiment and modeling. Environ Sci Technol 2013;47:159–67. https:// doi.org/10.1021/es3012222. i [182] Fredd CN, Fogler HS. Influence of transport and reaction on wormhole formation in porous media. Fluid Mech Transp Phenomena 1998;4:1933–49. https://doi. org/10.1002/aic.690440902. l [159] Foroutan M, Ghazanfari E, Amirlatifi A. Variation of failure properties creep response and ultrasonic velocities of sandstone upon injecting CO2-enriched brine. Geomech Geophys Geo-energ Geo-resour 2021;7:27. https://doi.org/ 10.1016/j.gete.2020.100217. i [183] Carman PC. Fluid flow through granular beds. Trans Inst Chem Eng 1937;15: 150–66. [160] Foroutan M, Ghazanfari E, Amirlatifi A, Perdrial N. Variation of pore-network, mechanical and hydrological characteristics of sandstone specimens through CO2- enriched brine injection. Geomech Energ Environ 2021;26:100217. https://doi. org/10.1007/s40948-021-00223-y. i [184] Hao Y, Smith M, Sholokhova Y, Carroll S. CO2-induced dissolution of low permeability carbonates. Part II: numerical modeling of experiments. Adv Water Resour 2013;62:388–408. https://doi.org/10.1016/j.advwatres.2013.09.009. [185] Juanes R, Spiteri EJ, Orr FMJ, Blunt MJ. Impact of relative permeability hysteresis on geological CO2 storage. Water Resour Res 2006;42. https://doi.org/ 10.1029/2005WR004806. References g y [161] Pearce JK, Dawson GK, Golab A, Knuefing L, Sommacal S, Rudolph V, Golding SD. A combined geochemical and μCT study on the CO2 reactivity of Surat Basin reservoir and cap-rock cores: porosity changes, mineral dissolution and fines migration. Int J Greenh Gas Control 2019;80:10–24. https://doi.org/10.1016/j. ijggc.2018.11.010. [186] Benson S, Pini R, Reynolds C, Krevor S. Relative permeability analyses to describe multi-phase flow in CO2 storage reservoirs. Glob Global CCS Institute Targeted Report. jgg [162] Lamy-Chappuis B, Angus D, Fisher Q, Grattoni C, Yardley BW. Rapid porosity and permeability changes of calcareous sandstone due to CO2-enriched brine injection. Geophys Res Lett 2014;41(2):399–406. https://doi.org/10.1002/ 2013GL058534. p [187] Egermann P, Bemer E, Zinszner B. An experimental investigation of the rock properties evolution associated to different levels of CO2 injection like alteration processes. Int Symp Society Core Analysts 2006:1–15. [188] Bemer E, Nguyen MT, Dautriat J, Adelinet M, Fleury M, Youssef S. Impact of chemical alteration on the poromechanical properties of carbonate rocks. Geophys Prospect 2016;64(4):810–27. https://doi.org/10.1111/1365- 2478.12387. [163] Lamy-Chappuis B, Angus D, Fisher QJ, Yardley BWD. The effect of CO2-enriched brine injection on the mechanical properties of calcite-bearing sandstone. Int Greenhouse Gas Control 2016;52:84–95. https://doi.org/10.1016/j. ijggc.2016.06.018. [189] Radilla1 G, Kacem M, Lombard JM, Fourar M. Transport properties of Lavoux limestone at various stages of CO2-like acid-rock alteration. Oil Gas Sci Technol IFP 2010;65(4):557–63. https://doi.org/10.2516/ogst/2009081. jgg [164] White P, Allis RG, Moore J, Chidsey T, Morgan C, Gwynn W, Adams M. Simulation of reactive transport of injected CO2 on the Colorado Plateau, Utah, USA. Chem Geol 2005;217(3–4):387–405. https://doi.org/10.1016/j. chemgeo.2004.12.020. l [190] Jiang F, Tsuji T. Changes in pore geometry and relative permeability caused by carbonate precipitation in porous media. Phys Rev E 2014;90(5):053306. htt ps://link.aps.org/doi/10.1103/PhysRevE.90.053306. [165] Brosse E, Magnier C, Vincent B. Modelling fluid-rock interaction induced by the percolation of CO2-enriched solutions in core samples: the role of reactive surface area. Oil Gas Sci Technol IFP 2005;60(2):287–305. https://doi.org/10.2516/ogst: 2005018. p p g y [191] Eberli GP, Baechle GT, Anselmetti FS, Incze ML. Factors controlling elastic properties in carbonate sediments and rocks. Lead Edge 2003;22(7):654–60. https://doi.org/10.1190/1.1599691. p g [192] Berryman JG, Pride SR, Wang HF. A differential scheme for elastic properties of rocks with dry or saturated cracks. Geophys J Int 2002;151(2):597–611. https:// doi.org/10.1046/j.1365-246X.2002.01801.x. [166] Fischer S, Liebscher A, De Lucia M, Hecht L. Reactivity of sandstone and siltstone samples from the Ketzin pilot CO2 storage site-Laboratory experiments and reactive geochemical modeling. References Experimental and numerical study of membrane properties and pore pressure transmission of Ghom shale. Measurement 2016;91: 93–100. https://doi.org/10.1016/j.measurement.2016.05.036. p p g [214] Atkinson BK. Subcritical crack propagation in rocks: theory, experimental results and applications. J Struct Geol 1982;4(1):41–56. https://doi.org/10.1016/0191- 8141(82)90005-0. [241] Ewy RT. Shale/claystone response to air and liquid exposure, and implications for handling, sampling and testing. Int J Rock Mech Min Sci 2015;80:388–401. https://doi.org/10.1016/j.ijrmms.2015.10.009. [215] Heap MJ, Baud P, Meredith PG, Bell AF, Main IG. Time-dependent brittle creep in Darley Dale sandstone. J Geophys Res Solid Earth 2009;114(B7). https://doi.org/ 10.1029/2008JB006212. p g j j [242] Feng G, Kang Y, Sun ZD, Wang XC, Hu YQ. Effects of supercritical CO2 adsorption on the mechanical characteristics and failure mechanisms of shale. Energy 2019; 15(173):870–82. https://doi.org/10.1016/j.energy.2019.02.069. [216] Zhu W, Baud P, Wong TF. Micromechanics of cataclastic pore collapse in limestone. J Geophys Res: Solid Earth 2010;115:B4. https://doi.org/10.1029/ 2009JB006610. [243] Minaeian V, Dewhurst DN, Rasouli V. Deformational behavior of a clay-rich shale with variable water saturation under true triaxial stress conditions. Geomech. Energy Environ. 2017;11:1–13. https://doi.org/10.1016/j.gete.2017.04.001. [217] Zhang X, Spiers CJ. Compaction of granular calcite by pressure solution at room temperature and effects of pore fluid chemistry. Int J Rock Mech Min Sci 2005;42 (7–8):950–60. https://doi.org/10.1016/j.ijrmms.2005.05.017. [244] Espinoza DN, Santamarina JC. Clay interaction with liquid and supercritical CO2: the relevance of electrical and capillary forces. Int J Greenhouse Gas Control 2012;10:351–62. https://doi.org/10.1016/j.ijggc.2012.06.020. i p g j j [218] Pietruszczak S, Lydzba D, Shao JF. Modelling of deformation response and chemo-mechanical coupling in chalk. Int J Numer Anal Methods GeoMech 2006; 30(10):997–1018. https://doi.org/10.1002/nag.510. ll [245] Shin H, Santamarina JC. Desiccation cracks in saturated fine-grained soils: particle-level phenomena and effective-stress analysis. Geotechnique 2011;61 (11):961–72. https://doi.org/10.1680/geot.8.P.012. p g g [219] Liteanu E, Spiers CJ. Influence of pore fluid salt content on compaction creep of calcite aggregates in the presence of supercritical CO2. Chem Geol 2009;265(1–2): 134–47. https://doi.org/10.1016/j.chemgeo.2008.12.010. [246] Finley RJ. An overview of the Illinois Basin–Decatur project. Greenhouse Gases Sci Technol 2014;4(5):571–9. https://doi.org/10.1002/ghg.1433. [220] Zhang X, Spiers CJ, Peach CJ. Compaction creep of wet granular calcite by pressure solution at 28◦C to 150◦C. J Geophys Res Solid Earth 2010;115(B9). https://doi.org/10.1029/2008JB005853. [247] Sarout J, Detournay E. Chemoporoelastic analysis and experimental validation of the pore pressure transmission test for reactive shales. Int J Rock Mech Min Sci 2011;48(5):759–72. https://doi.org/10.1016/j.ijrmms.2011.04.009. p g [221] Renard F, Gundersen E, Hellmann R, Collombet M, Le Guen Y. References Numerical modeling of carbon dioxide sequestration on the rate of pressure solution creep in limestone: Preliminary results. Oil Gas Sci Technol 2005;60:381–99. https://doi. org/10.2516/ogst:2005023. [248] Sinha S, Braun EM, Determan MD, Passey QR, Leonardi SA, Boros JA, Wood AC, Zirkle T, Kudva RA. Steady-state permeability measurements on intact shale samples at reservoir conditions – effect of stress, temperature, pressure, and type of gas. In: SPE Middle East oil and gas show and conference. Bahrain: Manama; 2013. https://doi.org/10.2118/164263-MS. [222] Xiao Y, Lasaga AC. Ab initio mechanical studies of the kinetics and mechanisms of silicate dissolution: H+ (H3O+) catalysis. Geochem Cosmochim Acta 1994;58: 5379–400. https://doi.org/10.1016/0016-7037(94)90237-2. [249] Tang AM, Cui YJ, Barnel N. Thermo-mechanical behaviour of a compacted swelling clay. Geotechnique 2008;58(1):45–54. https://doi.org/10.1680/ geot.2008.58.1.45. [223] Atkinson BK, Meredith PG. Stress corrosion cracking of quartz: a note on the influence of chemical environment. Tectonophysics 1981;77(1–2):T1–11. https:// doi.org/10.1016/0040-1951(81)90157-8. g [250] Neuzil CE. Permeability of clays and shales. Annu Rev Earth Planet Sci 2019;47 (1):247–73. https://doi.org/10.1146/annurev-earth-053018-060437. g [224] Dove PM. Geochemical controls on the kinetics of quartz fracture at subcritical tensile stresses. J Geophys Res Solid Earth 1995;100(B11):22349–59. https://doi. org/10.1029/95JB02155. [251] Houben ME, van Eeden JC, Barnhoorn A, Hangx SJ. Fracture-induced permeability in Whitby mudstone. Environ Sci Technol 2020;54(15):9564–72. https://doi.org/10.1021/acs.est.0c00557. g [225] Li HC, De Bruyn PL. Electrokinetic and adsorption studies on quartz. Surf Sci 1966;5:203–20. https://doi.org/10.1016/0039-6028(66)90082-3. g [252] Phillips T, Kampman N, Bisdom K, Inskip ND, den Hartog SA, Cnudde V, Busch A. Controls on the intrinsic flow properties of mudrock fractures: a review of their importance in subsurface storage. Earth Sci Rev 2020;211:103390. https://doi. org/10.1016/j.earscirev.2020.103390. [226] Oikawa Y, Takehara T, Tosha T. Effect of CO2 injection on mechanical properties of Berea Sandstone (ARMA 06-068). In: 42nd US rock MechanicsSymposium/2nd US-Canada rock mechanics symposium. San Francisco CA; 2008. l [253] Detwiler RL, Morris JP. Fracture initiation, propagation, and permeability evolution. Geological Carbon Storage: Subsurface Seals and Caprock Integrity 2018:119–35. https://doi.org/10.1002/9781119118657.ch5. [227] Schimmel MT, Hangx SJ, Spiers CJ. Effect of pore fluid chemistry on uniaxial compaction creep of Bentheim sandstone and implications for reservoir injection operations. Geomech Energy Environ 2022;29:100272. https://doi.org/10.1016/ j.gete.2021.100272. [254] Hajirezaie S, Peters CA, Cole DR, Sheets JM, Kim JJ, Swift AM, Crandall D, Cheshire MC, Stack AG, Anovitz LM. Sealing fractures to increase underground storage security: lessons learned from a multiscale multimodal imaging study of a syntaxial vein in a mudrock. Chem Geol 2022;614:121164. References Renewable and Sustainable Energy Reviews 178 (2023) 113270 mineralogical and mechanical changes in varyingly-cemented sandstones upon CO2 injection in deep saline aquifer environments. Energy 2017;133:404–14. https://doi.org/10.1016/j.energy.2017.05.154. l [233] Kim K, Makhnenko RY. Coupling between poromechanical behavior and fluid flow in tight rock. Transport Porous Media 2020;135:487–512. https://doi.org/ 10.1007/s11242-020-01484-z. [234] Horsrud P. Estimating mechanical properties of shale from empirical correlations. SPE Drill Complet 2001;16:68–73. https://doi.org/10.2118/56017-PA. p g j gy [208] Zhu W, Wong TF. The transition from brittle faulting to cataclastic flow: Permeability evolution. J Geophys Res Solid Earth 1997;102:3027–41. https:// doi.org/10.1029/96JB03281. [235] Du J, Hu L, Meegoda JN, Zhang G. Shale softening: observations, phenomenological behavior, and mechanisms. Appl Clay Sci 2018;161:290–300. https://doi.org/10.1016/j.clay.2018.04.033. [209] Dewers T, Newell P, Broome S, Heath J, Bauer S. Geomechanical behavior of cambrian Mount simon sandstone reservoir lithofacies, Iowa shelf, USA. Int J Greenhouse Gas Control 2014;21:33–48. https://doi.org/10.1016/j. ijggc.2013.11.010. p g j y [236] Giesting P, Guggenheim S, Koster van Groos AF, Busch A. X-Ray diffraction study of K- and Ca-exchanged montmorillonites in CO2 atmospheres. Environ Sci Technol 2012;46(10):5623–30. https://doi.org/10.1021/es3005865. jgg [210] Kivi IR, Ameri MJ, Molladavoodi H. Shale brittleness evaluation based on energy balance analysis of stress-strain curves. J Pet Sci Eng 2018;167:1–19. https://doi. org/10.1016/j.petrol.2018.03.061. [237] Giesting P, Guggenheim S, Koster van Groos AF, Busch A. Interaction of carbon dioxide with Na-exchanged montmorillonite at pressures to 640bars: implications for CO2 sequestration. Int J Greenh Gas Control 2012;8:73–81. https://doi.org/ 10.1016/j.ijggc.2012.01.011. g j p [211] Laubach SE, Lander RH, Criscenti LJ, Anovitz LM, Urai JL, Pollyea RM, Hooker JN, Narr W, Evans MA, Kerisit SN, Olson JE. The role of chemistry in fracture pattern development and opportunities to advance interpretations of geological materials. Rev Geophys 2019;57(3):1065–111. https://doi.org/ 10.1029/2019RG000671. [238] Heller R, Zoback M. Adsorption of methane and carbon dioxide on gas shale and pure mineral samples. J Unconv Oil Gas Resour 2014;8:14–24. https://doi.org/ 10.1016/j.juogr.2014.06.001. j j g [239] Oort EV, Hale AH, Mody FK, Roy S. Transport in shales and the design of improved water-based shale drilling fluids. SPE Drill Complet 1996;11:137–46. https://doi.org/10.2118/28309-PA. 03. [212] Rutter EH. The kinetics of rock deformation by pressure solution. Philos Trans R Soc London, Ser A 1976;283:203–19. https://www.jstor.org/stable/74639. il [213] Spiers CJ, Schutjens P. Densification of crystalline aggregates by fluid-phase diffusional creep. In: Deformation processes in minerals. Ceramics and Rocks Meredith; 1990. p. 334–53. https://doi.org/10.1007/978-94-011-6827-413. g [240] Kivi IR, Ameri MJ, Ghassemi A. References CO2-plume geothermal processes: a parametric study of salt precipitation influenced by capillary-driven backflow. Chem Eng J 2021;425:1385–8947. https://doi.org/10.1016/j. cej.2021.130031. [199] Cheng AHD. Poroelasticity. Theory and applications of transport in porous media, vol. 27; 2016. https://doi.org/10.1007/978-3-319-25202-5. l [200] Liteanu E, Spiers A, De Bresser JHP. The influence of water and supercritical CO2 on the failure behavior of chalk. Tectonophysics 2013;599:157–69. https://doi. org/10.1016/j.tecto.2013.04.013. j [173] Spycher N, Pruess K. A phase-partitioning model for CO2–brine mixtures at elevated temperatures and pressures: application to CO2-enhanced geothermal systems. Transport Porous Media 2010;82(1):173–96. https://doi.org/10.1007/ s11242-009-9425-y. g j [201] Gassmann F. Elastic waves through a packing of spheres. Geophysics 1951;16: 673–85. https://doi.org/10.1190/1.1437718. l y [174] Ott H, Andrew M, Snippe J, Blunt MJ. Microscale solute transport and precipitation in complex rock during drying. Geophys Res Lett 2014;41(23): 8369–76. https://doi.org/10.1002/2014GL062266. [202] Smith TM, Sondergeld CH, Rai CH. Gassmann fluid substitutions: a tutorial. Geophysics 2003;68(2):430–40. https://doi.org/10.1190/1.1567211. l [203] Han D, Batzle ML. Gassmann’s equation and fluid-saturation effects on seismic velocities. Geophysics 2004;69(2):398–405. https://doi.org/10.1190/1.1707059. [175] Smith MM, Hao Y, Carroll SA. Development and calibration of a reactive transport model for carbonate reservoir porosity and permeability changes based on CO2 core-flood experiments. Int J Greenhouse Gas Control 2017;57:73–88. https:// doi org/10 1016/j ijggc 2016 12 004 [204] Rinaldi AP, Rutqvist J. Modeling of deep fracture zone opening and transient ground surface uplift at KB-502 CO2 injection well, in Salah, Algeria. Int J Greenh Gas Control 2013;12:155–67. https://doi.org/10.1016/j.ijggc.2012.10.017. g j jgg [176] Luquot L, Gouze F. Experimental determination of porosity and permeability changes induced by injection of CO2 into carbonate rocks. Chem Geol 2009;265: 148–59. https://doi.org/10.1016/j.chemgeo.2009.03.028. [205] Zhu W, Baud P, Wong TF. Micromechanics of cataclastic pore collapse in limestone. J Geophys Res Solid Earth 2010;115:B04405. https://doi.org/ 10.1029/2009JB006610. l [177] Szymczak P, Ladd AJC. Wormhole formation in dissolving fractures. J Geophys Res Solid Earth 2009;114:B06203. https://doi.org/10.1029/2008JB006122. [206] Xie SY, Shao JF, Xu WY. Influences of chemical degradation on mechanical behavior of a limestone. Int J Rock Mech Min Sci 2011;48:741–7. https://doi.org/ 10.1016/j.ijrmms.2011.04.015. [178] Smith MM, Sholokhova Y, Hao Y, Carroll SA. CO2-induced dissolution of low permeability carbonates. Part I: characterization and experiments. Adv Water Resour 2013;62:370–87. https://doi.org/10.1016/j.advwatres.2013.09.008. j j [207] Rathnaweera TD, Ranjith PG, Perera MSA, Ranathunga AS, Wanniarachchi WAM, Yang SQ, Lashin A, Al Arifi N. An experimental investigation of coupled chemico- 22 A. Vafaie et al. References [267] Andreani M, Luquot L, Gouze P, Godard M, Hoise E, Gibert B. Experimental study of carbon sequestration reactions controlled by the percolation of CO2-rich brine through peridotites. Environ Sci Technol 2009;43(4):1226–31. https://doi.org/ 10.1021/es8018429. [290] Ishibashi T, Fang Y, Elsworth D, Watanabe N, Asanuma H. Hydromechanical properties of 3D printed fractures with controlled surface roughness: insights into shear-permeability coupling processes. Int J Rock Mech Min Sci 2020;128: 104271. https://doi.org/10.1016/j.ijrmms.2020.104271. l [268] Detwiler RL. Experimental observations of deformation caused by mineral dissolution in variable-aperture fractures. J Geophys Res Solid Earth 2008;113 (B8). https://doi.org/10.1029/2008JB005697. [291] Fazeli H, Nooraiepour M, Hellevang H. Microfluidic study of fracture dissolution in carbonate-rich caprocks subjected to CO2-charged brine. Ind Eng Chem Res 2019;59(1):450–7. https://doi.org/10.1021/acs.iecr.9b06048. p g [269] Yasuhara H, Elsworth D, Polak A. Evolution of permeability in a natural fracture: significant role of pressure solution. J Geophys Res Solid Earth 2004;109(B3). https://doi.org/10.1029/2003JB002663. g [292] Nolte DD, Pyrak-Nolte LJ. Monitoring fracture saturation with internal seismic sources and twin neural networks. J Geophys Res Solid Earth 2022;127(2): e2021JB023005. https://doi.org/10.1029/2021JB023005. p g [270] Pyrak-Nolte LJ, Morris JP. Single fractures under normal stress: the relation between fracture specific stiffness and fluid flow. Int J Rock Mech Min Sci 2000; 37(1–2):245–62. https://doi.org/10.1016/S1365-1609(99)00104-5. p g [293] Snippe J, Kampman N, Bisdom K, Tambach T, March R, Maier C, Phillips T, Inskip NF, Doster F, Busch A. Modelling of long-term along-fault flow of CO2 from a natural reservoir. Int J Greenh Gas Control 2022;118:103666. https://doi.org/ 10.1016/j.ijggc.2022.103666. [271] Min KB, Rutqvist J, Tsang CF, Jing L. Stress-dependent permeability of fractured rock masses: a numerical study. Int J Rock Mech Min Sci 2004;41(7):1191–210. https://doi.org/10.1016/j.ijrmms.2004.05.005. p g j j [272] Elkhoury JE, Detwiler RL, Ameli P. Can a fractured caprock self-heal? Earth Planet Sci Lett 2015;417:99–106. https://doi.org/10.1016/j.epsl.2015.02.010. j jgg [294] White AF, Brantley SL. The effect of time on the weathering of silicate minerals: why do weathering rates differ in the laboratory and field? Chem Geol 2003;202 (3–4):479–506. https://doi.org/10.1016/j.chemgeo.2003.03.001. [273] Zoback MD, Gorelick SM. Earthquake triggering and large-scale geologic storage of carbon dioxide. Proc Natl Acad Sci USA 2012;(26):10164–8. https://doi.org/ 10.1073/pnas.1202473109. 09. [295] Bossart P, Bernier F, Birkholzer J, Bruggeman C, Connolly P, Dewonck S, Fukaya M, Herfort M, Jensen M, Matray JM, Mayor JC, Moeri A, Oyama T, Schuster K, Shigeta N, Vietor T, Wieczorek K. Mont Terri rock laboratory, 20 Years of research: introduction, site characteristics and overview of experiments. Swiss J Geosci 2017;110(1):3–22. References A note on the scaling relations for opening mode fractures in rock. J Struct Geol 2010;32(10):1485–7. https://doi.org/10.1016/j.jsg.2010.09.007. p g p [283] Khather M, Saeedi A, Myers MB, Giwelli A. Effects of CO2-saturated brine on the injectivity and integrity of chalk reservoirs. Transport Porous Media 2020;135: 735–51. https://doi.org/10.1007/s11242-020-01498-7. [261] Szymczak P, Ladd AJC. The initial stages of cave formation: beyond the one- dimensional paradigm. Earth Planet Sci Lett 2011;301(3–4):424–32. https://doi. org/10.1016/j.epsl.2010.10.026. g j p [262] Steefel CI, Lichtner PC. Multicomponent reactive transport in discrete fractures: II: infiltration of hyperalkaline groundwater at Maqarin, Jordan, a natural analogue site. J Hydrol 1998;209(1–4):200–24. https://doi.org/10.1016/S0022- 1694(98)00173-5. p g [284] Menke HP, Bijeljic B, Blunt MJ. Dynamic reservoir-condition microtomography of reactive transport in complex carbonates: effect of initial pore structure and initial brine pH. Geochem Cosmochim Acta 2017;204:267–85. https://doi.org/10.1016/ j.gca.2017.01.053. [285] Krevor SCM, Pini R, Li B, Benson SM. Capillary heterogeneity trapping of CO2 in a sandstone rock at reservoir conditions. Geophys Res Lett 2011;38:L15401. https://doi.org/10.1029/2011GL048239.l [263] Jimenez-Martinez J, Hyman JD, Chen Y, Carey JW, Porter ML, Kang Q, Guthrie Jr G, Viswanathan HS. Homogenization of dissolution and enhanced precipitation induced by bubbles in multiphase flow systems. Geophys Res Lett 2020;47(7):e2020GL087163. https://doi.org/10.1029/2020GL087163. p g [286] Crandall D, Moore J, Gill M, Stadelman M. CT scanning and flow measurements of shale fractures after multiple shearing events. Int J Rock Mech Min Sci 2017;100: 177–87. https://doi.org/10.1016/j.ijrmms.2017.10.016. [264] Menefee AH, Welch NJ, Frgash LP, Hicks W, Carey JW, Ellis BR. Rapid mineral precipitation during shear fracturing of carbonate-rich shales. J Geophys Res Solid Earth 2020;125(6):e2019JB018864. https://doi.org/10.1029/ 2019JB018864. p g j j [287] Dang W, Wu W, Konietzky H, Qian J. Effect of shear-induced aperture evolution on fluid flow in rock fractures. Comput Geotech 2019;114:103152. https://doi. org/10.1016/j.compgeo.2019.103152. [265] Steefel CI, Hu M. Reactive transport modeling of mineral precipitation and carbon trapping in discrete fracture networks. Water Resour Res 2022;58(9): e2022WR032321. https://doi.org/10.1029/2022WR032321. g j pg [288] Wenning QC, Madonna C, Kurotori T, Pini R. Spatial mapping of fracture aperture changes with shear displacement using X-ray computerized tomography. J Geophys Res Solid Earth 2019;124(7):7320–40. https://doi.org/10.1029/ 2019JB017301. g [266] Gouze P, Noiriel C, Bruderer C, Loggia D, Leprovost R. X-ray tomography characterization of fracture surfaces during dissolution. Geophys Res Lett 2003;30 (5):1267. https://doi.org/10.1029/2002GL016755. [289] Head D, Vanorio T. Effects of changes in rock microstructures on permeability: 3- D printing investigation. Geophys Res Lett 2016;43(14):7494–502. https://doi. org/10.1002/2016GL069334. References https://doi.org/ 10 1016/j chemgeo 2022 121164 j g [228] Lyu Q, Long X, Ranjith PG, Tan J, Kang Y, Wang Z. Experimental investigation on the mechanical properties of a low-clay shale with different adsorption times in sub-/super-critical CO2. Energy 2018;147:1288–98. https://doi.org/10.1016/j. energy.2018.01.084. j g [255] Dockrill B, Shipton ZK. Structural controls on leakage from a natural CO2 geologic storage site: central Utah, USA. J Struct Geol 2010;32(11):1768–82. https://doi. org/10.1016/j.jsg.2010.01.007. gy [229] Jin X, Shah SN, Roegiers JC, Zhang B. An integrated petrophysics and geomechanics approach for fracability evaluation in shale reservoirs. SPE J 2015; 20:518–26. https://doi.org/10.2118/168589-PA. [256] Frery E, Gratier JP, Ellouz-Zimmerman N, Loiselet C, Braun J, Deschamps P, Blamart D, Hamelin B, Swennen R. Evolution of fault permeability during episodic fluid circulation: evidence for the effects of fluid-rock interactions from travertine studies (Utah–USA). Tectonophysics 2015;651:121–37. https://doi. org/10.1016/j.tecto.2015.03.018. [230] Dewhurst DN, Delle Piane C, Esteban L, Sarout J, Josh M, Pervukhina M, Clennell MB. Microstructural, geomechanical, and petrophysical characterization of shale caprocks. Geological Carbon Storage: Subsurface Seals and Caprock Integrity 2018:1–30. https://doi.org/10.1002/9781119118657.ch1. g y p g [231] Wigand M, Kaszuba JP, Carey JW, Hollis WK. Geochemical effects of CO2 sequestration on fractured wellbore cement at the cement/caprock interface. Chem Geol 2009;265(1–2):122–33. https://doi.org/10.1016/j. chemgeo 2009 04 008 g j [257] Kampman N, Busch A, Bertier P, Snippe J, Hangx S, Pipich V, Di Z, Rother G, Harrington JF, Evans JP, Maskell A. Observational evidence confirms modeling of the long-term integrity of CO2-reservoir caprocks. Nat Commun 2016;7:12268. https://doi.org/10.1038/ncomms12268. l g [232] Kaszuba JP, Janecky DR, Snow MG. Carbon dioxide reaction processes in a model brine aquifer at 200 C and 200 bars: implications for geologic sequestration of carbon. Appl Geochem 2003;18(7):1065–80. https://doi.org/10.1016/S0883- 2927(02)00239-1. p g [258] Noiriel C, Deng H. Evolution of planar fractures in limestone: the role of flow rate, mineral heterogeneity and local transport processes. Chem Geol 2018;497: 100–14. https://doi.org/10.1016/j.chemgeo.2018.08.026. 23 A. Vafaie et al. Renewable and Sustainable Energy Reviews 178 (2023) 113270 [259] Durham WB, Bourcier WL, Burton EA. Direct observation of reactive flow in a single fracture. Water Resour Res 2001;37(1):1–12. https://doi.org/10.1029/ 2000WR900228. [281] Pluymakers AM, Samuelson JE, Niemeijer AR, Spiers CJ. Effects of temperature and CO2 on the frictional behavior of simulated anhydrite fault rock. J Geophys Res Solid Earth 2014;119(12):8728–47. https://doi.org/10.1002/2014JB011575. ; ( ) p g [282] Dieterich JH. Time-dependent friction in rocks. J Geophys Res 1972;77:3690–7. https://doi.org/10.1029/JB077i020p03690. [260] Scholz CH. References https://doi.org/10.1007/978-3-319-70458-6_ 1. p [274] Vilarrasa V, Carrera J, Olivella S, Rutqvist J, Laloui L. Induced seismicity in geologic carbon storage. Solid Earth 2019;10(3):871–92. https://doi.org/ 10.5194/se-10-871-2019. [275] Rinaldi AP, Jeanne P, Rutqvist J, Cappa F, Guglielmi Y. Effects of fault-zone architecture on earthquake magnitude and gas leakage related to CO2 injection in a multi-layered sedimentary system. Greenhouse Gases Sci Technol 2014;4(1): 99–120. https://doi.org/10.1002/ghg.1403. [296] Birkholzer JT, Tsang CF, Bond AE, Hudson JA, Jing L, Stephansson O. 25 years of DECOVALEX-Scientific advances and lessons learned from an international research collaboration in coupled subsurface processes. Int J Rock Mech Min Sci 2019;122:103995. https://doi.org/10.1016/j.ijrmms.2019.03.015. g g g [276] Lang PS, Paluszny A, Zimmerman RW. Evolution of fracture normal stiffness due to pressure dissolution and precipitation. Int J Rock Mech Min Sci 2016;88:12–22. https://doi.org/10.1016/j.ijrmms.2016.06.004. [297] Sciandra D, Kivi IR, Vilarrasa V, Makhnenko R, Rebscher D. Hydro-mechanical response of opalinus clay in the CO2 long-term periodic injection experiment (CO2LPIE) at the Mont Terri rock laboratory. Geomech Geophys Geo-Energy Geo- Resour 2022. https://doi.org/10.1007/s40948-022-00442-x. p g j j [277] Pyrak-Nolte LJ, Nolte DD. Approaching a universal scaling relationship between fracture stiffness and fluid flow. Nat Commun 2016;7(1):1–6. https://doi.org/ 10.1038/ncomms10663. [298] Vilarrasa V, Rudqvist J. Thermal effects on geologic carbon storage. Earth Sci Rev 2017;165:245–56. https://doi.org/10.1016/j.earscirev.2016.12.011. [278] Stefanou I, Sulem J. Chemically induced compaction bands: triggering conditions and band thickness. J Geophys Res: Solid Earth 2014;119(2):880–99. https://doi. org/10.1002/2013JB010342. [299] Taron J, Elsworth D. Thermal–hydrologic–mechanical–chemical processes in the evolution of engineered geothermal reservoirs. Int J Rock Mech Min Sci 2009;46 (5):855–64. https://doi.org/10.1016/j.ijrmms.2009.01.007. [279] Hu MM, Hueckel T. Environmentally enhanced crack propagation in a chemically degrading isotropic shale. Geotechnique 2013;63(4):313–21. https://doi.org/ 10.1680/geot.SIP13.P.020. i [300] Viswanathan HS, Ajo-Franklin J, Birkholzer JT, Carey JW, Guglielmi Y, Hyman JD, Karra S, Pyrak-Nolte LJ, Rajaram H, Srinivasan G, Tartakovsky DM. From fluid flow to coupled processes in fractured rock: recent advances and new frontiers. Rev Geophys 2022;60(1):e2021RG000744. https://doi.org/10.1029/ 2021RG000744. [280] Schuler L, Ilgen AG, Newell P. Chemo-mechanical phase-field modeling of dissolution-assisted fracture. Comput Methods Appl Mech Eng 2020;362:112838. https://doi.org/10.1016/j.cma.2020.112838. 24 24
https://openalex.org/W3215247668	https://jurnal.ugm.ac.id/jks/article/download/64498/32604	Indonesian	null	Suling Dewa Sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara	Jurnal Kajian Seni	2,021	cc-by-sa	8,741	VOLUME 08, No. 01, November 2021: 15-35 VOLUME 08, No. 01, November 2021: 15-35 ABSTRACT This study aims to identify the role of Suling Dewa as a symbolic identity of the Karang Bajo community, North Lombok. The groups living in Bayan have different backgrounds. This phenomenon becomes interesting when among the four community groups in Wet Bayan, only the Karang Bajo people present different symbolic values for the existence of Suling Dewa. This theory used in this study is the theory of Burke and Jan E. Stets which states that identity is formed through symbols and the meaning of symbols as a perception. To see how the ideological reflection the author also refers to Thomson’s opinion on ideology and the use of symbolic forms and the attraction between interpretation, self-reflection, and identity criticism. This research identifies that the background predicate as spiritualists in the Karang Bajo community is a fundamental substance that gives birth to a symbolic identity of a supernatural bridge in Suling Dewa. Keywords: Suling Dewa, Karang Bajo Community, Symbolic Identity, Spiritual Predicate SULING DEWA SEBAGAI IDENTITAS SIMBOLIK MASYARAKAT SASAK KUTO-KUTE DI KARANG BAJO BAYAN LOMBOK UTARA M.A Nur Kholis, Wahyu Kurnia Pendidikan Sendratasik, Fakultas Pendidikan, Universitas Nahdlatul Ulama NTB nurkholissumardi@gmail.com M.A Nur Kholis, Wahyu Kurnia Pendidikan Sendratasik, Fakultas Pendidikan, Universitas Nahdlatul Ulama NTB nurkholissumardi@gmail.com Kata kunci: Suling Dewa, Masyarakat Karang Bajo, Identitas Simbolik, Predikat Spiritual ABSTRAK Penelitian ini bertujuan untuk mengidentifikasi peran Suling Dewa sebagai identitas simbolik masyarakat Karang Bajo Lombok Utara. Kelompok yang tinggal di Bayan memiliki latar belakang yang berbeda-beda. Fenomena ini menjadi menarik ketika di antara empat kelompok masyarakat di Wet Bayan, hanya masyarakat Karang Bajo yang menyuguhkan nilai-nilai simbolik yang berbeda atas keberadaan Suling Dewa. Teori yang digunakan dalam penelitian ini adalah teori Burke dan Jan E. Stets yang menyatakan bahwa identitas dibentuk melalui simbol dan pemaknaan simbol sebagai persepsi. Untuk melihat bagaimana refleksi ideologis penulis juga mengacu pada pendapat Thomson tentang ideologi dan penggunaan bentuk simbolik serta daya tarik antara interpretasi, refleksi diri, dan kritik identitas. Penelitian ini mengidentifikasi bahwa latar belakang predikat sebagai spiritualis dalam masyarakat Karang Bajo merupakan substansi fundamental yang melahirkan identitas simbolik sebuah jembatan supranatural di Suling Dewa. 15 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 PENGANTAR 16 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Kelompok-kelompok masyarakat adat Sasak Kuto-kute dengan pemahaman Islam Metu Telu dan hidup di wilayah adat Bayan Beleq sudah saling berdampingan dalam waktu yang sangat lama dalam ruang yang relatif dekat. Mereka dikenal sangat kuat dalam menjaga praktik adat dan tradisi Sasak Kuto-kute. Keempat kelompok masyarakat ini secara eksplisit tampak terbagi dalam ruang eksistensi kelompok masing-masing di wilayah Bayan Beleq. Mereka membuat kelompok sendiri dengan mendirikan pemukiman berbanjar dalam wilayah yang mereka sebut sebagai Kampu. Masyarakat adat yang tinggal di dalam sebuah Kampu masing-masing memiliki petinggi adat yang sangat dihormati dan dijunjung tinggi termasuk dalam hal ini adalah masyarakat Karang Bajo. mereka terikat dalam satu sub etnis yaitu Islam Metu Telu. Selain ikatan sub etnis ideologi, mereka juga memiliki ikatan komunitas yaitu Sasak Kuto-kute serta label orang-orang Wet Bayan yang melekat pada setiap individu. Suling Dewa terkorelasi berbagai macam ritus yang menjadi penyokong integritas sosial masyarakat (Nur Kholis:2017:50). Secara konseptual ritus- ritus antara masyarakat Karang Bajo dan tiga kelompok masyarakat lainnya bersifat homogen. Akan tetapi yang menjadi pembeda adalah eksistensi kesenian di dalam ritus yang menjalankan berbagai sistem dalam fungsionalismenya masing- masing, salah satunya yaitu Suling Dewa dan ritus – ritusnya. Hal di atas mengindikasikan bahwa masyarakat Karang Bajo memiliki perspektif yang berbeda terhadap Suling Dewa. Sebagai musik ritual dalam masyarakat Karang Bajo, Suling Dewa penuh dengan simbol-simbol yang berbicara tentang makna atas konstruksi pola pikir masyarakat Karang Bajo. Konstruksi pola pikir dalam mengekspresikan kelompok melalui bahasa simbol dalam konteks musik dan memiliki korelasi antara subjek dan objek. Pengekspresian ini menjadi penting di dalam kelompok masyarakat Wet Bayan untuk memunculkan eksistensi masyarakat Karang Bajo agar tetap ada di antara ketiga kelompok lainnya sebagai sebuah upaya membangun identitas. Perbedaan antara kelompok masyarakat adat di Bayan Beleq tidak hanya sebatas wilayah teritorial, akan tetapi bersifat kompleks dan mencakup hal yang bersifat mencakup aplikasi prekinesik, mikrokinesik dan kinesik di antara mereka. Hal ini pun dapat dibuktikan melalui primoldialisme yang subjektif terhadap praktik ritus Islam Metu Telu dan implikasi adat istiadat pada kehidupan sehari-hari. PENGANTAR secara kaku perkembangan teknologi. Masyarakat yang hidup di dalam wilayah Bayan Beleq yang tergolong kuat dengan adat tradisi menerima perkembangan terkini seperti sepeda motor, perabotan dapur dan lain sebagainya. Akan tetapi meskipun mereka menerima teknologi terkini, mereka tetap bersifat selektif dan memfilterisasi apa yang dirasa mengganggu integritas adat. Beberapa contoh yang tampak adalah bangunan rumah tempat tinggal mereka yang masih menggunakan bangunan tradisi, pakaian keseharian mereka yang menggunakan pakaian adat Sasak Kuto – kute dan ritus-ritus yang selalu mereka tegakan di dalam berbagai kegiatan keseharian. Suling Dewa merupakan instrumen musik kuno sakral masyarakat Sasak Kuto-kute yang berjumlah 4 buah dan hingga saat ini masih digunakan dalam ritus masyarakat Bayan. Secara geopolitik Bayan adalah sebuah kecamatan, desa sekaligus dusun terpencil di bawah kaki Gunung Rinjani yang berada di Pulau Lombok, Kabupaten Lombok Utara, Provinsi Nusa Tenggara Barat. Kendati demikian secara geokultural Bayan merupakan wilayah adat yang terbagi berdasarkan kelompok dan terlepas dari campur tangan pemerintah. Wilayah inti dari pembagian secara adat atas wilayah Bayan itu sendiri disebut dengan Bayan Beleq yang secara adat disebut sebagai Wet Bayan. Wilayah ini ditandai dengan berdirinya sebuah masjid kuno yang menjadi situs inti dari kegiatan empat kelompok masyarakat yang mendiami wilayah tersebut yaitu masyarakat Loloan, masyarakat Timuq Orong, masyarakat Bat Orong dan masyarakat Karang Bajo. Di luar wilayah adat Bayan Beleq, hidup masyarakat relatif terbuka dalam praktik adat istiadat Sasak Kuto – kute seperti yang mendiami wilayah Ancak, dan Anyar. Relatif terbuka yang dimaksudkan di sini adalah sikap masyarakat yang menerima perkembangan teknologi terbaru dan tidak terlalu menyibukkan diri dengan kegiatan adat, seperti membangun rumah menggunakan semen dan melakukan upacara adat hanya pada hari-hari besar tanggalan adat. Adapun yang dimaksud kuat dalam menjaga integritas adat di sini bukan berarti menolak Penulis melakukan penelitian dengan metode etnografi dan mendapati masyarakat adat yang mendiami wilayah Bayan yaitu Loloan, Timuq Orong, Bat Orong dan Karang Bajo dalam sekrup yang besar menyandang label yang homogen yaitu orang-orang Wet Bayan. Istilah penyebutan orang-orang Wet Bayan mengacu pada keberadaan masjid kuno Bayan yang menjadi situs sakral mereka. Setiap wilayah tempat berdirinya sebuah masjid kuno umumnya memiliki sekelompok masyarakat adat yang memiliki predikat linier antara subjek, objek dan tempat. Contoh masjid kuno Semokan memiliki masyarakat adat Semokan dengan desa adat Semokan. Kendati demikian orang-orang Bayan memiliki pengecualian tersendiri yaitu terdapat empat kelompok masyarakat dengan empat desa adat dalam satu wilayah Bayan dan dengan satu masjid kuno Bayan. PENGANTAR Salah satu contoh setiap kelompok masyarakat merasa bahwa Kampu mereka merupakan Kampu pelopor bagi lahirnya Wet Bayan, dan pada contoh lain masyarakat Loloan melarangan pemudanya menikahi gadis masyarakat Karang Bajo. Ideologi dalam sebuah identitas simbolik perlu dan penting dipertahankan untuk menjadikan sebuah musik menjadi meaning full yang pada akhirnya mampu Jika ditinjau dari segi kebiasaan perilaku, mereka relatif melakukan hal yang serupa. Hal ini dikarenakan 17 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 persepsi. Kedua, kemampuan seseorang dalam menggunakan simbol yang telas dimaknainya kemudian membuat orang lain memaknai sama dengannya. Ketiga, tindakan yang memiliki makna, sehingga ketika orang lain memaknai sama dengan apa yang dimaksudkan. Keempat, ide yang dikeluarkan seseorang berdasarkan situasi lingkungannya. Kelima adalah, bagaimana ide tersebut sesuai dengan individu dan orang lain yang berada dalam lingkungan yang sama, dan ide tersebut merupakan refleksi perasaan dan emosi yang dijadikan dasar dalam bertindak. Ini juga berarti seni musik dapat membentuk sebuah identitas melalui simbol-simbol yang terdapat di dalamnya bagi sebuah masyarakat. Pernyataan ini dilandaskan oleh sifat dari musik itu sendiri yang merupakan sebuah ide yang ditransformasi ke dalam objek bahasa simbolik. menciptakan sistem simbol. Seperti pernyataan Thomson yang mengatakan ideologi mengasumsikan sesuatu yang berbeda, yaitu memiliki karakter kritis. Ia memunculkan pertanyaan baru tentang penggunaan bentuk-bentuk simbol dan ketertarikan antara interpretasi, refleksi- diri, dan kritik (Thomson:2004:40). Ideologi dalam sebuah kelompok, terlebih etnis tertentu selalu tergambar dalam karakteristik identitas seninya. Kejadian ini merupakan bahasa lain, dari cara kelompok tersebut membahasakan budayanya. Akan tetapi ideologi adalah satu hal yang bersifat sensitif dan kaku, sehingga mudah mengundang konflik mana kala dibenturkan dengan ideologi yang berbeda. P e r b e d a a n p i l i h a n a n t a r a kelompok masyarakat Karang Bajo yang mengekspresikan Suling Dewa dan kelompok masyarakat Loloan, Timuq Orong serta Bat Orong yang juga turut melegitimasi eksistensi Suling Dewa sebagai instrumen sakral merupakan sebuah konflik identitas. Konflik ini telah mendorong lahirnya ideologi yang unik dalam masing-masing kelompok masyarakat, sehingga menyebabkan gesekan dan melahirkan integritas simbol baru melalui fenomena musik untuk menunjukkan masing-masing kelompok masyarakat menjadi ada. Hal di atas kemudian mendorong penulis merumuskan masalah dan membuat pertanyaan penelitian apakah Suling Dewa yang digunakan dalam ritus masyarakat Bayan Beleq benar mampu menciptakan identitas simbolik masyarakat Karang Bajo? Bagaimana Suling Dewa mampu menciptakan identitas simbolik bagi masyarakat Karang Bajo? PENGANTAR Mengapa masyarakat Karang Bajo menggunakan Suling Dewa dalam ritus sakral Islam Metu Telu dan tradisi Sasak Kuto-kute? Interaksi sosial merupakan dasar atau latar belakang terbentuknya identitas dalam masyarakat. Burke dan Stets (2009:60) menambahkan bahwa dalam membicarakan identitas ada lima hal yang menjadi perhatian. Pertama adanya simbol- simbol dan makna simbol sebagai bentuk PEMBAHASAN Suling Dewa hingga saat ini masih diakui sebagai musik ritual bagi seluruh kelompok masyarakat Wet Bayan. 18 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Masyarakat Karang Bajo hingga hari ini masih menolak sudut pandang kelompok masyarakat lain yang menganggap Suling Dewa bukan seni tersakral di Wet Bayan. Tidak ada seorang pun yang mampu memberikan alasan mengapa Suling Dewa bagi masyarakat Karang Bajo begitu disakralkan dan bertolak dengan ketiga kelompok masyarakat lainnya. Perbedaan integritas kelompok masyarakat di Wet Bayan mendorong masyarakat Karang Bajo memunculkan eksistensinya salah satunya melalui media Suling Dewa. Penulis memperkirakan pemilihan Suling Dewa sebagai sebuah ekspresi identitas simbolik bagi masyarakat Karang Bajo memiliki pola relasi yang kompleks yang saling berkaitan dan saling membentuk. Jika melihat sumber informasi yang beredar di masyarakat tentang masyarakat Karang Bajo, maka secara keseluruhan yang didapat adalah masyarakat Karang Bajo merupakan masyarakat pelaut yang datang dari arah timur bagian utara pulau Lombok. Tidak ada sumber tertulis berupa babad, ataupun hikayat yang bercerita tentang latar belakang kelompok masyarakat ini di Wet Bayan khususnya masyarakat Karang Bajo. Tidak terdapatnya cerita kelompok masyarakat adat Wet Bayan dalam situs tulis suku Sasak kemungkinan besar disebabkan oleh kehadiran kelompok masyarakat yang sudah sangat lama di Pulau Lombok dan memiliki tenggang waktu ratusan tahun hingga akhirnya suku Sasak mengenal tulisan. Pada akhirnya sumber terkuat yang dapat dijadikan acuan tentang kehadiran masyarakat Karang Bajo adalah sumber oral masyarakat beserta bukti-bukti simbol totemisme yang menguatkan pendapat bahwa masyarakat Karang Bajo pada awal mulanya merupakan orang-orang pelaut. Selain itu sangat minim data ditemukan dalam hal sejarah Sasak di bawah abad 13M yang berbicara khusus terkait masyarakat Sasak di lereng Rinjani. Hal ini dikarenakan tahun 1237M Gunung Samalas (sekarang gunung Rinjani) meletus dan letusannya mengalahkan Krakatau serta Tambora, sehingga membumi hanguskan peradaban di masanya (Sumardi:2017:64). Cerita rakyat kelahiran Suling Dewa yang secara konvensional disepakati Dewa Di dalam mimpi itu juga sang Jero Gamel disuruh menyampaikan kepada masyarakat untuk membawa uang bolong beserta benang sedangkan dia sendiri diberi petunjuk untuk masuk ke Gawah Sereru untuk mengambil satu bambu yang bersinar paling terang. Setelah semua mandat terlaksana, perintah selanjutnya adalah sang Jero Gamel harus meniup bambu tersebut dan meminta seorang wanita tua untuk mendampinginya mengikuti bunyi bambu tersebut menggunakan suara serta memberikan sirih, pinang, benang berikut uang bolong sebagai sesembahan. Setelah melakukannya selama satu hari satu malam akhirnya masyarakat terbebas dari wabah penyakit. Dewa Berdasarkan terminologinya Karang Bajo memiliki interpretasi yang beredar di masyarakat Wet Bayan yaitu Karang Bajo sebagai orang-orang yang berhubungan dengan suku Bajo Sulawesi. Masyarakat Karang Bajo secara garis besar diyakini sebagai masyarakat pelaut yang datang melalui pantai sebelah timur wilayah Lombok Utara. Hanya masyarakat Karang Bajo yang memiliki hubungan atas bentuk-bentuk totemisme hewan-hewan laut pada situs ukir dan ornamentasi yang dimiliki oleh Wet Bayan. Pernyataan ini dibenarkan oleh Kake Renadi serta Kake Alam Kundam seorang tokoh pemuda adat yang penulis wawancara di rumahnya. Mereka mengatakan bahwa ikan di dalam Mesigit Lokaq mewakili kelompok masyarakat Karang Bajo. Cerita rakyat kelahiran Suling Dewa yang secara konvensional disepakati oleh masyarakat Karang Bajo. Menurut 19 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 Inan Gending Mutringen dan Jero Gamel Nyakranom bahwa dahulu masyarakat pernah diserang oleh wabah penyakit yang mematikan. Tidak ada seorang pun yang mampu menemukan obat penawar bagi kesembuhan masyarakat, hingga seorang yang nantinya disebut sebagai Jero Gamel pertama mendapat bisikan untuk mengajak masyarakat bermigrasi ke lereng Gunung Rinjani karena di sana terdapat tanaman sirih beserta pinang. Di dalam mimpi itu juga sang Jero Gamel disuruh menyampaikan kepada masyarakat untuk membawa uang bolong beserta benang sedangkan dia sendiri diberi petunjuk untuk masuk ke Gawah Sereru untuk mengambil satu bambu yang bersinar paling terang. Setelah semua mandat terlaksana, perintah selanjutnya adalah sang Jero Gamel harus meniup bambu tersebut dan meminta seorang wanita tua untuk mendampinginya mengikuti bunyi bambu tersebut menggunakan suara serta memberikan sirih, pinang, benang berikut uang bolong sebagai sesembahan. Setelah melakukannya selama satu hari satu malam akhirnya masyarakat terbebas dari wabah penyakit. masyarakat Sasak hanya diperuntukkan bagi bakal kain sehingga dapat mencirikan bahwa mereka telah mengenal tata cara berpakaian. Korelasi latar belakang masyarakat Karang Bajo terhadap certita rakyat Suling Dewa adalah ketika masyarakat diperintahkan untuk pindah dan bermukim di bawah lereng Gunung Rinjani atau yang dahulu disebut dengan gunung Samalas (Sumardi:2017:64). Masyarakat Karang Bajo berpindah posisi menuju lereng gunung dan di dalam versi latar belakang masyarakat Karang Bajo menegaskan mereka adalah masyarakat pesisir. Inan Gending Mutringen dan Jero Gamel Nyakranom bahwa dahulu masyarakat pernah diserang oleh wabah penyakit yang mematikan. Tidak ada seorang pun yang mampu menemukan obat penawar bagi kesembuhan masyarakat, hingga seorang yang nantinya disebut sebagai Jero Gamel pertama mendapat bisikan untuk mengajak masyarakat bermigrasi ke lereng Gunung Rinjani karena di sana terdapat tanaman sirih beserta pinang. Masyarakat Masyarakat Karang Bajo dalam keutuhan Wet Bayan menunjukkan bahwa kelompok masyarakat mereka merupakan masyarakat spiritualis. Setiap kelompok masyarakat menunjukkan kedudukan dapat dilihat melalui salah satu fenomena yaitu masyarakat Bat Orong yang merasa kolektifnya sebagai trah kerajaan Bayan zaman dahulu sedang masyarakat Karang Bajo, mereka tidak peduli akan sistem imperalialisme seperti masyarakat Bat Orong, seperti yang dikatakan Max Weber tentang rasionalitas (2003:115), masyarakat Karang Bajo lebih mengutamakan rasionalitas afektual dan rasionalitas nilai dibanding dengan rasionalitas tujuan dan rasionalitas tradisional. Hal ini mendorong lahirnya perlakuan yang berbeda dari masing-masing kolektif tentang Suling Dewa, mulai dari cerita rakyat hingga fungsionalismenya. Dalam Cerita rakyat tentang Suling Dewa di atas menggambarkan kaitan dengan masyarakat Karang Bajo yang dipersepsikan sebagai orang pelaut. Uang bolong dalam certita rakyat Suling Dewa mengindikasikan bahwa sebelum tinggal di Wet Bayan yang berada di lereng Rinjani, mereka sudah mengenal alat tukar berupa uang, dan hal ini tercermin hingga hari ini dalam sesembahan mahar pernikahan. Sedangkan benang bagi 20 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara cerita rakyat masyarakat Karang Bajo, Suling Dewa hadir dengan metafora bahasa dan diyakini datang dari bisikan makhluk supranatural sedangkan dalam cerita rakyat pada kolektif lainnya cenderung memiliki jalan cerita yang sama namun berbeda ketika Jero Gamel mendapati mimpi dan diperintahkan untuk memotong bambu. Perbedaan cerita rakyat Suling Dewa antara masyarakat Karang Bajo dan kolektif lainnya mempertegas masyarakat Karang Bajo sangat identik dengan kiasan dunia ruh. makhluk supranatural. Secara etimologis Suling Dewa berarti seruling makhluk halus atau seruling supranatural. Arti ini disepakati oleh komunal Wet Bayan sebagai sebuah kebenaran. Namun yang menjadi keunikan, hanya masyarakat Karang Bajo yang memiliki pemahaman linier antara cerita rakyat Suling Dewa dengan istilah Suling Dewa dan latar belakangnya sebagai kolektif spiritualis yang sarat dengan dunia ruh. Masyarakat Karang Bajo memahami Suling Dewa sebagai sebenar-benarnya instrumen yang datang dari dunia supranatural. Mereka menganggap bahwa Suling Dewa merupakan bagian dari unsur-unsur supranatural yang harus dijalin kerja samanya guna mencapai kesempurnaan dalam menjalankan ritus. Secara garis besar seluruh kolektif di Wet Bayan cenderung kuat akan kepercayaan hal gaib. Kendati demikian, masyarakat Karang Bajo memiliki tingkatan yang lebih kuat dalam praktik spiritual. Masyarakat Hingga tersebar opini dalam kolektif yang berbeda bahwa masyarakat Karang Bajo ahli dalam berhubungan dengan dunia ruh sehingga dalam upacara besar kedudukan masyarakat Karang Bajo sebagai pemberi sembeq kepada peserta upacara yang terdiri dari seluruh kolektif Wet Bayan. Atas posisinya sebagai kolektif spiritual di Wet Bayan, masyarakat Karang Bajo membentuk Suling Dewa dalam berbagai nilai guna untuk menjalani ritus-ritus mereka yang tidak dimiliki oleh kolektif lainnya. Masyarakat Karang Bajo memberi nilai pada setiap aspek Suling Dewa mulai dari gending Suling Dewa, organologi, hingga praktisi. Berikut uraian tentang pemahaman nilai makna pada Suling Dewa bagi masyarakat Karang Bajo. a. Gending Suling Dewa Gending Suling Dewa terdiri dari 44 gending dan 5 di antaranya merupakan gending wajib yang dimainkan dalam setiap ritual yaitu Lembuneng Meloang, Pem Pang Poq, Kamboja, Lokoq Sebie dan Bao Daya. Kelima gending di atas mewakili Pancadewata. Buktinya adalah ketika keempat Suling Dewa digunakan dalam ritus tersakral oleh masyarakat Karang Bajo maka masing-masing Suling Dewa akan memainkan keempat gending di atas bersamaan untuk menciptakan dinding imajiner dan ditutup gending Pemahaman Suling Dewa Suling Dewa berasal dari kata Suling yang berarti seruling (alat tiup) dan Dewa yang berarti energi supranatural (spirit atau jin). Dalam kata lain istilah kedewayan memiliki arti kerasukan 21 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 Bao Daya. Dalam konsep pancadewata, Bao Daya diartikan titik tengah, pusat energi dan tempat berkumpulnya 4 arah mata angin. Ini membuktikan Bao Daya berperan khusus dalam Suling Dewa. Ketika ritual pengobatan (mendewa), dimainkan keempat gending wajib dan akhirnya ditutup dengan gending Bao Daya. Namun tidak jarang ketika mendewa berlangsung, pasien kerasukan, meminta memainkan gending di luar gending wajib, akan tetapi berapapun jumlah gending permintaan tetap ditutup oleh gending Bao Daya. rakyat Suling Dewa. Di dalam cerita rakyat diceritakan bahwa Jero Gamel mendapatkan bambu berkilau di lereng Rinjani, setelah dimainkan barulah dia mencari wanita tua untuk melagukan mantra. Jero Gamel juga menyatakan bahwa Bao Daya merupakan maki sedang gending yang lainnya merupakan nini. Hal ini sekaligus memberi bukti bahwa gending Bao Daya dirasa lebih dulu ada dan paling dekat dengan nilai ketuhanan dibandingkan gending yang lainnya dan transformasi dari energi Gunung Rinjani. Gending Bao Daya secara terminologi terdiri dari dua diksi kata yaitu Bao yang berarti teduh atau sejuk dan Daya yang berarti utara (bahasa di luar Sasak Kuto-kute), selatan (bahasa Sasak Kuto- kute) atau kekuatan. Secara etimologis Bao Daya berarti kekuatan atau energi Rinjani untuk menyejukkan. Dikatakan demikian karena Daya dalam bahasa Sasak umumnya berarti utara, namun masyarakat Sasak Kuto-kute menyebut Daya sebagai selatan. Hal ini terjadi karena arah Daya bagi masyarakat Sasak menunjuk pada letak posisi Gunung Rinjani. Sebab melihat keberadaan Kampu Karang Bajo terletak di lereng utara Gunung Rinjani maka mereka menyebut Daya sebagai selatan. Pada istilah yang lain dalam bahasa Sasak, Daya juga berarti kekuatan, “Ndeq Ku Dayaq” aku tidak punya kekuatan. masyarakat menitik arah Daya kepada Gunung Rinjani, karena Gunung Rinjani adalah sumber kekuatan bangsa Sasak. b. Organologi dan Instrumen Suling Dewa 08, No. 01, November 2021: 15-35 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 Gambar 2 Jero Gamel Anggalip dan Suling Dewa Replikanya. Sumber : Dokumentasi pribadi, 21 Februari 2014 dahulu harus melalui proses sakralisasi yang memiliki maksud mengikat antara Jero Gamel dan Inan Gending terhadap Suling Dewa-nya. Masyarakat Karang Bajo percaya, Suling Dewa hidup (subjek) dan harus dihargai layaknya makhluk, sehingga menjadi pelakunya harus sakralisasi terlebih dahulu. Sakralisasi antara pemain Suling Dewa dan instrumen Suling Dewa ini membuat seniman memiliki perlakuan khusus dalam sosial masyarakat. Perlakuan khusus ini serupa dengan konsep strata dalam Hindu, yaitu brahma, agamawan dan intelektual (seniman) merupakan kaum di atas para petani (rakyat biasa). Maka memanggil nama praktisi Suling Dewa, vokalis ataupun peniup suling, harus memberikan sisipan gelar di muka lalu disusul nama panggilan. Ketika seorang ahli waris Suling Dewa belajar meniup seruling replika, dia tidak disebut sebagai Jero Gamel, melainkan penyuling (istilah biasa). Gambar 2 Jero Gamel Anggalip dan Suling Dewa Replikanya. Sumber : Dokumentasi pribadi, 21 Februari 2014 Nyakranom mengatakan bahwa tanpa adanya angin maka bunyi tidak dapat terdengar dan makhluk tidak dapat hidup begitu juga Suling Dewa. Nyakranom mengatakan bahwa tanpa adanya angin maka bunyi tidak dapat terdengar dan makhluk tidak dapat hidup begitu juga Suling Dewa. Gelar peniup Suling Dewa saat proses upacara berlangsung yang semula disebut sebagai Jero Gamel berubah menjadi Amaq Maki sedangkan gelar vokalis Suling Dewa sebagai Inan Gending berubah menjadi Inaq Nini. Kedua gelar ini secara hierarki dalam konsep pemahaman masyarakat Karang Bajo lebih tinggi dari pada gelar sebelumnya. Maki dan nini juga merupakan konsep dualitas dalam masyarakat Karang Bajo. Ketika upacara dan ritus tersakral berlangsung, para praktisi Suling Dewa mendapat perlakuan lebih spesial dari hari-hari biasa. b. Organologi dan Instrumen Suling Dewa Memahami Suling Dewa, masyarakat Karang Bajo menganggapnya sebagai sebuah subjek hidup, bukan sekedar objek yang melalui manifestasi subjek. Maq Lokaq Walin Gumi Cameng salah satu petinggi masyarakat Karang Bajo. Beliau mengatakan: “Suling Dewa no no idup, mun ya mate, soraq nya bau ngatongan pujiq ta kon Nenek” Terjemahan: Suling Dewa itu hidup, jika dia mati, mana mungkin bisa menghantarkan doa atau permohonan ke Neneq. Sebelum membahas penting untuk memberi gambaran visual istilah-istilah dalam organologi Suling Dewa. 1) Seliper atau seleper adalah tempat lahirnya melodi permohonan yang menjadi perwakilan masyarakat. Seliper yang mengikat ujung kepala Suling Dewa dimaknai sebagai sebuah sapuq. Pemaknaan ini merujuk Eksistensi Bao Daya sebagai gending penutup juga berkaitan dengan cerita 22 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Jurnal Kajian Seni Manuscript Submission Te Seliper/Selepir/Seleper Pengosap Loang Bilo Awaq Poto Gambar 1 Istilah Organologi Gambar 1 Istilah Organologi. (Oleh Nur Kholis 2016) Pengosap Loang Awaq Gambar 1 Istilah Organologi Gambar 1 Istilah Organologi. (Oleh Nur Kholis 2016) sebagai kaki. Poto ini selalu menempel dengan bumi, mengisyaratkan hubungan dunia bawah, keberadaan prosesi di dunia nyata mengisyaratkan hubungan dunia tengah, dan bunyi representasi dari Suling Dewa menjadi hubungan dunia atas. 2 ) pada benteng pola pikir agar tidak terkontaminasi. ( 2) Loang yang terdiri dari enam buah. Melambangkan enam jumlah indra layaknya manusia yang memiliki tugas berbeda-beda. 3) Awaq mengartikannya sebagai sebuah tubuh dari Suling Dewa. Tubuh diartikan tercipta untuk taat, berbakti dan kembali kepada yang kuasa. 3) Awaq mengartikannya sebagai sebuah tubuh dari Suling Dewa. Tubuh diartikan tercipta untuk taat, berbakti dan kembali kepada yang kuasa. 6) Bilo, Bagian ini memiliki arti sederhana sebagai embrio atau bibit awal. Bilo merupakan sebutan bagi bambu yang telah disakralisasi dan merupakan bambu yang terpilih untuk menjadi replika Suling Dewa dan belum memiliki lubang. 6) 6) Bilo, Bagian ini memiliki arti sederhana sebagai embrio atau bibit awal. Bilo merupakan sebutan bagi bambu yang telah disakralisasi dan merupakan bambu yang terpilih untuk menjadi replika Suling Dewa dan belum memiliki lubang. 4) Pengosap, bagian ini merupakan dodot, bebet, dan slewoq bagi Suling Dewa. Dengan kata lain pengosap dipredikatkan sebagai pakaian Suling Dewa. 7) Embok atau yak. Bagian terakhir ini berarti nafas. Masyarakat memaknai embok sebagai sumber kehidupan bagi Suling Dewa. Jero Gamel 5) Poto diartikan sebagai pijakan Suling Dewa, sejenis anatomi kaki, akan tetapi masyarakat tidak menyebutnya 23 Jurnal Kajian Seni, Vol. c. Pemain Suling Dewa Suling Dewa memiliki delapan orang pemain. Empat vokalis (Inan Gending) dan empat peniup Suling Dewa (Jero Gamel). Formasi ini tidak bersifat profan dan dengan mudah didapatkan. Dalam hal pemberian istilah pada setiap pelaku, masyarakat Bayan memiliki aturan yang berbeda dari kebanyakan masyarakat umum yang kita ketahui tentang bagaimana cara memperlakukan seorang musisi atau pelaku seni. Sebelum pemain mampu diamanatkan menjadi praktisi Suling Dewa, mereka terlebih 24 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Suling Dewa sebagai motor dan Jero Gamel serta Inan Gending sebagai penggeraknya. Ketika tiga unsur ini bersinergi maka tercipta sebuah jembatan yang menghubungkan komunikasi dua arah antara alam yang terpisah dimensi. Kolektif Wet Bayan meyakini terdapat simbiosis mutualisme antara kedua alam, ketika secara hakikat manusia sebagai pelayan dan penjaga alam, lalu alampun melayani, memberikan kebutuhan dan menjaga manusia (maki dan nini). Kolektif Wet Bayan mempercayai Suling Dewa merupakan manifestasi jembatan yang mampu mengubungkan antara alam manusia dengan alam gaib. Mencakup sang pencipta dan makhluk supranatural. Maka dibangunlah konsep kerja antara dua alam yang berbeda. Berikut adalah gambaran kerja praktisi Suling Dewa dalam pemahaman Wet Bayan. Suling Dewa sebagai motor dan Jero Gamel serta Inan Gending sebagai penggeraknya. Ketika tiga unsur ini bersinergi maka tercipta sebuah jembatan yang menghubungkan komunikasi dua arah antara alam yang terpisah dimensi. Kolektif Wet Bayan meyakini terdapat simbiosis mutualisme antara kedua alam, ketika secara hakikat manusia sebagai pelayan dan penjaga alam, lalu alampun melayani, memberikan kebutuhan dan menjaga manusia (maki dan nini). Gambar 3 Suling Dewa, Jembatan, Alam Manusia dan Tuhan (Penulis, 2019) Suling Dewa - Jero Gamel - Inan Gending - Gending - Instrumen Musik Alam Manusia Alam Gaib Sistem Maki dan Nini dalam Suling Dewa Suling Dewa Jero Gamel Inan Gending Gending Instrumen Musik Habermas mengatakan bahwa musik merupakan sebuah hasil dari pengetahuan manusia dan pengetahuan terbentuk dari ekspresi kehidupan yang terdiri dari ekspresi linguistik, ekspresi tindakan dan ekspresi pengalaman (Habermas dalam Yanti:2016:30). Merujuk pendapat ini, Suling Dewa sebagai fenomena musik dalam masyarakat Karang Bajo terbentuk melalui seluruh elemen ekspresi tersebut. Masyarakat Karang Bajo yang pada latar belakangnya merupakan masyarakat spiritualis secara fundamental identik dengan dunia ruh, mengekspresikan kehidupannya ke dalam Suling Dewa. Ini dibuktikan melalui bagaimana mereka mempersepsikan Suling Dewa berbeda dari kolektif lain di Wet Bayan. Segala macam bentuk kemiripan yang menuju sifat homogenitas selalu terdistorsi melalui pemahaman ruh, seperti cerita rakyat Suling Dewa yang mana diketahui bahwa perbedaan cerita yang terdapat Alam Manusia Gambar 3 Suling Dewa, Jembatan, Alam Manusia dan Tuhan (Penulis, 2019) Gambar di atas menjelaskan antara alam manusia dan alam gaib terpisah melalui dimensi. Alam ghaib secara konseptual berhubungan dengan Suling Dewa, dan alam manusia secara konseptual berhubungan dengan Jerogamel dan Inan Gending. Kedua alam yang terpisah ini disatukan melalui jembatan yang dihasilkan melalui tiga rangkaian unsur yang berbeda yaitu 25 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 mewawancarai Penganak saat membahas konsep maki dan nini. Dia mengatakan “Neneq memang tunggal, betul, tetapi apakah anda merasa Islam tanpa ada Nabi Muhammad? Saya pernah berdebat dengan ulama, saya mengatakan bahwa jika agama ini adalah langit, maka adat ini adalah bumi. Tidak ada yang mengatakan itu langit tanpa adanya bumi.” pada masyarakat Karang Bajo terletak pada asal-usul bisikan gaib. Ekspresi seperti ini pada dasarnya linier terhadap representasi yang telah diinterpretasikan dan pada akhirnya menjadi sebuah persepsi baik dari segi individu ataupun komunal. Ekspresi yang dilakukan oleh masyarakat Karang Bajo kemudian menciptakan simbol melalui landasan ideologi yang mereka miliki. Ini merupakan proses dan upaya untuk mempertahankan eksistensinya di antara keempat kolektif yang hidup di Wet Bayan. Jurnal Kajian Seni Islam jelas dikatakan menjadi seorang D a r i p e r n y a t a a n P e n g a n a k menunjukkan ketunggalan Allah SWT dalam konteks individual digiring ke arah aspektual (agama Islam) yang mana dalam Islam jelas dikatakan menjadi seorang muslim harus mengakui eksistensi Nabi Muhammad SAW dan Allah SWT. Secara konsep maki nini pada akhirnya menyimpulkan bahwa jika ingin dikatakan orang beragama maka secara otomatis manusia harus menjadi orang beradat. Sistem Maki dan Nini dalam Suling Dewa Manuscript Submission Templa uslim harus mengakui eksistensi Nabi ara konsep maki nini pada akhirnya orang beragama maka secara otomatis nsep kuno suku Sasak tergambar jelas ektual dan universal. Tampak secara ut dan memanggil praktisi Suling Dewa Maki dan Nini merupakan konsep keseimbangan dalam masyarakat Karang Bajo yang berasal dari kepercayaan kuno Sasak. Konsep dualisme maki nini dapat dijumpai dalam setiap komunitas suku Sasak Lombok namun dengan istilah dan sistem yang berbeda. Pemahaman maki dan nini dalam masyarakat Karang Bajo mengurai setiap unsur individual akan menemui pelengkapnya pada tingkatan aspektual, dan universal. Pernyataan ini dipertegas ketika penulis Muhammad SAW dan Allah SWT. Se menyimpulkan bahwa jika ingin dikatak manusia harus menjadi orang beradat. Konsep maki nini sebagai sebuah di dalam unsur Suling Dewa secara a eksplisit masyarakat Karang Bajo menye pada saat ritual sedang berlangsung d suling dan Inaq Nini bagi vokalis, namun disebut sebagai nini sedangkan instrum lebih jelasnya lihat rantai kerja maki ni j di b h i i Konsep maki nini sebagai sebuah konsep kuno suku Sasak tergambar jelas di dalam unsur Suling Dewa secara aspektual dan universal. Tampak secara eksplisit masyarakat Karang ngan sebutan Amaq Maki bagi peniup ada tingkatan lebih jauh kedua praktisi n suling disebut sebagai maki. Untuk dalam pemahaman masyarakat Karang reaksi gabungan Inan Gending + Jero Gamel => Praktisi + Suling Dewa (instrumen) Nini Maki Nini Maki => Pra/Sarana + Ritus => Adat Tradisi + Agama => Manusia dan Alam Nini Maki Nini Maki Nini Maki Gambar 4 Rantai Kerja Maki dan Nini Gambar 4 Rantai Kerja Maki dan Nini (Penulis, 2019) Gambar 4 Rantai Kerja Maki dan Nini Rantai Kerja Maki dan Nini (Penulis, 2019) 26 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Bajo menyebut dan memanggil praktisi Suling Dewa pada saat ritual sedang berlangsung dengan sebutan Amaq Maki bagi peniup suling dan Inaq Nini bagi vokalis, namun pada tingkatan lebih jauh kedua praktisi disebut sebagai nini sedangkan instrumen suling disebut sebagai maki. Untuk lebih jelasnya lihat rantai kerja maki nini dalam pemahaman masyarakat Karang Bajo di Gambar 4. sebuah ideologi. Seperti kelompok Bat Orong yang feodalis, kelompok Loloan yang merasa berperan sebagai pelindung Wet Bayan dan kelompok Karang Bajo yang berperan sebagai ahli spiritual. Latar belakang spiritualis telah dikonvensi oleh masyarakat Karang Bajo sebagai sebuah ideologi bersama. Sistem Maki dan Nini dalam Suling Dewa Ini menjadikan mereka memiliki pemahaman tersendiri tentang Suling Dewa, kemudian membuat representamen tentang seluruh material yang ada di dalam Suling Dewa bagi masyarakat Karang Bajo berbeda dengan kelompok masyarakat lainnya. Meninjau rantai kerja di atas maka dapat dilihat bahwa maki nini merupakan sistem kerja proporsional dari relasi struktur yang saling melengkapi. Selanjutnya pemberian predikat maki dan nini dalam paradigma berpikir masyarakat Karang Bajo tidak terjadi begitu saja. Setiap unsur yang diberi predikat maki merupakan unsur yang secara konseptual dipandang lebih dekat terhadap nilai ketuhanan dan lebih dulu ada sedang yang menjadi predikat nini adalah penunjang keutuhan untuk mencapai kesempurnaan nilai. Predikat spiritualis menjadi sebuah konsep yang tidak lepas dari Wet Bayan yang menciptakan empat pilar kelompok masyarakat untuk menyokong eksistensi Wet-nya. Keberadaan kaum spiritualis ini juga tidak terlepas melalui eksistensi Islam Metu Telu sebagai kepercayaan yang dianut oleh masyarakat Wet Bayan. Sebagaimana rahasia umum yang telah kita ketahui bahwa dalam setiap kepercayaan, baik Hindu maupun Islam serta yang lainnya, wajib memiliki kedudukan khusus bagi spiritualis. Dalam Islam kaum spiritualis ini dikenal dengan sebutan Ulama sedangkan dalam istilah Hindu kaum ini dikenal dengan sebutan Brahmana. Spiritualis sebagai Ideologi Salah satu syarat mutlak untuk menciptakan kolektif atau komunitas dan kelompok adalah sebuah pemahaman, kesepakatan dan pola pikir yang homogen. Keseragaman berpikir ini dalam masyarakat Karang Bajo tercipta dalam predikat kaum spiritualis. Adat memiliki andil yang penting dalam Islam Metu Telu. Prinsip ini jelas tampak dalam konsep maki nini. Konsep saling melengkapi maki nini mendorong segala unsur individual masuk ke dalam sistem dualitas pada tahapan aspektual. Pada akhirnya konsep ini berbicara manusia tidak dapat dikatakan beragama tanpa memiliki adat, bagaikan Masyarakat Karang Bajo yang tergabung dalam kesamaan Wet bersama ketiga kelompok masyarakat lainnya memiliki integritas masing-masing. Mereka membentuk konstruksi kultural tersendiri menggunakan sumber-sumber wacana melalui fondasi material sebagai 27 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 masyarakat Karang Bajo. Karena secara kausal maki nini memberi gambaran bahwa tingkat ritual yang tersakral sejajar dengan tingkat instrumen yang digunakan (Suling Dewa), yang menjadi pembeda hanya terletak dalam persoalan yang mana lebih dekat dengan nilai ketuhanan atau yang mana lebih dulu ada. Ini otomatis menggambarkan Suling Dewa memiliki posisi yang sangat penting bagi masyarakat Karang Bajo karena berkedudukan sebagai nini dalam ritual (maki) yang penting dan tersakral. Akan tetapi ada pertanyaan baru bahwa mengapa Suling Dewa begitu dianggap penting. Hal tersebut kembali kepada bagaimana pola berpikir dan pola persepsi masyarakat Karang Bajo dalam mengekspresikan Suling Dewa. Untuk lebih jelas akan dibahas pada sub bab selanjutnya. langit yang tidak mendapat predikat langit tanpa eksistensi bumi. Maki nini juga mendorong posisi agama dan adat menjadi seimbang. Dalam dikotomi oposisi biner ini, masyarakat Wet Bayan mengistilahkan agama sebagai maki dan adat sebagai nini berlandaskan asas entitas mana yang lebih dekat nilai ketuhanannya. Jika secara agama memiliki konsep kaum spiritual, maka secara adatpun mereka membuat ahli spiritual dalam bentuk kolektif, yaitu masyarakat Karang Bajo. Kedudukan Suling Dewa dalam Ritus Maq Lokaq Pande sebagai petinggi adat masyarakat Karang Bajo mengatakan tidak ada satu orang pun yang melakukan ritual bersifat menyucikan (ngaponin) tanpa menghadirkan Suling Dewa sebagai pelindung energi supranatural negatif. Mengacu pernyataan ini jika meninjau rantai kerja maki nini, jelas bahwa ritus berposisi maki, dan Suling Dewa dengan Jero Gamel berserta Inan Gending merupakan nini yang melengkapi maki. Namun permasalahan di atas tidak hanya sekedar pada permukaan maki dan nini. Lebih dalam bagaimana maki dan nini bekerja dalam aplikasinya. Kedudukan Suling Dewa dalam Ritus Kedudukan Suling Dewa dalam Ritus Maq Lokaq Pande sebagai petinggi adat masyarakat Karang Bajo mengatakan tidak ada satu orang pun yang melakukan ritual bersifat menyucikan (ngaponin) tanpa menghadirkan Suling Dewa sebagai pelindung energi supranatural negatif. Mengacu pernyataan ini jika meninjau rantai kerja maki nini, jelas bahwa ritus berposisi maki, dan Suling Dewa dengan Jero Gamel berserta Inan Gending merupakan nini yang melengkapi maki. Spiritualis sebagai Ideologi Sebuah subjek untuk menjadi nini bagi sebuah objek atau subjek maki harus bersifat proporsional, sebab jika tidak, maka subjek atau objek yang bersifat maki akan menolak subjek atau objek yang bersifat nini, sehingga sistem tidak akan berjalan. Proporsionalitas ini menggiring titik cerah jawaban atas pertanyaan mengapa Suling Dewa harus dihadirkan dalam ritual tersakral Suling Dewa Sebagai Identitas Simbolik Suling Dewa Sebagai Identitas Simbolik Posisi Wet Bayan sudah memasuki tahap heterogen. Memiliki empat kelompok masyarakat dan Kampu, posisi ini mendapati konsekuensi logis dari sifat multikulturalisme. Sifat ini bukanlah sebuah hasil akhir dari perkembangan struktur budaya. Dalam konteks multikultur setiap komunal menciptakan pola memunculkan (becoming) untuk menjadi ada. Namun permasalahan di atas tidak hanya sekedar pada permukaan maki dan nini. Lebih dalam bagaimana maki dan nini bekerja dalam aplikasinya. Sebuah subjek untuk menjadi nini bagi sebuah objek atau subjek maki harus bersifat proporsional, sebab jika tidak, maka subjek atau objek yang bersifat maki akan menolak subjek atau objek yang bersifat nini, sehingga sistem tidak akan berjalan. Proporsionalitas ini menggiring titik cerah jawaban atas pertanyaan mengapa Suling Dewa harus dihadirkan dalam ritual tersakral Suling Dewa sebagai sebuah seni musik Wet Bayan, secara konseptual tidak hanya ditemui dalam masyarakat Karang Bajo. Setiap kelompok masyarakat Wet Bayan mengakui dan mengklaim kepemilikan atas eksistensi Suling Dewa. Ini menjadikan secara fundamental 28 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Suling Dewa bersifat relatif, dengan esensi objektif yang seragam dalam pemahaman Wet Bayan. Kerelatifan Suling Dewa dalam paham kolektif Wet Bayan menjadi sebuah perspektif baru bagi masyarakat Karang Bajo dalam memunculkan kelompoknya. Mengapa demikian? Karena di antara keempat kolektif hanya Karang Bajo yang mengekspresikan Suling Dewa secara berbeda. Dewa. Serupa dengan yang dikatakan Jameson (2007:80) bahwa identitas budaya dilahirkan melalui struktur budaya yang tersusun berdasarkan: 1) pola persepsi, 2) pola berpikir dan 3) perasaan, sedangkan identitas sosial terbentuk melalui struktur sosial yang tersusun berdasarkan pola-pola perilaku sosial. Identitas budaya sangat diperlukan bagi sekelompok orang yang hidup di lingkungan yang bersifat heterogen dan multikultur. Karena tanpa identitas budaya sebuah kelompok tidak dapat menjadi eksis di antara kebudayaan lainnya. Klaim Suling Dewa hingga sekarang masih sebagai aksioma dalam setiap kolektif. Mereka memiliki animo yang kuat atas eksistensi Suling Dewa, hingga banyak pendapat apriori muncul dalam kesejarahan Suling Dewa. Seperti di saat seluruh kolektif sepakat Jero Gamel mendapatkan mimpi perintah memotong bambu, masyarakat Karang Bajo hadir dengan pendapat Jero Gamel mendapat bisikan untuk mengambil bambu yang paling bercahaya. Cerita rakyat ini dengan sendirinya menggiring pewacanaan bisikan gaib dan bambu yang bercahaya merupakan simbol spiritualis masyarakat Karang Bajo. Hal ini serupa seperti apa yang dikatakan oleh Langer bahwa prinsip seni dan penciptaannya diekspresikan melalui simbol-simbol. Suling Dewa Sebagai Identitas Simbolik Seni memiliki khaidah sendiri berupa ekspresi atau bentuk ekspresi yang memiliki ciri-ciri hubungan secara simbolis dengan kehidupan (Langer:2006:40). Proses membangkitkan identitas yang dilakukan oleh masyarakat Karang Bajo juga turut serta berdampak pada Suling Dewa. Sebab musik sebagai sebuah seni merupakan salah satu dari unsur kebudayaan. Seperti yang dikatakan oleh Koentjaraningrat (2002:30) bahwa seni adalah salah satu dari 7 unsur kebudayaan. Maka mana kala sebuah kelompok membentuk identitas dalam kebudayaannya maka hal tersebut akan berdampak pada progres keseniannya, ini dapat dilihat ketika masyarakat Karang Bajo membentuk identitas bagi kelompoknya maka logis jika mereka juga membentuk identitas melalui simbol- simbol dalam Suling Dewa. Suling Dewa sebagai ekspresi simbolik dari hasil pola persepsi dan pola berpikir masyarakat Karang Bajo yang dibangun berdasarkan transformasi ideologi masyarakat dalam sebuah objek. Thomson mengatakan ideologi mengasumsikan sesuatu yang berbeda, yaitu memiliki karakter kritis. Ia Labelitas masyarakat Karang Bajo sebagai ahli spiritual memiliki nilai intrinsik berupa memanifestasikan pola pikir dan persepsi masyarakat Karang Bajo yang diekspresikan melalui Suling 29 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 memunculkan pertanyaan baru tentang penggunaan bentuk-bentuk simbol dan ketertarikan antara interpretasi, refleksi- diri, dan kritik (Thomson:2004:40). Suling Dewa sebagai bagian dari fenomena musik merefleksikan ideologi masyarakat Karang Bajo melalui bahasa simbolik (bunyi) dan subjek atau objek simbolik yang pada akhirnya membangun identitas bagi masyarakat Karang Bajo. Oleh sebab Suling Dewa merupakan musik ritual maka unsur-unsur yang terkandung di dalamnya tidak akan terlepas dari makna dan nilai antara teks serta konteks atas eksistensinya di tengah masyarakat Karang Bajo. dalam bentuk ideologi akan terefleksi sebagai bentuk bagian musik. Meninjau gending Suling Dewa melalui ideologi masyarakat Karang Bajo sebagai kaum spiritualis yang sarat dengan hal gaib (dunia ruh) maka akan menjadi saling berkaitan. Aspek nini pada gending sebagai pijakan alam dunia pada kenyataannya juga berstatus nini dan aspek maki pada gending mengisyaratkan fondasi- fondasi jembatan yang saling kait mengait. Gending-gending Suling Dewa pada masyarakat Karang Bajo bersifat seperti puzzel anak tangga yang saling berhubungan. Terdapat 44 gending pakem Suling Dewa di dalam masyarakat Karang Bajo. 43 gending memiliki lirik kecuali satu gending, yaitu gending Bao Daya yang tidak memiliki lirik (instrumental). Burke dan Jan E. Stets mengatakan bahwa dalam membicarakan identitas ada lima hal yang menjadi perhatian. Pertama adanya simbol- simbol dan makna simbol sebagai bentuk persepsi. Kedua, kemampuan seseorang dalam menggunakan simbol yang telah dimaknainya kemudian membuat orang lain memaknai sama dengannya. Suling Dewa Sebagai Identitas Simbolik Ketiga, tindakan yang memiliki makna, sehingga ketika orang lain memaknai sama dengan apa yang dimaksudkan. Keempat, ide yang dikeluarkan seseorang berdasarkan situasi lingkungannya. Kelima adalah, bagaimana ide tersebut sesuai dengan individu dan orang lain yang berada dalam lingkungan yang sama, dan ide tersebut merupakan refleksi perasaan dan emosi yang dijadikan dasar dalam bertindak (Burke dan Jan E. Stets: Makna yang tercipta dalam Suling Dewa selalu berkaitan dengan dunia gaib sebagai ekspresi kehidupan spiritual masyarakat Karang Bajo. Namun segala hal dalam interpretan Suling Dewa yang berlandaskan kegaiban selalu dilengkapi oleh alam manusia sebagai bukti terapan sistem dualitas maki dan nini tidak boleh lepas. Ini membuktikan bahwa tematikal gaib dalam Suling Dewa bersifat aspektual. Jika mengamati keseluruhan gending Suling Dewa dalam masyarakat Karang Bajo maka akan ditemui hal yang unik, yaitu 43 gending memiliki lirik dan satu gending bersifat instrumental (gending Bao Daya). Mengacu pada prinsip maki nini, Bao Daya merupakan gending bersifat maki sedang gending lainnya seperti Lokoq Sebia, Kamboja, Pem Pang Poq dan Lembuneng Meloang bersifat nini. Mengingat Suling Dewa merupakan musik ritual masyarakat Karang Bajo, maka pola persepsi dan pola berpikir 30 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Berikut adalah contoh gambar dari gending Suling Dewa dan tete. eni 2009:60). Jika pernyataan Burke dan Stets diurai maka akan mendapat gambaran sebagai berikut: Jurnal Kajian ao Daya Gending Bao Daya o Daya 43 gending Suling Dewa Gambar 5 Tete (Jembatan Tradisional Suk (Penulis, 2018) Gambar 5 Tete (Jembatan Tradisional Suku Sasak) (Penulis, 2018) Gending Bao Daya 43 Gending Suling Dewa 1. M a s y a r a k a t K a r a n g B a j o menyimbolkan dan memaknai Suling Dewa sebagai persepsi jembatan penghubung alam manusia dan alam ruh. 2. Simbol jembatan yang melekat pada Suling Dewa dimaknai seragam oleh seluruh kelompok masyarakat Karang Bajo. 3. Tindakan masyarakat Karang Bajo yang mensakralisasi Suling Dewa sehingga memandangnya sebagai subjek hidup. 4. Ide tentang Suling Dewa yang lahir akibat heterogenitas di dalam lingkungan Wet Bayan. 5. Seluruh ide di atas merupakan refleksi dan perasaan masyarakat Karang Bajo yang menjadi dasar tentang cara mereka mensakralkan Suling Dewa. Seperti menempatkannya ke dalam ritus-ritus tersakral. Gending B ao Daya Gending Bao Daya Simbolisasi Suling Dewa sebagai sebuah jembatan tidak hanya bersifat aspektual, tetapi juga bersifat material. KESIMPULAN Suling Dewa menjadi identitas simbolik masyarakat Karang Bajo dibuktikan melalui sistem maki nini dan corak Suling Dewa yang identik dengan dunia ruh serta menjadi manifestasi ideologi spiritualis kolektif Karang Bajo. Pelaku Suling Dewa menciptakan identitas simbolik kolektif Karang Bajo melalui simbol jembatan (tete) penghubung alam manusia dan alam gaib berdasarkan penerapan konsep maki nini. Mereka menggunakan Suling Dewa dalam ritus tersakral Islam Metu Telu karena sadar keterbatasan manusia untuk menjangkau hal-hal yang bersifat supranatural. Fenomena ini menjadi alasan yang kuat atas kehadiran Suling Dewa dalam ritus tersakral masyarakat Karang Bajo dan sekaligus menjadi identitas simbolik masyarakat Karang Bajo Bayan Lombok Utara. Pola masyarakat Karang Bajo dalam menciptakan mekanisme refleksi ideologi sebagai simbol-simbol dalam Suling Dewa menjadikan kesenian ini benar mampu menjadi identitas simbolik bagi masyarakat Karang Bajo di Wet Bayan. Hagel hingga G. H. Mead, mengatakan identitas terkait dengan pengakuan yang saling menguntungkan atau yang dikenal dengan istilah mutual recognition (Kellner:1995:80). Selaras dengan pernyataan kedua tokoh ini, dapat dibuktikan bahwa kehadiran masyarakat Karang Bajo dengan membentuk Suling Dewa sebagai identitas simboliknya juga turut serta Suling Dewa Sebagai Identitas Simbolik Jero Gamel Nyakranom dan Jero Gamel Anggalip mengatakan “Bao Daya no no ya wah kah penutuq ne, ya pengapit gending okon Suling Dewa” artinya gending Bao Daya itu adalah gending pengapit dan gending pengakhir dari Suling Dewa. Jika melihat 43 gending Suling Dewa sebagai tete (jembatan) dan gending Bao Daya sebagai pengapit, maka pola ini menggambarkan pola gambaran tete seperti konstruksi nyata. Ga gending B yang ber Dewa me 43 gending Suling De 43 Gending Suling Dewa Gambar 5 Tete (Jembatan Tradisional Su (Penulis 2018) Gambar 5 Tete (Jembatan Tradisional Suku Sasak) (Penulis, 2018) Gambar 5 Tete (Jembatan Tradisional Su (Penulis 2018) Gambar 5 Tete (Jembatan Tradisional Suku Sasak) (Penulis, 2018) bar di atas menjelaskan sisi kiri dan ka o Daya, sedangkan pijakan pada tete dis hir dengan ansteceden. Ini menunjukka pakan sebuah kesatuan gending yang sali di khi i d di B D Gambar 5 di atas menjelaskan sisi kiri dan kanan tete disimbolkan melalui gending Bao Daya, sedangkan pijakan pada tete disimbolkan melalui 43 gending yang berakhir dengan ansteceden. Ini 31 erpo Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 menunjukkan bahwa 43 gending Suling Dewa merupakan sebuah kesatuan gending yang saling topang- menopang dan isi-mengisi yang di akhiri dengan gending Bao Daya. Inan Gending sebagai vokalis tidak mampu menjadi jembatan karena dia berposisi sebagai manifestasi alam manusia. Sedangkan Suling Dewa secara terminologi merupakan suling makhluk halus dan manifestasi energi gaib sehingga disimbolkan sebagai jembatan yang mampu mengantarkan doa. Oleh sebab penjelasan inilah ritus tersakral milik masyarakat Karang Bajo tidak berani dilakukan tanpa kehadiran Suling Dewa, karena pada dasarnya upacara dan ritual selalu berhubungan dengan kegiatan manusia yang melakukan komunikasi dengan alam supranatural. Masyarakat menyadari kediriannya sebagai manusia, maka mereka membutuhkan media penyambung dunia mereka dan dunia supranatural yaitu Suling Dewa untuk mencapai sebuah kesempurnaan. menguntungkan kelompok masyarakat lainnya. Artinya di sini Wet Bayan semakin bergerak menemukan integritas empat pilar kelompok masyarakatnya. A Amaq : Ayah sebagai orang tua, dalam pandangan masyarakat Bayan Amaq juga merupakan gelar strata sosial yang lebih tinggi dari Raden. Amaq : Ayah sebagai orang tua, dalam pandangan masyarakat Bayan Amaq juga merupakan gelar strata sosial yang lebih tinggi dari Raden. DAFTAR PUSTAKA Heriyawati,Yanti, Seni Pertunjukan dan Ritual, Yogyakarta: Penerbit Ombak, 2016. Koentjaraningrat, Pengantar Ilmu Antropologi. Jakarta: Penerbit Aksara Baru, 1979. K. Langer, Suzzane, “Problematika Seni” Terjemahan F.X Widaryanto, Bandung: STSI Bandung, 2006. Stets, Burke and Jan E, Identity Theory, New Yok: Oxford University Press, 2009. 32 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara Thomson, Jhon B, Kritik Idieologi Global: Teori Sosial Kritis tentang Relasi Ideologi dan Komunikasi Massa, Yogyakarta: Ircisod, 2004. Kake Alam, 30 tahun, Tukang, Luk, Kecamatan Bayan, Lombok Utara. Kae Renadi, 28 tahun, Wiraswasta, Anyar, Kecamatan Bayan, Lombok Utara. Sumardi, Nur Kholis, Evolusi Gendang Beleq Lombok, dalam Gondang: Jurnal Seni dan Budaya Vol.1 No.2 Universitas Medan, 2017. Kake Sutyadi, 28 tahun, Petani, Kampu Karang Bajo, Kecamatan Bayan Kabupaten Lombok Utara. Ma Lokaq Pande, 80 tahun, Tetua Adat, Kampu Karang Bajo, Kecamatan Bayan, Kabupaten Lombok Utara. Jameson, Daphne, A., Reconceptualizing Cultural Identity and Its Role in Intercultural Business Communication d a l a m J o u r n a l B o u s i n e s s Communication Vol.44 Issue 3, Sage Publishing, 2007. Ma Lokaq Walin Gumi, 70 tahun, Tetua Adat, Trantapan, Kecamatan Bayan, Kabupaten Lombok Utara. Penghulu Adat Agung, 58 tahun, Tetua Adat, Bayan, Kecamatan Bayan, Kabupaten Lombok Utara. Kellner,Douglass Media Culture: Cultural Studies, Indentity and Politics betweenthe Modern and the Posmodern, London and New York: Routledge, 1995. Penganak, 90 tahun, Tetua Adat, Semokan, Sukadana, Kapubaten Lombok Utara Kholis, Muhammad Arsyad Nur, Suling Dewa dalam Ritual Mendewa Suku Sasak, Skripsi S-1 Etnomusikologi ISI Yogyakarta, Yogyakarta, 2017. Raden Jambe (Penghulu Raden), 45 Tahun, Pedagang, Kampu Timuq Orong, Bayan Beleq, Kecamatan Bayan, Kabupaten Lombok Utara Pip Jones, Pengantar Teori-Teori Social: Dari Teori Fungsionalisme Hingga Post Modernisme (trj) Saifuddin, Jakarta: Pustaka Obor, 2003. K Kamboja : Salah satu gending dalam Suling Dewa yang juga merujuk pada bunga kamboja. Bebet : Kain tenun yang dililit di pinggang seperti sebuah sabuk. Bilo: Bambu dalam bahasa Bayan. Bilo: Bambu dalam bahasa Bayan. D Daya : Kekuatan. Arti lain utara namun khusus dalam bahasa Sasak Kuto- kute diartikan selatan. Dodot : Pakaian adat Sasak, Bedodot mengenakan pakain adat. E Embok : Nafas G Gawah Sereru : Hutan Bambu Gending Suling Dewa : Komposisi Musik secara utuh yang disajikan dalam permainan Suling Dewa H Hikayat: Naskah lama Sasak yang bertuliskan Arab Melayu dan bercerita tentang perjalanan seseorang ataupun pituah-pituan nilai hidup. I Inan Gending : Gelar Vokalis Suling Dewa yang sudah melalui proses sakralisasi. Inaq : Bunda sebagai orang tua, dalam pandangan masyarakat Bayan Inaq juga merupakan gelar strata sosial yang lebih tinggi dari Raden dan istilah kebangsawanan wanita lainnya. Inaq Nini : Sebuah gelar yang dinobatkan kepada Vokalis Suling Dewa khusus di waktu prosesi ritus. I l M t T l S kt I l l k l S k Kampu : Wilayah adat yang menjadi pemukiman masyarakat adat. Daya : Kekuatan. Arti lain utara namun khusus dalam bahasa Sasak Kuto- kute diartikan selatan. Kedewayan : Kesurupan Kuto-kute : Sub Suku Sasak yang sebagian besar mendiami wilayah Lombok Utara. Arti lain begitu-begini, sub bahasa Lombok Utara. Dodot : Pakaian adat Sasak, Bedodot mengenakan pakain adat. L Lembuneng Meloang : Lebah Kayu (Xylocopa violacea) yang sedang membuat lubang. Salah satu gending dalam Suling Dewa. Lembuneng Meloang : Lebah Kayu (Xylocopa violacea) yang sedang membuat lubang. Salah satu gending dalam Suling Dewa. Loang : Lubang H Lokoq Sebia : Sungai Cabai. Salah satu gending dalam Suling Dewa. Lokoq Sebia : Sungai Cabai. Salah satu gending dalam Suling Dewa. NARASUMBER Inan Gending Inaq Mutringen, 90 tahun, Seniman, Senaru, Kecamatan Bayan, Kabupaten Lombok Utara. Amaq Maki : Sebuah gelar yang dinobatkan kepada pemain Suling Dewa khusus di waktu prosesi ritus. Amaq Maki : Sebuah gelar yang dinobatkan kepada pemain Suling Dewa khusus di waktu prosesi ritus. Jero Gamel Anggalip, 86 tahun, Seniman dan Peternak, Telaga Banyaq, Kecamatan Bayan, Kabupaten Lombok Utara. B Jero Gamel Nyakranom, 93 tahun, Seniman, Senaru, Kecamatan Bayan, Kabupaten Lombok Utara. Babad : Kumpulan tulisan cerita lampau Sasak yang terhimpun dan dibukukan dalam susunan daun lontar. Babad : Kumpulan tulisan cerita lampau Sasak yang terhimpun dan dibukukan dalam susunan daun lontar. Bao : Teduh 33 Jurnal Kajian Seni, Vol. 08, No. 01, November 2021: 15-35 Bao Daya : Teduh di utara, khusus Sasak Kuto-kute artinya teduh di selatan. Arti lain kekuatan meneduhkan atau menjadikan teduh. Salah satu gending dalam Suling Dewa. J Jero Gamel : Gelar Peniup Suling Dewa yang sudah melalui proses sakralisasi. N Selepir,Seliper,Seleper : Ikat dan pemecah udara pada organologi Suling Dewa. Selepir,Seliper,Seleper : Ikat dan pemecah udara pada organologi Suling Dewa. Ndeq Ku Dayaq : Aku tidak berdaya, aku tidak mampu, aku tidak memiliki kekuatan. Sembeq : Berkat yang diberikan antara subjek-objek dan atau subjek-subjek. Sembeq : Berkat yang diberikan antara subjek-objek dan atau subjek-subjek. Slewoq : Kain tenun lembaran Sasak yang digunakan menupi pinggang hingga kaki. Neneq: Tuhan Neneq: Tuhan Nini : Sifat feminim dalam sistem dualisme Sasak. Nini : Sifat feminim dalam sistem dualisme Sasak. Suling: Seruling. P Pancadewata : Sistem arah, 5 titik dan 5 warna dalam Hindu. Pancadewata : Sistem arah, 5 titik dan 5 warna dalam Hindu. Suling Dewa : Suling sakral khas Bayan. Arti lain Seruling energi alam. Suling Dewa : Suling sakral khas Bayan. Arti lain Seruling energi alam. Pem Pang Poq : Cabang ranting pohon mangga. Salah satu gending Suling Dewa. T Tete: Jembatan khas Sasak. W Wet : Wilayah, wilayah adat. Wet Bayan : Wilayah adat masyarakat Bayan T Tete: Jembatan khas Sasak. W Wet : Wilayah, wilayah adat. Wet Bayan : Wilayah adat masyarakat Bayan M Maki : Sifat maskulin dalam sistem dualisme Sasak. Maki : Sifat maskulin dalam sistem dualisme Sasak. I Mendewa : Ritus dan atau melakukan ritus pemanggilan, pengusiran dan pengobatan yang berkaitan dengan mahluk supranatural. Inan Gending : Gelar Vokalis Suling Dewa yang sudah melalui proses sakralisasi. Inaq : Bunda sebagai orang tua, dalam pandangan masyarakat Bayan Inaq juga merupakan gelar strata sosial yang lebih tinggi dari Raden dan istilah kebangsawanan wanita lainnya. Mesigit Lokaq : Masjid Kuno Mesigit Lokaq : Masjid Kuno Maq Lokaq Pande : Gelar status petinggi adat Wet Bayan yang bertugas dalam hal alam ghaib. Maq Lokaq Walin Gumi : Gelar status petinggi adat Wet Bayan yang bertugas dalam hal alam dunia. Maq Lokaq Walin Gumi : Gelar status petinggi adat Wet Bayan yang bertugas dalam hal alam dunia. Inaq Nini : Sebuah gelar yang dinobatkan kepada Vokalis Suling Dewa khusus di waktu prosesi ritus. Islam Metu Telu : Sekte Islam lokal Suku Sasak. 34 M.A Nur Kholis, Wahyu Kurnia, Suling Dewa sebagai Identitas Simbolik Masyarakat Sasak Kuto-Kute di Karang Bajo Bayan Lombok Utara N Ndeq Ku Dayaq : Aku tidak berdaya, aku tidak mampu, aku tidak memiliki kekuatan. Neneq: Tuhan Nini : Sifat feminim dalam sistem dualisme Sasak. P Pancadewata : Sistem arah, 5 titik dan 5 warna dalam Hindu. Pem Pang Poq : Cabang ranting pohon mangga. Salah satu gending Suling Dewa. Penganak : Salah satu gelar petinggi adat Kampu Semokan. Pengosap : Ornamentasi kain pada Suling Dewa. Poto : Ujung S Sasak Kuto-kute : Sub Sasak yang berbahasa Kuto-kute dan dominan berada di wilayah Lombok Utara. Sapuq : Ikat kepala Sasak, yang dibentuk menggunakan kain, untuk menutupi kepala. T Penganak : Salah satu gelar petinggi adat Kampu Semokan. Pengosap : Ornamentasi kain pada Suling Dewa. Y Yak : Nafas yang berhembus Y Yak : Nafas yang berhembus S S Y Sasak Kuto-kute : Sub Sasak yang berbahasa Kuto-kute dan dominan berada di wilayah Lombok Utara. Sasak Kuto-kute : Sub Sasak yang berbahasa Kuto-kute dan dominan berada di wilayah Lombok Utara. Sapuq : Ikat kepala Sasak, yang dibentuk menggunakan kain, untuk menutupi kepala. Sapuq : Ikat kepala Sasak, yang dibentuk menggunakan kain, untuk menutupi kepala. 35 35
https://openalex.org/W3194143449	https://link.springer.com/content/pdf/10.1007/s00120-021-01620-7.pdf	German	null	Einflussfaktoren bei der Wahl der Androgendeprivationstherapie für Patienten mit hormonsensitiven Prostatakarzinom	Der Urologe	2,021	cc-by	4,958	g Urologe 2022 · 61:173–182 https://doi.org/10.1007/s00120-021-01620-7 Angenommen: 13. Juli 2021 Online publiziert: 17. August 2021 © Der/die Autor(en) 2021 Einﬂussfaktoren bei der Wahl der Androgendeprivationstherapie für Patienten mit hormonsensitiven Prostatakarzinom Ergebnisse der ProComD-Studie J. Lehmann1 · C. W. Kluike2 · A. Haider3 · K. S Haider3 · S. Baumann4 · M. Flesch5 · M. Gedamke6 · D. Kägebein7 1Urologische Gemeinschaftspraxis Prüner Gang, Gesundheitszentrum Kiel-Mitte, Kiel, Deutschland 2Urologie am Wasserturm, Lüneburg, Deutschland 3Praxis für Urologie und Andrologie, Bremerhaven, Deutschland 4Praxisgemeinschaft für Urologie, Leipzig, Deutschland 5Marienkrankenhaus, Soest, Deutschland 6msc, Kiel, Deutschland 7Ferring Arzneimittel GmbH, Kiel, Deutschland Urologe 2022 · 61:173–182 https://doi.org/10.1007/s00120-021-01620-7 Angenommen: 13. Juli 2021 Online publiziert: 17. August 2021 © Der/die Autor(en) 2021 Ergebnisse der ProComD-Studie J. Lehmann1 · C. W. Kluike2 · A. Haider3 · K. S Haider3 · S. Baumann4 · M. Flesch5 · M. Gedamke6 · D. Kägebein7 1Urologische Gemeinschaftspraxis Prüner Gang, Gesundheitszentrum Kiel-Mitte, Kiel, Deutschland 2Urologie am Wasserturm, Lüneburg, Deutschland 3Praxis für Urologie und Andrologie, Bremerhaven, Deutschland 4Praxisgemeinschaft für Urologie, Leipzig, Deutschland 5Marienkrankenhaus, Soest, Deutschland 6msc, Kiel, Deutschland 7Ferring Arzneimittel GmbH, Kiel, Deutschland DieAndrogendeprivationstherapie(ADT) mit GnRH(Gonadotropin-Releasinghor- mon)-Agonisten oder dem GnRH-Anta- gonisten Degarelix ist seit Jahrzehnten der Behandlungsstandard beim fortge- schrittenen hormonnaiven Prostatakar- zinom (PCa). Entscheidungskriterien für den Einsatz einer der beiden Substanz- klassen wurden bislang nicht identiﬁ- ziert. Diese Studie gibt Hinweise, wie Ba- sischarakteristika und Komorbiditäten der PCa-Patienten die Therapieentschei- dung beeinﬂussen. 17]. PCa-Patienten weisen aufgrund ihres meist höheren Alters häuﬁg Komorbidi- täten auf – insbesondere ﬁnden sich oft kardiovaskuläre Vorerkrankungen in ihrer Anamnese [5]. Randomisierte klinische Studien, Me- taanalysen und Real-world-Evidenzstudi- en haben gezeigt, dass das Risiko kardio- vaskulärer Ereignisse unter einer ADT bei Patienten, die mit einem GnRH-Antago- nisten behandelt werden, im Vergleich zu den mit GnRH-Agonisten behandelten Pa- tienten reduziert ist [1, 3, 8, 12, 16, 20, 21]. Dies impliziert ein unterschiedliches Risiko für kardiovaskuläre Ereignisse in Ab- hängigkeit der ADT mit GnRH-Agonisten oder -Antagonisten [18]. Der Urologe 2 · 2022 173 Originalien Originalien Einﬂussfaktoren bei der Wahl der Androgendeprivationstherapie für Patienten mit hormonsensitiven Prostatakarzinom Ergebnisse der ProComD-Studie Hintergrund In der Therapie des hormonnaiven PCa wird der medikamentöse Androgenent- zug in der Regel entweder durch die Be- handlung mit GnRH-Agonisten oder -Ant- agonisten erreicht [25]. Das Outcome der ADT wird stark von Komorbiditäten beein- ﬂusst, und Beobachtungen deuten darauf hin, dass Patienten, die sich einer ADT un- terziehen, einem erhöhten Risiko für be- handlungsbedingte unerwünschte Ereig- nisse ausgesetzt sein können, insbeson- dere für kardiovaskuläre Ereignisse [7, 13, Die Kenntnis über bestehende Komor- biditäten der Patienten zum Zeitpunkt der Therapieentscheidung kann deshalb von wesentlicher Bedeutung für die eﬀektive undsichereBehandlungderPCa-Patienten sein. Voraussetzung für die Kenntnis über Komorbiditäten ist eine hohe Qualität der Anamnese. DieQualitätder Anamnese, die Entscheidungskriterien für eine Therapie mit GnRH-Agonisten oder Degarelix sowie das Ausmaß der interdisziplinären Zusam- QR-Codescannen&Beitragonlinelesen Der Urologe 2 · 2022 173 Originalien Zusammenfassung Hintergrund: Die Androgendeprivationstherapie (ADT) mit einem GnRH-Agonisten (Gonadotropin-releasing-Hormon) oder -Antagonisten stellt einen zentralen Bestandteil der Behandlung des Prostatakarzinoms (PCa) dar. Über die Faktoren, Hintergrund: Die Androgendeprivationstherapie (ADT) mit einem GnRH-Agonisten (Gonadotropin-releasing-Hormon) oder -Antagonisten stellt einen zentralen Bestandteil der Behandlung des Prostatakarzinoms (PCa) dar. Über die Faktoren, welche die Wahl der ADT beeinﬂussen, ist bis jetzt wenig bekannt. menarbeit während der PCa-Behandlung sind jedoch bisher nicht bekannt. menarbeit während der PCa-Behandlung sind jedoch bisher nicht bekannt. Schlüsselwörter Schlüsselwörter Schlüsselwörter Degarelix · GnRH-Antagonist · Therapieentscheidung · Komorbiditäten · Behandlungspfade Degarelix wurde als 1-Monats-Formu- lierung mit einer Startdosis von 240mg und einer Erhaltungsdosis von 80mg ver- abreicht.AlsGnRH-AgonistenwurdenLeu- prorelinacetat, Goserelinacetat, Busereli- nacetat sowie Triptorelinacetat (jeweils als 1-, 3- oder 6-Monats-Fomulierung) einge- setzt. Diagnostik, Medikation und Patien- tenüberwachung lagen ausschließlich im Ermessen des Arztes entsprechend seiner üblichen Behandlungspraxis. Die Visiten einschließlich ihrer Dokumentation waren Teil der klinischen Routine und fanden nichtzuimVorausdeﬁniertenZeitpunkten statt. wie Risikofaktoren zum Zeitpunkt der Ent- scheidungsﬁndung sowie die hierfür her- angezogenen Informationsquellen. 2) Ba- sierend auf 1) sollte festgestellt werden, wie diese Kenntnisse die behandelnden Ärzte in ihrer Therapieentscheidung be- einﬂusst haben. Ein weiteres Ziel war die Bestimmung der Häuﬁgkeiten verschiede- ner Komorbiditäten zu Therapiebeginn. Methodik p Die häuﬁgste konsultierte Informationsquelle bzgl. vorhandener Komorbiditäten ist die Anamnese durch den behandelnden Urologen selbst (68,5% in beiden Gruppen). Bei Patienten mit kardiovaskulären Vorerkrankungen wurde zusätzlich der Arztbrief (45,8% Degarelix vs. 38,9% GnRH-Agonisten) oder der Anamnese-Fragebogen (38,9% Degarelix vs. 20% GnRH-Agonisten) herangezogen. Patienten und Behandlung Patienten mit hormonnaivem PCa (lokal, lokal fortgeschritten oder metastasiert) und der Indikation einer ADT wurden nach der Therapieentscheidung des Arz- tes in die Studie eingeschlossen. Nicht erlaubt war der Einschluss von Patienten, die eine andere hormonelle Behandlung ihres PCa erhielten. Ausgenommen hier- von waren Patienten mit bereits erfolgter hormoneller neoadjuvanter oder adjuvan- ter Behandlung im Rahmen einer kurativ intendierten Primärtherapie, wenn diese Behandlung nicht länger als 6 Mona- te durchgeführt wurde und mindestens 6 Monate vor Studieneinschluss abge- schlossen war. Eine Mindestdauer der ADT war nicht vorgegeben. Ziel der Studie j g Ziele der Arbeit: Faktoren, welche die Wahl der ADT bei Patienten mit hormonsensi- tivem PCa beeinﬂussen, werden identiﬁziert. Vom Urologen zur Identiﬁzierung von Begleiterkrankungen genutzte Informationsquellen sowie deren Prävalenzen werden bestimmt. Ziel der Studie war es, Faktoren zu identiﬁ- zieren, welche bei Patienten mit hormon- naivem PCa die Therapieentscheidung bzgl. der medikamentösen ADT (Degare- lix oder GnRH-Agonisten) beeinﬂussen. Darüber hinaus wurden die Häuﬁgkeit von Begleiterkrankungen und die von den Urologen hierfür herangezogenen Informationsquellen bestimmt. Methoden: Die zweiarmige, prospektive, nicht-interventionelle Studie „ProComD“ wurde von Sept. 2014 bis Juni 2019 an 80 Studienzentren in Deutschland durchgeführt. Patienten mit hormonnaivem PCa und Notwendigkeit einer ADT wurden nach erfolgter Therapieentscheidung in die Studie eingeschlossen. Fragen bezüglich Informationsquelle und Therapieentscheidung wurden vom Arzt direkt im elektronischen Datenerfassungssystem (eCRF) beantwortet. Ergebnisse: Es wurden Daten von 413 Patienten ausgewertet (Degarelix n= 268; GnRH- Agonisten n= 145). Ausschlaggebend für die Therapieentscheidung waren für beide Behandlungsgruppen u.a. die Faktoren Komorbiditäten (bei 42% aller Patienten), Compliance (64%) und Alter (81%). Datenerhebung Die Basischarakteristika wurden in Routi- nevisite 1 vom behandelnden Arzt erfragt oder aus der Krankenakte entnommen. Folgedaten in den weiteren Routinevisiten wurden alle 3 (± 4 Wochen) Monate für das erste Jahr und alle 6 Monate (± 6 Wo- chen) für die folgenden 3 Jahre erhoben. Die Daten wurden in ein elektronisches Datenerfassungssystem übertragen. Studiendesign Die zweiarmige, prospektive, nicht-inter- ventionelle Studie (NIS) ProComD wur- de vom 17. September 2014 bis 30. Ju- ni 2019 an 80 Studienpraxen in Deutsch- land durchgeführt. Die Patienten wurden mit der Verschreibung einer medikamen- tösen ADT (Degarelix oder GnRH-Agonist) unmittelbar in die Studie aufgenommen und bis zu 48 Monate lang nachbeobach- tet. Die NIS wurde von der Ethikkommissi- onderLandesärztekammerNordrheinhin- sichtlich berufsethischer und -rechtlicher Gesichtspunkte begutachtet (Berufsrecht- liche Beratung nach § 15 der Berufsord- nung). Die Patienten gaben nach Aufklä- rung schriftlich ihr Einverständnis, bevor sie an der Studie teilnahmen. Schlussfolgerungen: Komorbiditäten zählen neben dem Alter und der Compliance zu den wichtigen Faktoren, die die Wahl der ADT beeinﬂussen. Ziele Primäres Ziel der Studie war es, Faktoren zu identiﬁzieren, die zur Entscheidung für eine medikamentöse ADT bei Patienten mit hormonnaivem PCa mit und ohne Ko- morbiditäten beitragen. Folgende Aspek- te wurden dabei insbesondere bewertet: 1)KenntnissedesUrologeninBezugaufdie Krankengeschichte der Patienten, Begleit- erkrankungen und Begleitmedikation so- Fragen bezüglich der konsultierten In- formationsquellen und der Faktoren, die die Therapieentscheidung beeinﬂussten, wurden vom Arzt ebenfalls direkt im elek- tronischen Datenerfassungssystem (eCRF) innerhalb eines Fragebogens beantwor- 174 Der Urologe 2 · 2022 Hier steht eine Anzeige. K Hier steht eine Anzeige. Originalien Tab. 1 Demographische Daten undandere Basischarakteristika Degarelix (n= 268) GnRH-Agonisten (n= 145) Gesamt (n= 413) Alter (Jahre) Mittelwert±SD 73,6± 8,3 74,9± 8,35 74,0± 8,34 Median 75 76 76 Größe (m) Mittelwert±SD 1,75± 0,07 1,74± 0,07 1,75± 0,07 Median 1,75 1,74 1,75 Gewicht (kg) Mittelwert±SD 82,9± 13,1 83,0± 15,8 82,9± 82,9 Median 82,0 82,0 82,0 Body Mass Index (BMI) Mittelwert±SD 27,2± 4,0 27,2± 4,3 27,2± 4,1 Median 26,7 26,5 26,6 Testosteron (ng/dl) Mittelwert±SD 385,4 359,2 378,1 Median 317,3 330,2 320,0 PSA (ng/ml) Mittelwert±SD 129,0± 382,8 111,6± 309,5 122,7± 357,8 Median 21,1 9,8 15,3 Range 0–4565 0–2289 0–4565 PSA-Untergruppen <10ng/ml 91 (34,0%) 69 (47,6%) 160 (38,7%) 10–20ng/ml 28 (10,4%) 16 (11,0%) 44 (10,7%) <20–50ng/ml 43 (16,0%) 16 (11,0%) 59 (14,3%) >50ng/ml 82 (30,6%) 36 (24,8%) 118 (28,6%) Gleason-Score 2–4 1 (0,4%) 1 (0,7%) 2 (0,5%) 5–6 38 (14,2%) 17 (11,7%) 55 (13,3%) 7 75 (28,0%) 50 (34,5%) 125 (30,3%) 8–10 139 (51,9%) 71 (49,0%) 210 (50,8%) Keine Angabe 15 (5,6%) 6 (4,1%) 21 (5,1%) Mittelwert±SD 7,8± 1,2 7,7± 1,2 7,7± 1,2 Median 8,0 8,0 8,0 Krankheitsstadium Lokal begrenzt (T1–2, N0/N x, M0) 50 (18,7%) 46 (31,7%) 96 (23,2%) Lokal fortgeschritten (T3–4, N0/N x, M0) 29 (10,8%) 13 (9,0%) 42 (10,2%) Fortgeschritten bzw. metastasiert (N1 und/oder M1) 92 (34,3%) 40 (27,6%)a 132 (32,0%) Nicht kategorisierbar (Tx und/oder Mx) 97 (36,2%) 46 (31,7%) 143 (34,6%) ap< 0,0001, p-Werte basieren auf einem Binomialtest geführt. Diesebestand aus allenPatienten, die für die jeweilige Behandlung einge- teilt wurden. Darüber hinaus musste die erste Routinevisite sowie mindestens ei- ne weitere Follow-up-Visite dokumentiert sein. Patienten, die die ADT-Behandlung abbrachen, wurden für die Safety-Analy- sen weiterhin untersucht. Soweit Vergleiche zwischen Behand- lungsgruppendurchführtwurden,wurden diese als adjustierte Unterschiede (metri- sche Variablen) oder Raten (kategorische Variablen), jeweils mit einem Konﬁdenzin- tervallvon95%,dargestellt.DieVergleiche wurden hinsichtlich Unterschiede in den Baselineparameter mittels einer Regressi- onsanalyse adjustiert. Ziele Um Korrelationen zwischen der Wahl der Therapie und vorbestehenden oder neuen Komorbiditäten zu untersuchen, wurde ein χ2-Test verwendet. Beim Auf- treten von kleinen erwarteten Frequenz- werten wurde ein exakter Fisher-Test (ggf. unter Verwendung einer Monte-Carlo-An- näherung) angewandt. Um Unterschiede zwischen den Behandlungsgruppen zu untersuchen, wurde für metrische Varia- blen ein Student’s t-Test oder Wilcoxon- Rangsummentest und für kategoriale Variablen ein χ2- (>2 Kategorien) oder ein Binomialtest angewendet. Alle ab- geleiteten p-Werte sind deskriptiv zu interpretieren. Das Signiﬁkanzniveau wur- de für alle angewandten Testverfahren auf 5% festgelegt. Die Studie benötigte keine bestätigen- den statistischen Tests, da sie dazu ange- legt war, die Qualität der Anamnese und die Eﬀektivität und Sicherheit der Behand- lung mit Degarelix und GnRH-Agonist zu berichten. Alle untersuchten Hypothesen waren zweiseitig. Für ordinalskalierte Kor- relationen wurde der Spearman-Koeﬃzi- ent verwendet, für metrische Korrelatio- nen der Pearson-Koeﬃzient. tet. Begleiterkrankungen wurden gemäß MedDRA (Vers. 19.0) und Begleitmedika- tion nach ATC-Code kodiert. Komorbiditäten, Risikofaktoren, Begleitmedikation Komorbiditäten, Risikofaktoren, Begleitmedikation Insgesamt wurde bei 75,5% (312/413) aller Patienten vom Urologen beim Ba- sisbesuch mindestens eine Komorbidität dokumentiert,miteinemsigniﬁkantenUn- terschied zwischen der Degarelix-Gruppe (76,5%, 205/268) und der GnRH-Agonis- tengruppe (73,8%, 107/145; p< 0,0001). Die zu Studienbeginn am häuﬁgsten dokumentierten Begleiterkrankungen (Kodierung gemäß MedDRA) waren für beide Behandlungsgruppen vaskuläre Er- krankungen (einschließlich Hypertonie; 60,3%, 249/413), gefolgt von Stoﬀwech- sel- und Ernährungserkrankungen(37,0%, 153/413) und kardialen Erkrankungen (24,2%, 100/413), wobei in der Degarelix- Gruppe der Anteil der Patienten signi- ﬁkant höher (p< 0,0001) war als in der GnRH-Agonistengruppe (. Abb. 1). Tab. 3 Risikofaktoren Degarelix (n= 268) GnRH-Agonisten (n= 145) Gesamt (n= 413) Kardiovaskuläre Risikofaktoren 192 (71,6%) 100 (69,0%)c 292 (70,7%) Hypertonie 159 (59,3%) 78 (53,8%)c 237 (57,4%) Diabetes mellitus 59 (22,0%) 33 (22,8%)a 92 (22,3%) Hyperlipidämie 64 (23,9%) 28 (19,3%)b 92 (22,3%) Kardiovaskuläre Vorerkrankung 90 (33,6%) 43 (29,7%)a 133 (32,2%) Herzinsuﬃzienz 43 (16,0%) 16 (11,0%)b 59 (14,3%) Periphere arterielle Verschlusskrankheit 33 (12,3%) 8 (5,5%)c 41 (9,9%) Koronare Herzkrankheit 8 (3,0%) 5 (3,4%) 13 (3,1%) Schweres kardiovaskuläres Ereignis 40 (14,9%) 22 (15,2%)a 62 (15,0%) Myokardinfarkt 23 (8,6%) 11 (7,6%)a 34 (8,2%) Zerebrovaskuläres Ereignis 19 (7,1%) 11 (7,6%) 30 (7,3%) Chronische Nierenerkrankung 32 (11,9%) 6 (4,1%)c 38 (9,2%) Chronische Lungenerkrankung 21 (7,8%) 6 (4,1%)a 27 (6,5%) Familiäre Belastung Kardiovaskuläre Erkrankung bei Ver- wandten 1. Grades 50 (18,7%) 26 (17,9%)b 76 (18,4%) Diabetes mellitus bei Verwandten 1. Grades 28 (10,4%) 20 (13,8%) 48 (11,6%) Lebensstil BMI >30 51 (19,0%) 24 (16,6%)a 75 (18,2%) Alkoholmissbrauch 5 (1,9%) 3 (2,1%) 8 (1,9%) Raucher 30 (11,2%) 10 (6,9%)a 40 (9,7%) Ehemaliger Raucher 29 (10,8%) 23 (15,9%) 52 (12,6%) Mehrfachnennungen möglich ap< 0,05, bp< 0,001, cp< 0,0001, p-Werte basieren auf einem Binomialtest Kardiovaskuläre Vorerkrankungen la- gen bei 32,2% (133/413) aller Patienten vor. Herzinsuﬃzienz (16,0 vs. 11,0%, p< 0,001), periphere arterielle Verschluss- krankheit (pAVK; 12,3 vs. 5,5%, p<0,0001) in der Anamnese wurden in der Degarelix- Gruppe signiﬁkant häuﬁger dokumentiert als in der GnRH-Agonistengruppe. Bei Behandlungsbeginn hatten bereits 15,0% der Patienten ein schwerwiegen- des kardiovaskuläres Ereignis in ihrer Ana- mnese. 8,6% der Patienten in der De- garelix-Gruppe hatten bereits einen Myo- kardinfarkt erlitten im Vergleich zu 7,6% in der GnRH-Agonistengruppe (p< 0,05; . Tab. 3). Das Krankheitsstadium bei Einschluss in die Studie war signiﬁkant unterschied- lich zwischen den beiden Behandlungs- gruppen (p= <0,0001). Subgruppen Es konnten die Daten von 413 Pati- enten ausgewertet werden (Degarelix n= 268; GnRH-Agonisten n= 145). Fol- gende GnRH-Agonisten wurden zur Be- handlung eingesetzt: Buserelin (11,1%; Statistische Analysen wurden mittels SAS, Version 9.3 (SAS Institute Inc., Cary, NC, USA)durchgeführt.DieﬁnalenAuswertun- gen wurden für die FAS-Population durch- Folgende Subgruppen wurden deﬁniert: Patienten mit/ohne Metastasen, Patienten miteinem Basis-PSA-Wert0–50ng/ml und >50ng/ml und Patienten ohne/mit kar- 176 Der Urologe 2 · 2022 Tab. 2 Subgruppen Subgruppe Degarelix (n= 268) GnRH-Agonisten (n= 145) Gesamt (n= 413) Metastasen M+ 99 (36,9%) 36 (24,8%)a 135 (32,7%) PSA >50ng/ml 82 (30,6%) 36 (24,8%)a 118 (28,6%) Kardiovaskuläre Vorerkrankung(en) (inklu- sive Hypertonie) 179 (66,8%) 90 (62,1%)a 269 (65,1%) ap< 0,0001, p-Werte basieren auf einem Binomialtest 24,8%) und „Patienten mit Metastasen“ (36,9 vs. 24,8%) jeweils signiﬁkant höher (p< 0,0001; . Tab. 2). Der Urologe 2 · 2022 177 Komorbiditäten, Risikofaktoren, Begleitmedikation In der Degarelix- Gruppe war das metastasierte PCa das am häuﬁgsten dokumentierte Krankheitssta- dium (34,3%), während in der GnRH-Ago- nistengruppe das lokal begrenzte PCa am häuﬁgsten dokumentiert wurde (31,7%; . Tab. 1). n= 46), Leuprorelin (19,4%; n= 80), Trip- torelin (4,4%; n= 18) und Goserelin (0,2%; n= 1). Es wurden keine statistisch signiﬁ- kantenundklinischrelevantenUnterschie- de zwischen den Behandlungsgruppen hinsichtlichder Demographieund anderer Basischarakteristika beobachtet (. Tab. 1). n= 46), Leuprorelin (19,4%; n= 80), Trip- torelin (4,4%; n= 18) und Goserelin (0,2%; n= 1). Es wurden keine statistisch signiﬁ- kantenundklinischrelevantenUnterschie- de zwischen den Behandlungsgruppen hinsichtlichder Demographieund anderer Basischarakteristika beobachtet (. Tab. 1). Für 407 Patienten waren Daten zur Ge- samtdauer der ADT innerhalb der Studie vorhanden. Die durchschnittliche Dauer der ADT betrugfür dieDegarelix-Patienten 22,4 Monate (SD= 13,7) und für die Pati- enten, die mit einem Agonisten behandelt wurden 27,0 Monate (SD= 13,9). n= 46), Leuprorelin (19,4%; n= 80), Trip- torelin (4,4%; n= 18) und Goserelin (0,2%; n= 1). Es wurden keine statistisch signiﬁ- kantenundklinischrelevantenUnterschie- de zwischen den Behandlungsgruppen hinsichtlichder Demographieund anderer Basischarakteristika beobachtet (. Tab. 1). Für 407 Patienten waren Daten zur Ge- samtdauer der ADT innerhalb der Studie vorhanden. Die durchschnittliche Dauer der ADT betrugfür dieDegarelix-Patienten 22,4 Monate (SD= 13,7) und für die Pati- enten, die mit einem Agonisten behandelt wurden 27,0 Monate (SD= 13,9). Kardiovaskuläre Risikofaktoren hatten 70,7% aller Patienten. In der Degarelix- Gruppe wurden im Vergleich zur GnRH- Agonistengruppe die anamnestischen Ri- sikofaktoren Hypertonie (59,3 vs. 53,8%, p< 0,0001) und Hyperlipidämie (23,9 vs. 19,3%,p= 0,0002)signiﬁkanthäuﬁgerdo- kumentiert (. Tab. 3). Für 407 Patienten waren Daten zur Ge- samtdauer der ADT innerhalb der Studie vorhanden. Die durchschnittliche Dauer der ADT betrugfür dieDegarelix-Patienten 22,4 Monate (SD= 13,7) und für die Pati- enten, die mit einem Agonisten behandelt wurden 27,0 Monate (SD= 13,9). Für 81% (334/413) der Patienten wur- de die Begleitmedikation dokumentiert. Die Ergebnisse zeigten einen Unterschied von mindestens 5% zwischen den Be- handlungsgruppen für „Antiandrogene“ (13,4% Degarelix vs. 39,3% GnRH-Ago- Der prozentuale Anteil der Degare- lix-Patienten war in den Subgruppen „Patienten mit kardiovaskulären Vorer- krankungen“ (66,8 vs. 62,1%), „Patienten mit Baseline-PSA >50ng/ml“ (30,6 vs. Der Urologe 2 · 2022 177 Der Urologe 2 · 2022 177 Originalien Abb. 1 8 Begleiterkrankungen zu Studienbeginn nach MedDRA(häuﬁgste Erkrankungen,Mehrfachnennung möglich, *p< 0,0001,p-Werte basieren auf einemBinomialtest) Abb. 2 8 Informationsquelle überdie Begleiterkrankungen (Mehrfachantworten möglich) Originalien Abb. 1 8 Begleiterkrankungen zu Studienbeginn nach MedDRA(häuﬁgste Erkrankungen,Mehrfachnennung möglich, p< 0,0001,p-Werte basieren auf einemBinomialtest) Abb. Komorbiditäten, Risikofaktoren, Begleitmedikation 1 8 Begleiterkrankungen zu Studienbeginn nach MedDRA(häuﬁgste Erkrankungen,Mehrfachnennung möglich, **p< 0,0001,p-Werte basieren auf einemBinomialtest) Abb. 2 8 Informationsquelle überdie Begleiterkrankungen (Mehrfachantworten möglich) Abb. 2 8 Informationsquelle überdie Begleiterkrankungen (Mehrfachantworten möglich) 178 Der Urologe 2 · 2022 Abb. 3 8 Patientenindividuelle Faktoren, die dieTherapieentscheidung beeinﬂussten Abb. 3 8 Patientenindividuelle Faktoren, die dieTherapieentscheidung beeinﬂussten Insgesamt wurden 117 UE während der Behandlungszeitgemeldet, vondenen 95 auf die ADT zurückgeführt wurden. Die Rate aller UE betrug 32,1% (n= 86) in der Degarelix-Gruppe und 21,7% (n= 31) in der GnRH-Agonistengruppe. Zu den häu- ﬁgsten UE, die auf die ADT zurückgeführt wurden, zählten Hitzewallungen (Degare- lix 9,3% vs. GnRH-Agonisten 15,2%) und Reaktionen an der Injektionsstelle (De- garelix 11,6% vs. GnRH-Agonisten 0%). Die Rate schwerwiegender UE betrug 2,2% (n= 6) in der Degarelix-Gruppe und 0% (n= 0) in der GnRH-Agonistengrup- pe. Folgende schwerwiegende UE traten unter Degarelix-Behandlung auf (Deﬁni- tion nach MedDRA PT): „Apathy, General physical health deterioration, Chills (2×), Supraventricular tachycardia, Anemia“. nisten; p= 0,0000), „Antihypertensiva“ (56,0% Degarelix vs. 49,0% Agonis- ten) und „Antithrombotika/Herztherapie“ (30,2% Degarelix vs. 24,8% Agonisten). war. Ausschlaggebend für die Therapie- entscheidung waren für beide Behand- lungsgruppen die Faktoren Compliance (64%), Komorbiditäten (42%) und Al- ter (81%), wobei die Komorbiditäten bei den Degarelix-behandelten Patienten einen prozentual höheren Einﬂuss auf die Therapieentscheidung hatten, als bei Patienten, die mit einem GnRH-Agonis- ten behandelt wurden (45,5 vs. 35,9%, p= 0,0577). Von geringerer Bedeutung für die Therapieentscheidung waren Arznei- mittelunverträglichkeiten (Degarelix 9,0% vs. GnRH-Agonisten 6,2%) oder Wechsel- wirkungen mit anderen Medikamenten (6,3 vs. 5,5%) (. Abb. 3). Die Analyse der Subgruppen ergab keine statistisch signiﬁkanten Unterschiede. Informationsquelle Die Anamnese durch den Urologen war mit 68,5% die Hauptinformationsquelle für Komorbiditäten mit geringem Unter- schied zwischen der Degarelix-Gruppe (69,8%, 187/268) und der GnRH-Agonis- tengruppe (66,2%, 96/145). Bei Patienten mit kardiovaskulären Vorerkrankungen dienten zusätzlich auch der Anamnese- fragebogen (Degarelix 38,9% vs. GnRH- Agonisten 20,0%) sowie externe Arztbrie- fe (Degarelix 45,8% vs. GnRH-Agonisten 38,9%) als Informationsquelle (. Abb. 2). Diskussion Nach den neuesten EAU-Guidelines übersteigt die kardiovas- kuläre Mortalität sogar den Prostatakrebs als häuﬁgste Todesursache [19]. Eine be- stehende kardiovaskuläre Erkrankung ist demnach eine der wichtigsten Herausfor- derungen bei der Behandlung des PCa. Bei der Wahl der ADT sollte die Entscheidung daher auf Präparate fallen, die mit einem geringeren Risiko in Verbindung stehen, diese Begleiterkrankungen zu verstärken [18]. In einem Kontext, in dem ein besse- res Gesamtüberleben ohne Verschlechte- rung der Lebensqualität eine Herausforde- rung darstellt, sind die potenziellen kar- diovaskulären Nebenwirkungen der ADT von entscheidender Bedeutung. Um die behandlungsbedingten Nebenwirkungen einer ADT patientenindividuell besser ein- schätzen und bewältigen zu können, müs- sen Ärzte und Patienten in ständigem Dia- log stehen und regelmäßig Informationen austauschen. Da sich in der ProComD-Stu- dieKomorbiditätenals wichtiges Entschei- dungskriterium für die ADT herauskristalli- sierthabenundDegarelixbeiPatientenmit kardiovaskulären Vorerkrankungen bevor- zugt eingesetzt wurde, besteht die Annah- me, dass die bestehenden Kenntnisse den Behandlungsalltag bereits beeinﬂussen. Eine große Beobachtungsstudie aus Italien mit 9785 Patienten kommt zu einem ähnlichen Ergebnis [20]. Die In- zidenz kardiovaskulärer Ereignisse war signiﬁkant höher bei Patienten, die mit GnRH-Agonisten statt mit Degarelix be- handelt wurden (8,8 vs. 6,2, p= 0,002). Alter, Bluthochdruck, Dyslipidämie, beste- hende kardiovaskuläre Erkrankung und frühere kardiovaskuläre Ereignisse stell- ten unabhängige Risikofaktoren für ein neues kardiovaskuläres Ereignis dar und wurden in der multivariaten Analyse be- rücksichtigt. Patienten, die mit Degarelix behandelt wurden, hatten ein geringeres Risiko, ein kardiovaskuläres Ereignis zu erleiden als Patienten unter Agonisten- therapie (HR [95%-KI]: 0,76 [0,60–0,95], p= 0,018; [20]). Dies wird bestätigt durch die kürzlich publizierten Ergebnisse einer Real-world-Evidenzstudie aus Großbri- tannien. Das relative Risiko, ein kardiales Ereignis unter ADT zu erleben, war bei De- garelix signiﬁkant geringer als bei GnRH- Agonisten (RR 6,9 vs. 17,7%; 0,39 [95%- KI 0,191, 0,799]; p= 0,01; [8]). Komorbiditäten spielten bei der The- rapieentscheidung für Degarelix eine größere Rolle als bei einer Entscheidung für einen Agonisten (45,5% Degarelix vs. 35,9%GnRH-Agonisten).BeiPatientenmit kardiovaskulären Vorerkrankungen wurde Degarelix häuﬁger eingesetzt als GnRH- Agonisten. Das Patientenkollektiv war in dieser Studie zu klein, um zuverlässige AussagenbezüglichdesEinﬂussesderADT auf das kardiovaskuläre Risiko ableiten zu können. Verschiedene Publikationen [2, 3, 15, 23] und aktuelle Reviews [5, 6, 10, 24] deuten jedoch auf die Vorteile der GnRH-Antagonisten bei Patienten mit kardiovaskulären Vorerkrankungen hin. Inder ProComD-Studiezeigtesich, dass die Anamnese durch den Urologen über beide Arme hinweg als primäre Informati- onsquelle zu Begleiterkrankungen diente. Diskussion Alle innerhalb der Studie dokumentierten UE wurden nach der MedDRA Version 19.0 kodiert.DieUEwurdeninnichtschwerwie- gende und schwerwiegende UE eingeteilt. Bei der Erstuntersuchung gaben die Ärz- te für 96,1% der Patienten an, dass die Therapieentscheidung für beide Behand- lungsgruppen gleichermaßen einfach In der ProComD-Studie konnten Faktoren identiﬁziert werden, die für die Therapie- entscheidung bei Patienten mit hormon- naivem PCa herangezogen werden. Ko- Der Urologe 2 · 2022 179 Originalien morbiditäten, Alter und Compliance ha- ben dabei eine hohe Relevanz. Für 76% der Patienten wurde in dieser Studie min- destens eine Komorbidität dokumentiert. 32% der Patienten in der ProComD-Studie hatten kardiovaskuläre Vorerkrankungen. Ähnlich hoch (36%) war der Anteil in einer kürzlich veröﬀentlichten Versorgungsfor- schungsstudie [12]. Das durchschnittliche Alter der Patienten lag in der Studie bei 74 Jahren. Ineiner epidemiologischenStu- die in Deutschland lag das durchschnittli- che Alter der PCa-Patienten bei 72 Jahren [11]. se. Patienten, die mit Relugolix behandelt wurden, hatten ein um 54% geringeres relatives Risiko für schwerwiegende un- erwünschte kardiovaskuläre Ereignisse im Vergleich zu Leuprorelin. In der Subgrup- pe der Patienten mit einem schwerwie- genden kardiovaskulären Ereignis in der AnamnesewarderUnterschiednochdeut- licher mit einer 80%igen Reduktion des relativen Risikos [21]. Da Patienten mit kardiovaskulären Ri- sikofaktoren oder Komorbiditäten beson- ders von einem GnRH-Antagonisten pro- ﬁtieren, steht die Identiﬁzierung dieser im Vordergrund. Eine interdisziplinäre ka- nadische Arbeitsgruppe gibt Empfehlun- gen zur Identiﬁzierung von PCa-Patien- ten, die von einem optimalen Manage- ment ihrer kardiovaskulären Erkrankung und/oder einer Änderung der kardialen Ri- sikofaktoren proﬁtieren können. Dies um- fasst ein einfaches Screeningtool (STAMP), mit dem Hochrisikopatienten leicht identi- ﬁziert werden können [14]. Für die Identi- ﬁzierung und ggf. Therapie von Patienten mit hohem kardiovaskulärem Risiko exis- tiert auch in Deutschland eine S3-Leitlinie [9]. In einer weiteren kürzlich publizierten prospektiven Phase-II-Studie mit 80 Pa- tienten wurde der Zusammenhang zwi- schen kardiovaskulären Vorerkrankungen und der Auswirkung der Behandlung mit GnRH-Agonisten gegenüber Degarelix un- tersucht [16]. Von den Patienten, die einen GnRH-Agonisten erhielten, erlitten in den ersten 12 Monaten 20% ein schwerwie- gendes kardiovaskuläres oder zerebrovas- kuläres Ereignis im Vergleich zu 3% der Degarelix-Patienten (p= 0,013), was einer absolutenRisikoreduktionvon18,1%bzw. einer relativen Risikoreduktion von 88% entspricht [16]. Die eindringliche Empfeh- lung der Autoren lautet daher, Patienten mitkardiovaskulärenVorerkrankungenauf diese Daten hinzuweisen, bevor sie sich ei- ner ADT unterziehen. Die Ergebnisse von zwei großen rando- misierten Studien mit PCa-Patienten unter ADT [4, 22] konnten zeigen, dass kardio- vaskuläreEreignissediezweithäuﬁgsteTo- desursache (34 bzw. 27%) nach dem PCa (36 und 41%) sind. Keywords y Degarelix · GnRH antagonist · Treatment decision · Comorbidities · Treatment options AllebeschriebenenUntersuchungenamMenschen oderanmenschlichemGewebewurdenmitZustim- mungderzuständigenEthikkommission,imEinklang mitnationalemRechtsowiegemäßderDeklaration vonHelsinkivon1975(inderaktuellen,überarbei- tetenFassung)durchgeführt.Vonallenbeteiligten PatientenliegteineEinverständniserklärungvor. Literatur 1. Abufaraj M, Iwata T, Kimura S et al (2020) Diﬀerential impact of Gonadotropin-releasing hormone antagonist versus agonist on clinical safety and oncologic outcomes on patients with metastatic prostate cancer: a meta-analysis of randomizedcontrolledtrials.EurUrol79:44–53 Open Access.DieserArtikelwirdunterderCreative CommonsNamensnennung4.0InternationalLizenz veröﬀentlicht,welchedieNutzung,Vervielfältigung, Bearbeitung,VerbreitungundWiedergabeinjegli- chemMediumundFormaterlaubt,sofernSieden/die ursprünglichenAutor(en)unddieQuelleordnungsge- mäßnennen,einenLinkzurCreativeCommonsLizenz beifügenundangeben,obÄnderungenvorgenom- menwurden. PD Dr. med. J. Lehmann Urologische Gemeinschaftspraxis Prüner Gang, Gesundheitszentrum Kiel-Mitte Prüner Gang 15., 24103 Kiel, Deutschland lehmannkiel@arcor.de PD Dr. med. J. Lehmann 2. Agarwal M, Canan T, Glover G et al (2019) Cardiovascular eﬀects of androgen deprivation therapyinprostatecancer.CurrOncolRep21:91 3. AlbertsenPC,Klotz L,TombalB etal(2014)Cardio- vascular morbidity associatedwith gonadotropin releasinghormoneagonistsandanantagonist.Eur Urol65:565–573 4. Calais Da Silva FE, Bono AV, Whelan P et al (2009) Intermittent androgen deprivation for locally advanced and metastatic prostate cancer: results from a randomised phase 3 study of the South European Uroncological Group. Eur Urol 55:1269–1277 Danksagung. Die Autoren danken Dr. Juliane Schreier und Dr. Tanja Domke (co.medical, Berlin) für die Unterstützung bei der Manuskripterstellung. DieindiesemArtikelenthaltenenBilderundsonstiges Drittmaterialunterliegenebenfallsdergenannten CreativeCommonsLizenz,sofernsichausderAbbil- dungslegendenichtsanderesergibt.Soferndasbe- treﬀendeMaterialnichtunterdergenanntenCreative CommonsLizenz stehtunddiebetreﬀendeHandlung nichtnachgesetzlichenVorschriftenerlaubtist,istfür dieobenaufgeführtenWeiterverwendungendesMa- terialsdieEinwilligungdesjeweiligenRechteinhabers einzuholen. 5. CeredaV,FalboPT,MannaGetal(2020)Hormonal prostate cancer therapies and cardiovascular disease:asystematicreview.HeartFailRev.https:// doi.org/10.1007/s10741-020-09984-2 5. CeredaV,FalboPT,MannaGetal(2020)Hormonal prostate cancer therapies and cardiovascular disease:asystematicreview.HeartFailRev.https:// doi.org/10.1007/s10741-020-09984-2 Förderung. Die Studie wurde unterstützt von der Ferring Arzneimittel GmbH, Kiel. D. Kägebein ist Angestellter der Ferring Arzneimittel GmbH. 6. Corona G, Filippi S, Bianchi N et al (2020) Cardiovascular risks of androgen deprivation therapy for prostate cancer. World J Mens Health 38:e47 Factors inﬂuencing the choice of androgen deprivation therapy for patients with hormone-sensitive prostate cancer. Results of the ProComD study Background: Androgen deprivation therapy (ADT) with a GnRH agonist or the GnRH antagonist degarelix is a central component in the treatment of prostate cancer (PCa). Little is currently known regarding the decision criteria. Knowledge of these could improve the success of treatment in the future. Diskussion Besonders bei Patienten mit kardiovasku- lären Vorerkrankungen waren außerdem externe Arztbriefe und der Anamnesefra- gebogenvonBedeutung.Hierwurdedeut- lich, dass der Anamnesefragebogen bei Patienten mit kardiovaskulären Vorerkran- kungen in der Degarelix-Gruppe häuﬁger als Informationsquelle eingesetzt wurde als in der GnRH-Agonistengruppe. Die neuen Phase-III-Daten der HERO- Studiezum oralenGnRH-AntagonistenRe- lugolix bestätigen die Vorteile bei Patien- ten mit kardiovaskulären Vorerkrankun- gen. Bei Studieneinschluss hatten 80% der Patienten mindestens einen kardio- vaskulären oder zerebrovaskulären Risiko- faktor [21] und 14% ein schwerwiegendes kardiovaskuläres Ereignis in der Anamne- Die ProComD-Studie unterliegt den üb- lichen Limitationen einer NIS. Darüber hi- naus war keine Mindestdauer der ADT vor- gegeben. Dadurch konnten auch Patien- ten, die nur eine kurzzeitige ADT (z.B. als begleitendadjuvanteTherapieimRahmen 180 Der Urologe 2 · 2022 Abstract einer kurativ intendierten Primärtherapie) erhielten, eingeschlossen werden. Bei die- sen spielen evtl. andere Einﬂussfaktoren bei der Wahl der ADT eine Rolle als bei Pa- tienten, die für eine Langzeitbehandlung intendiert sind. Factors inﬂuencing the choice of androgen deprivation therapy for patients with hormone-sensitive prostate cancer. Results of the ProComD study Factors inﬂuencing the choice of androgen deprivation therapy for patients with hormone-sensitive prostate cancer. Results of the ProComD study Fazit für die Praxis Objectives: To identify factors inﬂuencing the treatment decision in patients with hormone-sensitive prostate cancer receiving ADT and to determine the incidence of concomitant disease in both treatment groups. 4 Die Faktoren Komorbiditäten, Alter und Compliance beeinﬂussen die Therapie- entscheidung. Methods: The two-arm, prospective, non-interventional study “ProComD” was conducted from September 2014 to June 2019 at 80 study centers in Germany. After the therapy decision was made, patients with hormone-sensitive prostate cancer needing ADT were included in the study. Data were collected during routine visits. Results: Data from 413 patients were evaluated (degarelix N= 268; GnRH agonists N= 145). Key factors inﬂuencing the therapy decision for both treatment options included comorbidities (42% of all patients), compliance (64%), and age (81%). The source of information consulted most frequently regarding existing comorbidities was the patient’s medical history conducted by the treating urologist themselves (65% in both groups). For patients with pre-existing cardiovascular diseases, the doctor’s letter (45.8% degarelix vs. 38.9% GnRH agonists) or the medical history questionnaire (38.9% degarelix vs. 20% GnRH agonists) was additionally taken into account. 4 Bestehende kardiovaskuläre Risikofak- toren und Begleiterkrankungen wurden häuﬁger bei den Degarelix-behandelten Patienten dokumentiert. 4 Anamnesefragebögen sowie externe Arztbriefe sind die häuﬁgsten vom Uro- logen verwendeten Informationsquellen bezüglich vorhandener kardiovaskulärer Begleiterkrankungen. 4 Kenntnisse über Komorbiditäten, Risiko- faktoren und Begleitmedikation könnten die Therapieentscheidung bezüglich ei- ner Androgendeprivationstherapie (ADT) erleichtern. Conclusion: Comorbidities along with age and compliance are among the key factors inﬂuencing the treatment decisions made by urologists. Korrespondenzadresse PD Dr. med. J. Lehmann Urologische Gemeinschaftspraxis Prüner Gang, Gesundheitszentrum Kiel-Mitte Prüner Gang 15., 24103 Kiel, Deutschland lehmannkiel@arcor.de Keywords Degarelix · GnRH antagonist · Treatment decision · Comorbidities · Treatment options Keywords l Einhaltung ethischer Richtlinien Interessenkonﬂikt. J.Lehmann:Ferring,Janssen, Astellas,Sanoﬁ/AventisundApogepha.C.W.Kluike, A.Haider,K.S.Haider,S.Baumann,M.Flesch,M.Ge- damkeundD.Kägebeingebenan,dasskeinInteres- senkonﬂiktbesteht. WeitereDetailszurLizenz entnehmenSiebitteder Lizenzinformationaufhttp://creativecommons.org/ licenses/by/4.0/deed.de. WeitereDetailszurLizenz entnehmenSiebitteder Lizenzinformationaufhttp://creativecommons.org/ licenses/by/4.0/deed.de. 7. D’amico AV, Denham JW, Crook J et al (2007) Inﬂuence of androgen suppression therapy for prostate cancer on the frequency and timing of fatal myocardial infarctions. J Clin Oncol 25:2420–2425 7. D’amico AV, Denham JW, Crook J et al (2007) Inﬂuence of androgen suppression therapy for prostate cancer on the frequency and timing of fatal myocardial infarctions. J Clin Oncol 25:2420–2425 Der Urologe 2 · 2022 181 Originalien OrganisationforResearchandTreatmentofCancer (EORTC)Trial30891.JClinOncol24:1868–1876 8. Davey P, Kirby MG (2020) Cardiovascular risk proﬁles of GnRH agonists and antagonists: real- world analysis from UK general practice. World J Urol.https://doi.org/10.1007/s00345-020-03433- 8. Davey P, Kirby MG (2020) Cardiovascular risk proﬁles of GnRH agonists and antagonists: real- world analysis from UK general practice. World J Urol.https://doi.org/10.1007/s00345-020-03433- 3 23. Tschöpe C, Kherad B, Spillmann F et al (2016) Cardiovascular risk of androgen deprivation therapy for treatment of hormone-dependent prostate cancer : Diﬀerences between GnRH antagonistsandGnRHagonists.Herz41:697–705 9. Deutsche Gesellschaft Für Allgemeinmedi- zin Und Familienmedizin E.V. (2017) S3- Leitlinie Hausärztliche Risikoberatung zur kardiovaskulären Prävention. AWMF on- line. https://leitlinien.dgk.org/2017/awmf-s3- leitlinie-hausaerztliche-risikoberatung-zur- kardiovaskulaeren-praevention. Zugegriﬀen: 11. Jan.2021 24. Van Poppel H, Abrahamsson PA (2020) Consi- derations for the use of gonadotropin-releasing hormoneagonistsandantagonistsinpatientswith prostatecancer.IntJUrol27:830–837 25. VanPoppelH,NilssonS(2008)Testosteronesurge: rationale for gonadotropin-releasing hormone blockers?Urology71:1001–1006 10. Freedland SJ, Abrahamsson PA (2020) Androgen deprivation therapy and side eﬀects: are GnRH antagonistssafer?AsianJAndrol22:1–8 11. Hermann S, Kraywinkel K (2019) Epidemiologie des Prostatakarzinoms in Deutschland. Onkologe 25:294–303 12. Hupe M, Hammerer P, Ketz M et al (2019) RetrospektiveGKV-Versorgungsforschungsstudie überGnRH-Antagonisten/-Agonistenzurinitialen TherapiedesfortgeschrittenenProstatakarzinoms – Verordnungsmuster und Krankenhauskosten in Deutschland.AktuelleUrol51:275–284 13. KeatingNL,O’malleyAJ,SmithMR(2006)Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol24:4448–4456 14. Kenk M, Grégoire JC, Coté MA et al (2020) Optimizing screening and management of cardiovascularhealthinprostatecancer:Areview. CanUrolAssocJ14:E458–E464 15. Klotz L, Miller K, Crawford ED et al (2014) Disease control outcomes from analysis of pooled individual patient data from ﬁve comparative randomised clinical trials of degarelix versus luteinisinghormone-releasinghormoneagonists. EurUrol66:1101–1108 16. Margel D, Peer A, Ber Y et al (2019) Cardiovascular morbidity in a randomized trial comparing GnRH agonist and GnRH antagonist among patients with advanced prostate cancer and preexisting cardiovasculardisease.JUrol202:1199–1208 17. Merseburger A, Bro Falkenberg A, Kornilova OJ (2019)Newstudysuggestspatientswithadvanced prostate cancer on androgen deprivation therapy need more dialogue with health care provider, especially around cardiovascular risk. World J Urol 37:1085–1093 18. Einhaltung ethischer Richtlinien Merseburger AS, Sedding D, Hüter K (2016) Cardiovascular risk patients under androgen deprivation therapy: lower risk with GnRH ant- agonists compared to LHRH agonists? Urologe A 55:218–225 19. Mottet N, Van Den Bergh R, Briers E (2018) EAU – ESTRO–ESUR–SIOGGuidelinesonProstateCancer 2018.EuropeanAssociationofUrologyGuidelines. EuropeanAssociationofUrologyGuidelinesOﬃce, Arnhem 20. Perrone V, Degli Esposti L, Giacomini E et al (2020) Cardiovascular risk proﬁle in prostate cancerpatientstreatedwithGnRHagonistsversus antagonists: an Italian real-world analysis. Ther ClinRiskManag16:393–401 21. Shore ND, Saad F, Cookson MS et al (2020) Oral relugolix for androgen-deprivation therapy in advanced prostate cancer. N Engl J Med 382:2187–2196 22. Studer UE, Whelan P, Albrecht W et al (2006) Immediate or deferred androgen deprivation for patients with prostate cancer not suitable for local treatment with curative intent: European 182 Der Urologe 2 · 2022 182
W4391822290.txt	https://journals.wlb-stuttgart.de/ojs/index.php/wfr/article/download/10325/10203	de		Rezension von: Borst, Otto, Ein Stück Deutscher Kulturgeschichte	Württembergisch Franken	2,024	cc-by	550	Neue Bücher 234 Hans-Ulrich Simon: Mörike-Chronik. Der Stuttgart: Metzler 1981. 415 S. Verfasser, wissenschaftlicher Mitarbeiter des Deutschen Literaturarchivs Marbach, aus den heute erreichbaren Daten zu Mörikes Leben graphie bereitgestellt. Konkrete hat handschriftlichen und gedruckten Quellen sämtliche wesentliche und Werk erfaßt und die Grundlagen für eine moderne Mörikebio- Die nüchterne Datensammlung beschränkt sich auf das Faktisch- (Lebensumstände, Entwicklungs- und Bildungsgang, Freundes- und Bekannten- kreis, Aufenthaltsorte, Reisen, Besuche, Lektüre, Entstehung der Lyrik und Prosaarbeiten) und gewinnt gerade dadurch, daß Verknüpfung und Bewertung Aufgabe des Lesers bleiben, ihre Eindringlichkeit. Die gen physiognomische Entwicklung des Dichters kann an 15 Abbildun- beobachtet werden. Wissenschaftler und Mörikeleser werden die Chronik mit ihren Gö umfangreichen Personen- und Werkregistern dankbar benutzen. Otto Borst: Ein Stück Deutscher Kulturgeschichte. Esslingen: Schreiber. 53 S. Die Bücher des Esslinger Verlags J. F. Schreiber gehörten einst (und gehören noch) zum Bestand jeder auch noch so kleinen Kinderbibliothek. Vor kurzem beging der bekannte Verlag seinen 150. Geburtstag. Aus diesem Anlaß schrieb Otto Borst mit gewandter Feder einen kurzen Rückblick auf die Verlagsgeschichte, die er mit Fug und Recht als einen Teil deutscher Kulturgeschichte ansieht. U. Erich Straßner: Fränkischer Volkshumor. Schwanksagen, Schildbürgergeschichten und Ortsneckereien aus Franken (= Veröffentlichungen der Gesellschaft für fränkische Geschichte, XII. Reihe, Quellen und Forschungen zur fränkischen Volkskunde, 2). Neustadt an der Aisch: Degener in Kommission Erich Straßner, 1933 in 1979. XII, 258 S. Treuchtlingen geboren, lebt seit nunmehr einem Jahrzehnt als Professor in Tübingen in der »schwäbischen Provinz«. Einer breiteren Öffentlichkeit ist der Linguist bekannt geworden als Kritiker der Nachrichtensprache in Hörfunk und Fernsehen, deren geringe Allgemeinverständlichkeit er bemängelt. In den Jahren 1962 und 1963 hat der Redakteur des Ostfränkischen Wörterbuchs von Erlangen aus mit Fragebogen nach Überund Spottnamen und nach entsprechenden Stücklein, Streichen und Schwänken gefahndet. Diese Umfrage erbrachte mehr als 3000 Necknamen-Orte, und sie ergab, daß fast jeder Ort in Ober-, Mittel- und Unterfranken mit einem oder mehreren Übernamen von den Nachbarn versehen worden war. In Erkundungsfahrten wurde dann noch der örtlichen Erzähltradition nachgespürt. »Dabei festzustellen, daß war hinter vielen Übernamen, die heute durchaus bei den Bewohnern der betroffenen Orte und in der näheren oder weiteren Umgebung bekannt sind, kein Inhalt Kenntnis mehr einer steht, kein Bezug auf ein Vorkommnis, Schwankgeschichte. Der Name wird nur das zum noch Namen tradiert, führte, keine ohne daß ein Spott oder gar Neid, Mißgunst oder ähnliches dahinter stünden. Bei anderen erzählt man sich zwar die alten Geschichten, die einst Spott hervorriefen, aber durchaus handfester wohlwollend, vielleicht mit leichtem Schmunzeln über die >gute alte Zeit<« (S. XI). Als »sanften, nicht mehr aggressiven Spott« kennzeichnet Erich Straßner die Necknamen heute, deren Traditionsträger kaum mehr die Menschen sind, eher die schriftliche Überlieferung in Kalendern, Zeitungen und Zeitschriften. In thematischer Reihung folgen dann die Beispiele, vielfach im mundartlichen Ausdruck. Ein Ortsregister schlüsselt das umfangreiche Material dieser Arbeit auf, die als unschätzbares Quellenwerk Die thematische jede Anerkennung verdient. Einteilung ist im wesentlichen übernommen von Hugo Mosers Habilitations- schrift »Schwäbischer Volkshumor- Neckereien in Stadt und Land, von Ort zu Ort«, die 1981 in zweiter ergänzter Auflage im Konrad Theiss Verlag, Stuttgart, erschienen ist. Hugo Moser hatte noch spekuliert, der Ostfranke sei weniger spottlustig als der Rheinfranke und der Schwabe. Erich Straßners Bestandsaufnahme der Ortsneckereien, Eulenspiegeleien und Schildbürgergeschichten in Ostfranken widerlegt diese Schreibtischthese. Dennoch ehrt mit seiner Arbeit Erich Straßner den Vorgänger Hugo Moser, und er sorgt zugleich dafür, daß
https://openalex.org/W2082588358	https://pubs.rsc.org/en/content/articlepdf/2015/ja/c5ja00055f	English	null	Chlorine isotope determination via the monitoring of the AlCl molecule by high-resolution continuum source graphite furnace molecular absorption spectrometry – a case study	Journal of analytical atomic spectrometry	2,015	cc-by	9,560	F. V. Nakadi,ab M. A. M. S. da Veiga,ab M. Aramend´ıa,ac E. Garc´ıa-Ruiza and M. Resanoa F. V. Nakadi,ab M. A. M. S. da Veiga,ab M. Aramend´ıa,ac E. Garc´ıa-Ruiza and M. Resanoa This work investigates the possibility of obtaining isotopic information via the monitoring of the absorption spectra of a gaseous diatomic molecule generated in a graphite furnace and using a high-resolution (approx. 1.5 pm per pixel) monochromator (HR CS GFMAS). To test this concept, Cl was chosen as the analyte and AlCl as the target species. The results demonstrate that, unlike what occurs with atomic spectra, under optimum conditions it is possible to acquire isotopic information by HR CS GFMAS in a straightforward way, as it is feasible to observe band heads for each Cl isotope (actually, for Al35Cl and Al37Cl) that are separated, i.e., they act like two diﬀerent molecules absorbing at diﬀerent wavelengths. The method proposed, based upon the addition of both Pd and Al and the selection of peak height values, enables Cl isotopic analysis with precision values around 2% RSD for solutions with Cl contents at the mg L1 level. Accurate values, within this uncertainty, can be directly obtained without requiring any method for mass bias correction. The potential of isotope dilution for calibration is also explored, and it is proven how this approach can help in providing accurate results in situations where the occurrence of chemical interferences, a case frequently encountered for the HR CS GFMAS technique, hampers the use of other calibration approaches, as demonstrated for water analysis. Received 20th February 2015 Accepted 7th April 2015 DOI: 10.1039/c5ja00055f www.rsc.org/jaas plasma mass spectrometry (MC-ICPMS),14 ion chromatography MC-ICPMS,15 gas chromatography-MS,16 and high performance liquid chromatography-MS/MS,17 have evaluated the isotope content of Cl in diﬀerent matrices. These techniques rely on the high sensitivity and selectivity of mass spectrometry, which typically requires rigorous sample preparation to avoid matrix interferences.18 aDepartment of Analytical Chemistry, Arag´on Institute of Engineering Research (I3A), University of Zaragoza, Pedro Cerbuna 12, 50009, Zaragoza, Spain. E-mail: mresano@unizar.es bWhile on leave from Departamento de Qu´ımica, Faculdade de Filosoa, Ciˆencias e Letras de Ribeir˜ao Preto, Universidade de S˜ao Paulo, 14040-901, Ribeir˜ao Preto-SP, Brazil cCentro Universitario de la Defensa-Academia General Militar de Zaragoza, Carretera de Huesca s/n, 50090, Zaragoza, Spain aDepartment of Analytical Chemistry, Arag´on Institute of Engineering Research (I3A), University of Zaragoza, Pedro Cerbuna 12, 50009, Zaragoza, Spain. E-mail: mresano@unizar.es J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1531 Cite this: J. Anal. At. Spectrom., 2015, 30, 1531 Received 20th February 2015 Accepted 7th April 2015 DOI: 10.1039/c5ja00055f www.rsc.org/jaas JAAS PAPER ticle. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. s article is licensed under a Creative Commons Attribution 3.0 Unported Licence. JAAS cCentro Universitario de la Defensa-Academia General Militar de Zaragoza, Carretera de Huesca s/n, 50090, Zaragoza, Spain bWhile on leave from Departamento de Qu´ımica, Faculdade de Filosoa, Ciˆencias e Letras de Ribeir˜ao Preto, Universidade de S˜ao Paulo, 14040-901, Ribeir˜ao Preto-SP, Brazil 2.2 Standards, reagents and samples Two reference materials with a certied Cl isotopic composition (CRMs) were used in this work: SRM 975a Isotopic Standard for Chlorine (35Cl atom percent: 75.774 0.028%; 37Cl atom percent: 24.226 0.028%; absolute abundance ratio 35Cl/37Cl: 3.1279 0.0047), available from the National Institute of Standards and Technology (NIST, USA), and ERM-AE642 Cl in water (37Cl content: 4.375 0.026 106 mol g1; absolute abundance ratio 35Cl/37Cl: 0.01914 0.00048), available from the Institute for Reference Materials and Measurements (IRMM, Belgium). A stock standard Cl solution, 1000 mg L1 (Merck, Germany) was also used. In all of these materials, Cl was present as NaCl. There is another advantage of HR CS instrumentation that should not be overlooked. That is the potential to monitor not only atomic but also complex and transient molecular spectra, as generated from a graphite furnace, which oﬀers more sensitivity than a ame as the atomizer. The resulting tech- nique, high-resolution continuum source graphite furnace molecular atomic spectrometry (HR CS GFMAS) opens new ways to determine non-metals,34–36 for which it is hardly feasible to get access to their main atomic lines because they are located in the vacuum UV. But the use of HR CS GFMAS can also be benecial for isotopic analysis, because, in the same way as explained before for LAMIS and emission spectrometry, the isotopic shis found when monitoring molecular absorption spectra are expected to be much larger than those found for atomic absorption spectra. Other Cl solutions were prepared with diclofenac sodium (2-[(2,6-dichlorophenyl)amino]benzeneacetic acid sodium salt, Fluka, Switzerland), KCl, NH4Cl and HCl (Merck). High-purity water (Trace Select Ultra, Cl level lower than 1 mg kg1, Fluka) was used for the preparation of all solutions. A stock standard Pd solution of 10 g L1 (Merck) was diluted to 2 g L1, and this solution was applied as the chemical modier. A stock stan- dard Al solution of 1000 mg L1 (Merck) was also used to generate the AlCl molecule. A 10 g Ca L1 solution was prepared by dissolving CaCO3 (Merck) in 5% (v/v) HNO3 (Merck), and used for an interference study. All the reagents were of analytical grade purity or higher. This work aims at exploring this idea, the evaluation of the absorption of diatomic molecules, as a new possibility to acquire isotope data by means of HR CS GFMAS. 1 Introduction Despite its toxicity, chlorine is a widely used industrial commodity. Chlorine is used in many diﬀerent elds, such as production of chlorinated solvents and other chemicals, water purication, plastics production and pulp and paper manu- facture,1 thus making it necessary to control the presence of this element at trace levels in a wide variety of samples.2–9 Some interesting alternatives to MS techniques, which may be worth exploring, are appearing in the literature recently. For instance, laser ablation molecular isotopic spectrometry (LAMIS), developed by Russo et al., explores the concept of wavelength isotopic shis between diatomic molecules.19,20 This technique can be seen as an evolution of LIBS, but instead of aiming at the monitoring of the emission spectra of elemental species, which are well-known to show very small isotopic variations, its goal is to evaluate and quantify isotopic shis occurring for diatomic molecules, in which one of the atoms is the target analyte. These shis are signicantly larger than their atomic counterparts,19 thus permitting to determine boron,19,21,22 carbon,19,23 hydrogen,19,24 oxygen,19 and strontium25 isotopes, without requiring extensive sample preparation or a vacuum environment. In addition to controlling the elemental content, investi- gating the isotopic composition of Cl can be of interest in diﬀerent situations, as it may help, for instance, to track the source or the fate of diﬀerent Cl species in the environment, or in biological uids, or help in elucidating geochemical processes.10 Some recent articles, based on the use of diﬀerent techniques such as gas source mass spectrometry (oen referred to as IR-MS11), thermal ionization mass spectrom- etry,12,13 laser ablation multicollector-inductively coupled In principle, not only emission but also absorption could be used to obtain isotopic information. When the goal is to monitor atomic absorption, high-resolution continuum source J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1531 This journal is © The Royal Society of Chemistry 2015 Paper View Article Online Paper View Article Online View Article Online JAAS atomic absorption spectrometry (HR CS AAS) oﬀers very signicant advantages when compared to classic line source instrumentation,26 such as improved possibilities to detect and correct for interferences and lower limits of detection.27–32 In addition, this technique oﬀers enough resolution (down to approx. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. So far, only one study has intended to attain isotopic infor- mation using HR CS AAS.33 In such pioneering work, Wiltsche et al. used a ame as the atomizer and tried to take advantage of the large mass diﬀerence existing between the two isotopes of B, 10B and 11B, to determine B ratios in steel samples. The authors found a less sensitive atomic line (around 208.9 nm) that shows a higher isotopic shithan the most sensitive one (249.8 nm): 208.95898 nm for 10B and 208.95650 nm for 11B. This diﬀerence is still too small to observe two fully resolved peaks with a HR CS FAAS instrument. However, it can still be appreciated that the B absorption prole shis as a function of the boron isotopic ratio: the higher the % of 10B, the higher the wavelength at which the peak maximum is found. Thus, it is in principle possible to determine the B isotopic composition. Nevertheless, since the isotopic shiof boron is very small, it is necessary to correct for possible instabilities of the monochromator in order to achieve suﬃcient precision and accuracy, which was done using a suitable internal standard (Fe or Ni) that was monitored in truly simultaneous mode. This approach was found satis- factory to distinguish diﬀerent enrichment levels of 10B (within 5%) in steel samples. Moreover, an ICP mass spectrometer, NexION 300X (Perkin Elmer, USA) was used to validate the results obtained for the water samples by means of HR CS GFMAS. 1 Introduction 1 pm) to quantify diﬀerent portions of an atomic line; however, even with such an instrument, it is very challenging to detect the small energy variations occurring for diﬀerent isotopes, owing to a number of well-known factors (i.e., colli- sional broadening, Doppler and Stark eﬀects) that result in the broadening of the peaks. This journal is © The Royal Society of Chemistry 2015 2.1 Instrumentation All the AlCl measurements were performed with a high-reso- lution continuum source atomic absorption spectrometer, ContrAA 700, commercially available from Analytik Jena AG (Jena, Germany). The optical system comprises a xenon short- arc lamp (GLE, Berlin, Germany) operating in “hot-spot” mode as the radiation source, a high-resolution double echelle monochromator (DEMON) and a linear CCD array detector with 588 pixels, 200 of which are used for analytical purposes (monitoring of the analytical signal and BG correc- tion), while the rest are used for internal functions, such as correcting for uctuations in the lamp intensity. More details on this type of instrumentation can be found elsewhere.26 This instrument is also equipped with a transversely heated graphite tube atomizer, pyrolytic graphite tubes, and both automated solid sampling and liquid sampling accessories. In this work, only liquids were introduced in the graphite furnace. 2.2 Standards, reagents and samples In order to investigate this new concept, which to the best of our knowledge has not been reported before in the literature, Cl has been selected as the analyte, owing to its relevance, as discussed before. Al has been chosen as the pair to create the AlCl mole- cule, which will be the target of the HR CS GFMAS monitoring, with the goal of producing isotopic transitions that could be spectrally resolved. Finally, two CRM samples with certied Cl levels were analyzed in this work: ION-915, natural water from Lake Supe- rior (lot 1109), and KEJIM-02, sowater from Kejimkujik Lake (lot 0914), both available from Environment Canada (Canada). Five mineral water samples were acquired at local stores and analyzed as well. This journal is © The Royal Society of Chemistry 2015 1532 \| J. Anal. At. Spectrom., 2015, 30, 1531–1540 View Article Online JAAS Paper Table 1 Instrumental parameters used to determine Cl by monitoring the AlCl molecule using HR CS GFMAS Electronic transition X1S+ / A1P Wavelengths 262.238 nm (Al35Cl) 262.222 nm (Al37Cl) Number of detector pixels summed per line 3 (4.65 pm) Reactant/chemical modier Al (10 mg)/Pd (20 mg) Sample volume 10 mL Temperature program Step Temperature/C Ramp/C s1 Hold/s Ar gas ow/L min1 Drying 90 3 20 2.0 Drying 110 5 20 2.0 Pyrolysis 500 300 25 2.0 Vaporization 2200 1500 5 0 Cleaning 2500 500 4 2.0 Table 2 Instrumental parameters used to determine Cl by means of ICPMS RF power/W 1600 Plasma gas ow rate (Ar)/L min1 18 Nebulizer gas ow rate (Ar)/L min1 1.0 Auxiliary gas ow rate (Ar)/L min1 1.2 Scan mode Peak hopping RPq 0.25 Dwell time/ms 150 Sweeps/reading 1 Readings/replicate 20 Number of replicates 10 Nuclides monitored 35Cl+ Paper Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Table 1 Instrumental parameters used to determine Cl by monitoring the AlCl molecule using HR CS GFMAS Electronic transition X1S+ / A1P Wavelengths 262.238 nm (Al35Cl) 262.222 nm (Al37Cl) Number of detector pixels summed per line 3 (4.65 pm) Reactant/chemical modier Al (10 mg)/Pd (20 mg) Sample volume 10 mL Temperature program Table 1 Instrumental parameters used to determine Cl by monitoring the AlCl molecule using HR CS GFMAS element,36 which probably indicates that its determination is not straightforward. 2.2 Standards, reagents and samples So far, only three articles have reported on the use of this technique to determine Cl. The rst two articles selected AlCl as the target molecule,4,37 while the most recent one opted for SrCl.9 In this work, we have selected AlCl, not only because it is a fairly sensitive and stable molecule (bond dissociation energy, 511 kJ mol1), but also because Al is monoisotopic and, therefore, the potential isotopic shishould only obey Cl variations. Furthermore, the isotopic shis observed for this molecule are suﬃciently large to be appreci- ated by means of HR CS GFMAS, as will be proved below. Step Temperature/C Ramp/C s1 Hold/s Ar gas ow/L min1 Drying 90 3 20 2.0 Drying 110 5 20 2.0 Pyrolysis 500 300 25 2.0 Vaporization 2200 1500 5 0 Cleaning 2500 500 4 2.0 The theoretical isotopic shican be derived from the equa- tion available in the classic book of Herzberg,38 in a similar way as described by Russo et al.,19 as shown below: Dv ¼ ð1 rÞ u0 e v0 þ 1 2 u00 e v00 þ 1 2 1 r2 " u0 ex0 e v0 þ 1 2 2 u00 ex00 e v00 þ 1 2 2# þ 1 r3 " u0 ey0 e v0 þ 1 2 3 u00 ey00 e v00 þ 1 2 3# (1) (1) where Dn is the isotopic shiin cm1, v is the vibrational quantum number, ue is the harmonic frequency, uexe and ueye are the rst and second anharmonic constants, respectively; r ¼ (m/mi)1/2, where m is the reduced mass of the molecule and i corresponds to the heavier isotope. The number of apostrophes denotes the electronic levels (two for the lower one, and one for the upper one) involved in the electronic transition. The only diﬀerence with the equation used in ref. 19 is that we observed that, for the AlCl molecule and for the type of transitions investigated in our work (transitions between diﬀerent vibrational levels), it is necessary to add a third term. In theory, the number of terms is innite, but with these three terms shown above an excellent agreement between theoretical and experimental shis can be obtained (R2 ¼ 0.99987). This journal is © The Royal Society of Chemistry 2015 J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1533 2.3 Analysis of the samples Samples were directly analyzed without any dilution step by means of HR CS GFMAS, using the instrumental parameters shown in Table 1. Both a calibration curve prepared with ve Cl standards covering the interval 1–10 mg L1 and isotope dilu- tion (using a spike of 0.509 35Cl/37Cl molar ratio and a Cl content of 1.09 mmol L1) were used for quantitation. Table 3 Experimental and theoretical isotopic shifts found for the analytical band heads of the AlCl molecule (X1S+ / A1P electronic transition) by HR CS GFMAS for diﬀerent vibrational transitions (v0, v00). The isotopic shift was calculated using eqn (1) and converted to wavelength for correlation with experimental data. The relative sensitivity was compared with the most sensitive line (261.418 nm) and, when two separate peaks were observed, the Al35Cl and Al37Cl signals were summed On the other hand, mineral water samples were also analyzed by means of ICPMS for validation purposes. The ICPMS parameters are shown in Table 2. In this case, samples were diluted using high-purity water and the calibration curve was constructed using ve Cl standards that covered the interval between 100 and 1000 mg L1. In all cases, three diﬀerent sample aliquots were analyzed. la/nm lexp/nm v0,v00 Dlcalc/pm Dlexp/pm Relative sensitivity/% 261.44 261.418 0,0 1.37 — 100 261.70 261.695 1,1 4.86 4.8 62 261.82 261.819 2,2 9.61 9.6 30 262.24 262.238 3,3 16.0 15.6 26 262.70 262.697 4,4 24.2 24.3 16 263.22 263.216 5,5 34.8 35.3 7.4 263.81 263.807 6,6 48.0 48.1 3.2 264.49 264.490 7,7 64.3 64.5 1.1 a Data obtained from ref. 44. la/nm lexp/nm v0,v00 Dlcalc/pm Dlexp/pm Relative sensitivity/% 3.1 Cl monitoring by HR CS GFMAS 3.1.1 Theoretical and experimental Cl isotopic monitoring. Contrary to what occurs with other non-metals such as F, P and S, the number of articles devoted to Cl monitoring by HR CS GFMAS is surprisingly low considering the importance of this This journal is © The Royal Society of Chemistry 2015 J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1533 Paper View Article Online View Article Online JAAS Paper In order to better evaluate the potential of these transitions for isotopic analysis, a series of 2D spectra (abs. versus wave- length) were acquired for a Cl solution with a 35/37 molar ratio close to unity (except for Fig. 1B, as will be discussed below). These spectra are shown in Fig. 1. As previously discussed, Fig. 1A shows that for the 261.418 nm transition only a broad band can be detected, as the signals for the two isotopes are not resolved. This transition is the one typically selected for Cl elemental analysis.4,37 The next transition in terms of sensitivity, located at 261.695 nm, still did not show two diﬀerent peaks, but it could be seen that the peak maximum shis as a function Table 3 shows this comparison. The isotopic shiof the 261.418 nm transition could not be observed because the shi is too small for the resolution provided by the instrument (every pixel covers 1.47 pm in this wavelength region). For the rest of the transitions, the diﬀerence between the theoretical and experimental shis was always lower than 2%. It is important to stress that, unfortunately, the isotopic shi is inversely proportional to the sensitivity of the transition. This is probably due to the fact that the higher shis occur for higher vibrational levels, which exhibit a lower population even at the relatively high temperature of a graphite furnace. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Cl (1 : 1, 35Cl/37Cl molar ratio solution) at wavelengths of: (A) 261.418 nm; (B) 261.695 nm; (C) 261.819 nm; (D) 262.238 n Cl (1 : 1, 35Cl/37Cl molar ratio solution) at wavelengths of: (A) 261.418 nm; (B) 261.695 nm; (C) 261.819 nm; (D) 262.238 nm; (E) 216 nm; (G) 263.807 nm; and (H) 264.490 nm. They were obtained for 400 ng of Cl, 10 mg of Al and 20 mg of Pd. 3.1 Cl monitoring by HR CS GFMAS However, 1B spectra: in black, the one obtained for 400 ng of Cl of CRM NIST 975a (35Cl, 75.774%), and in red, the one obtained for 400 ng (37Cl, 98.122%). Fig. 1 Spectra of AlCl (1 : 1, 35Cl/37Cl molar ratio solution) at wavelengths of: (A) 261.418 nm; (B) 261.695 nm; (C) 261.819 nm; (D) 262.238 nm; (E) 262.697 nm; (F) 263.216 nm; (G) 263.807 nm; and (H) 264.490 nm. They were obtained for 400 ng of Cl, 10 mg of Al and 20 mg of Pd. However, 1B shows two diﬀerent spectra: in black, the one obtained for 400 ng of Cl of CRM NIST 975a (35Cl, 75.774%), and in red, the one obtained for 400 ng of Cl of CRM AE642 (37Cl, 98.122%). This journal is © The Royal Society of Chemistry 2015 1534 \| J. Anal. At. Spectrom., 2015, 30, 1531–1540 View Article Online Fig. 2 Optimization of the (A) Al and (B) Pd amount used to monitor Al35Cl by HR CS GFMAS for 200 ng of Cl at the 262.238 nm line. 20 mg of Pd were added during the Al optimization, and 10 mg of Al were added during the Pd optimization. The three central pixels were summed for quantitation. Error bars represent the standard deviation (n ¼ 5). Fig. 3 Optimization of the pyrolysis and vaporization temperatures to monitor AlCl by HR CS GFMAS for 200 ng of Cl (37Cl/35Cl molar ratio, 1.08) at the 262.222 (Al37Cl) and 262.238 (Al35Cl) nm transitions when adding 10 mg of Al and 20 mg of Pd. A pyrolysis temperature of 300 C was set for the vaporization study, and a vaporization temperature of 2400 C for the pyrolysis study. The three central pixels of each transition were summed for quantitation. Error bars represent the standard deviation (n ¼ 5). JAAS Fig. 2 Optimization of the (A) Al and (B) Pd amount used to monitor Al35Cl by HR CS GFMAS for 200 ng of Cl at the 262.238 nm line. 20 mg of Pd were added during the Al optimization, and 10 mg of Al were added during the Pd optimization. The three central pixels were summed for quantitation. Error bars represent the standard deviation (n ¼ 5). JAAS Paper of the isotopic composition. This aspect can be appreciated in Fig. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. The 261.238 nm transition (Fig. 1D) shows a better situation, as the peaks, while still not fully resolved, do not seem to overlap suﬃciently to inuence the ratio obtained in a signi- cant way, since the ratio of both peaks is close to unity. This spectrum hints for the rst time at the possibility of carrying out isotopic analysis by HR CS MAS with good resolution and in a straightforward manner, as the band head of each Cl isotope is separated, i.e., they act like two diﬀerent molecules absorbing at diﬀerent wavelengths. The situation for the next transitions (Fig. 1E–H) shows an even higher separation between the peaks, but at the cost of a decreasing signal. This can be problematic as the eﬀect of the background structure may begin to aﬀect the ratios obtained. Still, these transitions may be useful for higher Cl levels. For determinations at the mg L1 level, the transition shown in Fig. 1D was considered as the most suitable and was further used throughout the work. Fig. 2 Optimization of the (A) Al and (B) Pd amount used to monitor Al35Cl by HR CS GFMAS for 200 ng of Cl at the 262.238 nm line. 20 mg of Pd were added during the Al optimization, and 10 mg of Al were added during the Pd optimization. The three central pixels were summed for quantitation. Error bars represent the standard deviation (n ¼ 5). 3.1.2 Optimization of the working parameters. Once the most suitable wavelength was selected, other aspects were investigated. Two factors that can be critical for the formation of the AlCl molecule in the graphite furnace are (i) the amount of Al added, that should be enough to aim at the maximum Cl conversion into AlCl, and (ii) the amount of the chemical modier added, which may help in stabilizing both Al, and particularly, Cl species. Pd was chosen as the chemical modier and the amount of both Al and Pd were optimized, as shown in Fig. 2. Fig. 3 Optimization of the pyrolysis and vaporization temperatures to monitor AlCl by HR CS GFMAS for 200 ng of Cl (37Cl/35Cl molar ratio, 1.08) at the 262.222 (Al37Cl) and 262.238 (Al35Cl) nm transitions when adding 10 mg of Al and 20 mg of Pd. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. A pyrolysis temperature of 300 C was set for the vaporization study, and a vaporization temperature of 2400 C for the pyrolysis study. The three central pixels of each transition were summed for quantitation. Error bars represent the standard deviation (n ¼ 5). Only a small AlCl absorbance variation was observed for Al contents greater than 5 mg (Fig. 2A). Nonetheless, an amount of 10 mg was chosen for further studies to ensure the correct formation of the AlCl molecule. As for the Pd content (see Fig. 2B), it is clear that it has a signicant inuence on the AlCl signal. Low Pd levels lead to poor sensitivity, while very high contents also provide a very low signal, probably because AlCl is no longer formed in signicant amounts owing to the strong Pd competition. The optimum values range from 5 to 20 mg of Pd. A value of 20 mg was chosen for further studies. Fig. 3 Optimization of the pyrolysis and vaporization temperatures to monitor AlCl by HR CS GFMAS for 200 ng of Cl (37Cl/35Cl molar ratio, 1.08) at the 262.222 (Al37Cl) and 262.238 (Al35Cl) nm transitions when adding 10 mg of Al and 20 mg of Pd. A pyrolysis temperature of 300 C was set for the vaporization study, and a vaporization temperature of 2400 C for the pyrolysis study. The three central pixels of each transition were summed for quantitation. Error bars represent the standard deviation (n ¼ 5). The temperature program was subsequently optimized as well (see Fig. 3). In the presence of Pd, a stable signal was obtained for a pyrolysis temperature of up to 500 C, with good agreement between the theoretical and experimental isotopic ratios. As for the vaporization temperature, a plateau is reached for temperatures of 2100 C or higher. The use of 2200 C guarantees maximum sensitivity, a good peak denition while still preserving the lifetime of the graphite parts. these optimum conditions was recorded. The temporal signal proles obtained for Al35Cl are shown in Fig. 4. As can be seen, no signicant diﬀerence between them was observed (variations of both the peak area and the peak height were smaller than 6%). The same conclusion was reached for the Al37Cl transition. This journal is © The Royal Society of Chemistry 2015 3.1 Cl monitoring by HR CS GFMAS 1B, which compares the signal obtained for a spike enriched in 37Cl with that occurring for another solution richer in 35Cl. This is similar to the situation found by Wiltsche et al. when monitoring atomic lines for B isotopic analysis.33 The next transition, 261.819 nm (Fig. 1C), already shows two diﬀerent peaks, the leone corresponding to Al37Cl and the right one to Al35Cl, although they are not well resolved. That would aﬀect the ratio nally obtained, as the tailing of the Al37Cl peak has a clear eﬀect on the Al35Cl signal, thus resulting in a 35/37 ratio that is biased high. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Finally, the AlCl signal obtained for ve very diﬀerent Cl species (diclofenac sodium, HCl, KCl, NH4Cl and NaCl) under This journal is © The Royal Society of Chemistry 2015 J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1535 Paper View Article Online View Article Online Paper JAAS Paper Fig. 4 Comparison of the temporal peak proﬁles obtained for ﬁve diﬀerent Cl species (250 ng of Cl) when monitoring the 262.238 nm Al35Cl transition. The three central detector pixels were summed to represent the proﬁle. 3.2 Quantication of the isotopic shi The values in blue were obtained using the peak area, and the values in green using the peak height, while a lighter color intensity indicates a higher number of detector pixels (1, 3, 5 or 7) selected in both cases. In the case of using the peak height, the absorbance recorded by those detector pixels at the time at which the maximum peak height appears was summed. Error bars represent the standard deviation (n ¼ 5). (B) Comparison of the variations observed for a series of 20 replicates for a solution (200 ng of Cl) of CRM NIST 975a when monitoring the individual 262.222 (Al37Cl) and 262.238 (Al35Cl) nm transitions (left y-axis) or when calculating the Cl ratio from the same set of data (right y-axis). The red dashed line represents the certiﬁed value. The blue line and the blue interval surrounding it represent the average value of the 20 replicates and its uncertainty (as the standard deviation), respectively. Fig. 5 (A) Evaluation of the accuracy and precision obtained for the quantiﬁcation of the Cl ratio for a solution (200 ng of Cl) of CRM NIST 975a (35/37 ¼ 3.1279 0.0047) for diﬀerent AlCl transitions as a function of the way in which the signal is processed and quantiﬁed. The values in blue were obtained using the peak area, and the values in green using the peak height, while a lighter color intensity indicates a higher number of detector pixels (1, 3, 5 or 7) selected in both cases. In the case of using the peak height, the absorbance recorded by those detector pixels at the time at which the maximum peak height appears was summed. Error bars represent the standard deviation (n ¼ 5). (B) Comparison of the variations observed for a series of 20 replicates for a solution (200 ng of Cl) of CRM NIST 975a when monitoring the individual 262.222 (Al37Cl) and 262.238 (Al35Cl) nm transitions (left y-axis) or when calculating the Cl ratio from the same set of data (right y-axis). The red dashed line represents the certiﬁed value. The blue line and the blue interval surrounding it represent the average value of the 20 replicates and its uncertainty (as the standard deviation), respectively. Fig. 3.2 Quantication of the isotopic shi 3.2.1 Signal evaluation. Traditionally, in GFAAS it is pref- erable to integrate the absorbance peak along time, as it typi- cally leads to better precision, and it may minimize the inuence of the matrix on the nal result. However, when the aim is to determine the isotopic composition, the situation could be diﬀerent. Since there is no information on this issue, a study was performed evaluating the signals both in terms of the peak area and of the peak height, and also investigating the optimum number of detector pixels to be considered, which is another parameter that can signicantly inuence the nal result.39–41 Fig. 5A shows the results obtained during this study of the main ve AlCl transitions for which two separate Al37Cl and Al35Cl were detected, as discussed in Section 3.1.1. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Fig. 4 Comparison of the temporal peak proﬁles obtained for ﬁve diﬀerent Cl species (250 ng of Cl) when monitoring the 262.238 nm Al35Cl transition. The three central detector pixels were summed to represent the proﬁle. These results were obtained via analysis of a CRM (NIST 975a) with a certied Cl ratio that is shown in the gure (dotted line). As can be seen, the results for the rst transition evaluated (around 261.82 nm) are biased high. This was expected because the two peaks for Al37Cl and Al35Cl are not well resolved in this case (see Fig. 1C). However, the diﬀerence with the certied value is lower when the peak height is selected. This diﬀerence This conrms that the method developed shows potential for the monitoring of Cl isotopes regardless of the particular form in which Cl is found in the samples. This conrms that the method developed shows potential for the monitoring of Cl isotopes regardless of the particular form in which Cl is found in the samples. Fig. 5 (A) Evaluation of the accuracy and precision obtained for the quantiﬁcation of the Cl ratio for a solution (200 ng of Cl) of CRM NIST 975a (35/37 ¼ 3.1279 0.0047) for diﬀerent AlCl transitions as a function of the way in which the signal is processed and quantiﬁed. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. The following transitions oﬀer results that tend to be biased low. As discussed before, the problem of such transitions is that the sensitivity decreases with every one of them and the inu- ence of the structured background on the ratio becomes more pronounced (see Fig. 1E–G). This inuence is more signicant when the peak area is selected instead of the peak height, as could be anticipated. The precision, as shown by the error bars, is also signicantly better when using the peak height. There- fore, overall, the use of the peak height seems to be preferable for this type of analysis, and was adopted for further work. As for the number of pixels, it seems to be a critical aspect only if the Al37Cl and Al35Cl peaks are closely located. As a general rule, a number of 3 pixels was selected (central pixel 1), but this is a parameter that should be evaluated also depending on the analyte content for every particular situation, as higher contents may lead to peak broadening and a higher chance of spectral overlap. 3.2.2 Correction for spectral overlap in the case of high 37/ 35 ratios. Despite the satisfactory results obtained so far for solutions showing a 35Cl/37Cl molar ratio close to the natural one, when performing tracer experiments or using isotope dilution, ratios that diﬀer from the natural one are oen used. That could represent a potential problem, because the tailing of the Al37Cl may inuence the Al35Cl signal (see Fig. 1) in cases when spikes with a high 37/35 ratio are used. Thus, in order to check the robustness of the method developed, solutions with very diﬀerent ratios were prepared by mixing diﬀerent amounts of two reference materials (NIST 975a and ERM-AE642), and their isotope ratio was calculated experimentally by HR CS GFMAS using the optimum parameters discussed before and summarized in Table 1. The results obtained are shown in Fig. 6A. As can be seen, a very good agreement between theo- retical and experimental values was obtained for 37/35 ratios up to 1.5. However, spikes showing a higher 37/35 ratio tend to provide results that are biased low, which indicates a potential overlap, as discussed above. The precision that can be achieved under these optimum conditions was evaluated by isotope analysis of a solution of Fig. 3.2 Quantication of the isotopic shi 5 (A) Evaluation of the accuracy and precision obtained for the quantiﬁcation of the Cl ratio for a solution (200 ng of Cl) of CRM NIST 975a (35/37 ¼ 3.1279 0.0047) for diﬀerent AlCl transitions as a function of the way in which the signal is processed and quantiﬁed. The values in blue were obtained using the peak area, and the values in green using the peak height, while a lighter color intensity indicates a higher number of detector pixels (1, 3, 5 or 7) selected in both cases. In the case of using the peak height, the absorbance recorded by those detector pixels at the time at which the maximum peak height appears was summed. Error bars represent the standard deviation (n ¼ 5). (B) Comparison of the variations observed for a series of 20 replicates for a solution (200 ng of Cl) of CRM NIST 975a when monitoring the individual 262.222 (Al37Cl) and 262.238 (Al35Cl) nm transitions (left y-axis) or when calculating the Cl ratio from the same set of data (right y-axis). The red dashed line represents the certiﬁed value. The blue line and the blue interval surrounding it represent the average value of the 20 replicates and its uncertainty (as the standard deviation), respectively. This journal is © The Royal Society of Chemistry 2015 1536 \| J. Anal. At. Spectrom., 2015, 30, 1531–1540 View Article Online Paper JAAS NIST 975a with a Cl concentration of 20 mg L1. The results obtained aer 20 replicates are shown in Fig. 5B. As can be seen, the typical variation observed for a particular individual isotopic AlCl transition is rather high (approx. 8%), but when isotope ratios are calculated a much better value is obtained, as all the correlated sources of noise are compensated. In this way, a 2% RSD value can be considered as typical for Cl isotopic analysis under these conditions. This value is clearly insuﬃcient if the goal is to monitor natural isotopic variations, which for Cl are expected to be much lower, but it may be enough for performing tracer experiments or to use isotope dilution for calibration, thus opening new possibilities for the HR CS GFMAS technique. J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1537 This journal is © The Royal Society of Chemistry 2015 3.2 Quantication of the isotopic shi Furthermore, it can be stressed that, unlike what occurs with most MS techniques, in which at least correction for the instrumental mass bias is required, the results demonstrate that an accurate ratio, within the uncertainty mentioned above, can be directly obtained without performing any type of correction. increases with the number of detector pixels chosen for quan- titation (which is logical, because the degree of peak overlap is reduced when less pixels are selected). The precision obtained is also better for the peak height. The results obtained for the next transition (around 262.24 nm, spectrum shown in Fig. 1D) are signicantly better, and they are in good agreement with the reference value using both the peak area and the peak height. This is the transition that was initially selected as most promising, and these results conrmed its selection for Cl monitoring at the mg L1 level. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. (A) Raw spectrum obtained at the time that provides the peak maximum; (B) eﬀect of the Al37Cl signal tail on the Al35Cl peak; (C) application of the asymptotic regression model (y ¼ a bcx) using Origin 9.1 software; (D) estimation of the actual Al35Cl peak. Downloaded on 10/24/2024 5:05:29 AM. ive Commons Attribution 3.0 Unported Licence. Fig. 7 Spectral overlap correction performed at 262.238 nm for a Cl (500 ng) spike with a 37Cl/35Cl molar ratio ¼ 2.5. (A) Raw spectrum obtained at the time that provides the peak maximum; (B) eﬀect of the Al37Cl signal tail on the Al35Cl peak; (C) application of the asymptotic regression model (y ¼ a bcx) using Origin 9.1 software; (D) estimation of the actual Al35Cl peak. Fig. 8 Study of the Ca interference. (A) Al35Cl absorbance signal, as recorded at 262.238 nm, and (B) Al35Cl/Al37Cl absorbance ratio as a function of the Ca content, for solutions containing 100 ng of Cl, 10 mg of Al and 20 mg of Pd. Error bars represent 2 s (n ¼ 5). interferences, because the presence of other elements in the matrix may result in the formation of a molecule diﬀerent from the target one at some point of the pyrolysis and/or vaporization steps. This fact can aﬀect the absorbance nally obtained and make it diﬃcult to develop a straightforward calibration procedure. Therefore, it is important to choose a target molecule that is stable even at high temperatures, such as the case of AlCl (bond dissociation energy, 511 kJ mol1), to minimize this eﬀect. However, that is not the only issue to consider, because the concentration of the competing species may also play a role. One element that is usually present at high levels in many types of samples and can interact with Cl is Ca (CaCl bond dissociation energy, 409 kJ mol1). Thus, this potential issue was investigated in more detail for varying amounts of Ca. Fig. 8A shows that, indeed, the presence of a high amount of Ca may represent a problem. As can be seen, for low levels of Ca the signal of AlCl is hardly aﬀected. However, for Ca/Cl molar ratios higher than 2, the AlCl signal tends to rise. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. 6 Correlation between the experimental (as obtained via HR CS GFMAS) and theoretical (as prepared by properly mixing two certiﬁed reference materials) 37Cl/35Cl ratio when monitoring the 262.222 nm (Al37Cl) and 262.238 nm (Al35Cl) transitions for solutions containing 200 ng of Cl. (A) without any mathematical deconvolution and (B) after mathematical deconvolution, as shown in Fig. 7. Error bars represent the standard deviation (n ¼ 5). This problem could be solved using signal deconvolution. However, most deconvolution approaches are not really suitable for this particular situation, because they are based on Gaussian or Lorentzian functions, which assume symmetric peak proles. That is clearly not the case in this work, as the peaks show a longer right tail (see Fig. 1). Thus, a diﬀerent model was evaluated to circumvent this problem, following the steps shown in Fig. 7. Fig. 7A shows the initial peak prole, which shows a potential overlap of the Al37Cl tail on the Al35Cl signal. Fig. 7B shows the same data in more detail, focusing in the range between pixel 91 and pixel 150, and also showing the individual detector pixels recorded. Aer examining this area, those values that could be aﬀected by the presence of the Al35Cl molecule are removed, and the rest of the data is tted using an asymptotic regression model (Fig. 7C). Then, nally, the contribution of the Al37Cl signal can be subtracted. Fig. 6 Correlation between the experimental (as obtained via HR CS GFMAS) and theoretical (as prepared by properly mixing two certiﬁed reference materials) 37Cl/35Cl ratio when monitoring the 262.222 nm (Al37Cl) and 262.238 nm (Al35Cl) transitions for solutions containing 200 ng of Cl. (A) without any mathematical deconvolution and (B) after mathematical deconvolution, as shown in Fig. 7. Error bars represent the standard deviation (n ¼ 5). Fig. 6B shows the results obtained when this correction is applied to those values with a 37/35 ratio higher than 1.5. As can be seen, a good agreement between theoretical and experi- mental data is now obtained for the entire interval investigated. This journal is © The Royal Society of Chemistry 2015 J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1537 View Article Online JAAS Fig. 7 Spectral overlap correction performed at 262.238 nm for a Cl (500 ng) spike with a 37Cl/35Cl molar ratio ¼ 2.5. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Even though ultimately the AlCl molecule is formed (which is logical because it is more stable and there is more Al available – 10 mg – than Ca), Ca seems to act like a chemical modier,42 helping in forming AlCl with more eﬃciency. However, for Ca/Cl ratios higher than 90 the signal for AlCl begins to decrease, until it reaches a point where AlCl is practically not formed owing to the higher availability of Ca atoms. Fig. 8 Study of the Ca interference. (A) Al35Cl absorbance signal, as recorded at 262.238 nm, and (B) Al35Cl/Al37Cl absorbance ratio as a function of the Ca content, for solutions containing 100 ng of Cl, 10 mg of Al and 20 mg of Pd. Error bars represent 2 s (n ¼ 5). Fig. 8 Study of the Ca interference. (A) Al35Cl absorbance signal, as recorded at 262.238 nm, and (B) Al35Cl/Al37Cl absorbance ratio as a function of the Ca content, for solutions containing 100 ng of Cl, 10 mg of Al and 20 mg of Pd. Error bars represent 2 s (n ¼ 5). While it could be interesting to further investigate on the vaporization mechanism of AlCl in the presence of Ca and Pd, such aspect is out of the scope of the present work. What is relevant here is to stress that the presence of a high but unknown amount of Ca in the sample will make it very diﬃcult to obtain accurate results when constructing the calibration curve with aqueous standards, as this element will have a clear impact on the AlCl absorbance nally obtained. Therefore, this model can be considered as suitable for spectral overlap correction. Therefore, this model can be considered as suitable for spectral overlap correction. This journal is © The Royal Society of Chemistry 2015 3.4 Cl determination in water using ID for calibration To explore the robustness of an ID method in comparison with a conventional calibration strategy, two diﬀerent natural water CRMs (ION-915 and KEJIM-02) were selected for analysis. For the latter approach, a calibration curve was prepared with 5 points ranging from 1 to 10 mg L1. The calibration curve showed a R2 coeﬃcient of 0.9998, a limit of detection (LOD) of 0.30 mg L1 and a limit of quantication (LOQ) of 1.0 mg L1. The results obtained for the samples are shown in Table 4. As can be seen, a result that is in good agreement with the Cl certied value was obtained for KEJIM-02, but the same was not true for the other sample, ION-915, for which a result biased high was found. This fact could be explained considering the Ca/Cl content of the samples, which is much higher in the case of ION-915 (13.7 mg L1 Ca) than that of KEJIM-02 (0.852 mg L1 Ca), thus likely inuencing the absolute AlCl absorbance signal, as demonstrated in the previous section. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Table 5 Cl determination by means of HR CS GFMAS (using ID for calibration) and of ICPMS in ﬁve diﬀerent commercial mineral water samples. Uncertainties are expressed as 95% conﬁdence intervals (n ¼ 3) Sample Cl concentration/mg L1 Label ID-HR CS GFMAS ICPMS Veri 1.9 1.99 0.15 1.87 0.07 Fonter 5.1 5.27 0.35 5.06 0.38 Sol´an de Cabras 7.4 7.33 0.30 7.17 0.30 Aquarel Abetos 8.5 8.52 0.17 8.08 0.52 Fontecabras 49.5 49.9 2.5 54.5 2.8 4 Conclusion This work explores for the rst time the potential of HR CS GFMAS for isotope analysis, focusing on Cl monitoring aer the formation of the AlCl molecule. The results obtained conrm that, by monitoring molecular instead of atomic spectra, it is possible to observe an isotopic shithat is large enough to appreciate two diﬀerent peaks, each one corresponding to an isotope, which are suﬃciently resolved to allow their quanti- cation, regardless of the chemical form in which Cl is present in the sample. Isotope analysis brings a new possibility to circumvent this kind of problem, which so far has been limited to MS tech- niques. Instead of using the cumbersome standard addition method, or matrix-matched standards, the use of isotope dilu- tion (ID) can help in successfully overcoming this type of interference. In fact, Fig. 8B shows that, despite the high uc- tuations observed for the individual AlCl signal for varying amounts of Ca, the value obtained for the Al35Cl/Al37Cl ratio remains very stable, which clearly indicates that ID may provide satisfactory results in this case. The method developed, based on the addition of Pd and Al, and the monitoring of peak height values, enables the isotopic analysis of Cl at the mg L1 level with precision values around 2% RSD, not requiring any method for mass bias correction. However, if a spike with a high 37/35 ratio is to be monitored and a spectral overlap is detected, it is necessary to perform a deconvolution to correct for this interference. Furthermore, it was demonstrated that using ID for cali- bration is feasible in this context and provides a novel – so far only used by MS techniques – and a very powerful strategy to determine Cl in samples in which otherwise chemical interfer- ences are detected, as demonstrated for water analysis. Acknowledgements This work has been funded by the Spanish Ministry of Economy and Competitiveness (project CTQ2012-33494) and the Arag´on Government (Fondo Social Europeo). The authors are also grateful to Capes/PDSE for the mobility grant under process BEX 4250/14-1. This journal is © The Royal Society of Chemistry 2015 1 R. B. Evans, Lung, 2004, 183, 151–167. 2 M. D. Huang, H. Becker-Ross, S. Florek, U. Heitmann and M. Okruss, Spectrochim. Acta, Part B, 2006, 61, 959–964. 3 A. Doyle, A. Saavedra, M. L. B. Trist˜ao, L. A. N. Mendes and R. Q. Auc´elio, Spectrochim. Acta, Part B, 2013, 86, 102–107. 4 M. Fechetia, A. L. Tognon and M. A. M. S. da Veiga, Spectrochim. Acta, Part B, 2012, 71–72, 98–101. 3.3 Ca interference One of the main diﬀerences between HR CS GFMAS and HR CS GFAAS is that the former is more prone to suﬀer from chemical This journal is © The Royal Society of Chemistry 2015 1538 \| J. Anal. At. Spectrom., 2015, 30, 1531–1540 View Article Online JAAS Paper Table 4 Determination of Cl by HR CS GFMAS when constructing the curve with Cl aqueous standard solutions or by means of isotope dilution (ID). Experimental uncertainties are expressed as 95% conﬁ- dence intervals (n ¼ 3) estimation of the LOD are discussed elsewhere.43 The LOD was estimated to be of 0.25 mg L1. As also shown in Table 4, the concentrations found by means of ID for both samples were in good agreement with the certied values, proving that this calibration technique can circumvent the interference detected in a simple way. To further test the robustness of the ID approach, ve commercial mineral water samples covering a wide range of Cl levels were also subjected to analysis. The results obtained are shown in Table 5. As can be seen, an excellent agreement with the expected values (as indicated by the producer in the label) was always attained. No diﬀerence could be drawn by means of the t-Student test (95% condence interval). The results obtained by means of ICPMS are also provided as a means of further validation. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. E. I. Muller and E. M. M. Flores, Anal. Bioanal. Chem., 2013, 405, 7615–7642. 28 B. Welz, S. Mor´es, E. Carasek, M. G. R. Vale, M. Okruss and H. Becker-Ross, Appl. Spectrosc. Rev., 2010, 45, 327–354. 9 ´E. R. Pereira, B. Welz, A. H. D. Lopez, J. S. de Gois, G. F. Caramori, D. L. G. Borges, E. Carasek and J. B. de Andrade, Spectrochim. Acta, Part B, 2014, 102, 1–6. 29 M. Resano and E. Garc´ıa-Ruiz, Anal. Bioanal. Chem., 2011, 399, 323–330. 10 Handbook of Environmental Isotope Geochemistry, ed. M. Baskaran, Springer-Verlag, Berlin-Heidelberg, 2012. 30 M. Resano, M. R. Fl´orez and E. Garc´ıa-Ruiz, Spectrochim. Acta, Part B, 2013, 88, 85–97. 31 B. Welz, M. G. R. Vale, ´E. R. Pereira, I. N. B. Castilho and M. B. Dessuy, J. Braz. Chem. Soc., 2014, 25, 799–821. 11 N. C. Sturchio, J. K. B¨ohlke, A. D. Beloso Jr, S. H. Streger, L. J. Heraty and P. B. Hatzinger, Environ. Sci. Technol., 2007, 41, 2796–2802. 32 M. Resano, M. Aramend´ıa and M. A. Belarra, J. Anal. At. Spectrom., 2014, 29, 2229–2250. 12 A. Sun, Q. Xu, S. Xu, H. Shen, J. Sun and Y. Zhang, Chem. Geol., 2014, 381, 21–25. 33 H. Wiltsche, K. Prattes, M. Zischka and G. Knapp, Spectrochim. Acta, Part B, 2009, 64, 341–346. 13 C. Luo, Y. Xiao, H. Wen, H. Ma, Y. Ma, Y. Zhang, Y. Zhang and M. He, Appl. Geochem., 2014, 47, 141–149. 34 B. Welz, F. G. Lepri, R. G. O. Araujo, S. L. C. Ferreira, M. D. Huang, M. Okruss and H. Becker-Ross, Anal. Chim. Acta, 2009, 647, 137–148. 14 J. Fietzke, M. Frische, T. H. Hansteen and A. Eisenhauer, J. Anal. At. Spectrom., 2008, 23, 769–772. 35 D. J. Butcher, Anal. Chim. Acta, 2013, 804, 1–15. 15 Y. Zakon, L. Halicz and F. Gelman, Anal. Chem., 2014, 86, 6495–6500. 36 M. Resano, M. R. Fl´orez and E. Garc´ıa-Ruiz, Anal. Bioanal. Chem., 2014, 406, 2239–2259. 16 C. Wiegert, M. Mandalakis, T. Knowles, P. N. Polymenakou, C. Aeppli, J. Mach´aˇckov´a, H. Holmstrand, R. P. Evershed, R. D. Pancost and ¨O. Gustafsson, Environ. Sci. Technol., 2013, 47, 6449–6456. 37 U. Heitmann, H. Becker-Ross, S. Florek, M. D. Huang and M. Okruss, J. Anal. At. Spectrom., 2006, 21, 1314–1320. 38 G. Herzberg, Molecular Spectra and Molecule Structure. I. Spectra of Diatomic Molecules, D. Van Nostrand, 2nd edn, New York, 1950. 17 M. Wu, Y. Qian, J. References 1 R. B. Evans, Lung, 2004, 183, 151–167. 2 M. D. Huang, H. Becker-Ross, S. Florek, U. Heitmann and M. Okruss, Spectrochim. Acta, Part B, 2006, 61, 959–964. 3 A. Doyle, A. Saavedra, M. L. B. Trist˜ao, L. A. N. Mendes and R. Q. Auc´elio, Spectrochim. Acta, Part B, 2013, 86, 102–107. 4 M. Fechetia, A. L. Tognon and M. A. M. S. da Veiga, Spectrochim. Acta, Part B, 2012, 71–72, 98–101. 1 R. B. Evans, Lung, 2004, 183, 151–167. 2 M. D. Huang, H. Becker-Ross, S. Florek, U. Heitmann and M. Okruss, Spectrochim. Acta, Part B, 2006, 61, 959–964. 3 A. Doyle, A. Saavedra, M. L. B. Trist˜ao, L. A. N. Mendes and R. Q. Auc´elio, Spectrochim. Acta, Part B, 2013, 86, 102–107. Alternatively, the same samples were also analyzed using the conditions shown in Table 1 and ID for calibration, as described in Section 2.3. The calculations for the ID technique and for the Alternatively, the same samples were also analyzed using the conditions shown in Table 1 and ID for calibration, as described in Section 2.3. The calculations for the ID technique and for the 4 M. Fechetia, A. L. Tognon and M. A. M. S. da Veiga, Spectrochim. Acta, Part B, 2012, 71–72, 98–101. This journal is © The Royal Society of Chemistry 2015 J. Anal. At. Spectrom., 2015, 30, 1531–1540 \| 1539 View Article Online JAAS Paper 5 J. S. de Gois, ´E. R. Pereira, B. Welz and D. L. G. Borges, Spectrochim. Acta, Part B, 2015, 105, 12–17. 25 X. Mao, A. A. Bol'shakov, I. Choi, C. P. McKay, D. L. Perry, O. Sorkhabi and R. E. Russo, Spectrochim. Acta, Part B, 2011, 66, 767–775. p 6 M. U. Rahim, X. Gao and H. Wu, Fuel, 2014, 129, 314–317. 26 B. Welz, H. Becker-Ross, S. Florek and U. Heitmann, High- Resolution Continuum Source AAS. The Better Way to Do Atomic Absorption Spectrometry, Wiley-VCH, Weinheim, 2005. 7 T. A. Labutin, A. M. Popov, S. M. Zaytsev, N. B. Zorov, 7 T. A. Labutin, A. M. Popov, S. M. Zaytsev, N. B. Zorov, M. V. Belkov, V. V. Kiris and S. N. Raikov, Spectrochim. Acta, Part B, 2014, 99, 94–100. M. V. Belkov, V. V. Kiris and S. N. Raikov, Spectrochim. Acta, Part B, 2014, 99, 94–100. 8 P. A. Mello, J. S. Barin, F. A. Duarte, C. A. Bizzi, L. O. Diehl, 27 M. This journal is © The Royal Society of Chemistry 2015 References Resano, F. Vanhaecke and M. T. C. de Loos-Vollebregt, J. Anal. At. Spectrom., 2008, 23, 1450–1475. Open Access Article. Published on 08 April 2015. Downloaded on 10/24/2024 5:05:29 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. M. Boyd, S. E. Hrudey, X. C. Le and X.-F. Li, J. Chromatogr. A, 2014, 1359, 156–161. 39 U. Heitmann, B. Welz, D. L. G. Borges and F. G. Lepri, Spectrochim. Acta, Part B, 2007, 62, 1222–1230. 18 Isotopic Analysis: Fundamentals and Applications Using ICP- MS, ed. F. Vanhaecke and P. Degryse, Wiley-VCH, Weinheim, 2012. 40 M. Resano, J. Brice˜no and M. A. Belarra, Spectrochim. Acta, Part B, 2009, 64, 520–529. 19 R. E. Russo, A. A. Bol'shakov, X. Mao, C. P. McKay, D. L. Perry and O. Sorkhabi, Spectrochim. Acta, Part B, 2011, 66, 99–104. 41 J. Brice˜no, M. A. Belarra, K. A. C. de Schamphelaere, S. Vanblaere, C. R. Janssen, F. Vanhaecke and M. Resano, J. Anal. At. Spectrom., 2010, 25, 503–510. 20 R. E. Russo, X. Mao, J. J. Gonzalez, V. Zorba and J. Yoo, Anal. Chem., 2013, 85, 6162–6177. 21 X. Mao, A. A. Bol'shakov, D. L. Perry, O. Sorkhabi and R. E. Russo, Spectrochim. Acta, Part B, 2011, 66, 604–609. 42 R. Nowka, K. Eichardt and B. Welz, Spectrochim. Acta, Part B, 2000, 55, 517–524. 43 M. Resano, M. Aramend´ıa and F. Vanhaecke, J. Anal. At. Spectrom., 2006, 21, 1036–1044. 22 A. Sarkar, X. Mao and R. E. Russo, Spectrochim. Acta, Part B, 2014, 92, 42–50. 23 M. Dong, X. Mao, J. J. Gonzalez, J. Lu and R. E. Russo, Anal. Chem., 2013, 85, 2899–2906. 44 R. W. B. Pearse and A. G. Gaydon, The Identication of Molecular Spectra, Chapman & Hall, 3rd edn, London, 1963. 24 A. Sarkar, X. Mao, G. C. Y. Chan and R. E. Russo, Spectrochim. Acta, Part B, 2013, 88, 46–53. 1540 \| J. Anal. At. Spectrom., 2015, 30, 1531–1540 This journal is © The Royal Society of Chemistry 2015
https://openalex.org/W3211081710	https://vbn.aau.dk/ws/files/452988498/Christiansen_2021_Special_issue_recurrent_pregnancy_loss.pdf	English	null	Special Issue Recurrent Pregnancy Loss: Etiology, Diagnosis, and Therapy	Journal of clinical medicine	2,021	cc-by	2,387	Aalborg Universitet Special issue recurrent pregnancy loss Etiology, diagnosis, and therapy Christiansen, Ole Bjarne Published in: Journal of Clinical Medicine DOI (link to publication from Publisher): 10.3390/jcm10215040 Creative Commons License CC BY 4.0 Publication date: 2021 Document Version Publisher's PDF, also known as Version of record Link to publication from Aalborg University Citation for published version (APA): Christiansen, O. B. (2021). Special issue recurrent pr Clinical Medicine , 10(21), Article 5040. https://doi.org Citation for published version (APA): Christiansen, O. B. (2021). Special issue recurrent pregnancy loss: Etiology, diagnosis, and therapy. Journal of Clinical Medicine , 10(21), Article 5040. https://doi.org/10.3390/jcm10215040 General rights C i h d General rights Copyright and moral rights for the publications made accessible in the public 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. - Users may download and print one copy of any publication from the public portal for the purpose of private study or research. - You may not further distribute the material or use it for any profit-making activity or commercial gain - Users may download and print one copy of any publication from the public portal for the purpose of private study Y t f th di t ib t th t i l it f fit ki ti it i l i - Users may download and print one copy of any publication from the public portal for the purpose of private study or research. - You may not further distribute the material or use it for any profit-making activity or commercial gain y p py y p p p p p p - You may not further distribute the material or use it for any profit-making activity or commercial gain - You may freely distribute the URL identifying the publication in the public portal - y p py y p p p - You may not further distribute the material or use it for any profit-making activity or co Y f l di t ib t th URL id tif i th bli ti i th bli t l Take down policy If you believe that this document breaches copyright please contact us at vbn@aub.aau.dk 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 at vbn@aub.aau.dk providing details, and we will remove access to the work immediately and investigate your claim. Citation for published version (APA): Christiansen, O. B. (2021). Special issue recurrent pregnancy loss: Etiology, diagnosis, and therapy. Journal of Clinical Medicine , 10(21), Article 5040. https://doi.org/10.3390/jcm10215040 Aalborg Universitet Special issue recurrent pregnancy loss Etiology, diagnosis, and therapy Christiansen, Ole Bjarne Published in: Journal of Clinical Medicine DOI (link to publication from Publisher): 10.3390/jcm10215040 Creative Commons License CC BY 4.0 Publication date: 2021 Document Version Publisher's PDF, also known as Version of record Link to publication from Aalborg University Citation for published version (APA): Christiansen, O. B. (2021). Special issue recurrent pregnancy loss: Etiology, diagnosis, and therapy. Journal of Clinical Medicine , 10(21), Article 5040. https://doi.org/10.3390/jcm10215040 Aalborg Universitet Editorial Special Issue Recurrent Pregnancy Loss: Etiology, Diagnosis, and Therapy Ole Bjarne Christiansen Department of Obstetrics and Gynecology, Aalborg University Hospital, 9000 Aalborg, Denmark; olbc@rn.dk The deﬁnition of recurrent pregnancy losses (RPL) varies between guidelines from different national and international scientiﬁc societies, but overall, a history of two or more (or alternatively, three or more) conﬁrmed pregnancy losses is required for the diagnosis. Depending on the deﬁnition, 0.5% to 2.5% of all women suffer from RPL. p g Most clinicians and researchers in the ﬁeld of RPL recognize that the cause of the condition is multifactorial, with factors such as repeated fetal chromosome anomalies, the presence of antiphospholipid antibodies and clinical thyroid disease being recognized as risk factors/causes of RPL. Conversely, there is limited consensus about the importance of other factors such as maternal natural killer (NK) cells, heritable thrombophilia factors, cervical insufﬁciency and immunity against male-speciﬁc (HY) antigens. The controversies about which risk factors to screen for and which treatments should be used are highlighted in the review by Vomstein et al. [1]. Several papers in this Special Issue highlight and try to clarify the controversial issues in the area of RPL. Strobel et al. [2] report that the NK cell number and NK cytotoxicity in peripheral blood are higher in patients with primary RPL compared to secondary RPL. This is an important new ﬁnding which may explain controversies in previous studies of NK cells in RPL patients where samples taken from primary RPL patients are typically compared with samples from parous controls. If a previous birth really affects the number of NK cells permanently, as suggested in the study by Strobel et al. [2], only RPL patients and controls with similar parity should be compared in future studies of NK cells. Citation: Christiansen, O.B. Special Issue Recurrent Pregnancy Loss: Etiology, Diagnosis, and Therapy. J. Clin. Med. 2021, 10, 5040. https:// doi.org/10.3390/jcm10215040 Received: 29 September 2021 Accepted: 26 October 2021 Published: 28 October 2021 Citation: Christiansen, O.B. Special Issue Recurrent Pregnancy Loss: Etiology, Diagnosis, and Therapy. J. Clin. Med. 2021, 10, 5040. https:// doi.org/10.3390/jcm10215040 Downloaded from vbn.aau.dk on: October 24, 2024 Downloaded from vbn.aau.dk on: October 24, 2024 Journal of Clinical Medicine Journal of Clinical Medicine Journal of Clinical Medicine 2021, 10, 5040 2 of 3 mechanisms through which intravenous immunoglobulin acts in the potential prevention of RPL. mechanisms through which intravenous immunoglobulin acts in the potential prevention of RPL. A subset of RPL patients experience second trimester loss or preterm birth under a clinical picture of painless cervical dilatation or rupture of the membranes, traditionally called cervical insufﬁciency. Vaginal cerclage often fails in these cases, but a cohort study by Chung et al. [5], where 21.1% of the included patients had suffered three or more previous pregnancy losses, suggest that modiﬁed laparoscopic cervico-isthmic cerclage can increase the fetal survival rate. However, randomized trials are still needed to document the real beneﬁts of this procedure. In the review by Wierzchowska-Opoka et al. [6], an overview is given of the procedures, outcomes and risk factors for failure after the application of emergency cerclage. This is, again, an area with no randomized controlled trials, although these would be difﬁcult to carry out. y Several studies, especially from Danish research groups, have reported that a previous birth of a boy predisposes patients to subsequent RPL compared with a previous birth of a girl, and there are some indications that immunity against male-speciﬁc HY antigens is responsible for this clinical observation. The family-based study by Noergaard-Pedersen et al. [7] suggests that not only the previous birth of a boy, but also a history of an older brother can predispose to secondary RPL: the explanation for this observation may be the persistence of microchimaeric cells transferred through the placenta of the common mother from the older brother to the younger sister. This is the ﬁrst (although indirect) evidence that microchimeric cells may play a role in the pathogenesis of RPL, although more studies are needed to conﬁrm this. During recent years, several studies have reported an increased incidence of obstetric and perinatal complications in women with a diagnosis of RPL. Most studies have been small, and there is still no strong consensus that RPL women are a high-risk population for adverse obstetrical and perinatal outcome. The large cohort study by Ticconi et al. [8] reports an increased incidence of intrauterine fetal death, preterm birth, fetal growth re- striction preeclampsia, and placental abruption in RPL patients compared with a non-RPL control group. In the register-based study by Roepke et al. [9] based on all RPL patients in Sweden and a randomly selected register-based control group, these observations are con- ﬁrmed. The risks of stillbirth, preterm birth, being small for gestational birth, preeclampsia, and placental abruption were all increased compared with controls with adjusted odds ratios between 1.45 and 2.47. There is now such overwhelming documentation that RPL patients are a high-risk population for obstetrical and perinatal complications that closer surveillance of the patients in late pregnancy must be introduced in clinical practice if not already done. The study by Shiina [10] falls outside the subject area of the above-mentioned studies since it does not deal with risk factors for RPL or treatment of miscarriage, but rather reviews the knowledge about a rare neo-vascular lesion that can develop after delivery or miscarriage. The signiﬁcance of this rare lesion is as yet uncertain. Overall, the articles in this Special Issue of the Journal of Clinical Medicine should provide the reader with an overview of highlights from the very active research going on in this important area. Funding: This research received no external funding. Conﬂicts of Interest: The authors declare no conﬂict of interest. Guidelines agree that RPL patients should be screened for the anti-phospholipid (aPL) antibodies (lupus anticoagulant, anti-cardiolipin antibodies and β2-glycoprotein-I antibod- ies), whereas screening for other aPL antibodies such as antiphosphatidyl (aPS)/prothrombin (PT) antibodies is not recommended. The study by Pleguezuelo et al. [3] seems to indicate that aPS/PT antibodies may have a stronger association to RPL than the traditional aPL antibodies, with limited overlap between them. The results of the study call for more studies to clarify the importance of the non-traditional aPL antibodies. Received: 29 September 2021 Accepted: 26 October 2021 Published: 28 October 2021 There has been a growing focus on the role of peripheral blood extracellular vesicles in many late pregnancy complications, such as preeclampsia and gestational diabetes, but they have not previously been investigated in RPL. The placenta is a source of many kinds of extracellular vesicles during pregnancy, and the composition of these biomarkers during pregnancy may reﬂect the growth and “health status” of the trophoblast tissue, which may predict later pregnancy complications. In the study by Rajaratnam et al. [4], extracellular vesicles were measured before and several times during early pregnancy in a cohort of women with RPL. It was found that the levels of one type of extracellular vesicle with the surface marker CD9 (a general marker for extracellular microvesicles) were lower before pregnancy and increased more rapidly during early pregnancy in RPL patients who went on to miscarry again compared with those who gave birth. This calls for further studies on extracellular vesicles as disease markers in RPL. The study by Rajaratnam et al. [4] also found that infusions with intravenous immunoglobulin profoundly increased the levels of almost all kinds of extracellular vesicles in RPL patients, which may be one of the Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Copyright: © 2021 by the author. 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/). Copyright: © 2021 by the author. 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/). https://www.mdpi.com/journal/jcm J. Clin. Med. 2021, 10, 5040. https://doi.org/10.3390/jcm10215040 J. Clin. Med. 1. Vomstein, K.; Feil, K.; Strobel, L.; Aulitzky, A.; Hofer-Tollinger, S.; Kuon, R.-J.; Toth, B. Immunological Risk Factors in Recurrent Pregnancy Loss: Guidelines versus Current State of Art. J. Clin. Med. 2021, 10, 869. [CrossRef] [PubMed] 1. Vomstein, K.; Feil, K.; Strobel, L.; Aulitzky, A.; Hofer-Tollinger, S.; Kuon, R.-J.; Toth, B. Immunological Risk Factors in Recurrent Pregnancy Loss: Guidelines versus Current State of Art. J. Clin. Med. 2021, 10, 869. [CrossRef] [PubMed] 2. Strobel, L.; Vomstein, K.; Kyvelidou, C.; Hofer-Tollinger, S.; Feil, K.; Kuon, R.-J.; Ebner, S.; Troppmair, J.; Toth, B. Different Background: Natural Killer Cell Proﬁles in Secondary versus Primary Recurrent Pregnancy Loss. J. Clin. Med. 2021, 10, 194. [CrossRef] [PubMed] g y 2. Strobel, L.; Vomstein, K.; Kyvelidou, C.; Hofer-Tollinger, S.; Feil, K.; Kuon, R.-J.; Ebner, S.; Troppmair, J.; Toth, B. Different Background: Natural Killer Cell Proﬁles in Secondary versus Primary Recurrent Pregnancy Loss. J. Clin. Med. 2021, 10, 194. [CrossRef] [PubMed] References References J. Clin. Med. 2021, 10, 5040 3 of 3 3. Pleguezuelo, D.E.; Cabrera-Marante, O.; Abad, M.; Rodriguez-Frias, E.A.; Naranjo, L.; Vazquez, A.; Villar, O.; Gil-Etayo, F.J.; Serrano, M.; Perez-Rivilla, A.; et al. Anti-Phosphatidylserine/Prothrombin Antibodies in Healthy Women with Unexplained Recurrent Pregnancy Loss. J. Clin. Med. 2021, 10, 2094. [CrossRef] [PubMed] g y 4. Rajaratnam, N.; Ditlevsen, N.E.; Sloth, J.K.; Baek, R.; Joergensen, M.M.; Christiansen, O.B. Extracel Biomarker in Recurrent Pregnancy Loss? J. Clin. Med. 2021, 10, 2549. [CrossRef] [PubMed] 5. Chung, H.; Lee, S.; Song, C.; Jang, T.-K.; Bae, J.-G.; Kwon, S.-H.; Shin, S.-J.; Cho, C.-H. Modiﬁed Laparoscopic, Transabdominal, Cervico—Isthmic Cerclage for the Surgical Management of Recurrent Pregnancy Loss due to Cervical Factors. J. Clin. Med. 2021, 10, 693. [CrossRef] [PubMed] 6. Wierzchowska-Opoka, M.; Kimber-Trojnar, Z.; Leszczynska-Gorzelak, B. Emergency Cervical Cerclage. J. Clin. Med. 2021, 10, 1270. [CrossRef] [PubMed] 7. Noergaard-Pedersen, C.; Kesmodel, U.; Christiansen, O.B. Women with Recurrent Pregnancy Loss h Brother and a Previous Birth of a Boy; is Male Microchimerism a Risk Factor? J. Clin. Med. 2021, 10, 26 7. Noergaard-Pedersen, C.; Kesmodel, U.; Christiansen, O.B. Women with Recurrent Pregnancy Loss have more often an Older Brother and a Previous Birth of a Boy; is Male Microchimerism a Risk Factor? J. Clin. Med. 2021, 10, 2613. [CrossRef] [PubMed] 8. Ticconi, C.; Pietropolli, A.; Specchia, M.; Nicastri, E.; Chiramonte, C.; Piccione, E.; Scambia, G.; Di Simone, N. Pregnancy-Related Complications in Women with Recurrent Pregnancy Loss: A Prospective Cohort Study. J. Clin. Med. 2020, 9, 2833. [CrossRef] [PubMed] y J [ ] [ ] 8. Ticconi, C.; Pietropolli, A.; Specchia, M.; Nicastri, E.; Chiramonte, C.; Piccione, E.; Scambia, G.; Di Simone, N. Pregnancy-Related Complications in Women with Recurrent Pregnancy Loss: A Prospective Cohort Study. J. Clin. Med. 2020, 9, 2833. [CrossRef] [PubMed] 9. Roepke, E.R.; Christiansen, O.B.; Källén, K.; Hansson, S.R. Women with a History of Recurrent Pregna Population for Adverse Obstetrical Outcome: A Retrospective Cohort Study. J. Clin. Med. 2021, 10, 179 10. Shiina, Y. Overview of Neo-Vascular Lesions after Delivery or Miscarriage. J. Clin. Med. 2021, 10, 1 10. Shiina, Y. Overview of Neo-Vascular Lesions after Delivery or Miscarriage. J. Clin. Med. 2021, 10, 1084. [CrossRef] [PubMed]
https://openalex.org/W4251779134	http://scielo.iics.una.py/pdf/spmi/v7n2/2312-3893-spmi-7-02-109.pdf	es		Postsplenectomy sepsis syndrome	Revista virtual de la Sociedad Paraguaya de Medicina Interna	2,020	cc-by	4,358	Doi:10.18004/rvspmi/2312-3893/2020.07.02.109 REPORTE DE CASO Síndrome de sepsis post esplenectomía Postsplenectomy sepsis syndrome a Fabricio Picoita1, bChristian Mora2, cVerónica Pinto2, dOscar Tabares2, e Fausto Gady Torres Toala3 1 Universidad Central del Ecuador. Hospital San Vicente de Paul. Quito, Ecuador 2 RESUMEN El síndrome de sepsis post esplenectomía define a todo cuadro de sepsis, sepsis grave, neumonía grave y meningitis secundaria a bacterias encapsuladas, virus, parásitos, protozoos, en pacientes con antecedentes de asplenia o hipoesplenismo. Se presenta el caso de un paciente masculino de 55 años con antecedente de esplenectomía hace 30 años sin inmunización posterior quien acude a una institución privada de la ciudad de Quito, Ecuador por atención tras presentar desde hace varios meses a nivel de piel petequias sin causa aparente. De manera súbita y tras la ingesta alimentaria presentó y desarrollo un síndrome gastrointestinal importante que motivó su ingreso a Emergencia y posteriormente a la Unidad de cuidados intensivos. Tras su estudio se encasilla al paciente bajo el cuadro de síndrome de sepsis post esplenectomía. La esplenectomía es un procedimiento quirúrgico frecuente en el país y es dado por varias causas entre ellas traumatismos abomínales severos. El caso que se presenta y desarrolla en esta oportunidad es un claro reflejo de la desinformación y falta de apego al seguimiento que todo paciente con asplenia e hipoesplenia debería tener. Esta patología tiene un índice de mortalidad elevado en las primeras horas y pude causar severa morbilidad o discapacidad si no se brinda el soporte necesario. Palabras claves: sepsis, esplenectomía, Streptococcus pneumoniae ABSTRACT Post-splenectomy sepsis syndrome defines all symptoms of sepsis, severe sepsis, severe pneumonia, and meningitis secondary to encapsulated bacteria, viruses, parasites, protozoa, in patients a Médico Tratante de la Unidad de Terapia Intensivos del Hospital de Especialidades Eugenio Espejo, Nova Clínica S.A, Profesor alterno Postgrado de Terapia Intensiva b Médico Residente de Terapia Intensiva Nova Clínica S.A. c Médico tratante Medicina Interna, Hospital Clínica Moderna d Médico Residente Terapia Intensiva, Hospital Clínica Moderna e Investigador principal Makroscopio Servicios de Salud, Docente PUCE Autor correspondiente Dr. Fausto Gady Torres Toala Correo electrónico: gtorres@suportamed.com ORCID: 0000-0002-8006-4447 Artículo recibido: 1 febrero 2020 Artículo aceptado: 23 abril 2020 Este es un artículo publicado en acceso abierto bajo una Licencia Creative Commons CC-BY 4.0 Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 109 Síndrome de sepsis post esplenectomía with a history of asplenia or hyposplenia. The case of a 55-year-old male patient, with a history of splenectomy performed 30 years ago without subsequent immunization, is presented. The patient attends a private institution in the city of Quito, Ecuador for care after presenting petechiae for several months at the skin level without apparent cause. Suddenly and after food intake, he presented and developed a major gastrointestinal syndrome that led to his admission into the Emergency Department and later to the Intensive Care Unit. After his study, the patient is classified under the picture of post-splenectomy sepsis syndrome. Splenectomy is a frequent surgical procedure in the country and is caused by various causes, including severe abdominal trauma. The case presented and developed on this occasion is a clear reflection of the misinformation and lack of adherence to followup that all patients with asplenia and hyposplenia should have. This pathology has a high mortality rate in the first hours and can cause severe morbidity or disability if the necessary support is not provided. Keywords: sepsis, splenectomy, Streptococcus pneumoniae INTRODUCCIÓN El síndrome de sepsis grave post esplenectomía (OPSI) define a todo cuadro de sepsis, neumonía grave y meningitis secundario a bacterias encapsuladas, virus, parásitos, protozoos, en pacientes con antecedentes de asplenia o hipoesplenismo (1). El hipoesplenismo y asplenia secundaria a una esplenectomía confiere un elevado riesgo de adquirir infecciones por microorganismos encapsulados (Streptococcus pneumoniae, Haemophilus influenzae tipo B, Neisseria meningitidis), 12 a 25 veces mayor en relación a la población general y que se asocian a una mortalidad superior al 80% (1-4). La incidencia en niños con asplenia o hipoesplenismo ha sido estimada en 1 de cada 175 al año, y en adultos 1 por cada 400-500 esplenectomizados al año. El mayor riesgo para desarrollar OPSI es durante los 2 primeros años posteriores a la esplenectomía, pero este se puede presentar hasta los 40 años posteriores. Por lo tanto, se considera que el riesgo de presentar sepsis en paciente asplénico es muy alto y permanente. El riesgo es mayor en menores de 16 años y pacientes cuya enfermedad subyacente se comporta como inmunodeficiencia, enfermedad hemática maligna o que recibe tratamientos que altere el estado inmunitario (5,6). Las complicaciones más frecuentes de la esplenectomía son infecciosas. Las mismas suelen iniciar en el postoperatorio, con una incidencia de hasta 40%, además. Además, pueden presentarse infecciones graves como OPSI que suele reportarse entre 3 a 5%. El riesgo de infección, aunque comienza inmediatamente en el postoperatorio y en los primeros dos años después del procedimiento, persiste toda la vida. El manejo riguroso de las inmunizaciones, la profilaxis antibiótica y la educación del paciente son indispensables (7-12). PRESENTACIÓN DEL CASO Paciente masculino de 55 años con antecedente de esplenectomía hace 30 años sin inmunización posterior ni seguimiento, quien presenta desde hace varios meses a nivel de piel petequias sin causa aparente. De manera súbita, tras la ingesta de comida copiosa (mariscos), presentó deposiciones líquidas en moderada cantidad con abundante moco, así como náusea que llega al vómito de contenido alimentario por 8-10 ocasiones. Se administró paracetamol sin lograr aliviar el cuadro, razón por la que acudió al servicio de emergencia de institución privada de la ciudad de Quito, Ecuador donde presenta, además de la sintomatología descrita, desorientación y dificultad respiratoria evidenciada por disnea con cianosis central. Debido a esto se decidió su ingreso a la Unidad de Terapia Intensiva donde presenta una presión arterial de 80/50 mm Hg, frecuencia cardiaca de 125 latidos por minuto, frecuencia respiratoria 30 respiraciones por minuto, temperatura axilar 110 Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 Síndrome de sepsis post esplenectomía 36,8 °C, acrocianosis (Gráfico 1), saturación O2 al aire ambiente 75% que con apoyo de O2 por mascarilla incrementa hasta 85%. El control gasométrico muestra tendencia a la acidosis respiratoria y se observa inadecuada mecánica ventilatoria por lo que requirió asistencia ventilatoria. Gráfico 1. Necrosis distal miembro superior al inicio de la sepsis Estudios de laboratorio al ingreso demuestran hemoglobina 17,5 g/dL, hematocrito 56,3%, plaquetas 95.000/mm3, leucocitos 14.000/mm3, neutrófilos 90%, linfocitos 5%, cayados 4%, tiempo de protrombina 29,0 segundos, tiempo de tromboplastina parcial 76,9 segundos, INR 3,02, glucosa 119 mg/dL, creatinina 4,20 mg/dL, BUN 35,47 mg/dL, lactato 10, bilirrubina total 4,05 mg/dL, bilirrubina indirecta 0,60 mg/dL, bilirrubina directa 3,45 mg/dL, procalcitonina mayor a 100 ng/mL (tabla 1). Se considera el diagnóstico de choque séptico post esplenectomía más disfunción multiorgánica. Se implementa terapéutica guiada por metas basadas en la campaña “Sobreviviendo a la sepsis” con infusión endovenosa de cristaloides, requerimiento de soporte con fármaco vasopresor a base de norepinefrina. Se inició esquema antibiótico empírico con meropenem 1 gramo endovenoso cada 8 horas y vancomicina 1 gramo endovenoso diario. Tabla 1. Seguimiento de la respuesta inflamatoria por laboratorio. En reporte de hemocultivos se evidencia Streptococcus pneumoniae, por lo que se decide desescalar antibioticoterapia a base de penicilina cristalina. Con los datos obtenidos se documenta que el paciente presenta un síndrome de OPSI por la sintomatología previa y el estado de choque desencadenado por el proceso infeccioso en un paciente esplenectomizado sin profilaxis neumocócica. Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 111 Síndrome de sepsis post esplenectomía El paciente permaneció con apoyo vasopresor durante 72 horas, soporte de ventilación mecánica invasiva por 10 días, inició terapia de remplazo renal (diálisis) a los 7 días de inicio de cuadro clínico por perpetuación de fallo renal. En su evolución desarrolla cuadro de pancreatitis aguda. A los siete días posteriores al retiro de ventilación mecánica invasiva presenta signos de insuficiencia respiratoria severa por lo que fue necesario el reingreso a ventilación mecánica invasiva permaneciendo en la misma por 10 días con diagnóstico de neumonía nosocomial tardía más sepsis por catéter. Se inició terapia antimicrobiana empírica en base a ureidopenicilina, quinolonas y oxazolidinonas. La filiación microbiológica de hemocultivos realizados a los 7 días reporta Staphylococcus aureus resistente a meticilina, se continua antibioticoterapia propuesta. Como parte de su evolución y dentro de las complicaciones del síndrome desarrolla púrpura cutánea en cara, brazos y piernas desde su ingreso que posteriormente se tornan de características necróticas en manos, tercio distal de antebrazo, así como en tercio medio de muslo hasta tercio inferior de piernas (gráfico 2). Por la severidad de la necrosis y ante la necesidad de plantear amputación a nivel de extremidades se realiza angiotomografía de miembros superiores e inferiores que reporta vascularización arterial en el lado derecho hasta la región de la muñeca principalmente por la arteria radial hasta la región del metacarpo y en el lado izquierdo hasta el tercio medio del antebrazo. En los miembros inferiores se logra visualizar de manera adecuada la vasculatura distal por medio de las tibiales anterior y posterior de ambos lados delimitándose parcialmente los arcos plantares. Existe importante edema de los miembros inferiores y superiores, con alteración de realce en las áreas distales, compatible con defectos de perfusión. Gráfico 2. Lesiones necróticas de manos y piernas Debido a la progresión rápida de necrosis distal de miembros superiores e inferiores se decide realizar amputación hasta tercio medio distal de brazo y mano derecha, con formación de muñón y amputación hasta tercio medio distal de húmero izquierdo con formación de muñón, previa autorización de paciente y familiares. A pesar de las limpiezas quirúrgicas de lesiones instauradas en miembros inferiores (gráfico 3) al igual que en los miembros superiores, la progresión de necrosis es rápida, por lo que se decide realización de amputación hasta tercio medio bilateral con formación de muñón respectivo. En el postquirúrgico mediato el paciente presenta evolución favorable, sin complicaciones ni presencia de infecciones en muñones ni heridas quirúrgicas. Desarrolla nuevo evento de repunte de infección, se considera nuevo episodio de sepsis asociada a catéter que se documenta en hemocultivo con la presencia de nuevo germen Gram negativo de tipo Enterobacter aerogenes sensible a tigacilina, con lo que completo esquema antibiótico. Presento buena evolución y fue dado de alta vivo. 112 Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 Síndrome de sepsis post esplenectomía Gráfico 3. Lesiones necróticas de miembros inferiores DISCUSIÓN El bazo es el principal órgano linfoide en el cuerpo. Los sinusoides que posee el bazo en su estructura filtran la sangre que circula a través de capilares los que secuestran eritrocitos senescentes y dañados de la circulación sanguínea, además de patógenos y toxinas exógenas. Las células mononucleares que se encuentran dentro de esta red capilar fagocitan a las bacterias, virus, parásitos y hongos circulantes, en particular los microorganismos no opsonizados. A dicho nivel se encuentra una gran cantidad de linfocitos B y T con un sistema de opsoninas importante que incluyen los receptores de tipo Toll, lectinas de tipo C, I y receptores scavenger tipo II, los receptores de los macrófagos con estructura colágena, los CD36, el receptor de manosa y otros como los factores del complemento. La ausencia de todos estos receptores predispone al paciente al desarrollo de bacteriemias fuera de la circulación esplénica, los antígenos polisacáridos son poco inmunogénicos en comparación con los antígenos proteicos (3,13-15). Esto contribuye a que la cobertura de polisacáridos de las bacterias pueda evadir la respuesta a receptores que participan en la presentación y opsonización de los microorganismos desempeñando un papel crítico en el control temprano de las infecciones bacterianas, virales, fúngicas y parasitarias (7,11,16). Con su fagocitosis posterior, los pacientes asplénicos pueden desarrollar sepsis con cualquier tipo de microorganismo especialmente relacionada con microorganismos encapsulados como el Streptococcus pneumoniae (mismo que participa en más del 50% de los casos), Haemophilus influenzae, Neisseria meningitidis, parásitos o bacterias intraeritrocíticos, tales como Babesia, Ehrlichia, Plasmodium, Enterococcus, Bacteroides, Salmonella, Bartonella y otras especies que ocasionalmente se han informado ser responsables de OPSI. Recientemente, también se han reportado casos debido a bacterias Gram negativas como Capnocytophaga canimorsus, transmitida a través de mordeduras de perros, gatos u otros animales. Todos estos patógenos resisten a la fagocitosis pero puede ser rápidamente superados en presencia de pequeñas cantidades de anticuerpos de tipo específico excretados a nivel del bazo. En la asplenia o el hipoesplenismo la producción de anticuerpos ante un nuevo antígeno está deteriorada y las bacterias proliferan rápidamente (7,13,17). El patógeno más peligroso para individuos asplénicos es el S. pneumoniae, al menos el 87% de infecciones son causados por esta bacteria y tan solo el 30% de pacientes han recibido la vacuna antes Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 113 Síndrome de sepsis post esplenectomía de presentar OPSI. La inmunización frente a esta bacteria ha sido aceptada como una estrategia preventiva frente a infecciones en la población asplénica. Es el agente más frecuente asilado en hemocultivos (50-90%) de los pacientes con sepsis tras esplenectomía, con tendencia aumentar con la edad y en niños con hipofunción esplénica por la enfermedad de células falciformes. En general son los serotipos 6,14,18,19 y 23. Presenta una tasa de sepsis letal de 2-7%, lo que supone un riesgo 540 veces superior al de la población general, con especial incidencia en menores de 15 años. Actualmente disminuye la incidencia por la vacunación infantil con anti-HIB (4,13,18-20). Cuadro clínico Los síntomas son tardíos para el síndrome OPSI. Estos incluyen fatiga, pigmentación de la piel, pérdida de peso, dolor abdominal, diarrea, estreñimiento, náusea, cefalea. La neumonía y meningitis son los síntomas más frecuentes en sepsis severa, para lo que es imprescindible realizar un adecuado diagnóstico diferencial (tabla 2). Tabla 2. Enfermedades asociadas con hipoesplenismo o atrofia esplénica (15) Causas congénitas: Hipoesplenismo aislado congénito Síndrome de Ivemark Síndrome de poliendocrinopatíacandidiasis-distrofia ectodérmica Autoinmunitaria Hipoparatiroidismo Síndrome de Stormorken Enfermedades gastrointestinales: Enfermedad celiaca Enfermedad inflamatoria intestinal Enfermedad de Whipple Dermatitis herpetiforme Linfangiectasia intestinal Enteritis ulcerativa crónica idiopática Enfermedades infecciosas: VIH-SIDA Meningitis neumocócica Malaria Enfermedades hepáticas Hepatitis crónica activa Cirrosis biliar primaria Cirrosis hepática e hipertensión portal Enfermedades onco-hematológicas: Hemoglobina S Trasplante de médula ósea Enfermedad de injerto contra huésped crónica Leucemia aguda Enfermedades mieloproliferativas crónicas Formas iatrogénicas Exposición a metildopa Esteroides a dosis altas Nutrición parenteral total Irradiación a bazo Enfermedades autoinmunitarias: Lupus eritematoso generalizado Artritis reumatoide Glomerulonefritis Granulomatosis de Wegener Síndrome Goodpasture Síndrome de Sjögren Poliarteritis nodosa Tiroiditis Sarcoidosis Alteraciones en la circulación esplácnica: Trombosis de la arteria esplénica Trombosis de la vena esplénica Trombosis del tronco celiaco Misceláneas: Amiloidosis El cuadro es variado: puede ser de presentación lenta que simula un resfriado común en la fase temprana de la evolución, o un cuadro de sepsis sin foco evidente, de pródromo corto que progresa a sepsis grave. Una vez instalada la infección presenta una evolución muy agresiva y de rápido 114 Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 Síndrome de sepsis post esplenectomía progreso, pasando por el coma y la muerte entre 24 y 48 horas, presentando choque séptico, hipoglicemia, acidosis marcada, hiponatremia, distrés respiratorio y coagulación vascular diseminada. Cuadros sobreañadidos muy comunes encontrados en autopsias son el Síndrome de Waterhouse-Friderichsen, insuficiencia de las glándulas suprarrenales debido a sangrado dentro de dichas glándulas y es causado por una infección meningocócica severa u otra infección bacteriana grave (2,3,21,22). Tratamiento El punto crítico en el manejo es el reconocimiento temprano del paciente en riesgo, seguido por un tratamiento inmediato y agresivo. Todos los pacientes asplénicos con fiebre de origen desconocido deben ser tratados como una emergencia médica. La penicilina intravenosa ha sido el tratamiento de elección, debido a que provee una excelente actividad contra neumococo y meningococo. Con el cambio en la patente de resistencia, algunos autores proponen el inicio del tratamiento con un antibiótico de amplio espectro, enfatizando el empleo de drogas con actividad contra S. pneumoniae, H. influenzae y N. meningitidis. En este sentido, la ceftriaxona constituye la droga de elección (13,21,23-26). El déficit de volumen intravascular se debe corregir en forma agresiva. Otras modalidades terapéuticas, tales como el empleo de vasopresores o heparina para el manejo de la CID pueden estar indicadas en casos seleccionados. Se ha recomendado el empleo de corticoides hasta poder establecer la integridad de la función suprarrenal y en pacientes con meningitis (27). Si el paciente sobrevive al episodio séptico inicial, las complicaciones de la CID pueden hacer necesaria la amputación de partes distales de las extremidades debido a la necrosis isquémica. La cirugía está indicada en presencia de gangrena de los miembros que pueden actuar como focos secundarios de infección. Las áreas de necrosis de la piel serán tratadas de la misma manera que en los quemados, con desbridamiento e implante secundario. En ocasiones puede ser necesario realizar fasciotomías. Las intervenciones quirúrgicas rara vez están indicadas en el manejo inicial, y las amputaciones, si son necesarias, se realizarán cuando el paciente haya superado la etapa de resucitación (13,28). Prevención En la última década diversas publicaciones han llamado la atención sobre la relevancia de esta complicación infecciosa y la importancia de distintas medidas preventivas, que se han agrupado de forma sistemática en forma de guías por parte de comités y sociedades científicas. Dichas medidas preventivas incluyen la profilaxis antibiótica, la administración de vacunas y la educación sanitaria de los pacientes. La utilización de antibióticos profilácticos es muy controvertida, debido a la falta de datos clínicos que avalen su eficacia, a la dificultad que supone tomar antibióticos de forma continuada durante años y al problema creciente de la resistencia antibiótica. En este sentido, debe tenerse en cuenta que la profilaxis debe ir dirigida fundamentalmente contra el neumococo, cuyas tasas de resistencia a la penicilina y otros antibióticos se hallan en franco aumento. De hecho, ya se han comunicado diversos casos de sepsis fulminante causados por cepas de neumococo con resistencia múltiple. Algunos autores recomiendan la profilaxis antibiótica en los niños, por lo menos durante dos años después de la esplenectomía, pero sin una clara evidencia científica; en los adultos, la tendencia creciente es la de no realizar profilaxis, sino recomendar la automedicación inmediata con antibióticos activos frente a neumococo ante la presencia de fiebre o cualquier sospecha de infección, antes de acudir a la consulta médica (4,7,8). Una pauta razonable consiste en la toma de 3 g de amoxicilina oral; es obvio que las personas en riesgo deben disponer del antibiótico en su domicilio. La inmunización mediante vacunas es una estrategia preventiva aceptada de forma universal en estos pacientes. La vacunación antineumocócica, que es la más importante, se realiza mediante la administración de la vacuna polisacárida 23 valente, que incluye más del 75% de serotipos causantes de infección en esta población. Es recomendable vacunar al paciente por lo menos dos semanas antes Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 115 Síndrome de sepsis post esplenectomía de la esplenectomía, ya que se cree que la eficacia de la vacuna es menor si se administra después de la intervención, así como si se administra a pacientes sometidos a quimioterapia. Asimismo, se recomienda la revacunación cada 3 a 5 años (18). Existen pocos estudios controlados que demuestren la eficacia de la vacuna antineumocócica en esta población. Los dos estudios más convincentes, uno en 280 niños y otro en 200 adultos, ambos realizados en Dinamarca, demostraron que la vacunación sistemática de todos los pacientes se asoció a la ausencia de casos de sepsis fulminante. Los únicos cuatro casos de sepsis observados en un período de 13 años fueron producidos por serotipos de neumococo no incluidos en la vacuna. En general, se acepta que la respuesta de anticuerpos no es óptima en los pacientes esplenectomizados y que en alrededor del 20% de ellos es débil o inexistente, como ocurre en otros pacientes inmunodeprimidos y en la población muy anciana. La detección de fallos vacunales, mediante la determinación seriada del título de anticuerpos, podría ser útil para identificar población no protegida y plantear estrategias alternativas. En el futuro, el uso de la vacuna antineumocócica conjugada, con mayor capacidad inmunogénica, puede suponer una ventaja. Asimismo, es conveniente plantearse la administración de las vacunas conjugadas frente al meningococo A y C y frente a H. influenzae tipo B, que es ya de uso habitual en la infancia en muchos países (13,19). CONCLUSIÓN La esplenectomía es un procedimiento quirúrgico frecuente en el país y es dado por varias causas entre ellas por traumatismos abdominales severos. El caso que lo presentamos y desarrollamos en esta oportunidad es un claro reflejo de la desinformación y falta de apego al seguimiento que todo paciente con asplenia e hipoesplenia debe tener y sobre todo conocer cuáles son los signos de alarma para poder acudir y recibir atención oportuna para evitar una catástrofe orgánica que pueda causar la muerte o dejar serias secuelas funcionales como es el resultado de nuestro paciente. Debería ser una política y por qué no desarrollar un programa nacional de registro, información, educación, vigilancia y prevención de todos los pacientes con esplenectomía o hipoesplenismo para reducir el riesgo de esta grave complicación, porque como se ha insistido en este trabajo, la morbilidad la discapacidad y mortalidad resultan inaceptablemente elevadas para una condición prevenible y tratable cuando se diagnostica de manera temprana y oportuna. Por lo mencionado sería importante y creemos muy necesario se desarrolle en todo el país los criterios de profilaxis, tratamiento y seguimiento de los enfermos esplenectomizados por cualquier etiología o con hipoesplenismo, para de una manera más oportuna poder brindar una prevención adecuada y tratamiento oportuno a nuestros pacientes. Conflictos de interés Ninguno destacado por los autores. REFERENCIAS BIBLIOGRÁFICAS 1. Waghorn DJ. Overwhelming infection in asplenic patients: current best practice preventive measures are not being followed. J Clin Pathol [Internet]. 2001 [cited 2019 Dic 22]; 54(3):214-8. Available from: https://jcp.bmj.com/content/54/3/214 2. Picoita Camacho FF, Mora Ch, Tabares O. Síndrome de sepsis grave post esplenectomía (OPSI), reporte de un caso más revisión de la literatura. Nova Clínica S.A. U Investiga [Internet]. 2017 [citado 22 Dic 2019]; 4(2):44 – 54. Disponible en: http://revistasojs.utn.edu.ec/index.php/ lauinvestiga/article/view/254 116 Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 Síndrome de sepsis post esplenectomía 3. Jenks PJ, Jones E. Infections in asplenic patients. Clin Microbiol Infect [Internet]. 1996 [cited 2019 Dic 22]; 1(4):266-72. Available from: https://www.clinicalmicrobiologyandinfection.com/ action/showPdf?pii=S1198-743X%2815%2960286-5 4. Kanhutu K, Jones P, Cheng AC, Grannell L, Best E, Spelman D. Spleen Australia guidelines for the prevention of sepsis in patients with asplenia and hyposplenism in Australia and New Zealand. Intern Med J. 2017; 47(8):848-55. 5. Sumaraju V, Smith LG, Smith SM. Infectious complications in asplenic hosts. Infect Dis Clin North Am. 2001; 15(2):551-65, x. 6. Kyaw MH, Holmes EM, Toolis F, Wayne B, Chalmers J, Jones IG, Campbell H. Evaluation of severe infection and survival after splenectomy. Am J Med. 2006; 119(3):276.e1-7. 7. Ashorobi D, Fernandez R. Asplenia [Internet]. En: StatPearls. Treasure Island (FL): StatPearls Publishing; 2019 [cited 2019 Dic 22]. Available from: http://www.ncbi.nlm.nih.gov/books/ NBK538171/ 8. Rubin LG, Schaffner W. Clinical practice. Care of the asplenic patient. N Engl J Med. 2014; 371(4):349-56. 9. Holdsworth RJ, Cuschieri A, Irving AD. Postsplenectomy sepsis and its mortality rate: Actual versus perceived risks. Br J Surg [Internet]. 1991 [cited 2019 Dic 22]; 78(9):1031-8. 10. Schwartz PE, Sterioff S, Mucha P, Melton LJ, Offord KP. Postsplenectomy sepsis and mortality in adults. JAMA. 1982; 248(18):2279-83. 11. Bisharat N, Omari H, Lavi I, Raz R. Risk of infection and death among post-splenectomy patients. J Infect. 2001; 43(3):182-6. 12. Ahmed R, Devasia AJ, Viswabandya A, Lakshmi KM, Abraham A, Karl S, et al. Long-term outcome following splenectomy for chronic and persistent immune thrombocytopenia (ITP) in adults and children. Ann Hematol. 2016; 95(9):1429-34. https://doi.org/10.1007/s00277-016-2738-3 13. Davidson RN, Wall RA. Prevention and management of infections in patients without a spleen. Clin Microbiol Infect. 2001; 7(12):657-60. 14. Pernar LIM, Tavakkoli A. Chapter 136 - Anatomy and physiology of the spleen. En: Yeo ChJ, editor. Shackelford's surgery of the alimentary tract, 2 Volume Set. 8th Edition. Philadelphia: Content Repository Only!; 2019. p. 1591-7. 15. Kapila V, Tuma F. Physiology, Spleen [Internet]. En: StatPearls. Treasure Island (FL): StatPearls Publishing; 2019 [citado 2019 dic 22]. Available from: http://www.ncbi.nlm.nih.gov/ books/NBK537307/ 16. Buzelé R, Barbier L, Sauvanet A, Fantin B. Medical complications following splenectomy. J Visc Surg. 2016; 153(4):277-86. 17. Moffett SL. Overwhelming postsplenectomy infection: managing patients at risk. JAAPA. 2009; 22(7):36-9, 45. 18. Bonanni P, Grazzini M, Niccolai G, Paolini D, Varone O, Bartoloni A, et al. Recommended vaccinations for asplenic and hyposplenic adult patients. Hum Vaccin Immunother [Internet]. 2017 [cited 2019 Dic 22]; 13(2):359-68. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC5328222/pdf/khvi-13-02-1264797.pdf 19. Centers for Disease Control and Prevention. Vaccination of Adults: Asplenia and adult vaccination. [Internet]. USA: CDC; 2016 [cited 2019 Dic 22]. Available from: https://www.cdc.gov/ vaccines/adults/rec-vac/health-conditions/asplenia.html 20. Lamsfus-Prieto JÁ, Membrilla-Fernández E, Garcés-Jarque JM. Prevención de la sepsis en pacientes esplenectomizados. Cir Esp [Internet]. 2007 [citado 22 Dic 2019]; 81(5):247-51. Disponible en: http://webs.fcm.unc.edu.ar/wp-content/blogs.dir/30/files/2018/03/ Prevenci%C3%B3n-de-la-sepsis-en-pacientes-esplenectomizados.pdf 21. Abildgaard N, Nielsen JL. Pneumococcal septicaemia and meningitis in vaccinated splenectomized adult patients. Scand J Infect Dis. 1994; 26(5):615-7. 22. Carrillo Esper R, Peña Pérez CA, Carrillo Córdova JR, Carrillo Córdova LD, Carrillo Córdova CA, Carrillo Córdova DM. Síndrome de sepsis grave postesplenectomía secundario a meningococcemia. Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118 117 Síndrome de sepsis post esplenectomía Rev Fac Med (Méx) [Internet]. 2014 [citado 22 Dic 2019]; 57(2):31-41. Disponible en: http://www.scielo.org.mx/scielo.php?script=sci_abstract&pid=S0026-17422014000200031&lng =es&nrm=iso&tlng=es 23. Palange P, Rohde G, Editors. ERS handbook of respiratory medicine. European Respiratory Society; 2019. 24. Luu S, Spelman D, Woolley IJ. Post-splenectomy sepsis: preventative strategies, challenges, and solutions. Infect Drug Resist [Internet]. 2019 [cited 2019 Dic 22]; 12:2839-51. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6748314/ 25. Sinwar PD. Overwhelming post splenectomy infection syndrome – Review study. Int J Surg [Internet]. 2014 [cited 2019 Dic 22];12(12):1314-6. Available from: https://reader.elsevier.com/ reader/sd/pii/S1743919114009650?token=DC9FC2BC7DBB1EED6C0D6657033644CD7B411F25B ABC5CE534F7E072CA8EA611DF68F527DB2FF1C8B1C2BB4F7A84C319 26. Hashimoto N. Management of overwhelming postsplenectomy infection syndrome. Clin Surg [Internet]. 2016 [cited 2019 Dic 22]; 1: 1148-1148. Available from: file:///C:/Users/Acer/ Downloads/cis-v1-id1148.pdf 27. Dendle C, Sundararajan V, Spelman T, Jolley D, Woolley I. Splenectomy sequelae: an analysis of infectious outcomes among adults in Victoria. Med J Aust [Internet]. 2012 [cited 2019 Dic 22]; 196(9):582-6. Available from: https://www.mja.com.au/system/files/issues/196_09_210512/ den10909_fm.pdf 28. Bennett JE, Dolin R, Blaser MJ, editors. Enfermedades infecciosas: Principios y práctica. 8ª Edición. España : Elsevier; 2015. 118 Rev. virtual Soc. Parag. Med. Int. setiembre 2020; 7 (2):109-118
https://openalex.org/W2088307367	https://ir.lib.hiroshima-u.ac.jp/47064/files/37404	English	null	Generation of Dipeptidyl Peptidase-IV-Inhibiting Peptides from<i>β</i>-Lactoglobulin Secreted by<i>Lactococcus lactis</i>	BioMed research international	2,014	cc-by	6,705	Hindawi Publishing Corporation BioMed Research International Volume 2014, Article ID 393598, 8 pages http://dx.doi.org/10.1155/2014/393598 Hindawi Publishing Corporation BioMed Research International Volume 2014, Article ID 393598, 8 pages http://dx.doi.org/10.1155/2014/393598 1 Interdisciplinary Graduate School of Science and Technology, Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan 1 Interdisciplinary Graduate School of Science and Technology, Shinshu University, 8304 Minamiminowa, Kamiin Nagano 399-4598, Japan 2 Japan Society for the Promotion of Science (JSPS), Japan 5 Frontier Agriscience and Technology Center (FAST), Shinshu University, 8304 Minamiminowa, Kamiina, Nagan 6 5 Frontier Agriscience and Technology Center (FAST), Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4 6Graduate School of Agriculture, Department of Interdisciplinary Genome Sciences and Cell Metabolism, Institute for Biomedical Sciences, Interdisciplinary Cluster for Cutting Edge Research (ICCER), Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan Graduate School of Agriculture, Department of Interdisciplinary Genome Sciences and Cell Metabolism, 6Graduate School of Agriculture, Department of Interdisciplinary Genome Sciences and Cell Metabolism, Institute for Biomedical Sciences, Interdisciplinary Cluster for Cutting Edge Research (ICCER), Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan f g p f p y Institute for Biomedical Sciences, Interdisciplinary Cluster for Cutting Edge Research (ICCER), Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan Correspondence should be addressed to Takeshi Shimosato; shimot@shinshu-u.ac.jp Correspondence should be addressed to Takeshi Shimosato; shimot@shinshu-u.ac.jp Received 2 July 2014; Accepted 15 July 2014; Published 3 August 2014 Academic Editor: Julio Villena Academic Editor: Julio Villena Copyright © 2014 Suguru Shigemori 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. Previous studies showed that hydrolysates of 𝛽-lactoglobulin (BLG) prepared using gastrointestinal proteases strongly inhibit dipeptidyl peptidase-IV (DPP-IV) activity in vitro. In this study, we developed a BLG-secreting Lactococcus lactis strain as a delivery vehicle and in situ expression system. Interestingly, trypsin-digested recombinant BLG from L. lactis inhibited DPP-IV activity, suggesting that BLG-secreting L. lactis may be useful in the treatment of type 2 diabetes mellitus. Research Article Generation of Dipeptidyl Peptidase-IV-Inhibiting Peptides from 𝛽-Lactoglobulin Secreted by Lactococcus lactis Suguru Shigemori,1,2 Kazushi Oshiro,3 Pengfei Wang,3 Yoshinari Yamamoto,3 Yeqin Wang,1 Takashi Sato,4 Yutaka Uyeno,5 and Takeshi Shimosato1,3,6 1 Interdisciplinary Graduate School of Science and Technology, Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan 2 Japan Society for the Promotion of Science (JSPS), Japan 3 Graduate School of Agriculture, Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan 4Department of Internal Medicine and Clinical Immunology, Graduate School of Medicine, Yokohama City University, 3-9 Fukuura Kanagawa, Yokohama 236-0004, Japan 5 Frontier Agriscience and Technology Center (FAST), Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399 6Graduate School of Agriculture, Department of Interdisciplinary Genome Sciences and Cell Metabolism, Institute for Biomedical Sciences, Interdisciplinary Cluster for Cutting Edge Research (ICCER), Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan Correspondence should be addressed to Takeshi Shimosato; shimot@shinshu-u.ac.jp Received 2 July 2014; Accepted 15 July 2014; Published 3 August 2014 Academic Editor: Julio Villena Copyright © 2014 Suguru Shigemori et al. This is an open access article distributed under the Creative Common License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original wo cited. Previous studies showed that hydrolysates of 𝛽-lactoglobulin (BLG) prepared using gastrointestinal proteases st dipeptidyl peptidase-IV (DPP-IV) activity in vitro In this study we developed a BLG-secreting Lactococcus lactis strai 2. Materials and Methods dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE) sample buffer was added to each sample. The cell and supernatant fractions were boiled for 5 min or placed at room temperature overnight, respectively, and then subjected to SDS-PAGE (15% (v/v) polyacrylamide). 2.1. Bacterial Strains and Growth Conditions. L. lactis NZ9000 (NZ9000; MoBiTec, Goettingen, Germany), derived from L. lactis subsp. cremoris MG1363 (MG1363), which harbors the regulatory genes nisR and nisK integrated into the pepN gene, was used as a host strain. NZ9000 was grown at 30∘C without shaking in M17 broth (OXOID, Hampshire, UK) supplemented with 0.5% glucose (GM17). rNZ9000 strains were grown in GM17 with 20 𝜇g/mL of chloramphenicol. Electrophoresed proteins were transferred from the gel onto a Hybond-P polyvinylidene difluoride membrane (GE Healthcare, Buckinghamshire, UK). Western blotting was performed with primary antibody (Ab) against the 6x His- tag (1:1,000; BioLegend Inc., San Diego, CA, USA), followed by incubation with a horseradish peroxidase-conjugated secondary Ab (1:5,000; Sigma, St. Louis, MO, USA). The resulting blots were visualized using an ECL Western Blotting Analysis System (GE Healthcare). Signals were detected using a lumino image analyzer (ImageQuant LAS 4000 mini, GE Healthcare) and analyzed using ImageQuant TL software (GE Healthcare). 2.2. Construction of rNZ9000 Strains. To create the secre- tion vector for L. lactis, the sequence of usp45, the lac- tococcal signal peptide, was inserted into pNZ8148#2:CYT [13] between the nisin-inducible promoter (𝑃nis) and 6x His-tag sequences; the resulting plasmid was designated pNZ8148#2:SEC (Figure 1(a)). The pNZ8148#2:CYT vector is a modified form of the L. lactis expression vector pNZ8148 (MoBiTec).h 2.2. Construction of rNZ9000 Strains. To create the secre- tion vector for L. lactis, the sequence of usp45, the lac- tococcal signal peptide, was inserted into pNZ8148#2:CYT [13] between the nisin-inducible promoter (𝑃nis) and 6x His-tag sequences; the resulting plasmid was designated pNZ8148#2:SEC (Figure 1(a)). The pNZ8148#2:CYT vector is a modified form of the L. lactis expression vector pNZ8148 (MoBiTec).h 2.2. Construction of rNZ9000 Strains. To create the secre- tion vector for L. lactis, the sequence of usp45, the lac- tococcal signal peptide, was inserted into pNZ8148#2:CYT [13] between the nisin-inducible promoter (𝑃nis) and 6x His-tag sequences; the resulting plasmid was designated pNZ8148#2:SEC (Figure 1(a)). The pNZ8148#2:CYT vector is a modified form of the L. lactis expression vector pNZ8148 (MoBiTec).h 2.4. Purification of rBLG. Expression of rBLG in NZ9000:SEC-BLG was induced in a 4 L large-scale culture according to the above-mentioned methods. 1. Introduction Incretins drive insulin secretion in pancreatic 𝛽cells and suppress pancreatic glucagon production [10, 11]. Thus, DPP- IV inhibitors, such as vildagliptin, saxagliptin, sitagliptin, alogliptin, and linagliptin, are used in the management of type 2 diabetes mellitus (T2D). 𝛽-Lactoglobulin (BLG) is the major whey protein in the milk of cows and other ruminants, as well as some nonruminants, but it is not contained in human breast milk [1]. BLG is widely recognized as a high-functional protein [2]. Previous studies have consistently demonstrated that hydrolysates of BLG prepared using gastrointestinal proteases such as trypsin or pepsin have an inhibitory effect on dipeptidyl peptidase- IV (DPP-IV, also known as CD26) activity [3–8]. DPP-IV is a serine protease that is ubiquitously expressed in vivo, and its endogenous physiological substrates are incretins [9]. The incretins, primarily glucose-dependent insulinotropic polypeptide and glucagon-like peptide-1, are gastrointestinal hormones released from intestinal L and K cells, respectively. We hypothesized that intestinal mucosa-targeted oral delivery of BLG might be a useful strategy for managing T2D. Generally recognized as safe (i.e., GRAS), genetically modified lactic acid bacteria (LAB) have recently emerged as vehicles for the efficient delivery of therapeutic proteins and as stable producers of these proteins in situ [12]. In this study, we engineered a Lactococcus (L.) lactis strain to secrete BLG and validated the DPP-IV-inhibiting activity of trypsin- digested recombinant BLG (rBLG). BioMed Research International 2 2. Materials and Methods Following induction of rBLG expression, cells were harvested by centrifugation for 20 min at 3,000 ×g and 4∘C. The cell pellet was washed twice with phosphate buffer (50 mM NaH2PO4, 300 mM NaCl, pH 8.0) and resuspended with 80 mL of phosphate buffer containing EDTA-free proteinase inhibitor cocktail (Roche Diagnostics, Indianapolis, IN, USA). To disrupt the cells, the bacterial suspension was homogenized with 0.2 mm glass beads for 3 min using a beads beater (Beads Crasher 𝜇T-12, Taitec, Saitama, Japan) operating at 3,200 rpm. Soluble proteins released from the bacterial cells were collected by centrifugation for 15 min at 20,400 ×g and 4∘C.hi The gene encoding BLG (accession number: EU883598) was synthesized and subcloned into pGEM-T easy opti- mized for MG1363 codon-usage by Eurofins Genomics (Tokyo, Japan). The sequence of the BLG gene was subse- quently cloned between the KpnI and HindIII restriction sites in pNZ8148#2:SEC, generating pNZ8148#2:SEC-BLG (Figure 1(b)).h The pNZ8148#2:SEC-BLG vector was introduced into NZ9000 by electroporation using a Gene Pulser Xcell elec- troporation system (Bio-Rad Laboratories Inc., CA, USA) fol- lowing the manufacturer’s instructions. The resulting recom- binant strain was designated NZ9000:SEC-BLG. NZ9000 was also electroporated with the empty plasmid pNZ8148#2:SEC to generate an NZ9000 vector control strain (NZ9000:SEC- VC). The rBLG was purified from the prepared lysate using a HisTrap HP column (1 mL, precharged with Ni2+; GE Healthcare) under native conditions. The phosphate buffer- equilibrated column was filled with the prepared lysate and then the flow-through was collected. Next, nonadsorbed proteins were eluted using wash buffer (phosphate buffer containing 20 mM imidazole). Adsorbed proteins were eluted with elution buffers (phosphate buffer containing 31, 63, 125, 250, or 500 mM imidazole). The prepurification crude lysate and the flow-through and eluted fractions were analyzed by Western blotting with an anti-6x His-tag Ab. Eluted fractions were dialyzed against Tris-buffered saline (TBS; 50 mM Tris-HCl (pH 8.0), 138 mM NaCl, and 2.7 mM KCl) and frozen at −80∘C until use. The protein concentration of the dialyzed samples was measured with a BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA) before the samples were frozen. 2.3. Small-Scale Expression of rBLG. A 1.9 mL aliquot of fresh medium was inoculated with 0.1 mL of an overnight culture and incubated at 30∘C without shaking. 2. Materials and Methods When the turbidity of the culture reached an optical density at 600 nm (OD600) of 0.4 (1.5 to 2 h later), various concentrations of nisin were added and the culture was incubated for an additional 3 to 4 h. The OD600 of the culture was measured at the end of the incubation period. p Cells were isolated from the nisin-induced culture via centrifugation for 1 min at 8,000 ×g and 4∘C. Protein extracts were prepared from the cells using previously described methods [14]. Proteins remaining in the supernatant were isolated and concentrated using trichloroacetic acid (TCA) precipitation. Briefly, 300 𝜇L of cold 100% (w/v) TCA solution was added to 1.5 mL of supernatant and the mixture was placed on ice for 3 h. The resulting precipitates were pelleted by centrifugation at 20,400 ×g and 4∘C. To completely remove the TCA, the protein pellet was washed twice with ice-cold acetone. The resulting pellet was dried at 55∘C and then dissolved in a small amount of TE buffer (10 mM Tris-HCl (pH 8.0), 1 mM EDTA). An equal volume of 2x sodium 2.5. Trypsin Digestion of BLG. Commercial BLG (cBLG; Sigma) was dissolved in TBS at a concentration of 1 mg/mL. Purified TBS-dialyzed rBLG was diluted to 50 𝜇g/mL with TBS. Trypsin (from porcine pancreas, Wako Pure Chemical Industries Ltd., Osaka, Japan) was added at an enzyme to substrate ratio of 1 : 20 (w/w) and the mixture was incubated at 37∘C for 24 h with gentle agitation. 2. Materials and Methods AAG GAG GCA CTC ACC ATG AAA AAA AAG ATT ATC TCA GCT ATT TTA ATG TCT ACA GTG ATA CTT TCT GCT GCA GCC CCG TTG TCA GGT GTT TAC GCT GCC ATG GAA AGA GGA TCG CAT CAT CAC CAC CAC CAT GGA TCT GGC TCT GGA TCT GGT ATC GAG GGA AGG CCT TAT AAT GGA ACT GGA TCC GCA GGT ACC CCG GGT CGA CCT GCA GCC AAG CTT AAT TAG CTG AGC TTG GAC TCC TGT CTA GAG AGC TCA AGC TTT CTT TGA ACC AAA ATT AG TGC GAG CTC RBS Start 6x His-tag FXa Stop KpnI HindIII SPusp45 SPusp45 T cm repC repA pNZ8148#2:SEC 3,284 bp Pnis (a) T cm repC repA pNZ8148#2:SEC 3,284 bp Pnis (a) (a) FXa FXa His- tag His- tag RBS RBS MCS BLG KpnI HindIII SPusp45 SPusp45 Pnis Pnis (b) (b) Figure 1: Maps of vectors used in this study. (a) Schematic representation of the secretion vector, pNZ8148#2:SEC. The DNA sequence between the RBS and the stop codon is presented in the rectangle. (b) The codon-optimized BLG gene (537 bp) was cloned into pNZ8148#2:SEC (above) between the KpnI and HindIII restriction enzyme recognition sites, and the resulting plasmid was designated pNZ8148#2:SEC-BLG (below). 𝑃nis: nisin-inducible promoter; RBS: ribosome binding site; SP𝑢𝑠𝑝45: sequence of the signal peptide of the usp45 gene product; His- tag: 6x histidine-tag; FXa: factor Xa recognition site; MCS: multiple cloning site; T: terminator; rep: replication gene; cm: chloramphenicol acetyltransferase gene; BLG: 𝛽-lactoglobulin. by heating the sample at 90∘C for 5 min. The reaction mixture was kept frozen at −80∘C until used in experiments. 2.7. cBLG-Immunized Mice. Pathogen-free female BALB/c mice (4 weeks of age, 𝑛= 3) were purchased from Japan SLC (Shizuoka, Japan) and housed under temperature- and light- controlled conditions. Mice were fed a standard diet (MF, Oriental Yeast Co., Ltd., Tokyo, Japan) and sterile water ad libitum. After a 1-week preliminary acclimatization period, mice were sensitized with 100 𝜇g of cBLG emulsified with 0.1 mL of complete Freund’s adjuvant (Difco Laboratories, Detroit, MI, USA) and injected intraperitoneally. At 2 and 4 weeks after the first sensitization, mice were boosted with 100 𝜇g of cBLG emulsified with 0.1 mL of incomplete Freund’s adjuvant (Difco Laboratories). All experimental procedures 2.6. Analysis of DPP-IV Inhibition. The DPP-IV inhibition assay was performed using a DPP-IV Drug Discovery Kit according to the manufacturer’s instructions. 2. Materials and Methods Digestion was stopped BioMed Research International 3 AAG GAG GCA CTC ACC ATG AAA AAA AAG ATT ATC TCA GCT ATT TTA ATG TCT ACA GTG ATA CTT TCT GCT GCA GCC CCG TTG TCA GGT GTT TAC GCT GCC ATG GAA AGA GGA TCG CAT CAT CAC CAC CAC CAT GGA TCT GGC TCT GGA TCT GGT ATC GAG GGA AGG CCT TAT AAT GGA ACT GGA TCC GCA GGT ACC CCG GGT CGA CCT GCA GCC AAG CTT AAT TAG CTG AGC TTG GAC TCC TGT CTA GAG AGC TCA AGC TTT CTT TGA ACC AAA ATT AG TGC GAG CTC RBS Start 6x His-tag FXa Stop KpnI HindIII SPusp45 SPusp45 AAG GAG GCA CTC ACC ATG AAA AAA AAG ATT ATC TCA GCT ATT TTA ATG TCT ACA GTG ATA CTT TCT GCT GCA GCC CCG TTG TCA GGT GTT TAC GCT GCC ATG GAA AGA GGA TCG CAT CAT CAC CAC CAC CAT GGA TCT GGC TCT GGA TCT GGT ATC GAG GGA AGG CCT TAT AAT GGA ACT GGA TCC GCA GGT ACC CCG GGT CGA CCT GCA GCC AAG CTT AAT TAG CTG AGC TTG GAC TCC TGT CTA GAG AGC TCA AGC TTT CTT TGA ACC AAA ATT AG TGC GAG CTC RBS Start 6x His-tag FXa Stop T cm repC repA pNZ8148#2:SEC KpnI HindIII SPusp45 SPusp45 3,284 bp Pnis (a) FXa FXa His- tag His- tag RBS RBS MCS BLG KpnI HindIII SPusp45 SPusp45 Pnis Pnis (b) Figure 1: Maps of vectors used in this study. (a) Schematic representation of the secretion vector, pNZ8148#2:SEC. The DNA sequence b the RBS and the stop codon is presented in the rectangle. (b) The codon-optimized BLG gene (537 bp) was cloned into pNZ8148 (above) between the KpnI and HindIII restriction enzyme recognition sites, and the resulting plasmid was designated pNZ8148#2:SE (below). 𝑃nis: nisin-inducible promoter; RBS: ribosome binding site; SP𝑢𝑠𝑝45: sequence of the signal peptide of the usp45 gene produ tag: 6x histidine-tag; FXa: factor Xa recognition site; MCS: multiple cloning site; T: terminator; rep: replication gene; cm: chloramp acetyltransferase gene; BLG: 𝛽-lactoglobulin. were conducted in accordance with the Regulations for Animal Experimentation of Shinshu University. were conducted in accordance with the Regulations for Animal Experimentation of Shinshu University. 2.9. Statistical Analysis. ANOVA and post hoc tests were performed using the ystat2004.xls statistical software package (Igakutosho Shuppan, Tokyo, Japan). One-way ANOVA with the post hoc Dunnett test or Student-Newman-Keuls test was used to determine the significance of differences in DPP- IV inhibitory assay and immunoreactivity assay, respectively. Differences were considered significant at 𝑃< 0.05. 2.8. Immunoreactivity Assay. To investigate whether rBLG is immunoreactive, splenocytes were isolated from cBLG- immunized mice and the level of interleukin- (IL-) 13 mRNA expression was measured using real-time quantitative PCR (qPCR) following stimulation with purified rBLG. Spleno- cytes were prepared using standard methods [15, 16] and then cultured at a final concentration of 1 × 107 cells/well (total volume of 2 mL per well). Splenocytes were stimulated with 50 or 100 𝜇g/mL of cBLG or purified rBLG. After 72 h of incubation, cells were harvested and IL-13 mRNA expression was assessed using qPCR as previously described [16]. Fluorescent qPCR was performed with SYBR Premix Ex Taq (TaKaRa Bio Inc., Tokyo, Japan) and IL-13-specific primers, with each reaction containing 5 ng of cDNA in a total volume of 25 𝜇L. The 𝛽-actin-specific and IL-13-specific primers were purchased from TaKaRa Bio Inc. The PCR cycling conditions were 10 s at 95∘C followed by 45 cycles of 5 s at 95∘C and 30 sec at 60∘C. As a control, poly (A)+ RNA samples were used as templates to check for the presence of contaminating genomic DNA. Reaction sensitivity and amplification of contaminant products (e.g., extension of self- annealed primers) were evaluated by amplifying serial dilu- tions of the cDNA template. For cross-sample comparisons of results obtained using various treatments, cytokine mRNA levels were first normalized to 𝛽-actin mRNA levels. Results are presented as the mean and SD of three independent experiments. 2. Materials and Methods Fluorescence resulting from degradation of substrate (H-Gly-Pro-AMC) by DPP-IV was detected using a Fluoroskan Ascent FL Microplate Fluorometer (Thermo Scientific) at excitation and emission wavelengths of 355 and 460 nm, respectively. Data are expressed as DPP-IV activity (%) remaining in test samples versus the control (no sample added). 4 BioMed Research International 45.1 32.2 26.8 17.2 NZ9000:SEC- NZ9000:SEC- VC BLG Nisin + + − − MW (kDa) (a) 45.1 32.2 26.8 17.2 NZ9000:SEC- NZ9000:SEC- VC BLG Nisin + + − − MW (kDa) (b) Figure 2: Secretion of rBLG by NZ9000. Two rNZ9000 strains were cultured with (+) or without (−) 10 ng/mL of nisin. Protein extracts from cells (a) or culture supernatants (b) were analyzed by Western blotting with a 6x His-tag Ab. White and black arrowheads indicate the secretory rBLG precursor (pre-rBLG, 27 kDa) and the secretory form of rBLG (24.3 kDa), respectively. 45.1 32.2 26.8 17.2 NZ9000:SEC- NZ9000:SEC- VC BLG Nisin + + − − MW (kDa) (b) 45.1 32.2 26.8 17.2 NZ9000:SEC- NZ9000:SEC- VC BLG N + + − − MW (kDa) (a) (b) Figure 2: Secretion of rBLG by NZ9000. Two rNZ9000 strains were cultured with (+) or without (−) 10 ng/mL of nisin. Protein extracts from cells (a) or culture supernatants (b) were analyzed by Western blotting with a 6x His-tag Ab. White and black arrowheads indicate the secretory rBLG precursor (pre-rBLG, 27 kDa) and the secretory form of rBLG (24.3 kDa), respectively. 3. Results and Discussion All fractions were analyzed by Western blotting with a 6x His-tag Ab. White and black arrowheads indicate the secretory rBLG precursor (pre-rBLG, 27 kDa) and the secretory form of rBLG (24.3 kDa), respectively. Figure 4: Purification of rBLG secreted by NZ9000. Expression of rBLG by NZ9000 was induced in a 4 L large-scale culture. The cell extract (crude lysate) was prepared as described in Section 2 and then passed through a HisTrap HP column (flow-through). The column was washed to remove nonadsorbed protein (wash) and then eluted (elution) with wash buffer containing 20 mM imidazole and elution buffers containing 31 to 500 mM imidazole. All fractions were analyzed by Western blotting with a 6x His-tag Ab. White and black arrowheads indicate the secretory rBLG precursor (pre-rBLG, 27 kDa) and the secretory form of rBLG (24.3 kDa), respectively. 3.2. Optimal Conditions for rBLG Secretion by NZ9000. Previous studies suggested that the amount of a protein expressed in an engineered LAB may be increased in a nisin- concentration-dependent manner using the NICE system [24]. We addressed this possibility in our previous research [13, 14]. In this study, we show that rBLG expression and secretion are nisin-concentration dependent (Figure 3(b)). Inhibition of bacterial proliferation (Figure 3(a)) and an increase in the intensity of a band corresponding to pre- rBLG in the culture supernatant (Figure 3(b)) were observed in NZ9000:SEC-BLG cultured with nisin at concentrations greater than 16 ng/mL. in significant cell damage. Based on these results, 15 ng/mL was determined to be the optimal nisin concentration for maximizing rBLG secretion. 3.3. Purification of rBLG Produced by NZ9000. Expression of rBLG by strain NZ9000:SEC-BLG was induced in a 4 L large-scale culture, after which the protein was isolated and purified. The cellular crude lysate was passed through a HisTrap HP column and the flow-through was collected. The column was washed to remove nonadsorbed proteins and the adsorbed proteins were eluted with various concentrations of imidazole (20 to 500 mM). To confirm the purification of rBLG, all fractions were analyzed by Western blotting with an anti-6x His-tag Ab (Figure 4). Bands corresponding to pre-rBLG and/or rBLG were observed in each of the eluted fractions. Pre-rBLG and rBLG were particularly abundant in the fractions eluted with 63 and 125 mM imidazole. However, highly purified rBLG samples (i.e., fractions eluted Nisin, a typical type-A(I) lantibiotic produced by some L. 3. Results and Discussion 3.1. Construction of the rBLG-Secreting NZ9000 Strain. Genetically engineered LAB have recently emerged as suit- able vehicles for the delivery of therapeutic proteins (e.g., antigens of pathogenic bacteria, allergen, and immunomod- ulators such as cytokines) to the intestinal mucosa [17]. Many heterologous gene expression systems for LAB have been developed [18–20]. The nisin-controlled gene expression (NICE) system, which utilizes a nisin-inducible promoter to express downstream genes [21], is one of the most efficient expression systems and is commonly used with L. lactis and other LAB [22, 23]. In the current study, we constructed the secretion plasmid pNZ8148#2:SEC (Figure 1(a)), which is a modified form of pNZ8148, the standard plasmid used in the NICE system. The pNZ8148#2:SEC plasmid contains sequences encoding the nisin-inducible promoter (𝑃nis), usp45 signal peptide (SP𝑢𝑠𝑝45), 6x His-tag, a Factor Xa recognition site, and a terminator. This plasmid was cloned from the unmu- tated BLG gene (Figure 1(b)) and then introduced into NZ9000, resulting in the generation of NZ9000:SEC-BLG. The NZ9000:SEC-BLG strain was induced with nisin, and BioMed Research International 5 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 2 4 8 16 31 63 125 250 500 OD600 Nisin (ng/mL) (a) Nisin (ng/mL) 0 2 4 8 16 31 63 125 250 500 Cell Sup (b) Figure 3: Determination of the optimal nisin concentration for secretion of rBLG by NZ9000. (a) Strain NZ9000:SEC-BLG was cultured with various concentrations of nisin (0 to 500 ng/mL). After 3 h of incubation, the OD600 of each sample was measured. The mean and SD from three independent experiments are shown. (b) Protein extracts from cells (CELL, above) or culture supernatants (SUP, below) were analyzed by Western blotting with a 6x His-tag Ab. Representative images from three independent experiments are shown. White and black arrowheads indicate the secretory rBLG precursor (pre-rBLG, 27 kDa) and the secretory form of rBLG (24.3 kDa), respectively. BioMed Research International BioMed Research International 5 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 2 4 8 16 31 63 125 250 500 OD600 Nisin (ng/mL) (a) Nisin (ng/mL) 0 2 4 8 16 31 63 125 250 500 Cell Sup (b) (b) (a) Figure 3: Determination of the optimal nisin concentration for secretion of rBLG by NZ9000. (a) Strain NZ9000:SEC-BLG was cultured with various concentrations of nisin (0 to 500 ng/mL). 3. Results and Discussion After 3 h of incubation, the OD600 of each sample was measured. The mean and SD from three independent experiments are shown. (b) Protein extracts from cells (CELL, above) or culture supernatants (SUP, below) were analyzed by Western blotting with a 6x His-tag Ab. Representative images from three independent experiments are shown. White and black arrowheads indicate the secretory rBLG precursor (pre-rBLG, 27 kDa) and the secretory form of rBLG (24.3 kDa), respectively. Imidazole (mM) Wash Elution Crude lysate Flow-through 20 31 63 125 250 500 Figure 4: Purification of rBLG secreted by NZ9000. Expression of rBLG by NZ9000 was induced in a 4 L large-scale culture. The cell extract (crude lysate) was prepared as described in Section 2 and then passed through a HisTrap HP column (flow-through). The column was washed to remove nonadsorbed protein (wash) and then eluted (elution) with wash buffer containing 20 mM imidazole and elution buffers containing 31 to 500 mM imidazole. All fractions were analyzed by Western blotting with a 6x His-tag Ab. White and black arrowheads indicate the secretory rBLG precursor (pre-rBLG, 27 kDa) and the secretory form of rBLG (24.3 kDa), respectively. Imidazole (mM) Wash Elution Crude lysate Flow-through 20 31 63 125 250 500 rBLG expression and secretion were validated by Western blotting with an anti-6x His-tag Ab. A strong secretory rBLG precursor signal (pre-rBLG, 27 kDa) and a weak band repre- senting the secretory form of rBLG (24.3 kDa) were observed on Western blotting of the cell extract of nisin-induced NZ9000:SEC-BLG (Figure 2(a)). Moreover, a single band corresponding to rBLG was detected in the culture super- natant of nisin-induced NZ9000:SEC-BLG (Figure 2(b)). No signal was detected in analyses of protein extracts of the nisin- induced NZ9000:SEC-VC and noninduced NZ9000 strains (Figure 2). We clearly demonstrated that NZ9000:SEC-BLG intracellular expression of pre-rBLG is dependent on nisin stimulation and that this protein is subsequently secreted by the cell’s secretory machinery [18]. Figure 4: Purification of rBLG secreted by NZ9000. Expression of rBLG by NZ9000 was induced in a 4 L large-scale culture. The cell extract (crude lysate) was prepared as described in Section 2 and then passed through a HisTrap HP column (flow-through). The column was washed to remove nonadsorbed protein (wash) and then eluted (elution) with wash buffer containing 20 mM imidazole and elution buffers containing 31 to 500 mM imidazole. 3. Results and Discussion lactis strains, exhibits bactericidal activity against a wide range of Gram-positive bacteria [25]. The mechanism of nisin’s antimicrobial action is thought to involve (i) inhibition of cell wall biosynthesis by scavenging of the peptidoglycan precursor lipid II and (ii) lysis of the cell membrane by the formation of pores [26]. Our observations suggest that although the level of rBLG production by NZ9000:SEC-BLG increases in a nisin-concentration-dependent manner, expo- sure to high nisin concentrations (i.e., >16 ng/mL) results BioMed Research International 6 105 100 95 90 85 80 Control 1 10 100 DPP-IV activity (%) ∗ ∗∗ ∗∗ cBLG (𝜇g/mL) (a) 105 100 95 90 85 80 DPP-IV activity (%) ∗∗ ∗∗ Control cBLG rBLG (b) Figure 5: Inhibition of DPP-IV activity by trypsin-digested rBLG. Various concentrations of trypsin-digested cBLG (a) or trypsin hydrolysates of 50 𝜇g/mL of either cBLG or purified rBLG (b) were evaluated for inhibitory activity against DPP-IV using a DPP-IV Drug Discovery Kit. Concentrations indicated are the protein concentrations before trypsin treatment. Data are expressed as percent activity remaining in test samples versus that in the control (no sample added). Black, dotted white, and dotted black bars indicate control, cBLG, and purified rBLG (rBLG), respectively. Values represent means and error bars indicate SD (𝑛= 3). ∗𝑃< 0.05; ∗∗𝑃< 0.01 versus control. 105 100 95 90 85 80 Control 1 10 100 DPP-IV activity (%) ∗ ∗∗ ∗∗ cBLG (𝜇g/mL) (a) 105 100 95 90 85 80 DPP-IV activity (%) ∗∗ ∗∗ Control cBLG rBLG (b) (b) (a) Figure 5: Inhibition of DPP-IV activity by trypsin-digested rBLG. Various concentrations of trypsin-digested cBLG (a) or trypsin hydrolysates of 50 𝜇g/mL of either cBLG or purified rBLG (b) were evaluated for inhibitory activity against DPP-IV using a DPP-IV Drug Discovery Kit. Concentrations indicated are the protein concentrations before trypsin treatment. Data are expressed as percent activity remaining in test samples versus that in the control (no sample added). Black, dotted white, and dotted black bars indicate control, cBLG, and purified rBLG (rBLG), respectively. Values represent means and error bars indicate SD (𝑛= 3). ∗𝑃< 0.05; ∗∗𝑃< 0.01 versus control. with 250 and 500 mM imidazole) were used in subsequent experiments. we previously developed [13, 14]. In this study, IL-13 mRNA expression was significantly enhanced in a dose-dependent manner in splenocytes stimulated with cBLG as compared with medium-stimulated control cells (Supplemental Figure 1 available online at http://dx.doi.org/10.1155/2014/393598). Conflict of Interests The authors declare that there is no conflict of interests regarding the publication of this paper. 4. Conclusions We developed a L. lactis strain (NZ9000 : SEC-BLG) that efficiently secretes rBLG. The rBLG produced by this strain demonstrates both strong allergenicity and inhibi- tion of DPP-IV activity following trypsin digestion. The NZ9000:SEC-BLG strain may prove useful in therapies for treating T2D and allergies to cow’s milk. 3.5. Immunoreactivity of rBLG. Vaccines using genetically modified LAB as mucosal antigen delivery vehicles have been tested in mouse models of allergies to house dust mites [27–29], eggs [30], birch pollen [31], Japanese cedar pollen [32], and cow’s milk [33]. These studies demon- strated suppression of the allergic response by LAB vaccines through induction of immune response by T-helper (Th)1 or regulatory T cells. Indeed, Adel-Patient et al. showed that oral pretreatment with a BLG-producing L. lactis strain prevents Th2-type immune responses through a reduction in BLG-specific IgE production and the upregulation of specific Th1-type IgG2a and fecal IgA production [34]. There- fore, we also investigated the immunoreactivity of L. lactis- produced rBLG using a simple in vitro assay system that 3. Results and Discussion Similar results were observed with cells stimulated with purified rBLG (Supplemental Figure 1). These results indicate that rBLG demonstrates immunoreactivity as strong as that of cBLG. Hence, NZ9000:SEC-BLG may be useful not only in managing T2D but also in therapies to treat or prevent allergies to cow’s milk. 3.4. Inhibition of DPP-IV Activity by Trypsin-Digested rBLG. Trypsin-digested cBLG inhibited the activity of the DPP-IV enzyme in a concentration-dependent manner (Figure 5(a)). The inhibitory activity of trypsin-digested rBLG against DPP- IV was similar to that of cBLG (Figure 5(b)). Uchida et al. also showed that DPP-IV inhibitory activity of trypsin-digested BLG was exerted in a concentration-dependent fashion, and IC50 value was 210 𝜇M [8]. Thus, highly concentrated tryptic hydrolysate of rBLG may efficiently inhibit DPP-IV activity. Other studies have shown that some peptides produced by trypsin digestion of BLG inhibit DPP-IV activity [6, 8]. Particularly, the pentapeptide IPAVF exhibits the strongest inhibitory activity [6]. The strong inhibition of DPP-IV activity observed with trypsin-digested rBLG in this study was probably because of the involvement of these peptides within the rBLG. References Li, G. X. Ma, and Y. J. Pan, “The nisin- controlled gene expression system: construction, application and improvements,” Biotechnology Advances, vol. 24, no. 3, pp. 285–295, 2006. [7] G. Tulipano, V. Sibilia, A. M. Caroli, and D. Cocchi, “Whey proteins as source of dipeptidyl dipeptidase IV (dipeptidyl peptidase-4) inhibitors,” Peptides, vol. 32, no. 4, pp. 835–838, 2011. [24] L. G. Berm´udez-Humar´an, P. Kharrat, J. Chatel, and P. Langella, “Lactococci and lactobacilli as mucosal delivery vectors for therapeutic proteins and DNA vaccines,” Microbial Cell Facto- ries, vol. 10, supplement 1, article S4, 2011. [8] M. Uchida, Y. Ohshiba, and O. Mogami, “Novel dipeptidyl peptidase-4-inhibiting peptide derived from 𝛽-lactoglobulin,” Journal of Pharmacological Sciences, vol. 117, no. 1, pp. 63–66, 2011. [25] E. Gross and J. L. Morell, “The structure of nisin,” Journal of the American Chemical Society, vol. 93, no. 18, pp. 4634–4635, 1971. [26] M. R. Islam, J. Nagao, T. Zendo, and K. Sonomoto, “Antimi- crobial mechanism of lantibiotics,” Biochemical Society Trans- actions, vol. 40, no. 6, pp. 1528–1533, 2012. [9] R. Thoma, B. L¨offler, M. Stihle, W. Huber, A. Ruf, and M. Hen- nig, “Structural basis of proline-specific exopeptidase activity as observed in human dipeptidyl peptidase-IV,” Structure, vol. 11, no. 8, pp. 947–959, 2003. [27] Y. Charng, C. Lin, and C. Hsu, “Inhibition of allergen- induced airway inflammation and hyperreactivity by recombi- nant lactic-acid bacteria,” Vaccine, vol. 24, no. 33-34, pp. 5931– 5936, 2006. [10] G. E. Lim and P. L. Brubaker, “Glucagon-like peptide 1 secretion by the L-cell: the view from within,” Diabetes, vol. 55, no. 2, pp. S70–S77, 2006. [28] A. Kruisselbrink, M. H. Den Bak-Glashouwer, C. E. G. Havenith, J. E. R. Thole, and R. Janssen, “Recombinant Lac- tobacillus plantarum inhibits house dust mite-specific T-cell responses,” Clinical and Experimental Immunology, vol. 126, no. 1, pp. 2–8, 2001. [11] L. K. Phillips and J. B. Prins, “Update on incretin hormones,” Annals of the New York Academy of Sciences, vol. 1243, pp. E55– E74, 2011. [12] C. Daniel, Y. Roussel, M. Kleerebezem, and B. Pot, “Recom- binant lactic acid bacteria as mucosal biotherapeutic agents,” Trends in Biotechnology, vol. 29, no. 10, pp. 499–508, 2011. [29] P. Rigaux, C. Daniel, M. Hisbergues et al., “Immunomodulatory properties of Lactobacillus plantarum and its use as a recom- binant vaccine against mite allergy,” Allergy, vol. 64, no. 3, pp. 406–414, 2009. [13] S. Shigemori, S. Yonekura, T. Sato, M. Nakanishi, H. References [17] J. M. Wells and A. Mercenier, “Mucosal delivery of therapeutic and prophylactic molecules using lactic acid bacteria,” Nature Reviews Microbiology, vol. 6, no. 5, pp. 349–362, 2008. [1] B. Hern´andez-Ledesma, I. Recio, and L. Amigo, “𝛽- Lactoglobulin as source of bioactive peptides,” Amino Acids, vol. 35, no. 2, pp. 257–265, 2008. [18] E. Morello, L. G. Berm´udez-Humar´an, D. Llull et al., “Lacto- coccus lactis, an efficient cell factory for recombinant protein production and secretion,” Journal of Molecular Microbiology and Biotechnology, vol. 14, no. 1–3, pp. 48–58, 2007. [2] D. E. W. Chatterton, G. Smithers, P. Roupas, and A. Brod- korb, “Bioactivity of 𝛽-lactoglobulin and 𝛼-lactalbumin— technological implications for processing,” International Dairy Journal, vol. 16, no. 11, pp. 1229–1240, 2006. [19] M. Bahey-El-Din, C. G. M. Gahan, and B. T. Griffin, “Lactococ- cus lactis as a cell factory for delivery of therapeutic proteins,” Current Gene Therapy, vol. 10, no. 1, pp. 34–45, 2010. [3] I. M. E. Lacroix and E. C. Y. Li-Chan, “Inhibition of dipeptidyl peptidase (DPP)-IV and 𝛼-glucosidase activities by pepsin- treated whey proteins,” Journal of Agricultural and Food Chem- istry, vol. 61, no. 31, pp. 7500–7506, 2013. [20] D. S. Pontes, M. S. P. de Azevedo, J. M. Chatel, P. Langella, V. Azevedo, and A. Miyoshi, “Lactococcus lactis as a live vector: heterologous protein production and DNA delivery systems,” Protein Expression and Purification, vol. 79, no. 2, pp. 165–175, 2011. [4] I. M. Lacroix and E. C. Li-Chan, “Isolation and characterization of peptides with dipeptidyl peptidase-IV inhibitory activity from pepsin-treated bovine whey proteins,” Peptides, vol. 54, pp. 39–48, 2014. [21] P. G. de Ruyter, O. P. Kuipers, and W. M. de Vos, “Controlled gene expression systems for Lactococcus lactis with the food- grade inducer nisin,” Applied and Environmental Microbiology, vol. 62, no. 10, pp. 3662–3667, 1996. [5] A. B. Nongonierma and R. J. Fitzgerald, “Dipeptidyl peptidase IV inhibitory and antioxidative properties of milk protein- derived dipeptides and hydrolysates,” Peptides, vol. 39, no. 1, pp. 157–163, 2013. [22] I. Mierau and M. Kleerebezem, “10 years of the nisin-controlled gene expression system (NICE) in Lactococcus lactis,” Applied Microbiology and Biotechnology, vol. 68, no. 6, pp. 705–717, 2005. [6] S. T. Silveira, D. Mart´ınez-Maqueda, I. Recio, and B. Hern´andez-Ledesma, “Dipeptidyl peptidase-IV inhibitory peptides generated by tryptic hydrolysis of a whey protein concentrate rich in 𝛽-lactoglobulin,” Food Chemistry, vol. 141, no. 2, pp. 1072–1077, 2013. [23] X. X. Zhou, W. F. Acknowledgments The authors thank Mr. Jiro Maeda (Faculty of Agriculture, Shinshu University) for excellent animal care and technical support. The study was supported by a grant from the Iijima Memorial Foundation for the Promotion of Food Science and Technology to Takeshi Shimosato. BioMed Research International 7 References Otani, and T. Shimosato, “Expression of a biologically active GFP-𝛼S1- casein fusion protein in Lactococcus lactis,” Current Microbiol- ogy, vol. 64, no. 6, pp. 569–575, 2012. [30] I. L. Huibregtse, V. Snoeck, A. de Creus et al., “Induction of ovalbumin-specific tolerance by oral administration of Lacto- coccus lactis secreting ovalbumin,” Gastroenterology, vol. 133, no. 2, pp. 517–528, 2007. [14] S. Shigemori, S. Yonekura, T. Sato, H. Otani, and T. Shimosato, “Expression of the immunoreactive buckwheat major allergenic storage protein in Lactococcus lactis,” Applied Microbiology and Biotechnology, vol. 97, no. 8, pp. 3603–3611, 2013. [31] C. Daniel, A. Repa, C. Wild et al., “Modulation of allergic immune responses by mucosal application of recombinant lactic acid bacteria producing the major birch pollen allergen Bet v 1,” Allergy, vol. 61, no. 7, pp. 812–819, 2006. [15] Y. Ito, S. Shigemori, T. Sato et al., “Class I/II hybrid inhibitory oligodeoxynucleotide exerts Th1 and Th2 double immunosup- pression,” FEBS Open Bio, vol. 3, pp. 41–45, 2013.i [32] K. Ohkouchi, S. Kawamoto, K. Tatsugawa et al., “Prophylactic effect of Lactobacillus oral vaccine expressing a Japanese cedar pollen allergen,” Journal of Bioscience and Bioengineering, vol. 113, no. 4, pp. 536–541, 2012. [16] T. Shimosato, M. Tohno, T. Sato et al., “Identification of a potent immunostimulatory oligodeoxynucleotide from Streptococcus thermophilus lacZ,” Animal Science Journal, vol. 80, no. 5, pp. 597–604, 2009. BioMed Research International 8 [33] M. S. P. de Azevedo, S. Innocentin, F. A. Dorella et al., “Immunotherapy of allergic diseases using probiotics or recom- binant probiotics,” Journal of Applied Microbiology, vol. 115, no. 2, pp. 319–333, 2013. [34] K. Adel-Patient, S. Ah-Leung, C. Creminon et al., “Oral admin- istration of recombinant Lactococcus lactis expressing bovine 𝛽-lactoglobulin partially prevents mice from sensitization,” Clinical and Experimental Allergy, vol. 35, no. 4, pp. 539–546, 2005. References Submit your manuscripts at http://www.hindawi.com Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Anatomy Research International Peptides International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com International Journal of Volume 2014 Zoology Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Molecular Biology International Genomics International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 The Scientific World Journal Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Bioinformatics Advances in Marine Biology Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Signal Transduction Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 BioMed Research International Evolutionary Biology International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Biochemistry Research International Archaea Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Genetics Research International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Advances in Virology Hindawi Publishing Corporation http://www.hindawi.com Nucleic Acids Journal of Volume 2014 Stem Cells International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Enzyme Research Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 International Journal of Microbiology Submit your manuscripts at http://www.hindawi.com Peptides International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com International Journal of Volume 2014 Zoology Genomics International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 The Scientific Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 BioMed Research International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Advances in Virology Hindawi Publishing Corporation http://www.hindawi.com Nucleic Acids Journal of Volume 2014 Stem Cells International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Genomics International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 The Scientific World Journal Hindawi Publishing Corporation http://www hindawi com Volume 2014 Submit your manuscripts at http://www.hindawi.com Submit your manuscripts at http://www.hindawi.com The Scientific World Journal Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Anatomy Research International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Molecular Biology International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Bioinformatics Advances in Marine Biology Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Evolutionary Biology International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Biochemistry Research International Archaea Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Genetics Research International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Enzyme Research Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 International Journal of Microbiology Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Anatomy Research International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Molecular Biology International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Bioinformatics Advances in Marine Biology Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Evolutionary Biology International Journal of Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Biochemistry Research International Archaea Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Genetics Research International Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Enzyme Research Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 International Journal of Microbiology
https://openalex.org/W2900962233	https://europepmc.org/articles/pmc6316058?pdf=render	English	null	TRPM8 Channels and Dry Eye	Pharmaceuticals	2,018	cc-by	5,036	Received: 17 September 2018; Accepted: 12 November 2018; Published: 15 November 2018 Received: 17 September 2018; Accepted: 12 November 2018; Published: 15 November 2018 Abstract: Transient receptor potential (TRP) channels transduce signals of chemical irritation and temperature change from the ocular surface to the brain. Dry eye disease (DED) is a multifactorial disorder wherein the eyes react to trivial stimuli with abnormal sensations, such as dryness, blurring, presence of foreign body, discomfort, irritation, and pain. There is increasing evidence of TRP channel dysfunction (i.e., TRPV1 and TRPM8) in DED pathophysiology. Here, we review some of this literature and discuss one strategy on how to manage DED using a TRPM8 agonist. Keywords: transient receptor potential; dry eye disease; TRPV; TRPM8 1. Emerging Concepts of Neural–Sensory Mechanisms in Dry Eye Dry eye disease (DED) is a multifactorial disorder of the ocular surface, and especially of the sensory and motor nerves that regulate the physiology of this surface [1]. It is an economic burden to society, as it affects 5–30% of the population in a wide range of age groups [2]. Considering the pathophysiology of DED, the treatment strategy has shifted from just hydrating and lubricating the ocular surface to modifying the underlying disease process. Traditionally, aqueous tear deﬁciency is considered one of the major symptoms of DED, which is caused by a deﬁcit of lacrimal and conjunctival tear secretion [3]. Recently, increased attention has focused on the neuronal regulation of glandular tear secretion [3–5]. Studies also show that thermal changes at the ocular surface activate cool neurons and may affect surface wetness [6–8]. pharmaceuticals Pharmaceuticals 2018, 11, 125; doi:10.3390/ph11040125 Jee Myung Yang 1,2 , Edward T. Wei 3, Seong Jin Kim 4 and Kyung Chul Yoon 1,* 1 Department of Ophthalmology, Chonnam National University Medical School and Hospital, Gwangju 61469, Korea; jeemang87@naver.com Gwangju 61469, Korea; jeemang87@naver.com 2 Graduate School of Medical Science and Engineering, Korea Advanced Institute of Science and Technolog Daejeon 34141, Korea School of Public Health, University of California, Berkeley, CA 94720, USA; koolicin@yahoo.com Department of Dermatology, Chonnam National University Medical School and Hospital, School of Public Health, University of California, Berkeley, CA 94720, USA; koolicin@yahoo.com 4 Department of Dermatology, Chonnam National University Medical School and Hospital, Gwangju 61469, Korea; seongkim@chonnam.ac.kr * Correspondence: kcyoon@jnu.ac.kr www.mdpi.com/journal/pharmaceuticals 2. TRP Channels Related to Cooling Sensation in DED Transient receptor potential (TRP) cation channels are associated with the perception of chemical irritation and temperature change [9]. These channels are classiﬁed into six subfamilies: (1) TRPC (canonical); (2) TRPV (vanilloid); (3) TRPM (melastatin); (4) TRPP (polycystin); (5) TRPML (mucolipin), and (6) TRPA (ankyrin) [10]. For the ocular surface, TRP channels have been identiﬁed in the cornea (TRPV1-4, TRPA1, TRPC4, and TRPM8), in the conjunctiva (TRPV1, TRPV2, and TRPV4), and in the eyelid TRPM8 (Figure 1) [10–13]. In addition, TRP receptors are differentially expressed in the corneal epithelium (TRPV1, TRPV3, TRPV4, TRPM8, and TRPC4), stroma (TRPV1, and TRPM8), and endothelium (TRPV1, TRPM8, and TRPA1) [11,14]. Within the TRP family, TRPM8 is a cold-sensing receptor (cold thermoreceptor), with a threshold of ~25 ◦C, and located on nerve endings of the ophthalmic branch of the trigeminal nerve [8]. TRPM8 receptors appear to be ﬁrst activated on the ocular surface after evaporation of the tear ﬁlm [15]. The receptor is highly sensitive to dynamic Pharmaceuticals 2018, 11, 125; doi:10.3390/ph11040125 www.mdpi.com/journal/pharmaceuticals 2 of 7 d icilin Pharmaceuticals 2018, 11, 125 to dynamic temperature temperature reduction and is also stimulated by cooling agents, such as menthol and icilin [16]. These cold-sensitive ocular thermoreceptors are the only TRP receptors that exhibit tonic, spontaneous activity. The discharge of these afferents may regulate basal tear secretion by sensing eye wetness [17]. spontaneous activity. The discharge of these afferents may regulate basal tear secretion by sensing eye wetness [17]. Figure 1. Illustration depicting identified transient receptor potential (TRP) channels in the human anterior surface. Figure 1. Illustration depicting identiﬁed transient receptor potential (TRP) channels in the human anterior surface. Figure 1. Illustration depicting identified transient receptor potential (TRP) channels in the human anterior surface. Figure 1. Illustration depicting identiﬁed transient receptor potential (TRP) channels in the human anterior surface. Dysfunction of TRPM8-mediated sensing of evaporation-induced temperature and osmolarity change in the corneal surface has been suggested as a possible pathophysiological mechanism in DED [15]. TRPM8 knock-out mice showed a reduction in basal tear secretion [8]. Corcoran et al. [18] showed that perception of cold was modified in patients with DED compared to healthy subjects. The altered sensitivity was not seen in corneal mechanoreceptors. Recently, Alcalde et al. [19] found in studies on mice that aging impairs activity of high-threshold cold thermoreceptors, which makes the ocular surface more sensitive to stimuli and ocular irritation, and to tearing. 2. TRP Channels Related to Cooling Sensation in DED These findings substantiate the hyperactivity of the cold thermoreceptor on the cornea as one of the risk factors that cause abnormal lacrimation and contribute to the high incidence of DED in aged people. It is not clear if the manifestations of changes seen in temperature sensitivity of DED patients are triggered by the disease process or just epiphenomenon. The consensus view at this time is that abnormal TRPM8 reactivity on the cornea triggers irritation. Dysfunction of TRPM8-mediated sensing of evaporation-induced temperature and osmolarity change in the corneal surface has been suggested as a possible pathophysiological mechanism in DED [15]. TRPM8 knock-out mice showed a reduction in basal tear secretion [8]. Corcoran et al. [18] showed that perception of cold was modiﬁed in patients with DED compared to healthy subjects. The altered sensitivity was not seen in corneal mechanoreceptors. Recently, Alcalde et al. [19] found in studies on mice that aging impairs activity of high-threshold cold thermoreceptors, which makes the ocular surface more sensitive to stimuli and ocular irritation, and to tearing. These ﬁndings substantiate the hyperactivity of the cold thermoreceptor on the cornea as one of the risk factors that cause abnormal lacrimation and contribute to the high incidence of DED in aged people. It is not clear if the manifestations of changes seen in temperature sensitivity of DED patients are triggered by the disease process or just epiphenomenon. The consensus view at this time is that abnormal TRPM8 reactivity on the cornea triggers irritation. 4. Modulation of TRP Channels in DED The cornea and conjunctiva of DED patients display abnormal hypersensitivity to normally harmless cold stimuli (cold allodynia) [31]. In this context, the use of TRPM8 antagonists could be a strategy of treatment. The utility of cooling for relieving dysesthesia and pain in DED is uncertain. In studies on humans, cooling relieved post-cataract surgery pain and cooled-artiﬁcial tears provided relief by decreasing corneal and conjunctival sensation measured by esthesiometry [32,33]. In addition, an ice pack applied to the orbit reduced the pain of injury, suggesting that TRPM8 activation is beneﬁcial for discomfort [33]. For conﬁrmation of the beneﬁts of cooling, it must be accurately stated how a TRPM8 agonist applied to the ocular surface will affect sensation or discomfort in patients with DED. In animal models of DED, menthol, a TRPM8 agonist, was shown to accentuate the cooling sensation [17]. However, TRPM8 agonists, such as menthol and icilin, have limited value in ocular studies in humans. After a brief episode of cooling, menthol vapors irritated the eye and menthol solutions caused signiﬁcant discomfort in patients [34]. Icilin, a more potent TRPM8 agonist than menthol, is not soluble in any ophthalmic vehicles and is, hence, difﬁcult to deliver to target receptors [12,35]. Borneol, a bicyclic monoterpenoid compound widely used in traditional Chinese medicine, has been introduced as a TRPM8 agonist that could be used in DED treatment by increasing corneal wetness in a temperature- and dose-dependent manner [36]. By studying human conjunctival epithelial cells, researchers found that thyronamine, an endogenous thyroid hormone metabolite, activated the TRPM8 channel and prevented the capsaicin-induced activation of the TRPV1 channel [37]. In addition, the effect of thyronamine has been validated in human corneal epithelial cells, suggesting this molecule as a novel endogenous modulator of TRPM8 in the ocular surface [38,39]. The antagonism of TRPM8 by N-(3-aminopropyl)-2-{[(3-methylphenyl)methyl]oxy}-N-(2-thienylmethyl) benzamide (AMTB) has also been considered for DED treatment, since evaporative cooling and hyperosmotic stimuli trigger dry eye pain as well as blinking [40]. However, an undesirable side effect of antagonists may be the reduction of tear secretion [8]. For example, the ocular application of the TRPM8 antagonist (N-(4-tertiarybutylphenyl)-4-(3-chloro-pyridin-2-yl)-tetrahydropyrazine-1(2H)-carbox-amide) (BCTC) decreases the response to corneal dryness by 45–80% [4]. An experimental human study with systemic dosing of a selective TRPM8 antagonist [(R)-3-[(1-(4-ﬂuorophenyl)ethyl)(quinolin-3-ylcarbonyl)amino] methylbenzoic acid] found no ocular symptoms when given to 22 volunteers [41]. 3 TRP Channels Related to Ocular Pain in DED 3. TRP Channels Related to Ocular Pain in DED 3. TRP Channels Related to Ocular Pain in DED According to a TFOS DEWS II report, the revised definition of DED included neurosensory abnormalities emphasizing the importance of neural regulation of tearing as well as pain sensing [1,15]. Researchers focus on corneal nociceptors (polymodal nociceptors) as initiators of DED pathophysiology [15,20]. The coding of sensory neural circuits has been intensively studied [21]. The distributions and morphological specialization related to the function of TRPV1, and TRPM8 ion channels on the cornea have been dissected [22]. For example, Alamri et al. [22] showed that corneal According to a TFOS DEWS II report, the revised deﬁnition of DED included neurosensory abnormalities emphasizing the importance of neural regulation of tearing as well as pain sensing [1,15]. Researchers focus on corneal nociceptors (polymodal nociceptors) as initiators of DED pathophysiology [15,20]. The coding of sensory neural circuits has been intensively studied [21]. The distributions and morphological specialization related to the function of TRPV1, and TRPM8 ion channels on the cornea have been dissected [22]. For example, Alamri et al. [22] showed that corneal polymodal nociceptors have TRPV1-positive small ramifying nerve endings, whereas the cold thermoreceptor has TRPM8-positive large complex endings. TRPV1 has also been detected in human 3 of 7 Pharmaceuticals 2018, 11, 125 corneal epithelial and conjunctival cells [13,23]. Generally, it is known that TRPV1 plays a role in DED, since it can be activated by hypertonic challenge, which in turn leads to an increased release of pro-inﬂammatory cytokines [24,25]. In addition, the functional transduction of the heat, irritation, and pain signal from the ocular surface has been validated for TRPV1 [21,26]. Polymodal nociceptors in the eye that preferentially contain neuropeptides are expressed through a TRPV1 channel, and those without neuropeptides are expressed through a TRPA1 channel [8,20,27,28]. These two channels are considered major detectors of external irritants, endogenous mediators, and heat on ocular surface [29]. In a murine dry eye model, the TRPV1 channel plays a major role in hypertonic saline-induced nocifensive behavior, while the TRPM8 channel is less important [30]. TRPM8 immunoﬂuorescence is dense in the cornea and eyelid skin, but not on the conjunctiva [12,26]. 4. Modulation of TRP Channels in DED Other strategies for modulating DED via TRP channels have been described and await proof of concept in clinical studies [42–44]. 5. Novel Application of a TRPM8 Agonist in DED The dialkylphosphorylalkane (Dapa) cooling agents, ﬁrst described in 1978 [36], are attractive for ocular applications because some of these analogs are soluble in water at 0.5 to 5 mg/mL and provide refreshing sensations of cooling. A new Dapa TRPM8 agonist that relieves signs and symptoms of DED was described [12]. The chemical is 1-diisopropylphosphorylnonane (CAS Registry Number 1503744–37–8-7), which is called cryosim-3 (C3) (Figure 2A). C3 is an ideal candidate since it has high 4 of 7 4 of 7 Pharmaceuticals 2018, 11, 125 Pharmaceuticals 2018, 11, x FOR selectivity for TRPM8, no overt irritation, and has an optimal duration of drug action on the ocular surface. Topical administration of C3 to the closed eyelids by wiping quickly induced coolness on the periocular surface, and the cooling sensation lasted for more than 40 min by a single application (Figure 2B). C3 improves the symptoms of DED and basal tear secretion signiﬁcantly without any adverse effects, such as ocular irritation or pain (Figure 2C,D). Corneal sensitivity measured by Cochet–Bonnet esthesiometry was not affected by application of 2 mg/mL solution of a related analog, 1-di-sec-butylphosphorylpentane, onto the closed eyelid by aerosol spray [unpublished data]. In addition, 5 µM of this analog did not inhibit hNav1.7 (sodium channels) in vitro, indicating the absence of lidocaine-like anesthetic activity. The method of drug delivery, that of wiping ~20 µL of solution per eye to the ocular margins where TRPM8 is expressed, avoids stimulation of the corneal polymodal neurons [20,45]. Minimizing bolus contact with the corneal nociceptors that cause sting, irritation, and pain is a critical factor that leads to the success of the wiping method of TRPM8 agonist delivery. candidate since it has high selectivity for TRPM8, no overt irritation, and has an optimal duration of drug action on the ocular surface. Topical administration of C3 to the closed eyelids by wiping quickly induced coolness on the periocular surface, and the cooling sensation lasted for more than 40 min by a single application (Figure 2B). C3 improves the symptoms of DED and basal tear secretion significantly without any adverse effects, such as ocular irritation or pain (Figure 2C,D). Corneal sensitivity measured by Cochet–Bonnet esthesiometry was not affected by application of 2 mg/mL solution of a related analog, 1-di-sec-butylphosphorylpentane, onto the closed eyelid by aerosol spray [unpublished data]. 5. Novel Application of a TRPM8 Agonist in DED In addition, 5 μM of this analog did not inhibit hNav1.7 (sodium channels) in vitro, indicating the absence of lidocaine-like anesthetic activity. The method of drug delivery, that of wiping ~20 μL of solution per eye to the ocular margins where TRPM8 is expressed, avoids stimulation of the corneal polymodal neurons [20,46]. Minimizing bolus contact with the corneal nociceptors that cause sting, irritation, and pain is a critical factor that leads to the success of the wiping method of TRPM8 agonist delivery. Figure 2. Chemical structure and function of C3. (A) Structure of 1-diisopropylphosphorylnonane, cryosim-3 (C3). (B,C) Visual analogue scale (VAS), (B) dry eye symptom score, and basal tear secretion (C) after single application of vehicle or C3. (D) Ocular surface disease index (OSDI) score, and basal tear secretion after application of vehicle of C3 four times a day for 2 weeks. * p < 0.05, ** p < 0.01, compared to baseline value or vehicle (Adapted from Yang et al. [12]). Figure 2. Chemical structure and function of C3. (A) Structure of 1-diisopropylphosphorylnonane, cryosim-3 (C3). (B,C) Visual analogue scale (VAS), (B) dry eye symptom score, and basal tear secretion (C) after single application of vehicle or C3. (D) Ocular surface disease index (OSDI) score, and basal tear secretion after application of vehicle of C3 four times a day for 2 weeks. * P < 0.05, ** P < 0.01, compared to baseline value or vehicle (Adapted from Yang et al. [12]). Figure 2. Chemical structure and function of C3. (A) Structure of 1-diisopropylphosphorylnonane, cryosim-3 (C3). (B,C) Visual analogue scale (VAS), (B) dry eye symptom score, and basal tear secretion (C) after single application of vehicle or C3. (D) Ocular surface disease index (OSDI) score, and basal tear secretion after application of vehicle of C3 four times a day for 2 weeks. * p < 0.05, ** p < 0.01, compared to baseline value or vehicle (Adapted from Yang et al. [12]). Figure 2. Chemical structure and function of C3. (A) Structure of 1-diisopropylphosphorylnonane, cryosim-3 (C3). (B,C) Visual analogue scale (VAS), (B) dry eye symptom score, and basal tear secretion (C) after single application of vehicle or C3. (D) Ocular surface disease index (OSDI) score, and basal tear secretion after application of vehicle of C3 four times a day for 2 weeks. 5. Novel Application of a TRPM8 Agonist in DED * P < 0.05, ** P < 0.01, compared to baseline value or vehicle (Adapted from Yang et al. [12]). References 1. Craig, J.P.; Nichols, K.K.; Akpek, E.K.; Caffery, B.; Dua, H.S.; Joo, C.K.; Liu, Z.; Nelson, J.D.; Nichols, J.J.; Tsubota, K.; et al. TFOS DEWS II Deﬁnition and Classiﬁcation Report. Ocul. Surf. 2017, 15, 276–283. [CrossRef] [PubMed] 1. Craig, J.P.; Nichols, K.K.; Akpek, E.K.; Caffery, B.; Dua, H.S.; Joo, C.K.; Liu, Z.; Nelson, J.D.; Nichols, J.J.; Tsubota, K.; et al. TFOS DEWS II Deﬁnition and Classiﬁcation Report. Ocul. Surf. 2017, 15, 276–283. [CrossRef] [PubMed] 2. Stapleton, F.; Alves, M.; Bunya, V.Y.; Jalbert, I.; Lekhanont, K.; Malet, F.; Na, K.-S.; Schaumberg, D.; Uchino, M.; Vehof, J.; et al. TFOS DEWS II Epidemiology Report. Ocul. Surf. 2017, 15, 334–365. [CrossRef] [PubMed] 2. Stapleton, F.; Alves, M.; Bunya, V.Y.; Jalbert, I.; Lekhanont, K.; Malet, F.; Na, K.-S.; Schaumberg, D.; Uchino, M.; Vehof, J.; et al. TFOS DEWS II Epidemiology Report. Ocul. Surf. 2017, 15, 334–365. [CrossRef] [PubMed] 3. Bron, A.J.; de Paiva, C.S.; Chauhan, S.K.; Bonini, S.; Gabison, E.E.; Jain, S.; Knop, E.; Markoulli, M.; Ogawa, Y.; Perez, V.; et al. TFOS DEWS II pathophysiology report. Ocul. Surf. 2017, 15, 438–510. [CrossRef] [PubMed] 3. Bron, A.J.; de Paiva, C.S.; Chauhan, S.K.; Bonini, S.; Gabison, E.E.; Jain, S.; Knop, E.; Markoulli, M.; Ogawa, Y.; Perez, V.; et al. TFOS DEWS II pathophysiology report. Ocul. Surf. 2017, 15, 438–510. [CrossRef] [PubMed] 4. Hirata, H.; Oshinsky, M.L. Ocular dryness excites two classes of corneal afferent neurons implicated in basal tearing in rats: Involvement of transient receptor potential channels. J. Neurophysiol. 2012, 107, 1199–1209. [CrossRef] [PubMed] 4. Hirata, H.; Oshinsky, M.L. Ocular dryness excites two classes of corneal afferent neurons implicated in basal tearing in rats: Involvement of transient receptor potential channels. J. Neurophysiol. 2012, 107, 1199–1209. [CrossRef] [PubMed] 5. Jones, L.; Downie, L.E.; Korb, D.; Benitez-del-Castillo, J.M.; Dana, R.; Deng, S.X.; Dong, P.N.; Geerling, G.; Hida, R.Y.; Liu, Y.; et al. TFOS DEWS II Management and Therapy Report. Ocul. Surf. 2017, 15, 575–628. [CrossRef] [PubMed] 5. Jones, L.; Downie, L.E.; Korb, D.; Benitez-del-Castillo, J.M.; Dana, R.; Deng, S.X.; Dong, P.N.; Geerling, G.; Hida, R.Y.; Liu, Y.; et al. TFOS DEWS II Management and Therapy Report. Ocul. Surf. 2017, 15, 575–628. [CrossRef] [PubMed] 6. Robbins, A.; Kurose, M.; Winterson, B.J.; Meng, I.D. Menthol activation of corneal cool cells induces TRPM8-mediated lacrimation but not nociceptive responses in rodents. Investig. Ophthalmol. Vis. Sci. 2012, 53, 7034–7042. [CrossRef] [PubMed] 6. Robbins, A.; Kurose, M.; Winterson, B.J.; Meng, I.D. References Menthol activation of corneal cool cells induces TRPM8-mediated lacrimation but not nociceptive responses in rodents. Investig. Ophthalmol. Vis. Sci. 2012, 53, 7034–7042. [CrossRef] [PubMed] 7. Belmonte, C.; Gallar, J. Cold thermoreceptors, unexpected players in tear production and ocular dryness sensations. Investig. Ophthalmol. Vis. Sci. 2011, 52, 3888–3892. [CrossRef] [PubMed] 7. Belmonte, C.; Gallar, J. Cold thermoreceptors, unexpected players in tear production and ocular dryness sensations. Investig. Ophthalmol. Vis. Sci. 2011, 52, 3888–3892. [CrossRef] [PubMed] 8. Parra, A.; Madrid, R.; Echevarria, D.; del Olmo, S.; Morenilla-Palao, C.; Acosta, M.C.; Gallar, J.; Dhaka, A.; Viana, F.; Belmonte, C. Ocular surface wetness is regulated by TRPM8-dependent cold thermoreceptors of the cornea. Nat. Med. 2010, 16, 1396–1399. [CrossRef] [PubMed] 9. McKemy, D.D.; Neuhausser, W.M.; Julius, D. Identiﬁcation of a cold receptor reveals a general role for TRP channels in thermosensation. Nature 2002, 416, 52–58. [CrossRef] [PubMed] 10. Eguchi, H.; Hiura, A.; Nakagawa, H.; Kusaka, S.; Shimomura, Y. Corneal Nerve Fiber Structure, Its Role in Corneal Function, and Its Changes in Corneal Diseases. Biomed. Res. Int. 2017, 2017. [CrossRef] [PubMed] 11. Reinach, P.S.; Mergler, S.; Okada, Y.; Saika, S. Ocular transient receptor potential channel function in health and disease. BMC Ophthalmol. 2015, 15. [CrossRef] [PubMed] 10. Eguchi, H.; Hiura, A.; Nakagawa, H.; Kusaka, S.; Shimomura, Y. Corneal Nerve Fiber Structure, Its Role in Corneal Function, and Its Changes in Corneal Diseases. Biomed. Res. Int. 2017, 2017. [CrossRef] [PubMed] 11 R i h PS M l S Ok d Y S ik S O l t i t t t ti l h l f ti i h lth 1. Reinach, P.S.; Mergler, S.; Okada, Y.; Saika, S. Ocular transient receptor potential channel function in he and disease. BMC Ophthalmol. 2015, 15. [CrossRef] [PubMed] 12. Yang, J.M.; Li, F.; Liu, Q.; Rüedi, M.; Wei, E.T.; Lentsman, M.; Lee, H.S.; Choi, W.; Kim, S.J.; Yoon, K.C. A novel TRPM8 agonist relieves dry eye discomfort. BMC Ophthalmol. 2017, 17, 101. [CrossRef] [PubMed] 13. Mergler, S.; Garreis, F.; Sahlmüller, M.; Lyras, E.-M.; Reinach, P.S.; Dwarakanath, A.; Paulsen, F.; Pleyer, U. Calcium regulation by thermo- and osmosensing transient receptor potential vanilloid channels (TRPVs) in human conjunctival epithelial cells. Histochem. Cell Biol. 2012, 137, 743–761. [CrossRef] [PubMed] 14. Reinach, P.S.; Chen, W.; Mergler, S. Polymodal roles of transient receptor potential channels in the control of ocular function. Eye Vis. 2015, 2, 5. [CrossRef] [PubMed] 15. Pharmaceuticals 2018, 11, 125 Pharmaceuticals 2018, 11, 125 Author Contributions: J.M.Y. and E.T.W. contributed to the writing of the manuscript. S.J.K. and K.C.Y. contributed to the review and editing of the manuscript. Funding: The study was partially supported by the Chonnam National University Hospital Biomedical Research Institute (CRI 18093-1) and Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2017R1A2B4003367). Acknowledgments: We would like to thank Peter Reinach for helpful comments, and anonymous reviewers for their valuable comments on the preparation of the manuscript. Conﬂicts of Interest: E.T.W. is listed on a patent application on the use of C3 for eye discomfort and nasal congestion (Di-Isopropyl-phosphinoyl-alkane (DIPA) compounds as topical agents for the treatment sensory discomfort. World Intellectual Property Organization, WO2015059432, 30 April 2015). The other authors declare that they have no competing interests. 6 Conclusions 6. Conclusions 6. Conclusions The treatment of DED becomes more challenging as people are frequently exposed to high evaporative environments. Working in front of the computer decreases the blinking rate, thereby increasing tear evaporation, which may cause dry eye-related symptoms, such as computer vision syndrome. TRP channels can be activated by trivial stresses that we encounter in daily life. So far, the translation of research findings for DED treatment has not been clearly defined in the field of TRP channels. It would be interesting to know whether TRP expressing sensory nerves and their function is preserved in patients with DED or DED-related conditions, and whether TRP modulation has the potential to treat DED in such patients. The manipulation of these TRP channels on the ocular surface may provide novel options for treating DED, especially to those refractory to conventional strategies The treatment of DED becomes more challenging as people are frequently exposed to high evaporative environments. Working in front of the computer decreases the blinking rate, thereby increasing tear evaporation, which may cause dry eye-related symptoms, such as computer vision syndrome. TRP channels can be activated by trivial stresses that we encounter in daily life. So far, the translation of research ﬁndings for DED treatment has not been clearly deﬁned in the ﬁeld of TRP channels. It would be interesting to know whether TRP expressing sensory nerves and their function is preserved in patients with DED or DED-related conditions, and whether TRP modulation has the potential to treat DED in such patients. The manipulation of these TRP channels on the ocular surface may provide novel options for treating DED, especially to those refractory to conventional strategies of surface lubrication and anti-inﬂammatory agents. We expect to further elucidate the C3 therapeutic strategy in DED patients, including patients with Sjogren’s syndrome and neuropathic pain. 5 of 7 References Belmonte, C.; Nichols, J.J.; Cox, S.M.; Brock, J.A.; Begley, C.G.; Bereiter, D.A.; Dartt, D.A.; Galor, A.; Hamrah, P.; Ivanusic, J.J.; et al. TFOS DEWS II pain and sensation report. Ocul. Surf. 2017, 15, 404–437. [CrossRef] [PubMed] 16. Bautista, D.M.; Siemens, J.; Glazer, J.M.; Tsuruda, P.R.; Basbaum, A.I.; Stucky, C.L.; Jordt, S.-E.; Julius, D. The menthol receptor TRPM8 is the principal detector of environmental cold. Nature 2007, 448, 204–208. [CrossRef] [PubMed] Bautista, D.M.; Siemens, J.; Glazer, J.M.; Tsuruda, P.R.; Basbaum, A.I.; Stucky, C.L.; Jordt, S.-E.; Julius, D The menthol receptor TRPM8 is the principal detector of environmental cold. Nature 2007, 448, 204–208 6 of 7 Pharmaceuticals 2018, 11, 125 17. Hirata, H.; Meng, I.D. Cold-sensitive corneal afferents respond to a variety of ocular stimuli central to tear production: Implications for dry eye disease. Investig. Ophthalmol. Vis. Sci. 2010, 51, 3969–3976. [CrossRef] [PubMed] 18. Corcoran, P.; Hollander, D.A.; Ousler, G.W.; Angjeli, E.; Rimmer, D.; Lane, K.; Abelson, M.B. Dynamic Sensitivity of Corneal TRPM8 Receptors to Menthol Instillation in Dry Eye Versus Normal Subjects. J. Ocul. Pharmacol. Ther. 2017, 33, 686–692. [CrossRef] [PubMed] 19. Alcalde, I.; Íñigo-Portugués, A.; González-González, O.; Almaraz, L.; Artime, E.; Morenilla-Palao, C.; Gallar, J.; Viana, F.; Merayo-Lloves, J.; Belmonte, C. Morphological and functional changes in TRPM8-expressing corneal cold thermoreceptor neurons during aging and their impact on tearing in mice. J. Comp. Neurol. 2018, 526, 1859–1874. [CrossRef] [PubMed] 20. Belmonte, C.; Acosta, M.C.; Merayo-Lloves, J.; Gallar, J. What Causes Eye Pain? Curr. Ophthalmol. Rep. 2015, 3, 111–121. [CrossRef] [PubMed] 21. Alamri, A.; Bron, R.; Brock, J.A.; Ivanusic, J.J. Transient receptor potential cation channel subfamily V member 1 expressing corneal sensory neurons can be subdivided into at least three subpopulations. Front. Neuroanat. 2015, 9, 74. [CrossRef] [PubMed] 22. Alamri, A.S.; Wood, R.J.; Ivanusic, J.J.; Brock, J.A. The neurochemistry and morphology of functionally identiﬁed corneal polymodal nociceptors and cold thermoreceptors. PLoS ONE 2018, 13, e0195108. [CrossRef] [PubMed] 23. Zhang, F.; Yang, H.; Wang, Z.; Mergler, S.; Liu, H.; Kawakita, T.; Tachado, S.D.; Pan, Z.; Capó-Aponte, J.E.; Pleyer, U.; et al. Transient receptor potential vanilloid 1 activation induces inﬂammatory cytokine release in corneal epithelium through MAPK signaling. J. Cell. Physiol. 2007, 213, 730–739. [CrossRef] [PubMed] 24. Pan, Z.; Wang, Z.; Yang, H.; Zhang, F.; Reinach, P.S. TRPV1 activation is required for hypertonicity-stimulated inflammatory cytokine release in human corneal epithelial cells. Investig. Ophthalmol. Vis. Sci. 2011, 52, 485–493. [CrossRef] [PubMed] 25. References Martínez-García, M.C.; Martínez, T.; Pañeda, C.; Gallego, P.; Jimenez, A.I.; Merayo, J. Differential expression and localization of transient receptor potential vanilloid 1 in rabbit and human eyes. Histol. Histopathol. 2013, 28, 1507–1516. [CrossRef] [PubMed] 6. Julius, D. TRP Channels and Pain. Annu. Rev. Cell Dev. Biol. 2013, 29, 355–384. [CrossRef] [PubMed] 27. Murata, Y.; Masuko, S. Peripheral and central distribution of TRPV1, substance P and CGRP of rat corneal neurons. Brain Res. 2006, 1085, 87–94. [CrossRef] [PubMed] 28. Canner, J.P.; Linsenmayer, T.F.; Kubilus, J.K. Developmental regulation of trigeminal TRPA1 by the cornea. Investig. Ophthalmol. Vis. Sci. 2014, 56, 29–36. [CrossRef] [PubMed] 29. Stucky, C.L.; Dubin, A.E.; Jeske, N.A.; Malin, S.A.; McKemy, D.D.; Story, G.M. Roles of transient receptor potential channels in pain. Brain Res. Rev. 2009, 60, 2–23. [CrossRef] [PubMed] 30. Bereiter, D.A.; Rahman, M.; Thompson, R.; Stephenson, P.; Saito, H. TRPV1 and TRPM8 Channels and Nocifensive Behavior in a Rat Model for Dry Eye. Investig. Ophthalmol. Vis. Sci. 2018, 59, 3739–3746. [CrossRef] [PubMed] 31. Situ, P.; Simpson, T.L.; Fonn, D.; Jones, L.W. Conjunctival and Corneal Pneumatic Sensitivity Is Associated with Signs and Symptoms of Ocular Dryness. Investig. Opthalmol. Vis. Sci. 2008, 49, 2971. [CrossRef] [PubMed] 2. Fujishima, H.; Yagi, Y.; Shimazaki, J.; Tsubota, K. Effects of artiﬁcial tear temperature on corneal sens and subjective comfort. Cornea 1997, 16, 630–634. [CrossRef] [PubMed] 33. Fujishima, H.; Yagi, Y.; Toda, I.; Shimazaki, J.; Tsubota, K. Increased comfort and decreased inﬂammation of the eye by cooling after cataract surgery. Am. J. Ophthalmol. 1995, 119, 301–306. [CrossRef] 34. Acosta, M.C.; Belmonte, C.; Gallar, J. Sensory experiences in humans and single-unit activity in cats evoked by polymodal stimulation of the cornea. J. Physiol. 2001, 534, 511–525. [CrossRef] [PubMed] 35. Wei, E.T.; Seid, D.A. AG-3-5: A chemical producing sensations of cold. J. Pharm. Pharmacol. 1983, 35, 110–112. [CrossRef] [PubMed] 36. Chen, G.L.; Lei, M.; Zhou, L.P.; Zeng, B.; Zou, F. Borneol is a TRPM8 agonist that increases ocular surface wetness. PLoS ONE 2016, 11, e0158868. [CrossRef] [PubMed] 37. Khajavi, N.; Reinach, P.S.; Slavi, N.; Skrzypski, M.; Lucius, A.; Strauß, O.; Köhrle, J.; Mergler, S. Thyronamine induces TRPM8 channel activation in human conjunctival epithelial cells. Cell Signal. 2015, 27, 315–325. [CrossRef] [PubMed] 7 of 7 Pharmaceuticals 2018, 11, 125 38. Khajavi, N.; Mergler, S.; Biebermann, H. 3-Iodothyronamine, a Novel Endogenous Modulator of Transient Receptor Potential Melastatin 8? Front. Endocrinol. 2017, 8, 198. [CrossRef] [PubMed] 39. © 2018 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/). References Lucius, A.; Khajavi, N.; Reinach, P.S.; Köhrle, J.; Dhandapani, P.; Huimann, P.; Ljubojevic, N.; Grötzinger, C.; Mergler, S. 3-Iodothyronamine increases transient receptor potential melastatin channel 8 (TRPM8) activity in immortalized human corneal epithelial cells. Cell Signal. 2016, 28, 136–147. [CrossRef] [PubMed] 40. Quallo, T.; Vastani, N.; Horridge, E.; Gentry, C.; Parra, A.; Moss, S.; Viana, F.; Belmonte, C.; Andersson, D.A.; Bevan, S. TRPM8 is a neuronal osmosensor that regulates eye blinking in mice. Nat. Commun. 2015, 6, 7150. [CrossRef] [PubMed] 41. Winchester, W.J.; Gore, K.; Glatt, S.; Petit, W.; Gardiner, J.C.; Conlon, K.; Postlethwaite, M.; Saintot, P.; Roberts, S.; Gosset, J.R.; et al. Inhibition of TRPM8 channels reduces pain in the cold pressor test in humans. J. Pharmacol. Exp. Ther. 2014, 351, 259–269. [CrossRef] [PubMed] 42. Moreno-Montañés, J.; Bleau, A.-M.; Jimenez, A.I. Tivanisiran, a novel siRNA for the treatment of dry eye disease. Expert Opin. Investig. Drugs 2018, 27, 421–426. [CrossRef] [PubMed] 43. Benitez-Del-Castillo, J.M.; Moreno-Montañés, J.; Jiménez-Alfaro, I.; Muñoz-Negrete, F.J.; Turman, K.; Palumaa, K.; Sádaba, B.; González, M.V.; Ruz, V.; Vargas, B.; et al. Safety and Efﬁcacy Clinical Trials for SYL1001, a Novel Short Interfering RNA for the Treatment of Dry Eye Disease. Investig. Ophthalmol. Vis. Sci. 2016, 57, 6447–6454. [CrossRef] [PubMed] 44. Khajavi, N.; Reinach, P.S.; Skrzypski, M.; Lude, A.; Mergler, S. L-carnitine reduces in human conjunctival epithelial cells hypertonic-induced shrinkage through interacting with TRPV1 channels. Cell. Physiol. Biochem. 2014, 34, 790–803. [CrossRef] [PubMed] 45. Rosenthal, P.; Borsook, D. The corneal pain system. Part I: The missing piece of the dry eye puzzle. Ocul. Surf. 2012, 10, 2–14. [CrossRef] [PubMed] © 2018 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/). © 2018 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/W4237135129	https://europepmc.org/articles/pmc6675085?pdf=render	English	null	Retraction: Symptom Experienced Three Years after Liver Transplantation under Immunosuppression in Adults	PloS one	2,019	cc-by	539	Retraction: Symptom Experienced Three Years after Liver Transplantation under Immunosuppression in Adults The PLOS ONE Editors Concerns have been raised that the transplants performed in the local context at the time of procedures reported in this article [1] may have involved organs/tissues procured from prison- ers [2]. Details as to the donor sources and methods of obtaining informed consent from donors were not reported in [1], and when following up on these concerns the authors did not clarify these issues or the cause(s) of donor death in response to journal inquiries. International ethi- cal standards call for transparency in organ donor and transplantation programs and clear informed consent procedures including considerations to ensure that donors are not subject to coercion [3, 4, 5]. In addition, the authors did not provide documentation when requested by the journal to confirm that the study had institutional ethical approval. The authors did not respond to inquiries about the availability of underlying data support- ing this study. Owing to the lack of documentation to demonstrate that this study had prospec- tive ethical approval, insufficient reporting, the unresolved concerns around the source of transplanted organs, the high probability that the transplant cases in the study included organs from prisoners given the study/procedure dates, and in compliance with international ethical standards for organ/tissue donation and transplantation, the PLOS ONE Editors retract this article. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 The authors did not respond to the retraction decision. OPEN ACCESS Citation: The PLOS ONE Editors (2019) Retraction: Symptom Experienced Three Years after Liver Transplantation under Immunosuppression in Adults. PLoS ONE 14(8): e0220872. https://doi. org/10.1371/journal.pone.0220872 Published: August 1, 2019 Copyright: © 2019 The PLOS ONE Editors. 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. Citation: The PLOS ONE Editors (2019) Retraction: Symptom Experienced Three Years after Liver Transplantation under Immunosuppression in Adults. PLoS ONE 14(8): e0220872. https://doi. org/10.1371/journal.pone.0220872 Citation: The PLOS ONE Editors (2019) Retraction: Symptom Experienced Three Years after Liver Transplantation under Immunosuppression in Adults. PLoS ONE 14(8): e0220872. https://doi. org/10.1371/journal.pone.0220872 3. World Health Organization. WHO guiding principles on human cell, tissue and organ transplantation. Transplantation 2010; 90:229–33. https://doi.org/10.1097/TP.0b013e3181ec29f0 PMID: 20664493 4. WMA—The World Medical Association. WMA Statement on organ and tissue donation. https://www. wma.net/policies-post/wma-statement-on-organ-and-tissue-donation/ 5. Stock P Policy and Ethics. The Transplantation Society. https://www.tts.org/about-tts-5/governance/ policy-a-ethics Published: August 1, 2019 References 1. Wang C, Wang G, Yi H, Tan J, Xu C, Fang X, et al. (2013) Symptom Experienced Three Years after Liver Transplantation under Immunosuppression in Adults. PLoS ONE 8(11): e80584. https://doi.org/ 10.1371/journal.pone.0080584 PMID: 24260427 2. Rogers W, Robertson MP, Ballantyne A, et al Compliance with ethical standards in the reporting of donor sources and ethics review in peer-reviewed publications involving organ transplantation in China: a scoping review BMJ Open 2019; 9:e024473. https://doi.org/10.1136/bmjopen-2018-024473 PMID: 30723071 Published: August 1, 2019 Copyright: © 2019 The PLOS ONE Editors. 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 / 1 PLOS ONE \| https://doi.org/10.1371/journal.pone.0220872 August 1, 2019
https://openalex.org/W2984569312	https://europepmc.org/articles/pmc6838316?pdf=render	English	null	Tailoring the surface area and the acid–base properties of ZrO2 for biodiesel production from Nannochloropsis sp.	Scientific reports	2,019	cc-by	9,630	Nurul Jannah Abd Rahman1, Anita Ramli1, Khairulazhar Jumbri1,3 & Yoshimitsu Uemura2,3 Bifunctional heterogeneous catalysts have a great potential to overcome the shortcomings of homogeneous and enzymatic catalysts and simplify the biodiesel production processes using low-grade, high-free-fatty-acid feedstock. In this study, we developed ZrO2-based bifunctional heterogeneous catalysts for simultaneous esterification and transesterification of microalgae to biodiesel. To avoid the disadvantage of the low surface area of ZrO2, the catalysts were prepared via a surfactant-assisted sol-gel method, followed by hydrothermal treatments. The response surface methodology central composite design was employed to investigate various factors, like the surfactant/ Zr molar ratio, pH, aging time, and temperature on the ZrO2 surface area. The data were statistically analyzed to predict the optimal combination of factors, and further experiments were conducted for verification. Bi2O3 was supported on ZrO2 via the incipient wetness impregnation method. The catalysts were characterized by a variety of techniques, which disclosed that the surfactant-assisted ZrO2 nanoparticles possess higher surface area, better acid–base properties, and well-formed pore structures than bare ZrO2. The highest yield of fatty acid methyl esters (73.21%) was achieved using Bi2O3/ ZrO2(CTAB), and the catalytic activity of the developed catalysts was linearly correlated with the total densities of the acidic and basic sites. The mechanism of the simultaneous reactions was also discussed. Biodiesel is an attractive alternative source of energy owing to its renewability, biodegradability, sustainability, and non-toxicity1. It is produced by transesterification of vegetable oil or animal fat with short-chain alcohols in the pres- ence of suitable chemical catalysts (homogeneous/heterogeneous) or enzymatic biocatalysts2–4. Today, there has been growing research interest in using microalgae as biodiesel feedstock because of its rapid growth rate, high photosyn- thetic efficiency, and high oil contents, as well as the minimum space needed for cultivation5,6. Industrially, conven- tional homogeneous catalysts are used in the transesterification process for the production of biodiesel2. However, the catalysts require extensive washing and purification steps, and they cause undesired saponification when dealing with high-free-fatty-acid (FFA) content feedstock7. The enzymatic transesterification of lipases is commonly associated with high production cost and fast deactivation at severe reaction conditions that limit its application at an industrial scale8. An alternative method to overcome these challenges is the utilization of heterogeneous catalysts.h Numerous studies have been reported on heterogeneous catalysis for biodiesel production. The most common key features of efficient and active heterogeneous transesterification catalysts are high surface area9,10, adequate acidic11–13 and basic14–17 densities, good crystallinity14, and -well-formed pore structure10,12,13. www.nature.com/scientificreports www.nature.com/scientificreports Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 Tailoring the surface area and the acid–base properties of ZrO2 for biodiesel production from Nannochloropsis sp. Nurul Jannah Abd Rahman1, Anita Ramli1, Khairulazhar Jumbri1,3 & Yoshimitsu Uemura2,3 Nurul Jannah Abd Rahman1, Anita Ramli1, Khairulazhar Jumbri1,3 & Yoshimitsu Uemura2,3 Recent developments in this field have led to renewed interest in bifunctional acid–base heterogeneous catalysts for simultaneous esterifi- cation and transesterification of low-grade high-FFA model feedstock7,18,19, such as microalgae lipid. Heterogeneous acidic catalysts are commonly used for the esterification step as the reaction is less affected by the presence of water and FFA. Instead, heterogeneous basic catalysts are employed in the second transesterification step because they are more active than acidic catalysts, which require shorter reaction time and lower reaction temperature20. Some stud- ies have reported the use of bifunctional heterogeneous catalyst for biodiesel production18,19,21–23. However, reports on their application using microalgae lipid as the biodiesel feedstock are still limited. 1Fundamental and Applied Sciences Department, Universiti Teknologi PETRONAS, 32610, Seri Iskandar, Perak, Malaysia. 2Centre for Biofuel and Biochemical Research, Universiti Teknologi PETRONAS, 32610, Seri Iskandar, Perak, Malaysia. 3These authors contributed equally: Khairulazhar Jumbri and Yoshimitsu Uemura. email: anita_ramli@utp.edu.my www.nature.com/scientificreports/ Zirconium dioxide (zirconia, ZrO2) is a well-known heterogeneous catalyst and catalyst support that exhibits unique characteristic of amphoteric nature which indicates its remarkable potential to perform simultaneous esterification–transesterification reactions of high-FFA feedstock to biodiesel19. ZrO2 has a high boiling point, high melting point, good thermal stability, and good corrosion resistance, making it an excellent heterogeneous catalyst even under harsh reaction conditions24,25. As a catalyst support, ZrO2 exhibits better chemical properties and higher stability than the traditional catalyst supports of γ-alumina and silica26. Among the common tech- niques of synthesizing ZrO2 are sol-gel27,28, precipitation24,29, microwave-assisted30, ultrasound-assisted31,32, and emulsion33 methods. However, one of the biggest challenges that has limited its performance in practical applica- tions so far24 is the development of a suitable synthetic route of ZrO2 with a high surface area, adequate acid–base properties, good crystalline structure, and well-developed porosity for the aforementioned purpose. p p g y p p y p p Several attempts have been adopted to improve the surface area of heterogeneous catalysts through the syn- thesis of nanoscale materials by surfactant-assisted methodologies34,35. The surfactant plays a decisive role in tailoring the properties of the heterogeneous catalysts, including its shape and size, which, in turn, depend on the nature of the surfactant, such as the length of the hydrophobic tail and the ions (cationic, anionic, or non-ionic) 36. Results and Discussionf Effect of the type of the surfactants and the reaction parameters on the surface area of ZrO2 and optimization study. The maximum surface area of ZrO2 was achieved using poly(ethylene oxide)-block-poly(propylene oxide)-block-poly(ethylene oxide) (Pluronic P123) and cetrimonium bromide (CTAB) as surfactants, yielding the ZrO2(P123) and ZrO2(CTAB) catalysts, respectively. The optimization of the pro- cess parameters was conducted by employing RSMCCD, which maintained the experimental conditions within the desired range of independent parameters. According to the literature, the most important parameters affect- ing the surface area of ZrO2 are the surfactant/Zr ratio (A), pH (B), aging time (C), and temperature (D) 39,40,41. The specific values of the independent parameters used in this study, along with the surface area obtained for ZrO2(P123) and ZrO2(CTAB) are cited in Supplementary Tables S1 and S2, respectively. Specifically, among the 30 experimental RSMCCD runs, ZrO2(P123) displayed a maximum surface area of 79 m2/g (Run 21), whereas ZrO2(CTAB) exhibited a maximum surface area of 295 m2/g (Run 20).h ( ) g The relationship between the independent parameters and the surface area obtained using the analysis of variance (ANOVA) test for ZrO2(P123) and ZrO2(CTAB) are summarized in Supplementary Tables S3 and S4, respectively. By fitting the data to various polynomial models, the ANOVA result shows that both ZrO2(P123) and ZrO2(CTAB) were suitably fitted to reduced cubic models. The obtained P-values (< 0.05) indicated that the suggested model terms have a significant effect on the response42. In particular, for ZrO2(P123) (R1), the signifi- cant terms were A, B, C, D, CD, B2, C2, D2, ACD, and A2B, and for ZrO2(CTAB) (R2), the significant terms were C, BC, CD, A2, B2, C2, D2, BCD, A2C, and A2D. Herein, the combined effect of aging time and temperature (CD) was one of the most significant terms toward the improvement of the ZrO2 surface area. Thus, this observation highlighted the importance of sufficient aging time for the effective distribution of Zr–OH and Zr–O–Zr in order to form a stable network gel. Moreover, it was proven that a suitable hydrothermal temperature leads to the development of internal pressure, increases in the motion velocity of the surfactants, and prevents the agglomeration of the Zr nanoparticles38.hfii gg p The high coefficient of determination (R2) obtained indicated the goodness of fit of the two generated models. Furthermore, the lack of fit values for both models were not significant which is the desirable result. Nurul Jannah Abd Rahman1, Anita Ramli1*, Khairulazhar Jumbri1,3 & Yoshimitsu Uemura2,3 Previous studies suggested the use of surfactant in the sol-gel technique because of its homogeneity and abil- ity to control the surface area, the pore volume, and the pore size distribution of the catalysts31,37. The synthe- sis of surfactant-assisted ZrO2 catalyst is governed by a variety of parameters, including the surfactant type and the synthetic conditions that affect its overall quality. For instance, Eltejaei et al. used poly(ethylene gly- col)-block-poly(propylene glycol)-block-poly(ethylene glycol) (PEG–PPG–PEG) as a non-ionic surfactant in the synthesis of tetragonal ZrO2, employing the precipitation method at basic Ph34. Alteration of the pH from 10 to 11 resulted in high surface area ZrO2 due to the increase in surface charge and nucleation that occurs at high pH values. In another study, Zhang et al. synthesized nano-sized tetragonal ZrO2 via hydrothermal treatment using cetrimonium bromide (CTAB) as the cationic surfactant. Hydrothermal energy, a non-conventional energy source for the synthesis of nanoparticles, prevents particle agglomeration and allows for uniform grain size and regular morphology38. g p gy In the present study, we developed an effective ZrO2-based bifunctional heterogeneous catalyst for simulta- neous esterification–transesterification of microalgae lipid to biodiesel. The effect of several process parameters on the surface area of ZrO2 prepared by a surfactant-assisted sol-gel method followed by a hydrothermal treat- ment using non-ionic and cationic surfactants under basic conditions was investigated. The optimization of the process parameters was achieved using response surface methodology central composite design (RSMCCD). Mathematical models were developed and validated to predict the maximum surface area of ZrO2. The acidic and basic properties of ZrO2 were tailored after modification with bismuth oxide (Bi2O3) via incipient wetness impregnation method. The synthesized catalysts were found to be active towards the conversion of microalgae lipid to biodiesel. Results and Discussionf + . + . − . + . + . − . − . + . − . − . − . − . + . − . + . + . R A B C D AB AC AD BC BD CD A B C D ACD BCD A C A D 2 274 42 2 23 3 68 25 83 7 54 1 23 2 48 3 33 8 25 5 90 47 18 52 12 52 36 39 59 27 33 2 76 8 15 22 65 38 17 (2) 2 2 2 2 2 2 (2) where, A is the surfactant/Zr molar ratio, B is the pH value, C is the aging time, and D is the temperature. The optimum reaction parameters suggested by RSMCCD for the highest surface area of ZrO2(P123) and ZrO2(CTAB) are summarized in Supplementary Table S5. Specifically, the optimum surface area of ZrO2(P123) was 79 m2/g using a surfactant/Zr molar ratio of 0.03, pH of 9.5, and aging time of 22 h at 110 °C. On the other hand, an optimum surface area of 295 m2/g was achieved for ZrO2(CTAB) using a surfactant/Zr molar ratio of 0.89, pH of 9.8, and aging time of 39 h at 110 °C. The deviation (%) values were calculated according to the deviation between the predicted and experimental values43. The results obtained were satisfactory and reliable, with accept- able proximity. Catalyst characterization. Figure 1(a–g) illustrate the low-angle and wide-angle X-ray diffraction (XRD) spectra of the synthesized catalysts. The presence of low-angle diffraction peaks at a 2θ range from 0.06° to 0.80° indicates that both ZrO2(P123) and ZrO2(CTAB) exhibited well-organized mesopore structures after calcination at 500 °C (Fig. 1a,b) 44,45. Figure 1c shows the wide-angle XRD diffractogram of bare ZrO2 with reflection peaks at 2θ = 30.2° (011), 35° (110), 50.3° (112), and 60.1° (121) that corresponded to tetragonal ZrO2 (t-ZrO2; ICDD: 98-015-7619). The remaining peaks at 2θ = 24.4° (110), 28.3° (11-1), 31.5° (111), 40.8° (102), and 45° (211) corresponded to monoclinic ZrO2 (m-ZrO2; ICDD: 98-006-8782). It was observed that the addition of Pluronic P123 (Fig. 1d) slightly increased the intensity of the peaks but did not shift the peaks position. Results and Discussionf Both R2 and adjusted R2 values were close to unity, indicating the accuracy of the models. The low values of the coeffi- cient of variation (CV%) for both models indicated the good precision and reliability of the experiments. The correlation between the predicted and actual surface areas of ZrO2(P123) and ZrO2(CTAB) are shown respectively in Supplementary Fig. S1 (a,b). The relationship between the predicted and actual values for both models was approximately linear, pointing out the reliability of the models developed to establish a correlation between the process parameters and the surface area. Accordingly, the final predicted surface areas of ZrO2(P123) and ZrO2(CTAB) were determined based on the given values of each factor, as shown in Eqs (1) and (2), respectively: Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ Figure 1. Low-angle XRD diffractograms of (a) ZrO2(P123) and (b) ZrO2(CTAB) and wide-angle XRD diffractograms (inset) of (c) bare ZrO2, (d) ZrO2(P123), (e) Bi2O3/ZrO2(P123), (f) ZrO2(CTAB), and (g) Bi2O3/ ZrO2(CTAB) catalysts. t and m refer to the tetragonal and monoclinic ZrO2, respectively. Figure 1. Low-angle XRD diffractograms of (a) ZrO2(P123) and (b) ZrO2(CTAB) and wide-angle XRD diffractograms (inset) of (c) bare ZrO2, (d) ZrO2(P123), (e) Bi2O3/ZrO2(P123), (f) ZrO2(CTAB), and (g) Bi2O3/ ZrO2(CTAB) catalysts. t and m refer to the tetragonal and monoclinic ZrO2, respectively. = . + . − . − . + . − . −. + . + . − . −. − . + . + . + . + . + . R A B C D AB AC AD BC BD CD A B C D ACD A B 1 64 14 1 17 2 51 1 79 2 74 0 0119 0 0056 0 0344 0 2769 0 4531 3 18 0 5026 1 88 2 66 1 55 0 6956 3 04 (1) 2 2 2 2 2 = . + . − . − . + . − . −. + . + . − . −. − . + . + . + . + + R A B C D AB AC AD BC BD CD A B C D ACD A B 1 64 14 1 17 2 51 1 79 2 74 0 0119 0 0056 0 0344 0 2769 0 4531 3 18 0 5026 1 88 2 66 1 55 0 6956 3 04 2 2 2 2 2 (1) = . + . − . Results and Discussionf Nitrogen adsorption/desorption isotherms of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ZrO2(CTAB) catalysts. framework increased the crystallite size of the catalyst because of the participation of Bi2O3 in the growth of the particles. Since ZrO2(CTAB) was amorphous, no XRD data related to crystallite size were obtained. However, the crystallite sizes of m-ZrO2 and t-ZrO2 in Bi2O3/ZrO2(CTAB) were determined at 10.7 and 9.3 nm, respectively. Rietveld quantitative analysis was applied as a powerful tool to quantify the crystalline components in the mul- tiphase structures47. As outlined in Table 1, the volume fractions of the monoclinic and tetragonal phases of bare ZrO2 were similar. Initially, ZrO2(P123) was predominantly in the tetragonal phase. However, loading of Bi2O3 on the ZrO2(P123) surface transformed the tetragonal to the monoclinic phase. By contrast, Bi2O3/ZrO2(CTAB) has a higher content of t-ZrO2 compared to m-ZrO2. In many reaction systems, t-ZrO2 has been reported to show high catalytic activity48 because of its low surface energy49 and its optimum geometrical arrangement that stabilizes a transition state complex between the reactants on the t-ZrO2 surface50. p 2 Figure 2 shows the nitrogen adsorption/desorption isotherms of the catalysts. According to IUPAC classi- fication, all of the catalysts exhibited a type IV isotherm with a hysteresis loop because of capillary condensa- tion attributed to the well-developed mesoporous system51. The shape of the hysteresis loop contributed to the characteristic specific pore structures in the catalysts. According to the obtained results, bare ZrO2, ZrO2(CTAB), and Bi2O3/ZrO2(CTAB) resembled the H2 type, typical for inorganic oxides with ink-bottle-shaped mesopores52. ZrO2(P123) and Bi2O3/ZrO2(P123) exhibited a H1 type of hysteresis loop, implying the existence of a cylindrical pore geometry, spherical particles compacted in uniform arrangement, and a high degree of pore size uniformity52,53. The Brunauer–Emmett–Teller (BET) surface area, the total pore volume, and the average pore size are out- lined in Table 2. Bare ZrO2 exhibited the lowest values of surface area and total pore volume. Regarding the surfactant-assisted nanoparticles, ZrO2(CTAB) exhibited a significantly larger BET surface area and total pore volume than ZrO2(P123). This is due to the higher CTAB/Zr molar ratio used in the synthesis of ZrO2(CTAB) and the effect of the surfactant’s cationic nature. After mixing water with CTAB, the cationic charges of CTAB were released and induced repulsive forces between the Zr particles, which resulted in a high pore volume54. Results and Discussionf The addition of a small amount of Bi2O3 on the surface of ZrO2(P123) has significantly enhanced the crystalline structure of the final catalyst by forming sharp and highly intense peaks at the same 2θ values (Fig. 1e). On the other hand, the addition of CTAB (Fig. 1f) resulted in two broad and unremarkable peaks centered at 2θ = 30.7° and 50.6°, indicating an amorphous structure. The amorphous structure of ZrO2(CTAB) was also crystallized to almost tetragonal phase after impregnation with Bi2O3 (Fig. 1g) due to its instability which allowed phase transitions46. Although the peaks of Bi2O3 could not be identified because of overlapping with the peaks of m-ZrO2, the cubic phase of Bi2O3 (ICDD: 98-000-2375) was present at 2θ = 27.2° (111), 31.5° (002), and 45.2° (022). Table 1 shows the average crystallite sizes and compositions of the m-ZrO2 and t-ZrO2 forms of the catalysts. Bare ZrO2 possessed larger average crystallite sizes for both monoclinic and tetragonal phases compared to the Pluronic- and CTAB-assisted nanoparticles. During the synthesis of ZrO2, the surfactant served as a soft tem- plate to prevent agglomeration of nanoparticles through various repulsive and attractive forces that developed between the surfactant and the nanoparticles36. The average crystallite sizes of ZrO2(P123) were found to be 11.5 and 10.3 nm for the monoclinic and tetragonal phases, respectively. The insertion of Bi2O3 into the ZrO2(P123) Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ Catalyst Crystallite size (nm) Phase composition (vol %) m-ZrO2 t-ZrO2 Vm Vt ZrO2 21.3 30.2 48.3 51.7 ZrO2(P123) 11.5 10.3 43.9 56.1 Bi2O3/ZrO2(P123) 14.5 12.3 77.9 22 ZrO2(CTAB) — — — — Bi2O3/ZrO2(CTAB) 10.7 9.3 21.6 78.4 Table 1. Average crystallite sizes and phase compositions of the m-ZrO2 and t-ZrO2 forms of the synthesized catalysts. Catalyst Crystallite size (nm) Phase composition (vol %) m-ZrO2 t-ZrO2 Vm Vt ZrO2 21.3 30.2 48.3 51.7 ZrO2(P123) 11.5 10.3 43.9 56.1 Bi2O3/ZrO2(P123) 14.5 12.3 77.9 22 ZrO2(CTAB) — — — — Bi2O3/ZrO2(CTAB) 10.7 9.3 21.6 78.4 Table 1. Average crystallite sizes and phase compositions of the m-ZrO2 and t-ZrO2 forms of the synthesized catalysts. Table 1. Average crystallite sizes and phase compositions of the m-ZrO2 and t-ZrO2 forms of the synthesized catalysts. Table 1. Average crystallite sizes and phase compositions of the m-ZrO2 and t-ZrO2 forms of the synthesize catalysts. Figure 2. Nitrogen adsorption/desorption isotherms of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ZrO2(CTAB) catalysts. Figure 2. Results and Discussionf After the impregnation with Bi2O3, an increase in the average pore size of Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB) was observed because of the shifting of the pore size plots to the right of the larger pore size area55.hi The acid–base bifunctional properties of the catalysts were proven by the NH3-TPD and CO2-TPD profiles, as depicted in Figs 4 and 5, respectively. The total acidic/basic sites of the catalysts along with their density are summarized in Table 3. The number of total acidic/basic sites was calculated based on the intensity of the NH3/ CO2 desorption peaks, and the density of each catalyst was obtained by dividing the number of total acidic/basic sites by the surface area. Meanwhile, the strength of the acidic/basic sites was denoted by the desorption temper- ature. For ZrO2, the desorption peaks below 250 °C could be attributed to weak acidic/basic sites, the adsorption peaks between 250 °C and 500 °C corresponded to acid/basic sites of medium strength, and the adsorption peaks over 500 °C represented strong acidic/basic sites56. Overall, it was proven that the surfactant-assisted nanopar- ticles increased the number and the density of the total acidic/basic sites compared to bare ZrO2. According to the NH3-TPD profile (Fig. 4), bare ZrO2 showed a small desorption peak at 489 °C, indicating medium acid strength. Regarding the surfactant-assisted nanoparticles, ZrO2(P123) exhibited a small desorption peak at 254 °C, and ZrO2(CTAB) showed a broader desorption peak centered at 266 °C, also indicating medium acidic strength. ZrO2(CTAB) exhibited higher total acidic sites compared to ZrO2(P123) because of the higher surface area of the catalyst39. Interestingly, the curves of Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB) with higher total acidic sites com- pared to their parents ZrO2 were shifted to the right. This trend was in general agreement with other findings on Bi2O3-modified La2O3 catalysts7. Nizah et al. found that the addition of Bi2O3 on the surface of La2O3 enhanced the acidic properties of the final catalysts7. On the basis of the CO2-TPD profile (Fig. 5), all catalysts, except for bare ZrO2, exhibited basic sites of weak strength at a desorption temperature between 112 °C and 118 °C. Instead, bare ZrO2 presented a small desorption peak at 487 °C, indicating medium basic strength. Results and Discussionf The higher pore volume along with smaller average pore size contributed to the formation of a higher total surface area. However, the total surface areas of ZrO2(P123) and ZrO2(CTAB) decreased by about 20% and 47%, respectively, after impregnation with Bi2O3 because of the pore filling effect. In addition, all of the catalysts exhibited a mesoporous structure with an average pore size between 5.6 and 13.98 m2/g. Figure 3 shows the pore size distribution plots Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ Catalyst BET surface area (m2/g) Total pore volume (cm3/g) Average pore size (nm) ZrO2 37 0.06 5.6 ZrO2(P123) 79 0.31 10.79 Bi2O3/ZrO2(P123) 63 0.26 13.98 ZrO2(CTAB) 295 0.58 5.69 Bi2O3/ZrO2(CTAB) 157 0.31 9.80 Table 2. BET surface area, total pore volume, and average pore size of the synthesized catalysts. Catalyst BET surface area (m2/g) Total pore volume (cm3/g) Average pore size (nm) ZrO2 37 0.06 5.6 ZrO2(P123) 79 0.31 10.79 Bi2O3/ZrO2(P123) 63 0.26 13.98 ZrO2(CTAB) 295 0.58 5.69 Bi2O3/ZrO2(CTAB) 157 0.31 9.80 Table 2. BET surface area, total pore volume, and average pore size of the synthesized catalysts. Table 2. BET surface area, total pore volume, and average pore size of the synthesized catalysts. Table 2. BET surface area, total pore volume, and average pore size of the synthesized catalysts. Figure 3. Pore distribution plots of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ZrO2(CTAB) catalysts. Figure 3. Pore distribution plots of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ZrO2(CTAB) catalysts. using the Barrett, Joyner, and Halenda method. All of the catalysts exhibited unimodal pore size distribution plots, with ZrO2(P123) and Bi2O3/ZrO2(P123) showing a narrower pore size distribution compared to the other cat- alysts, indicating their high degree of pore size uniformity. After the impregnation with Bi2O3, an increase in the average pore size of Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB) was observed because of the shifting of the pore size plots to the right of the larger pore size area55.hi using the Barrett, Joyner, and Halenda method. All of the catalysts exhibited unimodal pore size distribution plots, with ZrO2(P123) and Bi2O3/ZrO2(P123) showing a narrower pore size distribution compared to the other cat- alysts, indicating their high degree of pore size uniformity. Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 Results and Discussionf In addition, the results of the mapping analysis showed that the Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ Catalyst Total acidic site (mmol/g) Density of total acidic site (mmol/m2) Total basic site (mmol/g) Density of total basic site (mmol/m2) ZrO2 0.08 0.002 0.17 0.004 ZrO2(P123) 0.35 0.004 1.66 0.021 Bi2O3/ZrO2(P123) 0.41 0.006 1.49 0.024 ZrO2(CTAB) 16.12 0.055 6.68 0.022 Bi2O3/ZrO2(CTAB) 17.38 0.111 4.36 0.027 Table 3. Acidic and basic properties of the synthesized catalysts. Table 3. Acidic and basic properties of the synthesized catalysts. Bi2O3 particles were evenly dispersed on the surface of ZrO2 owing to the homogeneous structure of the Bi2O3/ ZrO2(P123) and Bi2O3/ZrO2(CTAB) catalysts. Catalytic activity toward biodiesel production from microalgae. The simultaneous esterification– transesterification of Nannochloropsis sp. lipid to biodiesel was selected as the model reaction to test the activity of the synthesized catalysts. The catalytic activity was evaluated based on the fatty acid methyl esters (FAME) yield, as shown in Fig. 7. In particular, Bi2O3/ZrO2(CTAB) afforded the highest FAME yield (73.21%), followed by ZrO2(CTAB) (71.65%), Bi2O3/ZrO2(P123) (67.01%), ZrO2(P123) (64.73%), and bare ZrO2 (25.48%). Although the gen- eral increase in the surface area of the catalyst by the surfactant-assisted nanoparticles resulted in better FAME yield compared to bare ZrO2, the single high surface area did not lead to high catalytic activity. It was found that the catalytic performance is as a result of the synergistic role of both the total acidic and basic site densities. Apart from the lowest surface area and acidic/basic site densities, the least FAME yield obtained by bare ZrO2 was also correlated with the small pore openings of the catalyst structure, which prevented the bulky triglyceride mole- cules from reaching the catalyst’s active site. Similar findings have been observed in other studies. Omar et al. found that the balanced acidity and basicity on the surface of Sr/ZrO2 catalysts contributed to high FAME yields from waste cooking oil19. In another study, a bifunctional catalyst of Bi2O3-modified La2O3 was employed for simultaneous esterification–transesterification of Jatropha oil to biodiesel7. The results of the current study also showed that the performance of the catalyst was associated with the high surface area and the strong acidic and basic sites, and the mixed oxide catalysts exhibited higher catalytic performance than their parent ZrO2. Similarly, Umdu et al. investigated the use of single metal oxides and mixed oxides toward the production of biodiesel from microalgae lipid17. Results and Discussionf It was demonstrated that pure CaO and MgO were inactive, but mixed oxides of CaO/Al2O3 and MgO/Al2O3 were catalytically active for transesterification under the same reaction conditions. Their high catalytic performance was attributed to the high density and the mild strength of their basic sites. Surfactant-assisted sol-gel method followed by hydrothermal treatment. There are three main reactions involved in the sol-gel process, namely, hydrolysis, condensation, and aging43. During hydrolysis, H2O is replaced by an OH group because of the loss of protons. The condensation reaction leads to the construction of M–OH–M (ol) or M–O–M (oxo) bridges after the elimination of the water molecules57. The aging process makes the gel more resistant to capillary stress and increases its mechanical properties57. Ward and Ko reported two main concepts of the sol-gel process58. In the first concept, a gel is formed because of the condensation of partially hydrolyzed species into a three-dimensional polymeric network. And in the second concept, the properties of the gel depend significantly on the synthesis conditions. The surfactants used in the current study played a deci- sive role in the development of the catalysts by generating a good porous structure that contributed to the high specific surface area2. Figure 8 shows the possible Pluronic P123 and CTAB templating routes for the formation of ZrO2 nanoparticles. Both CTAB and Pluronic P123 are well dispersed in polar solvents (especially water) to form micelles, which consist of a hydrophilic head (outward arrangement) and a hydrophobic tail (Fig. 8a). When added into the template solution, the zirconyl precursor assembles and attaches to the hydrophilic head, as shown in Fig. 8b. The zirconyl precursor solution is naturally acidic (pH <1). Under acidic conditions, hydrolysis occurs at a faster rate than condensation and results in a weak branched gel59. The addition of ammonia increases the pH of the solution. At this stage, the condensation accelerates compared to hydrolysis, thus forming a rigid gel. Upon stirring, the zirconyl precursor spreads uniformly in the template solution. Internal pressure is generated when the gel is transferred in the autoclave, where the employed heat treatment forces out the micelles and prevents the aggregation of the zirconyl particles, thus contributing to a high surface area7. Finally, removal of the template at 500 °C provides a large number of pores with a high pore volume of ZrO2 (Fig. 8c). Reaction mechanism of biodiesel production using a bifunctional acid–base catalyst. Results and Discussionf Apart from the weak strength basic sites, ZrO2(CTAB) displayed multiple desorption peaks at desorption temperatures between 468 °C and 525 °C, indicating medium to strong strength of the basic sites.h The morphologies of the synthesized catalysts are illustrated in Fig. 6(a–e). Figure 6a depicts the small pore openings of bare ZrO2, and Fig. 6b shows the spherical nanoparticles of ZrO2(P123) (10–20 nm) with ordered arrangement. ZrO2(CTAB) exhibited a rough catalyst surface, and irregular shapes of particles with large external pores were observed between the particles (Fig. 6d). In this analysis, the effects of the surfactant’s hydrophobic tail length were clearly determined as the long chain length of Pluronic P123 provided a better steric effect than CTAB and allowed the self-organization while significantly preventing the collapse of the pore network during Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ Figure 4. NH3-TPD profiles of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ ZrO2(CTAB) catalysts. Figure 4. NH3-TPD profiles of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ ZrO2(CTAB) catalysts. Figure 5. CO2-TPD profiles of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ ZrO2(CTAB) catalysts. Figure 5. CO2-TPD profiles of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ ZrO2(CTAB) catalysts. the drying process54. In addition, the uniform size and arrangement of the ZrO2(P123) particles explained the nar- rower pore size distribution plot of ZrO2(P123) compared to ZrO2(CTAB). As shown in Fig. 6c,e, the deposition of Bi2O3 on the outer surface of the catalysts resulted in agglomeration of the Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB) morphologies, respectively. This observation agreed with the large nanoparticle and crystallite sizes obtained previously for Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB). Supplementary Figs. S2(a,b) depicts the energy-dispersive X-ray (EDX) and mapping analyses that were applied to measure the elemental composition and distribution for Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB), respectively. On the basis of the EDX spectrum, three distinct phases of Zr, Bi, and O were clearly observed, which confirmed the presence of Bi2O3 on the surface of the ZrO2 cata- lyst. The concentrations of Bi in Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB) were also in agreement with the amount loaded during the preparation of the catalysts. Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 Results and Discussionf In het- erogeneous catalysis, the adsorption of the reactants and the desorption of the products occur on the catalyst’s surface. Therefore, both acidic and basic properties of the catalyst are important to achieve a simultaneous ester- ification and transesterification of high-FFA-content feedstock. Figure 9 shows the possible mechanism for the simultaneous reactions using the bifunctional catalyst. The five steps for a bifunctional catalytic reaction involve (1) diffusion of reactants, (2) physical adsorption of reactants, (3) surface reaction, (4) desorption of products, and (5) diffusion of products. In the first step, the FFA carbonyl group (fatty acid ester) and methanol are diffused from the bulk of solution to the internal catalyst’s surface through catalyst pores. In the second step, the FFA car- bonyl group is adsorbed on the acidic site (esterification), and a methanol molecule is adsorbed on the basic site (transesterification) of the catalyst’s surface, thereby affording a carbocation and an oxygen anion, respectively. In the third step, a tetrahedral intermediate is formed via a nucleophilic attack of the alcohol to the esters for both acidic and basic sites. In the fourth step, the hydroxyl group is distracted from the tetrahedral intermediate to Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ Figure 6. Field electron scanning electron microscope (FESEM) images of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ZrO2(CTAB) catalysts. Figure 6. Field electron scanning electron microscope (FESEM) images of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ZrO2(CTAB) catalysts. Figure 6. Field electron scanning electron microscope (FESEM) images of (a) bare ZrO2, (b) ZrO2(P123), (c) Bi2O3/ZrO2(P123), (d) ZrO2(CTAB), and (e) Bi2O3/ZrO2(CTAB) catalysts. Figure 7. Catalytic activity of the synthesized catalysts for biodiesel production from Nannochloropsis sp. lipid. igure 7. Catalytic activity of the synthesized catalysts for biodiesel production from Nannochloropsis sp. lipid. Figure 8. Possible Pluronic P123 and CTAB templating routes for the formation of ZrO2 catalysts. Figure 8. Possible Pluronic P123 and CTAB templating routes for the formation of ZrO2 catalysts. Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 9. The reaction mechanism of biodiesel production using a bifunctional acid–base catalyst. Figure 9. The reaction mechanism of biodiesel production using a bifunctional acid–base catalyst. form one molecule of water and one molecule of FAME on the acidic site. On the basic site, the C–O bond breaks to form one molecule of FAME and glycerol as by products. Results and Discussionf The products are desorbed from the catalyst surface as the reaction progresses. In the final step, all the products are diffused from the catalyst’s surface to the bulk of the solution and all the steps were repeated for cleavage of each fatty acid ester19,60. form one molecule of water and one molecule of FAME on the acidic site. On the basic site, the C–O bond breaks to form one molecule of FAME and glycerol as by products. The products are desorbed from the catalyst surface as the reaction progresses. In the final step, all the products are diffused from the catalyst’s surface to the bulk of the solution and all the steps were repeated for cleavage of each fatty acid ester19,60. Materials and Methods Materials. Zirconyl nitrate hydrate (ZrO(NO3)2.xH2O, 99%) and bismuth nitrate pentahydrate (Bi(NO3)3.5H2O, 98%) were purchased from Aldrich. The surfactants, namely, poly(ethylene oxide)-block-poly(propylene oxide)-block-poly(ethylene oxide) (Pluronic P123, 90%) and cetyltrimethylammonium bromide (CTAB, 90%), were supplied by Sigma. The aqueous ammonium solution (25%) was purchased from Merck. Marine microal- gae of Nannochloropsis sp. used as the biodiesel feedstock were supplied by Laboratory & Scientific Enterprise, Malaysia. Catalyst preparation. Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ Catalyst characterization. The structure of the surfactant-assisted ZrO2 catalysts was characterized by small-angle XRD analysis (Bruker AXS D8). The scanning was performed with a step of 0.02° in a 2θ range of 0° to 10°. The crystalline phases of the catalyst were characterized using wide-angle powder XRD analysis (PAN Analytical X′pert3 Powder & Empyrean) coupled with Cu-Kα radiation. The scanning was performed with a step of 0.02° and 2 s per step in a 2θ range of 10° to 80°. The crystallite sizes were defined by adopting the Debye– Scherrer formula based on the highest crystal peak. The structure of the crystalline phases was refined using the Rietveld method. The surface area, total pore volume, and pore size distribution were acquired from nitrogen adsorption–deso- rption isotherms using an adsorption porosimeter (Micromeritics ASAP 2020) at 78 K. Prior to the measurement, the catalyst was treated in vacuum at 200 °C to remove the moisture adsorbed from the catalyst surface and pores56.h y y p The acidic and basic properties of the catalyst were measured by temperature-programmed desorption (TPD; Thermo Scientific TPDRO 1100) of ammonia (NH3) and carbon dioxide (CO2). During the pre-treatment pro- cess, the samples were treated with helium (He) gas for 10 min at a rate of 20 mL/min. Then, the temperature was increased to 150 °C at a rate of 10 °C/min and was kept constant for 45 min. After being cooled down to 50 °C, the pre-treated samples were saturated with NH3 or CO2 at a rate of 20 mL/min and kept under these conditions for 60 min. Then, the samples were purged with He at a rate of 20 mL/min for 20 min to avoid physisorption and remove the remaining NH3 or CO2. Finally, the desorption of NH3 or CO2 was performed under He flow at a rate of 20 mL/ min, and the samples were heated up to 700 °C at a rate of 10 °C/min, where they were maintained for 60 min56.h p p y The morphology of the catalyst was captured using a FESEM microscope coupled with an EDX spectrometer Ziess Supra 55VP) operating at 5 kV. Biodiesel production and gas chromatography analysis. All of the catalytic reactions for biodiesel production were performed in a 50 mL three-necked flask equipped with a condenser and a stirrer. Catalyst preparation. In this study, constant reaction conditions were employed using a lipid/methanol ratio of 1:90 (g/mL) and a catalyst loading of 20 wt.% at 80 °C for 6 h. The upper layer, containing FAME, was separated from the heterogeneous catalyst by centrifugation. The biodiesel yield was measured using gas chromatography with a flame ionization detec- tor (GC-FID; Shimadzu GC-2010). BPX-20 was used as the column, with He as the carrier gas at a flow rate of 1.73 mL/min and a pressure of 83.9 kPa. The temperature of the column was first set at 150 °C and increased to 240 °C at a rate of 5 °C/min. Both the injector and FID temperatures were set at 250 °C. The biodiesel yield (%) was calculated using Eqs (3) and (4), = Σ − × × × FAME Content A A A C V m (%) 100 (3) FAME ISTD ISTD ISTD ISTD (3) Where, ∑AFAME is the total peak area of FAME, AISTD is the peak area of the internal standard, CISTD is the concentration of the internal standard (mg/mL), VISTD is the volume of the internal standard (mL), and m is the sample mass (mg). = × × FAME Yield FAME content from GC weight of biodiesel weight of microalgae lipid (%) 100 (4) (4) Conclusion RSMCCD RSMCCD was successfully employed to investigate the effect of various parameters on the surfactant-enhanced surface area of ZrO2. The high R2 obtained for ZrO2(P123) and ZrO2(CTAB) indicated that the empirical models derived from RSMCCD can effectively describe the relationship between the process parameters and the response (ZrO2 surface area). The physicochemical analyses revealed that ZrO2(P123) and ZrO2(CTAB) possess better surface area, pore structure, and acidic and basic properties compared to bare ZrO2. Moreover, the addition of Bi2O3 on ZrO2 improved the density of total acidic and basic sites of Bi2O3/ZrO2(P123) and Bi2O3/ZrO2(CTAB). Nevertheless, the success of the catalytic activity on simultaneous esterification and transesterification of microalgae lipid to biodiesel does not directly depend on the high surface area of the catalysts. In fact, the high density of the total acidic and basic sites is required for a high FAME yield. Thus, the development of bifunctional Bi2O3/ZrO2 cat- alysts has a high chance of simplifying the biodiesel production process using low-grade high-FFA feedstock. Received: 30 April 2019; Accepted: 15 October 2019; Published: xx xx xxxx Received: 30 April 2019; Accepted: 15 October 2019; Published: xx xx xxxx Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 Catalyst preparation. Catalyst preparation. A sol-gel method followed by a hydrothermal treatment was adopted for the synthe- sis of bare and surfactant-assisted ZrO2. On the basis of the typical synthesis method, 8.1 g of ZrO(NO3)2.xH2O was dissolved in 30 mL of distilled water, followed by the addition of 120 mL of absolute ethanol. The solution was vigorously stirred for about 20 min at room temperature to achieve homogenization. An aqueous ammo- nium solution (25%) was then added dropwise to the above solution until pH 8 was attained. The new solution was continuously stirred until gelling, and the obtained sample was transferred into a Teflon-lined autoclave. The vessel was sealed and heated in an oven at 120 °C for 24 h. The resulting gel was washed several times with distilled water and ethanol, dried at 120 °C, and calcined at 500 °C for 4 h. The surfactant-assisted ZrO2 samples were prepared similarly to the bare ZrO2, and the surfactant was added after the addition of absolute ethanol. The obtained calcined bare ZrO2, Pluronic P123-assisted ZrO2, and CTAB-assisted ZrO2 samples were designated as ZrO2, ZrO2(P123), and ZrO2(CTAB), respectively. Optimization study. A four-factor RSMCCD was implemented to study the effects of the independ- ent parameters, i.e., surfactant/Zr molar ratio (A: 0.01–0.05 for Pluronic P12340 and 0.6–1.0 for CTAB39), pH (B: 9–11), aging time (C: 12–48 h), and temperature (D: 80–120 °C), on the surface area of ZrO2. The statistical calculations were performed using Design Expert Version 11 (STAT-EASE Inc., Minneapolis, USA). The exper- imental values were compared with the predicted values to test the adequacy of the final reduced model. The recommended optimum conditions were experimentally implemented to validate the optimum surface area value predicted by the model. Impregnation with Bi2O3. To investigate the catalytic activities of mixed zirconia oxides, Bi2O3/ZrO2 sam- ples were prepared via a simple incipient wetness impregnation method. The representative ZrO2 was selected based on the optimum surface areas of ZrO2(P123) and ZrO2(CTAB). 5.8 g of Bi(NO3)3·5H2O (which corresponded to 5 wt.% of Bi2O3) was dissolved in distilled water, and 47.5 g of ZrO2(P123) and ZrO2(CTAB) were added (separately) into the metal solution, which was stirred for 24 h. Water was then removed by drying in an oven at 120 °C and subsequently by calcination at 500 °C for 5 h using a muffle furnace. The samples were designated as Bi2O3/ ZrO2(P123) and Bi2O3/ZrO2(CTAB). 6. Vicente, G., Carrero, A., Rodrïguez, R. & del Peso, G. L. Heterogeneous-catalysed direct transformation of microalga biomass into biodiesel-Grade FAMEs. Fuel 200, 590–598 (2017).i gi J p q p ( ) 2. Guldhe, A., Singh, P., Ansari, F. A., Singh, B. & Bux, F. Biodiesel synthesis from microalgal lipids using tungstated zirconia as a heterogeneous acid catalyst and its comparison with homogeneous acid and enzyme catalysts. Fuel 187, 180–188 (2017). g y p g y y ( ) 3. Heater, B. S., Lee, M. M. & Chan, M. K. Direct production of a genetically- encoded immobilized biodiesel catalyst. Sci. Rep. 8, 12783 (2018). www.nature.com/scientificreports/ Sol-Gel Sci. Technol. 67, 670–674 (2013). g y and oxygen vacancies. J. Sol-Gel Sci. Technol. 67, 670–674 (2013). yg 9. Chandra, N. et al. Synthesis and characterization of nano-sized zirconia powder synthesized by single emulsion-assisted direc precipitation. J. Colloid Interface Sci. 342, 327–332 (2010).f 29. Chandra, N. et al. Synthesis and characterization of nano- precipitation. J. Colloid Interface Sci. 342, 327–332 (2010).f p p f 0. Goharshadi, E. K. & Hadadian, M. Effect of calcination temperature on structural, vibrational, optical, and rheological properties o zirconia nanoparticles. Ceram. Int. 38, 1771–1777 (2012). 1. Guel, M. L. A., Jiménez, L. D. & Hernández, D. A. C. Ultrasound-assisted sol–gel synthesis of ZrO2. Ultrason. Sonochem. 35, 514–517 (2017). ( ) 2. Djerad, S., Geiger, B., Schott, F. J. P. & Kureti, S. Synthesis of nano-sized ZrO2 and its use as catalyst support in SCR. Catal. Commun 10, 1103–1106 (2009). ( ) 3. Zhang, H., Hardy, G. C., Khimyak, Y. Z., Rosseinsky, M. J. & Cooper, A. I. Synthesis of hierarchically porous silica and metal oxide beads using emulsion-templated polymer scaffolds. Chem. Mater. 16, 4245–4256 (2004).hlf g p p yf 34. Eltejaei, H. et al. The influence of preparation conditions on ZrO2 nanoparticles with different PEG - PPG - PEG surfactants by statistical experimental design. Mater. Lett. 65, 2913–2916 (2011).hf hl statistical experimental design. Mater. Lett. 65, 2913–2916 (2011) p g 35. Poyraz, A. S., Albayrak, C. & Dag, Ö. The effect of cationic surfactant and some organic–inorganic additives on the morphology of mesostructured silica templated by pluronics. Micropor. Mesopor. Mater. 115, 548–555 (2008). p y p p p 6. Sidhu, G. K. & Kumar, R. Role of anionic and cationic surfactants on the structural and dielectric properties of ZrO2 nanoparticles Appl. Surf. Sci. 392, 598–607 (2017). 37. Esposito, S. ‘Traditional’ sol–gel chemistry as a powerful tool for the preparation of supported metal and metal oxide catalysts. Materials (Basel) 12, 1–25 (2019). 8. Zhang, J., Song, B., Peng, W., Feng, Y. & Xu, B. CTAB-assisted hydrothermal synthesis of nano-sized tetragonal zirconium dioxide Mater. Chem. Phys. 123, 606–609 (2010). y 9. Shi, W., Liu, H., Ren, D., Ma, Z. & Sun, W. Hydrothermal synthesis, characterization and catalytic properties of nanoporous MoO3 ZrO2 mixed Oxide. Chem. Res. Chin. Univ. 22, 364–367 (2006). 40. Rezaei, M., Alavi, S. M., Sahebdelfar, S., Xinmei, L. & Yan, Z. F. www.nature.com/scientificreports/ Alumina supported/unsupported mixed oxides of Ca and Mg as heterogeneous catalysts fo transesterification of Nannochloropsis sp. microalga’s oil. Energy Convers. Manag. 88, 1193–1199 (2014).i i 6. Zeng, H. Y., Xu, S., Liao, M. C., Zhang, Z. Q. & Zhao, C. Activation of reconstructed Mg/Al hydrotalcites in the transesterification o microalgae oil. Appl. Clay Sci. 91–92, 16–24 (2014).i g pp y 7. Umdu, E. S., Tuncer, M. & Seker, E. Transesterification of Nannochloropsis oculata microalga’s lipid to biodiesel on Al2O3 supported CaO and MgO catalysts. Bioresour. Technol. 100, 2828–2831 (2009). g y 8. Farooq, M., Ramli, A. & Subbarao, D. Biodiesel production from waste cooking oil using bifunctional heterogeneous solid catalysts J. Clean. Prod. 59, 131–140 (2013).i 9. Wan Omar, W. N. N. W. & Amin, N. A. S. Biodiesel production from waste cooking oil over alkaline modified zirconia catalyst. Fue Process. Technol. 92, 2397–2405 (2011). 20. Borges, M. E. & Díaz, L. Recent developments on heterogeneous catalysts for biodiesel production by oil esterification and transesterification reactions: A review. Renew. Sustain. Energy Rev. 16, 2839–2849 (2012).i i 21. Salinas, D., Guerrero, S. & Araya, P. Transesterification of canola oil on potassium-supported TiO2 catalysts. Catal. Commun. 11, 773–777 (2010). 2. Wen, Z., Yu, X., Tu, S. T., Yan, J. & Dahlquist, E. Bioresource Technology biodiesel production from waste cooking oil catalyzed by TiO2 - MgO mixed oxides. Bioresour. Technol. 101, 9570–9576 (2010). 23. Macario, A. et al. Biodiesel production process by homogeneous/heterogeneous catalytic system using an acid–base catalyst. Appl. Catal. A: Gen. 378, 160–168 (2010). 24. Deshmane, V. G. & Adewuyi, Y. G. Synthesis of thermally stable, high surface area, nanocrystalline mesoporous tetragonal zirconium dioxide (ZrO2): effects of different process parameters. Micropor. Mesopor. Mater. 148, 88–100 (2012). ff 5. Liu, Y., Chi, W., Liu, H., Su, Y. & Zhao, L. Preparation of t-ZrO2 by a sol–gel process with carbon as a phase transformation promoter RSC Adv. 5, 34451–34455 (2015).h 6. Gopalan, R., Chang, C.-H. & Lin, Y. S. Thermal stability improvement on pore and phase structure of sol–gel derived zirconia. J Mater. Sci. 30, 3075–3081 (1995). 27. Bumajdad, A. et al. Controlled synthesis of ZrO2 nanoparticles with tailored size, morphology and crystal phases via organic–inorganic hybrid films. Sci. Rep. 8, 3695 (2018). g g yi p 8. Mahmood, Q., Afzal, A., Siddiqi, H. M. & Habib, A. Sol–gel synthesis of tetragonal ZrO2 nanoparticles stabilized by crystallite size and oxygen vacancies. J. References 1. Lau, P. C., Kwong, T. L. & Yung, K. F. Effective heterogeneous transition metal glycerolates catalysts for one-step biodiesel production from low grade non-refined Jatropha oil and crude aqueous bioethanol. Sci. Rep. 6, 23822 (2016). gi p q p 2. Guldhe, A., Singh, P., Ansari, F. A., Singh, B. & Bux, F. Biodiesel synthesis from microalgal lipids using tungstated zirconia as a heterogeneous acid catalyst and its comparison with homogeneous acid and enzyme catalysts. Fuel 187, 180–188 (2017). i 2. Guldhe, A., Singh, P., Ansari, F. A., Singh, B. & Bux, F. Biodiesel synthesis from microalgal lipids using tungstated zirconia as a heterogeneous acid catalyst and its comparison with homogeneous acid and enzyme catalysts. Fuel 187, 180–188 (2017). 3. Heater, B. S., Lee, M. M. & Chan, M. K. Direct production of a genetically- encoded immobilized biodiesel catalyst. Sci. Rep. 8, 12783 (2018). 4. Su, F., Li, G., Zhang, H. & Yan, Y. Enhanced performance of Rhizopus oryzae lipase immobilized on hydrophobic carriers and it application in biorefinery of rapeseed oil deodorizer distillate. Bioenerg. Res. 7, 935–945 (2014).i i 5. Cercado, A. P., Ballesteros, F. & Capareda, S. Ultrasound assisted transesterification of microalgae using synthesized novel catalyst. Sustain. Environ. Res. 28, 234–239 (2018). i 5. Cercado, A. P., Ballesteros, F. & Capareda, S. Ultrasound assisted transesterification of microalgae using synthesized novel catalyst Sustain. Environ. Res. 28, 234–239 (2018). 6. Vicente, G., Carrero, A., Rodrïguez, R. & del Peso, G. L. Heterogeneous-catalysed direct transformation of microalga biomass into biodiesel-Grade FAMEs. Fuel 200, 590–598 (2017). R bi h Ni h M F l P d i f bi di l f dibl J h il i ifi i i Bi O L O 6. Vicente, G., Carrero, A., Rodrïguez, R. & del Peso, G. L. Heterogeneous-catalysed direct transformation of microalga biomass into biodiesel-Grade FAMEs. Fuel 200, 590–598 (2017).i ( ) 7. Rabiah Nizah, M. F. et al. Production of biodiesel from non-edible Jatropha curcas oil via transesterification using Bi2O3-La2O catalyst. Energy Convers. Manag. 88, 1257–1262 (2014). Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 10 Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ www.nature.com/scientificreports/ 8. Firdaus, M. Y., Brask, J., Nielsen, P. M., Guo, Z. & Fedosov, S. Kinetic model of biodiesel production catalyzed by free liquid lipase from Thermomyces lanuginosus. J. Mol. Catal. B 133, 55–64 (2016).i h y g J ( ) 9. Tangy, A. et al. In-situ transesterification of Chlorella vulgaris using carbon-dot functionalized strontium–oxide as a heterogeneou catalyst under microwave irradiation. Energy Fuels 30, 10602–10610 (2016).ii y gy 10. Teo, S. H., Islam, A., Yusaf, T. & Taufiq-Yap, Y. H. Transesterification of Nannochloropsis oculata microalga’s oil to biodiesel using calcium methoxide catalyst. Energy 78, 63–71 (2014). l l f b b d l d d l l d d l l ll l f fi i q p calcium methoxide catalyst. Energy 78, 63–71 (2014). y gy 1. Ryu, Y. J. et al. Development of carbon-based solid acid catalysts using a lipid-extracted alga, Dunaliella tertiolecta, for esterification J. Microbiol. Biotechnol. 28, 732–738 (2018). ( ) 12. Carrero, A., Vicente, G., Rodríguez, R., del Peso, G. L. & Santos, C. Synthesis of fatty acids methyl esters (FAMEs) N hl i dit i l i h t id t l t Bi h E J 97 119 124 (2015) 12. Carrero, A., Vicente, G., Rodríguez, R., del Peso, G. L. & Santos, C. Synthesis of fatty acids methyl esters (FAMEs) from Nannochloropsis gaditana microalga using heterogeneous acid catalysts. Biochem. Eng. J. 97, 119–124 (2015). p g g g g y g ( ) 13. Carrero, A., Vicente, G., Rodríguez, R., Linares, M. & Del Peso, G. L. Hierarchical zeolites as catalysts for biodiesel production from Nannochloropsis microalga oil. Catal. Today 167, 148–153 (2011).i Nannochloropsis microalga oil. Catal. Today 167, 148–153 (2011). 4 Xu, S et al Mg-Fe mixed oxides as solid base catalysts for the transesterification of microalgae oil RSC Adv 5, 71278–71286 (2015) 14. Xu, S. et al. Mg-Fe mixed oxides as solid base catalysts for the tr 4. Xu, S. et al. Mg Fe mixed oxides as solid base catalysts for the transesterification of microalgae oil. RSC Adv. 5, 71278 71286 (2015) 5. Teo, S. H., Taufiq-Yap, Y. H. & Ng, F. L. Alumina supported/unsupported mixed oxides of Ca and Mg as heterogeneous catalysts for transesterification of Nannochloropsis sp. microalga’s oil. Energy Convers. Manag. 88, 1193–1199 (2014).i 5. Teo, S. H., Taufiq-Yap, Y. H. & Ng, F. L. www.nature.com/scientificreports/ Preparing catalytic materials by the sol–gel method. Ind. Eng. Chem. Res. 34, 421–433 (199 9. Teo, S. H., Islam, A. & Taufiq-Yap, Y. H. Algae derived biodiesel using nanocatalytic transesterification process. Chemical Engineering Research and Design 111, 362–370 (2016).iii 60. Yan, S., Salley, S. O. & Simon Ng, K. Y. S. Simultaneous transesterification and esterification of unrefined or waste oils over ZnO- La2O3 catalysts. Appl. Catal. A 353, 203–212 (2009). Acknowledgementsh g The authors acknowledge the support provided by Yayasan Universiti Teknologi PETRONAS (YUTP). The authors also extend the appreciation to Malaysia Ministry of High Education (MOHE) and Universiti Sains Islam Malaysia (USIM) for the scholarship awarded to N. J. A. R. Competing interestsh p g The authors declare no competing interests. Author contributionsh This project and the initial idea were conceived and originated by A.R. Experiments were performed by N.J.A.R. The manuscript was written by N.J.A.R. and scientific guidance was provided by K.J. and Y.U. The manuscript was reviewed by A.R. www.nature.com/scientificreports/ h k ff f d f Wu, H., Jia, J., Ma, B. & Li, J. Synthesis of bimodal mesoporous silica . Yang, L., Wu, H., Jia, J., Ma, B. & Li, J. Synthesis of bimodal mesopo 3. Yang, L., Wu, H., Jia, J., Ma, B. & Li, J. Synthesis of bimodal mesoporous silica with coexisting phases by co-hydrothermal aging route with P123 containing gel and F127 containing gel. Micropor. Mesopor. Mater. 253, 151–159 (2017). h k ff f d f h f b 53. Yang, L., Wu, H., Jia, J., Ma, B. & Li, J. Synthesis of bimodal mesoporous silica with coexisting phases by co-hydrothermal aging with P123 containing gel and F127 containing gel. Micropor. Mesopor. Mater. 253, 151–159 (2017). Ch k ff f d f h f g g g g p p ( ) 54. Jung, H.-N.-R., Han, W., Cho, H. H. & Park, H.-H. Effect of cationic and non-ionic surfactants on the microstructure of ambient pressure dried zirconia aerogel. Mater. Express 7, 291–298 (2017). 54. Jung, H.-N.-R., Han, W., Cho, H. H. & Park, H.-H. Effect of cationic and non-ionic surfactants on the microstructure o pressure dried zirconia aerogel. Mater. Express 7, 291–298 (2017).l gf pressure dried zirconia aerogel. Mater. Express 7, 291–298 (2017) p g p ( ) 5. Stichert, W. & Schüth, F. Influence of crystallite size on the properties of zirconia. Chem. Mater. 10, 2020–2026 (1998). p g p 55. Stichert, W. & Schüth, F. Influence of crystallite size on the properties of zirconia. Chem. Mater. 10, 2020–2026 (1998). , ,l y p p , ( ) 56. Abd Rahman, N. J., Ramli, A., Jumbri, K. & Uemura, Y. Biodiesel production from N. oculata microalgae lipid in the presence of Bi2O3/ZrO2 catalysts. Waste Biomass Valorization 0, 0 (2019). l y p p 56. Abd Rahman, N. J., Ramli, A., Jumbri, K. & Uemura, Y. Biodiesel production from N. oculata microalgae lipid in the Bi2O3/ZrO2 catalysts. Waste Biomass Valorization 0, 0 (2019). 2 3 2 y ( ) 7. Pierre, A. C. & Pajonk, G. M. Chemistry of aerogels and their applications. Chem. Rev. 102, 4243–4265 (2002). d l l b h l l h d d h ( 2 3 2 y ( ) 57. Pierre, A. C. & Pajonk, G. M. Chemistry of aerogels and their applications. Chem. Rev. 102, 4243–4265 (2002). 58. Ward, D. A. & Ko, E. I. www.nature.com/scientificreports/ Synthesis of mesoporous nanocrystalline zirconia with tetragonal crystallite phase by using ethylene diamine as precipitation agent. J. Mater. Sci. 42, 7086–7092 (2007).f 41. Rezaei, M., Alavi, S. M., Sahebdelfar, S. & Yan, Z. F. Effect of process parameters on the synthesis of mesoporous nanocrystalline zirconia with triblock copolymer as template. J. Porous Mater. 15, 171–179 (2008).i p y p 2. Wu, L. et al. Bentonite-enhanced biodiesel production by NaOH-catalyzed transesterification: process optimization and kinetics and thermodynamic analysis. Fuel 182, 920–927 (2016). y y 43. Kibar, M. E., Ozcan, O. & Yonel-Gumruk, E. Optimization, modeling and characterization of sol–gel process parameters for the synthesis of nanostructured boron doped alumina catalyst supports. Micropor. Mesopor. Mater. 229, 134–144 (2016). 44. Yang, P. & Margolese, D. I. Generalized syntheses of large-pore mesoporous metal oxides with semicrystalline frameworks, 396, 6–9 (1998) 5. Wang, P., Zhao, Z., Wang, Z., Chen, S. & Fan, G. Hydrothermal synthesis of mesoporous nanocrystalline tetragonal ZrO2 using Dehydroabietyltrimethyl ammonium bromine. BioResources 10, 1271–1284 (2015).hl 46. Chuah, G. K. & Jaenicke, S. The preparation of high surface area zirconia - Influence of precipitating agent and digestion. Appl. Catal. A 163, 261–273 (1997).i 47. Gauna, M. R. et al. Monoclinic-tetragonal zirconia quantification of commercial nanopowder mixtures by XRD and DTA. Ceram. Silik. 50, 318–325 (2015). 48. Refaat, A. A. Biodiesel production using solid metal oxide catalysts. Int. J. Environ. Sci. Technol. 8, 203–221 (2011). 49. Garvie, R. C. Stabilization of the tetragonal structure in zirconia microcrystals. J. Phys. Chem. 82, 218–224 (1978). Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9 www.nature.com/scientificreports/ 50. Stichert, W., Schüth, F., Kuba, S. & Knözinger, H. Monoclinic and tetragonal high surface area sulfated zirconias in bu isomerization: CO adsorption and catalytic results. J. Catal. 198, 277–285 (2001). y 51. Sing, K. S. W. Reporting physisorption data for gas/solid systems with special reference to the determination of surface area porosity. Pure Appl. Chem. 54, 2201–2218 (1982). p y pp , ( ) 52. Kruk, M. & Jaroniec, M. Gas adsorption characterization of ordered organic–inorganic nanocomposite materials. Chem. Mater. 13, 3169–3183 (2001). p y pp , ( ) 52. Kruk, M. & Jaroniec, M. Gas adsorption characterization of ordered organic–inorganic nanocomposite materials. Chem. Mater. 13, 3169–3183 (2001). 53. Yang, L., Wu, H., Jia, J., Ma, B. & Li, J. Synthesis of bimodal mesoporous silica with coexisting phases with P123 containing gel and F127 containing gel. Micropor. Mesopor. Mater. 253, 151–159 (2017). Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41598-019-52771-9. Supplementary information is available for this paper at https://doi.org/10.10 Correspondence and requests for materials should be addressed to A.R. Correspondence and requests for materials should be addressed to A.R. Reprints and permissions information is available at www.nature.com/reprints. 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 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) 2019 Scientific Reports \| (2019) 9:16223 \| https://doi.org/10.1038/s41598-019-52771-9
https://openalex.org/W4297369980	https://ojs.altstu.ru/index.php/PolzVest/article/download/175/207	Russian	null	ИССЛЕДОВАНИЕ ВИТАМИННОГО СОСТАВА РАСТОРОПШИ ПЯТНИСТОЙ	Polzunovskij vestnik	2,022	cc-by	3,855	Ползуновский вестник. 2022. № 3. С. 160‒165. Polzunovskiy vеstnik. 2022;3: 160‒165. EDN: VQMQMP doi: 10.25712/ASTU.2072-8921.2022.03.022 _______________  Кашолкина, Д. А., Болгова, М. А., Клейменова, Н. Л., Копылов, М. В., Болгова, И. Н., 2022 160 ПОЛЗУНОВСКИЙ ВЕСТНИК № 3 2022 160 ПОЛЗУНОВСКИЙ ВЕСТНИК № 3 2022 _________________________________________________________________________________ Для цитирования: Исследование витаминного состава расторопши пятнистой / Д.А. Кашолкина [и др.]. // Ползуновский вестник. № 3, 2022. С. 160 – 165. doi: 10.25712/ASTU.2072- 8921.2022.03.022. EDN: https://elibrary.ru/vqmqmp. _______________  Кашолкина, Д. А., Болгова, М. А., Клейменова, Н. Л., Копылов, М. В., Болгова, И. Н., 2022 1, 2, 3, 4 ,5 Voronezh State University of Engineering Technologies, Voronezh, Russia 1 dasha08082000@mail.ru 1, 2, 3, 4 ,5 Voronezh State University of Engineering Technologies, Voronezh, Russia 1 dasha08082000@mail.ru @ 2 bolgova.masha20@yandex.ru, https://orcid.org/0000-0001-7464-3883 3 klesha78@list.ru, http://orcid.org/0000-0002-1462-4055 4 kopylov-maks@yandex.ru, http://orcid.org/0000-0003-2678-2613 5 bolgovainessa@yandex.ru, http://orcid.org/0000-0003-0915-8405 2 bolgova.masha20@yandex.ru, https://orcid.org/0000-0001-7464-3883 3 klesha78@list.ru, http://orcid.org/0000-0002-1462-4055 4 kopylov-maks@yandex.ru, http://orcid.org/0000-0003-2678-2613 5 bolgovainessa@yandex.ru, http://orcid.org/0000-0003-0915-8405 Abstract. The subject of this scientific article is the seeds of milk thistle. Milk thistle is a well- known medicinal plant rich in valuable biologically active substances, macro and microelements, and vitamins. Milk thistle belongs to nutraceuticals, the study of the composition of the seeds of the plant is of practical importance for the creation of oils of functional value to meet the physiological needs of a person in useful substances necessary for the body. In addition, milk thistle seeds can be treated as a biologically active dietary supplement. The study of the vitamin composition, the content of micro- and macroelements in the seeds of milk thistle was the purpose of the study, carried out according to standard techniques using classical and spectral methods. In the course of experimental studies, the presence of β-carotene, vitamin B1, vitamin B2, vitamin B4 and vitamin E was found in the seeds of milk thistle. Choline (B4) has the highest content. The presence of α-tocopherol, or vitamin E, prevents the oxidation of body lipids, including polyunsaturated fatty acids and lipid components of cells and membrane organelles. It was also found that oilseeds contain iron, zinc, manganese, iodine, selenium, copper, chromium, potassium, calcium, magnesium, phosphorus. Obtained results which have shown that milk thistle seeds are rich in valuable biocompounds, are necessary for the analysis of seeds in order to design the composition of food products, taking into account the recommendations of the daily intake of vitamins and minerals. Thus, milk thistle seeds can be recommended as a potential source of functional supplements for preventive nutrition and therapeutic nutrition as part of complex therapy. Keywords: milk thistle vitamin composition oil macronutrients trace elements seeds Keywords: milk thistle, vitamin composition, oil, macronutrients, trace elements, seeds. Acknowledgements: the author expresses gratitude to his / her colleagues for their help, thanks for the financial support of the research. _________________________________________________________________________________ For citation: Kasholkina, D. A., Bolgova, M. A., Kleymenova, N. L., Kopylov, M. V.  Bolgova, I. N. (2022). Study of the vitamin composition of milk thistle. Polzunovskiy vеstnik, 3, 160-165. (In Russ.). doi: 10.25712/ASTU.2072-8921.2022.03.022. , 3, 4, 5 Воронежский государственный университет инженерных технологий, Воронеж, Россия dasha08082000@mail.ru , 3, 4, 5 Воронежский государственный университет инженерных технологий, Воронеж, Россия dasha08082000@mail.ru 2 bolgova.masha20@yandex.ru, https://orcid.org/0000-0001-7464-3883 3 klesha78@list.ru, http://orcid.org/0000-0002-1462-4055 4 kopylov-maks@yandex.ru, http://orcid.org/0000-0003-2678-2613 5 bolgovainessa@yandex.ru, http://orcid.org/0000-0003-0915-8405 p g 4 kopylov-maks@yandex.ru, http://orcid.org/0000-0003-2678-2613 5 bolgovainessa@yandex.ru, http://orcid.org/0000-0003-0915-8405 Аннотация. Предметом исследования данной научной статьи являются семена расто- ропши пятнистой. Расторопша – известное лекарственное растение, богатое ценными биологически активными веществами, микро- и макроэлементами, витаминами. Растороп- ша относится к нутрицевтикам, исследование состава семян которых имеет практическое значение для создания продуктов функциональной направленности с целью удовлетворения физиологической потребности человека в необходимых для организма полезных веществах. Помимо этого семена расторопши могут являться биологически активной добавкой в пи- тании. Изучение витаминного состава, содержания микро- и макроэлементов в семенах расторопши пятнистой являлось целью исследования, которое проводилось по стандарт- ным методикам с использованием классических и спектральных методов. В ходе экспери- ментальных исследований в семенах расторопши пятнистой установлено присутствие β- каротина, витамина В1, витамина В2, витамина В4 и витамина Е. Самое высокое содержа- ние имеет холин (В4). Также было установлено, что семена масличной культуры содержат железо, цинк, марганец, йод, селен, медь, хром, калий, кальций, магний, фосфор. Содержание железа, цинка преобладает среди микроэлементов, количество фосфора и кальция - среди макроэлементов. О богатстве семян расторопши ценными биосоединениями говорят полу- ченные результаты, которые необходимы для анализа семян с целью проектирования со- става пищевых продуктов с учетом рекомендаций суточного потребления витаминов и ми- нералов. Таким образом, семена расторопши можно рекомендовать как потенциальный ис- точник функциональных добавок при профилактическом питании и в лечебном питании в составе комплексной терапии. р Ключевые слова: расторопша пятнистая, витаминный состав, микроэлементы, мак- роэлементы, семена. Благодарности: автор выражает признательность коллегам за помощь, благодар- ность за финансовую поддержку исследования. _________________________________________________________________________________ Для цитирования: Исследование витаминного состава расторопши пятнистой / Д.А. Кашолкина [и др.]. // Ползуновский вестник. № 3, 2022. С. 160 – 165. doi: 10.25712/ASTU.2072- 8921.2022.03.022. EDN: https://elibrary.ru/vqmqmp. 160 ИССЛЕДОВАНИЕ ВИТАМИННОГО СОСТАВА РАСТОРОПШИ ПЯТНИСТОЙ Original article МЕТОДЫ Объектом исследования были выбраны семена расторопши пятнистой. Проведены исследования витаминного состава семян расторопши в соответствии с нормативными документами [10-14]. Следует отметить тот факт, что, несмот- ря на достаточно хорошо изученный состав плодов, в литературных источниках практи- чески отсутствует комплексная информация о химическом составе. Проведено исследование минеральных элементов семян расторопши пятнистой ме- тодами по ГОСТ 26573.2-2014, НСАМ № 512- МС [15,16]. В научной литературе ограничены све- дения о витаминном составе семян расто- ропши пятнистой и практически не встреча- ются данные о содержании микро- и макро- элементов. В связи с этим целью данной ра- боты является определение химического со- става в числовых показателях. Д. А. КАШОЛКИНА, М. А. БОЛГОВА, Н. Л. КЛЕЙМЕНОВА, М. В. КОПЫЛОВ, И. Н. БОЛГОВА Растение богато витаминами группы B и E, ß-каротином, микро- и макроэлементами. веществами связано с тем, что потребление продуктов, содержащих природные биологи- чески активные вещества, не удовлетворяет потребности организма и недостаточно для профилактики алиментарно-зависимых забо- леваний. Расторопша относится к нутрицевтикам, которые используются для лечебно- профилактического питания в виде биологи- чески активных добавок – флаволигнанов. Растение богато пищевыми волокнами, рас- тительными компонентами [5]. Семена расто- ропши являются идеальным ингредиентом как для фармацевтического, так и для косме- тического применения [6, 7]. Кроме того, се- мена могут быть использованы в качестве подходящего пищевого ингредиента в про- дуктах с низким содержанием клетчатки. Цель исследования – изучение витамин- ного состава, содержания микро- и макро- элементов в семенах расторопши пятнистой. ВВЕДЕНИЕ ряд необходимых для организма человека полезных и незаменимых веществ [1]. Повышение осведомленности потреби- телей о здоровом питании связано с развити- ем науки, а также с интерпретацией науки для широкой аудитории, которые были попу- ляризированы в журналах, телевизионных программах и в Интернете. В дополнение к сенсорным свойствам потребители ожидают, что пища положительно повлияет на их здо- ровье, благополучие и качество жизни. Расторопша пятнистая включает в себя комплекс биологически активных веществ, более значимые из них: силикристин, сили- дианин, силибин (флаволигнаны) [2]. Данные соединения определяют важные фармаколо- гические свойства растения – гепатопротек- торное, антиоксидатное и антитоксическое. Противоспалительные и антиканцерогенные свойства данного растения определяют со- держащиеся в семенах полифенолы [3]. POLZUNOVSKIY VESTNIK № 3 2022 Расторопша пятнистая в изобилии рас- тет около дорог в России и давно признана средством от заболеваний печени и желче- выводящих путей. Она классифицируется как лекарственное растение. Также является уникальным растением, которое содержит На основе этой масличной культуры из- готавливаются такие гепатопротекторные ле- карственные средства, как легалон, карсил, силимар, силибор [4]. 161 ИССЛЕДОВАНИЕ ВИТАМИННОГО СОСТАВА РАСТОРОПШИ ПЯТНИСТОЙ ИССЛЕДОВАНИЕ ВИТАМИННОГО СОСТАВА РАСТОРОПШИ ПЯТНИСТОЙ РЕЗУЛЬТАТЫ Результаты исследований витаминного состава семян расторопши пятнистой пред- ставлены на рисунке 1. В работе [8] продемонстрировано, что погодные условия в большей степени влияют на химический состав плодов расторопши, чем агротехнические условия. Содержание макроэлементов в г/кгдм было следующим: фосфор – 6,1, калий – 4,95; кальций – 7,6; магний – 2,6. Было обнаружено высокое со- держание железа – 82,3 мг/ кгдм. Рисунок 1 – Витаминный состав семян расторопши пятнистой Figure 1 – Vitamin composition of milk thistle seeds На рисунках 2 а, б и 3 представлено установленное в семенах содержание микро- и макроэлементов, соответственно. Авторы Apostol L., Iorga C.S., Moşoiu C.E., Mustățea G., & Cucu Ș.E. обнаружили, что се- мена расторопши отличаются высоким со- держанием минералов (мг/100 г): кальция (912), магния (433), железа (80,5), цинка (7,38) и меди (2,69). Из проведенных анали- зов содержания минералов можно отметить, что семена расторопши пятнистой представ- ляют собой материал с важным содержанием минералов, 100 г которого обеспечивают су- точное потребление некоторых из этих эле- ментов в соответствии с рекомендуемыми нормами потребления макро- и микронутри- ентов [9]. Рисунок 1 – Витаминный состав семян расторопши пятнистой Рисунок 1 – Витаминный состав семян расторопши пятнистой Figure 1 – Vitamin composition of milk thistle seeds На рисунках 2 а, б и 3 представлено установленное в семенах содержание микро- и макроэлементов, соответственно. Актуальность изучения этого растения является важной стратегической задачей для расширения разнообразия доступных ком- плексных биологически активных добавок в функциональных продуктах питания, которые в настоящее время находятся в центре по- вышенного внимания. Приобретение все большей популярно- сти функциональных пищевых продуктов, обогащенных витаминами, минеральными ПОЛЗУНОВСКИЙ ВЕСТНИК № 3 2022 ПОЛЗУНОВСКИЙ ВЕСТНИК № 3 2022 162 ОБСУЖДЕНИЕ а б Рисунок 2 – Содержание микроэлементов в семенах расторопши пятнистой (а, б) Figure 2 – Microelements content in milk thistle seeds (а, б) Образцы исследовали на содержание а б Рисунок 2 – Содержание микроэлементов в семенах расторопши пятнистой (а, б) Figure 2 – Microelements content in milk thistle seeds (а, б) Образцы исследовали на содержание макроэлементов, которые представлены на рисунке 3 В ходе и обнаружены: В1 (1,5 мг/10 витамин В4 мг/100 г). Ре ний позволяю пищу семена собны удовл требность в ных категори Анализ д кое содержа играет важн нервной сис рина в крови липидном о снижению ве [18]. Ранее а дование жир става коричне сравнении р культурам вы витаминов бо нах льна, в имеются отл можно испол гокомпонентн ных биологич Семена минеральные мы для пита образцов бы марганец (12 лен (220 мкг/ установили семенах раст ций (1660 м В ходе исследования семян расторопши обнаружены: β-каротин (0,8 мг/100 г), витамин В1 (1,5 мг/100 г), витамин В2 (1,1 мг/100 г), витамин В4 (10 мг/100 г) и витамин Е (4,7 мг/100 г). Результаты полученных исследова- ний позволяют рекомендовать для принятия в пищу семена расторопши, так как они спо- собны удовлетворить физиологическую по- требность в исследуемых веществах различ- ных категорий населения. Анализ данных показал, что самое высо- кое содержание имеет холин (В4), который играет важную роль в функционировании нервной системы, снижает уровень холесте- рина в крови, участвует в углеводном обмене, липидном обмене в печени, способствует снижению веса и уменьшает риск диабетов [18]. Ранее авторами было проведено иссле- дование жирно-кислотного и химического со- става коричневых и белых семян льна [17]. При сравнении результатов по двум масличным культурам выявили, что качественный состав витаминов более широко представлен в семе- нах льна, в количественном составе также имеются отличия. Эти данные впоследствии можно использовать при проектировании мно- гокомпонентных комплексных сбалансирован- ных биологически активных добавок. а а б а б б Рисунок 2 – Содержание микроэлементов в семенах расторопши пятнистой (а, б) Рисунок 2 – Содержание микроэлементов в семенах расторопши пятнистой (а, б) Семена расторопши пятнистой содержат минеральные элементы, которые необходи- мы для питания человека. При исследовании образцов были обнаружены микроэлементы: марганец (12 мг/100 г), йод (11 мкг/100 г), се- лен (220 мкг/100 г), хром (21 мкг/100 г). Также установили содержание макроэлементов в семенах растения: калий (920 мг/100 г), каль- ций (1660 мг/100 г), магний (420 мг/100 г), фосфор (9600 мг/100 г). Образцы исследовали на содержание макроэлементов, которые представлены на рисунке 3. СПИСОК ЛИТЕРАТУРЫ 15. ГОСТ 26573.2-2014 Премиксы. Методы опреде- ления марганца, меди, железа, цинка, кобальта: введ. 2016-01-01. Доступ из справ. – правовой системы «Кон- сультант плюс» (дата обращения: 04.05.2022). 1. Рамазанов А.Ш., Балаева Ш.А., Шахбанов К.Ш. Химический состав плодов и масла расторопши пятни- стой, произрастающей на территории Республики Даге- стан // Химия растительного сырья. 2019. №2. С. 113-118. doi:10.14258/jcprm.2019024441. 1. Рамазанов А.Ш., Балаева Ш.А., Шахбанов К.Ш. Химический состав плодов и масла расторопши пятни- стой, произрастающей на территории Республики Даге- стан // Химия растительного сырья. 2019. №2. С. 113-118. doi:10.14258/jcprm.2019024441. у ( р ) 16. НСАМ № 512-МС. Определение элементного состава образцов растительного происхождения (травы, листья) атомно-эмиссионным и масс-спектральным ме- тодами анализа – Москва. – URL: 16. НСАМ № 512-МС. Определение элементного состава образцов растительного происхождения (травы, листья) атомно-эмиссионным и масс-спектральным ме- тодами анализа – Москва. – URL: https://docs.cntd.ru/document/437145203 (дата обращения: 04.05.2022). листья) атомно-эмиссионным и масс-спектральным ме- тодами анализа – Москва. – URL: https://docs.cntd.ru/document/437145203 (дата обращения: 04.05.2022). j p 2. Смирнов С.О., Фазуллина О.Ф. Плоды расто- ропши пятнистой как перспективное сырье растительного происхождения в технологии производства биологически активных добавок к пище // Пищевая промышленность. 2018. №9. С. 8-12. https://docs.cntd.ru/document/437145203 (дата обращения: 04.05.2022). https://docs.cntd.ru/document/437145203 (дата обращения: 04.05.2022). 17. Болгова М.А., Клейменова Н.Л., Болгова И.Н., Копылов М.В. Исследование питательных веществ ко- ричневых и белых семян льна // Ползуновский вестник. 2021. №3. С. 13-20. doi: 10.25712/ASTU.2072- 8921.2021.03.002. 3. Milk thistle (Silybum marianum): A concise overview on its chemistry, pharmacological, and nutraceutical uses in liver diseases / L. Abenavoli, et al. // Phytotherapy Research. 2018. vol. 32. P. 2202-2213. doi:10.1002 / ptr.6171. 18. Пилипович А.А. Применение витаминов группы В в практике врача-невролога // Consilium Medicum. 2020. № 9. С 82-86. doi: 10.26442/20751753.2020.9.200438. 18. Пилипович А.А. Применение витаминов группы В в практике врача-невролога // Consilium Medicum. 2020. № 9. С 82-86. doi: 10.26442/20751753.2020.9.200438. 4. Кароматов И.Д., Асланова Д.К. Противоопухоле- вые свойства расторопши пятнистой // Биология и инте- гративная медицина. 2018. №10. С 56-63. 5. Nukabadi F.A., Hojjatoleslamy M., Abbasi H. Opti- mization of fortified sponge cake by nettle leaves and milk thistle seed powder using mixture design approach // Food Science & Nutrition. 2017. vol. 9. P. 757-771. doi:10.1007 / s10068-010-0087- х. А. КАШОЛКИНА, М. А. БОЛГОВА, Н. Л. КЛЕЙМЕНОВА, М. В. КОПЫЛОВ, И. Н. БОЛГОВА thistle seeds. Scientific Bulletin. Series F. Biotechnologies. vol. XXI. 2017. P. 165-169. thistle seeds. Scientific Bulletin. Series F. Biotechnologies. vol. XXI. 2017. P. 165-169. никах в том, что они представляют собой ком- плексные данные о витаминном составе и о содержании микро- и макроэлементов вместе. Полученные комплексные результаты необходимы при проектировании состава но- вых многоцелевых пищевых продуктов функ- ционального назначения не только с учетом жирнокислотного состава, но и с учетом ре- комендаций суточного потребления витами- нов и минералов. Эти же сведения можно использовать в косметическом и фармацев- тическом направлениях. никах в том, что они представляют собой ком- плексные данные о витаминном составе и о содержании микро- и макроэлементов вместе. 10. ГОСТ Р 58040-2017 Комплексы витаминно- 10. ГОСТ Р 58040-2017 Комплексы витаминно- минеральные: введ. 2018-09-01. Доступ из справ. – пра- вовой системы «Консультант плюс» (дата обращения: 04.05.2022). Полученные комплексные результаты необходимы при проектировании состава но- вых многоцелевых пищевых продуктов функ- ционального назначения не только с учетом жирнокислотного состава, но и с учетом ре- комендаций суточного потребления витами- нов и минералов. Эти же сведения можно использовать в косметическом и фармацев- тическом направлениях. 11. ГОСТ 31483-2012 Премиксы. Определение со- держания витаминов: B1 (тиаминхлорида), B2 (рибофла- вина), B3 (пантотеновой кислоты), B5 (никотиновой кис- лоты и никотинамида), B6 (пиридоксина), Bс (фолиевой кислоты), C (аскорбиновой кислоты) методом капилляр- ного электрофореза: введ. 2013-07-01. Доступ из справ. – правовой системы «Консультант плюс» (дата обращения: 04.05.2022). ) 12. ГОСТ 32042-2012 Премиксы. Методы опреде- ления витаминов группы В: введ. 2014-07-01. Доступ из справ. – правовой системы «Консультант плюс» (дата обращения: 04.05.2022). ЗАКЛЮЧЕНИЕ Принимая во внимание, что потребители все больше и больше становятся осведомле- ны о качестве продуктов питания, особенно с точки зрения питания, необходимо найти но- вые пищевые ресурсы, богатые биологически активными соединениями. В этом отношении семена расторопши отвечают ожиданиям по- требителей для получения продуктов пита- ния, богатых ценными биосоединениями. р ) 13. ГОСТ EN 12823-2-2014 Продукты пищевые. Определение содержания витамина А методом высоко- эффективной жидкостной хроматографии. Часть 2. Из- мерение содержания бета-каротина: введ. 2017-07-01. Доступ из справ. – правовой системы «Консультант плюс» (дата обращения: 04.05.2022). ( р ) 14. ГОСТ EN 12822-2014 Продукты пищевые. Определение содержания витамина Е (альфа-, бета-, гамма- и дельта-токоферолов) методом высокоэффек- тивной жидкостной хроматографии (Переиздание): введ. 2016-01-01. Доступ из справ. – правовой системы «Кон- сультант плюс» (дата обращения: 04.05.2022). Информация об авторах Д. А. Кашолкина  студент кафедры УК и ТВБ ФГБОУ ВО «ВГУИТ». Д. А. Кашолкина  студент кафедры УК и ТВБ ФГБОУ ВО «ВГУИТ». Д. А. Кашолкина  студент кафедры УК и ТВБ ФГБОУ ВО «ВГУИТ». М. А. Болгова  студент кафедры ТЖ, ПАХПП ФГБОУ ВО «ВГУИТ». Н. Л. Клейменова  кандидат технических наук, доцент кафедры УК и ТВБ ФГБОУ ВО «ВГУИТ». М. В. Копылов  кандидат технических наук, доцент кафедры ТЖ, ПАХПП ФГБОУ ВО «ВГУИТ». И. Н. Болгова  кандидат технических наук, доцент кафедры ТЖ, ПАХПП ФГБОУ ВО «ВГУИТ». 6. Копылов М.В., Болгова И.Н., Клейменова Н.Л., Терёхина А.В., Желтоухова Е. Ю. Разработка ресурсо- сберегающей технологии комплексной переработки мас- личных культур на сырьевые компоненты // Ползуновский вестник. 2019. № 2. С. 7-11. doi: 10.25712/ASTU.2072- 8921.2019.02.002. ФГБОУ ВО «ВГУИТ». Н. Л. Клейменова  кандидат технических наук, доцент кафедры УК и ТВБ ФГБОУ ВО «ВГУИТ». ФГБОУ ВО «ВГУИТ». Н. Л. Клейменова  кандидат технических наук, доцент кафедры УК и ТВБ ФГБОУ ВО «ВГУИТ». М. В. Копылов  кандидат технических наук, доцент кафедры ТЖ, ПАХПП ФГБОУ ВО «ВГУИТ». М. В. Копылов  кандидат технических наук, доцент кафедры ТЖ, ПАХПП ФГБОУ ВО «ВГУИТ». 7. Клейменова Н.Л. Жирнокислотный состав масла семян расторопши пятнистой, полученного методом хо- лодного прессования // Вестник воронежского государ- ственного университета инженерных технологий. 2020. Т. 82. № 4 (86). С. 102-106. doi:10.20914/2310-1202-2020-4- 102-106. ц ф р , И. Н. Болгова  кандидат технических наук, доцент кафедры ТЖ, ПАХПП ФГБОУ ВО «ВГУИТ». И. Н. Болгова  кандидат технических наук, доцент кафедры ТЖ, ПАХПП ФГБОУ ВО «ВГУИТ». ОБСУЖДЕНИЕ POLZUNOVSKIY VESTNIK № 3 2022 Рисунок 3 – Содержание макроэлементов в семенах расторопши пятнистой Figure 3 –The content of macronutrients in the seeds of milk thistle Семена характеризуются большим со- держанием микро- и макроэлементов. Сравнительный анализ полученных ре- зультатов проводили с не подвергавшимися переработке семенами. В ходе сравнения выявлено, что в исследуемых семенах со- держание микроэлементов больше: железа – в 1,8 раз, цинка – в 9,5 раз, меди – незначительно, а содержание макроэлемен- тов: кальция – в 1,8 раз, магния – практически такое же [9]. Имеются небольшие отличия и в качественном составе. Различия в определенной степени объ- ясняются различиями районов произрастания расторопши пятнистой, климатическими условиями и в составе почвы. Преимущества полученных сведений пе- ред представленными в литературных источ- 163 REFERENCES 8. Sadowska K., Jadwiga A., Woropaj-Janczak M. Ef- fect of weather and agrotechnical conditions on the content of nutrients in the fruits of milk thistle (Silybum marianum L. Gaertn.) // Hortorum Cultus. vol. 10(3). 2011. P. 197-207. 1. Ramazanov, A.Sh., Balaeva, Sh.A.  Shax- banov, K.Sh. (2019). Chemical composition of fruit and oil silybum marianum, growing in the territory of the republic of Dagestan. Chemistry of plant raw materials, (2), 113- 118. (In Russ.). DOI:10.14258/jcprm.2019024441. 1. Ramazanov, A.Sh., Balaeva, Sh.A.  Shax- banov, K.Sh. (2019). Chemical composition of fruit and oil silybum marianum, growing in the territory of the republic of Dagestan. Chemistry of plant raw materials, (2), 113- 118. (In Russ.). DOI:10.14258/jcprm.2019024441. ) ( ) 9. Apostol L., Iorga C.S., Moşoiu C.E., Mustățea G., & Cucu Ș.E. Nutrient composition of partially defatted milk 164 ПОЛЗУНОВСКИЙ ВЕСТНИК № 3 2022 ИССЛЕДОВАНИЕ ВИТАМИННОГО СОСТАВА РАСТОРОПШИ ПЯТНИСТОЙ from 1 July 2017. Moscow: Standarts Publishing House. (In Russ.). 2. Smirnov, S.O.  Fazullina, O.F. (2018). Holy thistle seeds as a promising vegetable raw material in the technology of production of biologically active food additive. Food industry, (9), 8-12. (In Russ.). 14. Foodstuffs. Determination of vitamin E (α-, β- , γ- и δ-tocopherols) content by high performance liquid chromatography: (2014). HOST EN 12822-2014. from 1 January 2016. Moscow: Standarts Publishing House. (In Russ.). 3. Abenavoli, L. Izza, A.A. Miliс, N., Cicala, C. Santini, A. Capasso, R. (2018). Milk thistle (Si- lybum marianum): A concise overview on its chemistry, pharmacological, and nutraceutical uses in liver diseases. Phytotherapy Research. 32. 2202-2213. DOI:10.1002 / ptr.6171. ) 15. Premixes. Methods for determination of man- ganese, copper, iron, zinc, cobalt: (2014). HOST 26573.2-2014. from 1 January 2016. Moscow: Standarts Publishing House. (In Russ.). 4. Karomatov, I.D. Aslanova, D.K. (2018). Antineoplastic properties of the milk thistle (Silybum marianum l. gaernt.) Biology and Integrative Medicine, (10), 56-63. (In Russ.). 16. Opredelenie elementnogo sostava obrazcov rastitel'nogo proiskhozhdeniya (travy, list'ya) atomno- emissionnym i mass-spektral'nym metodami analiza (2011). (NSAM No. 512-MS). Retrieved from – https://docs.cntd.ru/document/437145203. 5. Nukabadi, F.A., Hojjatoleslamy, M. Abbasi H.r (2017). Optimization of fortified sponge cake by nettle leaves and milk thistle seed powder using mixture design approach Food Science & Nutrition, (9), 757-771. DOI:10.1007 / s10068-010-0087- х. 17. Bolgova, M.A., Klejmenova, N.L., Bolgova, I.N.  Kopylov M.V. (2021). Issledovanie pitatel'nyh vesh- chestv korichnevyh i belyh semyan l'na Polzunovskij vestnik, (3), 13-20. (In Russ.). Doi: 10.25712/ASTU.2072- 8921.2021.03.002. 6. Kopylov, M.V.,  Bolgova, I.N.,Klejmenova, N.L., Teryohina, A.V. ZHeltouhova, E.Yu. (2019). Raz- rabotka resursosberegayushchej tekhnologii kompleksnoj pererabotki maslichnyh kul'tur na syr'evye komponenty Polzunovskij vestnik, (2), 7-11. (In Russ.). DOI: 10.25712/ASTU.2072-8921.2019.02.002. 18. Pilipovich, A.A. (2020). The use of B-vitamin group in the practice of a neurologist Consilium Medicum, (9), 82-86. (In Russ.). DOI: ( ), ( ) 10.26442/20751753.2020.9.200438. 7. Klejmenova, N.L. (2020). ZHirnokislotnyj sostav masla semyan rastoropshi pyatnistoj, poluchennogo metodom holodnogo pressovaniya // Vestnik voronezh- skogo gosudarstvennogo universiteta inzhenernyh tekhnologij, 4 (86), 102-106. (In Russ.). DOI.org/10.20914/2310-1202-2020-4-102-106. Information about the authors D. A. Kasholkina  Student of the Depart- ment of «Quality management and technology of aquatic bioresources department» of the FSBEI HE "VSUET". 8. Sadowska, K., Jadwiga, A.  Woropaj-Janczak, M. (2011). Effect of weather and agrotechnical conditions on the content of nutrients in the fruits of milk thistle (Si- lybum marianum L. Gaertn.) Hortorum Cultus, 10(3), 197- 207. M. A. Bolgova  Student of the Department of «Fat, processes and devices of chemical and food industries department» of the FSBEI HE "VSUET". 9. Apostol, L., Iorga, C.S., Moşoiu, C.E., Mustățea, G. & Cucu, Ș.E. (2017). Nutrient composition of partially defatted milk thistle seeds. Scientific Bulletin. Series F. Biotechnologies, (XXI), 165-169. N. L. Kleymenova  Candidate of Technical Sciences, Associate Professor of the Depart- ment of «Quality management and technology of aquatic bioresources department» of the FSBEI HE "VSUET". g ( ) 10. Vitamin-mineral complexes: (2017). HOST R 58040-2017. from 1 September 2018. Moscow: Standarts Publishing House. (In Russ.). 11. Premixes. Determination of vitamins: B1 (thia- minchloride), B2 (riboflavin), B3 (pantothenic acid), B5 (nicotinic acid and nicotinamide), B6 (pyridoxine), Bс (folic acid), C (ascorbic acid) content by method of capillary electrophoresis: (2012). HOST 31483-2012. from 1 July 2013. Moscow: Standarts Publishing House. (In Russ.). M. V. Kopylov  Candidate of Technical Sciences, Associate Professor of the Depart- ment of « Fat, processes and devices of chemi- cal and food industries department » of the FSBEI HE "VSUET". g ( ) 12. Premixes. Methods for determination of vitamin B complex: (2012). HOST 32042-2012. from 1 July 2014. Moscow: Standarts Publishing House. (In Russ.). I. N. Bolgova  Candidate of Technical Sci- ences, Associate Professor of the Department of « Fat, processes and devices of chemical and food industries department » of the FSBEI HE "VSUET". g ( ) 13. Foodstuffs. Determination of vitamin A by high performance liquid chromatography. Part 2. Measure- ment of β-carotene: (2014). HOST EN 12823-2-2014. Авторы заявляют об отсутствии конфликта интересов. Статья поступила в редакцию 14.06.2022; одобрена после рецензирования 25.07.2022; принята к б 15 08 2022 Статья поступила в редакцию 14.06.2022; одобрена после рецензирования 25.07.2022; принята к публикации 15.08.2022. у The article was received by the editorial board on 14 June 2022 approved after editing on 25 July 2022; ac- cepted for publication on 15 Aug 2022. 165 POLZUNOVSKIY VESTNIK № 3 2022
https://openalex.org/W2112992618	https://figshare.com/articles/journal_contribution/Supplementary_Figure_2_from_IKK-_Coordinates_Invasion_and_Metastasis_of_Ovarian_Cancer/22392801/1/files/39838377.pdf	Latin	null	IKK-ϵ Coordinates Invasion and Metastasis of Ovarian Cancer	Cancer research	2,012	cc-by	856	Ovarian cancer patient samples (n=65) B Hsu, et al - Supplemental Figure 2 Ovcar8 IKK-ε IKK-α IKK-β GAPDH none E1 B6 ctrl shRNA construct Ovcar5 E1 B6 ctrl shRNA construct IKK-ε tubulin Ovarian cancer patient samples (n=65) B Hsu, et al - Supplemental Figure 2 Ovcar8 IKK-ε IKK-α IKK-β GAPDH none E1 B6 ctrl shRNA construct Ovcar5 E1 B6 ctrl shRNA construct IKK-ε tubulin Hsu, et al - Supplemental Figure 2 Hsu, et al - Supplemental Figure 2 A CYP1B1 LYPD6B IKBKE SYTL1 FAM129A KYNU NNMT C1S EPSTI1 VCAM1 FN1 INHBA TDO2 UPP1 C15orf48 ASS1 SLPI ALOX5AP Ovarian cancer patient samples (n=65) UPP1 ALOX5AP TDO2 VCAM1 FN1 average NNMT C1S CYP1B1 FAM129A KYNU IKBKE SLPI ASS1 B Ovarian cancer patient samples (n=185) average Hsu, et al - Supplemental Figure 2 C Gene expression Tumor : normal ratio (log2) >0.5 Tumor : normal ratio (log2) >0.5 Copy number Gene Percent of tumors Gene Percent of tumors Ovcar8 IKK-ε IKK-α IKK-β GAPDH none E1 B6 ctrl shRNA construct Ovcar5 E1 B6 ctrl shRNA construct IKK-ε tubulin D Ovarian cancer patient samples (n=65) B su, et al - Supplemental Figure 2 Ovcar8 IKK-ε IKK-α IKK-β GAPDH none E1 B6 ctrl shRNA construct Ovcar5 E1 B6 ctrl shRNA construct IKK-ε tubulin O ( ) B ure 2 Ovcar8 IKK-ε IKK-α IKK-β GAPDH none E1 B6 ctrl shRNA construct struct B Ovcar5 E1 B6 ctrl shRNA construct IKK-ε tubulin CYP1B1 LYPD6B IKBKE SYTL1 FAM129A KYNU NNMT C1S EPSTI1 VCAM1 FN1 INHBA TDO2 UPP1 C15orf48 ASS1 SLPI ALOX5AP Ovarian cancer patient samples (n=65) UPP1 ALOX5AP TDO2 VCAM1 FN1 average NNMT C1S CYP1B1 FAM129A KYNU IKBKE SLPI ASS1 Ovarian cancer patient samples (n=185) average Gene expression Copy number IKK-ε IKK-α IKK-β GAPDH none E1 B6 ctrl shRNA construct E1 B6 ctrl shRNA construct IKK-ε tubulin CYP1B1 LYPD6B IKBKE SYTL1 FAM129A KYNU NNMT C1S EPSTI1 VCAM1 FN1 INHBA TDO2 UPP1 C15orf48 ASS1 SLPI ALOX5AP Ovarian cancer patient samples (n=65) UPP1 ALOX5AP TDO2 VCAM1 FN1 average NNMT C1S CYP1B1 FAM129A KYNU IKBKE SLPI ASS1 Ovarian cancer patient samples (n=185) average Gene expression Tumor : normal ratio (log2) >0.5 Tumor : normal ratio (log2) >0.5 Copy number UPP1 ALOX5AP TDO2 EPSTI1 FN1 NNMT C1S CYP1B1 C15ORF48 UPP1 IKBKE SLPI ASS1 30 47 59 71 51 57 52 80 50 21 10 16 13 6 6 21 6 FAM129A IKBKE LYPD6B NNMT Gene Percent of tumors Gene Percent of tumors control IKKε-B2 IKKε-B6 shRNA construct 1.0 0.6 0 0.4 0.8 ative ene ession old RNA ntrol) CYP1B1 EPSTI1 FN1 NNMT1 SLPI INHBA VCAM1 0.2 Ovcar5 F 1.2 0.8 0 0.4 1.0 Relative gene expression (fold shRNA control) CYP1B1 EPSTI1 FN1 NNMT1 SLPI 0.2 Ovcar8 0.6 Ovarian cancer patient samples (n=65) CYP1B1 LYPD6B IKBKE SYTL1 FAM129A KYNU NNMT C1S EPSTI1 VCAM1 FN1 INHBA TDO2 UPP1 C15orf48 ASS1 SLPI ALOX5AP average KYNU NNMT C1S EPSTI1 VCAM1 FN1 INHBA TDO2 UPP1 C15orf48 ASS1 SLPI ALOX5AP UPP1 ALOX5AP TDO2 VCAM1 FN1 average NNMT C1S CYP1B1 FAM129A KYNU IKBKE SLPI ASS1 Ovarian cancer patient samples (n=185) average Gene expression Copy number C15orf48 ASS1 SLPI ALOX5AP UPP1 ALOX5AP TDO2 VCAM1 FN1 average NNMT C1S CYP1B1 FAM129A KYNU IKBKE SLPI ASS1 Ovarian cancer patient samples (n=185) average Gene expression Tumor : normal ratio (log2) >0.5 Tumor : normal ratio (log2) >0.5 Copy number UPP1 ALOX5AP TDO2 EPSTI1 FN1 NNMT C1S CYP1B1 C15ORF48 UPP1 IKBKE SLPI ASS1 30 47 59 71 51 57 52 80 50 21 10 16 13 6 6 21 6 FAM129A IKBKE LYPD6B NNMT Gene Percent of tumors Gene Percent of tumors control IKKε-B2 IKKε-B6 shRNA construct 1.0 0.6 0 0.4 0.8 Relative gene xpression (fold shRNA control) CYP1B1 EPSTI1 FN1 NNMT1 SLPI INHBA VCAM1 0.2 Ovcar5 D F contr IKKε IKKε shRN constr 1.2 0.8 0 0.4 1.0 Relative gene expression (fold shRNA control) CYP1B1 EPSTI1 FN1 NNMT1 SLPI INHBA VCAM1 0.2 Ovcar8 0.6 UPP1 ALOX5AP TDO2 VCAM1 FN1 average NNMT C1S CYP1B1 FAM129A KYNU IKBKE SLPI ASS1 Ovarian cancer patient samples (n=185) Gene expression Tumor : normal ratio (log2) >0.5 Tumor : normal ratio (log2) >0.5 Copy number UPP1 ALOX5AP TDO2 EPSTI1 FN1 NNMT C1S CYP1B1 C15ORF48 UPP1 IKBKE SLPI ASS1 30 47 59 71 51 57 52 80 50 21 10 16 13 6 6 21 6 FAM129A IKBKE LYPD6B NNMT Gene Percent of tumors Gene Percent of tumors Gene expression Tumor : normal ratio (log2) >0.5 UPP1 ALOX5AP TDO2 EPSTI1 FN1 NNMT C15ORF48 IKBKE ASS1 30 47 59 71 51 57 52 80 50 Gene Percent of tumors D control IKKε-B2 IKKε-B6 shRNA construct 1.0 0.6 0 0.4 0.8 Relative gene expression (fold shRNA control) CYP1B1 EPSTI1 FN1 NNMT1 SLPI INHBA VCAM1 0.2 Ovcar5 E F control IKKε-B6.1 IKKε-B6.2 shRNA construct 1.2 0.8 0 0.4 1.0 Relative gene expression (fold shRNA control) CYP1B1 EPSTI1 FN1 NNMT1 SLPI INHBA VCAM1 0.2 Ovcar8 0.6 F control IKKε-B6.1 IKKε-B6.2 shRNA construct 1.2 0.8 0 0.4 1.0 Relative gene expression (fold shRNA control) CYP1B1 EPSTI1 FN1 NNMT1 SLPI INHBA VCAM1 0.2 Ovcar8 0.6 control IKKε-B2 IKKε-B6 shRNA construct 1.0 0.6 0 0.4 0.8 Relative gene expression (fold shRNA control) CYP1B1 EPSTI1 FN1 NNMT1 SLPI INHBA VCAM1 0.2 Ovcar5 E F E
W4386177321.txt	https://www.scielo.br/j/rcaat/a/yyFFnk6Wj9wrkrqkhq8wspM/?lang=en&format=pdf	en		Grain yield performance of corn in different plant arrangements	Revista Caatinga	2,023	cc-by	6,169	ISSN 0100-316X (impresso) ISSN 1983-2125 (online) http://dx.doi.org/10.1590/1983-21252023v36n306rc Universidade Federal Rural do Semi-Árido Pró-Reitoria de Pesquisa e Pós-Graduação https://periodicos.ufersa.edu.br/index.php/caatinga Grain yield performance of corn in different plant arrangements Desempenho produtivo do milho em diferentes arranjos de plantas Paulo H. Cazarim1* , Gabriel D. Shimizu1 , Lucas H. Fantin2 , Marcelo A. de A. Silva1 , Claudemir Zucareli1 1 Graduate Program in Agronomy, Universidade Estadual de Londrina, Londrina, PR, Brazil. 2Research and Development, Fundação Chapadão, Chapadão do Sul, MS, Brazil. ABSTRACT - Sowing arrangements composed of double-row spacing in corn can favor the interception of solar radiation by the canopy and, consequently, the yield performance of the crop. However, it is possible that the microclimate provided by this spacing, especially at high plant densities, favors the occurrence of leaf diseases. Thus, the objective was to evaluate the effect of 0.45 m and double-row spacing arrangements on the severity of foliar diseases and yield performance of corn grown in the first and second -crop seasons. Two independent experiments were conducted (with and without the fungicide fluxapyroxad + pyraclostrobin) in the first and second-crop seasons in a randomized block design arranged in a split-plot scheme with four repetitions. The plots consisted of spacing (0.45 m) and double-row (0.30 × 0.60 m), and the subplots, four plant densities (59,200, 74,000, 81,400, and 96,200 plants ha -1). In the plant density factor, in the second-crop season, there is a decrease in the severity of white spot as plant density is increased. Also, for the plant density factor, in the first-crop season, there may be a significant yield increase as the plant density is increased. RESUMO - Arranjos de plantas compostos pelo espaçamento fileira dupla no milho podem favorecer a interceptação da radiação solar pelo dossel e, consequentemente, o desempenho produtivo da lavoura. Entretanto, é possível que o microclima proporcionado por este espaçamento, sobretudo em altas densidades de plantas, favoreça a ocorrência de doenças foliares. Assim, objetivou-se avaliar o efeito de arranjos compostos pelos espaçamentos de 0,45 m e fileira dupla sobre a severidade de doenças foliares e o desempenho produtivo do milho cultivado em 1a e 2a safra. Foram conduzidos dois experimentos independentes (com e sem o fungicida fluxapiroxade + piraclostrobina) na 1a e 2a safra, em delineamento em blocos ao acaso com parcelas subdivididas, com quatro repetições. As parcelas consistiram nos espaçamentos (0,45 m) e fileira dupla (0,30 × 0,60 m), e as subparcelas, quatro densidades de plantas (59.200, 74.000, 81.400 e 96.200 plantas ha-1). No fator densidade de plantas, na 2a safra, há decréscimo na severidade da mancha branca conforme adensamento de plantas. Ainda, também para o fator densidade, na 1 a safra, pode haver incremento significativo na produtividade em função do adensamento de plantas. Keywords: Zea mays L. Inter-row spacing. Plant density. Leaf disease severity. Palavras-chave: Zea mays L. Espaçamento entrelinhas. Densidade de plantas. Severidade de doenças foliares. Conflict of interest: The authors declare no conflict of interest related to the publication of this manuscript. This work is licensed under a Creative Commons Attribution-CC-BY https://creativecommons.org/ licenses/by/4.0/ Received for publication in: February 26, 2022. Accepted in: May, 2, 2023. Corresponding author: <paulo_cazarim@hotmail.com> INTRODUCTION The adjustment of plant arrangement in corn cultivation considers variations in plant density and inter-row spacing (GALVÃO et al., 2014), and also spatial and temporal distribution of plants in the row (NOVAK; RANSOM, 2018), being a technique employed in the search for better yield performance, since it allows to allocate the maximum amount of plants per area, increasing the interception of photosynthetically active radiation, one of the determinants of grain yield (LU et al., 2017). Arrangements that employ 0.45 m spacing have been common in corn cultivation, but alternative arrangements are still being sought, such as those composed of double-row spacing. In the double-row configuration, two rows are spaced closer together and are alternated by wider spacing, which can provide greater interception of solar radiation (NOVACEK et al., 2013), especially in second-crop season conditions where light intensity is reduced and especially in higher plant density, since greater competition for the light within the canopy is seen (YANG et al., 2019). The increase in plant density in arrangements employing double-row spacing may trigger changes in air circulation, temperature, and relative humidity, conditioning a favorable microclimate for developing pathogens that cause leaf diseases, which would cancel the positive effect of a particular arrangement. However, this occurrence can be mitigated by applying fungicides (BRITO et al., 2013). Studies such as those by Silva et al. (2012) and Kappes, Andrade, and Arf (2013) evaluated the development of leaf diseases in different plant arrangements, however, none focused on double-row spacing. Thus, the present study aimed to evaluate the effect of 0.45 m and double-row spacing arrangements on the Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 532 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. severity of foliar diseases and the yield performance of corn grown in the first and second-crop seasons. MATERIAL AND METHODS Two independent experiments were conducted in the first and two in the second-crop seasons (one with and one without fungicide application in each crop season). Two were conducted in the first-crop season of 2019/2020 on Nitossolo Vermelho, at 23°20'27” S and 51°12'48” W, with an altitude of 572 m. The other two experiments were conducted in the second-crop season of 2020, on Latossolo Vermelho, at 23°40'04" S, 51°10'03” W, with an altitude of 650 m. The meteorological data of the sites are shown in Figures 1A and 1B. For the first-crop season, the values of accumulated precipitation and average temperatures (maximum, average, and minimum) were: 702 mm, 29.9°C, 24.9°C, and 19.9°C, respectively. For the second-crop season, 545 mm, 27.3°C, 20.9°C, and 14.5°C, respectively. 1 2 Figure 1. Daily precipitation and average, maximum, and minimum temperatures for the first (A) and second (B) crop seasons. S: sowing; V5, V7, V8: five, seven, and eight fully developed leaves, respectively; VT: tasselling; 1SA, 2SA, 3SA, 4SA, 5SA, 6SA, 7SA: first, second, third, fourth, fifth, sixth, and seventh leaf disease severity assessment, respectively; H: harvest. The chemical characteristics of the soil at a 0-20 cm soil depth determined before the installation of the experiments were: first-crop season - pH(CaCl2) = 5.25; H+Al = 4.75 cmolc dm-3; Al+3 = 0 cmolc dm-3; Ca+2 = 3.77 cmolc dm-3; Mg+2 = 1.30 cmolc dm-3; K+ = 0.65 cmolc dm-3; P = 18.66 mg dm-3; CEC(pH7.0) = 10.47 cmolcdm-3; CEC(effective) = 5.72 cmolc dm-3; OM = 3.20%; Second-crop season - pH(CaCl2) = 4.90; H+Al = 5.20 cmolc dm-3; Al+3 = 0 cmolc dm-3; Ca+2 = 8.13 cmolc dm-3; Mg+2 = 2.66 cmolc dm-3; K+ = 0.71 cmolc dm-3; P = 7.49 mg dm-3; CEC(pH7. 0) = 16.72 cmolcdm-3; CEC(effective) = 11.53 cmolc dm-3; OM = 3.21%. The experimental design was a randomized block design arranged in the split-plots scheme with four repetitions. The plots consisted of two inter-row spacings (0.45 m and double-row 0.30 × 0.60 m), and the subplots consisted of four plant densities (59,200, 74,000, 81,400, and 96,200 plants ha-1). The subplots comprised six 5 m long rows, with the evaluations occurring in the two central rows, discounting 1.0 m from the ends, totaling 2.7 m2. The corn hybrids used in the first and second-crop seasons were P3380H and LG36600 AgrisureViptera 3, respectively. The sowings of the first and second-crop seasons were performed on 10/26/2019 and 02/08/2020, respectively, in an area previously occupied by wheat. For the 0.45 m spacing, sowing was performed with the seeder in its original configuration. For double-row spacing (0.30 × 0.60 m), sowing was performed with a seeder at a spacing of 0.90 m, and then, with the help of a reference marker (wooden ruler of five meters), positioned 0.30 m parallel to each furrow, the other furrows were opened manually. The same reference marker was developed as a template so that the seeds were evenly distributed, also manually. The sowings were all performed based on the highest plant density, with the number of seeds already corrected previously according to the Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 533 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. germination test. The lowest densities were obtained by thinning at the growth stage of five fully developed leaves (V5), when the initial plant stand was recorded. The base fertilization used was adjusted according to soil analysis and following technical recommendations (PAULETTI; MOTTA, 2017), which consisted of 436 kg ha -1 of the NPK formulation 04-14-08 for the first-crop season, and 451 kg ha-1 of the same NPK formulation for the secondcrop season. In the first-crop season, the topdressing fertilization was performed twice, at the V5 growth stage, after thinning, and at V7 growth stage, applying 90 kg of N ha-1 at each stage, totaling 180 kg of N ha -1. For the second-crop season, it was applied at the V5 growth stage, after thinning, in the amount of 130 kg of N ha -1. The source used in both experiments was ammonium sulfate (21% N). The management of pests and weeds was carried out according to observations made during crop development. At the V8 growth stage, the commercial fungicide Orkestra ® SC was applied in the control experiment, composed of a mixture of the active ingredients fluxapyroxad (167 g L -1) and pyraclostrobin (333 g L-1) at a dose of 300 mL of the commercial product per hectare. A 20 L capacity electric backpack sprayer was used, with its flow rate adjusted to 166 L ha-1, equipped with a four-tip spray bar. Immediately after silking growth stage, phytometric evaluations were performed, including leaf area index (LAI), stem diameter (SD), ear insertion height (EIH), and plant height (PH), the last three of which were taken from the same plant. Following this, assessments of leaf disease severity were initiated. To estimate the leaf area index (LAI), the leaf area of ten plants (five random plants in each central row) of each subplot useful area was estimated. For this, the length (L) from the base to the tip of the leaf and the greatest width (W) of all photosynthetically active leaves were measured. The leaf area (A), expressed in cm2, was estimated using the expression: A = L × W × 0.75. The sum of the areas of all the plant leaves was used to determine the leaf area per plant. The leaf area index corresponded to the leaf area per plant divided by the soil surface occupied by it at each inter-row spacing combination (SANGOI et al., 2019). Stem diameter (SD) was determined by measuring in millimeters the largest and smallest diameter in the central part of the second or first elongated internode using a digital caliper, and then calculating the average of the measures. The ear insertion height (EIH) and plant height (PH) was determined by measuring the distance in centimeters between the soil surface and the top ear insertion node and between the soil surface and the base of the tassel, respectively. The measurements were performed on ten plants (five random plants in each central row) in each subplot useful area. Assessments of leaf disease severity, being naturally occurring diseases, were carried out weekly, for seven weeks for the first-crop season, and four weeks for the second-crop season, respectively. It was considered for assessment, the ear leaf of four plants, randomly taken in each assessment (MOTERLE; SANTOS, 2019) in each subplot useful area, two in each of the two central rows. The following diseases were evaluated: turcicum leaf blight (Exserohilum turcicum), gray leaf spot (Cercospora zeae-maydis), white spot (Pantoea ananatis and/or Phaeosphaeria maydis), and southern rust (Puccinia polysora). Severity scores were taken according to the following diagrammatic scales: turcicum blight (VIEIRA et al., 2014), gray leaf spot, white spot, and southern rust (CAPUCHO et al., 2010). From the severity scores obtained in the plant evaluations, the evolution of each disease was determined by calculating the area under the disease progress curve (AUDPC) according to the formula below proposed by Shaner and Finney (1977): 𝑛 AUDPC = (Yi + Yi+1 2) (ti+1 -ti ) 𝑖=1 where: [Yi and Yi+1] - severity values observed in two consecutive assessments; [ti+1 and ti] - interval between two assessments; [n] - total number of assessments. For the second-crop season, it was necessary to quantify the incidence of the stunt disease complex, according to Costa et al., (2019), performed 89 days after sowing, and the percentage of lodging at harvest, right after quantifying the final plant stand (Table 1). Plants with less than 45° of inclination about the ground were considered lodged, and then the percentage was calculated. Table 1. Desired (DS), initial (IS), and final (FS) plant stand in the first and second-crop seasons. Crop seasons DS -1 plants ha 59,200 74,000 81,400 96,200 First-crop season IS FS -1 -----------plants ha ----------59,200 74,000 81,200 94,600 59,000 73,600 80,700 93,200 Second-crop season FS/DS (%) 99 99 99 96 IS FS -1 -----------plants ha ----------59,200 73,800 81,200 93,900 58,100 73,300 80,000 92,000 FS/DS (%) 98 99 98 95 1 Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 534 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. Subsequently, the ears were harvested manually, when the average moisture content was close to 140 and 180 g of water per kilogram of grain, for the first and second-crop seasons, respectively. The water content in the grains was obtained by a Gehaka digital capacitance meter (G600). The harvested ears were quantified, and the number of ears per plant was determined with this value. Productive ears were considered as those that presented more than 15 formed grains. From the harvested ears, ten were randomly separated for the determination of the diameter (ED) and length (EL) of ears; 1000-grain weight (1000W) expressed in grams, determined from eight repetitions of one hundred grains; and yield (YLD) expressed in kg ha -1, determined from the mass of grains of all ears harvested in the useful area of the subplots, which were threshed manually. The last two had their moisture content corrected to 130 g of water per kilogram of grain. Besides the measurements mentioned above, the mass of grains per ear (MGE) was evaluated, determined by weighing the grains from the ten ears from which the yield components were measured, and averaging them, which was later used to calculate the number of grains per ear; the number of grains per ear (NGE), determined by the ratio between the mass of grains per ear, and the mass of one thousand grains; and number of grains per square meter (NG m-2), determined by multiplying the number of grains per ear by the number of ears per hectare, which resulted from multiplying the number of ears per plant by the final plant stand obtained, dividing the result by ten thousand. The data were tested for the assumptions of normality of errors and homogeneity of variances by the Shapiro-Wilk and Bartlett tests, respectively (p>0.05). Subsequently, they were submitted to the joint analysis of variance considering the mixed model in which block within fungicide and block within density were considered random effects, and the other effects were considered fixed effects. Interactions with the fungicide factor and the isolated fungicide were not considered in the interpretation since there is no valid error term, thus, it was only used for the need to analyze the results together or not. Wald's F test (p<0.05) analyzed the variation sources in the model. If significant, the means were compared by the Tukey test, following the emmeans package procedure (LENTH, 2018). When there was an interaction effect with density or an isolated effect, the linear or quadratic regression model was adjusted (p<0.05). Pearson correlation was performed between the yield components. R software (R CORE TEAM, 2020) was used for all analyses. RESULTS AND DISCUSSION for the first-crop season, and isolated significance was obtained only for the plant density factor. For the second-crop season, there was an interaction between the number of ears per plant with spacing and density; isolated significance between the stem diameter and spacing; and the remaining significances, also isolated, for plant density (Table 2). Stem diameter fitted quadratic equations for both crop seasons, according to the increasing plant density, with a minimum point outside the range studied for the first-crop season and close to 88,100 plants ha -1 for the second-crop season (Figures 2A and 2B). The behavior of reduction in the stem diameter according to the increase in plant density is based on the increase in intra-specific competition for light; physiological changes of hormonal origin, auxin in particular, which does not oxidize due to the greater shading, stimulating cell elongation; and alteration in the perception of photomorphogenic light (red and far-red) by the phytochrome, leading to an increase in the red/far-red ratio, which regulates the partitioning of photoassimilates and morphological adaptation. These factors together cause the vertical growth of the stem to predominate to the detriment of its diameter (SANGOI, 2001). For the second-crop season, stem diameter was greater at double-row spacing (17.66 mm) compared to 0.45 m (16.95 mm), so there was a significant difference at densities of 74,000 and 96,200 plants ha-1 (Figure 3). There was a quadratic adjustment for both spacings according to the increasing plant density, with minimum points near 91,100 plants ha-1 for the 0.45 m spacing and 85,200 plants ha -1 for the double-row spacing. The larger stem diameter at doublerow spacing may be linked to the higher interception of photosynthetically active radiation by the crop canopy around the vegetative stage of eight to nine leaves, as reported by Robles, Ciampitti, and Vyn (2012) and Novacek et al. (2013). Such an increase may have favored the photosynthetic efficiency of the plants, thus directing surplus photoassimilates to stem growth in diameter. The same authors also observed a greater stem diameter with doublerow spacing. For ear insertion height, there was a quadratic adjustment for both crop seasons according to the increasing plant density, with maximum and minimum points outside the range studied for the first and second-crop seasons, respectively (Figures 4A and 4B). The physiological explanations for the reduction in stem diameter go along with those that explain the increase in ear insertion height at higher densities. Courbier and Pierik (2019) point out that under such conditions, the greater intraspecific competition for light leads to the so-called etiolation effect, whereby plants increase their chances of growing above the canopy and avoiding shading, which justifies the greater ear insertion height. There was no interaction between the factors analyzed Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 535 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. Table 2. Summary of analysis of variance (p-value): stem diameter (SD); ear insertion height (EIH); plant height (PH); leaf area index (LAI); area under the disease progress curve: southern rust (AUDPC-S), turcicum leaf blight (AUDPC-T), gray leaf spot (AUDPC-G), white spot (AUDPC-W); stunt incidence (STI); plant lodging (LODG); number of ears per plant (EP); ear length (EL); ear diameter (ED); 1000-grain weight (1000W); number of grains per ear (NGE); number of grains per square meter (NG m -2); and grain yield (YLD). Spacing (S); plant density (D). Crop season Variable First-crop season Second-crop season 1 S D -11* S×D SD 0.72 EIH 0.66 2.6.10-02* 0.22 PH 0.41 0.48 0.54 2.9.10 -16* 0.08 LAI 0.09 AUDPC-S 0.24 0.08 0.90 EP 0.18 2.9.10-04* 0.74 EL 0.49 5.3.10-09* 0.79 -05* 2.10 0.71 ED 1.00 4.2.10 0.55 1000W 0.99 5.4.10-07* 0.28 NGE 0.75 4.8.10-05 * 0.67 NG m-2 0.07 1.4.10-11* 0.08 YLD 0.14 -07* 1.7.10 0.12 SD 3.8.10-02* 3.2.10-05* 0.54 EIH 0.86 5.3.10-03 0.69 PH 0.49 0.05 0.52 LAI AUDPC-T AUDPC-G 0.38 0.94 0.27 <2.10-16* 0.49 0.81 0.08 0.38 0.38 AUDPC-W 0.39 7.3.10-03 0.47 0.56 STI 0.26 0.51 LODG 0.09 0.12 0.38 EP 0.28 2.3.10-09*** 3.5.10-02* EL 0.96 2.2.10-02* 0.66 ED 0.43 0.97 0.79 1000W 0.27 0.68 0.63 NGE 0.92 0.10 0.40 NG m-2 0.82 0.11 0.13 YLD 0.33 0.63 0.94 : significant at 5% probability (p<0.05); : significant at 1% probability (p<0.01); * : significant at 0.1% probability (p<0.001). 1 Figure 2. Stem diameter (SD) according to the plant density: (A) first and (B) second-crop seasons. Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 536 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. 1 Figure 3. Stem diameter (SD) according to 0.45 m and double-row spacings. Figure 4. Ear insertion height (EIH) according to the plant density: (A) first and (B) second-crop seasons. Corroborating with the results obtained for stem diameter and ear insertion height, Fromme, Spivey, and Grichar (2019) working with increasing densities, from 43,000 to 117,000 plants ha-1, also observed the same pattern of response according to plant density increasing. Xue et al. (2017) point out that this growth pattern can have negative implications, as with a high center of gravity and smaller stem diameter, plants become more prone to events such as breakage and lodging. Thus, it is essential to maintain stem health to maintain their resistance under higher plant density conditions, as they become more susceptible to rots. There was a quadratic fit for leaf area according to the increasing plant density in both crop seasons, Figures 5A and 5B, with minimum points outside the range studied. The observed pattern of increase is since, under high plant density, the crop canopy per unit area covers a greater amount of soil. Bernhard and Below (2020), when increasing the density from 94,000 to 139,000 plants ha -1, observed an increase in leaf area index from 5.8 to 7.3. The area under the progress curve of the white spot in the second-crop season (Figure 6) fitted the quadratic equation according to the increase in plant density, with a minimum point outside the range studied. The severity values of the spot were within the range of 0.3 to 1.6% at the end of the fourth evaluation, at 89 days after sowing. Miller et al. (2016) point out that frequent precipitation is among the factors that favor the development of the disease, which did not occur during the period of evaluations, from 69 to 89 days after sowing (Figure 1B), which probably influenced its low severity. Also, considering that the pathogen is disseminated by wind (CARVALHO; PEREIRA; CAMARGO, 2016), it is speculated that the reduced air circulation within the canopy due to the higher plant densities and open leaf architecture pattern of the hybrid in question may have influenced its lower dissemination. Silva et al. (2012) and Kappes, Andrade, and Arf (2013) observed similar behavior when evaluating grey leaf spot and tropical rust, respectively. Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 537 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. Figure 5. Leaf area index (LAI) according to the plant density: (A) first and (B) second-crop seasons. Figure 6. Area under the progress curve of white spot (AUDPC-W) according to the plant density. The number of ears per plant in the first-crop season fitted the quadratic equation, according to the increase in plant density, with a minimum point outside the range studied (Figure 7A). In the second-crop season, the number of ears per plant was greater in double-row spacing (1.45) than in 0.45 m spacing (1.24) at a density of 59,200 plants ha -1 (Figure 7B). Also, in the second-crop season, there was a quadratic adjustment for both spacings according to the increase in plant density, with minimum points near 94,000 and 93,100 plants ha-1, for the double-row and 0.45 m spacings, respectively (Figure 7B). The greater number of ears per plant in the double-row spacing, at the density of 59,200 plants ha-1, can be justified by, besides a possible greater interception of solar radiation, a higher incidence of the corn stunt complex in the same arrangement. The average incidence rates for the 0.45 m and double-row spacings were 36.30 and 41.18%, respectively. There was no statistical difference among the four densities; however, at the density of 59,200 plants ha-1, the incidence values obtained were 36.82 and 49.53% for 0.45 m and double-row spacings, respectively. Despite the absence of significance, it is suspected that this difference may have been the reason for the higher number of ears per plant since among the physiological responses of corn plants attacked by the stunt complex, the production of multiple ears stands out (JONES; MEDINA, 2020). Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 538 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. Figure 7. Number of ears per plant (EP) according to the plant density: (A) first and (B) second-crop seasons. In general, in the absence of interactions, Sangoi (2001) points out that the factors responsible for reduced prolificacy are associated with competition between the ear and other plant organs for photoassimilates and in response to the hormonal balance of the auxin in specific. Large amounts of auxins are directed to the tassel stimulating its cell division, growth, and dry mass accumulation, once its primordium is transformed into a reproductive structure. High light intensity oxidizes and inactivates auxins, breaking apical dominance. Under high plant density conditions, less light falls on the point of tassel growth compared to low densities, and therefore there is less inactivation of auxins, which remain in higher concentration, thus promoting apical dominance and consequently favoring lower ear formation. The ear characteristics at the first-crop season, Figures 8A, 8C, and 8D, fitted quadratic equations according to the increasing plant density. For ear length, Figure 8A, the maximum point was near 61,000 plants ha -1; for ear diameter, Figure 8C, there was a tendency to decrease; for the number of grains per ear, Figure 8D, the maximum point was near 65,700 plants ha-1. In the second-crop season, Figure 8B, there was a quadratic adjustment for ear length, with a minimum point outside the studied range. Ventura and Dalchiavon (2018) also observed an ear length and diameter reduction in response to increased intraspecific competition. Reduction in the number of grains per ear was also observed by Zhang et al. (2018), who point out that such a result may stem from the low formation of floral primordia, reduced pollination due to asynchrony in flowering, and also high grain abortion rate after fertilization. Figure 8. Ear length (EL), ear diameter (ED), and number of grains per ear (NGE) according to the plant density: (A) first, (B) second, (C) first, (D) first-crop seasons. Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 539 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. The 1000-grain weight in the first-crop season, Figure 9A, fitted the quadratic equation according to the increase in plant density, with a minimum point outside the studied range. The lower development of the ears is partially justified by the reduction of ear development and grains of shorter length (TESTA; REYNERI, BLANDINO, 2016). Furthermore, based on the results of other authors, Haegele, Becker, and Below (2014) discuss this same behavior, stating that the shading provided by high plant density promotes a reduced photosynthetic rate, consequently reducing individual grain weight. It also indicates that hybrids with more stable grain weight in response to competitive conditions may be the most suitable for cultivation with higher plant densities. The number of grains per square meter in the first-crop season, Figure 9B, fitted the quadratic equation according to the increase in plant density, with a maximum point outside the studied range. According to Haegele, Becker, and Below (2014), an increase in the number of grains per unit area is expected, since it is influenced by the number of plants per area, and consequently, the number of ears per area. Figure 9. 1000-grain weight (1000W) and number of grains per square meter (NG m -2) according to the plant density: (A) and (B), first-crop season. The grain yield in the first-crop season, Figure 10, fitted the quadratic equation, according to the increase in plant density, with a maximum point near 89,000 plants ha -1, at which 12,609 kg ha-1 was obtained. In a study conducted by Pricinotto et al. (2019) under first-crop conditions, with density ranging from 40,000 to 120,000 plants ha -1, the yield was maximized with 93,400 plants ha -1, an amount close to that of the present study, in which they reached 11,858 kg ha-1. The same authors point out that under ideal edaphoclimatic conditions and appropriate management, it is possible to increase plant density and increase grain yield, thus better exploring the interaction between genotype and environment. Figure 10. Grain yield (YLD) according to the plant density. Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 540 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. Based on the results obtained, it was possible to observe that the increase in intra-specific competition according to the higher plant densities suppressed the grain yield per individual so that the number of ears per plant, the number of grains per ear, and the 1000-grain weight reduced, however, the number of grains per square meter increased. Thus, understanding the relationship between the yield components in grain yield formation in high plant-density corn crops allows more assertive decisions to be made during crop planning. When performing the correlation between grain yield and yield components 1000-grain weight and number of grains per square meter (Table 3), it was possible to observe that there was a high correlation in the first-crop season between the number of grains per square meter and grain yield (0.75), on the other hand, there was a negative correlation between 1000-grain weight and grain yield (-0.02). Such a result indicates that the number of grains per unit area was more important than grain mass for grain yield. The results obtained are in agreement with those of Ruffo et al. (2015), in which the authors point out that effort to reduce yield gaps should continue to focus on maximizing the number of ovules per ear and reducing the grain abortion rate to increase the number of grains per unit area. Table 3. Pearson correlation coefficient between the grain yield components 1000-grain weight (1000W) and number of grains per square meter (NG m-2) with grain yield (YLD). 1 Grain yield components YLD 1000W NG m-2 -0.02 ns 0.75 ns : not significant; **: significant at 1% (p<0.01). To the adopted inter-row spacing, it was possible to observe that the 0.45 m and double-row spacings were similar for diseases and grain yield. Thus, based on the data from this research, this configuration is not a justifiable alternative to be adopted by producers. On the other hand, in the inter-row spacing of 0.45 m, more studies with plant density involving management practices that aim to reduce the physiological stress provided may contribute to a greater base of information. It was observed under first-crop season conditions that the practice of adopting high plant densities is interesting as a strategy for obtaining high yields, however, studies are needed to search for the maximum point for each cultivar in each growing environment. For the second-crop season, increasing the number of plants per area did not prove advantageous, presenting only a numerical increase in grain yield, resulting largely from the smaller quantity and irregular distribution of precipitations. In second-crop conditions, factors other than the increase in plant density could provide better results. yield. Agronomy Journal, 112: 2456-2465, 2020. BRITO, A. H. et al. Controle químico da cercosporiose, mancha-branca e dos grãos ardidos em milho. Revista Ceres, 60: 629-635, 2013. CAPUCHO, A. S. et al. Desenvolvimento de metodologia para avaliação da mancha branca do milho. Sete Lagoas, MG: Embrapa Milho e Sorgo, 2010. 26 p. (Boletim de Pesquisa e Desenvolvimento, 26). CARVALHO, R. V.; PEREIRA, O. A. P.; CAMARGO, L. E. A. Doenças do milho. In: AMORIM. L. et al. (Eds.). Manual de Fitopatologia: doenças de plantas cultivadas. 5. ed. Ouro Fino, MG: Agronômica Ceres, 2016. v. 2, cap. 57, p. 549-560. COSTA, R. V. et al. Incidence of corn stunt disease in offseason corn hybrids in different sowing seasons. Pesquisa Agropecuária Brasileira, 54: 1-9, 2019. COURBIER, S.; PIERIK, R. Canopy light quality modulates stress responses in plants. Iscience, 22: 441-452, 2019. CONCLUSIONS In the plant density factor, in the second-crop season, there is a decrease in the severity of white spot as plant density is increased. Also, for the plant density factor, in the first-crop season, there may be a significant increase in grain yield as plant density is increased. FROMME, D. D.; SPIVEY, T. A.; GRICHAR, W. J. Agronomic response of corn (Zea mays L.) hybrids to plant populations. International Journal of Agronomy, 2019: 1-8, 2019. REFERENCES GALVÃO, J. C. C. et al. Sete décadas de evolução do sistema produtivo da cultura do milho. Revista Ceres, 61: 819-828, 2014. BERNHARD, B. J.; BELOW, F. E. Plant population and row spacing effects on corn: plant growth, phenology and grain HAEGELE, J. W.; BECKER, A. S. H.; BELOW, F. E. Row arrangement, phosphorus fertility, and hybrid contributions to Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 541 GRAIN YIELD PERFORMANCE OF CORN IN DIFFERENT PLANT ARRANGEMENTS P. H. CAZARIM et al. managing increased plant density of maize. Agronomy Journal, 106: 1838-1846, 2014. JONES, T. L.; MEDINA, R. F. Corn stunt disease: an ideal insect-microbial-plant pathosystem for comprehensive studies of vector-borne plant diseases of corn. Plants, 9: 1-16, 2020. KAPPES, C.; ANDRADE, J. A. C.; ARF, O. Efeito dos arranjos espaciais de plantas na sanidade de híbridos de milho. Scientia Agraria Paranaensis, 12: 53-65, 2013. LENTH, R. Emmeans: estimated marginal means, aka leastsquares mean. 2018. Disponível em: <https://cran.rproject.org/web/packages/emmeans/index.html>. Acesso em: 7 jan. 2020. LU, Y. et al. Increasing the planting uniformity improves the yield of summer maize. Agronomy Journal, 109: 1463-1475, 2017. MILLER, A. M. et al. Characterization of the inaA gene and expression of ice nucleation phenotype in Pantoea ananatis isolates from maize white spot disease. Genetics and Molecular Research, 15: 1-8, 2016. MOTERLE, L. M.; SANTOS, R. F. Época de aplicação de fungicida na cultura do milho segunda safra. Colloquium Agrariae, 15: 61-71, 2019. NOVACEK, M. J. et al. Twin rows minimally impact irrigated maize yield, morphology, and lodging. Agronomy Journal, 105: 268-276, 2013. NOVAK, L.; RANSOM, J. Factors impacting corn (Zea mays L.) establishment and the role of uniform establishment on yield. Agricultural Sciences, 9: 1317-1336, 2018. PAULETTI, V.; MOTTA, A. C. V. Manual de adubação e calagem para o estado do Paraná. 1. ed. Curitiba, PR: Núcleo Estadual Paraná da Sociedade Brasileira de Ciência do Solo – NEPAR-SBCS, 2017. 482 p. SANGOI, L. Understanding plant density effects on maize growth and development: an important issue to maximize grain yield. Ciência Rural, 31: 159-168, 2001. SANGOI, L. et al. Estratégias de manejo do arranjo de plantas visando otimizar a produtividade de grãos do milho. Revista Brasileira de Milho e Sorgo, 18: 47-60, 2019. SHANER, G.; FINNEY, R. The effect of nitrogen fertilization on the expression of slow mildewing resistance in Knox Wheat. Journal of Phytopathology, 67: 1051-1056, 1977. SILVA, R. R. et al. Influência da densidade de cultivo de dois genótipos de milho na severidade da mancha de cercospora e no rendimento de grãos na safrinha. Semina: Ciências Agrárias, 33: 1449-1454, 2012. TESTA, G.; REYNERI, A.; BLANDINO, M. Maize grain yield enhancement through high plant density cultivation with different inter-row and intra-row spacings. European Journal of Agronomy, 72: 28-37, 2016. VENTURA, M. F. B.; DALCHIAVON, F. C. Agronomic characteristics of corn grown in different population arrangements. Nativa, 6: 569-574, 2018. VIEIRA, R. A. et al. A new diagrammatic scale for the assessment of northern corn leaf blight. Crop Protection, 56: 55-57, 2014. XUE, J. et al. Research progress on reduced lodging of highyield and density-maize. Journal of Integrative Agriculture, 16: 2717–2725, 2017. YANG, Y. et al. Improving maize grain yield by matching maize growth and solar radiation. Scientific Reports, 9: 1-12, 2019. ZHANG, M. et al. How plant density affects maize spike differentiation, kernel set, and grain yield formation in Northeast China? Journal of Integrative Agriculture, 17: 1745-1757, 2018. PRICINOTTO, L. F. et al. Yield and biometric characteristics of maize submitted to plant population and trinexapac-ethyl doses. Revista Caatinga, 32: 667-678, 2019. R CORE TEAM. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2020. ROBLES, M.; CIAMPITTI, I. A.; VYN, T. J. Responses of maize hybrids to twin row spatial arrangement at multiple plant densities. Agronomy Journal, 104: 1747-1756, 2012. RUFFO, M. L. et al. Evaluating management factor contributions to reduce corn yield gaps. Agronomy Journal, 107: 495-505, 2015. Rev. Caatinga, Mossoró, v. 36, n. 3, p. 532 – 542, jul. – set., 2023 542
https://openalex.org/W4285004863	https://www.frontiersin.org/articles/10.3389/fgene.2022.874667/pdf	English	null	lncRNAs Functioned as ceRNA to Sponge miR-15a-5p Affects the Prognosis of Pancreatic Adenocarcinoma and Correlates With Tumor Immune Infiltration	Frontiers in genetics	2,022	cc-by	9,087	ORIGINAL RESEARCH published: 11 July 2022 doi: 10.3389/fgene.2022.874667 ORIGINAL RESEARCH published: 11 July 2022 doi: 10.3389/fgene.2022.874667 lncRNAs Functioned as ceRNA to Sponge miR-15a-5p Affects the Prognosis of Pancreatic Adenocarcinoma and Correlates With Tumor Immune Inﬁltration Yu Wang 1,2, Zhen Wang 1, KaiQiang Li 1, WeiLing Xiang 1, BinYu Chen 1, LiQin Jin 1,3* and Ke Hao 1* 1Laboratory Medicine Center, Allergy Center, Department of Transfusion Medicine, Zhejiang Provincial People’s Hospital (Afﬁliated People’s Hospital, Hangzhou Medical College), Hangzhou, Zhejiang, China, 2School of Laboratory Medicine and Life Science, Wenzhou Medical University, Wenzhou, Zhejiang, China, 3Department of Scientiﬁc Research, Zhejiang Provincial People’s Hospital, People’s Hospital of Hangzhou Medical College, Hangzhou, Zhejiang, China Pancreatic adenocarcinoma (PAAD) is one of the most common malignant tumors with poor prognosis worldwide. Mounting evidence suggests that the expression of lncRNAs and the inﬁltration of immune cells have prognostic value for patients with PAAD. We used Gene Expression Omnibus (GEO) database and identiﬁed six genes (COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2) that could affect the survival rate of pancreatic adenocarcinoma patients. Based on a series of in silico analyses for reverse prediction of target genes associated with the prognosis of PAAD, a ceRNA network of mRNA (COL1A2, ITGA2, LAMA3, LAMB3, and LAMC2)–microRNA (miR-15a-5p)–long non-coding RNA (LINC00511, LINC01578, PVT1, and TNFRSF14-AS1) was constructed. We used the algorithm “CIBERSORT” to assess the proportion of immune cells and found three overall survival (OS)–associated immune cells (monocytes, M1 macrophages, and resting mast cell). Moreover, the OS-associated gene level was signiﬁcantly positively associated with immune checkpoint expression and biomarkers of immune cells. In summary, our results clariﬁed that ncRNA-mediated upregulation of OS-associated genes and tumor-inﬁltration immune cells (monocytes, M1 macrophages M1, and resting mast cell resting) correlated with poor prognosis in PAAD. Specialty section: Specialty section: This article was submitted to Computational Genomics, a section of the journal Frontiers in Genetics Received: 17 February 2022 Accepted: 26 May 2022 Published: 11 July 2022 Keywords: pancreatic adenocarcinoma, ceRNA network, immune inﬁltration, therapeutic targets, prognosis Edited by: Manal S. Fawzy, Suez Canal University, Egypt Reviewed by: Rongying Ou, First Afﬁliated Hospital of Wenzhou Medical University, China Li Xueling, Inner Mongolia University, China Correspondence: LiQin Jin liqinjin@126.com Ke Hao haoke@hmc.edu.cn Edited by: Manal S. Fawzy, Suez Canal University, Egypt Reviewed by: Rongying Ou, First Afﬁliated Hospital of Wenzhou Medical University, China Li Xueling, I M li U i it Chi Reviewed by: Rongying Ou, First Afﬁliated Hospital of Wenzhou Medical University, China Li Xueling, I M li U i it Chi Data Collection and Differential Gene Expression Analysis The four pancreatic cancer gene chip data used in this study (GSE63111, GSE23397, GSE62165, and GSE43795) were derived from the NCBI-GEO database (https://www.ncbi.nlm.nih.gov/ geo). Among them, GSE63111 was included in 35 cases, including in seven tumor samples and 28 non-tumor samples; GSE23397 was included in 21 cases, including in 15 tumor samples and six non-tumor samples; GSE62165 was included in 131 cases, including in 118 tumor samples and 13 non-tumor samples; GSE43795 was included in 31 cases, including in 26 tumor samples and ﬁve non-tumor samples. We used the limma package in the R software to screen differentially expressed genes (DEGs) (Ritchie et al., 2015). The screening threshold was set to p < 0.05, and the absolute value of log-fold change \| log2FC\| ≥2. We used the online tool “Venny” to construct the Venn diagram of the DEGs and identiﬁed common DEGs (Oliveros, 2007-2015). effective biomarkers. With advances in high-throughput sequencing technology, more biomarkers can be discovered to determine prognosis and improve treatment strategies (Zhang et al., 2018; Ping et al., 2020; Qiu et al., 2020). Non-coding RNAs (ncRNAs) have an important role in the occurrence and development of cancer and are promising biomarkers and therapeutic targets (Cai et al., 2020; Mu et al., 2020; Tseng et al., 2020). lncRNAs are broadly deﬁned as non-coding RNA with a length greater than 200 nucleotides, which previously were thought to have limited protein-coding ability (Mi et al., 2020). However, accumulating evidence demonstrated that speciﬁc lncRNAs are related to cancer recurrence, progression, and metastasis (Lin et al., 2020a; Zhou et al., 2020a; Peng et al., 2020; Teng et al., 2020). Salmena et al. proposed the competitive endogenous RNA (ceRNA) hypothesis that lncRNA can affect mRNA through competitively combining different miRNAs (Salmena et al., 2011). miRNAs are a type of small single-stranded ncRNA with regulatory functions, with a length of 21–24 nucleotides (Zhang et al., 2020a; Lin et al., 2020b; Zhou et al., 2020b). One hypothesis of ceRNA considers that miRNAs can regulate multiple target genes, and the same target genes can be regulated by different miRNAs. Also, the ceRNA hypothesis also illustrates that the expression levels of lncRNA and miRNA are negatively correlated and positively correlated with mRNA expression (Wang and Liu, 2017; Li et al., 2020). Data Collection and Differential Gene Expression Analysis Previous studies have focused on elucidating the involvement of the ceRNA network in pancreatic adenocarcinoma progression, metastasis, and prognosis, such as LINC00958 and TP73-AS1 (Chen et al., 2019; Miao et al., 2021). In addition, tumor cells and aggressive immune cells participate in the occurrence and development of cancer (Hanahan and Weinberg, 2011). It has become clear that evaluating the type and degree of tumor- inﬁltrating immune cells is very important for predicting progression and prognosis. However, only a few studies focus on the interaction mechanism between ceRNA networks and immune inﬁltrating cells. Therefore, a deeper understanding of ceRNA networks and pancreatic adenocarcinoma immune inﬁltrating cells is needed. MATERIALS AND METHODS resection usually have a local or distant recurrence within 2 years after the surgery (Hessmann et al., 2020). Other treatment strategies such as radiation, combination chemotherapy, and biological therapy also have limited efﬁcacy (Iacobuzio-Donahue et al., 2009; Pipas et al., 2012). The most fundamental reason is that the pathogenesis and prognostic markers of PAAD are not clear. For better performing early diagnosis, treatment, and prognostic evaluation of PAAD, it is particularly important to ﬁnd effective biomarkers. resection usually have a local or distant recurrence within 2 years after the surgery (Hessmann et al., 2020). Other treatment strategies such as radiation, combination chemotherapy, and biological therapy also have limited efﬁcacy (Iacobuzio-Donahue et al., 2009; Pipas et al., 2012). The most fundamental reason is that the pathogenesis and prognostic markers of PAAD are not clear. For better performing early diagnosis, treatment, and prognostic evaluation of PAAD, it is particularly important to ﬁnd effective biomarkers. Functional Enrichment Analysis, Interaction Network Analysis, and Hub Gene Identiﬁcation To further elucidate the potential functional annotation and pathway enrichment related to DEGs, we used the clusterProﬁler package to perform the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) enrichment analyses (Yu et al., 2012). Signiﬁcance was deﬁned as p < 0.05. We used an online database Search Tool for the Retrieval of Interacting Genes (STRING, https://string- db.org/) to construct the protein–protein interaction (PPI) network of DEGs, and the conﬁdence score ≥0.9 was set as the threshold. We removed protein nodes that did not interact with other proteins. In addition, we used Cytoscape v.3.8.2 to analyze the PPI network for screening the hub genes. The CytoHubba (version: 0.1) plug-in was used to identify hub genes in the PAAD. Citation: Wang Y, Wang Z, Li K, Xiang W, Chen B, Jin L and Hao K (2022) lncRNAs Functioned as ceRNA to Sponge miR-15a-5p Affects the Prognosis of Pancreatic Adenocarcinoma and Correlates With Tumor Immune Inﬁltration. Front. Genet. 13:874667. doi: 10.3389/fgene.2022.874667 Pancreatic adenocarcinoma (PAAD) is one of the malignant tumors of the digestive system, and it is difﬁcult to diagnose and treat (Siegel et al., 2018). Due to insufﬁcient early symptoms, more than 50% of PAAD patients have metastatic disease at diagnosis (Balsano et al., 2019). Despite the signiﬁcant improvement in diagnostic imaging and surgical mortality in the past 20 years, the 5-year survival rate is still less than 5% (Philip et al., 2009). The treatment options for pancreatic cancer are limited. About 80% of patients with PAAD lose the chance of surgical resection (Wang et al., 2020). In addition, patients undergoing complete tumor July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 1 Wang et al. Pancreatic Adenocarcinoma; Therapeutic Targets Construction of the ceRNA Network According to the ceRNA hypothesis, we constructed ceRNA networks by backward prediction. First, we used multiple target gene prediction Survival Analysis and Veriﬁcation Survival Analysis and Veriﬁcation Kaplan–Meier survival analysis curves (p < 0.05) and a univariate Cox proportional hazards model were used to assess the prognostic value of all biomarkers. GEPIA (http://gepia.cancer- pku.cn/) is a web server for gene expression analyses based on TCGA and GTEx databases (Tang et al., 2017). Using this database to perform differential expression analysis between tumor samples and non-tumor samples. Kaplan–Meier Plotter (http://kmplot.com/analysis/) is an online server that can access the effects of genes on survival in multiple cancer types (Nagy et al., 2021). The statistical signiﬁcance of expression analyses and survival analysis is deﬁned as p < 0.05. Differential expressions of these hub genes at the transcription level were inspected by matching pancreatic tumors and adjacent normal pancreatic tissues from the microarray data we selected in the GEO database. To verify our results, We used the Human Protein In this study, we established a signiﬁcant ceRNA network that involved four lncRNAs (LINC00511, LINC01578, PVT1, and TNFRSF14-AS1), one miRNA (hsa-miR-15a-5p), and six mRNAs (COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2). Moreover, we used CIBERSORT to infer the proportion of immune cells in pancreatic adenocarcinoma samples (Chen et al., 2018). In addition, we determined the relationship of OS-associated gene expression with biomarkers of immune cells and immune checkpoints in PAAD. July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 2 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. as (HPA) online database (www.proteinatlas.org/) to rmine the expression of these genes at the translational l (Thul et al., 2017). Construction of the ceRNA Network According to the ceRNA hypothesis, we constructed ceRNA networks by backward prediction. First, we used multiple target gene prediction GURE 1 \| Screening of differentially expressed genes. Volcano plot of GSE63111 (A), GSE23397 (B), GSE62165 (C), and GSE43795 (D). (E,F) Upregulated and wnregulated DEGs in the GSE63111, GSE23397, GSE62165, and GSE43795. DEGs, differentially expressed genes. g et al. Pancreatic Adenocarcinoma; Therapeutic Targets Atlas (HPA) online database (www.proteinatlas.org/) to determine the expression of these genes at the translational level (Thul et al., 2017). Construction of the ceRNA Network According to the ceRNA hypothesis, we constructed ceRNA networks by backward prediction. First, we used multiple target gene prediction July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 3 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. FIGURE 2 \| GO enrichment analysis, KEGG pathway analysis, and PPI network of DEGs. Survival Analysis and Veriﬁcation (A) Bubble chart shows GO enrichment signiﬁcance items of DEGs. (B) Bubble chart shows enrichment of DEGs in signaling pathways. (C) PPI network of DEGs. PPI, protein–protein interaction. GO, Gene Ontology. KEGG, Kyoto Encyclopedia of Genes and Genomes. FIGURE 2 \| GO enrichment analysis, KEGG pathway analysis, and PPI network of DEGs. (A) Bubble chart shows GO enrichment signiﬁcance items of DEGs. (B) Bubble chart shows enrichment of DEGs in signaling pathways. (C) PPI network of DEGs. PPI, protein–protein interaction. GO, Gene Ontology. KEGG, Kyoto Encyclopedia of Genes and Genomes. survival analysis were performed on the identiﬁed lncRNA. p < 0.05 was considered statistically signiﬁcant. survival analysis were performed on the identiﬁed lncRNA. p < 0.05 was considered statistically signiﬁcant. programs to predict the upstream binding miRNAs of OS-associated genes, including miRbase, miRNet, starBase, Tarbase, PITA, and RNA22. Only the predicted miRNAs that appeared in more than two programs previously were regarded as potential miRNAs for subsequent analyses. Then, the mRNA–miRNA network was established by Cytoscape, and the CytoHubba plug-in was used to determine the top potential miRNAs. Next, Starbase (http://starbase. sysu.edu.cn/) was used to perform survival analysis and differential expression on miRNAs. The upstream lncRNAs of miRNAs identiﬁed earlier were predicted in the Starbase. Finally, expression analysis and Cibesort Estimation and Survival Analysis of Key Members of the Immune Cells CIBERSORT (http://cibersort.stanford.edu/) was utilized to evaluate the 22 kinds of immune cell types in PAAD. The p-value < 0.05 was chosen as the criterion to identify the immune-inﬁltrating cells July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 4 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. TABLE 1 \| Cox proportional hazard regression model for OS in patients with PAAD. Gene Coef HR se (coef) 95%CI_l 95%CI_u z_score P value ITGA2 0.335 1.398 0.107 1.134 1.725 3.130971659 0.00174229* COL1A2 0.149 1.161 0.075 1.001 1.346 1.977415373 0.047994698* LAMC2 0.296 1.345 0.081 1.146 1.577 3.635873935 0.00027704* ITGB6 0.309 1.362 0.078 1.170 1.587 3.97318082 7.09E-05* LAMB3 0.269 1.309 0.082 1.115 1.535 3.29746163 0.00097563* LAMA3 0.335 1.398 0.085 1.182 1.652 3.9240794 8.71E-05* CI, conﬁdence interval. p < 0.05, p < 0.01, and *p < 0.001. In the variable selection process, ﬁrst of all, the univariate Cox models including all DEGs were used to select potential prognostic genes. At the same time, the Kaplan–Meier analysis was performed based on all DEGs. The results of the Kaplan–Meier analysis (Figures 4A–F) suggested that the six genes have potential prognostic value. Eventually, the univariate Cox model is shown in this table only including six genes ﬁltrating by Kaplan–Meier analysis. TABLE 1 \| Cox proportional hazard regression model for OS in patients with PAAD. CI, conﬁdence interval. p < 0.05, p < 0.01, and p < 0.001. In the variable selection process, ﬁrst of all, the univariate Cox models including all DEGs were used to select potential prognostic genes. At the same time, the Kaplan–Meier analysis was performed based on all DEGs. The results of the Kaplan–Meier analysis (Figures 4A–F) suggested that the six genes have potential prognostic value. Eventually, the univariate Cox model is shown in this table only including six genes ﬁltrating by Kaplan–Meier analysis. FIGURE 3 \| Correlation between OS-related gene expression and tumor stage in PAAD patients (GEPIA). FIGURE 3 \| Correlation between OS-related gene expression and tumor stage in PAAD patients (GEPIA). io/timer/), which is an online server that can analyze immune inﬁltrates across multiple cancer types (Li et al., 2017). io/timer/), which is an online server that can analyze immune inﬁltrates across multiple cancer types (Li et al., 2017). possibly affected by the expression of OS-associated genes. To identify the interrelation between 22 types of immune cells, we performed a correlation-based heat map analysis between every two immune cells. Cibesort Estimation and Survival Analysis of Key Members of the Immune Cells To identify OS-associated immune cells, we used Kaplan–Meier survival analysis to determine the prognostic value. Immune Inﬁltrate Analysis and Immune Checkpoint Expression Analysis Identiﬁcation of Differential Genes Through the screening of gene expression microarrays associated with PAAD from the GEO database, four datasets (GSE63111, GSE23397, GSE62165, and GSE43795) were chosen. The DEGs in four datasets are shown in Figures 1A–D. The co-expressed DEGs were conﬁrmed by the Venn We used the GEPIA database to identify the interrelation between OS- associated genes and biomarkers of immune cells. To investigate the association between the expression of OS-associated genes and immune checkpoint expression, we applied TIMER (https://cistrome.shinyapps. July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 5 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. FIGURE 4 \| Survival analysis results of six DEGs. (A–F) Differential expression and survival analysis of the six selected genes via the GEPIA database. From the abovementioned ﬁgures, it was found that these six genes were differentially expressed in the normal group and cancer group and obviously reduce the OS of patients. OS, overall survival. FIGURE 4 \| Survival analysis results of six DEGs. (A–F) Differential expression and survival analysis of the six selected genes via the GEPIA database. From the abovementioned ﬁgures, it was found that these six genes were differentially expressed in the normal group and cancer group and obviously reduce the OS of patients. OS, overall survival. Diagram online tool, including 72 upregulated and 58 downregulated genes (Figures 1D–F). biological process (BP) group, DEGs were mainly enriched in extracellular matrix organization, extracellular structure organization, and external encapsulating structure organization (Figure 2A). In the cellular component (CC) group, DEGs were mainly enriched in the endoplasmic reticulum lumen and collagen- containing extracellular matrix (Figure 2A). In the molecular function (MF) group, DEGs were mainly enriched in endopeptidase activity, extracellular matrix structural constituent, Functional Enrichment Analysis of DEGs, and Protein-Protein Interaction Network Functional Enrichment Analysis of DEGs, and Protein-Protein Interaction Network Functional Enrichment Analysis of DEGs, and Protein-Protein Interaction Network To further investigate the biological functions of DEGs, we conducted GO and KEGG functional enrichment analyses. In the To further investigate the biological functions of DEGs, we conducted GO and KEGG functional enrichment analyses. In the July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 6 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. FIGURE 5 \| Comparison of the hub gene mRNA levels in paired adjacent normal tissues and PAAD tissues from the GEO database. (A) COL1A2, (B) ITGA2, (C) ITGB6, (D) LAMA3, (E) LAMB3, and (F) LAMC2. PAAD, pancreatic adenocarcinoma. FIGURE 5 \| Comparison of the hub gene mRNA levels in paired adjacent normal tissues and PAAD tissues from the GEO database. (A) COL1A2, (B) ITGA2, (C) ITGB6, (D) LAMA3, (E) LAMB3, and (F) LAMC2. PAAD, pancreatic adenocarcinoma. FIGURE 6 \| Veriﬁcation of hub DEG expression in PAAD and normal tissue using the HPA database. (A) COL1A2, (B) ITGA2, (C) ITGB6, (D) LAMA3, (E) LAMB3, and (F) LAMC2. and glycosaminoglycan binding (Figure 2A). Moreover, KEGG pathway analysis results demonstrated that DEGs were signiﬁcantly enriched in pancreatic secretion and protein digestion and absorption (Figure 2B). prognosis-related biomarkers were hub genes of the PPI network (Table 1). We identiﬁed these six genes (COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2) that could obviously affect the OS rate of PAAD patients. We analyzed the expression of OS-related genes with tumor stage for PAAD. The results showed that the expression levels of ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2 are signiﬁcantly correlated with the tumor stage of PAAD patients, while the expression levels of COL1A2 are not correlated with the tumor stage of PAAD patients (Figure 3). prognosis-related biomarkers were hub genes of the PPI network (Table 1). We identiﬁed these six genes (COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2) that could obviously affect the OS rate of PAAD patients. We analyzed the expression of OS-related genes with tumor stage for PAAD. The results showed that the expression levels of ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2 are signiﬁcantly correlated with the tumor stage of PAAD patients, while the expression levels of COL1A2 are not correlated with the tumor stage of PAAD patients (Figure 3). The PPI network of DEGs was analyzed by the STRING online database. Functional Enrichment Analysis of DEGs, and Protein-Protein Interaction Network A total of 68 nodes and 122 edges were obtained in the PPI network (Figure 2C), and this network was then analyzed using Cytoscape software. Fourteen DEGs were determined as hub genes through the degree algorithm in the CytoHubba plug-in. Construction of the ceRNA Network To construct a prognostic ceRNA in PAAD, we ﬁrst predicted upstream miRNAs that could potentially bind to six mRNAs with prognostic values and ﬁnally found 160 miRNAs (Figure 7). Cytohubba identiﬁed the top fourteen highly connected miRNAs (Figure 8A). The starBase database identiﬁed one miRNA, hsa- miR-15a-5p, which was obviously downregulated in PAAD, and its up-regulation was positively correlated with patient prognosis Survival Analysis and Veriﬁcation The difference between the tumor samples and the non- tumor samples was statistically signiﬁcant (Figures 4A–F). We used R software to analyze the expression of six hub genes at the transcriptional level. The results showed that the expression of COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2 in Cox regression, Kaplan–Meier, and the log-rank test (p < 0.05) were used to assess the correlation between DEGs and prognosis. Twenty-ﬁve genes were found to be signiﬁcant in the Kaplan–Meier analysis. However, only six promising July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 7 Wang et al. Pancreatic Adenocarcinoma; Therapeutic Targets FIGURE 7 \| Result of the use of mRNA to reverse-predict miRNA. Red represents highly expressed mRNA, and blue represents miRNA. RE 7 \| Result of the use of mRNA to reverse-predict miRNA. Red represents highly expressed mRNA, and blue represents miRNA (Figures 8B,C). Next, we use the starBase database to predict the upstream lncRNAs of hsa-miR-15a-5p. 286 possible lncRNAs were obtained. Then, the expression levels of these lncRNAs in PAAD were identiﬁed by GEPIA. As shown in Figure 9, LINC00511, LINC01578, PVT1, and TNFRSF14-AS1were signiﬁcantly upregulated in PAAD compared with normal controls (Figures 9A,C,E,G). Overexpressed LINC00511, LINC01578, PVT1, and TNFRSF14-AS1 indicated poor OS in patients with PAAD (Figures 9B,D,F,H). PAAD tissue was higher than that in adjacent normal pancreatic tissues (Figures 5A–F). Subsequently, we used the HPA database to explore the expression of the hub gene in protein at the translational level. The results illustrated that the protein level of COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2 was higher in PAAD tissues than in normal pancreatic tissues, matching their mRNA expression levels (Figures 6A–F). Composition and Co-Expression Analysis of the Immune Cells in Pancreatic Adenocarcinoma We use the CIBERSORT algorithm to evaluate the composition of tumor-inﬁltrating immune cells in PAAD tissue. The histogram (Figure 10A) and the heat map (Figure 10B) illustrate the July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 8 Wang et al. Pancreatic Adenocarcinoma; Therapeutic Targets FIGURE 8 \| Identiﬁcation of hsa-miR-15a-5p as a potential miRNA in PAAD. Cytoscape software was used to screen the miRNAs that were closely related to hub- mRNAs. Through the differential expression and survival analysis of the starbase database, one miRNA was selected. (A) miRNA–mRNA regulatory network was set up by Cytoscape software. (B) Expression of hsa-miR-15a-5p in PAAD and normal samples was examined by the starBase database. (C) Prognostic value of hsa-miR-15a- 5p in PAAD was evaluated by the Kaplan–Meier plotter. FIGURE 8 \| Identiﬁcation of hsa-miR-15a-5p as a potential miRNA in PAAD. Cytoscape software was used to screen the miRNAs that were closely related to hub- mRNAs. Through the differential expression and survival analysis of the starbase database, one miRNA was selected. (A) miRNA–mRNA regulatory network was set up by Cytoscape software. (B) Expression of hsa-miR-15a-5p in PAAD and normal samples was examined by the starBase database. (C) Prognostic value of hsa-miR-15a- 5p in PAAD was evaluated by the Kaplan–Meier plotter. prognostic tumor-inﬁltrating immune cells. (Figure 11D). Finally, Figure 11E showed the relationship of co-expression between hub genes and immune cells. The fraction of M1 macrophages was positively correlated with COL1A2 expression (R = 0.41, p < 0.001) (Supplementary Figure S1A). The M1 macrophages were positively correlated with LAMA3 expression (R = 0.28, p < 0.001) (Supplementary Figure S1B). The monocytes were negatively correlated with LAMA3 expression (R = −0.38) (Supplementary Figure S1C). The monocytes were negatively correlated with ITGA2 expression (R = −0.3, p < 0.001) (Supplementary Figure S1D). proportions of the 22 immune cells. The violin plot (Figure 10C) reveals the results of the Wilcoxon rank-sum test. The data showed that the fraction of follicular helper T cells (p < 0.001), monocytes (p = 0.017), M1 macrophages (p = 0.022), resting mast cells (p = 0.001), and activated mast cells (p = 0.008) was obviously different between the pancreatic adenocarcinoma samples and normal pancreatic samples. Then, we used Kaplan–Meier analysis to assess the correlation between the different immune cell subtypes and prognosis. OS-Associated Gene Expression in PAAD OS-Associated Gene Expression in PAAD We researched the relationship between the expression of OS- associated genes and biomarkers of immune cells. Monocytes, M1 macrophages, and resting mast cells are the signiﬁcant tumor- inﬁltrating immune cell. The results were strongly correlated with OS-associated immune cells (Table 2). PD1/PD-L1 and CTLA-4 are signiﬁcant immune checkpoints for immune escape. We evaluated the relationship of OS-associated genes with PD1, PD-L1, and CTLA-4 by the TIMER platform. As suggested in Figure 12, OS-associated gene expression was clearly positively associated with PD1, PD-L1, and CTLA-4 in PAAD, which was Composition and Co-Expression Analysis of the Immune Cells in Pancreatic Adenocarcinoma The fraction of monocytes (p = 0.00921), M1 macrophages (p = 0.0315), and resting mast cell (p = 0.0499) was found to be obviously associated with OS (Figures 11A–C). We perform co-expression analysis on July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 9 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. FIGURE 9 \| Continued. FIGURE 9 \| Continued. Association of Biomarkers of Immune Cells and Immune Checkpoints With adjusted by purity. These data indicate that immune escape might be concerned with the key gene–mediated poor prognosis of PAAD. DISCUSSION Pancreatic cancer is the deadliest form of cancer. Despite different types of treatment options that have been adopted for pancreatic cancer patients, the prognosis is still poor (Siegel et al., 2020). Molecularly targeted therapies targeting oncogenic genes associated with pancreatic cancer genesis and development are increasingly considered a promising way to cure pancreatic cancer (Paulson et al., 2013). Therefore, illustrating the July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 10 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. FIGURE 9 \| Expression analysis and survival analysis for upstream lncRNAs of hsa-miR-15a-5p in PAAD. Prediction of lncRNA by miRNA, and differential expression and prognostic analysis suggested that four lncRNAs have prognostic value. (A,C,E,G) Expression of LINC00511 (A), LINC01578 (C), PVT1 (E), and TNFRSF14-AS1 (G) in TCGA PAAD compared with “TCGA normal” or “TCGA and GTEx normal” data. (B,D,F,H) Overall survival (OS) analysis for LINC00511 (A), LINC01578 (C), PVT1 (E), and TNFRSF14-AS1 (G) in PAAD. p value < 0.05. FIGURE 9 \| Expression analysis and survival analysis for upstream lncRNAs of hsa-miR-15a-5p in PAAD. Prediction of lncRNA by miRNA, and differential expression and prognostic analysis suggested that four lncRNAs have prognostic value. (A,C,E,G) Expression of LINC00511 (A), LINC01578 (C), PVT1 (E), and TNFRSF14-AS1 (G) in TCGA PAAD compared with “TCGA normal” or “TCGA and GTEx normal” data. (B,D,F,H) Overall survival (OS) analysis for LINC00511 (A), LINC01578 (C), PVT1 (E), and TNFRSF14-AS1 (G) in PAAD. p value < 0.05. molecular mechanisms of the pathogenesis of pancreatic cancer and determining potential biomarkers are crucial to improve patient outcomes. screened by Cox regression and Kaplan–Meier analysis. According to the ceRNA hypothesis, we gradually determined upstream miRNAs and lncRNAs from mRNAs and ﬁnally successfully established the ceRNA network. All genes of the ceRNA network are associated with poor prognosis of PAAD. In addition, we found that hsa-miR-15a-5p and hub genes in ceRNA and immune cell subtypes (monocytes, M1 macrophages, and resting mast cell) could predict prognosis effectively. Finally, our study found that high expression of OS-associated genes was closely related to PD1, PD-L1, or CTLA-4 in PAAD, suggesting that targeting mRNAs might improve the efﬁcacy of immunotherapy in PAAD. Our study ﬁrst performed bioinformatics analysis based on GEO datasets, and 130 genes of differentially expressed genes were selected by analyzing the microarray (GSE63111, GSE23397, GSE62165, and GSE43795) related to pancreatic cancer, of which 72 were upregulated and 58 were downregulated. Frontiers in Genetics \| www.frontiersin.org DISCUSSION Enrichment analysis showed that DEGs were mainly concentrated in external encapsulating structure organization, extracellular structure organization, and extracellular matrix organization. Next, six key genes in the PPI network were July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 11 ang et al. Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. Pancreatic Adenocarcinoma; Therapeutic Targets URE 10 \| Proportion of the 22 immune cells detected by the CIBERSORT algorithm (A,B). Results of the Wilcoxon rank-sum test (C). Proportions of follicular er T cells (p < 0.001), monocytes (p = 0.017), M1 macrophages (p = 0.022), resting mast cells (p = 0.001), and activated mast cells (p = 0.008) were signiﬁcantly rent between pancreatic cancer samples and normal pancreatic samples. FIGURE 10 \| Proportion of the 22 immune cells detected by the CIBERSORT algorithm (A,B). Results of the Wilcoxon rank-sum test (C). Proportions of follicular helper T cells (p < 0.001), monocytes (p = 0.017), M1 macrophages (p = 0.022), resting mast cells (p = 0.001), and activated mast cells (p = 0.008) were signiﬁcantly different between pancreatic cancer samples and normal pancreatic samples. FIGURE 10 \| Proportion of the 22 immune cells detected by the CIBERSORT algorithm (A,B). Results of the Wilcoxon rank-sum test (C). Proportions of follicular helper T cells (p < 0.001), monocytes (p = 0.017), M1 macrophages (p = 0.022), resting mast cells (p = 0.001), and activated mast cells (p = 0.008) were signiﬁcantly different between pancreatic cancer samples and normal pancreatic samples. FIGURE 10 \| Proportion of the 22 immune cells detected by the CIBERSORT algorithm (A,B). Results of the Wilcoxon rank-sum test (C). Proportions of follicular helper T cells (p < 0.001), monocytes (p = 0.017), M1 macrophages (p = 0.022), resting mast cells (p = 0.001), and activated mast cells (p = 0.008) were signiﬁcantly different between pancreatic cancer samples and normal pancreatic samples. July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 12 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. FIGURE 11 \| Correlation between immune cell fraction and OS (A–C). Signiﬁcant co-expression patterns between proportions of immune cells (D). Signiﬁcant co- expression patterns between hub mRNAs and immune cells (E). FIGURE 11 \| Correlation between immune cell fraction and OS (A–C). Signiﬁcant co-expression patterns between proportions of immune cells (D). Signiﬁcant co- expression patterns between hub mRNAs and immune cells (E). DISCUSSION illustrated that the expression of ITGB6 is negatively related to prognosis in PAAD patients (Wu et al., 2019b), which is consistent with our survival analysis. LAMA3, LAMB3, and LAMC2 belong to the laminin gene family. Laminin is known to stimulate cell migration in cancers (Tripathi et al., 2008). Yang et al. Yang et al. (2019) found that four subunits of the laminin gene family (LAMA3, LAMA4, LAMB3, and LAMC2) were connected with the outcomes of PAAD patients. The ﬁndings in our study are consistent with previous reports as we found that the LAMA3, LAMB3, and LAMC2 genes were upregulated in patients with PAAD, with hazard ratios of 3.86, 2.18, and 3.06, respectively. Combined with patient survival curves, these genes were found to be poor prognostic markers in patients with PAAD. COL1A2 belongs to the family of ECM proteins. The ECM provides structural support, mediates cell-matrix adhesion, and integrates complex, multivalent signals for cells (Egusa et al., 2007; Hynes, 2009). Wu et al. Wu et al. (2019a) reduced the expression of COL1A2 to reduce the invasion and migration ability of pancreatic cancer cells and found that COL1A2 was highly expressed in pancreatic cancer patients (p < 0.05), which was related to poor prognosis in these patients. ITGA2 belongs to the integrin alpha chain family. Integrin is a membrane protein with cell adhesion and signal transduction functions (Desgrosellier and Cheresh, 2010). Studies have found that the expression of ITGA2 is an independent prognostic marker of PAAD and is related to the grading and aggressiveness of PAAD (Shang et al., 2019; Yan et al., 2020). They also found that the expression of ITGA2 could signiﬁcantly reduce the survival rate of PAAD patients, which is in accordance with our research. ITGB6 belongs to a family of β2 integrins. Some studies have Hsa-miR-15a-5p is a microRNA whose family has already become promising cancer-speciﬁc biomarkers (Shao et al., 2020). Compared with PAAD patients with high miR-15a-5p expression, July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 13 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. TABLE 2 \| Relationship between hub genes and biomarkers of immune cells in PAAD detected by GEPIA. DISCUSSION In addition, research shows a relationship between patient prognosis and tumor-associated macrophage inﬁltration in pancreatic cancer (Di Caro et al., 2016). In our present study, M1 macrophages which were chosen from CIEBERSORT and the co-expression correlation of COL1A2, LAMA3, and M1 macrophages. The correlation analysis results were consistent with the previous research that indicated that there is a regulatory connection between M1 macrophages and COL1A2 (Zhang et al., 2020b). Cells of the monocyte lineage contribute to tumor angiogenesis (Dalton et al., 2014). One study suggested monocytes are recruited into Pancreatic Ductal Adenocarcinoma tumors via IL35 and increased expression of genes that promote angiogenesis in IL35 products (Huang et al., 2018). Nevertheless, there are few research studies before about the connections of monocytes and Hsa-miR-15a-5p. Thus, we deduced that Hsa-miR-15a-5p, as a miRNA, might connect with monocytes and took active part in the progression of PAAD. It should be acknowledged that our study has some unavoidable limitations. First, the quantity of samples we gather from GEO databases is indeed limited, which suggests that clinico pathological information was not comprehensive and might bring about some potential errors or biases. In addition, only a limited number of links were predicted. The ceRNA network is highly complicated and diverse, and its complex regulatory network needs to be further elucidated. Finally, we did not consider the heterogeneity of the immune microenvironment which is allied to the immune inﬁltration location. The heterogeneity and complexity of the immune microenvironment may affect the generalization of results. However, to reduce the bias, we used numerous databases to reveal the expression of crucial biomarkers at the gene level, protein level, and tissue levels, and it showed that crucial biomarkers were upregulated in PAAD. This study provides valuable material for further research, in which we would further explore the expression levels of key members of the ceRNA network in clinical samples and the interaction mechanism between ceRNA and cell communication through experiments. In summary, through the integration of several bioinformatics analyses, we constructed a reverse mRNA model based on the LINC00511, LINC01578, PVT1, TNFRSF14-AS/hsa-miR-15a-5p/ COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2 ceRNA network, which can enhance our understanding of the development of PAAD. HR values of all genes in ceRNA are greater than 1, indicating that they are risk factors for PAAD, so they might be utilized as potential prognostic biomarkers and therapeutic targets for PAAD. DISCUSSION Immune Cells Biomarker COL1A2 ITGA2 ITGB6 LAMA3 LAMB3 LAMC2 R Value P Value R Value P Value R Value P Value R Value P Value R Value P Value R Value P Value Monocytes CD86 0.33a 3.7E-06*a 0.16a 0.033a 0.34a 1.1E-10*a 0.29a 2E-08a 0.32a 1.5E- 09a 0.38a 1.2E- 13a CD115(CSF1R) 0.26a 0.00048a 0.42a 2.2E-16a 0.33a 3.7E-10*a −0.16a 0.028a 0.23a 1E-05*a 0.33a 2.4E- 10a M1 macrophages NOS2 0.3a 3.2e-05a 0.21a 7.6e-05a 0.092 0.087 0.12a 0.022a 0.17a 0.0018a 0.15a 0.0063a PTGS2 0.15a 0.037a 0.28a 0.00012a 0.27a 0.00029a 0.19a 0.011a 0.29a 8.7E- 05*a 0.35a 1.6E- 06a IRF5 0.25a 6E-04a 0.12 0.11 0.36a 2.4E-12a 0.36a 3.2E- 12a 0.079 0.12 0.4a 4.4E- 15a Resting mast cell CD117(KIT) 0.22a 4.9E-05a 0.21a 0.0035a 0.24a 3.6E-06*a 0.18a 7E-04a 0.15a 0.0058a 0.22a 4.9E- 05*a CD203c (ENPP3) 0.18a 0.0016a 0.14a 0.0098*a −0.15a 0.045a −0.23a 0.0014a −0.31a 2.4E- 05a 0.082 0.12 SLC18A2 0.27a 0.00019a 0.17a 0.0016a 0.1 0.051 0.054 0.13 −0.23a 0.0016*a 0.12a 0.025a aThese results are statistically signiﬁcant. p value < 0.05; p value < 0.01; **p value < 0.001. TABLE 2 \| Relationship between hub genes and biomarkers of immune cells in PAAD detected by GEPIA. Figure 9E, F. Raman et al. identiﬁed a 5-gene panel (ADM, ASPM, DCBLD2, E2F7, and KRT6A) that performed well against previous signatures across multiple datasets and captures subtype-speciﬁc differences in patient prognosis (Raman et al., 2018). This study provides a framework for similar assessments in other cancers. Figure 9E, F. Raman et al. identiﬁed a 5-gene panel (ADM, ASPM, DCBLD2, E2F7, and KRT6A) that performed well against previous signatures across multiple datasets and captures subtype-speciﬁc differences in patient prognosis (Raman et al., 2018). This study provides a framework for similar assessments in other cancers. those patients with low miR-15a-5p levels have a shorter overall survival (Guo et al., 2020), which is consistent with the results of our Figure 8C. Functionally, exogenous miR-15a-5p expression decreases cell growth, epithelial-to-mesenchymal transition, and metastasis and increases cell cycle arrest in PAAD (Guo et al., 2014; Feng et al., 2019). Mechanistically, It has been conﬁrmed that miR-15a-5p exerts its anti-pancreatic cancer activity by directly targeting FGFR1, which is highly expressed and executes cancer- promoting actions in PAAD (Turner et al., 2008; Lehnen et al., 2013; Coleman et al., 2014). A study focusing on the mechanism of hsa- miR-15a-5p in pancreatic cancer independent from the CERS6-AS1 (Shen et al., 2021). Frontiers in Genetics \| www.frontiersin.org REFERENCES Guo, S., Fesler, A., Huang, W., Wang, Y., Yang, J., Wang, X., et al. (2020). Functional Signiﬁcance and Therapeutic Potential of miR-15a Mimic in Pancreatic Ductal Adenocarcinoma. Mol. Ther. - Nucleic Acids 19, 228–239. doi:10.1016/j.omtn.2019.11.010 Balsano, R., Tommasi, C., and Garajova, I. (2019). State of the Art for Metastatic Pancreatic Cancer Treatment: Where Are We Now? Anticancer Res. 39 (7), 3405–3412. doi:10.21873/anticanres.13484 Guo, S., Xu, X., Tang, Y., Zhang, C., Li, J., Ouyang, Y., et al. (2014). miR-15a Inhibits Cell Proliferation and Epithelial to Mesenchymal Transition in Pancreatic Ductal Adenocarcinoma by Down-Regulating Bmi-1 Expression. Cancer Lett. 344, 40–46. doi:10.1016/j.canlet.2013.10.009 Cai, L., Ye, Y., Jiang, Q., Chen, Y., Lyu, X., Li, J., et al. (2020). Author Correction: Epstein-Barr Virus-Encoded microRNA BART1 Induces Tumour Metastasis by Regulating PTEN-dependent Pathways in Nasopharyngeal Carcinoma. Nat. Commun. 11, 3437. doi:10.1038/s41467-020-17048-0 Hanahan, D., and Weinberg, R. A. (2011). Hallmarks of Cancer: The Next Generation. Cell. 144, 646–674. doi:10.1016/j.cell.2011.02.013 Chen, B., Khodadoust, M. S., Liu, C. L., Newman, A. M., and Alizadeh, A. A. (2018). Proﬁling Tumor Inﬁltrating Immune Cells with Cibersort. Methods Mol. Biol. 1711, 243–259. doi:10.1007/978-1-4939-7493-1_12 Hessmann, E., Buchholz, S. M., Demir, I. E., Singh, S. K., Gress, T. M., Ellenrieder, V., et al. (2020). Microenvironmental Determinants of Pancreatic Cancer. Physiol. Rev. Rev. 100, 1707–1751. doi:10.1152/ physrev.00042.2019 Chen, S., Chen, J. Z., Zhang, J. Q., Chen, H. X., Qiu, F. N., Yan, M. L., et al. (2019). Silencing of Long Noncoding RNA LINC00958 Prevents Tumor Initiation of Pancreatic Cancer by Acting as a Sponge of microRNA-330-5p to Down-Regulate PAX8. Cancer Lett. 446, 49–61. doi:10.1016/j.canlet.2018. 12.017 Huang, C., Li, Z., Li, N., Li, Y., Chang, A., Zhao, T., et al. (2018). Interleukin 35 Expression Correlates with Microvessel Density in Pancreatic Ductal Adenocarcinoma, Recruits Monocytes, and Promotes Growth and Angiogenesis of Xenograft Tumors in Mice. Gastroenterology 154, 675–688. doi:10.1158/0008-5472.CAN-17-161210.1053/j.gastro.2017. 09.039 Coleman, S. J., Chioni, A. M., Ghallab, M., Anderson, R. K., Lemoine, N. R., Kocher, H. M., et al. (2014). Nuclear Translocation ofFGFR1 andFGF2 in Pancreatic Stellate Cells Facilitates Pancreatic Cancer Cell Invasion. EMBO Mol. Med. 6, 467–481. doi:10.1002/emmm.201302698 Hynes, R. O. (2009). The Extracellular Matrix: Not Just Pretty Fibrils. Science 326, 1216–1219. doi:10.1126/science.1176009 Dalton, H. J., Armaiz-Pena, G. N., Gonzalez-Villasana, V., Lopez-Berestein, G., Bar-Eli, M., and Sood, A. K. (2014). Monocyte Subpopulations in Angiogenesis. Cancer Res. 74, 1287–1293. doi:10.1158/0008-5472.CAN-13-2825 Iacobuzio-Donahue, C. A., Fu, B., Yachida, S., Luo, M., Abe, H., Henderson, C. M., et al. (2009). 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 in the article/ Supplementary Material. FUNDING that hsa-miR-15a-5p regulates COL1A2, ITGA2, and LAMA3 so that monocytes and M1 macrophages are activated, which may have great effects on pancreatic cancer prognosis. This work was supported by grants from the National Natural Science Foundation of China, NO. 81600595, and NO. 82072366; the Medicine and Health Research Foundation of Zhejiang Province, Nos. 2019RC113, 2019KY017, 2019RC124, and 2020KY012; Key projects jointly constructed by the Ministry and the province of Zhejiang Medical and Health Science and Technology Project, Nos. WKJ-ZJ-2019 and WKJ-ZJ-2101. This work was supported by grants from the National Natural Science Foundation of China, NO. 81600595, and NO. 82072366; the Medicine and Health Research Foundation of Zhejiang Province, Nos. 2019RC113, 2019KY017, 2019RC124, and 2020KY012; Key projects jointly constructed by the Ministry and the province of Zhejiang Medical and Health Science and Technology Project, Nos. WKJ-ZJ-2019 and WKJ-ZJ-2101. This work was supported by grants from the National Natural Science Foundation of China, NO. 81600595, and NO. 82072366; the Medicine and Health Research Foundation of Zhejiang Province, Nos. 2019RC113, 2019KY017, 2019RC124, and 2020KY012; Key projects jointly constructed by the Ministry and the province of Zhejiang Medical and Health Science and Technology Project, Nos. WKJ-ZJ-2019 and WKJ-ZJ-2101. DISCUSSION Moreover, our study reveals the potential mechanism Mishra et al.s’ study represents the ﬁrst TCGA-based PDAC methylome data analysis and ﬁrst reported that several genes (B3GNT3, DMBT1, and DEPDC1B) and lncRNAs (PVT1, and GATA6-AS) are strongly correlated with pancreatic ductal adenocarcinoma survival (Kumar Mishra et al., 2019). The conclusions regarding PVT1 are consistent with our results in July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 14 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. GURE 12 \| Correlation of hub mRNA expression with PD-1, PD-L1, and CTLA-4 expression in PAAD. Spearman relationship of COL1A2 with an expression o -1 (A), PD-L1 (B), and CTLA-4 (C) in PAAD adjusted by purity using TIMER. Spearman relationship of ITGA2 with an expression of PD-1 (D), PD-L1 (E), and C in PAAD adjusted by purity using TIMER. Spearman relationship of LAMA3 with an expression of PD-1 (G), PD-L1 (H), and CTLA-4 (I) in PAAD adjusted by ng TIMER. Spearman relationship of LAMB3 with an expression of PD-1 (J), PD-L1 (K), and CTLA-4 (L) in PAAD adjusted by purity using TIMER. Spearma ationship of LAMC2 with an expression of PD-1 (M), PD-L1 (N), and CTLA-4 (O) in PAAD adjusted by purity using TIMER. FIGURE 12 \| Correlation of hub mRNA expression with PD-1, PD-L1, and CTLA-4 expression in PAAD. Spearman relationship of COL1A2 with an expression of PD-1 (A), PD-L1 (B), and CTLA-4 (C) in PAAD adjusted by purity using TIMER. Spearman relationship of ITGA2 with an expression of PD-1 (D), PD-L1 (E), and CTLA-4 (F) in PAAD adjusted by purity using TIMER. Spearman relationship of LAMA3 with an expression of PD-1 (G), PD-L1 (H), and CTLA-4 (I) in PAAD adjusted by purity using TIMER. Spearman relationship of LAMB3 with an expression of PD-1 (J), PD-L1 (K), and CTLA-4 (L) in PAAD adjusted by purity using TIMER. Spearman relationship of LAMC2 with an expression of PD-1 (M), PD-L1 (N), and CTLA-4 (O) in PAAD adjusted by purity using TIMER. July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 15 Pancreatic Adenocarcinoma; Therapeutic Targets Wang et al. AUTHOR CONTRIBUTIONS The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fgene.2022.874667/ full#supplementary-material The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fgene.2022.874667/ full#supplementary-material KH and LJ were responsible for study design; YW and ZW were involved in data collection; KH and LJ analyzed the data; YW wrote the manuscript. KH, LJ, ZW, KL, WX, and BC revised the manuscript. Supplementary Figure S1 \| Fraction of M1 macrophage cells was positively associated with COL1A2 expression (A) and LAMA3 expression (B). Monocyte cells were positively associated with LAMA3 (C) and ITGA2 (D). REFERENCES DPC4 Gene Status of the Primary Carcinoma Correlates with Patterns of Failure in Patients with Pancreatic Cancer. Jco 27, 1806–1813. doi:10.1200/jco.2008.17.7188 Desgrosellier, J. S., and Cheresh, D. A. (2010). Integrins in Cancer: Biological Implications and Therapeutic Opportunities. Nat. Rev. Cancer 10, 9–22. doi:10. 1038/nrc2748 Kumar Mishra, N., Southekal, S., and Guda, C. (2019). Survival Analysis of Multi-Omics Data Identiﬁes Potential Prognostic Markers of Pancreatic Ductal Adenocarcinoma. Front. Genet. 10, 624. doi:10.3389/fgene.2019. 00624 Di Caro, G., Cortese, N., Castino, G. F., Grizzi, F., Gavazzi, F., Ridolﬁ, C., et al. (2016). Dual Prognostic Signiﬁcance of Tumour-Associated Macrophages in Human Pancreatic Adenocarcinoma Treated or Untreated with Chemotherapy. Gut 65, 1710–1720. doi:10.1136/gutjnl- 2015-309193 Lehnen, N. C., von Mässenhausen, A., Kalthoff, H., Zhou, H., Glowka, T., Schütte, U., et al. (2013). Fibroblast Growth Factor Receptor 1 Gene Ampliﬁcation in Pancreatic Ductal Adenocarcinoma. Histopathology 63, 157–166. doi:10.1111/ his.12115 Egusa, H., Iida, K., Kobayashi, M., Lin, T. Y., Zhu, M., Zuk, P. A., et al. (2007). Downregulation of Extracellular Matrix-Related Gene Clusters during Osteogenic Differentiation of Human Bone Marrow- and Adipose Tissue-Derived Stromal Cells. Tissue Eng. 13, 2589–2600. doi:10.1089/ ten.2007.0080 Li, J., Wang, X., Wang, Y., and Yang, Q. (2020). H19 Promotes the Gastric Carcinogenesis by Sponging miR-29a-3p: Evidence from lncRNA-miRNA- mRNA Network Analysis. Epigenomics 12, 989–1002. doi:10.2217/epi-2020- 0114 Feng, H., Wei, B., and Zhang, Y. (2019). RETRACTED: Long Non-coding RNA HULC Promotes Proliferation, Migration and Invasion of Pancreatic Cancer Cells by Down-Regulating microRNA-15a. Int. J. Biol. Macromol. 126, 891–898. doi:10.1016/j.ijbiomac.2018.12.238 Li, T., Fan, J., Wang, B., Traugh, N., Chen, Q., Liu, J. S., et al. (2017). TIMER: A Web Server for Comprehensive Analysis of Tumor-Inﬁltrating Immune Cells. Cancer Res. 77, e108–e110. doi:10.1158/0008-5472.can-17- 0307 July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 16 Wang et al. Pancreatic Adenocarcinoma; Therapeutic Targets Cells by Sequestering MicroRNA-15a-5p and MicroRNA-6838-5p and Modulating HMGA1. Pancreas 50 (4), 617–624. doi:10.1097/MPA. 0000000000001806 Cells by Sequestering MicroRNA-15a-5p and MicroRNA-6838-5p and Modulating HMGA1. Pancreas 50 (4), 617–624. doi:10.1097/MPA. 0000000000001806 Lin, L., Que, Y., Lu, P., Li, H., Xiao, M., Zhu, X., et al. (2020). Chidamide Inhibits Acute Myeloid Leukemia Cell Proliferation by lncRNA VPS9D1-AS1 Downregulation via MEK/ERK Signaling Pathway. Front. Pharmacol. 11, 569651. doi:10.3389/fphar.2020.569651 Siegel, R. L., Miller, K. D., and Jemal, A. (20182018)., 68. CA, 7–30. doi:10.3322/ caac.21442Cancer Statistics, 2018CA A Cancer J. Clin.1 Lin, L., Que, Y., Lu, P., Li, H., Xiao, M., Zhu, X., et al. (2020). REFERENCES Chidamide Inhibits Acute Myeloid Leukemia Cell Proliferation by lncRNA VPS9D1-AS1 Downregulation via MEK/ERK Signaling Pathway. Front. Pharmacol. 11, 569651. doi:10.3389/fphar.2020.569651 Siegel, R. L., Miller, K. D., and Jemal, A. (20202020). Cancer Statistics, 2020. CA A Cancer J. Clin. 70, 7–30. doi:10.3322/caac.21590 Tang, Z., Li, C., Kang, B., Gao, G., Li, C., and Zhang, Z. (2017). GEPIA: a Web Server for Cancer and Normal Gene Expression Proﬁling and Interactive Analyses. Nucleic Acids Res. 45 (W1), W98–W102. doi:10. 1093/nar/gkx247 Mi, X., Xu, R., Hong, S., Xu, T., Zhang, W., and Liu, M. (2020). M2 Macrophage- Derived Exosomal lncRNA AFAP1-AS1 and MicroRNA-26a Affect Cell Migration and Metastasis in Esophageal Cancer. Mol. Ther. - Nucleic Acids 22, 779–790. doi:10.1016/j.omtn.2020.09.035 Teng, F., Zhang, J.-X., Chang, Q.-M., Wu, X.-B., Tang, W.-G., Wang, J.-F., et al. (2020). lncRNA MYLK-AS1 Facilitates Tumor Progression and Angiogenesis by Targeting miR-424-5p/E2F7 axis and Activating VEGFR-2 Signaling Pathway in Hepatocellular Carcinoma. J. Exp. Clin. Cancer Res. 39, 235. doi:10.1186/s13046-020-01739-z Miao, H., Lu, J., Guo, Y., Qiu, H., Zhang, Y., Yao, X., et al. (2021). LncRNA TP73- AS1 Enhances the Malignant Properties of Pancreatic Ductal Adenocarcinoma by Increasing MMP14 Expression through miRNA -200a Sponging. J. Cell. Mol. Med. 25 (7), 3654–3664. doi:10.1111/jcmm.16425 Thul, P. J., Åkesson, L., Wiking, M., Mahdessian, D., Geladaki, A., Ait Blal, H., et al. (2017). A Subcellular Map of the Human Proteome. Science 356 (6340). doi:10. 1126/science.aal3321 Mu, M., Niu, W., Zhang, X., Hu, S., and Niu, C. (2020). LncRNA BCYRN1 Inhibits Glioma Tumorigenesis by Competitively Binding with miR-619-5p to Regulate CUEDC2 Expression and the PTEN/AKT/p21 Pathway. Oncogene 39, 6879–6892. doi:10.1038/s41388-020-01466-x Tripathi, M., Nandana, S., Yamashita, H., Ganesan, R., Kirchhofer, D., and Quaranta, V. (2008). Laminin-332 Is a Substrate for Hepsin, a Protease Associated with Prostate Cancer Progression. J. Biol. Chem. 283, 30576–30584. doi:10.1074/jbc.M802312200 Nagy, Á., Munkácsy, G., and Győrffy, B. (2021). Pancancer Survival Analysis of Cancer Hallmark Genes. Sci. Rep. 11 (1), 6047. doi:10.1038/s41598-021- 84787-5 Oliveros, J. C. (2007-2015). Venny. An Interactive Tool for Comparing Lists with Venn’s Diagrams. https://bioinfogp.cnb.csic.es/tools/venny/index.html. Tseng, L.-L., Cheng, H.-H., Yeh, T.-S., Huang, S.-C., Syu, Y.-Y., Chuu, C.-P., et al. (2020). Targeting the Histone Demethylase PHF8-Mediated PKCα-Src-PTEN axis in HER2-Negative Gastric Cancer. Proc. Natl. Acad. Sci. U.S.A. 117, 24859–24866. doi:10.1073/pnas.1919766117 Paulson, A. S., Tran Cao, H. S., Tempero, M. A., and Lowy, A. M. (2013). Therapeutic Advances in Pancreatic Cancer. Gastroenterology 144, 1316–1326. doi:10.1053/j.gastro.2013.01.078 Turner, C. A., Calvo, N., Frost, D. REFERENCES O., Akil, H., and Watson, S. J. (2008). The Fibroblast Growth Factor System Is Downregulated Following Social Defeat. Neurosci. Lett. 430, 147–150. doi:10.1016/j. neulet.2007.10.041 Peng, W., Zhang, C., Peng, J., Huang, Y., Peng, C., Tan, Y., et al. (2020). lnc- FAM84B-4 Acts as an Oncogenic lncRNA by Interacting with Protein hnRNPK to Restrain MAPK Phosphatases-DUSP1 Expression. Cancer Lett. 494, 94–106. doi:10.1016/j.canlet.2020.08.036 Wang, L., and Liu, J. (2017). Research Progress of Competing Endogenous RNA. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi 34, 967–971. doi:10.7507/1001-5515.201606080 Philip, P. A., Mooney, M., Jaffe, D., Eckhardt, G., Moore, M., Meropol, N., et al. (2009). Consensus Report of the National Cancer Institute Clinical Trials Planning Meeting on Pancreas Cancer Treatment. Jco 27 (33), 5660–5669. doi:10.1200/JCO.2009.21.9022 Wang, X., Li, X., Wei, X., Jiang, H., Lan, C., Yang, S., et al. (2020). PD-L1 Is a Direct Target of Cancer-FOXP3 in Pancreatic Ductal Adenocarcinoma (PDAC), and Combined Immunotherapy with Antibodies against PD-L1 and CCL5 Is Effective in the Treatment of PDAC. Sig Transduct. Target Ther. 5, 38. doi:10.1038/s41392-020-0144-8 Ping, Y., Zhou, Y., Hu, J., Pang, L., Xu, C., and Xiao, Y. (2020). Dissecting the Functional Mechanisms of Somatic Copy-Number Alterations Based on Dysregulated ceRNA Networks across Cancers. Mol. Ther. - Nucleic Acids 21, 464–479. doi:10.1016/j.omtn.2020.06.012 Wu, J., Liu, J., Wei, X., Yu, Q., Niu, X., Tang, S., et al. (2019). A Feature-Based Analysis Identiﬁes COL1A2 as a Regulator in Pancreatic Cancer. J. Enzyme Inhibition Med. Chem. 34 (1), 420–428. doi:10.1080/14756366.2018. 1484734 Pipas, J. M., Zaki, B. I., McGowan, M. M., Tsapakos, M. J., Ripple, G. H., Suriawinata, A. A., et al. (2012). Neoadjuvant Cetuximab, Twice-Weekly Gemcitabine, and Intensity-Modulated Radiotherapy (IMRT) in Patients with Pancreatic Adenocarcinoma. Ann. Oncol. 23, 2820–2827. doi:10.1093/ annonc/mds109 Wu, M., Li, X., Zhang, T., Liu, Z., and Zhao, Y. (2019). Identiﬁcation of a Nine- Gene Signature and Establishment of a Prognostic Nomogram Predicting Overall Survival of Pancreatic Cancer. Front. Oncol. 9, 996. doi:10.3389/ fonc.2019.00996 Qiu, Y. L., Zheng, H., Devos, A., Selby, H., and Gevaert, O. (2020). A Meta- Learning Approach for Genomic Survival Analysis. Nat. Commun. 11, 6350. doi:10.1038/s41467-020-20167-3 Yan, W., Liu, X., Wang, Y., Han, S., Wang, F., Liu, X., et al. (2020). Identifying Drug Targets in Pancreatic Ductal Adenocarcinoma through Machine Learning, Analyzing Biomolecular Networks, and Structural Modeling. Front. Pharmacol. 11, 534. doi:10.3389/fphar.2020.00534 Raman, P., Maddipati, R., Lim, K. H., and Tozeren, A. (2018). Zhou, R., Sun, H., Zheng, S., Zhang, J., Zeng, D., Wu, J., et al. (2020). A Stroma- related lncRNA Panel for Predicting Recurrence and Adjuvant Chemotherapy Beneﬁt in Patients with Early-stage Colon Cancer. J. Cell. Mol. Med. 24, 3229–3241. doi:10.1111/jcmm.14999 REFERENCES Pancreatic Cancer Survival Analysis Deﬁnes a Signature that Predicts Outcome. PLoS One 13 (8), e0201751. doi:10.1371/journal.pone.0201751 Yang, C., Liu, Z., Zeng, X., Wu, Q., Liao, X., Wang, X., et al. (2019). Evaluation of the Diagnostic Ability of Laminin Gene Family for Pancreatic Ductal Adenocarcinoma. Aging 11 (11), 3679–3703. doi:10.18632/aging.102007 Ritchie, M. E., Phipson, B., Wu, D., Hu, Y., Law, C. W., Shi, W., et al. (2015). Limma Powers Differential Expression Analyses for RNA-Sequencing and Microarray Studies. Nucleic Acids Res. 43 (7), e47. doi:10.1093/nar/ gkv007 Yu, G., Wang, L.-G., Han, Y., and He, Q.-Y. (2012). clusterProﬁler: an R Package for Comparing Biological Themes Among Gene Clusters. OMICS A J. Integr. Biol. 16, 284–287. doi:10.1089/omi.2011.0118 Salmena, L., Poliseno, L., Tay, Y., Kats, L., and Pandolﬁ, P. P. (2011). A ceRNA Hypothesis: the Rosetta Stone of a Hidden RNA Language? Cell. 146, 353–358. doi:10.1016/j.cell.2011.07.014 Zhang, J., Sun, H., Liu, S., Huang, W., Gu, J., Zhao, Z., et al. (2020). Alteration of Tumor- Associated Macrophage Subtypes Mediated by KRT6A in Pancreatic Ductal Adenocarcinoma. aging 12 (22), 23217–23232. doi:10.18632/aging.104091 Shang, M., Zhang, L., Chen, X., and Zheng, S. (2019). Identiﬁcation of Hub Genes and Regulators Associated with Pancreatic Ductal Adenocarcinoma Based on Integrated Gene Expression Proﬁle Analysis. Discov. Med. 28 (153), 159–172. Zhang, Q., Song, G., Yao, L., Liu, Y., Liu, M., Li, S., et al. (2018). miR-3928v Is Induced by HBx via NF-Κb/egr1 and Contributes to Hepatocellular Carcinoma Malignancy by Down-Regulating VDAC3. J. Exp. Clin. Cancer Res. 37, 14. doi:10.1186/s13046-018-0681-y Shao, Y., Guo, X., Zhao, L., Shen, Y., Niu, C., Wei, W., et al. (2020). A Functional Variant of the miR-15 Family Is Associated with a Decreased Risk of Esophageal Squamous Cell Carcinoma. DNA Cell. Biol. 39 (9), 1583–1594. doi:10.1089/dna.2020.5606 Zhang, Y., Huang, Y.-X., Wang, D.-L., Yang, B., Yan, H.-Y., Lin, L.-H., et al. (2020). LncRNA DSCAM-AS1 Interacts with YBX1 to Promote Cancer Progression by Forming a Positive Feedback Loop that Activates FOXA1 Transcription Network. Theranostics 10, 10823–10837. doi:10.7150/thno. 47830 Shen, R., Wang, X., Wang, S., Zhu, D., and Li, M. (2021). Long Noncoding RNA CERS6-AS1 Accelerates the Proliferation and Migration of Pancreatic Cancer July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 17 Wang et al. 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. Frontiers in Genetics \| www.frontiersin.org July 2022 \| Volume 13 \| Article 874667 REFERENCES Pancreatic Adenocarcinoma; Therapeutic Targets 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. 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. Zhou, R., Sun, H., Zheng, S., Zhang, J., Zeng, D., Wu, J., et al. (2020). A Stroma-related lncRNA Panel for Predicting Recurrence and Adjuvant Chemotherapy Beneﬁt in Patients with Early-stage Colon Cancer. J. Cell. Mol. Med. 24, 3229–3241. doi:10. 1111/jcmm.14999 Copyright © 2022 Wang, Wang, Li, Xiang, Chen, Jin and Hao. 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. 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. July 2022 \| Volume 13 \| Article 874667 Frontiers in Genetics \| www.frontiersin.org 18
https://openalex.org/W2416796424	https://europepmc.org/articles/pmc4741190?pdf=render	English	null	Crystal structure of the flavoenzyme PA4991 from<i>Pseudomonas aeruginosa</i>	Acta crystallographica. Section F, Structural biology communications	2,016	cc-by	6,347	research communications Crystal structure of the flavoenzyme PA4991 from Pseudomonas aeruginosa ISSN 2053-230X ISSN 2053-230X Agata Jacewicz, Robert Schnell, Ylva Lindqvist and Gunter Schneider* Department of Medical Biochemistry and Biophysics, Karolinska Institutet, S-171 77 Stockholm, Sweden. *Correspon- dence e-mail: gunter.schneider@ki.se Received 4 December 2015 Accepted 18 December 2015 Edited by Z. S. Derewenda, University of Virginia, USA Received 4 December 2015 Accepted 18 December 2015 The locus PA4991 in Pseudomonas aeruginosa encodes an open reading frame that has been identiﬁed as essential for the virulence and/or survival of this pathogenic organism in the infected host. Here, it is shown that this gene encodes a monomeric FAD-binding protein of molecular mass 42.2 kDa. The structure of PA4991 was determined by a combination of molecular replacement using a search model generated with Rosetta and phase improvement by a low- occupancy heavy-metal derivative. PA4991 belongs to the GR2 family of FAD- dependent oxidoreductases, comprising an FAD-binding domain typical of the glutathione reductase family and a second domain dominated by an eight- stranded mixed -sheet. Most of the protein–FAD interactions are via the FAD- binding domain, but the isoalloxazine ring is located at the domain interface and interacts with residues from both domains. A comparison with the structurally related glycine oxidase and glycerol-3-phosphate dehydrogenase shows that in spite of very low amino-acid sequence identity (<18%) several active-site residues involved in substrate binding in these enzymes are conserved in PA4991. However, enzymatic assays show that PA4991 does not display amino- acid oxidase or glycerol-3-phosphate dehydrogenase activities, suggesting that it requires different substrates for activity. Edited by Z. S. Derewenda, University of Virginia, USA Edited by Z. S. Derewenda, University of Virginia, USA Keywords: molecular replacement; Rosetta model; phasing; oxidoreductase; FAD. Keywords: molecular replacement; Rosetta model; phasing; oxidoreductase; FAD. PDB reference: FAD-dependent oxidoreductase PA4991, 5ez7 PDB reference: FAD-dependent oxidoreductase PA4991, 5ez7 Supporting information: this article has supporting information at journals.iucr.org/f 2.1. Cloning, gene expression and protein purification The coding sequence of PA4991 was ampliﬁed by PCR from P. aeruginosa PAO1 genomic DNA using appropriate ampliﬁcation primers adapted to the ligation-independent cloning method and cloned into pNIC28BSA4 (GenBank Accession No. EF198106) as described previously (Moynie et al., 2013). The recombinant protein carries a TEV-cleavable His6 tag at the N-terminus. For a typical protein preparation, Escherichia coli BL21(DE3) cells harbouring the expression construct were grown at 37C in 3 l LB medium supplemented with kanamycin (30 mg l1) until the OD600 value reached 0.3, followed by growth for 1 h at 20C, after which expression was induced by the addition of 0.2 mM IPTG. The culture was grown under inductive conditions at 20C for 24 h. y FAD-dependent amino-acid oxidase activity, which results in the release of H2O2, was assayed using a colorimetric method described previously (Su et al., 2011). All 20 protei- nogenic amino acids, the d-amino acids d-Ala and d-Glu, several nonstandard amino acids (l-2-aminopimelate, d-2- aminopimelate, meso-diaminopimelic acid, -amino-adipate, l-norleucine, S-carboxy-l-cysteine, S-aminoethyl-l-cysteine, N-acetyl-l-cysteine, O-acetylserine and Gly-Gly) and nine amines (6-aminohexanoic acid, 1,5-diaminopentane, 1,6- diaminohexane, 1,8-diaminooctane, taurine, betaine, spermi- dine, spermine and sarcosine) were tested as potential substrates. The reaction mixtures consisted of 50 mM Tris– HCl, 100 mM NaCl pH 7.0, 2 mM of the putative substrate and PA4991 at 0.5 mM concentration. The detection limits for both assays were in the range 1–5 turnovers per minute. After harvesting by centrifugation at 4000g and 4C, the E. coli cells were resuspended in lysis buffer consisting of 20 mM Tris–HCl, 0.2 M NaCl, 10 mM imidazole pH 8.0. Lysozyme and DNAse were added to concentrations of 0.04 and 0.004 mg ml1, respectively, and the lysis mixture was incubated at 22C for 1 h. The cells were subsequently disrupted by sonication on ice and the lysate was clariﬁed by centrifugation (25 000g) at 4C. The supernatant was loaded onto 1.5 ml Ni2+ resin (Qiagen) equilibrated with lysis buffer and the column was washed with 50 ml of a solution consisting of 20 mM Tris–HCl, 0.2 M NaCl, 10 mM imidazole pH 8.0. The protein was eluted with a 10–500 mM imidazole gradient in 20 mM Tris–HCl buffer pH 8.0 containing 0.2 M NaCl. PA4991-containing fractions were combined and 0.7 mg TEV protease was added to the sample and incubated overnight at room temperature. research communications onto a 5 ml HiTrap Q HP column (GE Healthcare) equili- brated with 4% 40 mM Tris–HCl buffer containing 1 M NaCl. The recombinant protein was eluted with a 4–100% gradient of this buffer. The fractions containing PA4991 were collected and the sample was concentrated to a volume of 2 ml using a Vivaspin (Sartorius, Go¨ttingen, Germany) centrifugation concentrator device with a 10 kDa molecular-weight cutoff. The sample was then loaded onto a Superdex 200 column (GE Healthcare, Uppsala, Sweden) with running buffer consisting of 10 mM Tris–HCl pH 8.0, 80 mM NaCl. The fractions containing pure (>95%) PA4991 were collected and concen- trated to 40 mg ml1 and aliquots were ﬂash-frozen in liquid nitrogen. Selenomethionine-substituted PA4991 was produced according to the metabolic inhibition method (Van Duyne et al., 1993) and was puriﬁed as described above. and P14 in the sputum of cystic ﬁbrosis patients (Turner et al., 2015). The proteins encoded by these two open reading frames were included in a larger study aimed at the characterization and assessment of potential targets in P. aeruginosa for early drug discovery (Moynie et al., 2013). PA4992 belongs to the family of aldo–keto reductases, folding into an (/)8 barrel. NADPH binds at the C-terminal end of the -strands, close to helix 8. PA4992 also contains a catalytic triad, Asp66, Tyr71 and Lys94, typical of this enzyme family (Moynie et al., 2013). The proteins encoded by these two open reading frames were included in a larger study aimed at the characterization and assessment of potential targets in P. aeruginosa for early drug discovery (Moynie et al., 2013). PA4992 belongs to the family of aldo–keto reductases, folding into an (/)8 barrel. NADPH binds at the C-terminal end of the -strands, close to helix 8. PA4992 also contains a catalytic triad, Asp66, Tyr71 and Lys94, typical of this enzyme family (Moynie et al., 2013). Here, we report the crystallographic structure analysis of PA4991. The structure was determined by a combination of molecular replacement using a search model generated with Rosetta (Kim et al., 2004) and phase improvement employing a low-occupancy heavy-metal derivative. We show that the enzyme belongs to a family of FAD-dependent oxido- reductases and compare its three-dimensional structure with those of other members of this family. Enzymatic assays suggest that PA4991 has a different substrate spectrum to its closest relatives glycerol-3-phosphate dehydrogenase and glycine oxidase. 2.1. Cloning, gene expression and protein purification After TEV cleavage, the sample was diluted ﬁve times with 40 mM Tris–HCl buffer pH 8.5 and concentrated to a total volume of 2 ml. The sample was loaded research communications Here, we report the crystallographic structure analysis of PA4991. The structure was determined by a combination of molecular replacement using a search model generated with Rosetta (Kim et al., 2004) and phase improvement employing a low-occupancy heavy-metal derivative. We show that the enzyme belongs to a family of FAD-dependent oxido- reductases and compare its three-dimensional structure with those of other members of this family. Enzymatic assays suggest that PA4991 has a different substrate spectrum to its closest relatives glycerol-3-phosphate dehydrogenase and glycine oxidase. 2.2. Enzyme assays Enzymatic assays using recombinant PA4991 were carried out to probe for glycerol-3-phosphate dehydrogenase activity spectrophotometrically at 340 nm using both possible co- substrates: NAD+ and NADP+. The reaction mixtures consisted of 50 mM Tris–HCl, 100 mM NaCl pH 7.0, 1 mM glycerol-3-phosphate, 0.5 mM NAD+ or NADP+ and PA4991 at 1 mM concentration. Similarly, the reverse reaction based on the reduction of dihydroxyacetone-3-phosphate with NADH or NADPH was assayed. 1. Introduction Infections by Gram-negative bacteria are particularly difﬁcult to treat owing to the limited repertoire of antibiotics against these pathogens, increasing drug resistance and their tendency to form persistent infections (Zavascki et al., 2010). The ubiquitous pathogen Pseudomonas aeruginosa is able to adapt to a variety of environmental conditions and is therefore associated with a broad range of pathologies, including respiratory-tract, blood and skin infections. P. aeruginosa infections are mainly found in patients with immunosuppres- sion, burns or cystic ﬁbrosis (Kerr & Snelling, 2009). Increasing challenges in the clinical treatment of these infec- tions underline the need to identify novel targets and lead compounds for the development of antibacterial agents, in particular those that target Gram-negative pathogens (Rice, 2008). High-throughput genomic approaches have been employed to identify genes essential for the virulence and/or survival of pathogenic organisms in the infected host. Randomized transposon mutagenesis screens using P. aeruginosa strains PAO1 and P14 provided an estimate of 300–400 essential genes in PAO1 (Jacobs et al., 2003) and 335 genes in P14 (Liberati et al., 2006). Amongst these genes are PA4991 and PA4992, which are located on the same operon. A recent study using a transposon-sequencing and Monte Carlo simulation- based approach provided compelling evidence that the two genes are essential for growth of P. aeruginosa strains PAO1 105 http://dx.doi.org/10.1107/S2053230X15024437 Acta Cryst. (2016). F72, 105–111 research communications research communications Molecular replacement using this truncated model ﬁnally gave one solution that was clearly above the background, but with very low scores: RFZ = 4.9, TFZ = 4.8, PAK = 0, LLG = 30. From the phases obtained some more residues could be built; however, we did not deem it possible to bootstrap the complete structure. We then used Phaser-EP (McCoy et al., 2007) and the obtained molecular- replacement model to locate and reﬁne Hg sites and carry out phase improvement with Parrot (Zhang et al., 1997). This procedure was repeated interspersed with the Buccaneer pipeline (Cowtan, 2006) and manual model building with Coot (Emsley et al., 2010) until a complete structural model of PA4991 was obtained and reﬁned. Phaser-EP was able to localize two correct Hg sites (included in the ﬁnal model with occupancy 0.35), but also included some false sites during this procedure. During the model-building procedure electron density appeared for the cofactor FAD, which had not been included in any of the search models, thus indicating that the structure solution was correct. The ﬁnal reﬁnement run with REFMAC5 (Winn et al., 2001) included TLS reﬁnement. MolProbity (Chen et al., 2010) was used for validation. The Values in parentheses are for the outer shell. For the preparation of heavy-metal derivatives, crystals were soaked overnight in solutions of various salts [Pt(NH3)2Cl2, HgCl2 and phenylmercuric chloride] dissolved at concentrations of 1 and 2 mM in 0.2 M trimethylamine N-oxide, 20% PEG 2000 monomethyl ether, 0.1 M Tris–HCl pH 8.5. Before data collection, crystals were transferred to a solu- tion consisting of 30% PEG 2000 monomethyl ether, 0.2 M trimethylamine N-oxide, 80 mM NaCl, 10 mM Tris–HCl pH 8.0, 5%(w/v) glycerol and ﬂash-cooled in liquid nitrogen. The X-ray diffraction data were collected at 100 K on beamline ID14-4 at ESRF, Grenoble, France. The data sets were inte- grated with MOSFLM (Leslie, 2006) and scaled using SCALA from the CCP4 suite (Winn et al., 2011). The best diffracting crystal (2.4 A˚ resolution) was obtained by soaking native PA4991 crystals in 1 mM phenylmercuric chloride overnight and these data were used in the crystallographic analysis (Table 1). MolProbity (Chen et al., 2010) was used for validation. The ﬁnal model contains one polypeptide chain of PA4991, one FAD molecule, two Hg2+ ions and 41 water molecules. The crystallographic data for PA4991 have been deposited in the Protein Data Bank (PDB entry 5ez7). research communications research communications Table 1 Data-collection and reﬁnement statistics for PA4991. Values in parentheses are for the outer shell. Data collection Space group P21 Unit-cell parameters (A˚ , ) a = 37.9, b = 79.8, c = 63.2, = 104.4 Wavelength (A˚ ) 1.0093 Resolution (A˚ ) 48.6–2.40 (2.53–2.40) Rmerge 0.072 (0.298) Rmeas 0.084 (0.345) Rp.i.m. 0.042 (0.180) Total No. of observations 106569 No. of unique reﬂections 14372 hI/(I)i 17.3 (5.3) Completeness (%) 99.8 (98.7) Multiplicity 7.4 (7.2) Wilson B factor (A˚ 2) 41.2 Reﬁnement Resolution (A˚ ) 48.6–2.40 No. of reﬂections 13625 Rwork/Rfree 0.173/0.228 No. of protein residues 391 No. of FAD molecules 1 No. of atoms Protein 2780 FAD 53 Hg2+ (0.35 occupancy) 2 Water molecules 41 B factors (A˚ 2) Overall 40.0 Protein 41.3 FAD 25.1 Hg2+ (0.35 occupancy) 50.9 Water 31.1 R.m.s deviations Bond lengths (A˚ ) 0.014 Bond angles () 1.72 Ramachandran plot (%) Favoured 96.4 Allowed 3.6 Table 1 Data-collection and reﬁnement statistics for PA4991. metal sites, but the resulting electron-density map was un- interpretable and the sites were later shown to be false. The structure was therefore determined by molecular replacement with Phaser-MR (Bunko´czi et al., 2013). Initially, coordinates of putative homologues with PDB codes 2q6u (Carrell et al., 2007) and 2r4j (Yeh et al., 2008), both displaying less than 18% sequence identity to PA4991, were obtained from the PDB. Search models were derived from these coordinates by removing side chains and loop regions differing between the two structures, but no molecular-replacement solutions were found. The amino-acid sequence of PA4991 was then submitted to the Robetta server (http://robetta.bakerlab.org/) that runs the Rosetta software (Kim et al., 2004), resulting in a best model with a score of 0.51. The ﬁve best models from Robetta were superimposed and all residues that showed signiﬁcant deviations between these models were removed. The truncated models were used individually or as a search ensemble in molecular replacement, but none of these protocols yielded a clear solution. A second round of model trimming left only those parts of the best model that had the highest conﬁdence as judged by Robetta, i.e. three strands and two helices of the FAD-binding domain, as a search model (i.e. 60 out of 391 amino-acid residues). research communications Figures were prepared with PyMOL (http://www.pymol.org). 2.3. Crystallization and data collection Crystallization screens (JCSG+ from Qiagen and Wizard I+II from Emerald Bio) were carried out with native and selenomethionine-substituted PA4991 at protein concentra- tions ranging from 11.3 to 19.3 mg ml1 using a Phoenix crystallization robot. Droplets of 100 nl protein solution (10 mM Tris–HCl pH 8.0, 80 mM NaCl) were mixed with the same amount of mother liquor and left to equilibrate at 4 or 20C. One hit for native PA4991 could be reproduced in 24- well format, resulting in crystals that were suitable for data collection. Droplets of 1 ml protein solution were mixed with 1 ml mother liquor (0.2 M trimethylamine N-oxide, 20% PEG 2000 monomethyl ether, 0.1 M Tris–HCl pH 8.5) and equili- brated against 1.0 ml mother liquor at 20C. Acta Cryst. (2016). F72, 105–111 106 Jacewicz et al. PA4991 from Pseudomonas aeruginosa 106 Jacewicz et al. PA4991 from Pseudomonas aeruginosa research communications judged from SDS gels. The UV–visible spectrum of PA4991 shows absorption maxima at 365 and 445 nm characteristic of ﬂavin-binding proteins (Fig. 1). Analytical gel chromato- graphy suggests that PA4991 is a monomer in solution. The protein elutes as a single peak corresponding to a molecular mass of 43.8 kDa when examined on an analytical size- exclusion chromatography column (Fig. 2), which is in good agreement with the mass of 42.2 kDa calculated from the amino-acid sequence. in useful derivatives for ab initio phase determination. We therefore resorted to molecular-replacement approaches, although the search models available in the PDB had a sequence identity of below 18%. Molecular-replacement runs with Phaser-MR (Bunko´czi et al., 2013) using the coordinates of such putatively related structures did not produce inter- pretable electron-density maps. The structure of PA4991 was eventually solved by a combination of molecular replacement using a truncated search model generated from the sequence of PA4991 with Rosetta (Kim et al., 2004) and phase improvement using a low-occupancy heavy-metal derivative obtained by soaking crystals with phenylmercuric chloride. The reﬁned model of PA4991 consists of one polypeptide chain comprising residues 1–391, one FAD molecule, 41 water molecules and two Hg2+ ions bound on opposite sides of the thiol group of Cys245, albeit with low occupancy. Three loop regions, residues 55–67, 106–125 and 248–249, lacked electron density and appeared to be disordered in the crystal, and therefore were not modelled. Crystallographic reﬁnement resulted in a model with statistics as expected for this reso- lution (Table 1). The 2Fo Fc map for the bound cofactor FAD illustrates the quality of the electron-density maps in well ordered regions of the enzyme structure (Fig. 3). judged from SDS gels. The UV–visible spectrum of PA4991 shows absorption maxima at 365 and 445 nm characteristic of ﬂavin-binding proteins (Fig. 1). Analytical gel chromato- graphy suggests that PA4991 is a monomer in solution. The protein elutes as a single peak corresponding to a molecular mass of 43.8 kDa when examined on an analytical size- exclusion chromatography column (Fig. 2), which is in good agreement with the mass of 42.2 kDa calculated from the amino-acid sequence. 3.2. Structure determination Attempts to determine the crystal structure of PA4991 using experimental phasing were not successful. We did not obtain diffracting crystals of the selenomethionine-substituted enzyme and soaking with heavy-metal salts did not result Figure 1 UV–visible absorption spectrum of recombinant PA4991 from P. aeruginosa. The spectrum shows the characteristic maxima for FAD- containing proteins at 365 and 445 nm. The sample consisted of 20 mM PA4991 in in a solution of 25 mM Tris–HCl buffer pH 8.0, 0.15 M NaCl. i 3.1. PA4991 encodes a flavin-containing protein Locus PA4991 had been annotated to encode a ﬂavin- binding protein of 391 amino acids (Winsor et al., 2011). The construct used in this study resulted in soluble, folded protein of yellow colour that could be puriﬁed to homogeneity as Initial attempts to determine the structure of PA4991 using anomalous data collected from crystals of the phenylmercuric chloride derivative were not successful. The program CRANK2 (Skuba´k & Pannu, 2013) located and reﬁned two 107 Jacewicz et al. PA4991 from Pseudomonas aeruginosa Acta Cryst. (2016). F72, 105–111 research communications 3.3. Overall structure PA4991 consists of two domains: the canonical FAD- binding domain of the glutathione reductase family (Dym & Eisenberg, 2001) and a second domain that is most likely involved in substrate binding (Fig. 4). The / FAD domain consists of three chain segments: residues 1–95, 153–227 and 313–391. The domain is dominated by a central six-stranded -sheet with the sixth strand antiparallel to the others. This sheet is ﬂanked by helices 1 and 7 on one side and helix 4 Figure 1 Figure 1 UV–visible absorption spectrum of recombinant PA4991 from P. aeruginosa. The spectrum shows the characteristic maxima for FAD- containing proteins at 365 and 445 nm. The sample consisted of 20 mM PA4991 in in a solution of 25 mM Tris–HCl buffer pH 8.0, 0.15 M NaCl. Figure 3 Simulated-annealing 2Fo Fc OMIT electron-density map of the bound FAD molecule contoured at 1. The map was calculated using PHENIX (Adams et al., 2010). 108 Jacewicz et al. PA4991 from Pseudomonas aeruginosa Acta Cryst. (2016). F72, 105–111 Figure 2 Elution proﬁle of PA4991 from analytical size-exclusion chromatography using a Superdex 200 10/300 column (GE Healthcare) equilibrated with 25 mM Tris–HCl buffer pH 8.0 containing 150 mM NaCl. The column was calibrated with ribonuclease A (13.7 kDa), chymotrypsinogen A (25 kDa), ovalbumin (43 kDa), albumin (67 kDa) and blue dextran (2 MDa). The calibration curve is shown in the inset. Figure 3 Simulated-annealing 2Fo Fc OMIT electron-density map of the bound FAD molecule contoured at 1. The map was calculated using PHENIX (Adams et al., 2010). Figure 2 Elution proﬁle of PA4991 from analytical size-exclusion chromatography using a Superdex 200 10/300 column (GE Healthcare) equilibrated with 25 mM Tris–HCl buffer pH 8.0 containing 150 mM NaCl. The column was calibrated with ribonuclease A (13.7 kDa), chymotrypsinogen A (25 kDa), ovalbumin (43 kDa), albumin (67 kDa) and blue dextran (2 MDa). The calibration curve is shown in the inset. Figure 2 Figure 2 g Elution proﬁle of PA4991 from analytical size-exclusion chromatography using a Superdex 200 10/300 column (GE Healthcare) equilibrated with 25 mM Tris–HCl buffer pH 8.0 containing 150 mM NaCl. The column was calibrated with ribonuclease A (13.7 kDa), chymotrypsinogen A (25 kDa), ovalbumin (43 kDa), albumin (67 kDa) and blue dextran (2 MDa). The calibration curve is shown in the inset. 108 Jacewicz et al. PA4991 from Pseudomonas aeruginosa 3.3. Overall structure g Elution proﬁle of PA4991 from analytical size-exclusion chromatography using a Superdex 200 10/300 column (GE Healthcare) equilibrated with 25 mM Tris–HCl buffer pH 8.0 containing 150 mM NaCl. The column was calibrated with ribonuclease A (13.7 kDa), chymotrypsinogen A (25 kDa), ovalbumin (43 kDa), albumin (67 kDa) and blue dextran (2 MDa). The calibration curve is shown in the inset. Figure 3 Simulated-annealing 2Fo Fc OMIT electron-density map of the bound FAD molecule contoured at 1. The map was calculated using PHENIX (Adams et al., 2010). 108 Jacewicz et al. PA4991 from Pseudomonas aeruginosa 108 Jacewicz et al. PA4991 from Pseudomonas aeruginosa Acta Cryst. (2016). F72, 105–111 108 research communications horikoshii with an r.m.s.d. of 3.3 A˚ , 16% sequence identity and a Z-score of 26.6 (Tsuge et al., 2005), E. coli glycerol-3-phos- phate dehydrogenase with an r.m.s.d. of 3.4 A˚ , 14% sequence identity and a Z-score of 26.4 (Yeh et al., 2008), glycine oxidase from Bacillus subtilis with an r.m.s.d. of 3.5 A˚ , 14% sequence identity and a Z-score of 26.2 (Settembre et al., 2003; Mo¨rtl et al., 2004), the -subunit of sarcosine oxidase from Corynebacterium sp. U-96 with an r.m.s.d. of 3.6 A˚ , 16% sequence identity and a Z-score of 26.1 (Moriguchi et al., 2010) and -glycerophosphate oxidase from Streptococcus sp. with an r.m.s.d. of 3.6 A˚ , 15% sequence identity and a Z-score of 25.3 (Colussi et al., 2008). All of these proteins belong to the GR family of ﬂavoenzymes and share the two-domain core with PA4991. However, glycerol-3-phosphate dehydrogenase and -glycerophosphate oxidase contain additional domains that are not found in PA4991. The quaternary structures of these enzymes cover a wide range of assemblies from mono- meric (glycerol-3-phosphate dehydrogenase) to hetero-octa- meric (l-proline dehydrogenase). on the other. This domain also contains the -meander motif (11, 12, 13) typical of this GR domain, but in PA4991 this sheet contains an additional short -strand comprising resi- dues 5–6, extending this structural feature to a four-stranded mixed -sheet. The second domain consists of peptide segments 96–152 and 223–312. The dominating feature of this domain is a mixed eight-stranded -sheet with topology 7, 8, 6, 16, 17, 18, 15, 19 ﬂanked by two helices, 5 and 6, on one side. 3.6. FAD-binding site FAD is bound noncovalently to PA4991 in an extended conformation in a manner similar to that observed in other members of the GR2 family (Dym & Eisenberg, 2001). Most of the interactions of the cofactor with the protein are through the FAD-binding domain, but the isoalloxazine ring is located at the domain interface and interacts with residues from both domains. The ADP moiety of FAD binds to the classic nucleotide-binding fold, including the diphosphate-binding 3.3. Overall structure Two of these strands, 15 and 19, extend from the FAD domain and are part of the binding site of the isoalloxazine ring of FAD. 3.4. Quaternary structure The asymmetric unit of the PA4991 crystals contains one monomer of the enzyme. An analysis of the crystal packing using PISA (Krissinel & Henrick, 2007) does not show any large protein–protein interfaces that would suggest a stable oligomeric structure. The largest interfaces of 460 and 540 A˚ 2 are more typical of crystal-packing interactions than of protein–protein interfaces in oligomeric proteins (Ponstingl et al., 2000). We conclude that PA4991 is a monomeric protein both in the crystal and in solution (Fig. 2). Jacewicz et al. PA4991 from Pseudomonas aeruginosa research communications ﬁngerprint motif GxGxxG, which in PA4991 is represented by the sequence Gly13-Gly14-Gly15-Ile16-Ala17-Gly18 at the end of strand 2 in the central -sheet of the domain. The diphosphate group points towards the N-terminus of helix 1 and the turn comprising residues 43–48. chain N atom of Leu346, the N3 atom and the main-chain O atom of Ile51, the N5 atom and the main-chain N atom of Gln49, the O2 atom and the main-chain amides of Leu346 and Ala347, and the O4 atom and the main-chain N atom of Ile51. With the exception of the interaction of Glu36 with the ribose ring, there is little sequence conservation in residues involved in FAD binding in other members of this structural family, most likely because the majority of the interactions are through main-chain atoms. The adenine moiety is bound in a pocket lined by Val12, Ser37, Ile177, Ala205, Gly208 and Leu212 and is sandwiched between the side chain of Ser37 and Ala205 (Fig. 5a). The C2 and C3 hydroxyl groups of the ribose ring form hydrogen bonds to a glutamic acid residue, Glu36, characteristic of this fold (Wierenga et al., 1986). The P1 phosphate group is bound to the main-chain N atom of Ser45, the side chains of Ser45 and Gln44 and two ordered water molecules. The P2 phos- phate group interacts with the main-chain N atom of Ala17 and three water molecules (Fig. 5a). Van der Waals inter- actions dominate in the binding of the ribityl moiety, but a hydrogen bond from the C40 hydroxyl group to the side chain of Ser45 also contributes to FAD recognition and binding. 110 Jacewicz et al. PA4991 from Pseudomonas aeruginosa 3.5. Relation to other FAD oxidoreductases A search of the Protein Data Bank for structurally related proteins using DALI (Holm & Rosenstro¨m, 2010) returned a number of FAD-binding proteins, all with a very low degree of sequence identity to PA4991. The top hits were the -subunit of the l-proline dehydrogenase complex from Pyrococcus Figure 4 Overall structure of PA4991. (a) The surface of the FAD-binding domain is shown in light blue and that of the substrate-binding domain in light yellow. Secondary-structure elements are labelled and shown as a cartoon model with red -helices and yellow -strands. Disordered regions are displayed as dotted lines. The FAD molecule is shown in a stick representation. (b) Electrostatic potential surface of PA4991. The FAD-binding pocket and entrance to the active site is highlighted with a yellow arrow. The bound FAD molecule is shown as a stick model. Figure 4 Overall structure of PA4991. (a) The surface of the FAD-binding domain is shown in light blue and that of the substrate-binding domain in light yellow. Secondary-structure elements are labelled and shown as a cartoon model with red -helices and yellow -strands. Disordered regions are displayed as dotted lines. The FAD molecule is shown in a stick representation. (b) Electrostatic potential surface of PA4991. The FAD-binding pocket and entrance to the active site is highlighted with a yellow arrow. The bound FAD molecule is shown as a stick model. Figure 4 Overall structure of PA4991. (a) The surface of the FAD-binding domain is shown in light blue and that of the substrate-binding domain in light yellow. Secondary-structure elements are labelled and shown as a cartoon model with red -helices and yellow -strands. Disordered regions are displayed as dotted lines. The FAD molecule is shown in a stick representation. (b) Electrostatic potential surface of PA4991. The FAD-binding pocket and entrance to the active site is highlighted with a yellow arrow. The bound FAD molecule is shown as a stick model. Jacewicz et al. PA4991 from Pseudomonas aeruginosa 1 109 Acta Cryst. (2016). F72, 105–111 research communications 3.7. Active site and enzyme assays On the re face of the isoalloxazine ring there is an open cleft that is accessible from the solvent. Accessibility of the cofactor in PA4991 appears in part to be controlled by the loop comprising residues 54–67, which is disordered in PA4991. Crystal structures of related enzymes with bound ligands, for instance glycerol-3-phosphate dehydrogenase with bound dihydroxyacetone phosphate (Yeh et al., 2008), show that the corresponding loops fold over the active site and sequester the bound ligand. In the vicinity of the isoalloxazine ring there are the side chains of His53, Arg316 and Lys345 that potentially could be involved in ligand binding and/or catalysis. While in most of the distantly related structural homologues only one or two of these residues are structurally conserved, all three residues are found in the active site of glycerol-3-phosphate dehydrogenase. Arg316 is involved in the binding of the phosphate group of the substrate analogues glyceric acid 2-phosphate and glyceraldehyde-3-phosphate (Yeh et al., The isoalloxazine ring is located in an accessible pocket with the re face of FAD facing the solvent. The electron-density maps do not indicate bending of the isoalloxazine ring as observed in the structure of -glycerophosphate oxidase determined at a similar resolution (Colussi et al., 2008), but are most consistent with a planar conformation of the aromatic ring system. The si face of the isoalloxazine ring is not accessible and packs against Ser48. The ring is anchored to the protein through a series of hydrogen bonds, all involving main- chain atoms of the polypeptide chain (Fig. 5b). Hydrogen bonds are formed between the N1 atom of FAD and the main- Figure 5 FAD–PA4991 interactions. PA4991 C, N and O atoms are coloured yellow, blue and red, respectively. Water molecules are shown as magenta spheres. Hydrogen bonds (3.2 A˚ cutoff) are shown as dotted lines. (a) Binding pocket of the adenosine moiety of FAD. The characteristic ﬁngerprint motif Gly13- Gly14-Gly15-Ile16-Ala17-Gly18 is shown in blue. (b) Surroundings of the isoalloxazine ring of FAD in the active site of PA4991. Fi 5 Figure 5 FAD–PA4991 interactions. PA4991 C, N and O atoms are coloured yellow, blue and red, respectively. Water molecules are shown as magenta spheres. Hydrogen bonds (3.2 A˚ cutoff) are shown as dotted lines. (a) Binding pocket of the adenosine moiety of FAD. The characteristic ﬁngerprint motif Gly13- Gly14-Gly15-Ile16-Ala17-Gly18 is shown in blue. Jacewicz et al. PA4991 from Pseudomonas aeruginosa research communications research communications 2008). The structurally equivalent Arg302 in glycine oxidase participates in the binding of the carboxyl group of the active- site ligands N-acetylglycine and glycolate, respectively (Settembre et al., 2003; Mo¨rtl et al., 2004). Bunko´czi, G., Echols, N., McCoy, A. J., Oeffner, R. D., Adams, P. D. & Read, R. J. (2013). Acta Cryst. D69, 2276–2286. Bunko´czi, G., Echols, N., McCoy, A. J., Oeffner, R. D., Adams, P. D. & Bunkoczi, G., Echols, N., McCoy, A. J., Oeffner, R. D., Adams, P. D. & Read, R. J. (2013). Acta Cryst. D69, 2276–2286. Read, R. J. (2013). Acta Cryst. D69, 2276–2286. Carrell, C. J., Bruckner, R. C., Venci, D., Zhao, G., Jorns, M. S. & Mathews, F. S. (2007). Structure, 15, 928–941. Chen, V. B., Arendall, W. B., Headd, J. J., Keedy, D. A., Immormino, R M K l G J M L W Ri h d J S & Ri h d ( ) The conservation of some of the active-site residues of glycerol-3-phosphate dehydrogenase and glycine oxidase in PA4991 suggested that the enzyme might show corresponding enzymatic activities. However, enzyme assays using glycerol-3- phosphate with NAD+ or NADP+ as co-substrate and dihy- droxyacetone phosphate as substrate and NADH or NADPH as co-substrates, respectively, did not show any enzymatic activity. We also tested 41 putative amines as potential substrates (including the 20 proteinogenic amino acids, d-amino acids, unconventional and modiﬁed amino acids and biogenic amines). No enzymatic activity could be detected in these cases either, suggesting that PA4991 shows a very different substrate speciﬁcity compared with its closest struc- tural homologues. Chen, V. B., Arendall, W. B., Headd, J. J., Keedy, D. A., Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S. & Richardson, D C (2010) Acta Cryst D66 12–21 , , , , , , y, , , R. M., Kapral, G. J., Murray, L. W., Richardson, J. S. & Richardson, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S. & Richardson, D. C. (2010). Acta Cryst. D66, 12–21. D. C. (2010). Acta Cryst. D66, 12–21. Colussi, T., Parsonage, D., Boles, W., Matsuoka, T., Mallett, T. C., ( ) y Colussi, T., Parsonage, D., Boles, W., Matsuoka, T., Mallett, T. C., Karplus, P. A. & Claiborne, A. (2008). Biochemistry, 47, 965–977. Cowtan, K. (2006). Acta Cryst. D62, 1002–1011. ( ) y Dym, O. & Eisenberg, D. (2001). Protein Sci. 10, 1712–1728. research communications Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. (2010). Acta Cryst. D66, 486–501. Holm, L. & Rosenstro¨m, P. (2010). Nucleic Acids Res. 38, W545– W549. Jacobs, M. A., Alwood, A., Thaipisuttikul, I., Spencer, D., Haugen, E., Ernst, S., Will, O., Kaul, R., Raymond, C., Levy, R., Chun-Rong, L., Guenthner, D., Bovee, D., Olson, M. V. & Manoil, C. (2003). Proc. Natl Acad. Sci. USA, 100, 14339–14344. r, K. G. & Snelling, A. M. (2009). J. Hosp. Infect. 73, 3 Kim, D. E., Chivian, D. & Baker, D. (2004). Nucleic Acids Res. 32, W526–W531. Krissinel, E. & Henrick, K. (2007). J. Mol. Biol. 372, 774–797. Leslie, A. G. W. (2006). Acta Cryst. D62, 48–57. 4. Conclusions C., Dorrestein, P. C., Park, J.-H., Augustine, A. M., Rice, L. B. (2008). J. Infect. Dis. 197, 1079 1081. Settembre, E. C., Dorrestein, P. C., Park, J.-H., Augustine, A. M., ( ) Begley, T. P. & Ealick, S. E. (2003). Biochemistry, 42, 2971–2981. Skuba´k, P. & Pannu, N. S. (2013). Nature Commun. 4, 2777. ( ) Su, G., Wei, Y. & Guo, M. (2011). Am. J. Anal. Chem. 2, 879–884. Tsuge, H., Kawakami, R., Sakuraba, H., Ago, H., Miyano, M., Aki, K., Katunuma, N. & Ohshima, T. (2005). J. Biol. Chem. 280, 31045– 31049. Turner, K. H., Wessel, A. K., Palmer, G. C., Murray, J. L. & Whiteley, M. (2015). Proc. Natl Acad. Sci. USA, 112, 4110–4115. Van Duyne, G. D., Standaert, R. F., Karplus, P. A., Schreiber, S. L. & Clardy, J. (1993). J. Mol. Biol. 229, 105–124. 3.7. Active site and enzyme assays (b) Surroundings of the isoalloxazine ring of FAD in the active site of PA4991. Figure 5 FAD–PA4991 interactions. PA4991 C, N and O atoms are coloured yellow, blue and red, respectively. Water molecules are shown as magenta spheres. Hydrogen bonds (3.2 A˚ cutoff) are shown as dotted lines. (a) Binding pocket of the adenosine moiety of FAD. The characteristic ﬁngerprint motif Gly13- Gly14-Gly15-Ile16-Ala17-Gly18 is shown in blue. (b) Surroundings of the isoalloxazine ring of FAD in the active site of PA4991. Acta Cryst. (2016). F72, 105–111 110 References Adams, P. D. et al. (2010). Acta Cryst. D66, 213–221. 4. Conclusions Liberati, N. T., Urbach, J. M., Miyata, S., Lee, D. G., Drenkard, E., Liberati, N. T., Urbach, J. M., Miyata, S., Lee, D. G., Drenkard, E., Wu, G., Villanueva, J., Wei, T. & Ausubel, F. M. (2006). Proc. Natl Acad. Sci. USA, 103, 2833–2838. In the course of the structure determination of PA4991, phasing using heavy-metal derivatives or selenomethionine substitution failed. Initially, molecular replacement also did not provide a solution owing to a lack of suitable templates in the Protein Data Bank. However, a combination of molecular replacement using a truncated model produced by the Rosetta software and a weak heavy-metal derivative provided sufﬁ- cient phase information for structure determination to succeed. This approach might also be applicable in other cases where phase information is lacking. A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D. Storoni, L. C. & Read, R. J. (2007). J. Appl. Cryst. 40, 658–674. y, , , , , , , Storoni, L. C. & Read, R. J. (2007). J. Appl. Cryst. 40, 658–674. Moriguchi, T., Ida, K., Hikima, T., Ueno, G., Yamamoto, M. & Suzuki, H. (2010). J. Biochem. 148, 491–505. Mo¨rtl, M., Diederichs, K., Welte, W., Molla, G., Motteran, L., Andriolo, G., Pilone, M. S. & Pollegioni, L. (2004). J. Biol. Chem. 279, 29718–29727. Moynie, L., Schnell, R., McMahon, S. A., Sandalova, T., Boulkerou, W. A., Schmidberger, J. W., Alphey, M., Cukier, C., Duthie, F., Kopec, J., Liu, H., Jacewicz, A., Hunter, W. N., Naismith, J. H. & Schneider, G. (2013). Acta Cryst. F69, 25–34. Analysis of the structure of PA4991 identiﬁes this enzyme as a member of the GR2 family of ﬂavoenzymes. Enzymatic assays did not support a function as a G3P dehydrogenase/ oxidase or an amino-acid oxidase as hypothesized based on the three-dimensional structure. The conservation of the active-site arginine residue involved in the binding of nega- tively charged groups of the substrate in these enzymes suggests that the unknown substrate of PA4991 might also carry a negatively charged phosphate or carboxyl group. The function of this essential enzyme in the virulence and survival of P. aeruginosa in the host nevertheless remains elusive. ( ) Ponstingl, H., Henrick, K. & Thornton, J. M. (2000). Proteins, 41, 47–57. Rice, L. B. (2008). J. Infect. Dis. 197, 1079–1081. Rice, L. B. (2008). J. Infect. Dis. 197, 1079–1081. Settembre, E. Acknowledgements Wierenga, R. K., Terpstra, P. & Hol, W. G. J. (1986). J. Mol. Biol. 187, 101–107. We acknowledge access to synchrotron radiation at beamline ID14-4, ESRF and thank the beamline staff for support. We also thank the Protein Science Facility at Karolinska Institutet for access to necessary equipment. This work was funded by the European Commission Seventh Framework Programme (FP7/2007–2013; AEROPATH) and the Swedish Science Council. Winn, M. D. et al. (2011). Acta Cryst. D67, 235–242. Winn, M. D., Isupov, M. N. & Murshudov, G. N. (2001). Acta Cryst. D57, 122–133. Winsor, G. L., Lam, D. K. W., Fleming, L., Lo, R., Whiteside, M. D., Yu, N. Y., Hancock, R. E. W. & Brinkman, F. S. L. (2011). Nucleic Acids Res. 39, D596–D600. Yeh, J. I., Chinte, U. & Du, S. (2008). Proc. Natl Acad. Sci. USA, 105, 3280–3285. Zavascki, A. P., Carvalhaes, C. G., Pica˜o, R. C. & Gales, A. C. (2010). Expert Rev. Anti Infect. Ther. 8, 71–93. References Adams, P. D. et al. (2010). Acta Cryst. D66, 213–221. References p f Zhang, K. Y. J., Cowtan, K. & Main, P. (1997). Methods Enzymol. 277, 53–64. 111 Acta Cryst. (2016). F72, 105–111 Jacewicz et al. PA4991 from Pseudomonas aeruginosa 11
https://openalex.org/W4231572654	https://www.qeios.com/read/668828/pdf	English	null	Bias	Definitions	2,019	cc-by	68	Bias John P. A. Ioannidis Qeios · Definition, August 9, 2019 Open Peer Review on Qeios Qeios ID: 668828 · https://doi.org/10.32388/668828 Source John P. A. Ioannidis. (2005). Why Most Published Research Findings Are False. PLoS Med, vol. 2 (8), e124. The combination of various design, data, analysis, and presentation factors that tend to produce research findings when they should not be produced. Qeios ID: 668828 · https://doi.org/10.32388/668828 1/1
https://openalex.org/W4361234389	https://figshare.com/articles/journal_contribution/Table_S4_from_microRNA-1246_Is_an_Exosomal_Biomarker_for_Aggressive_Prostate_Cancer/22417977/1/files/39864090.pdf	English	null	Table S5 from microRNA-1246 Is an Exosomal Biomarker for Aggressive Prostate Cancer	null	2,023	cc-by	686	Table S4 Clinicopathologic characteristics of prostate cancer patients Table S4 Clinicopathologic characteristics of prostate cancer patients A. Clinicopathologic characteristics of training cohort of PCa patients used for ex-miR- 1246 profiling Cohort 1 (N=44) Cohort 2 (N=25) Characteristic N (%) N (%) Age, Years Mean 64.3 61.2 Median 65.5 62 Range 46-78 42-76 T-stage pT2 4 (9) 16 (64) pT3 3 (7) 8 (32) pT4 36 (84) - Gleason Score 4-6 - 1 (4) 7 (3+4) 12 (28) 18 (72) 7 (4+3) 9 (21) 2 (8) 8-10 19 (44) 4 (16) PSA Median 7.05 (2.6-54) 5.6 (3-52) N-stage N0/NX 17 (39) 25 (100) N1 26 (59) - M-stage M0/MX 44 (100) 25 (100) M1 - - A. Clinicopathologic characteristics of training cohort of PCa patients used for ex-miR- 1246 profiling ologic characteristics of training cohort of PCa patients used for ex-miR- ng ologic characteristics of training cohort of PCa patients used for ex-miR- 44 (100) 25 (100) Pathological diagnosis Adenocarcinoma 44 (100) 25 (100) Pathological stage and N-stage unknown for 1 case, Gleason score for 4 cases, PSA for 3 cases Pathological diagnosis 25 (100) Adenocarcinoma * Pathological stage and N-stage unknown for 1 case, Gleason score for 4 cases, PSA for 3 cases * Pathological stage and N-stage unknown for 1 case, Gleason score for 4 cases, PSA for 3 cases B. Clinicopathologic characteristics of validation cohort of PCa patients used for ex-miR- 1246 profiling (N=43) B. Clinicopathologic characteristics of validation cohort of PCa patients used for ex-miR- 1246 profiling (N=43) Characteristic N(%) Age, Years Mean 66 Median 65 Range 48-80 Gleason Score 4-6 4 (9) 7 (3+4) 9 (21) 7 (4+3) 10 (23) 8-10 16 (37) PSA Median 13 (3.3-512.6) M-stage M1 43 (100) Pathological diagnosis Adenocarcinoma 43 (100) * Age unknown for 1 case, Gleason score for 4 cases, PSA for 8 cases B. Clinicopathologic characteristics of validation cohort of PCa patients used for ex-miR- 1246 profiling (N=43) B. Clinicopathologic characteristics of validation cohort of PCa patients used for ex-miR- 1246 profiling (N=43) * Age unknown for 1 case, Gleason score for 4 cases, PSA for 8 cases C. Clinicopathologic characteristics of prostate cancer patients used for miR-1246 profiling in clinical tissues Characteristic N (%) A Y C. Clinicopathologic characteristics of prostate canc in clinical tissues Characteristic N (%) Age, Years Mean 63.9 Median 63.5 Range 52-76 T-stage pT2 or lower 16 (44) pT3 16 (44) pT4 3 (8) Gleason Score 4-6 8 (22) 7 (3+4) 14 (39) 7 (4+3) 6 (17) 8-10 6 (17) PSA Median 7.55 (<0.1-61.1) N-stage N0/NX 3 (8) N1 M-stage M0/MX 36 (100) M1 - Pathological diagnosis Adenocarcinoma 36 (100) * Pathological stage unknown for 1 case, Gleason scor C. Clinicopathologic characteristics of prostate cancer patients used for miR-1246 profiling in clinical tissues Characteristic N (%) Age, Years Mean 63.9 Median 63.5 Range 52-76 T-stage pT2 or lower 16 (44) pT3 16 (44) pT4 3 (8) Gleason Score 4-6 8 (22) 7 (3+4) 14 (39) 7 (4+3) 6 (17) 8-10 6 (17) PSA Median 7.55 (<0.1-61.1) N-stage N0/NX 3 (8) N1 33 (92) M-stage M0/MX 36 (100) M1 - Pathological diagnosis Adenocarcinoma 36 (100) * Pathological stage unknown for 1 case, Gleason score for 2 cases, PSA for 1 case C. Pathological diagnosis Clinicopathologic characteristics of prostate cancer patients used for miR-1246 profiling in clinical tissues Characteristic N (%) Age, Years Mean 63.9 Median 63.5 Range 52-76 T-stage pT2 or lower 16 (44) pT3 16 (44) pT4 3 (8) Gleason Score 4-6 8 (22) 7 (3+4) 14 (39) 7 (4+3) 6 (17) 8-10 6 (17) PSA Median 7.55 (<0.1-61.1) N-stage N0/NX 3 (8) N1 33 (92) M-stage C. Clinicopathologic characteristics of prostate cancer patients used for miR-1246 profiling in clinical tissues hologic characteristics of prostate cancer patients used for miR-1246 profiling ues C. Clinicopathologic characteristics of prostate cancer patients used for miR-1246 profiling in clinical tissues 3 (8) 33 (92) Pathological diagnosis 36 (100) 36 (100) * Pathological stage unknown for 1 case, Gleason score for 2 cases, PSA for 1 case * Pathological stage unknown for 1 case, Gleason score for 2 cases, PSA for 1 case
https://openalex.org/W3205202086	https://www.nature.com/articles/s41599-021-00910-x.pdf	English	null	Big data for whose sake? Governing migration through artificial intelligence	Humanities & social sciences communications	2,021	cc-by	4,457	COMMENT Big data for whose sake? Governing migration through artiﬁcial intelligence T b Bi 1✉& E E K k 2 https://doi.org/10.1057/s41599-021-00910-x OPEN COMMENT Big data for whose sake? Governing migration through artiﬁcial intelligence T b Bi 1✉& E E K k 2 https://doi.org/10.1057/s41599-021-00910-x OPEN AI at the border The agency employed a suite of AI technologies to streamline administrative tasks like visa applica- tions and processing travel documents, identity cards, and work permits. In mid-2017, the United Nations initiated the Unite Ideas Internal Displacement Event Tagging and Extraction Clustering Tool challenge. The winner was the Data for Democracy team, which built a tool capable of tracking and analysing refugees and other people forced to ﬂee or evacuate their homes. Given the apparent power imbalances, technology remains an arena of political struggle. The same technologies can be used to promote peace and spark conﬂict. For this reason, the political struggle to hold corporations and governments accountable can still change the course of technological development. For instance, anxiety over fake news and manipulative campaigns undermining democracy and the trustworthiness of elections is spurring many NGOs and political movements to demand that digital campaigns observe democratic principles and to hold political parties’ feet to the ﬁre for their social media advertising. In sum, various means to challenge the dystopian vision of the future of humanity remain. Many organisations in democratic societies retain in their grasp the tools to hold governments and technology companies to account for the direction that techno- logical change takes. Even so, migrants and refugees lack the same access citizens have to the arena of political contestation to protect and advance their interests. For anyone wishing to observe the likely shape of a dystopian future of a tech-oriented society in which the people have ceded their political autonomy, a close look at the daily experiences of migrants and refugees vis-à- vis the large tech providers offers sufﬁcient clues. Given the apparent power imbalances, technology remains an arena of political struggle. The same technologies can be used to promote peace and spark conﬂict. For this reason, the political struggle to hold corporations and governments accountable can still change the course of technological development. For instance, anxiety over fake news and manipulative campaigns undermining democracy and the trustworthiness of elections is spurring many NGOs and political movements to demand that digital campaigns observe democratic principles and to hold political parties’ feet to the ﬁre for their social media advertising. The European Asylum Support Ofﬁce (EASO) uses machine learning to predict pressures on the asylum administrations of member states of the European Union and associated countries (i.e., Norway and Switzerland). AI at the border Moreover, given the new knowledge generated from migrants’ data has the potential to be exploited by states and corporations alike, migration scholars must not lose sight of the principles of ethical research and should retain a critical approach to the practices of these powerful actors. Governments are investing vast sums in developing new tech- nologies, and tech companies are at the forefront of commercia- lising various tech-based solutions to meet consumer demand. However, it is not difﬁcult to observe growing concerns about the challenges arising from these new frontier technologies. The gen- eral optimism in the ﬁrst decade of the twenty-ﬁrst century has turned into a more pessimistic outlook due to the actual and potential outcomes that arise when corporations and governments deploy these new technologies. Today, ‘Big Tech’ increasingly serves as a synonym for ‘Big Brother’, and these ﬁrms stand accused of being monopolistic predators hungry for ever more consumer data. The result is ‘surveillance capitalism’, in which consumers divulge their most intimate selves in exchange for convenience and the tantalising prospect of previously undreamed- of goods and services (Zuboff, 2019). Therefore, such a critical change in the public mood around these technologies has shifted the frame from a utopian vision of human progress to a picture of an emerging dystopian tomorrow in which humankind is sub- sumed into the digital realm. Technology offers even more ways to manipulate and control the citizen-as-consumer with the help of non-human ‘smart products’ (Meier, 2011). Thus, it has moved to the centre of the debate over the growing power of the dominant forces in society, supplementing and indeed catalysing their existing tools of hegemony, control, and oppression. h b l h l Big Data projects in migration/border management can be divided into two main categories. In the ﬁrst category, the main aim is to analyse Big Data to provide faster and more efﬁcient services. Projects in the second category are piloting Big Data analysis for speculative targets, such as AI lie detection or auto- mating asylum or visa decision-making processes. Distinguishing these two categories is signiﬁcant as they pose different kinds of problems and require different regulatory tools. One of the ﬁrst AI applications in the ﬁrst category was initiated in 2007 by the Hong Kong Immigration Department as a part of the eBrain project. AI at the border Hope-mA The study of human migration has recently taken a giant step forward through the use of Big Data and AI. Numerous states and international organisations involved in migration management have turned to these new technologies to enhance their ability to govern border spaces and control movement across them. The reasoning has centred on the apparent payoffs of increased efﬁ- ciency, improved preparedness, and enhanced border security. Speciﬁc applications that have been touted include the ability to predict population movements, process administration, and border checks and prevent unauthorised migration. Moreover, automated decision-making in visa and asylum applications and AI lie detectors for immigrants and refugees encroach on human rights owing to limited or no international regulation. These aspects spearhead the controversial discussions. A lthough human activity constantly generates massive amounts of data, these data can only be analysed by mainly the private sector and governmental institutes due to data accessibility restrictions. Big Data analytics and Artiﬁcial Intelligence (AI) technologies are promoted as cutting-edge solutions to ongoing and emerging social, economic and gov- ernance challenges. Such analysis serves several practical pur- poses, including marketing goods and services, ensuring the success of political campaigns, and identifying criminal activities. New methods of harnessing Big Data to serve the public good are promising but as yet not widespread. While recent developments in frontier technologies have brought many opportunities, they also raise signiﬁcant concerns and challenges for societies. Giant steps in AI, blockchain, robotics and Big Data have had tremendous impacts on the socio- economic development of countries (Palvia et al., 2018) as well as on the daily lives of people. However, they have also destabilised traditional approaches to work and life, not to mention indivi- duals’ so-called ‘algorithmic identities’ (Cheney-Lippold, 2011) and their sense of self and the future (Floridi, 2007). Big Data analysis is also providing social scientists with new insights in the ﬁeld of migration research. Such analysis enables the study of temporary or circular migration patterns and real- time monitoring of public opinion and media discourse on migration. It also sheds much-needed light on aspects of migra- tion we currently have limited knowledge about, such as the integration prospects of recently arrived migrants and the pos- sible future migration patterns. Hope-mongering AI: The promise and perils Hope-mongering AI: The promise and perils Big data for whose sake? Governing migration through artiﬁcial intelligence Tuba Bircan 1✉& Emre Eren Korkmaz 2 Although human activity constantly generates massive amounts of data, these data can only be analysed by mainly the private sector and governmental institutes due to data accessibility restrictions. However, neither migrants (as the producers of this data) nor migration scholars (as scientiﬁc experts on the topic) are in a position to monitor or control how governments and corporations use such data. Big Data analytics and Artiﬁcial Intelligence (AI) technologies are promoted as cutting-edge solutions to ongoing and emerging social, economic and governance challenges. Meanwhile, states increasingly rely on digital and frontier technologies to manage borders and control migratory movements, and the defence industry and military–intelligence sectors provide high-tech tools to support these efforts. Worryingly, during the design and testing of algorithmic tools, migrants are often portrayed as a security threat instead of human beings with fundamental rights and liberties. Thus, privacy, data protection, and conﬁdentiality issues continue to pose risks and challenges to migrant communities and raise important questions for the public and decision-makers alike. This comment seeks to shed light on the lack of effective regulation of AI and Big Data as they are applied in migration ‘management’. Additionally, from the perspective of privacy issues and immigrant rights (seeking asylum as a human right, it aims at advocating improved access to Big Data for scientiﬁc research which might act as a social control function for the smart border and existing/ongoing migration governance practices of countries. We argue that the use of Big Data and AI for migration governance requires much better collaboration between migrants (including the civil society and grassroots organisations solidarity that represent them), data scientists, migration scholars and policymakers if the potential of these technologies is to be reached in a way that is reasonable and ethical. Numerous critical privacy questions arise are regarding the legal requirements, conﬁdentiality, and rules of engagement as well as the ethical concerns of (mis)use of new technologies. When the secretive nature of the ongoing exploitation of migrant data by states and corporations is considered raising such questions is essential for progress. 1 Vrije Universiteit Brussel, Brussels, Belgium. 2 Oxford University, Oxford, UK. ✉email: tuba.bircan@vub.be 1 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| (2021) 8:241 \| https://doi.org/10.1057/s41599-021-00910-x COMMENT HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| (2021) 8:241 \| https://doi.org/10.1057/s41599-021-00910-x AI at the border To do so, EASO builds on three types of data collected on past events—namely, data from tradi- tional countries of origin and transit (including social media monitoring), data on pressures at the EU’s external borders, and data on the outcomes of previous asylum applications in the EU. EASO’s algorithm predicts pressures up to four weeks in advance and suggests possible future medium-term scenarios using his- torical and current data.1 The European Data Protection Super- visor (EDPS) offered a sharp critique of the EASO’s social media monitoring of migrants and refugees:2 ‘Social media users mon- itoring is a personal data processing activity that puts individuals’ rights and freedoms at signiﬁcant risk. It involves uses of personal data that go against or beyond individuals’ reasonable expecta- tions. Such uses often result in personal data being used beyond their initial purpose, their initial context and in ways the indi- vidual could not reasonably anticipate’. The EDPS underscored the important principles of ‘purpose limitation’ and ‘data In sum, various means to challenge the dystopian vision of the future of humanity remain. Many organisations in democratic societies retain in their grasp the tools to hold governments and technology companies to account for the direction that techno- logical change takes. Even so, migrants and refugees lack the same access citizens have to the arena of political contestation to protect and advance their interests. For anyone wishing to observe the likely shape of a dystopian future of a tech-oriented society in which the people have ceded their political autonomy, a close look at the daily experiences of migrants and refugees vis-à- vis the large tech providers offers sufﬁcient clues. HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| (2021) 8:241 \| https://doi.org/10.1057/s41599-021-00910-x 2 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| https://doi.org/10.1057/s41599-021-00910-x COMMENT minimisation’, whereby personal data should be collected only for ‘speciﬁed, explicit and legitimate purposes. The EDPS’ warning also dovetails with the aims of this comment because as the public discusses the data mining and surveillance techniques to predict, manage and stop migratory movements, such legal and social pressure over ofﬁcial institutions and corporations will likely increase. information is requested are examples of such ‘consent’. More- over, migrants and refugees may ﬁnd themselves at the whim of politics through mechanisms like ‘function creep’, where the use of technology and information extends beyond the stated initial purpose. Big tech and migrants I l b l g g In recent years, global technology companies have also begun to pay more attention to migratory movements. Various tech cor- porations offer services and products to UN agencies, NGOs, states and migrants/refugees on a commercial basis (i.e., not as part of corporate social responsibility projects). Financial insti- tutions, telecom companies, mobile phone operators and tech ﬁrms seek to proﬁt from their investments. Thus, serving refugees and displaced populations has increasingly become a proﬁtable business. As a result of their technical superiority and ﬁnancial power, commercial actors act as suppliers to UN agencies and NGOs and introduce themselves as critical stakeholders, actively working with states and the UN agencies to manage and control migratory movements. However, neither migrants (as the pro- ducers of this data) nor migration scholars (as scientiﬁc experts on the topic) are in a position to monitor or control how gov- ernments and corporations use such data. The European Commission funded projects using AI to develop a lie-detection system to ramp up security at European borders (iBorderCtrl from 2016–2019)3 and create autonomous border surveillance systems (ROBORDER from 2017 to 2021).4 Moreover, in 2018, the mandate and operation areas of the European Union Agency for the Operational Management of Large-Scale IT Systems in the Area of Freedom, Security and Justice (eu-LISA) were expanded5 with a particular focus on the implementation of the EU’s asylum, border management and migration policies. The values of eu-LISA’s investments are identiﬁed as accountability, transparency, excellence, continuity, teamwork, and customer focus, which privilege the ‘customer’ (i.e., states and/or agencies) rather than migrants and fall well short on privacy and human rights aspects. Corporate involvement in migration, displacement and refugee issues is not limited to humanitarian emergencies, and corpora- tions’ keen interest in border management is a very concerning development. States increasingly rely on digital and frontier technologies to manage borders, and the defence industry and military–intelligence sector provide high-tech tools for this pur- pose. For instance, Europe’s largest arms sellers also market ‘smart’ border management tools (Akkerman, 2016). The ‘smart border’ constructed by the US along the southern frontier with Mexico and the EU’s digitalised border management systems (including Eurosur6 and Eurodac7) are examples of such tech- nology in action. However, during the design and testing of algorithmic tools, migrants are often portrayed as a security threat instead of human beings bearing fundamental rights and liberties. AI at the border Biometric surveillance and processing in public spaces (and advanced analyses in social media data) are other examples where individuals may be at risk but are given no information or insight into the collection or analysis of personal data. The current Afghan crisis offers a clear illustration of the risks and challenges here. On the one hand, we see evidence of the expanded deployment of high-tech surveillance systems by the EU and national authorities at external borders in anticipation of new migration waves, raising questions about the ethical implications of these technologies. On the other, the Taliban’s potential access to the biometric data of Afghan refugees registered by humani- tarian or military agencies calls urgent attention to the risks of ‘function creep’ in the uncertain ﬁeld of migration governance. Such Big Data monitoring projects are not only deployed by EU agencies. Similar approaches have been observed in border security and migration management projects in the US and Canada. For instance, US Customs and Border Protection (CBP) and Immigration and Customs Enforcement (ICE) buy com- mercial databases to track mobile phones to identify undocu- mented immigrants. It is also reported that from 2010 to 2014, CBP spent about $2.5 million to purchase cell-site simulator technology to develop fake towers that can detect and intercept mobile phone text and voice messages and pull the location and other information from mobile devices that are trying to connect to it (Ghaffary, 2020; Riotta, 2020). Also, as Akhmetova (2020) notes, ‘In 2018, the Canada Border Services Agency used private third-party DNA services such as Ancestry.com to establish the nationality of individuals subject to potential deportation. This is deeply concerning because one’s DNA is not related to immi- gration journey, ‘legality’, nationality and should bear no impact on one’s immigration/asylum applications. Another issue is the coercive nature of privacy invasion—individuals who gave their DNA samples to these companies might not have given consent nor knew that their data could be used by governments to assess immigration applications’. So far, no satisfactory response has been offered to these ser- ious political, ethical, and social questions. That this is the case reﬂects the minimal representation and bargaining power—and thus political muscle—of people on the move. AI at the border Although citizens can be victimised by the development of new technologies—for instance, through job losses to automation or the diminution of the public sphere—they still have ways to raise their concerns and apply pressure to governments and corporations. Refugees and undocumented migrants lack these opportunities. As commen- tators (Bigo, 2002; Maxmen, 2019) have pointedly asked: Who will defend migrants/refugees against the power of governments and corporations, whose capabilities for surveillance and oppression have expanded signiﬁcantly through the new technologies? Issues also arise with the second category, speculative Big Data pilot projects. For example, in 2018, Canada’s Immigration Department ran a pilot project (Artiﬁcial Intelligence Solution) to assess AI-supported decision-making for immigration and asy- lum applications. Agencies within the US Department of Homeland Security also use new technologies to automate migration-related and asylum-related decisions. However, human rights experts have criticised this approach for using vulnerable migrants and asylum-seekers as experimental subjects to train AI algorithms (Molnar, 2019). HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| (2021) 8:241 \| https://doi.org/10.1057/s41599-021-00910-x An over-reliance on AI for migration governance Implementing technological developments such as AI algorithms for migration management rests on large volumes of data from various sources; machine learning is one of the fundamental steps in improving these algorithms. This is why AI applications are constantly in need of ever more data. Beduschi (2020) identiﬁes three major challenges regarding the quality of the data used to train algorithms: (1) migrants’ data privacy, (2) algorithmic accountability and (3) fairness. The challenges that attend the rise of AI should not be overlooked. Moreover, researchers and pol- icymakers must avoid being distracted by the hype surrounding AI and focus on ensuring that comprehensive regulations are developed to protect the common good. In sum, the ‘inﬂux’ of Big Data and AI applications to address societal challenges raises profound questions about the fair dis- tribution of beneﬁts arising from these approaches. Digital tech- nologies and AI have the power to inﬂuence and shape democracy, as the recent misinformation campaigns in the US and many other countries lay bare. Considering the sensitivity of migration in many Western democracies, the risks of mis- information and disinformation make it all the more important that the scientiﬁc, political, and public discourse is as transparent as possible, especially since the stakeholder groups involved are often very opaque. Molnar (2019) argues that states deliberately lean into the lack of regulation in this space because it renders migrants more trackable and intelligible. Against this backdrop, the use of advanced tech- nology for surveillance and data collection are justiﬁed on national security grounds or the apparent ‘objectivity’ of data-driven pol- icymaking. or even—somewhat improbably—on humanitarian or development grounds. Concerns arise as to whether these ‘solu- tions’ are being tested on refugees and migrants because they can scarcely object and often lack even basic knowledge about what is involved. Nevertheless, as mentioned, improving algorithms requires vast amounts of data from a variety of sources and gov- ernments and corporations have increasingly turned to conﬂict zones and refugee camps as experimental ﬁelds. As a result, tech- nology ﬁrms are seeking to market their platforms to the EU and national governments as ‘migration prediction systems’ in order to boost sales and proﬁts (Taylor and Meissner, 2020). Big tech and migrants I l b l Thus, issues surrounding privacy, data protection and UN agencies and NGOs also collect data from migrants and refugees. The data is necessary to deliver services and make long- term plans. However, concerns arise about how such data is collected, where it is stored, whether and how people in vital need of assistance give their consent, and what measures are taken to prevent people’s data from being used against them. Moreover, the challenges go beyond discussions of consent since user priv- acy and secondary use are central issues as well. y y To illustrate, migrants and refugees often ﬁnd themselves providing ‘informed consent’ to primary data collection even though they often have insufﬁcient understanding of the impli- cations or potential threats. Visa and asylum applications and getting access to aid at refugee shelters where biometric 3 COMMENT HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| https://doi.org/10.1057/s41599-021-00910-x conﬁdentiality continue to pose risks and challenges to migrant communities. ‘preventing’ it, who beneﬁts the most from the AI and high technology utilisation for managing migration remains an open question. A central issue concerns who is absent from the decision-making table—namely, scholars and data scientists but also civil society and migrants themselves. To be able to overcome the existing challenges and be prepared for future complications in global migration governance, the discussions should rely on ‘human rights’ and there is a need for clear identiﬁcation for the stakeholders who will closely be involved with decisions and act as a controlling body. Notes 1 See European Migration Network, 2021, “Accurate, timely, interoperable? Data management in the asylum procedure”. https://ec.europa.eu/home-affairs/sites/ homeaffairs/ﬁles/00_eu_emn_study_data_management_ﬁnal_common_ template_ﬁnal_en.pdf 1 See European Migration Network, 2021, “Accurate, timely, interoperable? Data management in the asylum procedure”. https://ec.europa.eu/home-affairs/sites/ homeaffairs/ﬁles/00_eu_emn_study_data_management_ﬁnal_common_ template_ﬁnal_en.pdf 2 https://edps.europa.eu/sites/default/ﬁles/publication/19-11-12_reply_easo_ssm_ﬁnal_ reply_en.pdf 2 https://edps.europa.eu/sites/default/ﬁles/publication/19-11-12_reply_easo_ssm_ﬁnal_ reply_en.pdf 3 https://cordis.europa.eu/project/id/700626. 4 https://roborder.eu/the-project/aims-objectives/. 5 Regulation (EU) 2018/1726. 5 Regulation (EU) 2018/1726. 6 https://ec.europa.eu/home-affairs/policies/schengen-borders-and-visa/border- crossing/eurosur_en 6 https://ec.europa.eu/home-affairs/policies/schengen-borders-and-visa/border- crossing/eurosur_en 6 https://ec.europa.eu/home-affairs/policies/schengen-borders-and-visa/border- crossing/eurosur_en 7 https://www.eulisa.europa.eu/Activities/Large-Scale-It-Systems/Eurodac 7 https://www.eulisa.europa.eu/Activities/Large-Scale-It-Systems/Eurodac Issues surrounding privacy, data protection, and conﬁdentiality continue to pose risks and challenges to migrants. There is no effective auditing mechanism to ensure governments and cor- porations are accountable for using migrants’ data. In addition, several critical ethical questions arise about the legal require- ments, conﬁdentiality, and rules of engagement. These questions mostly refer to the possible (mis)use of new technologies for more conservative and preventive migration policies that violate human rights. An additional concern is the role of tech corporations as contractors for states and institutions, where data gathering for commercial purposes is combined with contracted analysis on these datasets. In other words, given the secretive way in which states and corporations exploit migrant data, which points out to a growing problem related to the exploitation of data for pre- ventive and restrictive migration governance. The use of Big Data and AI for migration governance requires much better colla- boration between migrants (including civil society and grassroots organisations that represent them), data scientists, migration scholars and policymakers if the potential of these technologies is to be reached in a way that is reasonable and ethical. An over-reliance on AI for migration governance Against this backdrop, informed consent and vulnerable groups’ ‘right to refuse’ intrusive data-gathering techniques are increasingly overlooked in the race for ever-more efﬁcient monitoring and ‘service delivery’ (as the infamous case of iris scanning of refugees in camps in Jordan bears out). Received: 29 December 2020; Accepted: 27 September 2021; Received: 29 December 2020; Accepted: 27 September 2021; Received: 29 December 2020; Accepted: 27 September 2021; HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| https://doi.org/10.1057/s41599-021-00910-x Zuboff S (2019) The Age of surveillance capitalism: the ﬁght for a human future at the new frontier of power. Hatchett, New York 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/. Competing interests p g The authors declare no competing interests. References Akhmetova R (2020) Efﬁcient discrimination: on how governments use artiﬁcial intelligence in the immigration sphere to create and fortify ‘invisible border walls’. University of Oxford, COMPAS Working Paper 20–149 y g p Akkerman M (2016) Border wars: the arms dealers proﬁting from Europe’s refugee tragedy. Transnational Institute and Stop Wapenhandel Akkerman M (2016) Border wars: the arms dealers proﬁting fro Beduschi A (2020) International migration management in the age of artiﬁcial intelligence. Migr Stud. https://doi.org/10.1093/migration/mnaa003 Bigo D (2002) Security and immigration: toward a critique of the governmentality of unease. Alternatives 27(1):63–92 Cheney-Lippold J (2011) A new algorithmic identity: soft biopolitics and the modulation of control. Theory Cult Soc 28(6):164–81 Floridi L (2007) A look into the future impact of ICT on our lives. Inf Soc 23(1):59–64 Ghaffary S (2020). The “smarter” wall: how drones, sensors, and AI are patrolling the border. Vox Media. https://www.vox.com/recode/2019/5/16/18511583/ smart-border-wall-drones-sensors-ai Maxmen A (2019) Can tracking people through phone-call data improve lives? Nature 569(7758):614–7 Meier P (2011) Do “Liberation technologies” change the balance of power between repressive states and civil society? Unpublished PhD thesis, Faculty of the Fletcher School of Law and Diplomacy Molnar P (2019) Technology on the margins: AI and global migration manage- ment from a human rights perspective. Camb Int Law J 8(2):305–30 Most discussions on politics, power and AI are in-depth but not widely spread. The applications in migration governance are often built on the challenging trade-offs for societal beneﬁts. Responsibility and accountability of the actors in AI applications linked to decision-making mechanisms are obscurely deﬁned. Hence, who can challenge the automated decisions for migration management is shrouded in mystery. Bearing in mind that the notion of ‘managing’ migration is typically a euphemism for Palvia P, Baqir N, Nemati H (2018) ICT for socio-economic development: a citi- zens’ perspective. Inf Manag 55(2):160–76 Riotta C (2020) Trump administration bought access to cell phone database to target immigrants, report says. Independent. https://www.independent.co.uk/ news/world/americas/trump-ice-immigration-cellphone-location-data-report- surveillance-tracking-a9324221.html Taylor L, Meissner F (2020) A crisis of opportunity: market-making, big data, and the consolidation of migration as risk. Antipode 52(1):270–90 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| (2021) 8:241 \| https://doi.org/10.1057/s41599-021-00910-x 4 COMMENT COMMENT Additional information Correspondence and requests for materials should be addressed to Tuba Bircan. Reprints and permission information is available at http://www.nature.com/reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. © The Author(s) 2021 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS \| (2021) 8:241 \| https://doi.org/10.1057/s41599-021-00910-x 5
https://openalex.org/W4249340617	https://www.qeios.com/read/G4EMS3/pdf	English	null	Renal Pelvis and Ureter Cancer Pathologic Regional Lymph Nodes TNM Finding v8	Definitions	2,020	cc-by	109	Qeios · Definition, February 8, 2020 Open Peer Review on Qeios Renal Pelvis and Ureter Cancer Pathologic Regional Lymph Nodes TNM Finding v8 National Cancer Institute Qeios ID: G4EMS3 · https://doi.org/10.32388/G4EMS3 Source National Cancer Institute. Renal Pelvis and Ureter Cancer Pathologic Regional Lymph Nodes TNM Finding v8. NCI Thesaurus. Code C140348. National Cancer Institute. Renal Pelvis and Ureter Cancer Pathologic Regional Lymph Nodes TNM Finding v8. NCI Thesaurus. Code C140348. A pathologic finding about one or more characteristics of renal pelvis and ureter cancer, following the rules of the TNM AJCC v8 classification system as they pertain to staging of regional lymph nodes. Qeios ID: G4EMS3 · https://doi.org/10.32388/G4EMS3 1/1
https://openalex.org/W2042124187	https://bmcinfectdis.biomedcentral.com/counter/pdf/10.1186/1471-2334-14-6	English	null	Screening of post-mortem tissue donors for Coxiella burnetii infection after large outbreaks of Q fever in The Netherlands	BMC infectious diseases	2,014	cc-by	6,432	Abstract Background: After the largest outbreaks of Q fever ever recorded in history occurred in the Netherlands, concern arose that Coxiella may be transmitted via donated tissues of latent or chronically infected donors. The Dutch Health Council recently advised to screen tissue donors, donating high risk tissues, for Coxiella infection. Methods: After validation of an enzyme immunoassay (EIA) test for IgG antibodies against phase 2 of C. burnetii for use on post-mortem samples, serum samples of 1033 consecutive Dutch post-mortem tissue donors were tested for IgG antibodies against phase 2 of C. burnetii. Confirmation of reactive results was done by immunofluorescence assay (IFA). All available tissues (corneas, heart valves, skin and bone marrow) from donors with IgG reactivity were tested for presence of Coxiella DNA by PCR. Risk factors for IgG reactivity were investigated. Results: After validation of the tests for use on post-mortem samples, 50/1033 donors (4.8%) screened positive for phase 2 anti-Coxiella IgG by EIA, and 31 were confirmed by IFA (3.0%). One donor showed a serological profile compatible with chronic infection. All tested tissues (25 corneas, 6 heart valves, 4 skin and 3 bone marrow) from donors with IgG reactivity tested negative for the presence of Coxiella DNA. Except for living in a postal code area with a high number of Q fever notifications, no risk factors for IgG reactivity were found. Conclusions: The strong correlation between notifications and seroprevalence confirms that the used assays are sufficiently specific for use on post-mortem samples, although one has to be aware of differences between batches. Thus, this study provides a validated method for screening tissue donors for infection with Coxiella burnetii that can be used in future outbreaks. ella burnetii, Q fever, Tissue transplantation, Serological screening, Zoonotic infections, Outbreaks Keywords: Coxiella burnetii, Q fever, Tissue transplantation, Serological screening, Zoonotic infectio © 2014 van Wijk 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. van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 Screening of post-mortem tissue donors for Coxiella burnetii infection after large outbreaks of Q fever in The Netherlands Marja J van Wijk1, D Willemijn Maas1, Nicole HM Renders2, Mirjam HA Hermans2, Hans L Zaaijer3,4 and Boris M Hogema3 arja J van Wijk1, D Willemijn Maas1, Nicole HM Renders2, Mirjam HA Hermans2, Hans L Zaaijer3, d Boris M Hogema3 * Correspondence: m.vanwijk@bislife.org 1BISLIFE Foundation, PO Box 309, 2333BD Leiden, The Netherlands Full list of author information is available at the end of the article Background farms, the number of reported cases declined in 2010 and returned to less than 100 cases per year in 2011. Q fever is a zoonotic disease caused by infection with the bacterium Coxiella burnetii. Airborne transmission from infected goats and sheep is the principle mode of transmission to humans. In the Netherlands Q fever out- breaks started in 2007 and increased in the subsequent years. Over 4000 cases and 25 casualties were reported between 2007 and 2010, making it the largest Q fever epidemic ever reported. After various measures were taken, including culling of pregnant goats at infected Most Coxiella infections pass asymptomatic (± 60%), but mild flu-like illness or severe disease with pneumo- nia or hepatitis occurs. Current data suggest that 1.5 to 2% of infected persons develop chronic Q fever, most often persons with underlying (cardiovascular) disease or immunocompromised individuals [1]. Chronic Q fever often presents as endocarditis, but currently in the Netherlands many cases of vascular infections occur [2]. Estimates of the case fatality rate for chronic Q fever vary from 5 to 50%, depending on clinical manifestations and treatment options [3]. Page 2 of 7 Page 2 of 7 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 In literature, no C. burnetii transmission through tissue transplantation has been described, but single ca- ses of transmission through blood transfusion [4] and through bone marrow transplantation to an immuno- compromised recipient [5] were reported as well as transmission through organ transplantation in animals [6]. C. burnetii has been shown to persist in various tis- sues after acute infection, most notably in bone marrow [7,8]. C. burnetii will not be detected by microbiological cultures, as employed in tissue banks. A pilot study indi- cated that serological testing on post-mortem blood is possible, with the test showing sufficient sensitivity and specificity to proceed to a larger scale of testing. C. bur- netii is not rendered harmless by storage, even at low temperatures, but some processing techniques may have a sterilising effect, e.g. glycerolisation, alcoholisation, peracetic acid treatment, irradiation or decellularisation. Different tissues may pose different risks, with heart valves and bone (in which Coxiella may grow during chronic infection) probably posing the highest risk [9]. Detection of Coxiella antibodies With progression of the Dutch outbreak, the risk of transmitting C. burnetii through tissue transplantation was minimized by implementing control measures based on a risk assessment [9]. These included exclusion of do- nors with increased risk of acute or chronic Q fever, based on occupational and geographical risk factors, and on clinical presentation and medical history. Further- more, the possibility for serological testing was investi- gated as well as processing strategies that can render the bacteria harmless. Currently, the Dutch Health Council has advised to screen tissue donors (with the exception of cornea donors) for signs of current or past infection with C. burnetii, the major concern after the outbreaks being that tissues from donors with chronic Q fever might be infectious [10]. The current study was under- taken prior to this advice with the following aims: Serum samples from the post-mortem tissue donors were tested for IgG antibodies against phase 2 of C. burnetii using the CE-marked Serion enzyme immuno- assay (EIA) test (Serion, Clindia Benelux, Leusden, the Netherlands). The cut-off values for EIA (borderline) po- sitivity were determined according to the manufacturer’s instructions. Borderline reactive samples were consi- dered positive. Confirmation of positive samples was performed using an immunofluorescence assay (IFA) for IgG antibodies against phase 1 and 2 of C. burnetii (Focus Diagnostics, Cypress, CA) following instructions of the manufacturer, using a cutoff dilution of 1:32. An IgG phase 1 antibody titer ≥1/1024 was considered suspect for chronic Q fever [11-13]. Both serologic tests have, to our knowledge, never been used for post- mortem blood samples. Cadaveric specimens are of lower quality than regular blood samples, showing more false- positive and false-negative reactions. Validation of the EIA and the IFA was therefore performed prior to the start of the study. Validation of a serological test for use with post- mortem samples and estimation of the seroprevalence of C. burnetii infection in Dutch tissue donors. To determine whether C. burnetii DNA can be detected in tissues for transplantation (cornea with scleral rim, skin, heart valves, bone marrow) after Coxiella infection. The Serion anti-Coxiella phase 2 IgG EIA was vali- dated by testing 45 randomly selected donated post- mortem samples. One of the 45 samples tested positive; this result was confirmed by IFA. Background The tissue with the lowest transmission risk through transplantation is probably cornea, which is avascular, and no Q fever symptoms have been described in the anterior eye [9]. Donor serum samples Serum samples were included from all Dutch post- mortem tissue donors between October 2010 and June 2011, from whom at least one tissue was approved at ini- tial assessment. All serum samples were obtained within 24 hours post-mortem, unless there was haemodilution or insufficient quality, in which cases (pre-transfusion) ante- mortem samples were used for testing. Standard donor se- lection criteria were applied during the study. Regarding Q fever the following donors were excluded: donors with proven acute Q fever; donors with signs of acute Q fever (such as flu-like symptoms, pneumonia without a clear cause or identified pathogen or hepatitis); and donors with a high risk of acute or chronic Q fever, such as donors with occupational hazard (i.e. farmers and veterinarians). Donor tissue samples Tissue samples from all donors who tested positive for IgG antibodies against phase 2 of C. burnetii were col- lected and stored for detection of Coxiella DNA (provided that permission for transplantation-related research had been given). Detection of Coxiella antibodies The average EIA signal (measured as the optical density signal to cutoff ratio) was not different between negative post-mortem samples and negative samples from 92 healthy blood donors from the Northwestern part of the country (OD/CO 0.107 ± 0.112 versus 0.104 ± 0.008, p = 0.87). The clinical This study provides information on which donor se- lection policies for tissue transplantation can be based, with an optimal balance between donor safety and tissue availability, providing a screening system that can be used during outbreaks of Q fever. Page 3 of 7 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 fourth of the cornea, ±50-100 mm3 of skin tissue, heart valve tissue or bone marrow was digested with 25 μL of proteinase K (20 mg/ml; Roche Diagnostics GmbH, Mannheim, Germany) and 225 μl digestion solution (0,5% SDS, 21 mM Tris–HCl). Paraffin-embedded tis- sues were cut into sections. Approximately 1–1.5 cm2 of sectioned tissue was put in 250 μL proteinase K diges- tion solution. Digestions were performed in a thermo- shaker at 55°C overnight at 1400 rpm. The subsequent day DNA was extracted using a NucliSens EasyMAG extraction system (bioMérieux, Boxtel, Netherlands). Sam- ples were processed according to manufacturer’s instruc- tions and eluted in 60 μL elution buffer. Ten μL of DNA isolate was added to the PCR, which was performed as previously described [15]. An internal control was added to each sample before EasyMAG isolation and a real-time PCR to detect the internal control target was run parallel to the C. burnetii PCR to monitor DNA extraction and PCR inhibition [15]. For cornea samples and paraffin- embedded samples an additional PCR to detect human albumin DNA was performed to ensure sufficient input material in each PCR. Albumin PCR cycle threshold values for cornea PCR’s varied from 19 to 20.3 and for paraffin-embedded tissues from 26.4 to 28.9. specificity in post-mortem samples could not be deter- mined, since no samples from post-mortem tissue do- nors historically proven to be reactive were available. Effects of the cadaveric nature of the samples on the sensitivity were assessed by spiking samples. A panel of 20 samples from post-mortem tissue donors, showing various degrees of haemolysis, was spiked with 1/8 vol- ume of serum from a healthy blood donor who tested positive for both anti-Coxiella phase 1 and 2 IgG [14]. Detection of Coxiella antibodies The average increase in OD caused by the spiking was slightly higher for post-mortem samples than for sera from 18 healthy blood donors (ΔOD = 0.556 ± 0.075 ver- sus 0.486 ± 0.080, respectively; p = 0.009). The signal in- crease was not significantly different in hemolytic samples compared to normal post-mortem samples, suggesting the EIA test is sufficiently robust for measuring post-mortem samples of relatively low quality. A small-scale validation of the IFA for measuring post- mortem serum samples was performed by measuring 27 random post-mortem samples. All samples tested nega- tive and no high background fluorescence was observed. Further validation was done by spiking samples suspec- ted to be false-positive in the EIA (see results section). 18 samples with a varying degree of EIA reactivity that were not confirmed by IFA were spiked with 1/8 volume of serum from a healthy blood donor positive for phase 1 and 2 IgG. The fluorescence intensity after spiking did not significantly differ between EIA-negative samples, borderline-positive samples and IEA-positive samples, suggesting that the negative IFA results (of the unspiked samples) were not caused by signal inhibition. Data collection D h Donor characteristics, such as age, gender, cause of death, clinical characteristics, and place of residence were recorded. Three criteria for increased geographical risk of Q fever were applied. First, living in a four-digit postal code area where at least one Q fever patient was reported in the preceding 3 months. Second, living within a five-kilometre radius of an infected farm, where C. burnetii was detected in the bulk tank milk. Third, living in a three-digit postal code area in which the Q fever incidence was higher than 20/100.000 inhabitants in any of the years 2007 to 2010. Approximately 15% of the Dutch population lives in this area, where 86.6% of the Q fever cases were reported. The data on Q fever in- cidence were obtained from the Dutch National Institute for Public Health and the Environment. The data on bulk tank milk positive farms were obtained from the Dutch Food and Consumer Product Safety Authority. The five-kilometer radius from infected farms to the residence of each donor was determined by measuring the distance between both postal codes. Detection of Coxiella DNA Donated tissues from donors positive for C. burnetii antibodies were tested for the presence of C. burnetii DNA by PCR. Corneas were frozen at −20°C until ana- lysis as soon as serology results were known. For cornea donors, a wedge of cornea, including scleral rim, was used for DNA extraction and PCR testing. For skin do- nors, skin samples, frozen in an 85% glycerol buffer, and skin storage solution were tested. For heart valve donors, aortic and pulmonary valves, aortic and pulmonary ar- tery wall and myocardium were tested. If available, samples were used from cryopreserved grafts, stored for transplantation. If no grafts for transplantation were available, sampling was done from formaldehyde-fixed remnants of the heart that are routinely stored for histo- logical examination. These samples were embedded in paraffin and cut into ribbons before DNA extraction. For musculoskeletal tissue donors, bone marrow samples were taken at retrieval and stored at −20°C for PCR testing. For all donors who tested positive for C. burnetii an- tibodies, additional clinical, occupational, geographical and -if available- histological data were gathered, to de- termine the likelihood of previous infection with C. bur- netii, and to establish the presence of signs of chronic Q fever. Information was gathered by reviewing charts, interviewing the general practitioner or next of kin and by reviewing autopsy results and results of histological To efficiently extract DNA from tissues, proteinase K digestion was performed prior to DNA extraction. One Page 4 of 7 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 positive. One donor (0.1%) had IgG phase 1 and 2 titers of both 1:4096 in the IFA, indicative for chronic Q fever [16]. There was a remarkable difference between the two kit lots used for this study. 12/547 (2.2%) of the donors tested (borderline) positive with the first kit lot (SBA.AG), while 38/486 (7.8%) tested positive with the second lot (SKA. CD, p < 0.001). All donors for whom IgG reactivity was not confirmed by IFA had been tested with the second lot. Usually IFA is more sensitive [17]. Thus, it is seems highly likely that the EIA results were false-positive. Spiking ex- periments confirmed this (see Methods section). Compar- ing signals from all measured samples showed elevated background signals for lot SKA.CD and increased num- bers of borderline and positive samples using this lot (Figure 1). Donor characteristics Between October 2010 and June 2011, 1033 consecutive donors donated at least one approved tissue, and were tested for IgG antibodies against phase 2 of C. burnetii. The mean age of the included donors was 65.6 ± 12.3 years (range 0–85) and 664 (64.3%) of them were male. 950 Donors donated corneas (92.0%), 187 skin (18.1%), 139 heart valves (13.5%), 13 thoracic aortas (1.3%), and 86 musculoskeletal tissues (8.3%). PCR testing f h For 29 of the 31 donors who tested positive in the EIA and IFA one or more tissues were available for PCR test- ing. For 2 donors the tissues were discarded after initial assessment at the bank for quality reasons. The number of available tissues for PCR testing and the results of PCR testing are presented for each type of tissue in Table 1. No Coxiella DNA was detected in any of the samples. Data analysis All d All data were entered in a database and analysed with the statistical software program SPSS 15.0 for Windows (SPSS Inc., Chicago, IL). For statistical analysis of dif- ferences between groups, ANOVA or Chi-squared tests were used, when appropriate. P-values ≤0.05 were con- sidered statistically significant. Characteristics of donors who tested positive for C. burnetii phase 2 IgG The mean age in IFA-confirmed seropositive donors was 67.0 ± 9.9 years (range 46–83 years). In the age group from 40–59 years 2.4% of the donors was seropositive, in the group from 60–79 years 3.2% and in group from 80–85 years 3.9%. 22/31 seropositive donors were male (71.0%). The donated tissues are shown in Table 1. Five donors had risk factors for chronic Q fever, namely vas- cular prosthesis (N = 1), heart valve disease (N = 1), ar- tificial heart valve (N = 1), aortic dissection/aneurysm (N = 2). None of these donors with risk factors for chro- nic Q fever had a serological profile indicating chronic disease, and in two donors autopsy was performed, which showed no abnormalities. In four seropositive donors the remnant hearts after heart valve donation were examined, which resulted in identification of one donor with a focal pericarditis. However, PCR testing of samples of the heart of this donor did not reveal any C. burnetii DNA. The donor who had a serological profile suspect for chronic Q Thirty-three donors (3.2%) lived in a postal code area, where in the previous three months a Q fever patient had been reported. Eighty donors (7.7%) lived within a five-kilometre radius of a previously contaminated farm, and 147 (14.3%) donors lived in the high incidence Q fever area. Ethics Consent for donation, including testing for transmittable infectious diseases, was obtained prior to donation from either the donor, as registered in the Dutch donor regis- try, or from the legal next of kin. Because of the Q fever outbreak in The Netherlands the Dutch Health Advisory board deemed additional testing for C. burnetii neces- sary. Since consent for testing for transmittable infec- tious diseases was obtained and all data were analysed anonymized no specific approval by an Ethical Commit- tee was needed for this study according to the Code of Conduct for Health Research, as implemented by the Dutch Federation of Biomedical Scientific Societies, which is followed by the involved organisations. These data suggest a problem with cross-reactivity with unknown antigens leading to false-positive reactiv- ity. Indeed, re-testing the 19 discordant samples with a new lot that became available after the study showed ten non-reactive, eight borderline and one positive sample in this lot. Based on these results, we consider only donors who tested positive in both EIA and IFA C. burnetii seropositive. Detection of Coxiella DNA examination of remnant hearts after heart valve dona- tion, if performed. Factors that were considered risk fac- tors for chronic Q fever are heart valve disease, heart valve or vascular prosthetics, aortic aneurysms and being immunocompromised [16]. Serological testing Fifty of the 1033 donors (4.8%) tested positive for IgG antibodies against phase 2 of C. burnetii by EIA. In 31 donors (3.0%) the reactivity was confirmed by IFA. Twenty-four samples were reactive for both phase 1 and 2, while seven samples were only reactive for phase 2 IgG. One sample showed very high background fluo- rescence with a few discernible bacteria and was scored Page 5 of 7 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 0 1 2 3 anti-phase 2 IgG OD/CO 0 20 40 60 80 100 Samples (%) lot SBA.AG (n=547) lot SKA.CD (n=486) Figure 1 Distribution of the anti-Coxiella phase 2 IgG signals observed with the two EIA kit lots used in this study. OD/CO = optical density divided by the cutoff. Samples with OD/CO ≥1 are considered positive. The borderline cutoff is lot-dependent and was 0.75 for lot SKA. CD and 0.72 for lot SBA.AG. anti-phase 2 IgG OD/CO lot SKA.CD (n=486) Figure 1 Distribution of the anti-Coxiella phase 2 IgG signals observed with the two EIA kit lots used in this study. OD/CO = optical density divided by the cutoff. Samples with OD/CO ≥1 are considered positive. The borderline cutoff is lot-dependent and was 0.75 for lot SKA. CD and 0.72 for lot SBA.AG. fever was a male donor, 68 years old, who died of an acute cardiac arrest due to a myocardial infarction. The donor did not live in a high risk area for Q fever and had no work related risk factors for Q fever, signs of past infection or risk factors for chronic infection. No autopsy was per- formed. He donated corneas and skin, which tested nega- tive for C. burnetii DNA. No seropositive patients lived in an area where in the last 13 weeks an acute Q fever pa- tient had been reported, but 6 seropositive donors lived within a five-kilometer radius of an infected farm. Eleven donors (35.5%) lived in a high incidence Q fever area. There was no significant difference in age between groups of seropositive or seronegative donors. There was no significant correlation between gender and seroposi- tivity; the percentage of males was 71.0% in seropositive versus 64.1% in seronegative donors, respectively. Serological testing burnetii DNA in donated tissues Serological testing There was no increased level of seropositivity in donors living in areas where in the previous 13 weeks acute Q fever patients had been reported or in donors living within a five-kilometer radius from an infected farm. In contrast, the percentage seropositivity was higher in donors living in high incidence Q fever areas as compared to donors living elsewhere (7.5% vs. 2.3%, P = 0.002). fever was a male donor, 68 years old, who died of an acute cardiac arrest due to a myocardial infarction. The donor did not live in a high risk area for Q fever and had no work related risk factors for Q fever, signs of past infection or risk factors for chronic infection. No autopsy was per- formed. He donated corneas and skin, which tested nega- tive for C. burnetii DNA. No seropositive patients lived in an area where in the last 13 weeks an acute Q fever pa- tient had been reported, but 6 seropositive donors lived within a five-kilometer radius of an infected farm. Eleven donors (35.5%) lived in a high incidence Q fever area. Table 1 Results of PCR testing for the presence of C. burnetii DNA in donated tissues Cornea Skin Heart valves Musculoskeletal EIA and IFA positive N = 31 Donated tissues 28 6 5 4 Available for PCR testing 25 6 4 3 PCR negative 24 +1a 6 4b 3 Results of PCR testing for the presence of C. burnetii DNA in donated tissues in donors who tested anti-Coxiella positive in EIA and IFA. a+1 indicates that one test gave an invalid result due to inhibition of the PCR. bFrom one patient no aortic valve and aortic wall tissue was available for PCR testing. Table 1 Results of PCR testing for the presence of C. burnetii DNA in donated tissues Cornea Skin Heart valves Musculoskeletal EIA and IFA positive N = 31 Donated tissues 28 6 5 4 Available for PCR testing 25 6 4 3 PCR negative 24 +1a 6 4b 3 Results of PCR testing for the presence of C. burnetii DNA in donated tissues in donors who tested anti-Coxiella positive in EIA and IFA. a+1 indicates that one test gave an invalid result due to inhibition of the PCR. bFrom one patient no aortic valve and aortic wall tissue was available for PCR testing. Table 1 Results of PCR testing for the presence of C. Discussion A f A few years after the large Q fever outbreaks in the Netherlands, concern about potential transmission through tissue transplantation is more focused on donors with si- lently incubating chronic Q fever than on acute infections. Risk assessments are hampered by limited knowledge about the magnitude of the outbreaks. Although the out- breaks have been studied extensively, most studies were limited to the outbreak areas. To our knowledge, no nation-wide data are available on the prevalence of anti- bodies against Coxiella after the outbreaks, and our study provides the first estimate of the general seroprevalence of Page 6 of 7 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 Page 6 of 7 3.0% (31/1033 donors). It should be added that the cohort may not be representative for the general population. The seroprevalence of 3.0% is twice as high as the prevalence prior to the series of outbreaks which was reported to be 1.5% in the general population, using the same EIA assay [18] and would even be higher if only EIA reactivity was considered (50/1033, 4.8%). can be drawn. The current study contained only one donor (0.1%) with a serological profile indicating chronic Q fever. The cornea and skin samples of this donor did not test positive for Coxiella DNA. It was not possible to confirm the diagnosis any further since the donor was deceased. Our data show that the combination of the Serion EIA and confirmation with IFA is an effective algorithm for screening post-mortem samples from tissue donors. The IFA test is considered the ‘golden standard’ for Q fever diagnostics, and is more sensitive than the EIA [17]. The EIA was chosen for the screening because it is less labour-intensive and more convenient for screening lar- ger numbers of samples. It is unlikely that the lower sen- sitivity of the EIA (as compared to the IFA) increases the risk of transplanting infected tissues, since all patients suffering from chronic Coxiella infection show high IgG titers, and the performance of the test during or shortly after acute infection is excellent. The major disadvantage of using a less sensitive screening test is that the preva- lence of Q fever may be underestimated due to waning of the IgG signal [18]. This underestimation will increase over time, assuming no new outbreaks occur. Discussion A f With the current study, that was undertaken at the end of/right after the outbreak period, this underestimation is ex- pected to be limited. An obvious advantage of a less sen- sitive screening test is that fewer tissues will be rejected based on screening results. Since Coxiella transmission is not expected years after acute infection, using the EIA as a screening method may provide the optimal balance between safety and availability of donated tissues. The seroprevalence was 7.5% (11/147 donors) in the area with a high number of Q fever notifications (>20/ 100.000 inhabitants during the outbreaks). The preva- lence in this high-risk area is lower than the previously reported 12% in healthy blood donors in the outbreak area [14] and the 10.7% reported by a hospital in the center of the outbreak [19]. In both cases the most likely explanation for the lower prevalence found among tissue donors is the definition of a larger high-risk area. In the age group of most donors (40 and older) in samples from 2006–2007 the prevalence was 2.1%. For this age group, the difference with the seroprevalence determined in this study is not significant. Prior to the outbreaks, the anti-Coxiella seroprevalence in the Netherlands was shown to increase with age [18]. A likely explanation why age was not identified as a risk factor for tissue donors is that few post-mortem tis- sue donors are younger than 40, and during the out- breaks it was unlikely that age affected the exposure risk since Q fever is an airborne infection. There was an unexpected difference in specificity be- tween the two EIA kit lots used in this study. It is not known whether the high number of false-positive reac- tions observed with one lot was partly due to the cada- veric nature of the specimens. The intended use of the EIA is the detection of recent or chronic Coxiella infec- tions and not the screening of donors, in whom the low incidence intrinsically leads to a lower positive predictive value. Tests used for screening of donors require a more stringent batch release than regular in-vitro diagnostics tests, but no Coxiella test is on the market for screening purposes. After the Q fever outbreaks in The Netherlands the Health Council of the Netherlands advised to test tissue donors donating tissues with a higher risk of transmis- sion for contamination with C. burnetii [10]. Discussion A f Because of the geographic spread with cases in almost the whole country in the course of the outbreak, the screening was performed nationwide. The necessity of a nationwide screening was confirmed by the results of this current study in which more than half of the seropositive donors (17 of 31) lived outside the risk areas for Q fever. No guideline was given by the Health Council as to when testing of donors can be stopped. The interval between initial infection and when chronic Q fever becomes manifest is reported to be years [13]. Since the main concern is transmission through tissues from donors with chronic Q fever, it may be reasonable to stop test- ing when the number of new chronic Q fever patients in The Netherlands drops back to pre-outbreak levels. After acute Coxiella infection, DNA and antigen can be detected for a long time in several tissues, in parti- cular in bone marrow where even in the absence of se- rological evidence for chronic infection, DNA can be detected up to decades after infection [7,8]. Animal ex- periments showed that the DNA-positive bone marrow is not infectious in susceptible mice [8]. In contrast, bone marrow from a patient with Q fever endocarditis was infectious in guinea pigs [20]. We detected no Coxiella DNA in any of the tissues available for PCR testing. The 24 PCR-negative corneas confirmed ex- pectations from a risk assessment that no bacteria are present in corneas after infection with C. burnetii. For skin, heart valves and bone marrow the numbers of tested donors were quite small and no final conclusions Authors’ contributions MJVW, DWM, NHMR, MHAH, HLZ and BMH were involved in the design of the study. NHMR and MHAH were involved in the molecular analyses. HLZ and BMH were involved in the development of serological testing and BMH was in charge of the serological testing. MJVW, DWM, BMH collected the data, analysed them, interpreted the results and drafted the manuscript. All authors reviewed and revised the manuscript and approved of the final version. 14. Hogema BM, Slot E, Molier M, Schneeberger PM, Hermans MH, Van Hannen EJ, Van der Hoek W, Cuijpers HT, Zaaijer HL: Coxiella burnetii infection among blood donors during the, Q-fever outbreak in the Netherlands. Transfusion 2009, 2012(52):144–150. 15. Schneeberger PM, Hermans MH, van Hannen EJ, Schellekens JJ, Leenders AC, Wever PC: Real-time PCR with serum samples is indispensable for early diagnosis of acute Q fever. Clin Vaccine Immunol 2010, 17:286–290. Conclusions This study provides a validated and effective test algo- rithm that can be used for the screening of post-mortem samples of tissue donors for antibodies against Coxiella Page 7 of 7 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 van Wijk et al. BMC Infectious Diseases 2014, 14:6 http://www.biomedcentral.com/1471-2334/14/6 5. Kanfer E, Farrag N, Price C, MacDonald D, Coleman J, Barrett AJ: Q fever following bone marrow transplantation. Bone Marrow Transplant 1988, 3:165–166. burnetii. Furthermore, this study provides a first estimate of 3.0% of the seroprevalence of antibodies against Coxiella in the Dutch deceased tissue donor popula- tion after the recent outbreaks of Q fever in The Netherlands. The fact that all corneas from donors with IgG reactivity tested negative for the presence of Coxiella DNA confirmed expectations from a risk assessment that no bacteria are present in corneas after infection with C. burnetii. The data provided by this study can be used to optimize the balance between safety and availability of do- nated tissues. 6. Criley JM, Carty AJ, Besch-Williford CL, Franklin CL: Coxiella burnetii infection in C.B-17 scid-bg mice xenotransplanted with fetal bovine tissue. Comp Med 2001, 51:357–360. 7. Harris RJ, Storm PA, Lloyd A, Arens M, Marmion BP: Long-term persistence of Coxiella burnetii in the host after primary Q fever. Epidemiol Infect 2000, 124:543–549. 8. Marmion BP, Storm PA, Ayres JG, Semendric L, Mathews L, Winslow W, Turra M, Harris RJ: Long-term persistence of Coxiella burnetii after acute primary Q fever. QJM 2005, 98:7–20. 9. Van Wijk MJ, Hogema BM, Maas DW, Bokhorst AG: A Q fever outbreak in the Netherlands: consequences for tissue banking. Transfus Med Hemother 2011, 38:357–364. 10. Health Council of the Netherlands: Q fever: risk of transmission via blood or other body material; 2011. http://www.gezondheidsraad.nl/sites/default/files/ 201115EQfever.pdf. Abbreviations EIA E i EIA: Enzyme immunoassay; IFA: Immunofluorescence assay; IgG: Immunoglobulin G; DNA: Desoxyribonucleic acid; PCR: Polymerase chain reaction; C. burnetii: Coxiella burnetii; OD: Optical density; CO: Cutoff; ΔOD: Change in optical density; SDS: Sodium dodecyl sulphate; Tris–HCl: Trisaminomethane hydrochloride. 11. Wegdam-Blans MCA, Nabuurs-Franssen MH, Horrevorts AM, Peeters MF, Schneeberger PM, Bijlmer HA: Laboratory diagnosis of acute Q fever. Ned Tijdschr Geneeskd 2010, 154:A2388. 12. Wegdam-Blans MC, Wielders CC, Meekelenkamp J, Korbeeck JM, Herremans T, Tjhie HT, Bijlmer HA, Koopmans MP, Schneeberger PM: Evaluation of commonly used serological tests for detection of Coxiella burnetii antibodies in well-defined acute and follow-up sera. Clin Vaccine Immunol 2012, 19:1110–1115. Author details 1 18. Schimmer B, Notermans DW, Harms MG, Reimerink JH, Bakker J, Schneeberger P, Mollema L, Teunis P, Van Pelt W, Van Duynhoven Y: Low seroprevalence of Q fever in The Netherlands prior to a series of large outbreaks. Epidemiol Infect 2012, 140:27–35. 1BISLIFE Foundation, PO Box 309, 2333BD Leiden, The Netherlands. 2Regional Laboratory Medical Microbiology and Infection Prevention, Jeroen Bosch Hospital, PO Box 90153, 5200ME ‘s Hertogenbosch, The Netherlands. 3Department of Blood-borne Infections and Viral Diagnostic Services, Sanquin blood supply foundation, PO Box 9892, 1006AN Amsterdam, The Netherlands. 4Department of Clinical Virology (CINIMA), Academic Medical Center, PO Box 22660, 1100DD Amsterdam, The Netherlands. 19. Kampschreur LM, Hagenaars JC, Wielders CC, Elsman P, Lestrade PJ, Koning OH, Oosterheert JJ, Renders NH, Wever PC: Screening for Coxiella burnetii seroprevalence in chronic Q fever high-risk groups reveals the magnitude of the Dutch Q fever outbreak. Epidemiol Infect 2013, 141:847–851. Netherlands. 4Department of Clinical Virology (CINIMA), Academic Medical Center, PO Box 22660, 1100DD Amsterdam, The Netherlands. 20. Peacock MG, Philip RN, Williams JC, Faulkner RS: Serological evaluation of O fever in humans: enhanced phase I titers of immunoglobulins G and A are diagnostic for Q fever endocarditis. Infect Immun 1983, 41:1089–1098. 20. Peacock MG, Philip RN, Williams JC, Faulkner RS: Serological evaluation of O fever in humans: enhanced phase I titers of immunoglobulins G and A are diagnostic for Q fever endocarditis. Infect Immun 1983, 41:1089–1098. Received: 23 July 2013 Accepted: 24 December 2013 Published: 6 January 2014 Received: 23 July 2013 Accepted: 24 December 2013 Published: 6 January 2014 doi:10.1186/1471-2334-14-6 Cite this article as: van Wijk et al.: Screening of post-mortem tissue donors for Coxiella burnetii infection after large outbreaks of Q fever in The Netherlands. BMC Infectious Diseases 2014 14:6. Acknowledgements We would like to thank the Euro Tissue bank in Beverwijk, Amnitrans EyeBank Rotterdam and the Heart Valve Bank Rotterdam for their participation in this study by providing all tissues for PCR testing. Furthermore, we would like to thank Ronald Huijsmans of the Jeroen Bosch Hospital in Den Bosch for organizing the molecular analyses and Michel Molier for performing IFA. 16. Raoult D, Tissot-Dupont H, Foucault C, Gouvernet J, Fournier PE, Bernit E: Q fever 1985–1998. Clinical and epidemiologic features of 1,383 infections. Medicine (Baltimore) 2000, 79:109–123. 17. Herremans T, Hogema BM, Nabuurs M, Peeters M, Wegdam-Blans M, Schneeberger P, Nijhuis C, Notermans DW, Galama J, Horrevoets A, Van Loo IH, Vlaminckx B, Zaaijer HL, Koopmans MP, Berkhout H, Socolovschi C, Raoult D, Stenos J, Nocholson W, Bijlmer H: Comparison of the performance of IFA, CFA, and ELISA assays for the serodiagnosis of acute Q fever by quality assessment. Diagn Microbiol Infect Dis 2013, 75:16–21. This study was funded mainly by BISLIFE. No financial support was received from outside the authors’ institutions and involved tissue banks. Competing interests h f l There are no financial or non-financial competing interests related to this manuscript to declare. 13. Wegdam-Blans MC, Kampschreur LM, Delsing CE, Bleeker-Rovers CP, Sprong T, van Kasteren ME, Notermans DW, Renders NH, Bijlmer HA, Lestrade PJ, Koopmans MP, Nabuurs-Franssen MH, Oosterheert JJ: Chronic Q fever: review of the literature and a proposal of new diagnostic criteria. J Infect 2012, 64:247–259. 1. ECDC: Risk assessment on Q fever. Stockholm; 2010. http://www.ecdc.europa. eu/en/publications/Publications/1005_TER_Risk_Assessment_Qfever.pdf. ISBN ISBN 978-92-9193-210-8. 2. Kampschreur LM, Dekker S, Hagenaars JCJP, Lestrade PJ, Renders NHM, De Jager-Leclercq MGL, Hermans MHA, Groot CAR, Groenewold RHH, Hoepelman AIM, Wever PC, Oosterheert JJ: Identification of risk factors for chronic Q fever, the Netherlands. Emerg Infect Dis 2012, 18:563–570. 3. Raoult D, Houpikian P, Dupont H, Riss J, Arditi-Djiane JJ, Brouqui P: Treatment of Q fever endocarditis: comparison of 2 regimens containing doxycycline and ofloxacin or Hydroxychloroquine. Arch Intern Med 1999, 159(2):167–173. l f l f 1. ECDC: Risk assessment on Q fever. Stockholm; 2010. http://www.ecdc.europa. eu/en/publications/Publications/1005_TER_Risk_Assessment_Qfever.pdf. ISBN ISBN 978-92-9193-210-8. References 1. ECDC: Risk assessment on Q fever. Stockholm; 2010. http://www.ecdc.europa. eu/en/publications/Publications/1005_TER_Risk_Assessment_Qfever.pdf. ISBN ISBN 978-92-9193-210-8. 2. Kampschreur LM, Dekker S, Hagenaars JCJP, Lestrade PJ, Renders NHM, De Jager-Leclercq MGL, Hermans MHA, Groot CAR, Groenewold RHH, Hoepelman AIM, Wever PC, Oosterheert JJ: Identification of risk factors for chronic Q fever, the Netherlands. Emerg Infect Dis 2012, 18:563–570. 3. Raoult D, Houpikian P, Dupont H, Riss J, Arditi-Djiane JJ, Brouqui P: Treatment of Q fever endocarditis: comparison of 2 regimens containing doxycycline and ofloxacin or Hydroxychloroquine. Arch Intern Med 1999, 159(2):167–173. 4 R lt D M i T Q f Cli I f t Di 1995 20 489 496 2. Kampschreur LM, Dekker S, Hagenaars JCJP, Lestrade PJ, Renders NHM, De Jager-Leclercq MGL, Hermans MHA, Groot CAR, Groenewold RHH, Hoepelman AIM, Wever PC, Oosterheert JJ: Identification of risk factors for chronic Q fever, the Netherlands. Emerg Infect Dis 2012, 18:563–570. 3. Raoult D, Houpikian P, Dupont H, Riss J, Arditi-Djiane JJ, Brouqui P: Treatment of Q fever endocarditis: comparison of 2 regimens containing doxycycline and ofloxacin or Hydroxychloroquine. Arch Intern Med 1999, 159(2):167–173. 4. Raoult D, Marrie T: Q fever. Clin Infect Dis 1995, 20:489–496.
https://openalex.org/W2794909794	https://link.springer.com/content/pdf/10.1007/s10265-018-1025-x.pdf	English	null	Spiral phyllotaxis underlies constrained variation in Anemone (Ranunculaceae) tepal arrangement	Journal of plant research	2,018	cc-by	8,729	Floral development –Re-evaluation of its importance– Floral development –Re-evaluation of its importance– Abstract Stabilization and variation of floral structures are indispensable for plant reproduction and evolution; however, the develop- mental mechanism regulating their structural robustness is largely unknown. To investigate this mechanism, we examined positional arrangement (aestivation) of excessively produced perianth organs (tepals) of six- and seven-tepaled (lobed) flowers in six Anemone species (Ranunculaceae). We found that the tepal arrangement that occurred in nature varied intraspecifically between spiral and whorled arrangements. Moreover, among the studied species, variation was commonly limited to three types, including whorls, despite five geometrically possible arrangements in six-tepaled flowers and two types among six possibilities in seven-tepaled flowers. A spiral arrangement, on the other hand, was unique to five-tepaled flowers. A spiral phyllotaxis model with stochasticity on initiating excessive primordia accounted for these limited variations in arrangement in cases when the divergence angle between preexisting primordia was less than 144°. Moreover, interspecific differences in the frequency of the observed arrangements were explained by the change of model parameters that represent meristematic growth and differential organ growth. These findings suggest that the phyllotaxis parameters are responsible for not only intraspecific stability but interspecific difference of floral structure. Decreasing arrangements from six-tepaled to seven- tepaled Anemone flowers demonstrate that the stabilization occurs as development proceeds to increase the component (organ) number, in contrast from the intuition that the variation will be larger due to increasing number of possible states (arrangements). Keywords Floral organ · Variation · Floral development · Mathematical model · Phyllotaxis · Ranunculaceae Journal of Plant Research (2018) 131:459–468 https://doi.org/10.1007/s10265-018-1025-x Journal of Plant Research (2018) 131:459–468 https://doi.org/10.1007/s10265-018-1025-x Spiral phyllotaxis underlies constrained variation in Anemone (Ranunculaceae) tepal arrangement Miho S. Kitazawa1,2 · Koichi Fujimoto2 Received: 10 November 2017 / Accepted: 26 February 2018 / Published online: 27 March 2018 © The Author(s) 2018 * Miho S. Kitazawa kitazawa@celas.osaka‑u.ac.jp Introduction Previously, we studied the variation in tepal arrange- ments of flowers with six tepals in wild populations of three Anemone species and found that variation was limited to three aestivation types, including trimerous double whorl type (floral diagram at the left end of middle row in Fig. 1), despite the option of five geometrically possible types (Ki d F ji 2016b) T i hi li i d Fig. 1 Possible arrangements for six- and seven-tepal flow- ers (middle and bottom rows, respectively) starting from quincuncial arrangements (top). E, I and A below floral dia- grams denote external, internal, and alternating arrangement of each tepal, respectively. The sequence of these letters indi- cates the positional relationship of tepals in a flower 460 Journal of Plant Research (2018) 131:459–468 perianth organs (tepals), which are usually white, pink, or purple but not green as sepals of core eudicots. As in other genera of Ranunculaceae, Anemone flowers show inter- and intra-specific diversity in the number of floral parts (Schöffel 1932; Yule 1902). In fact, this genus includes species, such as A. nemorosa and A. hepatica (Hepatica nobilis), with two cycles of trimerous whorls as well as spiral flowers, includ- ing A. nikoensis, A. flaccida, and A. scabiosa (A. × hybrida; Japanese anemone), with a typical tepal number of five (Kitazawa and Fujimoto 2016a). These species are not separately clustered but rather scattered in the phylogenetic tree (Hoot et al. 2012). Furthermore, trimerous-like arrange- ments can be stochastically found even in the species with pentamerous spiral flowers (Kitazawa and Fujimoto 2016b; Ren et al. 2010; Schöffel 1932). The development of some Anemone flowers indicates a transient pattern between spi- ral and trimerous whorls: first three tepal primordia arise sequentially, and then three tepal primordia initiate at once (Ren et al. 2010). Therefore, the Anemone flowers with six tepals can be considered a primitive form of the trimerous whorl, suggesting that whorled arrangement and spiral ini- tiation is closely related in developmental regulation. Previously we studied the variation in tepal arrange- Fig. 1 Possible arrangements for six- and seven-tepal flow- ers (middle and bottom rows, respectively) starting from quincuncial arrangements (top). E, I and A below floral dia- grams denote external, internal, and alternating arrangement of each tepal, respectively. The sequence of these letters indi- cates the positional relationship of tepals in a flower Fig. Introduction perianth organs (tepals), which are usually white, pink, or purple but not green as sepals of core eudicots. As in other genera of Ranunculaceae, Anemone flowers show inter- and intra-specific diversity in the number of floral parts (Schöffel 1932; Yule 1902). In fact, this genus includes species, such as A. nemorosa and A. hepatica (Hepatica nobilis), with two cycles of trimerous whorls as well as spiral flowers, includ- ing A. nikoensis, A. flaccida, and A. scabiosa (A. × hybrida; Japanese anemone), with a typical tepal number of five (Kitazawa and Fujimoto 2016a). These species are not separately clustered but rather scattered in the phylogenetic tree (Hoot et al. 2012). Furthermore, trimerous-like arrange- ments can be stochastically found even in the species with pentamerous spiral flowers (Kitazawa and Fujimoto 2016b; Ren et al. 2010; Schöffel 1932). The development of some Anemone flowers indicates a transient pattern between spi- ral and trimerous whorls: first three tepal primordia arise sequentially, and then three tepal primordia initiate at once (Ren et al. 2010). Therefore, the Anemone flowers with six tepals can be considered a primitive form of the trimerous whorl, suggesting that whorled arrangement and spiral ini- tiation is closely related in developmental regulation. perianth organs (tepals), which are usually white, pink, or purple but not green as sepals of core eudicots. As in other genera of Ranunculaceae, Anemone flowers show inter- and intra-specific diversity in the number of floral parts (Schöffel 1932; Yule 1902). In fact, this genus includes species, such as A. nemorosa and A. hepatica (Hepatica nobilis), with two cycles of trimerous whorls as well as spiral flowers, includ- ing A. nikoensis, A. flaccida, and A. scabiosa (A. × hybrida; Japanese anemone), with a typical tepal number of five (Kitazawa and Fujimoto 2016a). These species are not separately clustered but rather scattered in the phylogenetic tree (Hoot et al. 2012). Furthermore, trimerous-like arrange- ments can be stochastically found even in the species with pentamerous spiral flowers (Kitazawa and Fujimoto 2016b; Ren et al. 2010; Schöffel 1932). The development of some Anemone flowers indicates a transient pattern between spi- ral and trimerous whorls: first three tepal primordia arise sequentially, and then three tepal primordia initiate at once (Ren et al. 2010). Therefore, the Anemone flowers with six tepals can be considered a primitive form of the trimerous whorl, suggesting that whorled arrangement and spiral ini- tiation is closely related in developmental regulation. Introduction morphology with stable and clade-specific numbers of floral parts. The basic number is three in monocots, which usu- ally have two cycles of trimerous whorls in a flower, yield- ing a tepal number of six (floral diagram at the left end of middle row in Fig. 1), and in Magnoliids. Alternately, the basic number is four or five in eudicots. The evolutionary course and even the ancestral state are largely unclear for these plants. One way to increase our understanding of the evolutionary change of floral morphology is to clarify the developmental mechanism that has caused such differences between clades.l Diversification of number of floral parts (sepals and petals) is one of the major problems of floral evolution (Ronse De Craene 2016). From the ancestral state, plants evolved into a Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10265-018-1025-x) contains supplementary material, which is available to authorized users. * Miho S. Kitazawa kitazawa@celas.osaka‑u.ac.jp * Koichi Fujimoto fujimoto@bio.sci.osaka‑u.ac.jp 1 Center for Education in Liberal Arts and Sciences, Osaka University, Toyonaka, Osaka, Japan 2 Department of Biological Sciences, Graduate School of Science, Osaka University, Toyonaka, Osaka, Japan Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10265-018-1025-x) contains supplementary material, which is available to authorized users. * Miho S. Kitazawa kitazawa@celas.osaka‑u.ac.jp The arrangement of floral parts plays a central role in the precise determination of the number of floral parts. The two major types of floral organ arrangement are spiral and whorled arrangements. Spiral initiation is often found in various clades in angiosperms (Endress and Doyle 2007). In some clades, the aestivation, the overlapping arrangement of petals or sepals is quincuncial (top row in Fig. 1; Bentley * Koichi Fujimoto fujimoto@bio.sci.osaka‑u.ac.jp 1 Center for Education in Liberal Arts and Sciences, Osaka University, Toyonaka, Osaka, Japan 2 Department of Biological Sciences, Graduate School of Science, Osaka University, Toyonaka, Osaka, Japan (0121 460 Journal of Plant Research (2018) 131:459–468 1870). Quincuncial arrangement reflects the spiral initia- tion in at least some clades, such as Rosaceae and Ranuncu- laceae (Foster et al. 2003; Ren et al. 2010; De Craene 2010). Whorled arrangement, however, is more likely to stabilize floral part numbers than spiral arrangements, since the slight fluctuation in organ-fate determinants observed in interme- diate organ morphology (e.g., between tepals and stamens; Gonçalves et al. 2013; Jabbour et al. 2015) readily affects the number in spiral arrangements (Kitazawa and Fujimoto 2014; Wang et al. Introduction 2015). Whether the spiral arrangement or the whorled arrangement of floral organs is ancestral remains controversial (Endress and Doyle 2015). Spiral initiation of floral organs is sometimes considered as the ancestral state based on the morphology of basal angiosperms (Soltis et al. 2000). On the other hand, recent estimations of the ances- tral state indicated that the trimerous whorled arrangement is the ancestral state of the flowers (Sauquet et al. 2017); however, no strong evidence is available to support either of these ideas. Flowers in the family Ranunculaceae have been a focus of studies to elucidate the developmental basis of diversity of numbers of floral organ parts (Damerval and Becker 2017; Gonçalves et al. 2013; Wang et al. 2015). The variety of numbers (2, 3, 5, and their multiples) are observed among species and within a species; therefore, these flowers offer a good system to address the question of how the floral organ number changes. Flowers of Anemone, a genus in the family Ranuncu- l i f l i l i il d di perianth organs (tepals), which are usually white, pink, or purple but not green as sepals of core eudicots. As in other genera of Ranunculaceae, Anemone flowers show inter- and intra-specific diversity in the number of floral parts (Schöffel 1932; Yule 1902). In fact, this genus includes species, such as A. nemorosa and A. hepatica (Hepatica nobilis), with two cycles of trimerous whorls as well as spiral flowers, includ- ing A. nikoensis, A. flaccida, and A. scabiosa (A. × hybrida; Japanese anemone), with a typical tepal number of five (Kitazawa and Fujimoto 2016a). These species are not separately clustered but rather scattered in the phylogenetic tree (Hoot et al. 2012). Furthermore, trimerous-like arrange- ments can be stochastically found even in the species with pentamerous spiral flowers (Kitazawa and Fujimoto 2016b; Ren et al. 2010; Schöffel 1932). The development of some Anemone flowers indicates a transient pattern between spi- ral and trimerous whorls: first three tepal primordia arise sequentially, and then three tepal primordia initiate at once (Ren et al. 2010). Therefore, the Anemone flowers with six tepals can be considered a primitive form of the trimerous whorl, suggesting that whorled arrangement and spiral ini- tiation is closely related in developmental regulation. Introduction 1 Possible arrangements for six- and seven-tepal flow- ers (middle and bottom rows, respectively) starting from quincuncial arrangements (top). E, I and A below floral dia- grams denote external, internal, and alternating arrangement of each tepal, respectively. The sequence of these letters indi- cates the positional relationship of tepals in a flower 1870). Quincuncial arrangement reflects the spiral initia- tion in at least some clades, such as Rosaceae and Ranuncu- laceae (Foster et al. 2003; Ren et al. 2010; De Craene 2010). Whorled arrangement, however, is more likely to stabilize floral part numbers than spiral arrangements, since the slight fluctuation in organ-fate determinants observed in interme- diate organ morphology (e.g., between tepals and stamens; Gonçalves et al. 2013; Jabbour et al. 2015) readily affects the number in spiral arrangements (Kitazawa and Fujimoto 2014; Wang et al. 2015). Whether the spiral arrangement or the whorled arrangement of floral organs is ancestral remains controversial (Endress and Doyle 2015). Spiral initiation of floral organs is sometimes considered as the ancestral state based on the morphology of basal angiosperms (Soltis et al. 2000). On the other hand, recent estimations of the ances- tral state indicated that the trimerous whorled arrangement is the ancestral state of the flowers (Sauquet et al. 2017); however, no strong evidence is available to support either of these ideas. 1870). Quincuncial arrangement reflects the spiral initia- tion in at least some clades, such as Rosaceae and Ranuncu- laceae (Foster et al. 2003; Ren et al. 2010; De Craene 2010). Whorled arrangement, however, is more likely to stabilize floral part numbers than spiral arrangements, since the slight fluctuation in organ-fate determinants observed in interme- diate organ morphology (e.g., between tepals and stamens; Gonçalves et al. 2013; Jabbour et al. 2015) readily affects the number in spiral arrangements (Kitazawa and Fujimoto 2014; Wang et al. 2015). Whether the spiral arrangement or the whorled arrangement of floral organs is ancestral remains controversial (Endress and Doyle 2015). Spiral initiation of floral organs is sometimes considered as the ancestral state based on the morphology of basal angiosperms (Soltis et al. 2000). On the other hand, recent estimations of the ances- tral state indicated that the trimerous whorled arrangement is the ancestral state of the flowers (Sauquet et al. 2017); however, no strong evidence is available to support either of these ideas. Mathematical model settings To study the stochasticity in positional arrangement of the sixth and seventh tepals, we employed the inhibitory field model as described by Douady and Couder (1996). We assumed that the primordia sequentially appear one by one at the periphery of the circular meristem with a radius of R0, following the observation of early development of Anemone (Ren et al. 2010). Radial position of primordium j at the initiation of primordium i was set as (1) rj = R0 + (i −j)VT, (1) where VT is a centrifugal displacement of each primor- dium reflecting meristem growth with a velocity V during the time interval T of initiation of two successive primordia (Fig. 2a). To simplify the situation, the divergence angle φ between the successive i-th and i + 1-th primordia were fixed (1 ≦ i ≦ 4). The angle must be satisfied with 120° < φ < 180° to reproduce the quincuncial aestivation of flowers with five tepals (Fig. 1, top). The angular positions of sixth and later primordia θi (i ≧ 6) were determined stochastically based on the inhibitory effect from existing primordia (Fig. 2b, dashed line). Denoting an index of the current primordium as i (i ≧ 6) and those of existing primordia as j (1 ≦ j < i), the potential energy was formulated based on a previous study of floral phyllotaxis (Kitazawa and Fujimoto 2015, 2016b) as where VT is a centrifugal displacement of each primor- dium reflecting meristem growth with a velocity V during the time interval T of initiation of two successive primordia (Fig. 2a). To simplify the situation, the divergence angle φ between the successive i-th and i + 1-th primordia were fixed (1 ≦ i ≦ 4). The angle must be satisfied with 120° < φ < 180° to reproduce the quincuncial aestivation of flowers with five tepals (Fig. 1, top). The angular positions of sixth and later primordia θi (i ≧ 6) were determined stochastically based on the inhibitory effect from existing primordia (Fig. 2b, dashed line). Denoting an index of the current primordium as i (i ≧ 6) and those of existing primordia as j (1 ≦ j < i), the potential energy was formulated based on a previous study of floral phyllotaxis (Kitazawa and Fujimoto 2015, 2016b) as Introduction × hybrida, we were unable to identify the forms at many of our observation sites. Therefore, we used tepal color as the primary feature to define the forms. Based on tepal color, we could clearly distinguish the three groups: deep pink, pale pink, and white. Populations with deep pink tepals are further classified as types with broad, obovate tepals or with thin, numerous tepals (more than 10). While the latter was excluded from our study, we examined the pale-pink, white, and deep- pink broad tepal groups. Pulsatilla is nested in Anemone in molecular phylogenetic studies (Hoot et al. 2012; Jiang et al. 2017). We counted only the fresh flowers, because tepals became narrower at their basal parts and positional overlaps were lost as days went on after blooming. variation associated with the increasing number of tepals, we analyzed the position of the seventh as well as the sixth tepals of Anemone in both wild populations and a floral phyllotaxis model. We examined the possible consequences of tepal appearance and resultant arrangements from each position of the sixth tepal, assuming that the seventh tepal appears after the sixth tepal and any tepal does not overlap with others (Fig. 1). By comparing actual arrangements with the possible options, we demonstrate that the arrangement stabilizes as the number of components (tepals) increases. Introduction 1870). Quincuncial arrangement reflects the spiral initia- tion in at least some clades, such as Rosaceae and Ranuncu- laceae (Foster et al. 2003; Ren et al. 2010; De Craene 2010). Whorled arrangement, however, is more likely to stabilize floral part numbers than spiral arrangements, since the slight fluctuation in organ-fate determinants observed in interme- diate organ morphology (e.g., between tepals and stamens; Gonçalves et al. 2013; Jabbour et al. 2015) readily affects the number in spiral arrangements (Kitazawa and Fujimoto 2014; Wang et al. 2015). Whether the spiral arrangement or the whorled arrangement of floral organs is ancestral remains controversial (Endress and Doyle 2015). Spiral initiation of floral organs is sometimes considered as the ancestral state based on the morphology of basal angiosperms (Soltis et al. 2000). On the other hand, recent estimations of the ances- tral state indicated that the trimerous whorled arrangement is the ancestral state of the flowers (Sauquet et al. 2017); however, no strong evidence is available to support either of these ideas. Flowers in the family Ranunculaceae have been a focus of studies to elucidate the developmental basis of diversity of numbers of floral organ parts (Damerval and Becker 2017; Gonçalves et al. 2013; Wang et al. 2015). The variety of numbers (2, 3, 5, and their multiples) are observed among species and within a species; therefore, these flowers offer a good system to address the question of how the floral organ number changes. Previously, we studied the variation in tepal arrange- ments of flowers with six tepals in wild populations of three Anemone species and found that variation was limited to three aestivation types, including trimerous double whorl type (floral diagram at the left end of middle row in Fig. 1), despite the option of five geometrically possible types (Kitazawa and Fujimoto 2016b). To examine this limited Flowers of Anemone, a genus in the family Ranuncu- laceae, consist of multiple pistils, stamens, and surrounding 1 3 1 3 Journal of Plant Research (2018) 131:459–468 461 (Hepatica nobilis; Shiga and Okayama pref.), Pulsatilla cer- nua (Hyogo pref.), and three forms of Anemone × hybrida (Japanese anemone; we previously called A. scabiosa in Kitazawa and Fujimoto 2016a, b; Hokkaido, Tokyo, Mie, Shiga, Kyoto, Nara, Osaka and Hyogo pref.) in Japan. The frequency of each arrangement was measured as the sum of the multiple populations within the species (Table S1). Although there are several forms of A. Positional arrangement of perianth organs The arrangement of tepals was examined for flowers with five, six, and seven tepals, whereas the arrangement was not determined for flowers with more than seven tepals. Based on previous studies (Cunnell 1958; Kitazawa and Fujimoto 2016b; Morgan 1874; Schoute 1935), we sur- mised the aestivation by positional arrangements of tepals in mature flowers by identifying the arrangement of each organ with neighboring organs as either external, internal, or alternating (Fig. 1). For example, type 5-I represents quincuncial, whereas type 6-II represents three internal and three external tepals, exemplifying trimerous double whorls. For simplicity, reflected and rotated arrangements were not distinguished, and therefore, the position of the flower with respect to the main axis and the direction of the spiral was ignored. For five-tepaled flowers, there are four geometri- cally possible aestivations (Cunnell 1958). For six-tepaled flowers derived from quincuncial, there are five aestivation types (Fig. 1 middle row; Kitazawa and Fujimoto 2016b). Regarding flowers with seven tepals, since the arrangement type 6-II is radially symmetric, the resulting arrangement converge onto only the type 7-II (Fig. 1 lower). When the arrangements are type 6-I or 6-IV, the arrangement at the sixth tepal is bilateral. Therefore, three possible arrange- ments exist for seven-tepaled flowers: types 7-I, 7-II, and 7-V for type 6-I and types 7-II, 7-IV, and 7-V for type 6-IV. Type 6-III is a biradially symmetric arrangement, and there- fore, this arrangement can generate three arrangements of seven-tepaled flowers: types 7-II, 7-III, and 7-V. The non- symmetric arrangement type 6-V has five possibilities: types 7-I, 7-III, 7-IV, 7-V, and 7-VI. Results Fig. 2 Model settings. The first five primordia appear sequentially with fixed divergence angle φ at the edge of the floral meristem (a). The positions of the fifth and onward primordia are determined by the probability (Eq. 3) calculated from the inhibitory energy (Eq. 2) of the existing primordia (b). The potential and probability are cal- culated discretely with an interval of 0.1°. R0 = λ = 10, φ = 137.5°, VT = α = 0, and b = 10 Species‑dependent appearance of spiral pentamery and whorled trimery The condition VT = 0 corresponds to the situation described in the previous study (Kitazawa and Fujimoto 2016b), while for simplicity, the present model does not incorporate the angular displacement of primordia, which was introduced after primordia initiation in an earlier study (Kitazawa and Fujimoto 2015). The α accounts for the change of the direc- tion of auxin flux toward the inner tissue of primordia and/ or primordial boundary establishment and the increase of primordial volume. Since lower energy indicates a higher potential (probability) to place a new primordium, the prob- ability pθ,i is given by a monotonically decreasing sigmoidal function of the energy, respectively. In addition, α represents the difference of the inhibitory effect due to growth progression of pre-existing organs, whose positive/negative sign is identical to that from an earlier study (Kitazawa and Fujimoto 2015) but opposite of that from another one (Kitazawa and Fujimoto 2016b). The condition VT = 0 corresponds to the situation described in the previous study (Kitazawa and Fujimoto 2016b), while for simplicity, the present model does not incorporate the angular displacement of primordia, which was introduced after primordia initiation in an earlier study (Kitazawa and Fujimoto 2015). The α accounts for the change of the direc- tion of auxin flux toward the inner tissue of primordia and/ or primordial boundary establishment and the increase of primordial volume. Since lower energy indicates a higher potential (probability) to place a new primordium, the prob- ability pθ,i is given by a monotonically decreasing sigmoidal function of the energy, Plant samples (2) E휃,i = ∑ j exp(훼(i −j))exp(−dij∕휆) , We measured the positional arrangements of tepals in wild populations of A. nikoensis (Shiga, Osaka, Hyogo and Okay- ama prefectures), A. flaccida (Hokkaido, Shiga, Hyogo and Okayama pref.), A. soyensis (Hokkaido pref.), A. hepatica (2) where dij and λ are the distance between primordia i and j and a parameter representing effective distance of inhibition, 1 3 Journal of Plant Research (2018) 131:459–468 462 respectively. In addition, α represents the difference of the inhibitory effect due to growth progression of pre existing Fig. 2 Model settings. The first five primordia appear sequentially with fixed divergence angle φ at the edge of the floral meristem (a). The positions of the fifth and onward primordia are determined by the probability (Eq. 3) calculated from the inhibitory energy (Eq. 2) of the existing primordia (b). The potential and probability are cal- culated discretely with an interval of 0.1°. R0 = λ = 10, φ = 137.5°, VT = α = 0, and b = 10 where b is a parameter that indicates the steepness of the sigmoidal function, and µE and σE are the average and stand- ard deviation of energy 휇E ≡∫360 0 E휃,id휃∕360 and 휎E ≡ ( ∫360 0 (E휃,i −휇E)2d휃∕360 )1∕2 , respectively. C0 is a nor- malization constant that satisfies the function ∫360 0 p휃,id휃∕360 = 1 . Using this probability, we numerically obtained the stochastic arrangement of the primordia (Mira- bet et al. 2012; Refahi et al. 2016; Fig. 2b). malization constant that satisfies the function ∫360 0 p휃,id휃∕360 = 1 . Using this probability, we numerically obtained the stochastic arrangement of the primordia (Mira- bet et al. 2012; Refahi et al. 2016; Fig. 2b). Species‑dependent appearance of spiral pentamery and whorled trimery For five-tepaled flowers, almost all flowers underwent a unique aestivation type (n = 20,286 among 20,287 five-tep- aled Anemone flowers). That is, the flowers were quincuncial type 5-I (Fig. 3a), which is consistent with spiral initiation of organ primordia during floral development (Ren et al. 2010). For flowers with six tepals, we examined the frequency of aestivation types. The most frequent aestivation type was type 6-II, except for the A. × hybrida pale pink form, which exhibited the type 6-IV arrangement most frequently (Fig. 3a, b), as reported previously (Kitazawa and Fujimoto 2016b). Notably, whether pentamerous, quincuncial (type 5-I), or trimerous double whorls (type 6-II) was more fre- quent was species-dependent: quincuncial in A. nikoensis, A. flaccida, and A. × hybrida deep pink and pale pink and tri- merous double whorls in A. hepatica, A. soyensis, and Pul- satilla cernua (Fig. 3a). The frequencies of quincuncial and type 6-II were nearly equivalent in A. × hybrida white. The normalized frequency of pentamerous quincuncial further varied among the four observed plant types: A. × hybrida deep pink (99%), A. nikoensis (78%), A. flaccida (59%), and A. × hybrida pale pink (37%). Fig. 2 Model settings. The first five primordia appear sequentially with fixed divergence angle φ at the edge of the floral meristem (a). The positions of the fifth and onward primordia are determined by the probability (Eq. 3) calculated from the inhibitory energy (Eq. 2) of the existing primordia (b). The potential and probability are cal- culated discretely with an interval of 0.1°. R0 = λ = 10, φ = 137.5°, VT = α = 0, and b = 10 Fig. 2 Model settings. The first five primordia appear sequentially with fixed divergence angle φ at the edge of the floral meristem (a). The positions of the fifth and onward primordia are determined by the probability (Eq. 3) calculated from the inhibitory energy (Eq. 2) of the existing primordia (b). The potential and probability are cal- culated discretely with an interval of 0.1°. R0 = λ = 10, φ = 137.5°, VT = α = 0, and b = 10 respectively. In addition, α represents the difference of the inhibitory effect due to growth progression of pre-existing organs, whose positive/negative sign is identical to that from an earlier study (Kitazawa and Fujimoto 2015) but opposite of that from another one (Kitazawa and Fujimoto 2016b). Limited aestivation in six‑tepaled flowers In species showing higher frequency of type 6-II than quin- cuncial (A. soyensis, A. hepatica, and P. cernua), the fre- quency of type 6-II normalized to that of six-tepaled flow- ers was more than 90%, indicating outstanding robustness of type 6-II (Fig. 3a). On the other hand, in those showing equal or higher frequency of quincuncial (type 5-I) than type 6-II (A. flaccida, A. nikoensis, and A. × hybrida), types 6-I and 6-IV also appeared at comparable frequency to type 6-II, whereas the other two aestivation types (types 6-III and 6-V) were rare as reported previously (Kitazawa and Fujimoto 2016b; Fig. 3a, b). We confirmed the constrained appearance to three aestivation types of six-tepaled flowers by meas- uring a four times larger sample size (~ 9 × 103 six-tepaled (3) p휃,i = C0∕(1 + exp(b(E휃,i −(휇E −2휎E )))) , p휃,i = C0∕(1 + exp(b(E휃,i −(휇E −2휎E )))) , (3) 1 3 463 Journal of Plant Research (2018) 131:459–468 Fig. 3 Probability of perianth arrangements of flowers with five or six tepals (a, b) and those with seven tepals (c, d) measured for six Anemone species. The chart indicates the frequency of each arrange- ment normalized by all record counts of the species regardless of the tepal numbers (a, c), whereas the stacked bar chart shows the fre- quency of the arrangements normalized by the number of six-tepaled (b) and seven-tepaled (d) flowers. Photographs under charts (a, b) are representatives of each arrangement of each species and form, whose brightness is modified to clearly show the arrangements Journal of Plant Research (2018) 131:459–468 463 quency of the arrangements normalized by the number of six-tepaled (b) and seven-tepaled (d) flowers. Photographs under charts (a, b) are representatives of each arrangement of each species and form, whose brightness is modified to clearly show the arrangements Fig. 3 Probability of perianth arrangements of flowers with five or six tepals (a, b) and those with seven tepals (c, d) measured for six Anemone species. The chart indicates the frequency of each arrange- ment normalized by all record counts of the species regardless of the tepal numbers (a, c), whereas the stacked bar chart shows the fre- Fig. 3 Probability of perianth arrangements of flowers with five or six tepals (a, b) and those with seven tepals (c, d) measured for six Anemone species. Limited aestivation in seven‑tepaled flowers As a consequence of limited aestivation of flowers with six tepals, we analyzed the aestivation of seven-tepaled flowers that co-exist with the five- and six-tepaled flowers in wild populations (Fig. 3c). Geometrically, there are ten possi- ble aestivation types for seven-tepaled flowers (Fig. 1 lower row and Fig. S1). Even when we assume that the flowers with seven tepals are generated by adding two tepals to the inside of the quincuncial aestivation of five-tepaled flowers, six possible aestivation types remain (lower row in Fig. 1). Intriguingly, we found that two aestivation types (types 7-II and 7-V) appeared much more frequently than the other four types, dominating 98.3% of the seven-tepaled flowers in all observed species and forms (Fig. 3d). In addition, the fre- quency ratio of type 7-II to type 7-V was greater than 5, indicating outstanding robustness of type 7-II in all species and forms, except for A. nikoensis, where this frequency ratio was approximately 1.5, indicating similar robustness of the two aestivation types. il six possible aestivation types remain (lower row in Fig. 1). Intriguingly, we found that two aestivation types (types 7-II and 7-V) appeared much more frequently than the other four types, dominating 98.3% of the seven-tepaled flowers in all observed species and forms (Fig. 3d). In addition, the fre- quency ratio of type 7-II to type 7-V was greater than 5, indicating outstanding robustness of type 7-II in all species and forms, except for A. nikoensis, where this frequency ratio was approximately 1.5, indicating similar robustness of the two aestivation types. Next, we examined the dependency of centrifugal dis- placement due to growth VT (Fig. 4b). For the sixth tepal arrangements, as VT gets larger, the fraction of type 6-II increased until VT = 4 and then decreased as VT increased further. At the same time, the fraction of type 6-III increased, while the fraction of type 6-IV did not change and the frac- tion of type 6-I decreased to 10–20% at VT = 1. The position of the seventh tepal showed higher variation than for the case when VT = 0. For example, the position of the seventh tepal subsequent to the arrangement type 6-IV converged to two arrangements (types 7-II and 7-V), but another arrange- ment (type 7-IV) also appeared when VT was larger than 4 (Fig. 4b). Limited aestivation in seven‑tepaled flowers Overall, as VT increased, a higher degree of vari- ation in the arrangements with seven-tepaled flowers was observed. Therefore, the arrangements with seven tepals were limited to two dominant types (types 7-II and 7-V) at small VT but were more varied as VT increased. Limited aestivation in six‑tepaled flowers The chart indicates the frequency of each arrange- ment normalized by all record counts of the species regardless of the tepal numbers (a, c), whereas the stacked bar chart shows the fre- quency of the arrangements normalized by the number of six-tepaled (b) and seven-tepaled (d) flowers. Photographs under charts (a, b) are representatives of each arrangement of each species and form, whose brightness is modified to clearly show the arrangements 1 3 Journal of Plant Research (2018) 131:459–468 464 and Fujimoto 2016b). Three of the arrangements (types 6-I, 6-II, and 6-IV) observed in Anemone (Fig. 3a, b) appeared when φ < 144°, whereas the other two (types 6-III and 6-V) appeared when φ > 144° (Fig. 4a). Next, we checked the position of the seventh tepal for each arrangement and found that the seventh tepal position was also limited. Two arrangements (types 7-II and 7-V) that were dominant in Anemone (Fig. 3c, d), occupied nearly all the arrange- ments at φ < 144°, despite the six possible arrangements of seven tepals initiating from the quincuncial arrangement (Fig. 1, bottom row). At φ > 144°, the two arrangements decrease in frequency, and type 7-III, which was virtually not observed for Anemone (Fig. 3c, d), becomes dominant, increasing in frequency as φ gets larger (Fig. 4a). Type 7-IV (rank 3 except for A. hepatica) appeared and increased to a frequency of up to 6% when φ was approximately 144° (maximum value at φ = 146°) but decreased as φ was further increased. Type 7-I (rank 3 in A. hepatica and rank 4 in A. nikoensis) appeared at a very low frequency (maximum 1.4% at φ = 126°), decreased as φ neared 144, and then increased again (0.8% when φ = 156°). Type 7-VI (not observed in Anemone) was consistently nearly absent (0.4% at φ = 154° was the maximum) throughout the parameter range exam- ined. Therefore, we conclude that the dominant frequency of types 7-II and 7-V for seven-tepaled flowers as well as types 6-I, 6-II, and 6-IV for six-tepaled flowers is a natural consequence of spiral phyllotaxis with φ < 144°. flowers in all populations of observed species; Fig. 1) than in the previous report. These three types (types 6-I, 6-II, and 6-IV) dominated 99.3% of the six tepaled-flowers in all the observed species and forms (Fig. 3b). Limited aestivation in six‑tepaled flowers The frequency rank of the three types was type 6-II, type 6-IV, and type 6-I in ascending order for A. nikoensis and A. × hybrida white, whereas this order was types 6-II, 6-I, and 6-IV for A. flac- cida and types 6-IV, 6-II, and 6-I for A. × hybrida pale pink. Therefore, six-tepaled flowers of pentamerous species (A. flaccida, A. nikoensis, and A. × hybrida) exhibited limited variation of three aestivation types in a species- and form- dependent manner, while five-tepaled flowers of trimer- ous species (A. soyensis, A. hepatica, and P. cernua) were uniquely quincuncial. 3 A model for spiral phyllotaxis with stochasticity reproduced limited variation of Anemone tepal arrangements To understand the developmental mechanisms of the lim- ited positions of sixth and seventh tepals, we first exam- ined the arrangements of the six tepals at the time when the sixth tepal appeared after the quincuncial arrangement in a phyllotaxis model using computational analysis (Fig. 2). The simplest case includes a divergence angle φ = 144° and no growth (VT = 0; Fig. 4a), since the first five tepals are placed in the regular pentagon. The computational simula- tion was consistent with intuitive expectation. That is, the five possible arrangements of the six tepals appeared with equal probability (φ = 144° in Fig. 4a). The dependency on φ without growth was also consistent with a previous theoreti- cal study based on the energy landscape (Eq. 2; Kitazawa Finally, we examined the dependency of arrangements on another parameter, α representing thedifference of inhibi- tory effect due to growth progression on pre-existing organs (Fig. 4c; see “Materials and Methods” for definition of α). The limitation to three types (types 6-I, 6-II, and 6-IV) in Anemone flowers with six tepals and two types (types 7-II and 7-V) in those with seven tepals was reproduced for 0.05 ≲ α ≲ 0.1. As α represents a bias of inhibitory effect according to the tepal indices (Eq. 2), negative α had an 3 Journal of Plant Research (2018) 131:459–468 465 465 8 flowers with six tepals, -III increased, similar to ank of types 6-II, 6-IV, remained at 1, 2, and 3, t ith t l were more variable when α was small, as in the case when VT was larger. On the other hand, as α increased, the fraction of type 6-I increased, changing its rank among arrangements of six tepals from 3 to 2 and 1 (see α > − 0.2 in Fig. 4c), i t t ith i t d (Kit d F ji t effect similar to that of VT: For the flowers with six tepals, as α decreased, the fraction of type 6-III increased, similar to the case when VT increased, while rank of types 6-II, 6-IV, were more variable when α was small, as in the case when VT was larger. On the other hand, as α increased, the fraction of type 6-I increased, changing its rank among arrangements Fig. 4 Probability of perianth arrangements in spiral phyl- lotaxis model simulations. A model for spiral phyllotaxis with stochasticity reproduced limited variation of Anemone tepal arrangements A pair of stacked bars indicat- ing the arrangements at the appearance of sixth (left bar) and seventh tepal (right bar) is shown for each parameter value. The dependency on divergence angle φ (a), the growth length VT (b), and α (c) are shown. R0 = λ = 10, b = 10, VT = α = 0, and φ = 137.5° if the values are not specified 1 3 effect similar to that of VT: For the flowers with six tepals, as α decreased, the fraction of type 6-III increased, similar to the case when VT increased, while rank of types 6-II, 6-IV, and 6-I in six-tepaled arrangements remained at 1, 2, and 3, respectively (Fig. 4c), The arrangements with seven tepals were more variable when α was small, as in the case when VT was larger. On the other hand, as α increased, the fraction of type 6-I increased, changing its rank among arrangements of six tepals from 3 to 2 and 1 (see α > − 0.2 in Fig. 4c), consistent with a previous study (Kitazawa and Fujimoto effect similar to that of VT: For the flowers with six tepals, as α decreased, the fraction of type 6-III increased, similar to the case when VT increased, while rank of types 6-II, 6-IV, and 6-I in six-tepaled arrangements remained at 1, 2, and 3, respectively (Fig. 4c), The arrangements with seven tepals were more variable when α was small, as in the case when VT was larger. On the other hand, as α increased, the fraction of type 6-I increased, changing its rank among arrangements of six tepals from 3 to 2 and 1 (see α > − 0.2 in Fig. 4c), consistent with a previous study (Kitazawa and Fujimoto were more variable when α was small, as in the case when VT was larger. On the other hand, as α increased, the fraction of type 6-I increased, changing its rank among arrangements of six tepals from 3 to 2 and 1 (see α > − 0.2 in Fig. 4c), consistent with a previous study (Kitazawa and Fujimoto 1 3 Journal of Plant Research (2018) 131:459–468 466 parameter φ is likely to be less than 144 (Fig. 4a) in the field, and α is not a large positive number (Fig. 4c). A model for spiral phyllotaxis with stochasticity reproduced limited variation of Anemone tepal arrangements Within this parameter region, the paths towards type 7-II and type 7-IV were the most frequent in the simulation (Fig. 5, red lines); therefore, the frequent appearance of these two arrangements in association with the high frequency of type 6-I, type 6-II, and type 6-IV is consistent with the spiral phyllotaxis. Thus, we expect that if a population has a high frequency of type 6-III, then the population will exhibit a frequent appearance of type 7-III (Fig. 5, black lines). 2016b). In the arrangement with the seventh tepal, type 7-V was most dominant above a positive threshold value of α ~ 0.1, whereas the fraction of type 7-I subsequently increased to be the most dominant arrangement as α further increased. Of the parameter range examined, a positive α is the only condition for which the arrangement type 7-I appeared. In summary, a larger VT and smaller negative α cause greater variation in arrangements; therefore, the arrangements are not limited to a small number of types. A larger space between primordia caused by larger VT and a decrease of inhibition due to negative α both decrease the roughness of the energy landscape, resulting in gently slop- ing probability density over the edge of the floral meristem. Hence, limitation to small number of arrangement types implies that the development proceeds in compact packing of organ primordia. Possible developmental origins of the constrained variations When do these constrained arrangements emerge during floral development? In pentamerous Anemone species, tepal primordia initiate sequentially in a spiral order (e.g. A. taipaiensis, Fig. 1c and 7 in Ren et al. 2010), resulting in a quincuncial arrangement (type 5-I). On the other hand, in trimerous species (e.g. A. chinensis, Fig. 1f and 13 in Ren et al. 2010), three primordia emerge at the same time or in rapid succession, therefore it can be regarded as a trimer- ous whorl. Interestingly, in A. tomentosa with five or six tepals, tepal primordia initiate in a spiral order in a five- tepaled flower, whereas the primordia locate at the positions of trimerous double whorls (type 6-II) in a six-tepaled flower (Fig. 1d and 10 in Ren et al. 2010). Therefore, such stochas- tic appearance of trimerous whorls can occur by changing position and/or timing of primordia initiation. On the other hand, to our knowledge, the other arrangements presented here in mature flowers (types 6-I, 6-IV, 7-II, and 7-V) have not been reported at floral organ primordia initiation stage Conclusion In this study, we demonstrated that the arrangements of tepals in six- and seven-tepaled Anemone flowers are lim- ited to a small number of arrangement types in stark contrast with the intuitive notion that the variation would be greater due to an increasing number of possible states (arrange- ments). This limitation was explained by a mathematical model of spiral phyllotaxis. Our results indicate that the spi- ral nature underlies the stabilization process with increasing component (organ) number, yielding a canalization of organ arrangements (phyllotaxis) as development proceeds. Consistency of a stochastic spiral model with Anemone Both in field observations (Fig. 3) and numerical simulations (Fig. 4), we identified a clear bias of arrangement appear- ance in both six- and seven-tepaled flowers. In A. nikoensis and A. flaccida, type 6-I had a relatively high frequency in the flowers with six tepals (18.7 and 16.3%, respectively), whereas this frequency was low (less than 4%) in other species (Fig. 3b). These two species also showed higher frequencies of type 7-V among flowers with seven tepals compared with other species (39.8 and 16.4%, respectively; Fig. 3d). This increased frequency of type 6-I in association with the increase of type 6-IV was also found in the numeri- cal simulations with an α = − 0.1 (Fig. 4c). Therefore, we expect that with a small positive α, the older primordia have equal or slightly greater inhibitory effects on the initiation of a new primordium than newer primordia in these two species, whereas this effect appears to be smaller in other species. On the other hand, the present simulations did not account for a frequency of more than 90% for type 6-II in species A. soyensis, A. hepatica, and P. cernua (Fig. 3b), where the double-whorled arrangement was more frequent than quincuncial (Fig. 3a), suggesting that any other devel- opmental properties that were not incorporated into the pre- sent model regulate the stabilization of trimerous double whorls in Anemone. Exploring these mechanisms is an excit- ing future direction of floral phyllotaxis modelling. Acknowledgements This work supported by Grants-in-Aid for Scien- tific Research from MEXT to KF (16H01241, 17H06386). 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. Phyllotaxis of seven‑tepaled flowers is biased by those of six‑tepaled flowers Using the results of the numerical simulation, we can depict the possible pathways of development (Fig. 5). For exam- ple, type 7-VI can geometrically appear by adding one tepal inside the type 6-V (Fig. 5, dashed line), but this arrange- ment hardly appeared in simulations. Some paths appeared only for limited range of parameters. For example, the path from type 6-I to type 7-I only occurred when α was posi- tive (Fig. 5, blue solid line). As we observed three arrange- ments of six tepals [types 6-I, 6-II, and 6-IV (Fig. 3a, b)], the Fig. 5 Summary of the limited arrangement of five- (top), six- (middle), and seven- (bot- tom) tepaled flowers as they consistently appeared in Anemone and spiral phyllotaxis model (asterisks). Frame colors correspond to the simulation results in Fig. 4. The arrange- ment that was not found in the simulations is not boxed. Red paths indicate the arrange- ments that appear at φ < 144° and VT = α = 0, whereas a blue path indicates the arrangement appeared only at α > 0. The arrangements appeared at any other parameters are indicated by black paths, whereas those hardly appeared in simulations are indicated by dashed paths Fig. 5 Summary of the limited arrangement of five- (top), six- (middle), and seven- (bot- tom) tepaled flowers as they consistently appeared in Anemone and spiral phyllotaxis model (asterisks). Frame colors correspond to the simulation results in Fig. 4. The arrange- ment that was not found in the simulations is not boxed. Red paths indicate the arrange- ments that appear at φ < 144° and VT = α = 0, whereas a blue path indicates the arrangement appeared only at α > 0. The arrangements appeared at any other parameters are indicated by black paths, whereas those hardly appeared in simulations are indicated by dashed paths 1 3 3 Journal of Plant Research (2018) 131:459–468 467 in angiosperms. These arrangements would emerge at an early stage in floral development, or through the post-mer- istematic process, where internode length between succes- sive primordia are dynamically regulated (Kitazawa and Fujimoto 2015; Peaucelle et al. 2007). The developmental process of these arrangement types is an interesting future problem of experimental botany to estimate the importance of post-meristematic process on stabilizing floral structure. References Flora 217:64–74 1 3 468 Journal of Plant Research (2018) 131:459–468 Jiang N, Zhou Z, Yang J-B et al (2017) Phylogenetic reassessment of Refahi Y, Brunoud G, Farcot E et al (2016) A stochastic multicellular model identifies biological watermarks from disorders in self- organized patterns of phyllotaxis. Elife. https://doi.org/10.7554/ eLife.14093 Jiang N, Zhou Z, Yang J-B et al (2017) Phylogenetic reassessment of tribe Anemoneae (Ranunculaceae): non-monophyly of Anemone s.l. revealed by plastid datasets. PLoS One 12:e0174792 tribe Anemoneae (Ranunculaceae): non-monophyly of Anemone s.l. revealed by plastid datasets. PLoS One 12:e0174792 y p Kitazawa MS, Fujimoto K (2014) A developmental basis for stochastic- ity in floral organ numbers. Front Plant Sci 5:545 Ren YI, Chang H-L, Endress PK (2010) Floral development in Anemo- neae (Ranunculaceae). Bot J Linn Soc 162:77–100 yl g Kitazawa MS, Fujimoto K (2015) A dynamical phyllotaxis model to determine floral organ number. PLoS Comput Biol 11:e1004145 Sauquet H, von Balthazar M, Magallón S et al (2017) The ancestral flower of angiosperms and its early diversification. Nat Commun 8:16047f l Kitazawa MS, Fujimoto K (2016a) Relationship between the species- representative phenotype and intraspecific variation in Ranun- culaceae floral organ and Asteraceae flower numbers. Ann Bot 117:925–935 Schöffel K (1932) Untersuchungen über den blütenbau der Ranuncu- laceen. Planta 17:315–371l Kitazawa MS, Fujimoto K (2016b) Stochastic occurrence of trimery from pentamery in floral phyllotaxis of Anemone (Ranuncu- laceae). Acta Soc Bot Pol. https://doi.org/10.5586/asbp.3530 Schoute JC (1935) On corolla aestivation and phyllotaxis of floral phyl- lomes. Verh kon akad Wet Amsterdam Afd Natuurk 34:1–77 Soltis PS, Soltis DE, Zanis MJ, Kim S (2000) Basal lineages of angio- sperms: relationships and implications for floral evolution. Int J Plant Sci 161:S97–S107 ) p g p Mirabet V, Besnard F, Vernoux T, Boudaoud A (2012) Noise and robustness in phyllotaxis. PLoS Comput Biol 8:e1002389 Morgan AP (1874) Imbricative aestivation. Am Nat 8:705–713 p g p Mirabet V, Besnard F, Vernoux T, Boudaoud A (2012) Noise and robustness in phyllotaxis. PLoS Comput Biol 8:e1002389 robustness in phyllotaxis. PLoS Comput Biol 8:e1002389 Morgan AP (1874) Imbricative aestivation. Am Nat 8:705–713 p y p Morgan AP (1874) Imbricative aestivation. Am Nat 8:705–713 Wang P, Liao H, Zhang W et al (2015) Flexibility in the structure of spiral flowers and its underlying mechanisms. References Bentley R (1870) A manual of botany: including the structure, clas- sification, properties, uses, and functions of plants. Churchill, London Cunnell GJ (1958) Aestivation in Ranunculus repens L. New Phytol 57:340–352l Damerval C, Becker A (2017) Genetics of flower development in Ranunculales—a new, basal eudicot model order for studying flower evolution. New Phytol 216:361–366 l De Craene RLP (2010) Floral diagrams: an aid to understanding flower morphology and evolution. Cambridge University Press, Cambridgell De Craene RLP (2016) Meristic changes in flowering plants: how flow- ers play with numbers. Flora 221:22–37 Douady S, Couder Y (1996) Phyllotaxis as a dynamical self organizing process Part I: the spiral modes resulting from time-periodic itera- tions. J Theor Biol 178:255–273 l Although the number of possible arrangements increases as the number of tepals increases from six to seven, the number of actual arrangements decreased. Intuitively, the increase of components number is likely to correlate with an increase in variation. Our results suggest, however, that the process of development and increased component number can stabilize the overall structure. Our results also suggest a direct path from quincuncial arrangement to the arrange- ment of two cycles of trimerous whorls. One interesting unanswered question regarding floral evolution involves the path of development when the component number is fur- ther increased. For example, flowers with eight tepals may have an arrangement related to two cycles of tetramerous whorls. Understanding the stabilization of such an arrange- ment would illuminate the evolutionary pathway for precise determination and species diversity of floral part numbers. Endress PK, Doyle JA (2007) Floral phyllotaxis in basal angiosperms: development and evolution. Curr Opin Plant Biol 10:52–57 Endress PK, Doyle JA (2015) Ancestral traits and specializations in the flowers of the basal grade of living angiosperms. Taxon 64:1093–1116 Foster T, Johnston R, Seleznyova A (2003) A morphological and quantitative characterization of early floral development in apple (Malus × domestica Borkh.). Ann Bot 92:199–206 Gonçalves B, Nougué O, Jabbour F et al (2013) An APETALA3 homolog controls both petal identity and floral meristem pattern- ing in Nigella damascena L. (Ranunculaceae). Plant J 76:223–235i Hoot SB, Meyer KM, Manning JC (2012) Phylogeny and reclassifica- tion of Anemone (Ranunculaceae), with an emphasis on austral species. Syst Bot 37:139–152 Jabbour F, Nadot S, Espinosa F, Damerval C (2015) Ranunculacean flower terata: records, a classification, and some clues about floral developmental genetics and evolution. References Nat Plants 2:15188 Peaucelle A, Morin H, Traas J, Laufs P (2007) Plants expressing a miR164-resistant CUC2 gene reveal the importance of post-mer- istematic maintenance of phyllotaxy in Arabidopsis. Development 134:1045 l Yule GU (1902) Variation of the number of sepals in Anemone nemorosa. Biometrika 1:307–309 1 3 3
W2549437418.txt	https://link.springer.com/content/pdf/10.1007%2Fs00421-016-3499-3.pdf	en		Intraocular pressure and cerebral oxygenation during prolonged headward acceleration	European journal of applied physiology	2,016	cc-by	8,346	Eur J Appl Physiol (2017) 117:61–72 DOI 10.1007/s00421-016-3499-3 ORIGINAL ARTICLE Intraocular pressure and cerebral oxygenation during prolonged headward acceleration Ola Eiken1 · Michail E. Keramidas1 · Nigel A. S. Taylor2 · Mikael Grönkvist1 Received: 23 July 2016 / Accepted: 3 November 2016 / Published online: 11 November 2016 © The Author(s) 2016. This article is published with open access at Springerlink.com Abstract Purpose Supra-tolerance head-to-foot directed gravitoinertial load (+Gz) typically induces a sequence of symptoms/ signs, including loss of: peripheral vision—central vision— consciousness. The risk of unconsciousness is greater when anti-G-garment failure occurs after prolonged rather than brief exposures, presumably because, in the former condition, mental signs are not consistently preceded by impaired vision. The aims were to investigate if prolonged exposure to moderately elevated +Gz reduces intraocular pressure (IOP; i.e., improves provisions for retinal perfusion), or the cerebral anoxia reserve. Methods Subjects were exposed to 4-min +Gz plateaux either at 2 and 3 G (n = 10), or at 4 and 5 G (n = 12). Measurements included eye-level mean arterial pressure (MAP), oxygenation of the cerebral frontal cortex, and at 2 and 3 G, IOP. Results IOP was similar at 1 (14.1 ± 1.6 mmHg), 2 (14.0 ± 1.6 mmHg), and 3 G (14.0 ± 1.6 mmHg). During the G exposures, MAP exhibited an initial prompt drop followed by a partial recovery, end-exposure values being reduced by ≤30 mmHg. Cerebral oxygenation showed a Communicated by David C. Poole. * Ola Eiken ola.eiken@sth.kth.se 1 Department of Environmental Physiology, Swedish Aerospace Physiology Centre, School of Technology and Health, KTH, Royal Institute of Technology, Berzelius v 13, SE‑17165 Stockholm, Sweden 2 Centre for Human and Applied Physiology, School of Medicine, University of Wollongong, Wollongong, NSW, Australia similar initial drop, but without recovery, and was followed by either a plateau or a further slight decrement to a minimum of about −14 μM. Conclusion Gz loading did not affect IOP. That cerebral oxygenation remained suppressed throughout these G exposures, despite a concomitant partial recovery of MAP, suggests that the increased risk of unconsciousness upon G-garment failure after prolonged +Gz exposure is due to reduced cerebral anoxia reserve. Keywords Cerebral anoxia reserve · Cerebral blood flow · G tolerance · G-induced loss of consciousness · Oxyhaemoglobin saturation · Retinal anoxia reserve Abbreviations ABD With the full anti-G suit (including the abdominal bladder) pressurised AGE 39 Anti-G ensemble used in the Gripen 39 aircraft AGS Anti-G suit A-LOC Almost G-induced loss of consciousness DAP Diastolic arterial pressure ∆[Hb] Change in deoxygenated haemoglobin ∆[O2Hb] Change in oxygenated haemoglobin ∆[THb] Change in total haemoglobin G Dimensionless quantity denoting the ratio between the vector sum of gravitational and inertial forces and Earths gravity G-LOC G-induced loss of consciousness +Gz Gravitoinertial force in the head-to-foot direction Hb Haemoglobin HR Heart rate IOP Intraocular pressure MAP Mean arterial pressure NIRS Near infrared spectroscopy 13 62 NoABD With the anti-G suit (excluding the abdominal bladder) pressurised NoAGS Without the anti-G suit SACM Simulated aerial combat manoeuvres SAP Systolic arterial pressure SpO2 Capillary oxyhaemoglobin saturation Introduction A major challenge for pilots flying high-performance aircraft is to maintain adequate retinal and cerebral blood flow during headward acceleration. Thus, high gravitoinertial load in the head-to-seat direction (+Gz; henceforth, G denotes +Gz unless otherwise stated) creates exaggerated intravascular pressure gradients. At 9 G, for instance, heart-level arterial pressure needs to be elevated to 225– 275 mmHg to preserve blood perfusion of the brain and retina. The pilots’ arterial pressure is increased by muscular straining manoeuvres in combination with pressurisation of the anti-G suit and, in modern high-performance aircraft, of the breathing gas (Green 2016). Upon a gradual increase in G load, a sequence of symptoms and signs evolves that determines the G-tolerance threshold; the sequence starts with loss of peripheral vision, is followed by loss of central vision and eventually by loss of consciousness (G-LOC). That critical ischaemia is attained at lower G load for the retina than for the cerebral cortex is attributable to two mechanisms. First, to perfuse the retina, arterial pressure at eye level must be sufficient to overcome intraocular pressure (IOP) (Duane 1954; Lambert and Wood 1946; Newsom and Leverett 1968). Second, at increased G load, the “siphone effect” within intracranial vasculature is capable of facilitating perfusion of the cerebral cortex even when local arterial pressure drops to zero or sub-zero levels (Henry et al. 1951), thereby sustaining adequate tissue oxygenation. To date, the vast majority of studies regarding the influence of excessive G loads on vision and mental performance have been conducted with the G load being elevated from about 1 G to a level exceeding the tolerance threshold. In such conditions, the period of severe retinal or cerebral ischaemia is preceded by normal tissue oxygenation. We have, however, found that the sequence of signs and symptoms upon exposure to a supra-tolerance G load may be affected by the immediate G history preceding the high-G exposure. Thus, experienced fighter pilots were exposed to supra-tolerance G loads either by removing pressure in the G-protective system (anti-G suit and breathing mask) after a prolonged period (2 min) of sustained high-G loading or by rapidly applying such G loading, without pressurising the G-protective system; the pilots were instructed to terminate the G exposure 13 Eur J Appl Physiol (2017) 117:61–72 upon loss of peripheral vision. The risk of losing consciousness was several-fold higher when the pressure in the G-protective system was lost after 2 min of high-G loading, than when pressure failed to increase in conjunction with the onset of high G (Eiken and Grönkvist 2013). Presumably, the increased risk of G-LOC in the former condition is due to the fact that after a period of elevated but tolerable G load, warning symptoms in terms of reduced vision do not as distinctly precede G-LOC. This has obvious practical implications, since fighter pilots commonly use the loss of peripheral vision to gauge the tolerable G load. That a prolonged period of high-G load alters the tissue most susceptible to G-induced ischaemia from the peripheral retina to the cerebral cortex may, in turn, be either due to improved provisions for retinal perfusion (i.e., reduced IOP), or to a more pronounced drop of the tissue oxygen reserve in the cerebral cortex than in the retina. Accordingly, the aims of the present study were to investigate if prolonged exposure to moderately elevated G force reduces either the IOP or the cerebral anoxia reserve. These possibilities were addressed in two separate series of experiments. For practical and safety reasons, IOP was determined at ≤3G (series 1), whereas cerebral oxygenation was predominantly investigated at 4 and 5 G (series 2), because it is known that such G loading may induce substantial hypoxaemia (Barr 1963). Since the anti-G suit, and in particular its abdominal bladder, appears to aggravate the G-induced hypoxaemia (Barr 1963; Eiken et al. 2011), we investigated effects of prolonged G exposures with and without the abdominal bladder included in the suit inflation. Cerebral oxygenation was investigated using a transcranial Near Infrared Spectroscopy (NIRS) technique; the relative drop in frontal cortex oxygenation, as determined by NIRS, has been shown to be a valid risk predictor of G-LOC (Benni et al. 2003; Ryoo et al. 2004; Tripp et al. 2009). Methods Two series of experiments were performed, first, with exposures at 2 and 3 G, addressing the question of changes in IOP and second, with exposures at 4 and 5 G, focussing on the question of cerebral anoxia reserve (Fig. 1). Altogether, 17 healthy individuals (16 men and one woman) participated as subjects, five of whom took part in both experimental series. They gave their written, informed consent prior to enrolling, and were aware that they were free to withdraw from the study at any time. The protocol and experimental procedures were approved by the Regional Human Ethics Committee in Stockholm, Sweden. The experiments were performed in the 7.25-m radius human-use centrifuge (ASEA, Västerås, Sweden) at the Eur J Appl Physiol (2017) 117:61–72 63 (a) Latin-square order 2 Gz 1 Gz 4-min 4-min 3 Gz 2 Gz 1 Gz 3-min Interval ~10-min 1 Gz 4-min 4-min 1 Gz 3-min (b) 4-min 4-min 4 Gz 4-min 4-min 4-min 1 Gz 1 Gz 3-min Interval ~10-min 1 Gz 4-min 4-min 1 Gz 3-min Latin-square order 4 Gz 1 Gz Interval ~10-min 3 Gz 1 Gz Interval 3-min ~10-min 1 Gz 4-min 1 Gz Interval 3-min ~10-min 1 Gz 4-min 5 Gz 5 Gz 4-min 4-min 1 Gz Interval 1 Gz 1 Gz 3-min ~10-min 4-min 3-min Fig. 1 Schematic illustration of the experimental design. The shaded zones correspond with the treatment periods (G loading), with the light shading referring to complete anti-G-suit (AGS) pressurisation, and the dark shading identifying either no AGS pressurisation (a), or partial AGS pressurisation (no abdominal bladder; b) Royal Institute of Technology in Stockholm. During each experiment, the subject was seated in the centrifuge gondola, in a “mock” of the Gripen 39 seat, which has a back angle reclining 28° from the vertical. The centrifuge was controlled by an “open-loop” system, employing pre-set G-time profiles. Gz was measured using an analogue accelerometer mounted in the gondola at a position approximately corresponding to the vertical level of the subjects’ heart. Throughout each experiment, the subject was monitored via closed-circuit television, and was able to communicate with the experimenter by means of a two-way intercom system. The anti-G ensemble used in the Gripen 39 aircraft (AGE 39) has been described in detail elsewhere (Eiken et al. 2007, 2011). It comprises a pressure-breathing system and an extended coverage pneumatic anti-G suit, which contains a single-compartment inflatable bladder, fully covering the lower abdomen as well as the legs, from below the gluteal region and the inguinal ligaments to the ankles. For the purpose of the present experiments, a modified version of the anti-G suit was used, in which it was possible to pressurise the abdominal and leg portions of the bladder separately (cf. Eiken et al. 2011). As in the aircraft, pressurisation of the anti-G suit commenced at 2 G, with pressure increasing linearly with increasing G load to reach a maximum of 19.1 kPa (143 mmHg) at 5.0 G; at G loads ≤2 G, the suit was supplied with a “ready pressure” of 1.3 kPa. Pressurisation of the airways commenced at 4 G, with a stipulated pressure of 1.34 kPa at 5 G. AGE pressures were controlled through a G valve/breathing regulator (Eros, F-5341, Eros, Plaisir Cedex, France). Experimental protocols Series 1 Ten subjects (nine men and one woman; mean ± SD age: 36 ± 14 years, height: 181 ± 6 cm, body mass: 74 ± 13 kg) participated. On a separate occasion, preceding the experimental day, each subject was trained in performing the IOP measurements in the centrifuge gondola at 1 G (stationary), using the procedure described below. On the experimental day, the subject was instrumented and then positioned in the gondola wearing the anti-G suit. He/ she was investigated in four conditions: 2 G, once with the full AGS (2-G AGS) and once without the AGS pressurised (2-G NoAGS), and at 3 G with (3-G AGS) and without pressurised AGS (3-G NoAGS). The order of these conditions was alternated and balanced among subjects in a Latin-square manner. Each experimental trial started with 4 min of baseline measurements at 1 G. Thereafter, the G load was increased by 0.5 G/s to a plateau of either 2 or 3 G. After 4 min at the plateau, the load was decreased to 1 G by 0.5 G/s, and measurements continued for another 3 min. Successive trials were interspersed by at least a 10-min recovery period. Series 2 Twelve male subjects (age: 30 ± 10 years, height: 183 ± 4 cm, body mass: 81 ± 7 kg) participated. Each subject was instrumented and equipped with the complete AGE 39. He was investigated in four conditions: 4 Gz, once 13 64 with the full AGS (including the abdominal bladder) pressurised (4-G ABD) and once without the abdominal bladder pressurised (4-G NoABD), and at 5 G with and without the abdominal bladder pressurised (5-G ABD and 5-G NoABD, respectively). Note that in the 5-G exposures the breathing gas was also pressurised (see above). The order of the conditions was similarly alternated among subjects in a balanced fashion. Each G exposure started with 4 min of baseline measurements at 1 G, followed by a G-load increase (0.5 G/s) to a 4-min G plateau of 4 or 5 G. After 4 min, the load was decreased to 1 G by 0.5 G/s, and measurements continued for another 3 min. The G exposures were interspersed by at least 10 min of recovery. Physiological measurements Heart rate (HR) and arterial pressure HR was derived from electrocardiographic recordings using a cardiometer (Datex-Engström, Instrumentation Corp, Helsinki, Finland), and with the electrodes positioned in a precordial 5-lead arrangement. Systolic (SAP), diastolic (DAP) and mean (MAP) arterial pressures were measured using a volumeclamp technique (Portapres, TNO, Amsterdam, The Netherlands) with the pressure cuff placed around the middlephalanx of the third finger of the right hand, and with the reference pressure transducer taped to the skin of the temple, at the level of the eyes. The right arm was supported by an armrest adjusted, so that the distal portions of the fingers were at a vertical level corresponding to the jugulum sterni. Tonometry IOP was measured using a non-contact tonometer (Pulsair, Intellpuff, Keeler, UK), that was mounted on a custom-made horizontal slide, positioned in the centrifuge gondola with the tonometer transducer at the vertical and transversal levels of the subject’s left eye. The subject could move the tonometer in the sagittal direction, along a frictionless track; the tonometer was spring-loaded, so that it retracted away from the face when the subject released his/her grip of it. The tonometer was provided with a video camera (Plexgear USB Microscope) and a miniature video screen (7 inch TFT LCD Monitor), enabling the subject, using his/her right eye, to place the tonometer in the correct position for IOP measurement of his/her left eye. IOP readings were monitored via a standard CCD video camera. The subject was instructed to conduct as many IOP measurements as possible during the course of each intervention. Typically, due to nystagmus, IOP measurements could not be performed during, and for a few seconds following, each G transition. Capillary oxyhaemoglobin saturation (SpO2) SpO2 was measured continuously throughout all trials using a pulse oxymeter (Nellcor Puritan Bennett Inc., Pleasanton, CA, USA), with the transducer placed on the second finger of 13 Eur J Appl Physiol (2017) 117:61–72 the right hand. The oxymeter has an accuracy of ±2% units across the range 70–100% and an acceptable resilience to motion artefacts. Cerebral oxygenation Throughout each experiment, the oxygenation of the frontal cerebral cortex was measured using continuous-wave NIRS (NIRO-200NX, Hamamatsu, Japan). The transducer unit was positioned over the left prefrontal cortex between the first frontal-polar (Fp1) and the third frontal (F3) locations, as determined using the modified international 10–20 system for electroencephalograms. To minimise confounding influence of skin blood flow, the unit, which consists of an emitter and a detector, was taped to the skin at a fix inter-optode distance of 4.0 cm (Hampson and Piantadosi 1988). To reduce intrusion of external light and loss of transmitted NIR light from the measuring area, the transducer unit was covered with an opaque bandage. The NIR light is comprised of three wavelengths (735, 810 and 850 nm), and changes in tissue oxygenation were calculated as oxygenated (∆[O2Hb]) and deoxygenated (∆[HHb]) haemoglobin (Hb), respectively. In addition, total Hb (∆[THb]), which is the sum of ∆[O2Hb] and ∆[HHb], and reflects changes in regional blood volume (Van Beekvelt et al. 2001), was calculated continuously. The theory, limitations, and reliability of cerebral oxygenation obtained employing NIRS have been reviewed elsewhere (Boushel et al. 2001; Ferrari et al. 2004). The NIRS signal was recorded at 5 Hz and expressed relative to the 4-min baseline period preceding each trial. Analyses Data obtained during and after the G plateaux were averaged every 15 s. However, because of large inter-individual variability regarding the number of IOP measurements during the G exposures, IOP values were averaged for each min within subjects. Statistical analyses were performed using Statistica 8.0 (StatSoft, Tulsa, OK, USA). All data are reported as mean ± SD, unless otherwise indicated. Analysis of the normal distribution of the data was performed with the Kolmogorov–Smirnov test. The statistical significance of differences was evaluated by a two-way (condition × time) general linear model repeated measures analysis of variance (ANOVA). Mauchly’s test was conducted to assess the sphericity, and the Greenhouse-Geisser correction was used to adjust the degrees of freedom when the assumption of sphericity was not satisfied. The Tukey honestly significant difference post hoc test was employed to identify specific differences between means when ANOVA revealed a significant F ratio for interaction or main effects. Probabilities <0.05 were regarded as being statistically significant. Eur J Appl Physiol (2017) 117:61–72 Results 65 (a) 80 2G NoAGS 2G AGS Arterial pressure and heart rate In all trials, MAP dropped instantly upon G exposure, reaching a nadir value ~15 s into the G plateau (Fig. 2a). That drop was followed by a gradual increase with a complete recovery (p > 0.05) to baseline MAP during both trials conducted with the AGS pressurised (after ~120 s at 2 G and ~75 s at 3-G AGS). Without suit pressurisation (NoAGS trials), only a partial recovery was observed (p < 0.05). Throughout the G plateau, MAP was consistently lower in the 3-G NoAGS than in the other trials (p < 0.001). Upon returning to 1 G, MAP promptly reverted to baseline values in all trials (p > 0.05). During the G exposure, SAP decreased in both NoAGS trials, but remained unchanged in the AGS trials (Table 1). Each G exposure reduced DAP and elevated HR (p < 0.05), with changes being more pronounced in the 3-G NoAGS than in the other trials (p < 0.01; Table 1). Capillary oxyhaemoglobin saturation SpO2 remained unchanged during the 2-G NoAGS trial (p > 0.05; Fig. 2b). During the 3-G AGS trial, SpO2 dropped and levelled off within ~75 s; whereas in the 2-G AGS and 3-G NoAGS trials, SpO2 decreased to a stable level within ~90 s (p < 0.05). Following those G exposures, SpO2 gradually recovered in a trial-dependent manner, such that it had resumed baseline values within ~30, ~45, and ~90 s in the 2-G AGS, 3-G NoAGS, and 3-G AGS trials, respectively. During the G plateau and the initial part of recovery, the reduction in SpO2 was substantially greater in the 3-G AGS than in the other trials (p ≤ 0.01). SpO2 was also lower in the 3-G NoAGS than in the 2-G NoAGS trial (p = 0.02). Cerebral oxygenation In all trials, Δ[O2Hb] dropped instantly upon G exposure, reaching a nadir value ~15 s into the plateau period (p < 0.01; Fig. 2c). During the remaining course of the G plateau, Δ[O2Hb] did not exhibit any further significant change in any of the trials (p > 0.05). Throughout the G plateau, Δ[O2Hb] was consistently lower in the 3-G NoAGS than in the other trials (p < 0.05), and lower in the 3-G AGS than in both 2-G trials (p ≤ 0.05). Upon return to 1 G, Δ[O2Hb] rebounded promptly; during the recovery phase, Δ[O2Hb] was greater in the 3-G NoAGS trial than in the other trials (p ≤ 0.05). During the G exposure, Δ[HHb] did not change in any of the trials (p > 0.05; Table 1). Δ[THb] decreased by ~135, ~303. and ~261% in 2-G NoAGS, 3-G NoAGS, and 3-G AGS trials, respectively (p < 0.05; Table 1). Δ[THb] was reduced by ~134% in the 2-G AGS trial, although that MAP(mmHg) (mmHg) MAP 3G NoAGS 3G AGS 60 50 40 * * * * * * * * * ** * * * * 30 * (b) 100 98 SpO22 (%) (%) SpO Series 1 70 96 94 92 (c) 7 * 5 * * * * * * 3 [OHbO M) 2Hb] 2 (( M) All experiments were conducted without adverse events; none of the subjects experienced G-LOC. 1 -1 -3 -5 -7 * B T * * *** * * * * 4-min Plateau 3-min Recovery Fig. 2 Mean (SE) arterial pressure (MAP; a), capillary oxyhaemoglobin saturation (SpO2; b) and changes from resting values in cerebral oxyhaemoglobin (Δ[O2Hb]; c) obtained during the 4-min G plateaux at 2 and 3 G, and the 3-min recovery with (AGS) and without (NoAGS) anti-G-suit pressurised. B baseline phase, T transition (G onset) phase. Significantly different from 2-G NoAGS, 2-G AGS, and 3-G AGS, †significantly different from 2-G NoAGS, ‡significantly different from 2-G AGS, §significantly different from 2-G NoAGS, 2-G AGC, and 3-G NoAGS (p ≤ 0.05). n = 10 difference was not statistically significant (p = 0.88). The drop in Δ[THb] was greater in the 3-G NoAGS than in the 3-G AGS trial (p < 0.001). Intraocular pressure No statistically significant difference in IOP was detected between 1-G baseline periods and the 4-min periods at the 2- and 3-G plateaux (p = 0.09; Table 2). No difference was observed between the trials (p = 0.68; Table 2). 13 126 ± 24 49 ± 10 77 ± 15 1.26 ± 1.83 −3.48 ± 5.28 3.06 ± 4.45 Series 2 −4.77 ± 5.04§ 6.11 ± 5.23 0 123 ± 26 125 ± 23 124 ± 26† 47 ± 7 45 ± 8 34 ± 11 79 ± 16 71 ± 12 89 ± 11* 2.44 ± 2.88 0 −0.64 ± 2.94 0 122 ± 27 119 ± 18 114 ± 24 122 ± 22 125 ± 18 107 ± 29* 46 ± 8 45 ± 8 35 ± 9* 46 ± 8 46 ± 5 22 ± 9‡ †† 75 ± 14 69 ± 11 79 ± 13 73 ± 13 71 ± 11 103 ± 13§ 0.76 ± 1.83 0 −0.58 ± 1.91 −0.30 ± 1.18 0 −0.20 ± 2.89 B G plateau 13 Significantly different from 2-G NoAGS, 2-G AGS, and 3-G NoAGS Significantly different from 2-G NoAGS, 2-G AGS, and 3-G AGS Significantly different from 3-G AGG Significantly different from 2-G AGS and 3-G NoAGS Significantly different from 3-G NoAGS and 3-G AGS ‡ § # †† Significantly different from the baseline and recovery n = 10 † −1.62 ± 3.55 0 −3.51 ± 4.10* 2.49 ± 3.57 Δ[THb] (μM) SAP (mmHg) DAP (mmHg) HR (beats min−1) Δ[HHb] (μM) 0 127 ± 22 114 ± 29* 46 ± 9 33 ± 11* 72 ± 12 86 ± 12# 0 −1.96 ± 3.10 B G plateau B 2-G NoAGS R 2-G AGS G plateau R 0.47 ± 1.91 B 3-G NoAGS R 3-G AGS G plateau R Eur J Appl Physiol (2017) 117:61–72 Table 1 Mean (± SD) values of systolic (SAP) and diastolic (DAP) arterial pressure, heart rate (HR), and changes in cerebral frontal cortex deoxyhaemoglobin (Δ[HHb]) and total haemoglobin (Δ[THb]) obtained during the 4-min baseline (B), the 4-min plateau at 2 and 3 G, and the 3-min recovery (R), with (AGS) and without (NoAGS) the anti-G-suit pressurised 66 Arterial pressure and heart rate In all trials, MAP dropped instantly upon G exposure, reaching a nadir value ~15 s into the G plateau phase (p < 0.001; Fig. 3a). Thereafter, MAP gradually recovered, especially in the ABD trials, but remained below baseline values throughout the G plateau (p < 0.001). MAP was higher in the 4-G ABD than in either of the other trials (p < 0.05). It was also slightly higher in the 5-G ABD than in the 5-G NoABD trial, although that difference was not significant (p = 0.61). During the recovery phase, MAP reverted to baseline values within ~45 s, and no differences were observed between the trials (p > 0.05). During the G exposures, SAP and DAP were substantially reduced (p ≤ 0.05); no differences were observed between the trials (p > 0.05; Table 3). HR was elevated in all trials, and that increase was greater in the 5-G than the 4-G trials (p ≤ 0.001; Table 3). Capillary oxyhaemoglobin saturation In all trials, SpO2 dropped and levelled off within ~60 s of high-G onset (p < 0.001), and recovered within ~90 s upon returning to 1 G (Fig. 3b). SpO2 was lower in the 5-G NoABD than in the 4-G NoABD trial (p = 0.02). Pressurisation of the AGS abdominal bladder aggravated the G-induced reduction in SpO2, with SpO2 being substantially lower in the 5-G ABD than in the other trials (p < 0.001). It was also lower in the 4-G ABD than in the 4-G NoABD and 5-G NoABD trials (p < 0.01). Cerebral oxygenation In all trials, Δ[O2Hb] dropped promptly upon each G exposure (p < 0.001; Fig. 3c), and the reduction was more pronounced at 5 G than at 4 G (p < 0.001). During both 4-G trials and also during the 5-G NoABD trial, Δ[O2Hb] dropped and levelled off within 15 s of the G onset. In the 5-G ABD trial, however, the initial drop in Δ[O2Hb] was followed by a further gradual decrease with the nadir value being attained at the end of the G exposure. Upon return to 1 G, Δ[O2Hb] rebounded within ~30 s in all trials; yet, in the 5-G ABD trial, the recovery of Δ[O2Hb] was slightly slower than in the 4-G trials (p < 0.05). During the recovery phase, Δ[O2Hb] was higher following 5-G NoABD than following the other trials (p < 0.05). During these G exposures, Δ[HHb] did not change in any of the trials (p > 0.05; Table 3). Δ[THb] was substantially reduced in all trials (p < 0.001), and especially within the 5-G trials (p < 0.01; Table 3). Discussion Present results demonstrated that IOP was but minimally reduced by prolonged exposures to moderately elevated Eur J Appl Physiol (2017) 117:61–72 67 Table 2 Mean values ± SD as well as (total number of measurements) of intraocular pressure obtained during the 4-min baseline, the 4-min plateau at 2 and 3 G, and the 3-min recovery period, with (AGS) and without (NoAGS) the anti-G-suit pressurised Baseline 2-G NoAGS 2-G AGS 3-G NoAGS 3-G AGS 14.1 ± 1.6 (291) 14.1 ± 1.6 (320) 14.1 ± 1.6 (362) 14.1 ± 1.7 (308) G plateau Recovery 1 min 2 min 3 min 4 min 14.1 ± 1.6 (61) 14.1 ± 1.6 (52) 14.1 ± 1.6 (36) 14.1 ± 1.57 (89) 14.0 ± 1.56 (74) 14.1 ± 1.58 (70) 14.0 ± 1.6 (91) 14.0 ± 1.6 (67) 14.1 ± 1.6 (65) 14.0 ± 1.6 (94) 14.1 ± 1.6 (101) 14.0 ± 1.6 (80) 14.1 ± 1.7 (51) 14.1 ± 1.66 (84) 14.1 ± 1.6 (69) 14.1 ± 1.6 (69) 14.2 ± 1.6 (239) 14.2 ± 1.6 (210) 14.1 ± 1.6 (220) 14.1 ± 1.7 (207) n = 10 gravitoinertial load in the head-to-foot direction. By contrast, each high-G exposure induced a prompt reduction in cerebral oxygenation, the magnitude of which was dependent both on the size of the step-change in G load and on whether, and what kind of, G-protective garment the subject was using. In the 4- and 5-G exposures, the initial depression of cerebral oxygenation prevailed at a steady level, or oxygenation even tended to drop further during the course of the 4-min G plateaux. This response occurred despite the fact that the initial decline in local arterial pressure was followed by a partial recovery during the latter two-thirds of the plateau. These results should be viewed in the context of our previous study (Eiken and Grönkvist 2013), which showed that the risks of near and actual loss of consciousness are several-fold greater if pressure in the anti-G system is lost after a 2-min period of sustained high-G load than if the system fails to elevate pressure during onset of such G loading, presumably because in the former condition loss of consciousness is not always preceded by impeded vision. This raises the question of whether prolonged exposure to elevated but tolerable G loads either reduces the risk of G-induced loss of vision, or increases the risk of G-induced cerebral dysfunction. Intraocular pressure at increased G load Judging from our finding that the IOP was virtually unaffected by a threefold increase in G load, it appears unlikely that prolonged exposure to high, but tolerable, G loads might reduce the susceptibility to decrements in peripheral and central vision upon exposure to supra-tolerance loads. Thus, it has long been recognised that the G-induced loss of peripheral and foveal visions, commonly referred to by pilots as “grey out” and “black out”, respectively, are caused by retinal ischaemia resulting from insufficient perfusion pressure in intraocular blood vessels (Duane 1954; Lambert and Wood 1946; Newsom and Leverett 1968); once MAP drops below IOP, retinal perfusion ceases (Riva et al. 1986). A sustained reduction of arterial pressure may, in turn, reduce IOP (Mitchell et al. 2005), but the magnitude of any IOP drop induced by a sustained decrease in arterial pressure, will amount to a mere fraction of the actual drop in arterial pressure (Mitchell et al. 2005; Nicholson et al. 1968). Consequently, the only conceivable period during which an MAP-induced drop in IOP might predispose to increased ocular perfusion pressure, is the transient period between regaining MAP and the ensuing readjustment of the IOP to the elevated MAP. At the outset of the present experiments, we reasoned that improved provisions for retinal blood flow might occur during prolonged G exposure, either by way of a transient overshoot in retinal perfusion pressure following the recovery of MAP after its transient initial drop, or by another unforeseen mechanism. It appears, however, that IOP is rather resilient to G-induced decrements in local arterial pressure. Thus, even at 3 G without pressurised anti-G suit, IOP exhibited a minimal drop, despite MAP stabilising at about 20 mmHg below the baseline (1-G) level. Whether the G-induced drops in MAP were of sufficient magnitude and duration to significantly affect IOP remains to be established. Bakke et al. (2009) showed that a 2-min isometric exercise bout resulting in an MAP increase from 80 to 120 mmHg was paralleled by a 4-mmHg increase in IOP, whereas a decrease in local MAP of 20–25 mmHg accompanying a postural change from recumbent to sitting results in a rapid decrease in IOP (Krieglstein and Langham 1975), that may vary considerably in magnitude, with average changes reported in different studies ranging from 0.3 to 2.9 mm Hg (cf. Anderson and Grant 1973; Jain and Marmion 1976; Krieglstein and Langham 1975). In addition, long-term changes in arterial pressure appear to result in IOP changes of similar magnitude, with a 1-mmHg change in IOP for every 30-mmHg change in SAP (Mitchell et al. 2005). The reason for the large inter-study variations in IOP response to a 20–25 mmHg drop in local arterial pressure remains to be settled, as does the reason for the minute IOP change in response to a 20-mmHg drop in local MAP in the present 3-G exposures. The non-contact tonometry technique used in the present study is regarded a reproducible and sensitive means of measuring IOP (Almubrad and Ogbuehi 2010; Jain and Marmion 1976). 13 13 1.50 ± 6.55 0.99 ± 2.98 0.87 ± 6.43 −15.46 ± 11.04† 0 0 Significantly different from 2G-NoAGS, 2G-AGS, and 3G-NoAGS † * Significantly different from the baseline and recovery 0 n = 12 0 1.11 ± 2.75 −8.05 ± 4.52* 3.04 ± 2.66 5.37 ± 5.52 0 1.21 ± 2.77 −7.12 ± 4.66* 87 ± 13 1.72 ± 2.72 2.96 ± 5.64 81 ± 10 1.58 ± 5.48 5.31 ± 7.03 10.04 ± 14.14 −15.23 ± 10.95† 0 83 ± 11 96 ± 12 119 ± 16† 0 47 ± 13 93 ± 11 119 ± 14† 120 ± 18 16 ± 13 103 ± 23* 45 ± 11 121 ± 20 127 ± 23 52(15) 17 ± 25* 106 ± 31* 46 ± 12 121 ± 16 123 ± 21 R G plateau B R 48 ± 9 109 ± 15* 0 The second main aim of the present experiments was to investigate the effects of sustained moderate elevations of G load on the cerebral anoxia reserve. Oxygen delivery to the frontal cortex is determined by the local blood flow and the arterial oxygen content. As expected, all sustained G-load Δ[THb] (μM) Cerebral oxygenation at increased G load Δ[HHb] (μM) Fig. 3 Mean (SE) arterial pressure (MAP; a), capillary oxyhaemoglobin saturation (SpO2; b) and changes from resting values in cerebral oxyhaemoglobin (Δ[O2Hb]; c) obtained during the 4-min G plateaux at 4 and 5 G, and the 3-min recovery, with (ABD) and without (NoABD) the abdominal bladder of the anti-G-suit inflated. B baseline phase, T transition (G onset) phase. Significantly different from 4-G NoABD and 4-G ABD, †significantly different from 4-G NoABD, 4-G ABD, and 5-G NoABD, ‡significant difference between 4-G NoABD and 4-G ABD, §significant difference between 4-G NoABD and 5-G NoABD, ¶significant difference between 4-G ABD and 5-G NoABD, significant difference between 4-G ABD and 5-G ABD, ††significant difference between 5-G NoABD and 5-G ABD; (p ≤ 0.05). n = 12 80 ± 14 3-min Recovery 88 ± 11 4-min Plateau 110 ± 14 * * * * ** * * * ** * *** 81 ± 13 B T HR (beats·min−1) * -16 24 ± 11* ** * * ** * * * ** * * ** * -12 109 ± 23* * 44 ± 8 -8 121 ± 13 * 44 ± 10 * * 124 ± 23 * -4 14 ± 10* [O M) HbO 2Hb] 2 ( (M) 0 105 ± 19* *** * 4 42 ± 8 * * * 123 ± 17 * * * 8 DAP (mmHg) (c) SAP (mmHg) 80 G plateau 85 B 90 R SpO (%) SpO (%) 2 2 95 G plateau (b) 100 B 15 R 25 G plateau 35 B 45 5-G ABD MAP (mmHg) MAP (mmHg) 5G ABD 55 5-G NoABD 4G ABD 5G NoABD 65 4-G ABD 4G NoABD 75 4-G NoABD (a) Eur J Appl Physiol (2017) 117:61–72 Table 3 Mean (±SD) values of systolic (SAP) and diastolic (DAP) arterial pressure, heart rate (HR), and changes in cerebral frontal cortex deoxyhaemoglobin (Δ[HHb]) and total haemoglobin (Δ[THb]) obtained during the 4-min baseline (B), the 4-min plateau at 4 and 5 G, and the 3-min recovery (R), with (ABD) and without (NoABD) the abdominal bladder of the anti-G-suit inflated 68 Eur J Appl Physiol (2017) 117:61–72 increments induced an immediate drop in eye-level MAP, which then partially recovered, but nevertheless stabilised considerably below its 1-G baseline value, during the latter two-thirds of the G plateaux. By contrast, frontal cortex oxygenation (Δ[O2Hb]) decreased promptly upon high-G exposure and did not recover during the ensuing course of the G plateau; during the 5-G exposure, with pressurisation of the full AGS, including the abdominal bladder, the initial drop in frontal cortex oxygenation was even followed by a slight but further drop in oxygenation. That Δ[O2Hb] remained reduced throughout the G exposure may be attributed to two mechanisms, reduced local blood flow and hypoxaemia. In the 4- and 5-G trials, the steady-state eye-level MAP varied between 30 and 45 mmHg and was dependent not only on the G load but also of whether the G-suit abdominal bladder was pressurised, with a distinctly higher MAP at a given G load, when the bladder was pressurised. Cerebral blood flow is predominantly autoregulated, vascular myogenic tone being governed by a complex interplay between responses to local changes in transmural pressure and the chemical environment, but also being modified by regional vascular conducted responses (cf Jensen and Holstein-Rathlou 2013). Particularly in dynamic conditions, the interplay between autoregulatory mechanisms is not fully understood (cf. Panerai et al. 2002). Regardless, it is clear that the cerebral vasculature regulates flow across a wide range of perfusion pressures, whereas under 1-G conditions, its autoregulatory capacity is exceeded once local arterial pressure drops below about 60 mmHg (Testart 1990). Thus, it is reasonable to assume that in the present 4- and 5-G exposures, cerebral blood flow remained reduced throughout the high-G plateaux. The reduction in cerebral blood flow resulting from the G-induced drop in local arterial pressure might have been modulated by two mechanisms, acting in opposite directions. On the one hand, at increased G loads, cerebral perfusion pressure is influenced by the siphon effect (Henry et al. 1951). Since intracranially, venous walls are not as collapsible as in other regions of the body, local venous pressure may assume subatmospheric levels during headward acceleration. Hence, antegrade flow may prevail in cerebral vasculature even in the face of a G-induced drop in local arterial pressure to or below 0 mmHg (Henry et al. 1951). Even though the existence of the siphon effect is well established (Green 2016), it is difficult to quantify its role at different G loads, since data of intracerebral venous pressures in humans exposed to such conditions are scarce. On the other hand, it can be presumed that the G-induced hypoxaemia led to hyperpnoea and therefore to hypocapnia. Although hypocapnia constitutes a potent stimulus for cerebral vasoconstriction (Shapiro et al. 1970), it appears unlikely that it was capable of overriding the vasodilatory 69 effect of the G-induced drop in local precapillary transmural pressure. Thus, the prompt overshoot in frontal cortex oxygenation upon cessation of G exposure probably reflected a concomitant surge in cerebral blood flow and hence suggests that local precapillary resistance vessels were dilated upon release of the G load. Similar overshoot responses in Δ[O2Hb], reflecting a post-ischaemic reactive hyperaemia phase, have been noted following cessation of short-duration high +Gz loads (Kobayashi et al. 2012) and in particular following exposures inducing G-LOC (Ryoo et al. 2004). The present G-induced hypoxaemia is attributable to pulmonary atelectasis and consequent shunting of deoxygenised blood in dependent portions of the lungs (Barr 1962, 1963; Glaister 1970). Thus, the G-dependent intrathoracic hydrostatic pressure gradient results in basal redistribution of the intrapulmonary blood volume and compression of basal alveoli (Barr 1963; Dussault et al. 2016; Glaister 1970). Our finding that the G-induced hypoxaemia was exaggerated by inflation of the AGS abdominal bladder is also in agreement with results from previous studies (Barr 1963) and supports the notion that G-induced hydrostatic compression atelectasis is aggravated by transmission of pressure from the AGS to the abdominal cavity and the lower thorax (Eiken et al. 2011). It should be noted that the magnitude and time courses of the present G-induced hypoxaemia episodes, as reflected by the decrements in capillary oxygen haemoglobin saturation in a fingertip, should be interpreted with caution. As obvious from the SpO2 curve upon G offset, changes in oxyhaemoglobin saturation induced in the pulmonary circulation will appear in the fingertip with a delay of at least 30 s. In the cerebral vessels, the delay time is substantially shorter, and, in all likelihood, the magnitude of the saturation drop is larger than in the capillaries of a finger (cf. Lindholm et al. 2007). It appears that in the present 4- and 5-G exposures (but not in the 2- and 3-G exposures), the hypoxaemia was sufficiently severe to counteract any recovery of cerebral deoxygenation despite the considerable recovery of cerebral perfusion pressure during the latter part of the exposure. In fact, in the condition most relevant from an operational viewpoint, namely 5 G with full anti-G garment, ∆[O2Hb] continued to drop by about 40% during the course of the G load (from −10 to −14 μM). Judging from the considerable difference in ∆[O2Hb] between the 4- and 5-G ABD trials, it seems reasonable to assume that the continuous drop in ∆[O2Hb] during the course of the G plateau might be more pronounced at loads exceeding 5 G. Thus, progressive reduction of the cerebral anoxia reserve during the exposure to sustained high G whilst wearing full anti-G-protective garment remains a plausible explanation to our previous finding that the risk of G-LOC 13 70 Eur J Appl Physiol (2017) 117:61–72 2G NoAGS 17.0 (b) 16.0 15.0 15.0 14.0 13.0 12.0 13.0 11.0 B 1 2 3 4 R 3G NoAGS 17.0 B (d) 16.0 15.0 15.0 14.0 13.0 12.0 1 2 3 4 R 4 R 3G AGS 17.0 16.0 IOP (mmHg) IOP (mmHg) 14.0 12.0 11.0 (c) 2G AGS 17.0 16.0 IOP (mmHg) IOP (mmHg) (a) 14.0 13.0 12.0 11.0 11.0 B 1 2 3 4 R B 1 2 3 Fig. 4 Individual mean values, and group averages (bold curves), of intraocular pressure (IOP) obtained during the 4-min G plateaux at 2 and 3 G, and the 3-min recovery with (AGS) and without (NoAGS) anti-G-suit pressurised. B baseline phase, R recovery phase. n = 10 was several-fold higher when the pressure in the G-protective system was lost after a prolonged period of sustained exposure to +6 Gz than when pressure failed to increase in conjunction with the onset of +6 Gz (Eiken and Grönkvist 2013). In this connection, it should be noted that the retinal anoxia reserve, by contrast, appears rather resilient to hypoxaemia. Thus, a preceding period of moderate hypoxia does not seem to affect the latency period from complete blocking of retinal blood flow to loss of vision (Lambert and Bjurstedt 1962). Information is scarce regarding oxygenation responses of the frontal cortex to prolonged periods of sustained highG loads. In the present 5-G exposures, ∆[O2Hb] dropped by about 14 μM, which should be compared to the 20–23 μM decrements reported in conjunction with brief +Gz exposures resulting in G-LOC (Kurihara et al. 2007; Tripp et al. 2009). It should be noted, however, that although ∆[O2Hb] appears to be a valid predictor of A-LOC and G-LOC (Kurihara et al. 2007; Ryoo et al. 2004; Tripp et al. 2009), the relative drop in ∆[O2Hb] at which A-LOC/G-LOC occurs varies considerably between studies [e.g., from −6 to −7% in the study by Ryoo et al. (2004) to −32 to −33% in the study by Kurihara et al. (2007)]. Presumably, these 13 inter-study variations are mainly attributable to methodological differences, including spectrometer performance, inter-optode distance, and transducer position. Regardless, inter-study comparisons of ∆[O2Hb] in response to G exposure should be conducted with caution. Study limitations and delimitations Arguably, the present study lacked sufficient statistical power to rule out the possibility that increased +Gz loading may reduce IOP. From a practical viewpoint, it appears, however, that any IOP change induced by slight-to-moderate +Gz elevations is too small to have significant consequences (Fig. 4). Present G exposures consisted of 4-min plateaux at ≤ +5 Gz. From an operational viewpoint, it is important that these data are supplemented with data from G exposures comprising simulated aerial combat manoeuvres (SACM), during which the G loading is alternated (e.g., every 15 s) between moderate (e.g., 4.5 G) and high (e.g., 7 G). Judging from previous studies, long-duration SACM will induce similar levels of hypoxaemia (Balldin and Siegborn 1992), and hence presumably also of frontal cortex hypoxia, as those observed in the present long-duration, constant-load Eur J Appl Physiol (2017) 117:61–72 G exposures. In-flight recordings from a few F-15 pilots seem to suggest that, during ACM, cerebral oxygenation decreases in response to both the peak load and duration of the ACM (Kobayashi et al. 2002). Consequently, it is reasonable to assume, but remains to be established, that currently observed gradually diminishing cerebral oxygenation at elevated G load might also occur during long-duration SACM. Furthermore, it remains to be investigated whether, and in what manner, pressure breathing, G load, and G-suit pressurisation interact as regard both intrapulmonary shunting of deoxygenized blood and cerebral oxygenation. In the present study, the increase in G load from 4 to 5 G was accompanied not only by an increase in anti-G-suit pressure, but also by application of positive pressure breathing. Conclusions Prolonged exposure to light-moderate +Gz elevation did not affect intraocular pressure but induced a sustained reduction in cerebral oxygenation, despite a concomitant partial recovery of mean arterial pressure. That cerebral oxygenation remained suppressed throughout these exposures is attributable to pulmonary shunting of deoxygenized blood into the systemic arteries, and suggests that the increased risk of G-LOC upon G-garment failure after prolonged G exposure is due to a reduction in the cerebral anoxia reserve. Acknowledgements This study was supported by a grant from the Swedish Armed Forces (Grant No: 9220907). Compliance with ethical standards Conflict of interest The authors declare that they have no conflicts of interest. 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. References Almubrad TM, Ogbuehi KC (2010) On repeated corneal applanation with the Goldmann and two non-contact tonometers. Clin Exp Optom 35:587–596 Anderson DR, Grant WM (1973) The influence of position on intraocular pressure. Invest Ophtalmol 12:204–212 Bakke EF, Hisdal J, Semb SO (2009) Intraocular pressure increases in parallel with systemic blood pressure during isometric exercise. Invest Ophtalmol Vis Sci 50:760–764 71 Balldin UI, Siegborn J (1992) G endurance during heat stress and balanced pressure breathing. Aviat Space Environ Med 63:177–180 Barr P-O (1962) Hypoxemia in man induced by prolonged acceleration. Acta Physiol Scand 54:128–137 Barr P-O (1963) Pulmonary gas exchange in man as affected by prolonged gravitational stress. Acta Physiol Scand Suppl 207:1–46 Benni PB, Li JK, Chen B, Cammarota J, Amory DW (2003) Correlation of NIRS determined cerebral oxygenation with severity of pilot +Gz acceleration symptoms. Adv Exp Med Biol 530:381–389 Boushel R, Langberg H, Olesen J, Gonzales-Alonzo J, Bulow J, Kjaer M (2001) Monitoring tissue oxygen availability with near infrared spectroscopy (NIRS) in health and disease. Scand J Med Sci Sports 11:213–222 Duane TD (1954) Observations on the fundus oculi during black-out. Arch Ophtalmol 51:343–355 Dussault C, Gontier E, Verret C, Soret M, Boussuges A, Hedenstierna G, Montmerle-Borgdorff SS (2016) Hyperoxia and hypergravity are independent risk factors of atelectasis in healthy sitting humans: a pulmonary ultrasound and SPECT/CT study. J Appl Physiol 121:66–77 Eiken O, Grönkvist M (2013) Signs and symptoms during supra tolerance +Gz exposures, with reference to G-garment failure. Aviat Space Environ Med 84:196–205 Eiken O, Kölegård R, Bergsten E, Grönkvist M (2007) G protection: interaction of straining maneuvers and positive pressure breathing. Aviat Space Environ Med 78:392–398 Eiken O, Bergsten E, Grönkvist M (2011) Mechanisms for G protection afforded by the abdominal bladder of a full coverage anti-G suit with and without pressure breathing. Aviat Space Environ Med 82:972–977 Ferrari M, Mottola L, Quaresima V (2004) Principles, techniques, and limitations of near infrared spectroscopy. Can J Appl Physiol 29:463–487 Glaister DH (1970) The effect of gravity and acceleration on the lung. Slough England, Technivision Services. AGARDograph No 133 Green NDC (2016) Long duration acceleration. In: Gradwell D, Rainford DJ (eds) Ernsting’s aviation and space medicine, 5th edn. CRC Press, Boca Raton, pp 131–156 Hampson NB, Piantadosi CA (1988) Near infrared monitoring of human skeletal muscle oxygenation during forearm ischemia. J Appl Physiol 64:2449–2457 Henry JP, Gauer OH, Kety SS, Kramer K (1951) Factors maintaining cerebral circulation during gravitational stress. J Clin Invest 30:292–300 Jain MR, Marmion VJ (1976) Rapid pneumatic and Mackey-Marg applanation tonometry to evaluate the postural effect on intraocular pressure. Br J Ophtalmol 60:687–693 Jensen LJ, Holstein-Rathlou N-H (2013) The vascular conducted response in cerebral blood flow regulation. J Cereb Blood Flow Metab 33:649–656 Kobayashi A, Tong A, Kikukawa A (2002) Pilot cerebral oxygenation status during air-to-air combat maneuvering. Aviat Space Environ Med 73:919–924 Kobayashi A, Kikukawa A, Kimura M, Inui T, Miyamoto Y (2012) Cerebral near-infrared spectroscopy to evaluate anti-G straining maneuvers in centrifuge training. Aviat Space Environ Med 83:790–794 Krieglstein G, Langham ME (1975) Influence of body position on the intraocular pressure of normal and glaucomatous eyes. Ophtalmologica 171:132–145 Kurihara K, Kikukowa A, Kobayashi A, Nakadate T (2007) Frontal cortical oxygenation changes during gravity-induced loss of consciousness in humans: a near-infrared spatially resolved spectroscopic study. J Appl Physiol 103:1326–1331 13 72 Lambert EH, Bjurstedt H (1962) Effect of variations of oxygen and carbon dioxide tensions in inspired air on the latency of black out produced by pressure on the eyeball. Fed Proc 11:87 Lambert EH, Wood EH (1946) The problem of black out and unconsciousness in aviators. Med Clin N Am 30:833–844 Lindholm P, Bloggs SL, Gennser M (2007) Pulse oximetry to detect hypoxemia during apnea: comparison of finger and ear probes. Aviat Space Environ Med 78:770–773 Mitchell P, Lee AJ, Wang JJ, Rochchina E (2005) Intraocular pressure over the clinical range of blood pressure: blue mountains eye study findings. Am J Physiol Cell Physiol 4:C784–C794 Newsom WA, Leverett SD (1968) Retinal circulation in man during centrifugal acceleration. Am Acad Ophtalmol Otolaryngol 72:39–48 Nicholson AN, Macnamara WD, Borland RG (1968) Responsiveness of the cortex and visual pathways during transient hypotension. Electroencephalogr Clin Neurophysiol 25:330–337 Panerai RB, Hudson V, Fan L, Mahony P, Yeoman PM, Hope T, Evans DH (2002) Assessment of dynamic cerebral autoregulation based on spontaneous fluctuations in arterial blood pressure and intracranial pressure. Physiol Meas 23:59–72 13 Eur J Appl Physiol (2017) 117:61–72 Riva CE, Grunwals JE, Petrig BL (1986) Autoregulation of human retinal blood flow. An investigation with laser Doppler velocimetry. Invest Ophtalmol Vis Sci 27:1706–1712 Ryoo HC, Sun HH, Shender BS, Hrebien L (2004) Consiousness monitoring using near-infrared spectroscopy during high +Gz exposures. Med Eng Phys 26:745–753 Shapiro W, Wasserman AJ, Baker JP, Patterson JL Jr (1970) Cerebrovascular response to acute hypocapnic and eucapnic hypoxia in normal man. J Clin Invest 49:2362–2368 Testart J (1990) Patophysiology of cerebral arterial circulation. Application to cerebro arterial surgery. J Mal Vasc 15:194–201 Tripp LD, Warm JS, Matthews G, Chiu PY, Bracken BR (2009) On the tracking course of cerebral oxygen saturation and pilot performance during gravity-induced loss of consciousness. Hum Factors 51:775–784 Van Beekvelt MC, Colier WN, Wevers RA, Van Engelen BG (2001) Performance of near-infrared spectroscopy in measuring local O2 consumption and blood flow in skeletal muscle. J Appl Physiol 90:511–519
https://openalex.org/W2157778521	https://zenodo.org/records/1268888/files/article.pdf	English	null	Agility improvement through cooperative diversity in cognitive radio	GLOBECOM '05. IEEE Global Telecommunications Conference, 2005.	2,005	public-domain	4,766	matter experts for publication in the IEEE GLO his full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE GLO his full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE GLOBECOM 2005 proceedings. wed at the direction of IEEE Communications Society subject This work was supported by NSF under Grant CCR-0121565 and the U. S. Army Research Laboratory under the Collaborative Technology Alliance Program, Cooperative Agreement DAAD19-01-20-0011. † Corresponding Author: email : guru@ece.gatech.edu, phone/fax : 404- 385-3012/894-7883 I. INTRODUCTION The concept of software deﬁned radio (SDR) was intro- duced in [1]. In SDR, the software embedded in a radio cell phone would deﬁne the parameters under which the phone should operate in real-time as its user moves from place to place. Cognitive radio (CR) is even smarter than SDR. CR is a radio that is aware of and can sense its environment; learn from its environment; and perform functions that best serve its user. Since the cognitive (unlicensed) users are utilizing the licensed band, they must detect the presence of licensed (primary) users in a very short time and must vacate the band for the primary users. Thus one of the major challenges that confronts this technology is : how do the cognitive (unlicensed) radios sense the presence of the primary (licensed) user? One may expect this to be trivial but as shown in [2], there are fundamental limits to the detection capabilities of CR networks. In this paper, we show improvement in spectrum sensing capabilities through cooperation between individual cognitive users. Agility Improvement through Cooperative Diversity in Cognitive Radio Ghurumuruhan Ganesan†and Ye (Geoffrey) Li School of Electrical and Computer Engineering Georgia Institute of Technology, Atlanta, Georgia 30332–0250 transmitter without any processing, achieves full diversity. In this paper, we show that by allowing the users to cooperate under the AF protocol leads to much higher agility of the cognitive network. Also, we consider only the case where there are utmost two users per carrier and there is centralized scheduling to pair the users. Analysis of decentralized multi- user single-carrier networks can be found in [7]. Abstract— In this paper, we illustrate the beneﬁts of coop- eration in cognitive radio. Cognitive (unlicensed) users need to continuously monitor spectrum for the presence of primary (licensed) users. We show that by allowing the cognitive radios operating in the same band to cooperate we can reduce the detection time and thus increase the overall agility. We ﬁrst consider the case of two cognitive users and show how the inherent asymmetry in the network can be exploited to increase the agility. We then extend our protocol to study multi-user multi-carrier cognitive network. We compare our cooperation scheme with the non-cooperation scheme and derive expressions for agility gain. We show that our cooperation scheme reduces the detection time for the cognitive users by as much as 35% . The organization of the paper is as follows: In Section II, we formulate the problem for the simple two user case and show that the inherent asymmetry of the network can lead to faster detection with cooperation. We also describe the detector we shall use for detecting the presence of the primary user. In Section III, we develop expressions for asymptotic agility gain for the two user network and show improvement in agility. In Section IV, we extend our results to multi-user multi-carrier networks and state precise conditions under which agility gain is achieved. Finally, in Section V, present our conclusion. Index Terms— cognitive radio, cooperative diversity, agility gain, detection time. U.S. Government work not protected by U.S. copyright II. PROBLEM FORMULATION This full text paper was peer reviewed at the direction of IEEE Communications Society subject • The primary user location is known to both U1 and U2. • The primary user location is known to both U1 and U2. B. Energy Detector • Both the users when acting as relays have no transmit power constraint. The reasons for choosing energy detector (ED) are twofold: (1) We want to show the effect of cooperation in cognitive networks. Hence the choice of detector is not critical. (2) We model the signal as a random variable with known power. Hence ED is optimal [8]. It can be shown that given h12 = h21, the random variables H and W in (2) are Rayleigh distributed with zero-mean and variances, In [7], we incorporate power constraint for the relay user and study cooperative detection schemes for multi-user single- carrier networks. A. Cooperation Scheme σ2 H = P1 + P2h (3) (3) In Figure 2, we illustrate the protocol used by U1 and U2 to transmit data to some common receiver. We assume a slotted transmission system where U1 and U2 transmit in successive slots according to the AF protocol [5] in the same frequency band. Accordingly in time slot T1, U1 transmits and in T2, U2 relays the information of the previous slot. Unknown to both these users, a primary user has started using the band. This primary user must be detected as quickly as possible. At time slot T1, U1 transmits its data. Hence the signal received by U2 in the time-slot T1 is given by, and σ2 W = 1 + h (4) (4) respectively, where h = \|h12\|2 E{\|h12\|2}, P1 = E{\|h2 p1\|} and P2 = E{\|h2 p2\|}. Since h12 is complex Gaussian, it is easily seen that h is exponentially distributed with unity mean and variance. The ED forms the statistics respectively, where h = \|h12\|2 E{\|h12\|2}, P1 = E{\|h2 p1\|} and P2 = E{\|h2 p2\|}. Since h12 is complex Gaussian, it is easily seen that h is exponentially distributed with unity mean and variance. The ED forms the statistics T(Y ) = \|Y \|2 y2 = θhp2 + ah12 + w1, (1) (1) and compares with a threshold λ which is determined by some prespeciﬁed probability of false alarm α. Deﬁne where hpi denotes the channel gain between the primary user and user Ui and hij denotes the channel gain between users Ui and Uj. In (1), a denotes the message sent from U1 to the common receiver and θ indicates the presence of the primary user. Hence θ = 1 implies the presence of the primary user and θ = 0 implies its absence. In our cooperation protocol, we allow the user U1 also to listen during the relay slot T2. Thus the signal received by U1 from U2 during relay slot is given by ϕ(t; a, b) = ∞ 0 e−h− t a+bh dh (5) (5) for positive t, a and b. Also let Fi(t) denote the cumulative density function (cdf) of the random variable T under hypoth- esis Hi, i = 0, 1. For H0 it can be shown that, for positive t, a and b. A. Cooperation Scheme Also let Fi(t) denote the cumulative density function (cdf) of the random variable T under hypoth- esis Hi, i = 0, 1. For H0 it can be shown that, F0(t) = P(T(Y ) > t\|H0) = ∞ 0 P(T(Y ) > t\|H0, h)f(h) dh = ∞ 0 e− t 1+h e−h dh = ϕ(t; 1, 1). y1 = βh12(ah12 + w1 + θhp2) + θhp1 + w2, where β refers to the scaling factor [5] used by U2 to relay the information to the receiver. The scaling factor is usually chosen by U2 to satisfy its power constraint [5]. In this paper, however, we assume that the relay user has no power constraint. Hence we can choose β as we like. For simplicity, we choose 1 = 0 e− t 1+h e−h dh = ϕ(t; 1, 1). Similarly, it can be shown that Similarly, it can be shown that F1(t) = ϕ(t; P1 + 1, P2 + 1). β = 1 E{\|h12\|2}. Hence, for a given probability of false alarm α, we need to ﬁnd the threshold λ such that Hence, for a given probability of false alarm α, we need to ﬁnd the threshold λ such that Throughout the paper, we assume that the channels are recip- rocal symmetric, i.e., hij = hji. Thus, after the message component is cancelled from the received signal, the user U1 is left with the signal ϕ(λ; 1, 1) = α. (6) (6) In fact λ can be uniquely determined since ϕ in (5) is strictly decreasing in t. Similarly the probability of detection by U1 with cooperation from U2 is found to be In fact λ can be uniquely determined since ϕ in (5) is strictly decreasing in t. Similarly the probability of detection by U1 with cooperation from U2 is found to be Y = θH + W, where H = hp1 + √βhp2h12 and W = w1 + √βh12w2. The detection problem can then be stated as: p(1) c = ϕ(λ; P1 + 1, P2 + 1). (7) p(1) c = ϕ(λ; P1 + 1, P2 + 1). (7) Given the signal, Compared to the non-cooperative case, the average SNR is increased by cooperation provided P2 > P1. From (3) and (4), the instantaneous SNR is given by Compared to the non-cooperative case, the average SNR is increased by cooperation provided P2 > P1. II. PROBLEM FORMULATION We now describe the channel model of the system we use in the paper. We assume that all users experience Rayleigh fading that is independent from user to user. If a signal sent at the transmitter is x, the received signal y is given by y = hx + w, where h is the fading coefﬁcient modelled as a complex Gaussian random variable and w is additive white Gaussian noise. Unless otherwise mentioned all noise coefﬁcients in this paper are assumed to be with zero-mean and unit-variance. The main requirement of a cognitive radio architecture is to detect the presence of the primary or licensed users as quickly as possible. For that reason, the cognitive users should continuously monitor the spectrum for the primary (licensed) users. Consider the situation shown in Figure 1, where two cognitive radio users U1 and U2 operating in a ﬁxed TDMA mode for sending data to a common receiver and a primary user starts using the band soon after. Then the two cognitive users need to vacate the band as soon as possible to make way for the primary user. However, the detection time becomes signiﬁcant if one of the users is in the boundary of decodability of the primary user as shown in Figure 1. The signal received from the primary user is so weak that the cognitive user U1 takes a lot of time to sense its presence. In this section, we describe our protocol which employs cooperation to reduce the detection time thereby improving the sensitivity of the cognitive receiver. Throughout the paper, we assume that : Cooperative networks achieve diversity gain by allowing the users to cooperate [3] [4]. Cooperative schemes with orthogonal transmission in a TDMA system have been recently proposed in [5] and [6]. It has been shown in [5] that two user single hop networks in which one of the user acts as a relay for the other, result in lower outage probabilities. In particular, it is shown that the amplify-and-forward (AF) protocol [5], in which the relay transmits the signal obtained from the U.S. Government work not protected by U.S. copyright 2505 IEEE Globecom 2005 matter experts for publication in the IEEE GLOBECOM 2005 proceedings. This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE GLOBECOM 2005 proceedings. A. Cooperation Scheme From (3) and (4), the instantaneous SNR is given by Y = θH + W (2) the detector decides on E{γc\|h} = P1 + P2h 1 + h . H1 : θ = 1 or or Hence the average SNR in case of cooperation is given by H0 : θ = 0. γc = ∞ 0 E{γc\|h}f(h) dh = ∞ 0 P1 + P2h 1 + h e−h dh We now utilize a simple energy detector [8] to show advantage of our proposed scheme. U.S. Government work not protected by U.S. copyright 5 2506 U.S. Government work not protected by U.S. copyright 2506 IEEE Globecom 2005 matter experts for publication in the IEEE GLOBECOM 2005 proceedings eviewed at the direction of IEEE Communications Society subject (2) p(2) n > p(2) c . For the non-cooperative case the average SNR is given by For the non-cooperative case the average SNR is given by Thus when the primary user location is known, we allow only U2 to help U1 and not vice-versa. In that case, the total average detection time of our cooperation scheme is given by γnc = P1. Thus the SNR gain is, Thus the SNR gain is, γ = γc γnc = P2 P1 (1 −F) + F, F = ∞ 0 (1 + h)−1e−h dh < 1. γ = γc γnc = P2 P1 (1 −F) + F, Tc = 2 −p(1) c +p(2) n 2 p(1) c + p(2) n −p(1) c p(2) n . where where F = ∞ 0 (1 + h)−1e−h dh < 1. We deﬁne the agility gain of our cooperation scheme over the non-cooperation scheme when there are two users as µn/c(2) ∆= Tn Tc . In fact, γ −1 = (1 −F)( P2 P1 −1) > 0 if P2 > P1, which implies a SNR gain. Note that the agility gain µ(2) is a function of P1 and P2. However, without loss of generality we set P1 = 1 and consider the agility gain as a function of P2 alone. It can be shown that as P2 →∞, the asymptotic agility gain is given by, In Figure 3, we have shown the performance curve of the energy detector with and without cooperation. In the simulation, P1 = 1 and P2 = 2.7. Note that we assume that the relay user has no power constraint. A. Cooperation Scheme Hence in the cooperative mode, the relay terminal adjusts its transmit power so that the channel gain between the users U1 and U2 is kept constant irrespective of the position of U2. This serves as an upper bound on the performance of cooperation schemes. It can be seen that for the same probability of false alarm, we have higher detection probability through cooperation. This is the effect of cooperation in the network. µ∞ n/c(2) = lim P2→∞µn/c(2) = 2 √α. The non-cooperation scheme presented above assumes ex- tremely selﬁsh users. Hence cognitive user U1 detects the presence of the primary user without the help of U2. Also once detected U1 vacates the band without informing U2. Consider now a more practical situation where both the users get informed once either U1 or U2 detects the primary user. This is facilitated by transmitting the detection information to the common receiver which in turn informs the other cognitive users. Note that this is a partially cooperative network where the cognitive users still detect the primary user without any cooperation. In such a case, it can be shown that the time taken to vacate the band is given by, III. AGILITY OF THE TWO USER COGNITIVE RADIO NETWORK In this section, we determine the time taken by the two- user network described in Section II to detect the presence of the primary user. We show that under the cooperation scheme described in Section II, the average detection time is reduced thus implying an increase in agility. Let τn be the number of slots taken by user U1 in a non- cooperative network to detect the presence of the primary user. This detection time τn can be modelled as a geometric random variable, i.e., Tp = 2 −p(1) n +p(2) n 2 p(1) n + p(2) n −p(1) n p(2) n . The asymptotic agility gain in this case can be shown to be, The asymptotic agility gain in this case can be shown to be, Pr{τn = k} = (1 −p(1) n )k−1p(1) n , µ∞ p/c(2) = lim P2→∞ Tp Tc = 4 3√α −α. where p(i) n denotes the probability of detection by user Ui in a single slot under the non-cooperation scheme. For the system model of our paper, it can be shown that In Figure 4, we have plotted the agility gain µn/c(2) of a two user asymmetric network as a function of the asymmetry P2 for different values of false alarm probability α. As the network becomes more and more asymmetric, agility gain increases. From the ﬁgure, we note that for α = 0.1 as much as 35% reduction in detection time is achievable. Hence if one of the cognitive users is very close to the primary user, then we are guaranteed fast detection even if other user is very far away. p(1) n = α 1 P1+1 and p(2) n = α 1 P2+1 . The total time taken by both users to vacate the band can be shown to be The total time taken by both users to vacate the band can be shown to be Tn = 2 1 p(1) n + 1 p(2) n − 1 p(1) n + p(2) n −p(1) n p(2) n . If p(i) c denotes the detection time for Ui under the cooperation scheme described in Section II then from (7) we have If p(i) c denotes the detection time for Ui under the cooperation scheme described in Section II then from (7) we have IV. MULTIUSER COGNITIVE NETWORK In this section we assume that there are 2n cognitive users. The total bandwidth B is equally divided into n sub-bands each of bandwidth ∆n = B n . There are two cognitive users working on each sub-band following the cooperation protocol described in Section II. It must be noted that the primary user, if present, uses the whole bandwidth B. If P1 denotes the power received from the primary user at cognitive user U1 when it uses the whole band, the signal power received at U1 p(1) c = ϕ(λ; P1 + 1, P2 + 1), where λ satisﬁes (6). It can be similarly shown that p(2) c = ϕ(λ; P2 + 1, P1 + 1). We have shown that if P2 > P1 then (1) p(1) c > p(1) n and We have shown that if P2 > P1 then ( ) ( ) We have shown that if P2 > P1 then (1) p(1) c > p(1) n and IEEE Globecom 2005 2507 U.S. Government work not protected by U.S. copyright matter experts for publication in the IEEE GLOBECOM 2005 proceedings. wed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE GLOBECOM 2005 proceedings. This full text paper was peer reviewed at the direction of IEEE Communications Society subject It can be easily shown that for all n It can be easily shown that for all n now is given by P ′ 1 = P1 n . The noise power at U1 is similarly scaled by a factor of n. Suppose that users U1 and U2 form a link in a particular sub-band with U2 acting as a relay for U1. The detection problem for that sub-band is identical to (2) except that 1 µp/c(n) > 1 and and lim n→∞µp/c(n) = 1. σ2 H = 1 n(P1 + P2h) Hence for all ﬁnite cognitive user population we have agility gain, though, asymptotically, there is little difference between partial cooperation and total cooperation. and and σ2 W = 1 n(1 + h) where, as before, h denotes the scaled channel gain between the pair of users. Thus the detection problem for multi-carrier networks is similar to that of single-carrier networks except for a scaling factor of n. IV. MULTIUSER COGNITIVE NETWORK Hence for a given probability of false alarm α we need to ﬁnd the threshold λn such that For the totally non-cooperative network, we shall assume that the probability of detection is equal for all the sub-bands in consideration. Though this is strictly not true, we make this simplifying assumption because we are concerned not in exploiting the asymmetry of the network but the presence of multiple cognitive users. Let p′ = 1 −q′ denote the detection probability for all the n sub-bands under the totally non- cooperation scheme. It can be shown that the total time taken for detection is then given by ϕ λn; 1 n, 1 n = α. It can be easily shown that λn = λ n where λ is given by (6). Thus for example, the probability of detection by U1 with cooperation from U2 is found to be Tn(n) = n k=1 n k (−1)k+1 1 1 −(q′)k . p(1) c = ϕ λn; P1 + 1 n , P2 + 1 n = ϕ(λ; P1 + 1, P2 + 1). As before, we have deﬁned the agility gain to be As before, we have deﬁned the agility gain to be It follows that the detection probability for a particular user is the same independent of the bandwidth the user is occupying. It follows that the detection probability for a particular user is the same independent of the bandwidth the user is occupying. It can be seen that the probability that the cognitive users in a sub-band detect the primary user under the cooperation scheme and the non-cooperation scheme are given by µn/c(n) = Tn(n) Tc(n) It can be seen that the probability that the cognitive users in a sub-band detect the primary user under the cooperation scheme and the non-cooperation scheme are given by and have proved that lim n→∞ µn/c(n) ln n = Θ − 1 ln(1 −α) . p = p(1) c + p(2) n −p(1) c p(2) n p′ = p(1) n + p(2) n −p(1) n p(2) n p = p(1) c + p(2) n −p(1) c p(2) n and Thus the agility gain grows as ln n. In Figure 5, we have plotted the agility gain µn/c(n) of our multi-user multi-carrier cooperation scheme for various values of the probability of false alarm, α. IV. MULTIUSER COGNITIVE NETWORK Since α is very small, ln(1 −α) ≈−α, we see that the agility gain curves are well approximated as µn/c(n) ≈ln n α for large n. p′ = p(1) n + p(2) n −p(1) n p(2) n respectively. Since U2 acts as a relay for U1 and not vice- versa, it is obvious that from the discussion in the Section III, p > p′. Note that there are n sub-bands occupying the band. For k = 1, 2, ..., n, let pk and p′ k(< pk) denote the probability of detection of the kth sub-band to detect the presence of the primary user with and without cooperation respectively. As before we shall consider two cases of non-cooperation: V. CONCLUSION before we shall consider two cases of non-cooperation: (i) All the n sub-bands detect and vacate independently of other sub-bands (totally non-cooperative). (i) All the n sub-bands detect and vacate independently of other sub-bands (totally non-cooperative). In this paper, we have shown the beneﬁts of cooperation in increasing the agility of cognitive radio networks. We ﬁrst considered a simple two user cooperative cognitive network and showed improvement in agility by exploiting the inherent asymmetry. We then extended our cooperation scheme to multi-user multi-carrier networks and stated precise condi- tions under which agility gain is achieved. We analyzed two cooperative schemes employing varying degrees of cooper- ation : (1) totally non-cooperative, where each user detects the primary user and vacates without informing the other users and (2) partially cooperative, where the ﬁrst user to detect the primary user informs the other cognitive users through a common base station. We found that for multi-carrier networks, asymptotically there is no difference between the cooperation scheme described in Section II and partially coop- erative scheme. Compared to totally non-cooperative scheme, however, the agility gain of our cooperation scheme was found to be Θ(ln n) where n is the number of sub-bands. (ii) The sub-bands detect the primary user independently, but the ﬁrst sub-band to detect the primary user informs the other sub-bands through a common base station (partially cooperative). (ii) The sub-bands detect the primary user independently, but the ﬁrst sub-band to detect the primary user informs the other sub-bands through a common base station (partially cooperative). In case of (ii), it can be shown that the average number of slots taken for detection under cooperation and non-cooperation are given, respectively, by Tc(n) = 1 1 −n k=1(1 −pk) Tp(n) = 1 1 −n k=1(1 −p′ k). and and Deﬁne the agility gain to be µp/c(n) = Tp(n) Tc(n). IEEE Globecom 2005 U.S. Government work not protected by U.S. copyright U.S. Government work not protected by U.S. copyright 2508 Primary user : User 1 : User 2 Ordinary link dability of P Relay link time luation and future 8, pp. 25–31, Apr. l limits in cognitive mputing, Oct. 2004. eration in diversity vol. 51, pp. 1927– entation aspects and 51, pp. V. CONCLUSION 1939–1948, it i i l 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 false alarm probability (α) detection probability with coop without coop Fig. 3. Performance Curves 0 0.5 1 1.5 2 2.5 3 3.5 0.95 1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 asymmetry (log10(P2)) agility gain (µp/c(2)) α = 0.1 α = 0.15 α = 0.2 Fig. 4. Agility gain in single carrier two user asymmetric networks U1 U2 P P : Primary user U1 : User 1 U2 : User 2 Ordinary link Boundary of decodability of P Relay link Fig. 1. Cooperation in cognitive network 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 false alarm probability (α) detection probability with coop without coop Fig. 3. Performance Curves U1 U2 P Bound 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 false alarm probability (α) detection probability with coop without coop Fig. 3. Performance Curves Relay link Fig. 1. Cooperation in cognitive network Fig. 3. Performance Curves 0 0.5 1 1.5 2 2.5 3 3.5 0.95 1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 asymmetry (log10(P2)) agility gain (µp/c(2)) α = 0.1 α = 0.15 α = 0.2 Fig. 4. Agility gain in single carrier two user asymmetric networks 0 0.5 1 1.5 2 2.5 3 3.5 0.95 1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 asymmetry (log10(P2)) agility gain (µp/c(2)) α = 0.1 α = 0.15 α = 0.2 Fi 4 A ili i i i l i i k Fig. 2. Relay protocol used REFERENCES [1] I. J. Mitola, “Software radios: Survey, critical evaluation and future directions,” IEEE Aerosp. Electron. Syst. Mag., vol. 8, pp. 25–31, Apr. 1993. [2] A. Sahai, N. Hoven and R. Tandra, “Some fundamental limits in cognitive radio,” in Allerton Conf. on Commun., Control and Computing, Oct. 2004. [3] A. Sendonaris, E. Erkip and B. Aazhang, “User cooperation in diversity - Part I: System description,” IEEE Trans. Commun., vol. 51, pp. 1927– 1938, Nov. 2003. [4] ——, “User cooperation in diversity - Part II: Implementation aspects and performance analysis,” IEEE Trans. Commun., vol. 51, pp. 1939–1948, Nov. 2003. Fig. 4. Agility gain in single carrier two user asymmetric networks [5] J. N. Laneman and D. N. C. Tse, “Cooperative diversity in wireless networks: Efﬁcient protocols and outage behaviour,” IEEE Trans. Inform. Theory, vol. 50, pp. 3062–3080, Dec. 2004. 0 50 100 150 200 10 15 20 25 30 35 40 45 50 55 60 cognitive population (n) agility gain (µn/c(n)) α = 0.1 α = 0.2 α = 0.15 Fig. 5. Agility gain in multi-carrier multi-user symmetric networks [6] J. N. Laneman and G. W. Wornell, “Distributed space-time coded pro- tocols for exploiting cooperative diversity in wireless networks,” IEEE Trans. Inform. Theory, vol. 49, pp. 2415–2425, Oct. 2003. [7] G. Ganesan and Y. G. Li, “Cooperative spectrum sensing in cognitive radio networks,” in IEEE DYSPAN 2005, St. Louis, Missouri, Nov. 2005, accepted for publication. [8] H. V. Poor, An Introduction to Signal Detection and Estimation, 2nd ed. Springer-Verlag, 1994. Fig. 5. Agility gain in multi-carrier multi-user symmetric networks U.S. Government work not protected by U.S. copyright U.S. Government work not protected by U.S. copyright IEEE Globecom 2005 IEEE Globecom 2005 2509
https://openalex.org/W4300961823	http://ejournal.unp.ac.id/index.php/voteknika/article/download/117644/106856	Indonesian	null	Rancang Bangun Sistem E-Tracer Study Alumni SMKN 1 Lembah Melintang Untuk Mengetahui Output Pendidikan Berbasis Web	Voteteknika	2,022	cc-by-sa	3,859	Rancang Bangun Sistem E-Tracer Study Alumni SMKN 1 Lembah Melintang untuk Mengetahui Output Pendidikan Berbasis Web Muhammad Yusuf1, Ahmaddul Hadi2 1Pendidikan Teknik Informatika Fakultas Teknik Universitas Negeri Padang 2Departemen Teknik Elektronika Fakultas Teknik Universitas Negeri Padang Jln. Prof. Dr. Hamka –Kampus UNP, Air Tawar, Padang Corresponding author e-mail: inspirationyusuf@gmail.com Muhammad Yusuf1, Ahmaddul Hadi2 1Pendidikan Teknik Informatika Fakultas Teknik Universitas Negeri Padang 2Departemen Teknik Elektronika Fakultas Teknik Universitas Negeri Padang Jln. Prof. Dr. Hamka –Kampus UNP, Air Tawar, Padang Corresponding author e-mail: inspirationyusuf@gmail.com Jurnal Vocational Teknik Elektronika dan Informatika http://ejournal.unp.ac.id/index.php/voteknika/index Vol. 10, No. 3, September 2022 Jurnal Vocational Teknik Elektronika dan Informatika http://ejournal.unp.ac.id/index.php/voteknika/index Vol. 10, No. 3, September 2022 P- ISSN: 2302-3295, E-ISSN : 2716-3989 ABSTRACT Tracer study is an effort that can provide useful information to evaluate the learning outcomes of educational institutions. This information is used as a means of continuous development in ensuring the quality, quality, and progress of an educational institution. Tracer studies are very important to be carried out in order to obtain valid information from alumni related to the output of an educational institution which includes the suitability of knowledge gained during upper secondary education or in lectures or in the world of work. Based on that, a system is needed that will later help survey alumni data to be faster, more efficient, and also flexible. The method used in this study is waterfall model research. While the system architecture used is a Model-View-Controller-based architecture by utilizing the OOP (Object Oriented Programming) technique which will be built using the Yii2 framework using PHP and MySQL programming languages as a database storage medium. The final output of this study is to create an information system that is useful for recording alumni, one of which can help the accreditation process at SMKN 1 Lembah Melintang. Keywords: E-Tracer Study, Information Systems, Yii2 framework, Waterfall Keywords: E-Tracer Study, Information Systems, Yii2 framework, Waterfall ABSTRAK Tracer study merupakan sebuah upaya yang dapat menyediakan informasi yang bermanfaat untuk mengevaluasi hasil pembelajaran dari institusi pendidikan. Informasi ini digunakan sebagai sarana pengembangan yang berkelanjutan dalam menjamin kualitas, mutu, dan kemajuan sebuah institusi pendidikan. Tracer study sangat penting dilakukan guna mendapatkan informasi valid dari alumni yang berkaitan dengan output sebuah institusi pendidikan yang meliputi kesesuaian ilmu yang didapat pada saat pendidikan menengah atas atau dalam perkuliahan maupun dalam dunia kerja. Oleh karena itu dibutuhkan sebuah sistem yang nantinya akan membantu survey data alumni menjadi lebih cepat, efisien, dan juga fleksibel. Adapun metode yang digunakan dalam penelitian ini adalah penelitian model waterfall. Sedangkan arsitektur sistem yang digunakan adalah arsitektur berbasis Model-View-Controller dengan memanfaatkan teknik OOP (Object Oriented Programming) yang akan dibangun menggunakan framework Yii2 dengan menggunakan bahasa pemrograman PHP dan MySQL sebagai media penyimpanan database. Adapun output akhir dari penelitian ini adalah menciptakan sistem informasi yang berguna untuk mendata alumni yang salah satunya dapat membantu proses akreditasi di SMKN 1 Lembah Melintang. Kata kunci : : E-Tracer Study, Sistem Informasi, framework Yii2, Waterfall. VoteTEKNIKA Vol. 10, No. 3, September 2022 VoteTEKNIKA Vol. 10, No. 3, September 2022 A Vol. 10, No. 3, September 2022 penting dalam menentukan nilai akreditasi yang didapat oleh institusi pendidikan tersebut [1]. mulai dari level kebutuhan sistem lalu menuju ke tahap analisis, desain, coding (perancangan), testing, dan maintenance (Yahya Dwi Wijaya,2019). Model waterfall ini berkembang secara sistematis dari satu tahap ke tahap yang lain dalam mode layaknya air terjun. Motode ini dilakukan secara bertahap sehingga tidak terfokus pada tahap tertentu. Tracer study merupakan sebuah usaha yang dapat menyediakan informasi untuk penilaian dari hasil pendidikan. Informasi ini akan digunakan sebagai sarana pengembangan yang lebih jauh dalam menjamin kemajuan dan kualitas sebuah pendidikan. Tracer study sangat penting dilakukan guna mendapatkan informasi valid dari alumni yang berkaitan dengan output sebuah institusi pendidikan yang meliputi kesesuaian ilmu yang didapat pada saat pendidikan menengah atas atau dalam perkuliahan maupun dunia kerja. Pemodelan dirancang sebagai gambaran dari operasi sistem nyata secara ideal untuk menjelaskan atau menunjukkan hubungan-hubungan penting yang saling terkait, sehingga dapat dikatakan bahwa pemodelan adalah yang sangat penting dikarenakan dapat menyederhanakan suatu masalah dan dapat diuji kesesuaian dari model yang dibentuk dan dirancang. Tugas akhir ini menggunakan pemodelan Unified Modeling Language (UML) yang memungkinkan pengembangan menyajikan beberapa tampilan sistem menggunakan berbagai diagram grafis, seperti use case diagram, context diagram, class diagram, dan activity diagram. Pengelolaan data alumni SMKN 1 Lembah Melintang masih menggunakan sistem manual atau dengan dokumen cetak, buku, atau kuisioner dari google forms. Kuisioner digunakan untuk mengisi data alumni secara manual dan kurang efektif dalam hal perubahan data. Oleh Karena itu diperlukan suatu sistem informasi yang dapat menginformasikan data secara akurat dari para alumni yang ada dan juga bisa mengupadate data yang ada, pengelolaan data alumni yang masih menggunakan dokumen cetak, mengakibatkan pengelolaan data tersebut belum efeketif. Untuk mengatasi permasalahan tersebut ditemukan sebuah solusi yaitu dengan menciptakan pemrograman sistem e-tracer study berbasis web. a. Framework Yii2 Framework PHP berbasis komponen digunakan untuk membangun aplikasi berbasis web berskala kecil hingga berskala besar secara profesional dan elegan. Yii framework akan memudahkan perancangan dan pembuatan aplikasi e- tracer study alumni baik skala kecil sampai skala besar secara kompeten dan elegan dengan menggunakan fitur-fitur yang telah disediakan yii2 framework. Yii mengimplementasikan pola desain model-view-controller (MVC), yang diadopsi secara luas dalam pemrograman Web. MVC bertujuan untuk memisahkan logika bisnis dari pertimbangan antar muka pengguna agar para pengembang bisa lebih mudah mengubah setiap bagian tanpa mempengaruhi yang lain. Dalam MVC, model menggambarkan informasi (data) dan aturan bisnis; view(tampilan) berisi elemen antar muka pengguna seperti teks, input form; sementara controller mengatur komunikasi antar model dan view. Permasalahan ini dimodelkan menggunakan metode Unified Modeling Language (UML) yang menjelaskan tentang analisis dan desain perangkat lunak dalam bentuk grafik yang dikembangkan dengan menggunakan pemrograman berorientasi objek[2]. Perancangan sistem informasi ini menggunakan framework Yii2 dengan teknik MVC (Model-View-Controller) dan metode OOP (Object Oriented Programming). Tujuan rancang bangun ini adalah untuk menghasilkan sistem informasi yang mempermudah pengelolaan data alumni salah satunya untuk menunjang kemajuan akreditasi SMKN 1 Lembah Melintang. Konsep MVC secara garis besar memiliki alur kerja dengan pertama kali user mengakases aplikasi URL tertentu. Yii dalam sisi server dapat memanggil konfigurasi untuk menjalankan aplikasi sesuai konfigurasi tersebut. Selanjutnya, aplikasi mengecek request dari user (routing) untuk kemudian menentukan controller mana yang diminta. Kemudian aplikasi meng-create controller dengan memanggil komponen yang diperlukan. Selain itu, aplikasi meng-create action juga sesuai dengan request user, lalu dilakukan pengecekan, seperti hak akses terhadap action, jika kondisinya ditolak maka aplikasi akan menampilkan penolakan dalam bentuk informasi pesan error ke user. Namun, jika kondisi pengecekan dizinkan maka akan ada pemanggilan model dan kemudian proses rendering view, setelah itu hasilnya akan dikirim ke user. I. PENDAHULUAN pendidikan yang di tempuh sebelumnya. Informasi yang diperoleh dari alumni memiliki peranan penting terhadap kesuksesan maupun kemajuan dari institusi pendidikan itu sendiri. Kegunaan data alumni selain berperan sebagai penentu kualitas dari sebuah institusi pendidikan, data alumni juga berperan Data alumni merupakan bagian terpenting dari sebuah sistem informasi yang harus digali informasinya, untuk mengetahui keberadaan mereka setelah menyelesaikan sekolah dari institusi II. METODE Metode yang digunakan dalam mengaplikasikan penelitian ini ialah metode waterfall. Metode waterfall atau metode air terjun merupakan model yang dikembangkan untuk pengembangan perangkat lunak. Metode waterfall pertama kali diperkenalkan oleh Royce pada tahun 1970 (Petersen et.al,2009 dalam Iwan Binanto,2007) dengan 7 (tujuh) tahapan yang berturut. Metode waterfall mengalami banyak perbaikan dan perubahan diantaranya adalah perubahan langkah-langkah dari 7 (tujuh) menjadi 5 (lima) tahapan (Sommerville,2011 dalam Iwan Binanto,2007). Metode waterfall adalah metode yang melakukan pendekatan secara sistematis dan urut P-ISSN: 2302-3295 2 Vol. 10, No. 3, September 2022 VoteTEKNIKA Vol. 10, No. 3, September 2022 VoteTEKNIKA Gambar 3 Use Case Diagram E-Tracer Study Keuntungan menggunakan konsep MVC diantaranya merapikan struktur aplikasi dan mudah dipahami untuk proyek aplikasi yang kompleks. Maintenance lebih mudah ketika terjadi perubahan data, proses maupun tampilan. Memudahkan dalam tracking dan handling error serta dapat membagi pekerjaan jika proyek aplikasi dikerjakan oleh tim [3]. Gambar 1. Alur Kerja Aplikasi Menggunakan Framework yii Entity Relationship Diagram (ERD) Gambar 3. Use Case Diagram E-Tracer Study Pada gambar 3 dapat dilihat bahwa admin mempunyai hak akses terhadap sistem yang ada. Sedangkan admin mempunyai hak untuk menginputkan data admin akan tetapi tidak bisa mengakses data operator. Operator yang telah terdaftar mendapatkan user dan password untuk melakukan aktifitas di sistem. Admin berperan mengelola informasi, mengelola pengumuman, data alumni, riwayat alumni, data tracer study, dan download data. Data yang telah dimasukkan oleh admin dapat dilihat oleh pengguna alumni. Alumni yang telah terdaftar mendapatkan user dan password untuk masuk ke sistem. Alumni dalam sistem dapat melihat informasi, melihat pengumuman, edit alumni, dan dapat mengisi untuk kebutuhan kuisioner. Kepala sekolah akan menerima semua laporan sistem E- tracer study alumni SMKN 1 Lembah Melintang. d Context Diagram Gambar 1. Alur Kerja Aplikasi Menggunakan Framework yii y b. Entity Relationship Diagram (ERD) b. Entity Relationship Diagram (ERD) Entity Relationship Diagram (ERD) dengan menggunakan pemahaman yang terdiri dari kumpulan objek dasar yaitu hubungan antar entitas ERD yang digunakan dalam metodologi informasi untuk menggambarkan sistem yang terdiri dari hubungan entitas ataupun disebut juga sebuah diagram yang menggambarkan relasi antar rancangan data. Entity Relationship Diagram dan perancangan sistem informasi tracer study alumni pada SMKN 1 Lembah Melintang dapat dilihat pada gambar di bawah ini: KA Vol. 10, No. 3, September 2022 data informasi atau berita, data pengumuman, data download dan data tracer study. Administrator juga menerima data alumni, data informasi atau berita, data pengumuman, data download dan data tracer study. Super admin melakukan penambahan pada admin. Sedangkan alumni melakukan transaksi data yang berupa data alumni, data tracer study, data pengumuman, dan data berita, selain itu alumni menerima data alumni, data tracer study, data pengumuman, dan data berita. Gambar 7. Halaman Utama Operator p 3. Halaman Utama Website E-Tracer Study p 3. Halaman Utama Website E-Tracer Study Halaman utama website e-tracer study merupakan halaman utama yang memiliki fungsi menu-menu pada website tracer study meliputi tracer study, tentang, pengumuman, dan login. 1. Hasil g p g Gambar 8. Halaman Utama E-Tracer Study Berikut ini merupkana hasil rancangan tampilan login pada sistem e-tracer study alumni SMK N 1 Lembah Melintang: 1. Halaman Login 1. Halaman Login Gambar 5. Tampilan Login Gambar 8. Halaman Utama E-Tracer Study Perancangan Perancangan Perancangan context diagram menggambarkan satu lingkaran besar yang dapat mewakili seluruh proses yang terdapat suatu sistem. Context diagram merupakan gambaran ruang lingkup suatu sistem yang akan dibangun. Sehingga diketahui siapa saja yang memberi data masuk ke sistem serta kepada siapa saja informasi yang harus dihasilkan. Perancangan context diagram sistem tracer study alumni SMKN 1 Lembah Melintang sebagai berikut : Perancangan context diagram menggambarkan satu lingkaran besar yang dapat mewakili seluruh proses yang terdapat suatu sistem. Gambar 2. ERD E-Tracer Study c. Use Case Diagram Gambar 4. Context Diagram E-Tracer Study Pada gambar 4 terlihat bahwa admin melakukan transaksi data yang berupa data alumni, Gambar 4. Context Diagram E-Tracer Study Gambar 2. ERD E-Tracer Study c. Use Case Diagram Use case diagram merupakan gambaran interaksi antara pengguna dengan sistem[4.]. Perancangan use case diagram serta identifikasi aktor yang terkait pada sistem sebagai berikut : Gambar 4. Context Diagram E-Tracer Study Pada gambar 4 terlihat bahwa admin melakukan transaksi data yang berupa data alumni, Gambar 4. Context Diagram E-Tracer Study g y Pada gambar 4 terlihat bahwa admin melakukan transaksi data yang berupa data alumni, E-ISSN: 2716-3989 3 2. Pembahasan Poin dari rancangan user interface telah menampilkan halaman-halaman dan fitur-fitur dari sistem Electronic Tracer Study (E-Tracer study) SMKN 1 Lembah Melintang. Berikut pembahasan mengenai hasil rancangan dari tampilan tersebut. Poin dari rancangan user interface telah menampilkan halaman-halaman dan fitur-fitur dari sistem Electronic Tracer Study (E-Tracer study) SMKN 1 Lembah Melintang. Berikut pembahasan mengenai hasil rancangan dari tampilan tersebut. Gambar 5. Tampilan Login Gambar 5. Tampilan Login g 1. Halaman Login Tampilan halaman ini dikendalikan oleh Sitecontroller.php dengan memanggil view berupa login.php melalui actionLogin(). Yii secara built-in telah mendukung fitur authentication[3]. Authentication yaitu sebuah metode yang digunakan untuk melakukan validasi (tindakan pembuktian) terhadap identitas seorang pengguna ketika mengakses sebuah sistem. Proses ini memastikan bahwa informasi pengguna benar-benar asli atau pengguna yang mengakses benar pengguna yang dimaksud. Dalam proses pembuktian identitas pengguna sistem biasanya menggunakan sebuah akun yang akan diverifikasi berupa username dan password pada halaman login sistem. a. Halaman Utama 1. Halaman Utama Admin l i i j k 1. Halaman Utama Admin l i i j k 1. Halaman Utama Admin Halaman ini menunjukan pada halaman utama admin mempunyai fungsi untuk menampilkan menu-menu yang dapat diakses oleh akan dengan akses admin dan bisa melakukan penambahan akun. Halaman utama admin hanya dapat diakses level admin. Gambar 6. Halaman Utama Admin Halaman login pada e-tracer study alumni adalah halaman yang pertama kali tampil ketika url diakses. Halaman ini akan menerima pengguna untuk memasukkan username dan password. Pengguna- pengguna yang sudah memiliki akun maka dapat menginputkan username dan password. Jika username dan password benar, pengguna akan diarahkan ke halaman utama. Namun, jika salah, maka pengguna mendapatkan pesan error dan dapat mengulangi kembali memasukkan username dan password yang benar. Jika pengguna lupa password silakan ke menu lupa password. Gambar 6. Halaman Utama Admin 2. Halaman Utama Operator Halaman utama pada operator menunjukan halaman utama yang mempunyai fungsi unntuk menampilkan menu-menu yang dapat diakses dengan akses operator. 2) Halaman Detail Input Kuisioner Kuisioner baru yang telah dimasukan akan dan berhasil mengklik tombol create, maka halaman akan beralih ke detail kuisioner tersebut. Pada halaman detail kuisioner akan terdapat tiga tombol aksi yaitu : a) Tombol ubah, tombol yang digunakan untuk melakukan pengubahan data apabila terjadi kesalahan ketika memasukkan data. 2) Data kuisioner jawaban, berupa informasi, penambahan, penubahan, dan penghapusan data jawaban kuisioner yang diinputkan operator. 2) Data kuisioner jawaban, berupa informasi, penambahan, penubahan, dan penghapusan data jawaban kuisioner yang diinputkan operator. b) Tombol cetak, tombol yang digunakan untuk mencetak data 3) Data berita, berupa data dan informasi yang berkaitan dengan berita terkini yang di inputkan oleh operator dan dapat melakukan pengubahan dan penghapusan berita. c) Tombol hapus, tombol yang digunakankan untuk menghapus data dari daftar kuisioner yang telah dimasukkan. 4) Data pengumuman, berupa penambahan, pengubahan, dan penghapusan data pengumuman yang dapat dilakukan oleh operator. H l U K l S k l h 2. Halaman Utama Halaman utama adalah halaman yang akan diakses oleh pengguna ketika berhasil login. Halaman ini merupakan fitur authorization yang disediakan Yii2 P-ISSN: 2302-3295 4 Vol. 10, No. 3, September 2022 VoteTEKNIKA Vol. 10, No. 3, September 2022 Vot 3. Halaman CRUD Yii2 framework telah memudahkan developer dalam membuat sebuah CRUD. Yii2 memiliki tools kode generator bernama Gii yang berfungsi untuk meng-generate CRUD, sehingga membangun aplikasi menjadi lebih cepat. Tools Gii yang telah di generate dapat diskustomisasi sesuai dengan kebutuhan pemgembang (Hafid, 2016). Pada proses CRUD, utamanya telah menyediakan tabel penyimpanan pada database. Melalui generator gii, maka akan di- generate Model, View, dan Controller-nya. 1) Data admin, berupa informasi semua data admin yang terdaftar dan dapat melakukan penambahan, pengubahan, dan penghapusan data. 2) Data operator, berupa informasi data operator yang diinputkan oleh admin. 2) Data operator, berupa informasi data operator yang diinputkan oleh admin. 3) Data kepala sekolah, berupa data dan informasi yang berkaitan dengan kepala sekolah yang di inputkan oleh admin dan dapat melakukan pengubahan dan penghapusan data. a. Halaman CRUD Kuisioner b. Halaman Utama Operator Halaman utama operator merupakan halaman yang dapat diakses setelah login. Halaman ini yang akan ditampilkan ketika pengguna masuk atau login menggunakan level akun “operator”. Menu yang akan ditampilkan pada akun dengan level operator sebagai berikut. 1) Data kuisioner, berupa informasi data kuisioner yang telah dimasukan. Data kuisioner dapat dilakukan penambahan, pengubahan, dan penghapusan data kuisioner. 2) Halaman Detail Input Kuisioner Vol. 10, No. 3, September 2022 V (hafid, 2016). Proses authorization adalah proses menentukan action sebagai tindakan menentukan pengguna memiliki hak melakukan aktivitas tertentu terhadap sistem. Aktivitas pengguna akan berbeda- beda sesuai dengan hak akses dari masing-masing pengguna, sehingga pada setiap pengguna mendapatkan menu yang berbeda berdasarkan level akses masing- masing akun. grafik. Data grafik tracer study tersebut dapat dicetak jika dibutuhkan. 2) Data tracer study, berupa laporan informasi data kuisioner yang telah dimasukan menjadi bentuk data. Data tracer study tersebut dapat dicetak jika dibutuhkan. d. Halaman Utama Alumni Halaman utama alumni merupakan halaman dapat diakses setelah login. Halaman ini yang akan ditampilkan ketika pengguna masuk atau login menggunakan level akun „alumni‟. a. Halaman Utama Admin Halaman utama admin merupakan halaman dapat diakses setelah login. Halaman ini yang akan ditampilkan ketika pengguna masuk atau login menggunakan level akun “admin”. Menu yang akan ditampilkan pada akun dengan level admin sebagai berikut. 3) Halaman Index Kuisioner Halaman ini adalah halaman yang menampilkan semua data yang telah dimasukkan. Halaman ini diberi kolom pencarian. Pencarian berdasarkan kode dan judul kuisioner hanya dengan memasukan kode dan judul pada kolom yang disediakan. Pada tabel data kuisioner disediakan sebuah icon bentuk mata. Icon tersebut berfungsi untuk menampilkan kembali halaman detail kuisioner. y g p p c. Halaman Utama Kepala Sekolah 1) Halaman Create Kuisioner 4) Data alumni, berupa penambahan, pengubahan, dan penghapusan data alumni yang dapat dilakukan oleh admin. Halaman ini akan menampilkan seluruh form yang akan diisi untuk membuat kuisioner baru. Pengguna akan membutuhakan kode kuisioner. Pada sistem ini pencarian kode dilakukan dengan autocomplete, yaitu dengan memasukan nomor kode atau judul yang dibutuhkan pada field, maka otomatis akan ter-filter dan yang diinputkan telah terdaftar pada sistem. 4) Data alumni, berupa penambahan, pengubahan, dan penghapusan data alumni yang dapat dilakukan oleh admin. b. Halaman Utama Operator 1) Halaman Create Berita Halaman ini akan menampilkan seluruh form berita yang akan diisi untuk membuat berita baru. Pengguna akan membutuhkan id berita. Pada sistem ini pencarian id dilakukan dengan autocomplete, yaitu dengan memasukan nomor kode atau judul yang dibutuhkan pada field, maka otomatis akan ter-filter yang diinputkan telah terdaftar pada sistem. 2) Halaman Detail Input Berita 3) Halaman Index Kuisioner Jawaban Halaman ini merupakan halaman yang menampilkan semua data yang telah dimasukkan. Halaman ini diberi kolom pencarian. Pencarian berdasarkan id dan judul pengumuman hanya dengan memasukkan id dan judul pada kolom yang disediakan. Pada tabel data-data pengumuman disediakan sebuah icon bentuk mata. Icon tersebut berfungsi untuk menampilkan kembali halaman detail pengumuman. Halaman ini adalah halaman yang menampilkan semua data yang telah dimasukkan. Halaman ini diberi kolom pencarian. Pencarian berdasarkan kode dan judul kuisioner jawaban hanya dengan memasukkan kode dan judul pada kolom yang disediakan. Halaman ini adalah halaman yang menampilkan semua data yang telah dimasukkan. Halaman ini diberi kolom pencarian. Pencarian berdasarkan kode dan judul kuisioner jawaban hanya dengan memasukkan kode dan judul pada kolom yang disediakan. c. Halaman CRUD Berita b. Halaman Detail Kuisioner Jawaban 1) Halaman Create Kuisioner Jawaban Halaman ini akan menampilkan seluruh form yang akan diisi untuk membuat kuisioner jawaban baru. Pengguna akan membutuhkan kode kuisioner jawaban. Pada sistem ini pencarian kode dilakukan dengan autocomplete, yaitu dengan memasukkan nomor kode atau judul yang dibutuhkan pada field, maka otomatis akan ter-filter yang diinputkan telah terdaftar pada sistem. d. Halaman CRUD Pengumuman d. Halaman CRUD Pengumuman 1) Halaman Create Pengumuman Halaman create pengumuman ini akan menampilkan seluruh form pengumuman yang akan diisi untuk membuat informasi terbaru. 2) Halaman Input Pengumuman Informasi 2) Halaman Input Pengumuman Informasi p 2) Halaman Detail Input Kuisioner Jawaban pengumuman terbaru yang telah dimasukkan dan berhasil mengklik tombol create, maka halaman akan beralih ke detail berita tersebut. Pada halaman detail pengumuman akan terdapat tiga tombol aksi yaitu : jawaban baru yang telah dimasukkan akan berhasil mengklik tombol create, maka halaman akan berpindah ke detail kuisioner tersebut. Pada halaman detail kuisioner akan terdapat tiga tombol aksi yaitu : a) Tombol ubah, tombol yang digunakan untuk melakukan pengubahan data apabila terjadi kesalahan ketika memasukkan data. a) Tombol ubah merupakan tombol yang digunakan untuk melakukan pengubahan data apabila terjadi kesalahan ketika memasukan data. e) Tombol cetak merupakan tombol yang digunakan untuk mencetak data. b) Tombol cetak, tombol yang digunakan untuk mencetak data f) Tombol hapus merupakan tombol yang digunakan untuk menghapus data dari daftar pengumuman yang telah dimasukkan. f) Tombol hapus merupakan tombol yang digunakan untuk menghapus data dari daftar pengumuman yang telah dimasukkan. c) Tombol hapus, tompol yang digunakankan untuk menghapus data dari daftar kuisioner yang telah dimasukan. 3) Halaman Index Berita Halaman ini adalah halaman yang menampilkan semua data yang telah dimasukkan. Halaman ini diberi kolom pencarian. Pencarian berdasarkan id dan judul berita hanya dengan memasukkan id dan judul pada kolom yang disediakan. Pada tabel data berita disediakan sebuah icon bentuk mata. 3) Halaman Index Pengumuman 3) Halaman Index Kuisioner Jawaban A Vol. 10, No. 3, September 2022 halaman detail view. Pada halaman ini, pengguna mendapatkan data berupa file pdf yang siap untuk di cetak. Maka hasilnya akan berupa hardcopy. a) Tombol ubah, tombol yang digunakan untuk melakukan pengubahan data apabila terjadi kesalahan ketika memasukkan data. halaman detail view. Pada halaman ini, pengguna mendapatkan data berupa file pdf yang siap untuk di cetak. Maka hasilnya akan berupa hardcopy. b) Tombol cetak, tombol yang digunakan untuk mencetak data Halaman Cetak ini menggunakan sebuah komponen library tambahan pada yii2 untuk menghasilkan file PDF dengan cepat menggunakan konten HTML, yaitu extension yii2- mpdf. Extension ini sebagai cara mudah untuk mengintegrasikan dan mengekspor data serta menghasilkan laporan dalam bentuk PDF (www.github.com). c) Tombol hapus, tombol yang digunakan untuk menghapus data dari daftar berita yang telah dimasukkan. 3) Halaman Index Berita c. Halaman Utama Kepala Sekolah Halaman utama kepala sekolah merupakan halaman dapat diakses setelah login. Halaman ini yang akan ditampilkan ketika pengguna masuk atau login menggunakan level akun ”kepala sekolah‟. Menu yang akan ditampilkan pada akun dengan level kepala sekolah sebagai berikut. 4) Halaman Cetak Kuisioner 1) Data grafik tracer study, berupa laporan informasi data kuisioner yang telah dimasukan menjadi bentuk Halaman cetak data kuisioner akan dapat diakses melalui tombol cetak. Tombol cetak terdapat pada E-ISSN: 2716-3989 E-ISSN: 2716-3989 5 V. SARAN Sistem electronic tracer study (e-tracer study) alumni pada SMKN 1 Lembah Melintang ini diharapkan untuk pengembangan selanjutnya agar data-dari alumni dapat terintegrasi pada Badan Akreditasi Nasional Sekolah/Madrasah (BAN-S/M), sehingga tidak hanya terpaku pada SMKN 1 Lembah Melintang saja. Serta diharapkan juga kedepannya sistem e-tracer study alumni ini akan terdaftar ke digital signature dalam pengesahan tanda tangan secara hukum untuk menghindari pemalsuan data melalui internet. IV. KESIMPULAN Kesimpulan dari tugas akhir ini yaitu menghasilkan perancangan sistem electronic tracer study (e-tracer study) alumni pada SMKN 1 Lembah Melintang yang berbasis web mobile. Sistem e-tracer study merupakan sebuah sistem yang dibangun dengan tujuan memudahkan sekolah dalam mengelola data alumni, seperti penataan data alumni, pengelolaan informasi, dan sebagai salah satu pendukung dalam proses peningkatan akreditasi sekolah. Berita baru yang telah dimasukan akan dan berhasil mengklik tombol create, maka halaman akan beralah ke detail berita tersebut. Pada halaman detail berita akan terdapat tiga tombol aksi yaitu : P-ISSN: 2302-3295 6 Vol. 10, No. 3, September 2022 VoteTEKNIKA Vol. 10, No. 3, September 2022 VoteTEKNIKA DAFTAR PUSTAKA [1]Mukhlasin, Hafid. Membangun Aplikasi Profesional Berbasis Web Menggunakan Yii Framework. Buku Baik, Jakarta. 2016. [2]Anhar. Panduan Menguasai PHP dan MySQL secara Otodidak. Mediakita. Jakarta. 2010. [3]Arif, Muhammad. Pemodelan Sistem.: Deepublish. Yogyakarta. 2017. [3]Arif, Muhammad. Pemodelan Sistem.: Deepublish. Yogyakarta. 2017. [4] Wijaya, Yahya Dwi and Muna Wardah Astuti. “Sistem Informasi Penjualan Tiket Wisata Berbasis Web Menggunakan Metode Waterfall.” Seminar Nasional Teknologi Informasi dan Komunikasi. 2019. [5] Binanto, Iwan. “Analisa Metode Classic Life Cycle (Waterfall) Untuk Pengembangan Perangkat Lunak Multimedia”. SeNASTI. 2014. [6] Fundamental: Model-View-Controller (MVC) \| Panduan Definitif Yii 1.1 \| Yii PHP Framework Website: https://www.yiiframework.com/doc/guide/1.1/id basics.mvc diakses 24 Februari 2022. E-ISSN: 2716-3989 E-ISSN: 2716-3989 7
https://openalex.org/W2809351581	https://eprint.ncl.ac.uk/fulltext.aspx?url=250361/5E12F04A-053E-4465-B1A9-3643C1B45AA1.pdf&pub_id=250361	English	null	The INTENSE project: using observations and models to understand the past, present and future of sub-daily rainfall extremes	Advances in science and research	2,018	cc-by	8,352	17th EMS Annual Meeting: European Conference for Applied Meteorology and Climatology 2017 17th EMS Annual Meeting: European Conference for Applied Meteorology and Climatology 2017 17th EMS Annual Meeting: European Conference for Applied Meteorology and Climatology 2017 17th EMS Annual Meeting: European Conferenc Adv. Sci. Res., 15, 117–126, 2018 https://doi.org/10.5194/asr-15-117-2018 © Author(s) 2018. This work is distributed under the Creative Commons Attribution 4.0 License. S Annual Meeting: European Conference for Applied Meteorolo The INTENSE project: using observations and models to understand the past, present and future of sub-daily rainfall extremes European Conference for Applied Meteorology and Climatolog Stephen Blenkinsop1, Hayley J. Fowler1, Renaud Barbero2, Steven C. Chan1, Selma B. Guerreiro1, Elizabeth Kendon3, Geert Lenderink4, Elizabeth Lewis1, Xiao-Feng Li1, Seth Westra5, Lisa Alexander6, Richard P. Allan7, Peter Berg8, Robert J. H. Dunn3, Marie Ekström9, Jason P. Evans6, Greg Holland10, Richard Jones3, Erik Kjellström8, Albert Klein-Tank4, Dennis Lettenmaier11, Vimal Mishra12, Andreas F. Prein10, Justin Shefﬁeld13, and Mari R. Tye10 1School of Engineering, Newcastle University, Newcastle upon Tyne, UK 2National Research Institute of Science & Technology for Environment & Agriculture, Aix-en-Provence, France 3Met Ofﬁce Hadley Centre, Exeter, UK 4Royal Netherlands Meteorological Institute, De Bilt, the Netherlands 5School of Civil, Environmental and Mining Engineering, University of Adelaide, Adelaide, Australia 6Climate Change Research Centre, University of New South Wales, Sydney, Australia 7Department of Meteorology, University of Reading, Reading, UK 8Swedish Meteorological and Hydrological Institute, Norrköping, Sweden 9School of Earth and Ocean Sciences, Cardiff University, Cardiff, UK 10National Center for Atmospheric Research, Boulder, CO, USA 11Department of Geography, UCLA, Los Angeles, USA 12Indian Institute of Technology Gandhinagar, Gandhinagar, India 13Geography and Environment, University of Southampton, Southampton, UK Correspondence: Stephen Blenkinsop (stephen.blenkinsop@newcastle.ac.uk) Received: 14 February 2018 – Accepted: 5 June 2018 – Published: 19 June 2018 Stephen Blenkinsop1, Hayley J. Fowler1, Renaud Barbero2, Steven C. Chan1, Selma B. Guerreiro1, Elizabeth Kendon3, Geert Lenderink4, Elizabeth Lewis1, Xiao-Feng Li1, Seth Westra5, Lisa Alexander6, Richard P. Allan7, Peter Berg8, Robert J. H. Dunn3, Marie Ekström9, Jason P. Evans6, Greg Holland10, Richard Jones3, Erik Kjellström8, Albert Klein-Tank4, Dennis Lettenmaier11, Vimal Mishra12, Andreas F. Prein10, Justin Shefﬁeld13, and Mari R. Tye10 Stephen Blenkinsop1, Hayley J. Fowler1, Renaud Barbero2, Steven C. Chan1, Selma B. Guerreiro1, Elizabeth Kendon3, Geert Lenderink4, Elizabeth Lewis1, Xiao-Feng Li1, Seth Westra5, Lisa Alexander6, Richard P. Allan7, Peter Berg8, Robert J. H. Dunn3, Marie Ekström9, Jason P. Evans6, Greg Holland10, Richard Jones3, Erik Kjellström8, Albert Klein-Tank4, Dennis Lettenmaier11, Vimal Mishra12, Andreas F. Prein10, Justin Shefﬁeld13, and Mari R. 1 Introduction Changes in short-duration, heavy rainfall events are an im- portant ﬁngerprint of anthropogenic climate change (Hegerl et al., 2015) and are crucial to quantify as they are associ- ated with ﬂash ﬂooding which poses a signiﬁcant threat to lives, infrastructure, and natural ecosystems. Quantiﬁed esti- mates of extreme rainfall intensities at sub-daily timescales (down to a few minutes) and spatial scales of 1–10 km2 are needed for urban drainage design (Arnbjerg-Nielsen et al., 2013) and therefore for adaptation to future climate change. Observations conﬁrm basic physics predicting increased at- mospheric moisture with warming, fuelling intensiﬁcation of heavy rainfall (Trenberth et al., 2003). However, under- standing how changes to atmospheric moisture will combine with changes in circulation dynamics in a warming world to strengthen or weaken regional increases in intense rainfall re- mains a key challenge for climate change research (Pfahl et al., 2017). Achieving this requires the improved availability of high-quality, high-resolution rainfall observations (Wes- tra, et al., 2014; Wilby et al., 2017), advances in climate mod- elling on convection-permitting scales (Westra et al., 2014; Prein et al., 2015, 2017a) and an improved understanding of atmospheric processes that contribute to intense rainfall (O’Gorman, 2015; Lenderink and Fowler, 2017; Pfahl et al., 2017). – To provide new understanding of the inﬂuence of cli- mate model resolution and structure on the simula- tion of rainfall extremes for different climate regimes (Sect. 2.3); – To combine models and observations to improve un- derstanding of the drivers of rainfall extremes across multiple timescales, examining the inﬂuence of local thermodynamics and large-scale atmospheric circula- tion modes (Sect. 2.4); and – To use this knowledge to (i) provide a better understand- ing of the likely responses to warming of rainfall ex- tremes at different timescales and geographic locations, and (ii) use information from climate models in a more informed way to improve climate change adaptation de- cision making (Sect. 3). The project structure, main data requirements and outputs are summarised in Fig. 1. Subsequent sections in this pa- per provide a brief review of current understanding around these themes, incorporating the contributions made by IN- TENSE researchers and partners (reﬂected in the authorship of this paper) to this literature. More details of project out- puts, including publications, may be found on the INTENSE website1 and a summary is provided in Table 1. 1https://research.ncl.ac.uk/intense/outputs/ (last access: 12 June 2018). S. Blenkinsop et al.: The INTENSE project S. Blenkinsop et al.: The INTENSE project – To use this dataset to quantify the nature of global rain- fall extremes across multiple timescales and, where pos- sible, quantify recent change (Sect. 2.2); 1 Introduction This paper summarises how these challenges are being ad- dressed by the INTENSE (INTElligent use of climate mod- els for adaptatioN to non-Stationary hydrological Extremes) project to provide new knowledge relevant to adapting to hy- droclimatic risks under climate change. 2.1 Data collection and provision In situ precipitation data at sub-daily timescales are needed to quantify the characteristics of extreme events for important research questions including the detection of change and for practical applications such as urban drainage design. They are also valuable for the validation of radar and satellite prod- ucts (Hegerl et al., 2015) and the outputs from climate mod- els (Prein et al., 2015). However, such data are not as exten- sive, either in time or space, as those for daily rainfall to- tals (Westra et al., 2014; Hegerl et al., 2015), are subject to a range of sampling and instrument errors and inhomogeneities (Blenkinsop et al., 2017; Westra, et al., 2014; Wilby et al., 2017) and are less freely available compared with daily data (Hegerl et al., 2015). Consequently, compared with extreme rainfall on timescales of a day or longer, globally, sub-daily events have been little studied. The INTENSE project: using observations and models to understand the past, present and future of sub-daily rainfall extremes Tye10 1School of Engineering, Newcastle University, Newcastle upon Tyne, UK 2National Research Institute of Science & Technology for Environment & Agriculture, Aix-en-Provence, France 3Met Ofﬁce Hadley Centre, Exeter, UK 4Royal Netherlands Meteorological Institute, De Bilt, the Netherlands 5School of Civil, Environmental and Mining Engineering, University of Adelaide, Adelaide, Australia 6Climate Change Research Centre, University of New South Wales, Sydney, Australia 7Department of Meteorology, University of Reading, Reading, UK 8Swedish Meteorological and Hydrological Institute, Norrköping, Sweden 9School of Earth and Ocean Sciences, Cardiff University, Cardiff, UK 10National Center for Atmospheric Research, Boulder, CO, USA 11Department of Geography, UCLA, Los Angeles, USA 12Indian Institute of Technology Gandhinagar, Gandhinagar, India 13Geography and Environment, University of Southampton, Southampton, UK Correspondence: Stephen Blenkinsop (stephen.blenkinsop@newcastle.ac.uk) Received: 14 February 2018 – Accepted: 5 June 2018 – Published: 19 June 2018 ce for Applied Meteorology and Climatology 2017 ed Meteorology and Climatology 2017 Correspondence: Stephen Blenkinsop (stephen.blenkinsop@newcastle.ac.uk) Received: 14 February 2018 – Accepted: 5 June 2018 – Published: 19 June 2018 Abstract. Historical in situ sub-daily rainfall observations are essential for the understanding of short-duration rainfall extremes but records are typically not readily accessible and data are often subject to errors and inho- mogeneities. Furthermore, these events are poorly quantiﬁed in projections of future climate change making adaptation to the risk of ﬂash ﬂooding problematic. Consequently, knowledge of the processes contributing to intense, short-duration rainfall is less complete compared with those on daily timescales. The INTENSE project is addressing this global challenge by undertaking a data collection initiative that is coupled with advances in high-resolution climate modelling to better understand key processes and likely future change. The project has so far acquired data from over 23 000 rain gauges for its global sub-daily rainfall dataset (GSDR) and has provided evidence of an intensiﬁcation of hourly extremes over the US. Studies of these observations, combined with model simulations, will continue to advance our understanding of the role of local-scale thermodynamics and large-scale atmospheric circulation in the generation of these events and how these might change in the future. Published by Copernicus Publications. 118 2 The INTENSE project INTENSE is the ﬁrst major international effort to focus on global sub-daily rainfall extremes, enabling substantial ad- vances in quantifying observed historical changes and pro- viding the physical understanding of processes necessary for improved regional prediction of change. The project leads the global research effort within the Global Energy and Wa- ter Exchanges (GEWEX) Hydroclimatology Panel Cross- Cutting project on sub-daily precipitation extremes, address- ing the World Climate Research Programme “Grand Chal- lenge” on extremes (Alexander et al., 2016). To develop a more thorough understanding of the relationship between large-scale warming, atmospheric circulation and sub-daily extreme rainfall the project is addressing the research areas set out by Westra et al. (2014) through the following aims: The INTENSE project is therefore focussing on collect- ing sub-daily rain gauge records, which provide the most accurate representation of the amount of water reaching the ground at a given location. To date, the global sub-daily rain- fall dataset (GSDR) we have compiled comprises data from 23 687 gauges (Lewis et al., 2018a), having an average record – To undertake a unique, global-scale data collection ef- fort of sub-daily rainfall data and apply a set of quality control methods to construct a new, high-quality global dataset (Sect. 2.1); Adv. Sci. Res., 15, 117–126, 2018 www.adv-sci-res.net/15/117/2018/ 119 S. Blenkinsop et al.: The INTENSE project Figure 1. INTENSE project research themes, information ﬂows and outcomes. The collection of in situ sub-daily rain gauge data represents a core project activity (blue arrows). Thermodynamic and atmospheric/ocean processes are considered independently (red and green arrows) but also in an integrated manner (yellow arrows). These data and knowledge will be used to evaluate model simulations of extreme rainfall across different timescales and locations. Model outputs will then be used for the development of process-based downscaling methods and provide guidance on the use of projections (black arrow). Data in dashed boxes denotes externally produced data used in the project. Figure 1. INTENSE project research themes, information ﬂows and outcomes. The collection of in situ sub-daily rain gauge data represents a core project activity (blue arrows). Thermodynamic and atmospheric/ocean processes are considered independently (red and green arrows) but also in an integrated manner (yellow arrows). These data and knowledge will be used to evaluate model simulations of extreme rainfall across different timescales and locations. 2 The INTENSE project List of countries/data sources contributing at least 10 gauges with a minimum of 1 year of hourly data from the combined data collection exercise and ISD database. Numbers denote the number of gauges. (Abbreviations: US – USA, AU – Australia; UK – United Kingdom, JP – Japan, DE – Germany, NO – Norway, CH – Switzerland, IT – Italy, SE – Sweden, ES – Spain, MY – Malaysia, FR – France, ISD – Non-allocated gauges from the ISD database, FI – Finland, PT – Portugal, BE – Belgium, BR – Brazil, CZ – Czechoslovakia, NL – the Netherlands, SG – Singapore, IE – Republic of Ireland, RO – Romania, PA – Panama, CR – Costa Rica. rainfall indices based on the ETCCDI indices already avail- able for daily data (Peterson, 2005; Donat et al., 2013). These new indices have been identiﬁed following discussions with the climate observations and modelling communities and de- scribe important attributes of sub-daily extremes, where pos- sible, corresponding to existing daily ETCCDI indices. For example, indices of monthly maxima of hourly and multi- hourly rainfall will be provided along with monthly counts of threshold exceedances, as well as indices reﬂecting the di- urnal cycle (e.g. monthly index of the wettest hour). These indices are consistent with the naming and methodological conventions of the ETCCDI daily indices for easy applica- tion by users and the data will be made freely accessible for research purposes. length of 13 years at an hourly or multi-hourly resolution. Most gauge records commence after 1990 with some con- tinuing up to 2017 although there is a decrease in gauge availability in recent years. Just under half of these are from the Integrated Surface Database (ISD) (Smith et al., 2011), with the remainder being provided by national meteorologi- cal and environmental agencies. Figure 2 identiﬁes all those countries contributing at least 10 gauges with a minimum of 1 year of hourly data. Sub-hourly resolution data has also been obtained for the UK and Australia and similar resolu- tion data will be collected for the US to quantify extremes over shorter durations but also to improve estimates of hourly extremes that may be underestimated due to temporal aggre- gation (Morbidelli et al., 2017). 2 The INTENSE project Model outputs will then be used for the development of process-based downscaling methods and provide guidance on the use of projections (black arrow). Data in dashed boxes denotes externally produced data used in the project. Table 1. Key outputs from the INTENSE project researchers and partners at time of publication. Output Summary Global sub-daily rainfall dataset (GSDR) Hourly rainfall data collected from 23 countries (excluding ISD). Globally applicable quality control process Applied to hourly rainfall accumulations. Will be open-source, and freely available. Deﬁnition of 13 sub-daily rainfall indices Consistent with ETCCDI daily indices but reﬂecting the intensity, frequency and time of occurrence of sub-daily extremes. These will be freely available to the international community and hosted on the Deutscher Wetterdienst (DWD) GPCC website. 15 peer reviewed journal articles Published research progressing understanding of observed change of sub-daily extremes, thermo-dynamic and large-scale drivers, and the de- velopment and analysis of high resolution convection-permitting mod- els. Includes continental scale analysis of historical US hourly rainfall. Great Britain gridded 1 km hourly rainfall data (CEH-GEAR1hr) Hourly 1 km gridded rainfall dataset based on hourly gauge data and daily gridded dataset. Will be hosted and periodically updated along- side the CEH-GEAR dataset by the Centre for Ecology and Hydrology (CEH) in the UK. Table 1. Key outputs from the INTENSE project researchers and partners at time of publication. Adv. Sci. Res., 15, 117–126, 2018 www.adv-sci-res.net/15/117/2018/ 120 S. Blenkinsop et al.: The INTENSE project 120 S. Blenkinsop et al.: The INTENSE project Figure 2. List of countries/data sources contributing at least 10 gauges with a minimum of 1 year of hourly data from the combined data collection exercise and ISD database. Numbers denote the number of gauges. (Abbreviations: US – USA, AU – Australia; UK – United Kingdom, JP – Japan, DE – Germany, NO – Norway, CH – Switzerland, IT – Italy, SE – Sweden, ES – Spain, MY – Malaysia, FR – France, ISD – Non-allocated gauges from the ISD database, FI – Finland, PT – Portugal, BE – Belgium, BR – Brazil, CZ – Czechoslovakia, NL – the Netherlands, SG – Singapore, IE – Republic of Ireland, RO – Romania, PA – Panama, CR – Costa Rica. S. Blenkinsop et al.: The INTENSE project Figure 2. 2 The INTENSE project The collection process revealed that even where data are available they may not be adequate for the analysis of ex- treme events – for example, undocumented gauge break- downs, hourly values accumulated as daily totals, or mechan- ical failure of the rain gauges may produce erroneous rain- fall amounts (Blenkinsop et al., 2017). Furthermore, changes such as gauge location, site characteristics or equipment type may introduce inhomogeneities in climatic series (e.g. Bar- bero et al., 2017). We have therefore developed an automated quality control process for sub-daily data based on methods previously applied to the UK (Blenkinsop et al., 2017; Lewis et al., 2018b) but here used to produce a high-quality global dataset. 2.2 Quantifying the nature of extremes and observed change Analysis of this global dataset will help quantify the na- ture of extremes on multiple timescales, expanding regional analyses (e.g. Blenkinsop et al., 2017; Darwish et al., 2018; Forestieri et al., 2018a), and so improve understanding of current patterns of extreme rainfall globally. This will in- clude the spatial pattern of extreme intensities, their season- ality and diurnal cycles. Further, whilst existing studies of long-term trends and variability in sub-daily extremes have tended to cover relatively small spatial domains, INTENSE has already made progress analysing changes at national to continental scales, adding to the national scale studies of Sen Roy (2009) and Westra and Sisson (2011). Whilst US hourly Whilst data licensing restrictions mean that the GSDR dataset cannot immediately be made freely available, the data collected will be used to produce a set of sub-daily extreme Adv. Sci. Res., 15, 117–126, 2018 www.adv-sci-res.net/15/117/2018/ S. Blenkinsop et al.: The INTENSE project 121 tion models (Kendon et al., 2017). CPMs may also be neces- sary to provide improved quantiﬁcation of other hazards in- cluding lightning, hail and wind gusts (Kendon et al., 2017; Gadian et al., 2018). and daily seasonal maxima have increased, the percentage of stations showing signiﬁcant increasing annual maximum precipitation trends was generally higher for daily compared to hourly extremes (Barbero et al., 2017). Strong evidence though points to more widespread increases in the magni- tude and frequency of hourly extremes in winter compared to daily extremes. Changes in hourly extremes over the UK are currently difﬁcult to detect as substantial natural climate variability reduces the signal-to-noise ratio (Kendon et al., 2018). Although a recent increase in the intensity of UK sum- mer extremes has been observed, this is likely to be at least partly related to the natural variability of atmospheric and ocean modes. The ﬁrst study to look at changes in rainfall in long cli- mate change simulations at convection-permitting scale was carried out by the Met Ofﬁce Hadley Centre (Kendon et al., 2014). These simulations at a resolution of 1.5 km pro- vided evidence that intense summer rainfall could become heavier over the southern UK with almost ﬁve times more events exceeding 28 mm in one hour (indicative of poten- tial for surface water ﬂooding in the UK) in the future than in the current climate. 2.3 Improved modelling capabilities y CPM climate change simulations are now becoming avail- able across other parts of the globe. For example, using a larger North American domain, Prein et al. (2017c) used a 4 km resolution CPM to project increases in the maxi- mum rainfall rate in mesoscale convective systems (MCS) in the future. Although some consistent messages are emerging from these studies a better understanding of the differences between models is required, and it is therefore signiﬁcant that INTENSE is integrating with the main international effort to produce a coordinated set of convection permitting climate simulations through the CORDEX ﬂagship pilot study “Con- vective phenomena at high resolution over Europe and the Mediterranean”2. This will maximise the utility of the GSDR dataset and help to integrate knowledge on the key processes and drivers of extreme sub-daily rainfall. This pilot study is focused on Europe, with a number of different groups car- rying out < 3 km resolution climate simulations spanning a common Alpine domain plus, in some cases, additional Eu- ropean sub-domains. Traditional approaches to obtaining projections of regional precipitation change use an ensemble of regional climate models (RCMs) but these typically only provide outputs at a resolution of 10–50 km. Statistical downscaling methods, which include temporal disaggregation approaches, may be used to derive projections of change in sub-daily rainfall but these rely on the assumption that the relationship between rainfall at different timescales (e.g. daily to hourly) remains constant in the future. The INTENSE project has applied this statistical approach to produce projected future depth- duration-frequency (DDF) relationships for Sicily, Italy, for rainfall totals down to hourly durations (Forestieri et al., 2018b). However, although state-of-the-art climate models have a high degree of skill in simulating many features of our climate (Flato et al., 2013) they also have a number of known deﬁciencies. These include the simulation of intense rainfall due to their relatively coarse resolution that cannot explic- itly resolve convection which results in the consequent use of convective parameterisation schemes (e.g., Mishra et al., 2012; Mooney et al., 2017). The INTENSE project is linking its observational analyses with the development of very high resolution (< 5 km) convection-permitting models (CPMs) which allow dynamical representation of convection. 2http://www.cordex.org/blog/2018/01/26/endorsed-ﬂagship- pilot-studies/ (last access: 8 June 2018). 2.3 Improved modelling capabilities These have been shown to better simulate sub-daily rainfall charac- teristics (Kendon et al., 2012; Chan et al., 2014a; Prein et al., 2017b; Lind et al., 2016) including the relationship between temperature and extreme rainfall (Chan et al., 2016a), and produce different projections of change in rainfall intensity, duration and extremes to those from standard coarse resolu- 2.2 Quantifying the nature of extremes and observed change These simulations have also now been supplemented with corresponding data for the northern half of the UK (Chan et al., 2018a). As part of INTENSE, projected changes in rainfall extremes across timescales in these CPM simulations have been explored alongside ob- served trends, with implications for detection and attribution studies (Kendon et al., 2018). Data at sub-hourly timescales from these CPMs have also been investigated, with similar changes in summer 10 min rainfall over the southern UK pro- jected as for hourly timescales (Chan et al., 2016b), but the lack of observed data has precluded model validation to date. The sub-hourly datasets collated in INTENSE will begin to address this deﬁciency. Our results so far emphasise the importance of considering relevant region-speciﬁc atmospheric processes in the assess- ment of change, and the need to examine sufﬁciently long time periods and/or large enough datasets. The provision of quality-controlled data means that INTENSE is providing the opportunity to further extend these analyses using the GSDR dataset; approximately 17 % of the gauges obtained so far cover a period of over 30 years necessary for this type of analysis (Lewis et al., 2018a) although this falls to ∼4.5 % for periods of over 50 years. www.adv-sci-res.net/15/117/2018/ 2.4 Drivers of changes in intense rainfall in a changing climate It has been recognised that analyses of observed changes are more powerful if they make use of and diagnose phys- ical mechanisms that are, or may be, responsible for change (Hegerl et al., 2015). INTENSE is therefore investigating the www.adv-sci-res.net/15/117/2018/ Adv. Sci. Res., 15, 117–126, 2018 122 S. Blenkinsop et al.: The INTENSE project S. Blenkinsop et al.: The INTENSE project role of local thermodynamics and large-scale atmosphere- ocean modes as drivers of changes in intense rainfall through linking observational analyses with those based on CPMs. Observational evidence and climate models suggest that rainfall will intensify with temperature according to the Clausius–Clapeyron (CC) relation (a rate of ∼6–7 % ◦C−1) (Trenberth et al., 2003; Allen and Ingram, 2002; Allan et al., 2010; Pall et al., 2007) representing an important compo- nent of change in the global water cycle (Hegerl et al., 2015). However, larger-scale modes of atmospheric and ocean vari- ability such as the El Niño–Southern Oscillation (ENSO), North Atlantic Oscillation (NAO) and monsoon systems are also important drivers of regional precipitation and extremes, whilst moisture transport features such as atmospheric rivers have been associated with winter ﬂooding in many regions (e.g. Lavers and Villarini, 2013a, b). However, the interac- tion between these large-scale and thermodynamic processes and their effect on extreme precipitation is not yet well un- derstood. large-scale vertical velocity; the most intense events were typically associated with high vertical velocities which cause a convergence of moist air at a scale of hundreds of kilome- ters. The use of CPMs is also increasingly offering the po- tential to contribute to such process understanding; Chan et al. (2018b) used simulations over the southern UK to show that stability, and to a lesser degree relative vorticity and mean sea level pressure, displayed some skill as predictors of hourly extremes. Prein et al. (2017c) showed that the maxi- mum rainfall rate in MCS over North America is increas- ing in line with CC but the response in rainfall volume over mesoscale catchments is increasing by up to two times CC rates due to the spread of heavy rainfall areas in future MCS. This is due to changes in the MCS cloud physics and internal dynamics (e.g. vertical mass ﬂuxes). 2.4 Drivers of changes in intense rainfall in a changing climate We think that the key challenge to provide improved pro- jections of future change in sub-daily extremes will be using both the global observation datasets and climate models to understand the contributions from (and interactions between) large-scale circulation and local thermodynamics (Barbero et al., 2018b). As the project progresses, the increasing avail- ability of quality-controlled observations from different cli- mate regimes, particularly in tropical regions, and the in- creasing modelling capability with mesoscale atmospheric climate models is strengthening our ability to meet this chal- lenge. Our work is using the new observational dataset and mod- elling capabilities to examine how rainfall extremes respond to increasing temperature and moisture availability globally. Kendon et al. (2018) used a CPM simulation over the south- ern UK to identify the earlier emergence of future changes in hourly extremes compared with daily extremes which might be expected from enhanced scaling at shorter durations. Such super-CC scaling at hourly timescales has been demonstrated in some regions (e.g. Lenderink and van Meijgaard, 2008) however, CC scaling identiﬁed in US daily annual maxima contrasts with lower scaling during most seasons for hourly extremes (Barbero et al., 2017). In the UK, CC scaling is only observed for hourly extremes in summer (Blenkinsop et al., 2015) with lower scaling at other times of the year. To date, most of such studies have examined the rainfall relation- ship with temperature, but dew point temperature is emerg- ing as a more appropriate variable to use in scaling studies as it provides a direct measure of humidity (Lenderink et al., 2017), enabling a better physical understanding of the scal- ing relationship (e.g. Lochbihler et al., 2017; Barbero et al., 2018a). INTENSE is taking advantage of global scale ob- served datasets such as HadISD (Dunn et al., 2016) to per- form such analyses more widely. www.adv-sci-res.net/15/117/2018/ 3 Summary and outlook This may in part be indicative of dif- ﬁculties associated with capturing changes in localised con- vective storms (Barbero et al., 2017; Kendon et al., 2018) but also emphasises the need to assess the role of large- scale circulation patterns and their relationship with local thermodynamic drivers (Lenderink and Fowler, 2017; Bar- bero et al., 2018b). This interaction between drivers remains one of the fundamental challenges to be addressed by IN- TENSE but the global extent of the project also enables it to address, new, emerging questions, for example, ongoing work in the project is examining the contrasting behaviour of extreme rainfall over urban and rural areas, linking observa- tional analyses with simulations from CPMs. INTENSE outputs could also be used to better understand how river catchments respond to intense rainfall, however, initial work applying uncorrected CPM simulations to the estimation of river ﬂows across catchments in the southern UK indicates greater biases than those from coarser climate models and clear differences in projected ﬂood peaks (Kay et al., 2015). This demonstrates the need for further work investigating the application of new knowledge and tools de- veloped in INTENSE for the assessment of future risks as- sociated with hydrological hazards. INTENSE will provide stakeholders with up-to-date and reliable return period esti- mates for different rainfall durations (e.g. Forestieri et al., 2018a; Darwish et al., 2018) and so contribute to a better understanding of current ﬂood risk. Combining this type of information with CPM simulations of future projections (e.g. Chan et al., 2014b) will help to make societies more resilient to ﬂooding from intense rainfall. Pioneering work to gain understanding of the beneﬁts of CPMs in representing extreme rainfall on climate timescales by the Met Ofﬁce Hadley Centre, in collaboration with IN- TENSE researchers, has led to the inclusion of CPM ensem- ble simulations (consisting of 10+ members at 2.2 km res- olution spanning the UK) in the UK’s next set of ofﬁcial climate change projections (UKCP, 2016) to be released in 2018. This work at the Met Ofﬁce Hadley Centre also now includes extensions to examine larger model domains, in- cluding the recent completion of a 2.2 km resolution Euro- pean domain simulation. Analysis of these simulations in IN- TENSE, along with comparison of results with those from other international modelling centres across Europe, Aus- tralia and the US (e.g. 3 Summary and outlook These INTENSE datasets provide a signiﬁcant platform for future development by the wider scientiﬁc com- munity and efforts are underway to identify a mechanism for their ongoing maintenance and updating to ensure a long- term legacy of the project. The gauge data and new, derived dataset of global sub-daily extreme indices will be hosted by the German meteorological service (Deutscher Wetterdienst – DWD) though the availability of the former will be limited due to licensing arrangements. An improved international ca- pacity to both monitor change and share data therefore re- mains a signiﬁcant challenge (Hegerl et al., 2015). ous regions. These INTENSE datasets provide a signiﬁcant platform for future development by the wider scientiﬁc com- munity and efforts are underway to identify a mechanism for their ongoing maintenance and updating to ensure a long- term legacy of the project. The gauge data and new, derived dataset of global sub-daily extreme indices will be hosted by the German meteorological service (Deutscher Wetterdienst ensembles of CPM simulations more feasible. Even when ensemble large-domain CPM simulations become common- place, biases will still remain which will require correction with statistical downscaling methods. g A signiﬁcant area of focus should also include strength- ening the link between climate science and impact science (Westra et al., 2014) and the wider impact community. Cen- tral to this is the relationship between extreme rainfall and ﬂood risk. Changes in future rainfall intensity can impact on the ﬂooding of urban drainage systems and pollution of coastal waters, for example, in the UK through combined sewer overﬂow (CSO) spills. Our high-resolution observa- tions and the UK Met Ofﬁce CPM simulations have been used to demonstrate the need for improved guidance on es- timates of change in rainfall intensity (uplifts) for UK water and sewerage companies and the effect of future increases in intensity on increased spills (Dale et al., 2017). g – DWD) though the availability of the former will be limited due to licensing arrangements. An improved international ca- pacity to both monitor change and share data therefore re- mains a signiﬁcant challenge (Hegerl et al., 2015). A continental-scale analysis using the sub-daily dataset has not provided evidence of super-CC scaling over the US (Barbero et al., 2017). 3 Summary and outlook The INTENSE project is increasing our understanding of ex- treme short-duration rainfall events worldwide, linking ob- servations and models to better understand the mechanisms associated with extreme rainfall that can lead to ﬂash ﬂood- ing. This is partly derived from a data collection initiative that has currently yielded 23 687 stations to produce the global sub-daily rainfall dataset (GSDR) which is being used to characterise current extremes and, by linking with the lat- est CPM simulations, better understand drivers of change. Data collection is ongoing and the project is actively seek- ing additional contributions to extend coverage (see Lewis et al., 2018a for dataset coverage). We have used the gauge data to produce a 1 km resolution gridded dataset of histor- ical hourly rainfall for Great Britain which will be freely available to hydrologists, climate modellers and other practi- tioners (Lewis et al., 2018b). INTENSE will further explore options for combining gauge, radar and satellite data into gridded sub-daily products to add to existing merged prod- ucts as a key resource for the climate modelling community to validate model outputs (Prein et al., 2015, 2017d). This will help to address problems with the representativeness of rain gauges, as networks are insufﬁciently dense to capture local-scale convective storms. In the future, model outputs could also be incorporated to provide valuable information in areas with low observational coverage such as mountain- It is becoming increasingly clear however that future rain- fall extremes cannot be simply extrapolated from present day scaling relationships and that multiple other factors – land surface characteristics, temperature range, atmospheric dy- namics, large scale circulation and moisture availability – all play key roles (Lenderink and Fowler, 2017; Lochbihler et al., 2017; Prein et al., 2017d; Ali and Mishra, 2017; Bao et al., 2017; Barbero et al., 2018a). Barbero et al. (2018b) found that, apart from in summer, mid-latitude synoptic pat- terns (including the jet stream and cutoff lows) are major contributors to hourly annual maxima across the western US. Lenderink et al. (2017) also identiﬁed the prominent role of circulation features over the Netherlands as measured by the www.adv-sci-res.net/15/117/2018/ Adv. Sci. Res., 15, 117–126, 2018 S. Blenkinsop et al.: The INTENSE project 123 ous regions. www.adv-sci-res.net/15/117/2018/ References Chan, S. C., Kendon, E. J., Fowler, H. J., Blenkinsop, S., Roberts, N., and Ferro, C. A. T.: The value of high-resolution Met Of- ﬁce Hadley Centre regional climate models in the simulation of multi-hourly precipitation extremes, J. Climate, 27, 6155–6174, 2014a. Alexander, L., Zhang, X., Hegerl, G., Seneviratne, S., Behrangi, A., Fischer, E., Martius, O., Otto, F., Sillmann, J., and Vautard, R.: Implementation plan for WCRP grand challenge on understand- ing and predicting weather and climate extremes. World Climate Research Programme Report: Geneva, Switzerland, available at: http://www.wcrp-climate.org/gc-extremes-documents (last ac- cess: 22 January 2018), 2016. Chan, S. C., Kendon, E. J., Fowler, H. J., Blenkinsop, S., and Roberts, N. M.: Projected increases in summer and win- ter UK sub-daily precipitation extremes from high-resolution regional climate models, Environ. Res. Lett., 9, 084019, https://doi.org/10.1088/1748-9326/9/8/084019, 2014b. Ali, H. and Mishra, V.: Contrasting response of rainfall extremes to increase in surface air and dewpoint temperatures at urban loca- tions in India, Sci. Rep., 7, 1228, https://doi.org/10.1038/s41598- 017-01306-1, 2017. Chan, S. C., Kendon, E. J., Roberts, N. M., Fowler, H. J., and Blenkinsop, S.: Downturn in scaling of UK extreme rainfall with temperature for future hottest days, Nat. Geosci., 9, 24–28, https://doi.org/10.1038/ngeo2596, 2016a. Allan, R. P., Soden, B. J., John, V. O., Ingram, W., and Good, P.: Current changes in tropical precipitation, Environ. Res. Lett., 5, 025205, https://doi.org/10.1088/1748-9326/5/2/025205, 2010. Chan, S. C., Kendon, E. J., Roberts, N. M., Fowler, H. J., and Blenkinsop, S.: The characteristics of summer sub-hourly rainfall over the southern UK in a high-resolution con- vective permitting model, Environ. Res. Lett., 11, 094024, https://doi.org/10.1088/1748-9326/11/9/094024, 2016b. Allen, M. R. and Ingram, W. J.: Constraints on future changes in climate and the hydrologic cycle, Nature, 419, 224–232, 2002. Argueso, D., Evans, J. P., Fita, L., and Bormann, K. J.: Temperature response to future urbanization and climate change, Clim. Dy- nam., 42, 2183–2199, https://doi.org/10.1007/s00382-013-1789- 6, 2014. Chan, S. C., Kahana, R., Kendon, E. J., and Fowler, H. J.: Pro- jected changes in extreme precipitation over Scotland and North- ern England using a high-resolution regional climate model, Clim. Dynam., https://doi.org/10.1007/s00382-018-4096-4, on- line ﬁrst, 2018a. Arnbjerg-Nielsen, K., Willems, P., Olsson, J., Beecham, S., Pathi- rana, A., Bülow Gregersen, I., Madsen, H., and Nguyen, V. T. V.: Impacts of climate change on rainfall extremes and urban drainage systems: a review, Water Sci. Technol., 68, 16–28, 2013. Chan, S. C., Kendon, E. 3 Summary and outlook A sum- mary of INTENSE project outputs and activities may be found on the project website: https://research.ncl.ac.uk/intense/aboutintense/ (last access: 12 June 2018). We thank the three anonymous reviewers for their helpful comments and insight. Barbero, R., Westra, S., Lenderink, G., and Fowler, H. J.: Temperature-extreme precpipitation scaling: a two- way causality?, Int. J. Climatol., 38, e1274–e1279, https://doi.org/10.1002/joc.5370, 2018a. Barbero, R., Abatzoglou, J. T., and, Fowler, H. J.: Contribu- tion of large-scale midlatitude disturbances to hourly pre- cipitation extremes in the United States, Clim. Dynam., https://doi.org/10.1007/s00382-018-4123-5, online ﬁrst, 2018b. Benestad, R., Sillmann, J., Thorarinsdottir, T. L., Guttorp, P., Mesquita, M. d. S., Tye, M. R., Uotila, P., Fox Maule, C., Thejll, P., Drews, M., and Parding, K. M.: New vigour involving statisti- cians to overcome ensemble fatigue, Nat. Clim. Change, 7, 697– 703, 2017. Blenkinsop, S., Chan, S. C., Kendon, E. J., Roberts, N. M., and Fowler, H. J.: Temperature inﬂuences on intense UK hourly precipitation and dependency on largescale circulation, Environ. Res. Lett., 10, 054021, https://doi.org/10.1088/1748- 9326/10/5/054021, 2015. Edited by: Rasmus Benestad Reviewed by: three anonymous referees Edited by: Rasmus Benestad Reviewed by: three anonymous referees Edited by: Rasmus Benestad Reviewed by: three anonymous referees Blenkinsop, S., Lewis, E., Chan, S. C., and Fowler, H. J.: An hourly precipitation dataset and climatology of extremes for the UK, Int. J. Climatol., 37, 722–740, https://doi.org/10.1002/joc.4735, 2017. 3 Summary and outlook Ban et al., 2015; Argueso et al., 2014; Rasmussen et al., 2014), including CORDEX, will lead to a better understanding of how results from CPMs differ from coarser resolution models more generally (Kendon et al., 2017) to offer robust guidance on their use. Data availability. No data sets were used in this article. Data availability. No data sets were used in this article. Competing interests. The authors declare that they have no con- ﬂict of interest. Special issue statement. This article is part of the special issue “17th EMS Annual Meeting: European Conference for Applied Me- teorology and Climatology 2017”. It is a result of the EMS Annual Meeting: European Conference for Applied Meteorology and Cli- matology 2017, Dublin, Ireland, 4–8 September 2017. One of the key challenges limiting the application of CPMs is still the computational demand and consequent ﬁ- nancial cost they impose. This will in part be addressed by continued advances in computing capabilities but a more in- telligent use of these models, aided by novel statistical ap- proaches is necessary for the foreseeable future (Benestad et al., 2017). Chan et al. (2018b) identiﬁed large-scale predic- tors that have some skill in predicting when dynamical down- scaling using a CPM is needed. This could potentially allow the targeting of high-resolution models to simulate only peri- ods with a high likelihood of extremes, thus making larger Acknowledgements. Hayley J. Fowler leads the European Research Council INTENSE project (ERC-2013-CoG-617329) which has also funded Renaud Barbero, Stephen Blenkinsop, Selma B. Guerreiro, Geert Lenderink, Elizabeth Lewis and Xiao-Feng Li. Hayley J. Fowler is also funded by the Wolfson Foundation and the Royal Society as a Royal Society Wolfson www.adv-sci-res.net/15/117/2018/ Adv. Sci. Res., 15, 117–126, 2018 S. Blenkinsop et al.: The INTENSE project S. Blenkinsop et al.: The INTENSE project 124 Research Merit Award (WM140025) holder. NCAR is sponsored by the National Science Foundation. Richard Allan is funded by the NERC SINATRA project (NE/K00896X/1). Robert J. H. Dunn and Elizabeth Kendon were supported by the Joint UK BEIS/Defra Met Ofﬁce Hadley Centre Climate Programme (GA01101). The southern UK high-resolution model simulations were funded by the CONVEX project (NERC Changing Water Cycle: NE/I006680/1) and the northern UK simulations were funded by the NUT- CAT2050 project (NERC Changing Water Cycle: NE/K002392/1, NE/K002430/1, NE/K002406/1). The UK 1 km dataset was part funded by the NERC SINATRA project (NE/K00896X/1). S. Blenkinsop et al.: The INTENSE project and van Meijgaard, E.: Increase in hourly precipi- tation extremes beyond expectations from temperature changes, Nat. Geosci., 1, 511–514, 2008. Lenderink, G., Barbero, R., Loriaux, J. M., and Fowler, H. J.: Super- Clausius–Clapeyron scaling of extreme hourly convective pre- cipitation and its relation to large-scale atmospheric conditions, J. Climate, 30, 6037–6052, https://doi.org/10.1175/JCLI-D-16- 0808.1, 2017. Forestieri, A., Lo Conti, F., Blenkinsop, S., Cannarozzo, M., Fowler, H. J., and Noto, L.: Regional frequency analysis of extreme precipitation for Sicily (Italy), Int. J. Climatol., 38, e698–e716, https://doi.org/10.1002/joc.5400, 2018a. Lewis, E., Fowler, H. J., Alexander, L., Dunn, R., McClean, F., Bar- bero, R., Guerreiro, S. B., Li, X.-F., and Blenkinsop, S.: GSDR: A Global sub-daily rainfall dataset, J. Climate, submitted, 2018a. Forestieri, A., Arnone, E., Blenkinsop, S., Candela, A., Fowler, H., and Noto, L. V.: The impact of climate change on ex- treme precipitation in Sicily, Italy, Hydrol. Process., 32, 332– 348, https://doi.org/10.1002/hyp.11421, 2018b. Lewis, E., Quinn, N., Blenkinsop, S., Fowler, H. J., Freer, J., Tan- guy, M., Hitt, O., Coxon, G., Bates, P., and Woods, R.: A rule based quality control method for hourly rainfall data and a 1 km resolution gridded hourly rainfall dataset for Great Britain: CEH- GEAR1hr, J. Hydrol., in revision, 2018b. Gadian, A., Blyth, A., Bruyere, C. L., Burton, R., Done, J., Groves, J., Holland, G., Mobbs, S., Pozo, J. T., Tye, M., and Warner, J.: A case study of possible future summer convective precipitation over the UK and Europe from a regional climate projection, Int. J. Climatol, 38, 2314–2324, https://doi.org/10.1002/joc.5336, 2018. Lind, P., Lindstedt, D., Kjellström, E., and Jones, C.: Spatial and Temporal Characteristics of Summer Precipitation over Cen- tral Europe in a Suite of High- Resolution Climate Models, J. Climate, 29, 3501–3518, https://doi.org/10.1175/JCLI-D-15- 0463.1, 2016. Hegerl, G. C., Black, E., Allan, R. P., Ingram, W. J., Polson, D., Trenberth, K. E., Chadwick, R. S., Arkin, P. A., Sarojini, B. B., Becker, A., Dai, A., Durack, P. J., Easterling, D., Fowler, H. J., Kendon, E. J., Huffman, G. J., Liu, C., Marsh, R., New, M., Os- born, T. J., Skliris, N., Stott, P. A., Vidale, P. L., Wijffels, S. E., Wilcox, L. J., Willett, K. M., and Zhang, X.: Challenges in Quantifying Changes in the Global Water Cycle, B. Am. Mete- orol. Soc., 96, 1097–1115, https://doi.org/10.1175/BAMS-D-13- 00212.1, 2015. Lochbihler, K., Lenderink, G., and Siebesma, A. S. Blenkinsop et al.: The INTENSE project S. Blenkinsop et al.: The INTENSE project S. Blenkinsop et al.: The INTENSE project 125 Donat, M. G., Alexander, L. V., Yang, H., Durre, I., Vose, R., Dunn, R. J. H., Willett, K. M., Aguilar, E., Brunet, M., Caesar, J., Hewitson, B., Jack, C., Klein Tank, A. M. G., Kruger, A. C., Marengo, J., Peterson, T. C., Renom, M., Oria Rojas, C., Rustic- ucci, M., Salinger, J., Elrayah, A. S., Sekele, S. S., Srivastava, A. K., Trewin, B., Villarroel, C., Vincent, L. A., Zhai, P., Zhang, X., and Kitching, S.: Updated analyses of temperature and precipita- tion extreme indices since the beginning of the twentieth century: The HadEX2 dataset, J. Geophys. Res.-Atmos., 118, 2098–2118, https://doi.org/10.1002/jgrd.50150, 2013. Kendon, E. J., Ban, N., Roberts, N. M., Fowler, H. J., Roberts, M. J., Chan, S. C., Evans, J. P., Fosser, G., and Wilkinson, J. M.: Do convection-permitting regional climate models improve projec- tions of future precipitation change?, B. Am. Meterol. Soc., 98, 79–93, https://doi.org/10.1175/BAMS-D-15-0004.1, 2017. Kendon, E. J., Blenkinsop, S., and Fowler, H. J.: When Will We Detect Changes in Short-Duration Precipitation Extremes?, J. Climate, 31, 2945–2964, https://doi.org/10.1175/JCLI-D-17- 0435.1, 2018. Lavers, D. A. and Villarini, G.: The nexus between atmospheric rivers and extreme precipitation across Europe, Geophys. Res. Lett., 40, 3259–3264, https://doi.org/10.1002/grl.50636, 2013a. Dunn, R. J. H., Willett, K. M., Parker, D. E., and Mitchell, L.: Expanding HadISD: quality-controlled, sub-daily station data from 1931, Geosci. Instrum. Method. Data Syst., 5, 473–491, https://doi.org/10.5194/gi-5-473-2016, 2016. Lavers, D. A. and Villarini, G.: Atmospheric Rivers and Flood- ing over the Central United States, J. Climate, 6, 829–7836, https://doi.org/10.1175/JCLI-D-13-00212.1, 2013b. Flato, G., Marotzke, J., Abiodun, B., Braconnot, P., Chou, S.C., Collins, W., Cox, P., Driouech, F., Emori, S., Eyring, V., Forest, C., Gleckler, P., Guilyardi, E., Jakob, C., Kattsov, V., Reason, C., and Rummukainen, M.: Evaluation of Climate Models, in: Cli- mate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergov- ernmental Panel on Climate Change, edited by: Stocker, T. F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S. K., Boschung, J., Nauels, A., Xia, Y., Bex, V., and Midgley, P. M., Cambridge Uni- versity Press, Cambridge, UK and New York, NY, USA, 2013. Lenderink, G. and Fowler, H. J.: Understanding pre- cipitation extremes, Nat. Clim. Change, 7, 391–393, https://doi.org/10.1038/nclimate3305, 2017. Lenderink, G. References J., Roberts, N., Blenkinsop, S., and Fowler, H. J.: Large-scale predictors for extreme hourly pre- cipitation events in convection-permitting climate simulations, J. Climate, 31, 2115–2131, https://doi.org/10.1175/JCLI-D-17- 0404.1, 2018b. Ban, N., Schmidli, J., and Schar, C.: Heavy precipitation in a changing climate: Does short-term summer precipita- tion increase faster?, Geophys. Res. Lett., 42, 1165–1172, https://doi.org/10.1002/2014GL062588, 2015. Bao, J., Sherwood, S. C., Alexander, L. V., and Evans, J. P.: Future increases in extreme precipitation exceed observed scaling rates, Nat. Clim. Change, 7, 128–132, https://doi.org/10.1038/nclimate3201, 2017. Dale, M., Luck, B., Fowler, H. J., Blenkinsop, S., Gill, E., Bennett, J., Kendon, E. J., and Chan, S.: New climate change rainfall esti- mates for sustainable drainage, Engineering Sustainability, 170, 214–224, https://doi.org/10.1680/jensu.15.00030, 2017. Barbero, R., Fowler, H. J., Lenderink, G., and Blenkinsop, S.: Is the intensiﬁcation of precipitation extremes with global warm- ing better detected at hourly than daily resolutions?, Geophys. Res. Lett., 44, 974–983, https://doi.org/10.1002/2016GL071917, 2017. Darwish, M., Fowler, H. J., Blenkinsop, S., and Tye, M. R.: A regional frequency analysis of UK hourly and multi-hourly extreme precipitation, Int. J. Climatol., https://doi.org/10.1002/joc.5694, accepted, 2018. Adv. Sci. Res., 15, 117–126, 2018 www.adv-sci-res.net/15/117/2018/ S. Blenkinsop et al.: The INTENSE project S. Blenkinsop et al.: The INTENSE project 126 S. Blenkinsop et al.: The INTENSE project Pall, P., Allen, M. R., and Stone, D. A.: Testing the Clausius– Clapeyron constraint on changes in extreme precipitation under CO2 warming, Clim. Dynam., 28, 351–363, 2007. Sen Roy, S.: A spatial analysis of extreme hourly precip- itation patterns in India, Int. J. Climatol., 29, 345–355, https://doi.org/10.1002/joc.1763, 2009. Peterson, T. C.: Climate Change Indices, WMO Bulletin, 54, 83–86, 2005. Smith, A., Lott, N., and Vose, R.: The integrated surface database: recent developments and partnerships, B. Am. Meteorol. Soc., 92, 704–708, 2011. Pfahl, S., O’Gorman, P. A., and Fischer, E. M.: Understanding the regional pattern of projected future changes in extreme precipi- tation, Nat. Clim. Change, 7, 423–427, 2017. Trenberth, K. E., Dai, A., Rasmussen, R. M., and Parsons, D. B.: The changing character of precipitation, B. Am. Meteorol. Soc., 84, 1205–1218, 2003. Prein, A. F., Langhans, W., Fosser, G., Ferrone, A., Ban, N., Go- ergen, K., Keller, M., Tölle, M., Gutjahr, O., Feser, F., Brisson, E., Kollet, S., Schmidli, J., van Lipzig, N. P. M., and Leung, R.: A review on regional convection-permitting climate modeling: Demonstrations, prospects, and challenges, Rev. Geophys., 53, 323–361, 2015. UKCP: Technical approach to the delivery of new UK Climate Projections (UKCP18 project), UK Climate Projections, avail- able at: http://ukclimateprojections.metofﬁce.gov.uk/media.jsp? mediaid=88729&ﬁletype=pdf (last access: 23 January 2018), 2016. Prein, A. F., Rasmussen, R., and Stephens, G.: Challenges and ad- vances in convection-permitting climate modeling, B. Am. Me- teorol. Soc., 98, 1027–1030, 2017a. Westra, S. and Sisson, S. A.: Detection of non-stationarity in pre- cipitation extremes using a max-stable process model, J. Hydrol., 406, 119–128, https://doi.org/10.1016/j.jhydrol.2011.06.014, 2011. Prein, A. F., Liu, C., Ikeda, K., Bullock, R., Rasmussen, R. M., Hol- land, G. J., and Clark, M.: Simulating North American mesoscale convective systems with a convection-permitting climate model, Clim. Dynam., https://doi.org/10.1007/s00382-017-3993-2, on- line ﬁrst, 2017b. Westra, S., Fowler, H. J., Evans, J. P., Alexander, L. V., Berg, P., Johnson, F., Kendon, E. J., Lenderink, G., and Roberts, N. M.: Future changes to the intensity and frequency of short-duration extreme rainfall, Rev. Geophys., 52, 522–555, https://doi.org/10.1002/2014RG000464, 2014. Prein, A. F., Liu, C., Ikeda, K., Trier, S. B., Rasmussen, R. M., Hol- land, G. J., and Clark, M. P.: Increased rainfall volume from fu- ture convective storms in the US, Nat. Clim. Change, 7, 880–884, 2017c. Wilby, R. L., Clifford, N. www.adv-sci-res.net/15/117/2018/ S. Blenkinsop et al.: The INTENSE project P.: The spatial extent of rainfall events and its relation to pre- cipitation scaling, Geophys. Res. Lett., 44, 8629–8636, https://doi.org/10.1002/2017GL074857, 2017. Mishra, V., Dominguez, F., and Lettenmaier, D. P.: Ur- ban precipitation extremes: How reliable are regional climate models?, Geophys. Res. Lett., 39, L03407, https://doi.org/10.1029/2011GL050658, 2012. Mooney, P. A., Broderick, C., Bruyère, C. L., Mulligan, F. J., and Prein, A. F.: Clustering of Observed Diurnal Cycles of Precipita- tion over the United States for Evaluation of a WRF Multiphysics Regional Climate Ensemble, J. Climate, 30, 9267–9286, 2017. Kay, A. L., Rudd, A. C., Davies, H. N., Kendon, E. J., and Jones, R. G.: Use of very high resolution climate model data for hydrolog- ical modelling: baseline performance and future ﬂood changes, Climatic Change, 133, 193–208, https://doi.org/10.1007/s10584- 015-1455-6, 2015. Morbidelli, R., Saltalippi, C., Flammini, A., Cifrodelli, M., Pic- ciafuoco, T., Corradini, C., Casas-Castillo, C. M., Fowler, H. J., and Wilkinson, S. M.: Effect of temporal aggregation on the estimate of annual maximum rainfall depths for the design of hydraulic infrastructure systems, J. Hydrol., 554, 710–720, https://doi.org/10.1016/j.jhydrol.2017.09.050, 2017. Kendon, E. J., Roberts, N. M., Senior, C. A., and Roberts, M. J.: Re- alism of rainfall in a very high resolution regional climate model, J. Climate, 25, 5791–5806, 2012. Kendon, E. J., Roberts, N. M., Fowler, H. J., Roberts, M. J., Chan, S. C., and Senior, C. A.: Heavier summer downpours with climate change revealed by weather forecast resolution model, Nature Clim. Change, 4, 570–576, 2014. O’Gorman, P. A.: Precipitation extremes under climate change, Cur- rent Climate Change Reports, 1, 49–59, 2015. Adv. Sci. Res., 15, 117–126, 2018 www.adv-sci-res.net/15/117/2018/ Adv. Sci. Res., 15, 117–126, 2018 S. Blenkinsop et al.: The INTENSE project J., De Luca, P., Harrigan, S., Hillier, J. K., Hodgkins, R., Johnson, M. F., Matthews, T. K. R., Mur- phy, C., Noone, S. J., Parry, S., Prudhomme, C., Rice, S. P., Slater, L. J., Smith, K. A., and Wood, P. J.: The “dirty dozen” of freshwater science: detecting then reconciling hy- drological data biases and errors, WIREs Water, 4, e1209, https://doi.org/10.1002/wat2.1209, 2017. Prein, A. F., Rasmussen, R. M., Ikeda, K., Liu, C., Clark, M. P., and Holland, G. J.: The future intensiﬁcation of hourly precipitation extremes, Nat. Clim. Change, 7, 48–52, 2017d. Rasmussen, R., Ikeda, K., Liu, C., Gochis, D., Clark, M., Dai, A., Gutmann, E., Dudhia, J., Chen, F., Barlage, M., Yates, D., and Zhang, G.: Climate change impacts on the water balance of the Colorado headwaters: High-resolution regional climate model simulations, J. Hydrometeor., 15, 1091–1116, https://doi.org/10.1175/JHM-D-13-0118.1, 2014. Adv. Sci. Res., 15, 117–126, 2018 Adv. Sci. Res., 15, 117–126, 2018 www.adv-sci-res.net/15/117/2018/
https://openalex.org/W2025192477	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0020914&type=printable	English	null	Energy Metabolism in Human Pluripotent Stem Cells and Their Differentiated Counterparts	PloS one	2,011	cc-by	13,349	Abstract This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits tion, and reproduction in any medium, provided the original author and source are credited. Funding: This research was supported by grants from the National Institute of Health and Development, 1P01HD047675(GS), 1R01ES019566 (BVH), Fundac¸a˜o para a Cieˆncia e Tecnologia Portugal (FCT), PTDC/EBB-EBI/101114/2008 (JRS) and PA CURE (BVH). ASR is a recipient of a fellowship from FCT SFRH/BD/33463/2008. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: schattengp@upmc.edu * E-mail: schattengp@upmc.edu . These authors contributed equally to this work. . These authors contributed equally to this work. levels, namely differentiation potential, global gene expression, and epigenetic status. The various analyses have revealed that despite the high level of resemblance in terms of pluripotency these cell types are not identical [4,5,6,7,8]. Namely, gene expression profiling revealed that, while IPSCs are highly similar to ESCs, a unique gene expression signature appears in IPSCs regardless of their origin [4]. Furthermore, DNA methylation analysis revealed that the methylome of certain developmental genes in IPSCs was intermediate between that found in differentiated cells and ESCs. In other cases, the methylation pattern in IPSCs was exclusive, differing from both ESCs and differentiated cells [6] , suggesting both the occurrence of abnormal methylation during reprogram- ming and that some ‘‘epigenetic memory’’ from the somatic cells Sandra Varum1,2., Ana S. Rodrigues1,2,3., Michelle B. Moura4, Olga Momcilovic1, Charles A. Easley, IV1, Joa˜o Ramalho-Santos2, Bennett Van Houten4, Gerald Schatten1,5* 1 Pittsburgh Development Center, Magee Womens Research Institute, Pittsburgh, Pennsylvania, United States of America, 2 Center for Neuroscience and Cell Biology and Department of Life Sciences, University of Coimbra, Coimbra, Portugal, 3 PhD Programme in Experimental Biology and Biomedicine (PDBEB), Center for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal, 4 Department of Pharmacology and Chemical Biology, University of Pittsburgh Cancer Institute, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, United States of America, 5 Department of Obstetrics, Gynecology & Reproductive Sciences, and Cell Biology- Physiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, United States of America Energy Metabolism in Human Pluripotent Stem Cells and Their Differentiated Counterparts Sandra Varum1,2., Ana S. Rodrigues1,2,3., Michelle B. Moura4, Olga Momcilovic1, Charles A. Easley, IV1, Joa˜o Ramalho-Santos2, Bennett Van Houten4, Gerald Schatten1,5* June 2011 \| Volume 6 \| Issue 6 \| e20914 Abstract Background: Human pluripotent stem cells have the ability to generate all cell types present in the adult organism, therefore harboring great potential for the in vitro study of differentiation and for the development of cell-based therapies. Nonetheless their use may prove challenging as incomplete differentiation of these cells might lead to tumoregenicity. Interestingly, many cancer types have been reported to display metabolic modifications with features that might be similar to stem cells. Understanding the metabolic properties of human pluripotent stem cells when compared to their differentiated counterparts can thus be of crucial importance. Furthermore recent data has stressed distinct features of different human pluripotent cells lines, namely when comparing embryo-derived human embryonic stem cells (hESCs) and induced pluripotent stem cells (IPSCs) reprogrammed from somatic cells. Methodology/Principal Findings: We compared the energy metabolism of hESCs, IPSCs, and their somatic counterparts. Focusing on mitochondria, we tracked organelle localization and morphology. Furthermore we performed gene expression analysis of several pathways related to the glucose metabolism, including glycolysis, the pentose phosphate pathway and the tricarboxylic acid (TCA) cycle. In addition we determined oxygen consumption rates (OCR) using a metabolic extracellular flux analyzer, as well as total intracellular ATP levels by high performance liquid chromatography (HPLC). Finally we explored the expression of key proteins involved in the regulation of glucose metabolism. Conclusions/Findings: Our results demonstrate that, although the metabolic signature of IPSCs is not identical to that of hESCs, nonetheless they cluster with hESCs rather than with their somatic counterparts. ATP levels, lactate production and OCR revealed that human pluripotent cells rely mostly on glycolysis to meet their energy demands. Furthermore, our work points to some of the strategies which human pluripotent stem cells may use to maintain high glycolytic rates, such as high levels of hexokinase II and inactive pyruvate dehydrogenase (PDH). Citation: Varum S, Rodrigues AS, Moura MB, Momcilovic O, Easley CA IV, et al. (2011) Energy Metabolism in Human Pluripotent Stem Cells and Their Differentiated Counterparts. PLoS ONE 6(6): e20914. doi:10.1371/journal.pone.0020914 Editor: Marian Ludgate, Cardiff University, United Kingdom Received January 31, 2011; Accepted May 16, 2011; Published June 17, 2011 Received January 31, 2011; Accepted May 16, 2011; Publish Copyright: 2011 Varum et al. This is an open-access article distributed under the terms of the Creative Commons Attributi unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. um et al. Energy Metabolism in Pluripotent Cells reprogrammed may linger in IPSCs, with unknown consequences [9], although the extent of this phenomenon is questioned [10]. Recent data suggests that IPSCs may be more prone to genetic mutation and instability [11,12,13]. Recently, Zhao et al. [14] demonstrated immunogenicity differences between authentic ESCs and iPSCs. Given the potential of these cells, further studies are needed to scrutinize these differences and to understand their impact on differentiation potential and possible therapeutic applications. recently shown that some cancer lines cultured under hypoxia acquire an undifferentiated phenotype similar to that of ESCs [31]. The aerobic glycolysis feature of cancer cells may involve the regulation of PDH function, at least in some cases, and indeed modulation of PDH function has been therapeutically considered in cancer [32]. However these analogies must be considered with great care, as there are considerable metabolic variations between cancer cell lines [33,34]. Additionally, it would be of interest to determine if and how IPSCs acquire a more glycolytic metabolism upon reprogram- ming, or if they instead maintain the metabolic features of the original somatic cell type that was reprogrammed. In this study our aim was to answer these questions by further characterizing the energy metabolism of both hESCs and human IPSCs when compared with their differentiated somatic counterparts. Towards that goal we analyzed mitochondrial morphology, glucose-related gene expression, OCR, intracellular ATP levels, lactate produc- tion and protein levels of regulatory enzymes relevant in metabolism. The mammalian embryo resides in a hypoxic environment prior to implantation [12]. During the early stages of embryonic development there is a metabolic shift from oxidative phosphor- ylation (OXPHOS) to glycolysis, and oxidative metabolism is only fully reinstituted after implantation [15]. Studies in the mouse and hamster indicated that the number of mitochondria is higher in the trophectoderm (TE) than in the ICM. Furthermore, the ICM is characterized by spherical and depolarized mitochondria with low O2 consumption, whereas mitochondria of the TE are elongated and have both higher mitochondrial potential and higher O2 consumption [16,17,18]. In accordance to previous reports our results demonstrate that hESCs have mitochondria consistent with lower activity, at least when compared to differentiated cells. Furthermore we show that mitochondria in IPSCs are morphologically distinct from both hESC and differentiated somatic cell mitochondria. In addition IPSCs are not identical to hESCs in terms of glucose-related gene expression, although they cluster with hESCs rather than with their somatic counterparts. Energy Metabolism in Pluripotent Cells Moreover analysis of OCR, intracel- lular ATP and lactate levels confirm that human pluripotent stem cells rely mostly on glycolysis to meet their energy demands. Finally our study demonstrates that human pluripotent stem cells express high levels of hexokinase II and have an inactive PDH complex. In all parameters quantified IPSCs seem to be close to hESC, but at slightly lower levels of expression/activity. These results shed light on some of the mechanisms that human pluripotent cells use to maintain high levels of glycolysis under normoxia. Similarly to ICM cells, embryonic stem cells rely mostly on glycolysis for energy supply. Furthermore, the mitochondria in these cells are rather immature with perinuclear localization. As human pluripotent stem cells differentiate they acquire more mature mitochondria and undergo a metabolic switch from glycolysis to OXPHOS [16,19,20,21]. In agreement with these findings several authors have reported that hypoxia is beneficial for the maintenance of hESCs in a pluripotent state [22,23]. In addition low O2 tensions have been reported to increase the reprogramming efficiency of both mouse and human somatic cells [24]. It is known that glycolysis and OXPHOS function in a coordinated fashion, and that the former generates 18 times more ATP per molecule of oxidized glucose than the former [25]. The first glycolytic reaction is crucial as glucose is captured within the cell by conversion to glucose 6-phosphate. This is one of the rate limiting steps in glycolysis and is catalyzed by hexokinases. Cell types with high glycolytic rates have been reported to have high expression levels of hexokinases [26,27]. Glucose 6-phosphate can be further metabolized in glycolysis producing two molecules of pyruvate, NADH and ATP. In normoxic conditions pyruvate enters mitochondria and links glycolysis to aerobic respiration by entering the TCA cycle (which provides the mitochondrial electron transfer chain (ETC) with reducing agents for ATP synthesis) as acetyl-coenzyme A. However under low O2 tensions or in the presence of dysfunctional mitochondria pyruvate will be converted to lactate. PLoS ONE \| www.plosone.org Introduction Human embryonic stem cells (hESCs) are self-renewing and pluripotent cells derived from the inner cell mass (ICM) of a blastocyst prior to implantation. However, although they can be differentiated into any somatic cell lineage, they cannot be genetically matched to putative patients in possible cell replace- ment therapies. However, it has been recently shown that human somatic cells can be reprogrammed into hESC-like pluripotent stem cells, the so called induced pluripotent stem cells (IPSCs), by the ectopic expression of a combination of pluripotency factors and using a variety of approaches [1,2,3]. The differences and similarities of hESCs and IPSCs have been studied at several PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 1 Energy Metabolism in Pluripotent Cells Energy Metabolism in Pluripotent Cells The other lines were used in key experiments to consolidate the results trying to avoid any artifact that could be line specific or culture derived. cycles at 95uC for 15 seconds and 60uC for one minute. Each array contained wells that allowed us to control for genomic DNA contamination, and first cDNA synthesis efficacy. Twelve genes (ALDO B, FBP1, FBP2, G6PC, GCK, GYS2, HK3, PCK1, PDK4, PGK2, PKLR and PRPS1L1) were not considered in the analysis, as their Ct value was above 30. Gene expression differences were calculated using the -DDCt method. Gene expression was normalized to the housekeeping gene b-actin and fold change was calculated relative to the hESC line WA07. p values were determined using online software provided by SABiosciences. cycles at 95uC for 15 seconds and 60uC for one minute. Each array contained wells that allowed us to control for genomic DNA contamination, and first cDNA synthesis efficacy. Twelve genes (ALDO B, FBP1, FBP2, G6PC, GCK, GYS2, HK3, PCK1, PDK4, PGK2, PKLR and PRPS1L1) were not considered in the analysis, as their Ct value was above 30. Gene expression differences were calculated using the -DDCt method. Gene expression was normalized to the housekeeping gene b-actin and fold change was calculated relative to the hESC line WA07. p values were determined using online software provided by SABiosciences. Lactate Production Cells were cultured on matrigel or matrix free conditions for pluripotent cells or differentiated cells, respectively. Cells were rinsed with PBS and fixed with 2.5 % glutaraldehyde in PBS, pH 7.4, for 1 hour at room temperature. The fixative was removed and samples were washed 3 times with PBS at room temperature for 10 minutes each. Samples were post-fixed for 1 hour at 4uC in 1% osmium tetroxide with 1% potassium ferricyanide, followed by three washes with PBS (10 minutes each). Afterwards, samples were dehydrated using a graded ethanol series (30%, 50%, 70%, and 90%-10 minutes) with three changes in 100% ethanol (15 minutes each). Samples were then incubated three times with a 100% Epon solution (1 hour each). Finally, samples were incubated for 1 hour with Epon solution, maintained overnight at 37uC and afterwards at 60uC for 2 days. Sections were made using an ultra microtome and imaged in a 1011 CX Transmission electron microscope (JEOL, Tokyo, Japan). Lactate levels were determined using the Lactate assay kit (Biovision). In this assay lactate is enzymatically oxidized and the product of this reaction reacts with a lactate probe (provided) and emits fluorescence at Ex/Em = 535/587 nm. Both pluripotent and differentiated cells were plated and cultured for 48 hours. At this point, fresh media was added to the cells. Twenty-four hours later, 500 ml of medium was collected, and cell numbers were determined. Medium samples were processed following manufac- turer’s instructions. Fluorescence was determined using a micro- plate reader (Spectra Max M2; Molecular Devices, Sunnyvale, CA). Lactate concentrations were normalized to cell number. High Performance liquid Chromatography (HPLC) High Performance liquid Chromatography (HPLC) Pluripotent cells were cultured as mentioned above for seven days, whereas differentiated cells were cultured for two days. Cells were enzymatically dissociated with Accutase (Millipore) or TripLe Express (Sigma-Aldrich, St. Louis, MO) for pluripotent cells and differentiated cells, respectively. Adenosine triphosphate (ATP) was extracted with 0.6 M perchloric acid (Sigma) supplemented with 25 mM EDTA-Na (Sigma) followed by centrifugation. Supernatants were neutralized with 3M potassium hydroxide in 1.5M Tris. ATP separated by HPLC. The detection wavelength was 260 nm and the column was a Licrosphere 100 RP108 5 mm (Merck). The elution buffer for was 100 mM phosphate buffer (pH 6.5) supplemented with 1% methanol. Standard for ATP was purchased from Sigma- Aldrich. Transduction of pluripotent and differentiated cells using a baculovirus system y Mitochondria GFP labeling was performed using the organelle lights Mito reagent (CellLightTM Mitochondria-GFP by Molecular Probes/Invitrogen). This system is a baculovirus system that contains a leader sequence for the E1 alpha pyruvate dehydro- genase fused with GFP protein. Cells were plate on coverslips and cultured for three days or 24 hours, for pluripotent and differentiated cells, respectively. Cells were rinsed with PBS, and incubated in PBS containing 82% organelle lights reagent during 30 minutes at room temperature with gentle shaking. Organelle lights mixture was removed and cells were incubated with appropriated media containing 0.1% enhancer solution (provided) for two hours at 37uC. Enhancer solution was removed and replaced by cell culture media. Although the transfection efficiencies were low, 20 hours later, GFP protein could be observed in the mitochondria. Cells were then fixed with 4% paraformaldehyde (Sigma-Aldrich) and imaged using a TCS-SP@ laser scanning confocal microscope (Leica, Microsystems, Gmbh, Wetzlar, Germany). Cell culture In this work we used hESC lines WA07, WA09 and WA01 (WiCell Research Institute, Madison, WI), human IPSC lines HFF1 iPS, IMR-90 iPS (WiCell Research Institute) and AE iPS (derived in our laboratory). In terms of differentiated fibroblasts, both the IMR-90 human diploid fibroblast and the human foreskin fibroblast (HFF1) strains were obtained from American Type Culture Collection (ATTC, Manassas, VA); these lines were reprogrammed to form the IMR-90 iPS and HFF1 iPS lines noted above, respectively. Finally, differentiated H7TF fibroblasts were isolated from a WA07 ESC-derived teratoma, following injection of WA07 hESCs into immunocompromised mice. The pyruvate dehydrogenase (PDH) complex, localized in the mitochondrial matrix is the link between glycolysis and the TCA cycle. It is composed of several copies of three catalytic proteins (E1, E2 and E3) that catalyze the irreversible decarboxylation of pyruvate to acetyl-coenzyme A and NADH [28,29]. The E1-alpha subunit is considered the on/off switch of the PDH complex, as its phosphorylation by one of the four pyruvate dehydrogenase kinase isoforms (PDHK1-4) leads to PDH complex inactivation, whereas removal of one phosphate group by one of the two pyruvate dehydrogenase phosphatases (PDP1 and PDP2) activates the complex [28,29] HESC and IPSC lines were cultured in mTeSRTM (STEM- CELL Technologies, Vancouver, BC, Canada) on matrigel (BD Biosciences; San Jose, CA) coated dishes. Both H7TF and HFF1 were cultured in CF1 medium containing: 90% Dulbecco’s Modified Eagle’s medium; 10% fetal bovine serum (FBS); 1% MEM non-essential amino acids; 1% penicillin/streptomycin (Pen/Strep) and 1% 2 mM L-glutamine (all from Invitrogen, Carlsbad, CA). The IMR-90 line was cultured in 90% Eagle’s minimal essential medium (ATTC, Manassas, VA), 10% FBS and 1% Pen/Strep. All lines were maintained at 37uC, 21% O2 and 5% CO2. Throughout the work WA07, IMR-90 iPS, H7TF and IMR-90 lines were used in the majority of the experiments in order to compare cell lines that have the same genetic background. The exact mechanism by which hESCs maintain an anaerobic metabolism even in the presence of oxygen remains largely elusive. However, there is certain putative parallelism with what has been described for cancer cells, which also maintain glycolysis as a key metabolic pathway under normoxia, in detriment to OXPHOS, the so-called Warburg effect [30]. Furthermore it has been June 2011 \| Volume 6 \| Issue 6 \| e20914 2 Western blotting Cells were maintained on matrigel or matrix free conditions for pluripotent cells or fibroblasts, respectively. Pluripotent cells were dissociated using Accutase (Millipore), whereas for differentiated cells Triple E (Sigma-Aldrich) was used. Total cell extracts were lysed in RIPA buffer (Sigma) supplemented with 1 mM phenyl- methylsulphonyl fluoride (Sigma-Aldrich) PMSF and 2x Halt phosphatase inhibitor cocktail (Pierce, Rockford, IL). Mitochon- drial extracts were isolated using mitochondria isolation kit (Sigma-Aldrich), as previously described by Wieckowski and colleagues [36]. In brief, cells were washed once with PBS, and incubated on ice with the extraction buffer supplemented with cell lysis detergent (1:200) and PMSF (1:200) for 5 minutes. Samples were resuspended every minute. Afterwards, we added 2v/v of extraction buffer and centrifuged twice, at 4uC, for 10 minutes, at 2000 g. Pellets containing the cytosolic fraction were stored at 280uC. Supernatants were transferred to a new tube and centrifuged, at 4uC for 10 minutes, at 11000 g. Supernatants were discarded and the pellets containing the mitochondrial fraction, were washed with extraction buffer, and centrifuged at 4uC, for 20 minutes, at 11000 g. The debris fraction (supernatant) was stored, as well as, the mitochondrial fraction at 280uC. Protein quantification was carried out using the Bradford assay (Bio-Rad laboratories, Inc., CA) and 7 mg of protein were separated in 4%–15% SDS Page gels (Bio-Rad laboratories). RNA extraction, DNA clean up and Glucose RT2 profiler PCR array Oxygen consumption rates O2 consumption was determined using Seahorse XF24 extracellular Flux analyzer as previously described by Qian and Van Houten [37]. Briefly, cells were seeded in the 24-well XF24 cell culture plate in the respective culture media. For pluripotent cells, matrigel coated plates were used and cells were allowed to grow for three days, whereas for differentiated cells 300 000 cells were seeded and allowed to grow for 24 h. Thirty minutes prior to the run, culture media was replaced by unbuffered DMEM and plates were incubated at 37uC for 30 minutes for pH and temperature stabilization. Three mitochondrial inhibitors; oligo- mycin (1 mM), FCCP (300 mM) and rotenone (1 mM) were sequentially injected after measurements, 2, 4, and 6, respectively. After all the measurements were completed, cells were dissociated and counted. Immunocytochemistry Cells were cultured on coverslips for six days or two days for pluripotent and differentiated cells respectively. Cells were fixed in 2% formaldehyde, rinsed in PBS, permeabilized with PBS 0.1 % Triton and incubated with blocking solution (containing 0.3% BSA, 1x PBS and 5% of normal serum from the specie of the corresponding secondary antibody) for an hour at room temperature. The primary antibodies used were anti-hexokinase II (1:500, Cell Signaling and Technologies). Antibodies were incubated overnight at 4uC, followed by three washes of 15 minutes in PBS 0.15 Triton. Respective secondary antibodies were incubated for one hour at 37uC, followed by three washes in PBS 0.1% Triton. Coverslips were assembled in slides using the fluorescence mounting media Vectashield with DAPI (Vector Labs, Switzerland). Slides were imaged using a TCS-SP2 laser scanning confocal microscope (Leica). Metabolism-related gene expression in human pluripotent stem cells vs. differentiated cells To further investigate the metabolic pathways used by human pluripotent stem cells and differentiated somatic cells we used two hESC lines (WA07 and WA01), two IPSC lines (IMR-90 IPS and AE IPS), and two somatic lines (WA07 teratoma fibroblasts-H7TF, and IMR-90 fibroblasts) and performed the SABiosciences human glucose metabolism RT2 profiler PCR array. This array contains 84 key genes involved in the regulation of glucose and glycogen metabolism, along with five housekeeping genes. Overall we observed that pluripotent lines tend to cluster together rather than with their somatic counterparts (Fig. 2A, heatmap). Furthermore, we observed that there are no significant differences in gene expression between the two hESC lines (WA07 and WA01) used in this study (Table 1). Statistical analysis showed that when compared to WA07 cells IMR-90 iPSCs showed 25 genes with significantly different expression, whereas in AE iPSCs there were14 such genes (Table 1). For differentiated cells, IMR-90 cells showed 44 differentially expressed genes whereas for H7TF there were 18 (Table 1). We further focused our analysis on 60 genes known to be involved in; glycolysis, pentose phosphate pathway and TCA cycle (Fig. 2, Table 1). Due to their pleiotropic nature some of the genes were included in more than one group. In addition, for the analysis we considered genes that were differentially expressed at least two fold when compared with the WA07 line, and with p,0.05. Energy Metabolism in Pluripotent Cells Energy Metabolism in Pluripotent Cells the high nuclear-cytoplasmic ratio observed in the hESCs. Furthermore, the mitochondria in the WA07 showed a globular shape, whereas mitochondria in the H7TF line were mainly elongated and formed extensive reticular networks. Similarly, the mitochondria of the somatic lines HFF1 and IMR-90 were also distributed in the cytoplasm rather than cluster around the nucleus; their shape was mostly tubular and they formed extensive networks (Fig. 1A, bottom left and right). The primary antibodies used were: anti - Oxphos complex kit (1:500, Invitrogen); anti-PDH (1:500, Cell Signaling and Tech- nologies); anti- PDHK1(1:500, Cell Signaling and Technologies); anti- phospho PDH-E1a Ser 293 (1:250, EMD); anti-hexokinase II (1:500, Cell Signaling and Technologies), and anti-Nanog ( 1:500, Kamiya Biomedical Company, Seattle, WA), ECL Advanced Western Blot Detection kit (Amersham Biosciences, Piscataway, NJ) was used for detection. In order to better characterize mitochondrial morphological features in pluripotent versus differentiated somatic cells we performed transmission electron microscopy (TEM) for hESCs (WA07 line), IPSCs (IMR-90 and HFF1 IPSC lines) and fibroblasts cells (H7TF, IMR-90 lines) (Fig.1B). TEM analysis demonstrated that mitochondria in hESCs have few cristae and electron-lucid matrix (Fig. 1B, top left). These results are in accordance with previous reports [20,21]. TEM of cells obtained after differentiation of W07 hESCs (H7TF) revealed mitochondria with an elongated shape, and with a higher number of cristae, as well as a denser matrix (Fig.1B. bottom, left). Similar to H7TF, IMR-90 and lung fibroblasts (LF) showed mature mitochondria with numerous cristae and electron dense matrix (Fig.1B. bottom center and right, respectively). Interestingly, both IMR-90 and HFF1 IPSCs showed a mix of both elongated and globular mitochondria. Furthermore, the matrix of the mitochondria resembled that of differentiated cells (Fig.1B top, center and right, respectively). These results suggest that mitochondria morphology in human IPSCs is not identical to that found in hESCs, but IPSCs seem to have a mixed phenotype between that found in hESCs and differentiated cells. It would be of interest to perform histomorphometric analysis in order to fully validate this assumption. Statistical analysis Means and standard error of the mean (SEM) were calculated and statistically significant differences were determined by One- way ANOVA followed by Dunnett’s Multiple Comparison test. n refers to sample size. Statistical significance was determined at p,0.05. The p values for the Glucose array were calculated using SABiosciences online software. The test used was a two-tailed Student’s t-test, and statistical significance was determine at p,0.05. RNA extraction, DNA clean up and Glucose RT2 profiler PCR array Total RNA isolation using the TRIzol reagent (Invitrogen) and genomic DNA was eliminated with DNA-free kit (Ambion, Austin, TX) as previously described [35]. An additional step of DNA clean up was performed using the genomic elimination kit from SABiosciences (Frederick, MD). In brief, 1 mg of RNA was incubated with genomic DNA elimination mixture (containing 2 ml of 5X genomic DNA elimination buffer and nuclease free water) for 5 minutes at 42uC. The first cDNA strand was synthesized by incubating the product of the genomic DNA elimination reaction with 10 ml of the Reverse transcriptase (RT) mixture (5 x reverse transcriptase (RT) buffer, RT, primers and external control, and nuclease free water) for 15 minutes at 42uC. Reactions were immediately stopped by heating at 95uC for 5 minutes. RT products were further diluted with 91 ml of nuclease free water. cDNA samples were then loaded in the glucose gene expression panel (SABiosciences), run in a 7900HT Real Time system (Applied Biosystems, Foster City, CA) accordingly to the following thermal cycle conditions; 95uC for 10 minutes, and 40 PLoS ONE \| www.plosone.org PLoS ON June 2011 \| Volume 6 \| Issue 6 \| e20914 PLoS ONE \| www.plosone.org 3 Mitochondrial localization and morphology In order to determine mitochondria morphology and localiza- tion within the cell in pluripotent versus differentiated cells we transduced hESCs (WA07 line) and fibroblasts cells (H7TF, IMR- 90 and HFF1) with a baculovirus system containing a leader sequence for PDH E1 alpha subunit fused with the GFP protein. Mitochondria in the hESC line WA07 showed perinuclear localization (Fig. 1A, top left) whereas the mitochondria of their differentiated counterparts (H7TF) were distributed in the cytoplasm (Fig. 1A, top right). This might be a consequence of Regarding the glycolytic signature, six genes among all lines were either up regulated or down regulated when compared with the WA07 line (Fig. 2B and Table 1). Among these differentially expressed genes, ENO 1 and ENO 3 were downregulated (2.02 and 1.98 fold, respectively, p,0.05) in AE IPSC relatively to the WA07 line, whereas the expression of these did not significantly PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 June 2011 \| Volume 6 \| Issue 6 \| e20914 4 Energy Metabolism in Pluripotent Cells Figure 1. Mitochondrial morphology and localization in human pluripotent stem cells vs. differentiated cells. A) Transduct hESCs (WA07 line,) and differentiated cells (H7TF, HFF, and IMR-90 lines) with organelle lights MitoGFP was used investigate mitoch morphology and localization within the cell. Blue: DAPI; Green: pyruvate dehydrogenase GFP. Scale bar: 10 mm B) Transmission electron micro (TEM) was used to investigate mitochondria morphologic features in hESCs (WA07 line), IPSCs (HFF1 and IMR-90 IPSC lines), and fibroblast (H7TF, IMR-90 and HFF1). Scale bar: 500 nm. doi:10.1371/journal.pone.0020914.g001 Figure 1. Mitochondrial morphology and localization in human pluripotent stem cells vs. differentiated cells. A) Transduction of hESCs (WA07 line,) and differentiated cells (H7TF, HFF, and IMR-90 lines) with organelle lights MitoGFP was used investigate mitochondria morphology and localization within the cell. Blue: DAPI; Green: pyruvate dehydrogenase GFP. Scale bar: 10 mm B) Transmission electron microscopy (TEM) was used to investigate mitochondria morphologic features in hESCs (WA07 line), IPSCs (HFF1 and IMR-90 IPSC lines), and fibroblasts cells (H7TF, IMR-90 and HFF1). Scale bar: 500 nm. doi:10.1371/journal.pone.0020914.g001 Figure 1. Mitochondrial morphology and localization in human pluripotent stem cells vs. differentiated cells. A) Transduction of hESCs (WA07 line,) and differentiated cells (H7TF, HFF, and IMR-90 lines) with organelle lights MitoGFP was used investigate mitochondria morphology and localization within the cell. Blue: DAPI; Green: pyruvate dehydrogenase GFP. PLoS ONE \| www.plosone.org Mitochondrial localization and morphology An increase in gene expressio expression is represented by the green color. No differences in expression are depicted in black. Cl software. The various genes were grouped accordingly to their participation in different metabolic Figure 2. Metabolism-related gene express i G i i t d l Figure 2. Metabolism-related gene expression in human pluripotent stem cells vs. differentiated cells. A) Heat map of average gene expression. Gene expression is represented as log10 of Ct values. An increase in gene expression is depicted in red, whereas a decrease in gene expression is represented by the green color. No differences in expression are depicted in black. Clustering was performed using SABiosciences online software. The various genes were grouped accordingly to their participation in different metabolic pathways. B) Glycolysis-related genes with at least two fold difference when compared to the WA07 line. C) Pentose phosphate pathways-related genes with at least two fold differences when compared to the WA07 line. D) TCA cycle- related genes with at least two fold differences when compared to the WA07 line. Fold changes were calculated using the -DDCt method relative to the WA07 line. hESC lines are represented in blue, IPSC lines in red and somatic cell lines in green . The values represent means of three independent experiments. Statistical analysis was performed using Student’s t test (SABiosciences online software) and significance was determined at p,0.05. Statistically significant differences are represented in Table 1. doi:10.1371/journal.pone.0020914.g002 Figure 2. Metabolism-related gene expression in human pluripotent stem cells vs. differentiated cells. A) Heat map of average gene expression. Gene expression is represented as log10 of Ct values. An increase in gene expression is depicted in red, whereas a decrease in gene expression is represented by the green color. No differences in expression are depicted in black. Clustering was performed using SABiosciences online software. The various genes were grouped accordingly to their participation in different metabolic pathways. B) Glycolysis-related genes with at least two fold difference when compared to the WA07 line. C) Pentose phosphate pathways-related genes with at least two fold differences when compared to the WA07 line. D) TCA cycle- related genes with at least two fold differences when compared to the WA07 line. Fold changes were calculated using the -DDCt method relative to the WA07 line. hESC lines are represented in blue, IPSC lines in red and somatic cell lines in green . Mitochondrial localization and morphology Scale bar: 10 mm B) Transmission electron microscopy (TEM) was used to investigate mitochondria morphologic features in hESCs (WA07 line), IPSCs (HFF1 and IMR-90 IPSC lines), and fibroblasts cells (H7TF, IMR-90 and HFF1). Scale bar: 500 nm. doi:10.1371/journal.pone.0020914.g001 PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 5 Energy Metabolism in Pluripotent Cells expression in human pluripotent stem cells vs. differentiated cells. A) Heat map of averag ted as log10 of Ct values. An increase in gene expression is depicted in red, whereas a decrease i olor. No differences in expression are depicted in black. Clustering was performed using SABiosciences d accordingly to their participation in different metabolic pathways. B) Glycolysis-related genes with the WA07 line. C) Pentose phosphate pathways-related genes with at least two fold differences le- related genes with at least two fold differences when compared to the WA07 line. Fold change ve to the WA07 line. hESC lines are represented in blue, IPSC lines in red and somatic cell lines in gree ndent experiments. Statistical analysis was performed using Student’s t test (SABiosciences online so 0.05. Statistically significant differences are represented in Table 1. gure 2. Metabolism-related gene expression in human pluripotent stem cells vs. differentiated cells. A) Heat map of avera pression. Gene expression is represented as log10 of Ct values. An increase in gene expression is depicted in red, whereas a decrease pression is represented by the green color. No differences in expression are depicted in black. Clustering was performed using SABioscienc tware. The various genes were grouped accordingly to their participation in different metabolic pathways. B) Glycolysis-related genes wit o fold difference when compared to the WA07 line. C) Pentose phosphate pathways-related genes with at least two fold differenc mpared to the WA07 line. D) TCA cycle- related genes with at least two fold differences when compared to the WA07 line. Fold chan culated using the -DDCt method relative to the WA07 line. hESC lines are represented in blue, IPSC lines in red and somatic cell lines in gr ues represent means of three independent experiments. Statistical analysis was performed using Student’s t test (SABiosciences online d significance was determined at p,0.05. Statistically significant differences are represented in Table 1. :10.1371/journal.pone.0020914.g002 Figure 2. Metabolism-related gene expression in human pluripotent stem cells vs. di expression. Gene expression is represented as log10 of Ct values. PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 Mitochondrial localization and morphology WA01 AE iPS IMR-90 iPS H7TF IMR-90 Glycolytic signature in human pluripotent stem cells and differentiated cells GBE1 0.175241 0.182796 0.401636 0.059971 0.003255 GYS1 0.560387 0.305042 0.073519 0.361399 0.006032 UGP2 0.140358 0.583771 0.349007 0.009063 0.00436 AGL 0.716072 0.005425 0.192293 0.003653 0.000895 PGM1 0.118692 0.108458 0.557582 0.443997 0.255193 PGM2 0.314817 0.234121 0.248629 0.699586 0.002456 PGM3 0.219734 0.069024 0.263436 0.527696 0.591157 PYGL 0.12035 0.182944 0.63261 0.469592 0.834941 PYGM 0.105912 0.300795 0.072922 0.163273 0.161997 GSK3A 0.119705 0.144921 0.083618 0.408039 0.061973 GSK3B 0.434825 0.208689 0.086282 0.844473 0.094813 PHKA1 0.117699 0.046949 0.322249 0.283362 0.011132 PHKB 0.759246 0.227841 0.905398 0.490711 0.052339 PHKG1 0.857115 0.004105 0.098563 0.538787 0.712081 PHKG2 0.184464 0.330626 0.305181 0.085937 0.09916 Pentose phosphate signature in human pluripotent stem cells and differentiated cells G6PD 0.222918 0.12939 0.009571 0.000621 0.000546 H6PD 0.536777 0.772588 0.001435 0.022755 0.005125 PGLS 0.854679 0.021219 0.088223 0.276943 0.975412 PRPS1 0.245115 0.126676 0.019419 0.128065 0.018916 PRPS2 0.831567 0.056076 0.004843 0.465994 0.305017 RBKS 0.119577 0.899118 0.072344 0.913592 0.133126 RPE 0.121207 0.471102 0.264923 0.753951 0.037438 RPIA 0.300618 0.263414 0.010121 0.012123 0.000578 TALDO1 0.61781 0.043175 0.123504 0.065059 0.003374 TKT 0.152685 0.541674 0.001309 0.020472 0.010954 TCA signature in pluripotent stem cells and differentiated cells ACLY 0.107296 0.234368 0.061387 0.393416 0.246977 ACO1 0.399131 0.079385 0.17169 0.083256 0.001386 ACO2 0.211968 0.089029 0.006734 0.15926 0.006435 CS 0.598438 0.052026 0.45459 0.548219 0.003514 DLAT 0.23882 0.369543 0.531728 0.001807 0.000145 DLD 0.860033 0.027963 0.493081 0.671359 0.004943 DLST 0.628177 0.019154 0.018305 0.726826 0.036226 FH 0.189939 0.945714 0.123065 0.242483 0.105099 IDH1 0.112953 0.38673 0.274658 0.0133 0.009346 IDH2 0.109272 0.7786 0.031926 0.983299 0.759598 IDH3A 0.291665 0.144282 0.699497 0.35153 0.107709 IDH3B 0.126595 0.753791 0.063038 0.271417 0.002515 IDH3G 0.127383 0.888987 0.020491 0.160844 0.22335 MDH1 0.136175 0.888013 0.220693 0.24667 0.032701 MDH1B 0.162897 0.520214 0.113153 0.126206 0.019708 MDH2 0.741719 0.076133 0.030813 0.960944 0.012102 OGDH 0.20949 0.383583 0.033559 0.773533 0.003039 PC 0.10867 0.486329 0.070307 0.698611 0.144088 PCK1 0.606445 0.046384 0.011521 0.331536 0.339268 PCK2 0.10813 0.884791 0.082829 0.167213 0.250898 PDHA1 0.770854 0.041691 0.395379 0.045818 0.00137 PDHB 0.529193 0.012731 0.486834 0.018482 0.000693 PLoS ONE \| www.plosone.org 7 June 2011 \| Volume 6 \| Issue 6 \| e20914 7 Energy Metabolism in Pluripotent Cells WA01 AE iPS IMR-90 iPS H7TF IMR-90 SDHA 0.518039 0.053203 0.04001 0.445353 0.210469 SDHB 0.835404 0.151871 0.250403 0.557484 0.01685 SDHC 0.232507 0.009183 0.168632 0.033296 0.002187 SDHD 0.415562 0.059517 0.463437 0.475622 0.011049 SUCLA2 0.340556 0.193395 0.129939 0.172629 0.014183 SUCLG1 0.328695 0.025023 0.153357 0.198642 0.045376 SUCLG2 0.106127 0.465278 0.01718 0.621296 0.237038 Glycolysis; Pentose phosphate and TCA-related genes p values. Mitochondrial localization and morphology p values of fold change were determined relative to the WA07 line. Statistical significance was determined at p,0.05 by Student’s t test (SABiosciences online software). p values represented in red correspond to genes that are significantly up regulated, whereas p values depicted in green represent genes that are significantly down regulated. p values above 0.05 are represented in black. doi:10.1371/journal.pone.0020914.t001 Table 1. cont. Table 1. cont. Glycolysis; Pentose phosphate and TCA-related genes p values. p values of fold change were determined relative to the WA07 line. Statistical significance was determined at p,0.05 by Student’s t test (SABiosciences online software). p values represented in red correspond to genes that are significantly up regulated, whereas p values depicted in green represent genes that are significantly down regulated. p values above 0.05 are represented in black. doi:10.1371/journal.pone.0020914.t001 vary for the IMR-90 IPSC line. Instead IMR-90 IPSC displayed up regulation of BPGM (3.82 fold, p,0.01), and HK2 (3.15 fold, p,0.05) when compared to the WA07 line (Fig. 2B, Table 1). Both H7TF and IMR-90 lines showed down regulation of ENO 3 (4.72 fold, p,0.05 and 4.73 fold, p,0.01 for H7TF and IMR-90, respectively) in comparison with the WA07 (Fig. 2B, Table 1). In addition, the IMR-90 line showed decreased expression of PGM2 (2.38 fold, p,0.01) and up regulation of ALDO A (2.64 fold, p,0.001). vary for the IMR-90 IPSC line. Instead IMR-90 IPSC displayed up regulation of BPGM (3.82 fold, p,0.01), and HK2 (3.15 fold, p,0.05) when compared to the WA07 line (Fig. 2B, Table 1). Both H7TF and IMR-90 lines showed down regulation of ENO 3 (4.72 fold, p,0.05 and 4.73 fold, p,0.01 for H7TF and IMR-90, respectively) in comparison with the WA07 (Fig. 2B, Table 1). In addition, the IMR-90 line showed decreased expression of PGM2 (2.38 fold, p,0.01) and up regulation of ALDO A (2.64 fold, p,0.001). 90 IPSC lines. Both differentiated lines showed down regulation of PDP2 (3.52 fold, p,0.05 and 4.55 fold, p,0.01 for H7TF and IMR-90, respectively). In addition the IMR-90 displayed de- creased expression of PDK3 (5.32 fold, p,0.01). Mitochondrial contribution to the energy metabolism of human pluripotent stem cells and differentiated cells Both the H7TF and IMR-90 lines showed down regulation of DLAT (2.65 fold, p,0.01 and 4.65, p,0.001, for H7TF and IMR-90, respectively) and IDH1 (6.34 fold, p,0.05 and 14.13 fold, p,0.01) for H7TF and IMR-90, respectively) when compared with the WA07 line (Fig. 2C, Table 1). In addition, the IMR-90 line demonstrated decreased expression of CS (2.21 fold, p,0.01), DLD (4.65 fold, p,0.001), DLST (2.22 fold, p,0.05), IDH3B (3.44 fold, p,0.01), MDH1 (3.43 fold, p,0.05), MDH1B (6.52 fold, p,0.05), PDHA1 (2.78 fold, p,0.01), PDHB (2.88, p,0.001), SDHC (1.96, p,0.01) and SUCLA2 (2.52, p,0.05), and up regulation of ACO 1 (2.13 fold, p,0.01). In terms of the pentose phosphate pathway seven genes were differentially expressed among all the lines when compared with the WA07 line. No differences were observed between the WA07 line and both the WA01 and AE IPSC lines. The IMR-90 IPSC line showed up regulation of H6PD (4.65 fold, p,0.01), PRPS1 (2.18 fold, p,0.05), PRPS2 (2.17 fold, p,0.01), TKT (2.67 fold, p,0.01) in comparison with the WA07 line. Both the H7TF and IMR-90 lines displayed increased expression of G6PD (1.96 fold, p,0.001 and 2.82 fold, p,0.001, for H7TF and IMR-90, respectively) H6PD (7.33, p,0.05 and 7.82, p,0.01, respectively), and decreased expression of TKT (2.13 fold, p,0.05 and 2.65 fold, p,0.05, respectively) in comparison to the WA07 line. In addition the IMR-90 line showed decreased expression of PRPS1 (3.96 fold, p,0.05), RPE (2.50 fold, p,0.05) and RPIA (2.52 fold, p,0.05). p ) ( p ) ( p ) Regarding the TCA cycle signature we observed that 17 genes were differentially expressed when compared to the WA07 line (Fig. 2D, Table 1). Again no significant differences were observed between the two hESC lines. When compared to the WA07 line, the AE IPSC line displayed decreased expression of DLD (1.98 fold, p,0.05) (Fig. 2C). The IMR-90 IPSC line showed increased expression of ACO2 (2.09 fold, p,0.01), and SUCLG2 (4.08 fold, p,0.05) when compared with the WA07 line (Fig. 2C). Both the H7TF and IMR-90 lines showed down regulation of DLAT (2.65 fold, p,0.01 and 4.65, p,0.001, for H7TF and IMR-90, respectively) and IDH1 (6.34 fold, p,0.05 and 14.13 fold, p,0.01) for H7TF and IMR-90, respectively) when compared with the WA07 line (Fig. 2C, Table 1). Mitochondrial localization and morphology The values represent means of three independent experiments. Statistical analysis was performed using Student’s t test (SABiosciences online software) and significance was determined at p,0.05. Statistically significant differences are represented in Table 1. doi:10.1371/journal.pone.0020914.g002 June 2011 \| Volume 6 \| Issue 6 \| e20914 PLoS ONE \| www.plosone.org 6 Energy Metabolism in Pluripotent Cells Table 1. p values of fold changes for gene expression. Mitochondrial contribution to the energy metabolism of human pluripotent stem cells and differentiated cells To further understand mitochondrial contribution to the energy metabolism of pluripotent cells we analyzed protein expression of mitochondrial complexes II, III and V in pluripotent vs. differentiated somatic cells. We observed no statistical significant differences in mitochondrial complexes II, III and V between the various pluripotent lines. Intriguingly, we observed that pluripo- tent cells have higher protein levels of mitochondrial complexes, II, III and V than differentiated cells (Fig. 3A, B). Indeed, when compared to the WA07 line, mitochondrial complex V protein levels were decreased by 45% and 46% in IMR-90 and HFF1 fibroblasts, respectively (p,0.01 Fig. 3A,B). Furthermore, complex III protein levels were decreased by 76.8%, 38% and 27% in IMR-90, HFF1 and H7TF fibroblasts, respectively (p,0.01, Fig. 3A,B). Although, we observed a decreased expression trend for mitochondrial complex II in differentiated cells, there was not statistical significance. p ) In terms of the pentose phosphate pathway seven genes were differentially expressed among all the lines when compared with the WA07 line. No differences were observed between the WA07 line and both the WA01 and AE IPSC lines. The IMR-90 IPSC line showed up regulation of H6PD (4.65 fold, p,0.01), PRPS1 (2.18 fold, p,0.05), PRPS2 (2.17 fold, p,0.01), TKT (2.67 fold, p,0.01) in comparison with the WA07 line. Both the H7TF and IMR-90 lines displayed increased expression of G6PD (1.96 fold, p,0.001 and 2.82 fold, p,0.001, for H7TF and IMR-90, respectively) H6PD (7.33, p,0.05 and 7.82, p,0.01, respectively), and decreased expression of TKT (2.13 fold, p,0.05 and 2.65 fold, p,0.05, respectively) in comparison to the WA07 line. In addition the IMR-90 line showed decreased expression of PRPS1 (3.96 fold, p,0.05), RPE (2.50 fold, p,0.05) and RPIA (2.52 fold, p,0.05). Regarding the TCA cycle signature we observed that 17 genes were differentially expressed when compared to the WA07 line (Fig. 2D, Table 1). Again no significant differences were observed between the two hESC lines. When compared to the WA07 line, the AE IPSC line displayed decreased expression of DLD (1.98 fold, p,0.05) (Fig. 2C). The IMR-90 IPSC line showed increased expression of ACO2 (2.09 fold, p,0.01), and SUCLG2 (4.08 fold, p,0.05) when compared with the WA07 line (Fig. 2C). Mitochondrial contribution to the energy metabolism of human pluripotent stem cells and differentiated cells In addition, the IMR-90 line demonstrated decreased expression of CS (2.21 fold, p,0.01), DLD (4.65 fold, p,0.001), DLST (2.22 fold, p,0.05), IDH3B (3.44 fold, p,0.01), MDH1 (3.43 fold, p,0.05), MDH1B (6.52 fold, p,0.05), PDHA1 (2.78 fold, p,0.01), PDHB (2.88, p,0.001), SDHC (1.96, p,0.01) and SUCLA2 (2.52, p,0.05), and up regulation of ACO 1 (2.13 fold, p,0.01). In order to determine if the differences observed in mitochon- drial complex expression could be translated in terms of higher mitochondrial activity in pluripotent cells we determine O2 consumption rate (OCR) using the Seahorse XF24 extracellular flux analyzer. We have used a pharmacological profiling approach, by combining the use of three mitochondrial inhibitors (rotenone, FCCP and oligomycin) and the Seahorse instrument as previously described by Qian and Van Houten [37]. The OCR response to the chemical compounds were analyzed in one hESC line (WA01), one IPSC line (AE) and two differentiated cell lines (H7TF and IMR-90). The basal OCR (measure of OXPHOS) is the amount oxygen consumption that is linked to ATP synthesis in the mitochondria, and represents the mean basal levels of oxygen consumption minus the mean of the two values following oligomycin treatment (measurements 3 and 4). Oligomycin inhibits ATP synthase by binding to the oligomycin sensitivity-conferring protein (OSCP) at the F0 subunit of the ATP synthase. This binding blocks the proton conductance resulting in loss of electron transfer and O2 consumption. Addition of oligomycin resulted in decreased levels of OCR in all cell types (Fig. 3C). Nonetheless the kinetics and relative intensity of response varied among the various cell types With respect to genes involved in the regulation of TCA cycle activity two genes were differentially expressed among all lines and the WA07 line (Table 1). No significantly expressed genes were detected between WA07 line, and the WA01, AE IPSC and IMR- June 2011 \| Volume 6 \| Issue 6 \| e20914 8 PLoS ONE \| www.plosone.org Energy Metabolism in Pluripotent Cells Figure 3. Mitochondrial contribution to the energetic metabolism of human pluripotent stem cells and differentiated cells. Western blotting analysis of mitochondrial complexes II, III and V. b-actin was used as a loading control. B) Quantification of mitochondrial complex II III and V protein levels relative to the WA07 line C) Oxygen consumption rate (OCR) was determined by Seahorse XF24 analyzer The th Figure 3. Mitochondrial contribution to the energy metabolism of human pluripotent stem cells and differentiated cells Mitochondrial contribution to the energetic metabolism of human pluripotent stem cells and differentiated cells Western blotting analysis of mitochondrial complexes II, III and V. b-actin was used as a loading control. B) Quantification of mitochondrial compl II, III and V protein levels relative to the WA07 line. C) Oxygen consumption rate (OCR) was determined by Seahorse XF24 analyzer. The t mitochondrial inhibitors were sequentially injected (after measurement points, 2, 4, and 6, as indicated) and the final concentrations of each w oligomycin (1 mM); FCCP (300mM); rotenone (1 mM) D) Basal OCR (represents the mean of the first two measurements minus the mea measurements 3 & 4). E) Reserve capacity (FCCP induced levels, measurements 5 & 6 minus the basal level). F) Intracellular ATP levels determine HPLC. G) Lactate levels secreted to the media. The values are averages of three independent experiments. Statistical significance was determine one-way ANOVA followed by Dunnet’s multiple comparison test. Statistical significance was determined at P,0.05. Error bars: SEM. doi:10.1371/journal.pone.0020914.g003 PLoS ONE \| www.plosone.org 9 June 2011 \| Volume 6 \| Issue 6 \| e20 contribution to the energetic metabolism of human pluripotent stem cells and differentiated cells. A h d l l d b d l d l f f h d l l 3. Mitochondrial contribution to the energetic metabolism of human pluripotent stem cells and diffe Figure 3. Mitochondrial contribution to the energetic metabolism of human pluripotent stem cells and differentiated cells. A) Western blotting analysis of mitochondrial complexes II, III and V. b-actin was used as a loading control. B) Quantification of mitochondrial complexes II, III and V protein levels relative to the WA07 line. C) Oxygen consumption rate (OCR) was determined by Seahorse XF24 analyzer. The three mitochondrial inhibitors were sequentially injected (after measurement points, 2, 4, and 6, as indicated) and the final concentrations of each were: oligomycin (1 mM); FCCP (300mM); rotenone (1 mM) D) Basal OCR (represents the mean of the first two measurements minus the mean of measurements 3 & 4). E) Reserve capacity (FCCP induced levels, measurements 5 & 6 minus the basal level). F) Intracellular ATP levels determined by HPLC. G) Lactate levels secreted to the media. The values are averages of three independent experiments. Statistical significance was determined by one-way ANOVA followed by Dunnet’s multiple comparison test. Statistical significance was determined at P,0.05. Error bars: SEM. doi:10.1371/journal.pone.0020914.g003 Figure 3. Energy Metabolism in Pluripotent Cells Energy Metabolism in Pluripotent Cells with hESCs having the slowest and less pronounced response followed by IMR-90, AE IPSC and finally H7TF (Fig.3C). have an N-terminal hydrophobic domain that allows binding to the outer mitochondrial membrane. Pedersen and colleagues have shown that mitochondrial-bound hexokinase II is responsible for the high glycolytic profile of rat hepatoma cells [26,27]. In order to determine if hexokinase II could play a role in the maintenance of high glycolytic rates in pluripotent cells we performed western blotting analysis for hexokinase II in three hESC lines (WA01, WA07 and WA09), three IPSC lines (IMR-90 IPSC, AE IPS and HFF1 IPSC) three differentiated lines (H7TF, IMR-90 and HFF1). We observed that pluripotent cells overexpress hexokinase II when compared to differentiated cells (Fig 4A, B). We next isolated the cytoplasmic and mitochondrial fractions of hESCs (WA01 line), IPSCs (AE IPS and IMR-90 IPS) and fibroblasts cells (H7TF and IMR-90) and determined hexokinase protein levels in both fractions (Fig. 4C, D, respectively). We observed that hexokinase II in differentiated cells is preferentially localized in the cytoplasm, whereas pluripotent cells display high levels of hexokinase II in both cytoplasmic and mitochondrial fractions. Binding of hexokinase II to the outer mitochondrial membrane can enhance mitochondrial metabolism in two ways; first when bound to the outer mitochondrial membrane, hexokinase escapes the inhibitory effect of its product glucose 6-phosphate; second mitochondrial binding allows hexokinase II to gain access to ATP that permeates the voltage dependent anion channel [39,40]. The H7TF cells showed higher basal O2 consumption rates when compared to the other cell types (Fig. 4C and 4E). The lowest basal OCR was observed for hESCs (WA01) and the differentiated line IMR-90 (Fig. 4C and 4E). The AE IPSC line displayed intermediate basal OCR levels between those found for the H7TF and WA01 line (Fig. 4C and 4D). Likewise to what was observed in glucose-related gene expression there were marked differences between the IMR-90 and H7TF with respect to their basal OCR. Because H7TF cells are derived from teratomas one could question whether the OCR levels of these cells are representative of other differentiated lines. In order to address that question we performed OCR measurements for a normal diploid human skin fibroblast line (NDHF). We observed that OCR levels in these cells are similar to those observed in H7TF rather to those observed in the IMR-90 cells (Fig.S1). PDH complex regulation in human pluripotent stem cells and differentiated cells Key metabolic regulation in terms of glycolysis versus OXPHOS involves the PDH complex. We therefore analyzed PDHK1, phospho PDH and total PDH protein in pluripotent stem cells and differentiated cells. We observed that PDHK1 protein levels are up regulated in pluripotent cells when compared to differentiated cells (Fig.5 A, B). Statistical analysis revealed that there are no significant differences between the WA07 hESC line and the other pluripotent lines (Fig. 5B). Although, there was a decrease in PDHK1protein levels between the WA07 line and both H7TF and HFF1 fibroblast lines (49% and 34% decrease for H7TF and HFF1, respectively; p,0.05) no significant differences were observed for the IMR-90 line (Fig. 5B). Interestingly, we did not observe an increase in PDK1 gene expression in pluripotent vs. differentiated cells (Fig. S2 and Table S1). This result suggests that PDHK1 protein stabilization differs between pluripotent and differentiated cells. Consistent with higher PDHK1 protein levels, pluripotent cells also displayed higher phospho PDH levels when compared to differentiated cells (Fig.5 A and C). All three differentiated lines showed lower phospho PDH protein when compared with the hESC line WA07 ( 44%, 64% and 39% decrease for H7TF, IMR-90 and HFF1 fibroblast, respectively; p,0.01; Fig. 5C). Interestingly, both AE and IMR-90 IPSCs showed lower levels of phospho-PDH than the WA07 line (20% and 26% decrease for AE IPSCs and IMR-90 IPSCs, respectively; p,0.01; Fig. 5C). No significant differences were found for total PDH levels. These results suggest that the lower mitochondrial activity observed in pluripotent cells might, at least partially, be attributed to high levels of phospho PDH. Our results further suggest that besides PDHK1 other PDHKs are regulating the activity of the PDH complex in these cells, as no significant differences were observed for the PDHK levels in IMR-90 cells but these cells displayed higher levels of phospho PDH. Rotenone is a mitochondrial inhibitor that prevents electron transfer from complex I to ubiquinone by blocking the ubiquinone binding site [38]. Treatment with rotenone decreased OCR levels in all cells types although H7TF fibroblasts showed a more pronounced response to the treatment. Overall our results suggest that both hESCs and IPSCs have lower reliance in OXPHOS when compared to the H7TF cells. PDH complex regulation in human pluripotent stem cells and differentiated cells Interestingly IMR-90 fibroblasts showed low basal O2 consumption rates, but when challenged with FCCP had the ability to increase OXPHOS rates to a similar extent to that observed in other differentiated cell types (H7TF and NDHF). In contrast both hESCs and IPSCs showed a considerably reduced response to FCCP than differentiated cells (H7TF and NDHF). These results suggest that mitochondria in pluripotent cells, although functional, present a low activity rate, and that the differences in mitochondrial content noted previously do not directly account for differences in activity. Indeed we observed that mitochondria membrane potential is lower in hESCs when compared to their differentiated counterparts (data not shown). Total ATP levels are originated both from anaerobic and aerobic respiration. In order to determine intracellular ATP levels in human pluripotent stem cells and differentiated cells we performed HPLC (Fig. 3F). Our results showed that somatic cells have elevated intracellular ATP levels when compared to differentiated cells. In addition we measure lactate secretion to media (a good measure for glycolysis that does not add to the TCA cycle) in human pluripotent stem cells and differentiated cells (Fig.3G). As expected human pluripotent cells displayed higher lactate levels than differentiated cells. Overall these results suggest that human pluripotent stem cells rely mostly on glycolysis to meet their energy demands. Energy Metabolism in Pluripotent Cells FCCP is a mitochondrial uncoupler that dissipates the proton gradient, and therefore uncouples electron transport and mito- chondrial respiration from ATP synthesis. FCCP treatment has been shown to increase O2 consumption to its maximum in various cell types. In fact FCCP resulted in an increase of OCR levels in all cell types, with the most pronounced response being observed in H7TF, followed by IMR-90, AE IPS and WA01 cells (Fig. 3C). The difference between FCCP OCR and basal OCR is a good measure of respiratory capacity (also designed of reserve capacity) of these cells. Indeed we observed that the respiratory capacity of differentiated cells is higher than that observed for pluripotent lines (Fig.3E). Mitochondrial contribution to the energy metabolism of human pluripotent stem cells and differentiated cells Mitochondrial contribution to the energetic metabolism of human pluripotent stem cells and differentiated cells. A) Western blotting analysis of mitochondrial complexes II, III and V. b-actin was used as a loading control. B) Quantification of mitochondrial complexes II, III and V protein levels relative to the WA07 line. C) Oxygen consumption rate (OCR) was determined by Seahorse XF24 analyzer. The three mitochondrial inhibitors were sequentially injected (after measurement points, 2, 4, and 6, as indicated) and the final concentrations of each were: oligomycin (1 mM); FCCP (300mM); rotenone (1 mM) D) Basal OCR (represents the mean of the first two measurements minus the mean of measurements 3 & 4). E) Reserve capacity (FCCP induced levels, measurements 5 & 6 minus the basal level). F) Intracellular ATP levels determined by HPLC. G) Lactate levels secreted to the media. The values are averages of three independent experiments. Statistical significance was determined by one-way ANOVA followed by Dunnet’s multiple comparison test. Statistical significance was determined at P,0.05. Error bars: SEM. doi:10.1371/journal.pone.0020914.g003 June 2011 \| Volume 6 \| Issue 6 \| e20914 9 PLoS ONE \| www.plosone.org Discussion In this study we sought to understand the energetic metabolism of pluripotent stem cells and their differentiated counterparts. We Mammalian tissues harbor four hexokinases isoforms (I–IV) with particular kinetics and tissue specificity. Both hexokinase I and II PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 10 Energy Metabolism in Pluripotent Cells Figure 4. Hexokinase II expression in human pluripotent stem cells and differentiated cells. A) Western blotting analysis for hexokinase II protein levels. B) Quantification of Hexokinase II protein levels relative to the WA07 line. C–E) Hexokinase II protein levels in the cytoplasmic and mitochondrial fractions, respectively. b-actin was used as a loading control. Values are means of three independent experiments. Statistical significance was determined at p,0.05 by one-way ANOVA followed by Dunnet’s multiple comparison test. Error bars: SEM doi:10.1371/journal.pone.0020914.g004 Figure 4. Hexokinase II expression in human pluripotent stem cells and differentiated cells. A) Western blotting analy protein levels. B) Quantification of Hexokinase II protein levels relative to the WA07 line. C–E) Hexokinase II protein levels in t mitochondrial fractions, respectively. b-actin was used as a loading control. Values are means of three independent exp significance was determined at p,0.05 by one-way ANOVA followed by Dunnet’s multiple comparison test. Error bars: SEM doi:10.1371/journal.pone.0020914.g004 Figure 4. Hexokinase II expression in human pluripotent stem cells and differentiated cells. A) Western blotting analysis for hexokinase II protein levels. B) Quantification of Hexokinase II protein levels relative to the WA07 line. C–E) Hexokinase II protein levels in the cytoplasmic and mitochondrial fractions, respectively. b-actin was used as a loading control. Values are means of three independent experiments. Statistical significance was determined at p,0.05 by one-way ANOVA followed by Dunnet’s multiple comparison test. Error bars: SEM doi:10.1371/journal.pone.0020914.g004 PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 11 Energy Metabolism in Pluripotent Cells Figure 5. Pyruvate dehydrogenase regulation in human pluripotent stem cells and differentiated cells. A) Western blotting analysis of PDHK1, pPDH and total PDH protein levels. b-actin was used as a loading control. B–D) Protein levels quantification relative to the Wa07 line of PDHK1, pPDH and PDH, respectively. Values are means of three independent experiments. Statistical significance was determined at p,0.05 by one- way ANOVA followed by Dunnet’s multiple comparison test. Error bars: SEM Abbreviations: PDHK1, pyruvate dehydrogenase kinase one; pPDH: phospho Ser 293 pyruvate dehydrogenase subunit E1a; PDH: pyruvate dehydrogenase. Discussion doi:10.1371/journal.pone.0020914.g005 Figure 5. Pyruvate dehydrogenase regulation in human pluripotent stem cells and differentiated cells. A) Western blotting analysis of PDHK1, pPDH and total PDH protein levels. b-actin was used as a loading control. B–D) Protein levels quantification relative to the Wa07 line of PDHK1, pPDH and PDH, respectively. Values are means of three independent experiments. Statistical significance was determined at p,0.05 by one- way ANOVA followed by Dunnet’s multiple comparison test. Error bars: SEM Abbreviations: PDHK1, pyruvate dehydrogenase kinase one; pPDH: phospho Ser 293 pyruvate dehydrogenase subunit E1a; PDH: pyruvate dehydrogenase. doi:10.1371/journal.pone.0020914.g005 started by looking at mitochondrial morphology and localization in human pluripotent stem cells (both hESCs and IPSCs), and somatic cells. In agreement with previous reports our results demonstrate that mitochondria in hESCs appear immature with fewer cristae, and low electron density matrix, when compared to differentiated somatic cells [16,20]. Interestingly IPSCs displayed a mixed mitochondrial phenotype that still resembles the somatic cells of origin and is not quite that of ESCs.. Analyzing distinct mitochondrial features during the reprogramming process and how they compare with the acquisition and stabilization of pluripotency characteristics would therefore be of interest, and might provide novel insights into mitochondrial activity and metabolic shifts during reprogramming. generated using several somatic cells and various methodologies, demonstrating some cell-line differences in gene expression that are probably not due to a single factor. It would be interesting to determine if these differences are also reflected in terms of functional metabolic studies. Regarding the glycolytic signature of human pluripotent stem cells and somatic cells we observed that H7TF fibroblasts displayed similar glycolytic signature to their pluripotent counter- part the WA07 hESC line, with only four genes significantly downregulated. Interestingly the IMR-90 fibroblasts displayed 10 genes differentially expressed (six genes significantly downregulat- ed and four genes up regulated when compared to the WA07). Cell division requires not only ATP but biosynthetic precursors derived from glycolysis, the pentose phosphate pathway and the TCA cycle. For instance, glucose 6-phosphate can be diverted from the glycolytic pathway to the pentose phosphate pathway in order to generate ribose 5-phophate for de novo nucleotide synthesis. The pentose phosphate pathway can be divided in two major branches, an oxidative branch and a non-oxidative branch [41]. Our gene expression analysis demonstrated that both G6PD and H6PD are significantly up regulated in differentiated cells. PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 PLoS ONE \| www.plosone.org Discussion Importantly, human pluripotent stem cells did not display a strong response to FCCP treatment suggesting that aerobic respiration in these cells is somewhat impaired. Consistent with less active mitochondria, human pluripotent cells showed reduced ATP levels when compared with differentiated cells. In addition these cells secreted higher lactate levels indicative of higher glycolytic rates. Thus despite higher levels of genes encoding proteins from the TCA cycle and higher content in mitochondrial electron transport chain complexes, human pluripotent cells seem to have lower overall OXPHOS activity. It remains to be established why this is the case, but our data suggests some possibilities. On the other hand, the PDH complex is a crucial step in regulation of metabolism since it constitutes the link between anaerobic metabolism and the TCA cycle. Phosphorylation of the PDH E1a subunit leads to inactivation of the PDH complex and consequently results in lower levels of acetyl CoA to enter the TCA cycle. Tellingly, pluripotent lines have higher levels of phosphorylated PDH E1a. Phosphorylation of PDH complex can be carried out by four PDHKs and in this study we analyzed PDHK1 expression levels. Although we did not observe an increase in PDHK1 gene expression in pluripotent lines when compared to differentiated cells we did observe an increase in PDHK1 protein levels in pluripotent cells. These results suggest that PDHK1 protein stability differs between pluripotent and differenti- ated cells. Papandreou and colleagues have previously demonstrated that under hypoxic conditions hypoxia inducible factor (HIF-1) up regulation resulted in increased expression of PDHK1 leading to inactivation of the PDH complex. This in turn resulted in reduced substrate availability to enter the TCA cycle and led to decreased oxidative phosphorylation [55]. HIF-1a might therefore be considered a good target for future work, given that we observed that some of its targets are involved in the maintenance of this glycolytic profile. Overall our results demonstrate that human pluripotent cells have a greater reliance on glycolysis than differentiated cells. In addition our study suggests that this can be mediated by increasing hexokinase II levels and inactivation of the PDH complex. Discussion Furthermore, we observed that human plurip- otent cells have higher expression levels of genes related to the non-oxidative branch, such as TKT, and RPIA. These results suggest that human pluripotent cells preferentially use the non- oxidative branch of the pentose phosphate pathway in order to obtain ribose-5 phosphate which is important for nucleotide synthesis. A similar pattern has been previously described for cancer cells [43,44,45]. [ ] Interestingly, the TCA cycle signature revealed that most of these genes are downregulated in differentiated cells, when compared to pluripotent lines. Besides its role in oxidative catabolism of carbohydrates and fatty acids, the TCA cycle provides precursors for many biosynthetic pathways, including precursors for amino acid and nucleotide synthesis. It is possible that up regulation of TCA cycle-related genes in human pluripotent stem cells is related to the high proliferative rates and concomitant need for nucleotides rather than need for reducing agents for mitochondrial activity. This issue needs to be explored further. Regardless, higher levels of genes encoding TCA cycle proteins might be indicative of increased electron transport capacity and cellular respiration. Consistent with increased level of genes encoding proteins of the TCA cycle we observed that pluripotent lines have higher levels of mitochondrial electron transport complexes than differentiated cells. The transcription factor c-Myc is crucial for the maintenance of embryonic stem cell self-renewal and its ectopic expression allows the reprogramming of somatic cells to an embryonic stem cell like state [1,3]. In addition this transcription factor has been previously described to up regulate the expression of several glycolytic enzymes as well as to promote mitochondrial biogenesis [46,47,48,49]. Hence higher levels of complex II, III and V may be due to higher c-myc levels in pluripotent cells and may not be representative of high mitochondrial activity. In accordance with this hypothesis our OCR results showed that pluripotent lines consume lower levels of O2 than the differentiated line H7TF. However, IMR-90 fibroblasts displayed similar basal OCR levels than those found in pluripotent lines. These results could be related to the fetal origin of the IMR-90 fibroblasts, as the fetal lung is exposed to a rather hypoxic environment [50]. In agreement with this notion, when we induced mitochondrial respiration by the use of FCCP (reserve capacity) IMR-90 fibroblasts could increase OCR to similar levels to those found in H7TF. Discussion The enzymes encoded by these genes catalyze the first step in the oxidative branch of the pentose phosphate pathway that involves the conversion of glucose 6-phosphate to 6-phosphogluconate with Glucose-related gene expression analysis revealed that both hESC lines analyzed in this study have identical glucose-related metabolism signatures. In contrast IPSCs are not identical to hESCs, although they cluster together, rather than with their somatic counterparts. Interestingly both IPSC lines analyzed in this study showed different metabolic gene expression signatures, with IMR-90 IPSCs having higher transcripts levels. It should be noted that that the cocktail of transcription factors used for the reprogramming of both lines differ [1]. However recent papers [10] established a genome-wide map for DNA methylation and gene expression for a considerable number of IPSC lines June 2011 \| Volume 6 \| Issue 6 \| e20914 June 2011 \| Volume 6 \| Issue 6 \| e20914 12 Energy Metabolism in Pluripotent Cells in mouse embryonic lethality around E.D. 7.5 [51,52]. Although we did not observe significant differences at mRNA level, the total protein levels of hexokinase II were significantly elevated in human pluripotent stem cells. These results suggest that hexokinase II protein stabilization might be increased in human pluripotent lines. Higher levels of mitochondrial hexokinase II can be advantageous for glycolytic metabolism in two ways: first binding of hexokinase II the outer mitochondrial membrane allows this enzyme to escape inhibition by its product glucose-phosphate; and second it allows the enzyme to gain access to newly synthesized ATP required for the phosphorylation of glucose [39,40]. In addition, hexokinase II plays a key role in the prevention of cell death by binding to VDAC, therefore representing a link between glucose metabolism and apoptosis [53,54]. Hence it is possible that hexokinase II plays a role in preventing human pluripotent stem cell apoptosis as well. concomitant production of NADPH. Besides its role as a cofactor in many biosynthetic reactions, NADPH is essential to regenerate reduced glutathione from glutathione disulfide [42]. Hence G6PD and H6PD transcripts levels may reflect high requirement for NADPH, either as reducing agent for biosynthetic pathways or instead as a cofactor in the cell antioxidant defense. The latter hypothesis is supported by the fact that differentiated cells have a higher mitochondrial activity and therefore a higher potential for ROS production. PLoS ONE \| www.plosone.org References 1. Yu J, Vodyanik MA, Smuga-Otto K, Antosiewicz-Bourget J, Frane JL, et al. (2007) Induced pluripotent stem cell lines derived from human somatic cells. Science 318: 1917–1920. 20. Cho YM, Kwon S, Pak YK, Seol HW, Choi YM, et al. (2006) Dynamic changes in mitochondrial biogenesis and antioxidant enzymes during the spontaneous differentiation of human embryonic stem cells. Biochem Biophys Res Commun 348: 1472–1478. 2. Takahashi K, Yamanaka S (2006) Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126: 663–676. 21. Prigione A, Fauler B, Lurz R, Lehrach H, Adjaye J (2010) The senescence- related mitochondrial/oxidative stress pathway is repressed in human induced pluripotent stem cells. Stem Cells 28: 721–733. 3. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, et al. (2007) Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131: 861–872. p p 22. Ezashi T, Das P, Roberts RM (2005) Low O2 tensions and the prevention of differentiation of hES cells. Proc Natl Acad Sci U S A 102: 4783–4788. 23. Forristal CE, Wright KL, Hanley NA, Oreffo RO, Houghton FD (2010) Hypoxia inducible factors regulate pluripotency and proliferation in human embryonic stem cells cultured at reduced oxygen tensions. Reproduction 139: 85–97. 4. Chin MH, Mason MJ, Xie W, Volinia S, Singer M, et al. (2009) Induced pluripotent stem cells and embryonic stem cells are distinguished by gene expression signatures. Cell Stem Cell 5: 111–123. 5. Maherali N, Sridharan R, Xie W, Utikal J, Eminli S, et al. (2007) Directly reprogrammed fibroblasts show global epigenetic remodeling and widespread tissue contribution. Cell Stem Cell 1: 55–70. 24. Yoshida Y, Takahashi K, Okita K, Ichisaka T, Yamanaka S (2009) Hypoxia enhances the generation of induced pluripotent stem cells. Cell Stem Cell 5: 237–241. 6. Doi A, Park IH, Wen B, Murakami P, Aryee MJ, et al. (2009) Differential methylation of tissue- and cancer-specific CpG island shores distinguishes human induced pluripotent stem cells, embryonic stem cells and fibroblasts. Nat Genet 41: 1350–1353. 25. Semenza GL (2007) Life with oxygen. Science 318: 62–64. 26. Bustamante E, Pedersen PL (1977) High aerobic glycolysis of rat hepatoma cells in culture: role of mitochondrial hexokinase. Proc Natl Acad Sci U S A 74: 3735–3739. 7. Hawkins RD, Hon GC, Lee LK, Ngo Q, Lister R, et al. (2010) Distinct epigenomic landscapes of pluripotent and lineage-committed human cells. Acknowledgments We thank Phillip Graebing for HPLC analysis. Carrie Redinger, Stacie Oliver, David McFarland for help with cell culture. Diane Carlisle and Joa˜o Ferreira are thanked for critical discussion of results. Angela Palermo Lauff is thanked for her administrative assistance. Supporting Information Table S1 Average Ct values for gene expression. Differences were calculated using the -DDCt method and gene expression was normalized to the housekeeping gene b-actin. Figure S1 Comparison of OCR in two differentiated cell lines. A) Oxygen consumption rate (OCR) was determined by Seahorse XF24 analyzer for IMR-90 and NHDF lines, the former also listed in Figure 3. The mitochondrial inhibitors were sequentially injected at specific time points as illustrated in the figure. B) Basal OCR (represents the mean of the first three measurements minus the mean of measurements 4, 5 & 6). E) Reserve capacity (FCCP induced levels, measurements 7, 8 & 9 minus the basal level). Measurements show that NHDF cells show higher values than IMR-90 cells, similarly to what is described for the other differentiated cell line (H7TF) listed in Figure 3 (TIF) Average Ct values were determined using online software provided by SABiosciences. (PDF) References Cell Stem Cell 6: 479–491. 27. Bustamante E, Morris HP, Pedersen PL (1981) Energy metabolism of tumor cells. Requirement for a form of hexokinase with a propensity for mitochondrial binding. J Biol Chem 256: 8699–8704. 8. Deng J, Shoemaker R, Xie B, Gore A, LeProust EM, et al. (2009) Targeted bisulfite sequencing reveals changes in DNA methylation associated with nuclear reprogramming. Nat Biotechnol 27: 353–360. 28. Holness MJ, Sugden MC (2003) Regulation of pyruvate dehydrogenase complex activity by reversible phosphorylation. Biochem Soc Trans 31: 1143–1151. 29. Roche TE, Hiromasa Y (2007) Pyruvate dehydrogenase kinase regulatory mechanisms and inhibition in treating diabetes, heart ischemia, and cancer. Cell Mol Life Sci 64: 830–849. 9. Lister R, Pelizzola M, Kida YS, Hawkins RD, Nery JR, et al. (2011) Hotspots of aberrant epigenomic reprogramming in human induced pluripotent stem cells. Nature 471: 68–73. 30. Ferreira LM (2010) Cancer metabolism: the Warburg effect today. Exp Mol Pathol 89: 372–380. 10. Bock C, Kiskinis E, Verstappen G, Gu H, Boulting G, et al. (2011) Reference Maps of human ES and iPS cell variation enable high-throughput character- ization of pluripotent cell lines. Cell 144: 439–452. 31. Yeung TM, Gandhi SC, Bodmer WF (2011) Hypoxia and lineage specification of cell line-derived colorectal cancer stem cells. Proc Natl Acad Sci U S A 108: 4382–4387. 11. Hussein SM, Batada NN, Vuoristo S, Ching RW, Autio R, et al. (2011) Copy number variation and selection during reprogramming to pluripotency. Nature 471: 58–62. 32. Madhok BM, Yeluri S, Perry SL, Hughes TA, Jayne DG (2010) Dichloroacetate induces apoptosis and cell-cycle arrest in colorectal cancer cells. Br J Cancer 102: 1746–1752. 12. Fischer B, Bavister BD (1993) Oxygen tension in the oviduct and uterus of rhesus monkeys, hamsters and rabbits. J Reprod Fertil 99: 673–679. 33. Loiseau D, Morvan D, Chevrollier A, Demidem A, Douay O, et al. (2009) Mitochondrial bioenergetic background confers a survival advantage to HepG2 cells in response to chemotherapy. Mol Carcinog 48: 733–741. 13. Gore A, Li Z, Fung HL, Young JE, Agarwal S, et al. (2011) Somatic coding mutations in human induced pluripotent stem cells. Nature 471: 63–67. 14. Zhao T, Zhang ZN, Rong Z, Xu Y (2011) Immunogenicity of induced pluripotent stem cells. Nature. p py g 34. Jose C, Bellance N, Rossignol R (2010) Choosing between glycolysis and oxidative phosphorylation: A tumor’s dilemma? Biochim Biophys Acta. 15. Discussion Interestingly these metabolic strategies involving features of anaerobic metabolism under normoxia are also found in many types of tumor cells [34,56], and parallel assays in both pluripotent and tumor lines would be extremely interesting Importantly we demonstrate that despite the fact that both hESCs and IPSCs rely on glycolysis, these cell types are not identical in terms of glucose- related gene expression, mitochondrial morphology, and O2 consumption. This suggests that IPSC somatic cell reprogramming to IPSC may result in differences at the metabolic level, when compared to the pluripotent standard of hESC. While epigenetic and transcriptomic differences have been mentioned above [9,10] other significant genetic changes in IPSCs when compared to hESCs and differentiated cells were also recently described [11,12], including higher mutation rates and copy number variation. Interestingly, recent data suggests that IPSCmitochon- dria retain significant developmental plasticity upon IPSC generation, and somatic cell re-differentiation. [57]. In addition we also observe that not all the differentiated lines display the same metabolic profile and this might have an impact in the reprogramming efficiency of various somatic cell types, and on the characteristics of differentiated cells obtained from different IPSC lines. These results stress the importance of the complete characterization of these cell lines, and suggest that there may be other links to tumor cells besides those already described. Hexokinase II catalyzes the first reaction of glycolysis and it has been previously demonstrated that its expression is restricted to the inner cell mass of the blastocyst, and that HK2 knock down results PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 13 Energy Metabolism in Pluripotent Cells Energy Metabolism in Pluripotent Cells Figure S2 p-values for the glucose metabolism gene expression array in human pluripotent stem cells vs. differentiated cells. Statistical analysis was performed using SABi i li ft d St d t’ t t t li d Figure S2 p-values for the glucose metabolism gene expression array in human pluripotent stem cells vs. differentiated cells. Statistical analysis was performed using SABiosciences online software and Student’s t test was applied. Significance was determined at p,0.05. (PDF) Conceived and designed the experiments: SV ASR JR-S BVH GS. Performed the experiments: SV ASR MBM. Analyzed the data: SV ASR MBM JR-S BVH GS. Contributed reagents/materials/analysis tools: OM CAE. Wrote the paper: SV ASR JR-S BVH GS. MBM JR-S BVH GS. Contributed reagents/materials/analysis tools: OM CAE. Wrote the paper: SV ASR JR-S BVH GS. MBM JR-S BVH GS. Contributed reagents/materials/analysis tools: OM CAE. Wrote the paper: SV ASR JR-S BVH GS. Author Contributions Figure S2 p-values for the glucose metabolism gene expression array in human pluripotent stem cells vs. differentiated cells. Statistical analysis was performed using SABiosciences online software and Student’s t test was applied. Significance was determined at p,0.05. (PDF) Energy Metabolism in Pluripotent Cells of glucose by intramitochondrially generated ATP. J Biol Chem 263: 17422–17428. 48. Li F, Wang Y, Zeller KI, Potter JJ, Wonsey DR, et al. (2005) Myc stimulates nuclearly encoded mitochondrial genes and mitochondrial biogenesis. Mol Cell Biol 25: 6225–6234. 40. Mathupala SP, Ko YH, Pedersen PL (2009) Hexokinase-2 bound to mitochondria: cancer’s stygian link to the "Warburg Effect" and a pivotal target for effective therapy. Semin Cancer Biol 19: 17–24. 49. Wu KJ, Polack A, Dalla-Favera R (1999) Coordinated regulation of iron- controlling genes, H-ferritin and IRP2, by c-MYC. Science 283: 676–679. py 41. Wamelink MM, Struys EA, Huck JH, Roos B, van der Knaap MS, et al. (2005) Quantification of sugar phosphate intermediates of the pentose phosphate pathway by LC-MS/MS: application to two new inherited defects of metabolism. J Chromatogr B Analyt Technol Biomed Life Sci 823: 18–25. 50. Adams FH, Yanagisawa M, Kuzela D, Martinek H (1971) The disappearance fetal lung fluid following birth. J Pediatr 78: 837–843. 51. Cinelli P, Casanova EA, Uhlig S, Lochmatter P, Matsuda T, et al. (2008) Expression profiling in transgenic FVB/N embryonic stem cells overexpressing STAT3. BMC Dev Biol 8: 57. 42. Kletzien RF, Harris PK, Foellmi LA (1994) Glucose-6-phosphate dehydroge- nase: a "housekeeping" enzyme subject to tissue-specific regulation by hormones, nutrients, and oxidant stress. FASEB J 8: 174–181. 52. Heikkinen S, Pietila M, Halmekyto M, Suppola S, Pirinen E, et al. (1999) Hexokinase II-deficient mice. Prenatal death of homozygotes without distur- bances in glucose tolerance in heterozygotes. J Biol Chem 274: 22517–22523. 43. Tian WN, Braunstein LD, Pang J, Stuhlmeier KM, Xi QC, et al. (1998) Importance of glucose-6-phosphate dehydrogenase activity for cell growth. J Biol Chem 273: 10609–10617. g yg 53. Golshani-Hebroni SG, Bessman SP (1997) Hexokinase binding to mitochondria: a basis for proliferative energy metabolism. J Bioenerg Biomembr 29: 331–338. 53. Golshani-Hebroni SG, Bessman SP (1997) Hexokinase binding to mitochondria: b i f lif i b li J Bi Bi b 29 331 338 44. Boros LG, Torday JS, Lim S, Bassilian S, Cascante M, et al. (2000) Transforming growth factor beta2 promotes glucose carbon incorporation into nucleic acid ribose through the nonoxidative pentose cycle in lung epithelial carcinoma cells. Cancer Res 60: 1183–1185. 54. Majewski N, Nogueira V, Robey RB, Hay N (2004) Akt inhibits apoptosis downstream of BID cleavage via a glucose-dependent mechanism involving mitochondrial hexokinases. Mol Cell Biol 24: 730–740. 55. References Leese HJ, Barton AM (1984) Pyruvate and glucose uptake by mouse ova and preimplantation embryos. J Reprod Fertil 72: 9–13. 35. Varum S, Momcilovic O, Castro C, Ben-Yehudah A, Ramalho-Santos J, et al. (2009) Enhancement of human embryonic stem cell pluripotency through inhibition of the mitochondrial respiratory chain. Stem Cell Res 3: 142–156. 16. Ramalho-Santos J, Varum S, Amaral S, Mota PC, Sousa AP, et al. (2009) Mitochondrial functionality in reproduction: from gonads and gametes to embryos and embryonic stem cells. Hum Reprod Update 15: 553–572. 36. Wieckowski MR, Giorgi C, Lebiedzinska M, Duszynski J, Pinton P (2009) Isolation of mitochondria-associated membranes and mitochondria from animal tissues and cells. Nat Protoc 4: 1582–1590. 17. Stern S, Biggers JD, Anderson E (1971) Mitochondria and early development of the mouse. J Exp Zool 176: 179–191. 37. Qian W, Van Houten B (2010) Alterations in bioenergetics due to changes in mitochondrial DNA copy number. Methods 51: 452–457. 18. Houghton FD (2006) Energy metabolism of the inner cell mass and trophectoderm of the mouse blastocyst. Differentiation 74: 11–18. 38. Lambert AJ, Brand MD (2004) Inhibitors of the quinone-binding site allow rapid superoxide production from mitochondrial NADH:ubiquinone oxidoreductase (complex I). J Biol Chem 279: 39414–39420. 19. St John JC, Ramalho-Santos J, Gray HL, Petrosko P, Rawe VY, et al. (2005) The expression of mitochondrial DNA transcription factors during early cardiomyocyte in vitro differentiation from human embryonic stem cells. Cloning Stem Cells 7: 141–153. ( p ) 39. Arora KK, Pedersen PL (1988) Functional significance of mitochondrial bound hexokinase in tumor cell metabolism. Evidence for preferential phosphorylation PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 June 2011 \| Volume 6 \| Issue 6 \| e20914 14 Energy Metabolism in Pluripotent Cells Coussens M, Davy P, Brown L, Foster C, Andrews WH, et al. (2010) RNAi screen for telomerase reverse transcriptase transcriptional regulators identifies HIF1alpha as critical for telomerase function in murine embryonic stem cells. Proc Natl Acad Sci U S A 107: 13842–13847. 45. Cascante M, Centelles JJ, Veech RL, Lee WN, Boros LG (2000) Role of thiamin (vitamin B-1) and transketolase in tumor cell proliferation. Nutr Cancer 36: 150–154. p Proc Natl Acad Sci U S A 107: 13842–13847. 46. Kim JW, Zeller KI, Wang Y, Jegga AG, Aronow BJ, et al. (2004) Evaluation of myc E-box phylogenetic footprints in glycolytic genes by chromatin immuno- precipitation assays. Mol Cell Biol 24: 5923–5936. 56. Berridge MV, Herst PM, Tan AS (2010) Metabolic flexibility and cell hierarchy in metastatic cancer. Mitochondrion 10: 584–588. 57. Suhr ST, Chang EA, Tjong J, Alcasid N, Perkins GA, et al. (2010) Mitochondrial rejuvenation after induced pluripotency. PLoS One 5: e14095. p p y 47. Gordan JD, Thompson CB, Simon MC (2007) HIF and c-Myc: sibling rivals for control of cancer cell metabolism and proliferation. Cancer Cell 12: 108–113. PLoS ONE \| www.plosone.org June 2011 \| Volume 6 \| Issue 6 \| e20914 15
https://openalex.org/W1812064270	https://jtultrasound.biomedcentral.com/track/pdf/10.1186/2050-5736-3-S1-P51	English	null	Accounting for sliding motion in fast numerical simulations of abdominal HIFU applications for targets under respiratory motion	Journal of therapeutic ultrasound	2,015	cc-by	817	Background/introduction motion information. The mesh is split at the abdominal wall into individually moving parts to allow the sliding motion of the liver along the abdominal wall. The numerical simulation thereby resolves the sliding bound- ary and allows the prediction of the temperature rise also in the beam path between transducer and target. Fast US beam steering is simulated using ultrasound pressure field pre-computations for key states over the respiratory cycle. Any remaining location mismatch between closest pre-computation and actual target loca- tion is compensated using an additional transformation. The numerical simulation method is integrated into a HIFU-therapy planning tool. Based on image data the target volume, anatomical and risk structures are defined. Treatment planning is then performed guided by direct feedback from the numerical simulation including respiratory motion of the domain. Using the feedback from the simulation, the user can manually optimize transducer and target location and HIFU parameters. Non-invasive ablation and hyperthermia treatments of abdominal organs using High Intensity Focused Ultra- sound (HIFU) still impose severe challenges. Especially for liver and kidney treatments, the target motion and accessibility over the entire respiratory cycle make the application of HIFU difficult. To allow for a safe, effec- tive, and efficient treatment, detailed planning and monitoring of the intervention is needed. Current gen- eral-purpose physics simulation software cannot fulfill the real-time requirements of the therapy-monitoring application. Our goal is to improve this by numerical simulation to allow for fast and accurate treatment planning and real-time detection of motion-induced risks during HIFU-therapy. Results and conclusions The method accounts for patient-specific anatomy and motion information in real time. Sliding motion of the inner organs along the abdominal wall is accounted for by the simulation. The transformation to the static refer- ence frame results in an accurate scheme for long simula- tion durations. The developed method can be a possible important building block for HIFU-therapy planning and conduction in moving organs. Motion information is fed into the simulation by an unstructured moving point cloud. On top of this point cloud a tetrahedral mesh is build to interpolate the Methods We developed a method to numerically simulate HIFU using patient-specific motion information in real time. The bio-heat equation describing the temperature distri- bution in the patient is mathematically transformed to a static reference anatomy. The numerical solution can then be performed in the reference anatomy on a static computational domain. Patient-specific respiratory motion information is provided by an abdominal motion model that can be fed with sparse tracking data in the clinical setting. The respiratory motion is modelled allowing sliding motion of the inner organs along the abdominal wall. © 2015 Schwenke 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/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 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. Schwenke et al. Journal of Therapeutic Ultrasound 2015, 3(Suppl 1):P51 http://www.jtultrasound.com/content/3/S1/P51 Schwenke et al. Journal of Therapeutic Ultrasound 2015, 3(Suppl 1):P51 http://www.jtultrasound.com/content/3/S1/P51 Schwenke et al. Journal of Therapeutic Ultrasound 2015, 3(Suppl 1):P51 http://www.jtultrasound.com/content/3/S1/P51 Open Access 1Fraunhofer MEVIS, Bremen, Germany Full list of author information is available at the end of the article Acknowledgements (Funding) g g The research leading to these results has received funding from the European Community’s Seventh Framework Programme FP7/2007-2013 under grant agreement n° 611889. 1Fraunhofer MEVIS, Bremen, Germany Full list of author information is available at the end of the article Page 2 of 2 Schwenke et al. Journal of Therapeutic Ultrasound 2015, 3(Suppl 1):P51 http://www.jtultrasound.com/content/3/S1/P51 Authors’ details 1Fraunhofer MEVIS, Bremen, Germany. 2ETH Zurich, Zurich, Switzerland. 3InSightec Ltd, Tirat Carmel, Israel. 4Fraunhofer MEVIS/Jacobs University Bremen, Bremen, Germany. Published: 30 June 2015 doi:10.1186/2050-5736-3-S1-P51 Cite this article as: Schwenke et al.: Accounting for sliding motion in fast numerical simulations of abdominal HIFU applications for targets under respiratory motion. Journal of Therapeutic Ultrasound 2015 3(Suppl 1):P51. Figure 1 Figure 1 Figure 1 Authors’ details 1 Authors’ details 1Fraunhofer MEVIS, Bremen, Germany. 2ETH Zurich, Zurich, Switzerland. 3InSightec Ltd, Tirat Carmel, Israel. 4Fraunhofer MEVIS/Jacobs University Bremen, Bremen, Germany. Published: 30 June 2015 doi:10.1186/2050-5736-3-S1-P51 Cite this article as: Schwenke et al.: Accounting for sliding motion in fast numerical simulations of abdominal HIFU applications for targets under respiratory motion. Journal of Therapeutic Ultrasound 2015 3(Suppl 1):P51. doi:10.1186/2050-5736-3-S1-P51 Cite this article as: Schwenke et al.: Accounting for sliding motion in fast numerical simulations of abdominal HIFU applications for targets under respiratory motion. Journal of Therapeutic Ultrasound 2015 3(Suppl 1):P51. 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/W2052438828	https://europepmc.org/articles/pmc3774640?pdf=render	English	null	Effects of Constitutive β-Catenin Activation on Vertebral Bone Growth and Remodeling at Different Postnatal Stages in Mice	PloS one	2,013	cc-by	8,107	Min Jia1☯, Sixu Chen1☯, Bo Zhang1, Huaping Liang1, Jianquan Feng2, Zhaowen Zong1,2 Min Jia1☯, Sixu Chen1☯, Bo Zhang1, Huaping Liang1, Jianquan Feng2, Zhaowen Zong1 1 Department of Trauma Surgery, State Key Laboratory of Trauma, Burn and Combined injury, Daping Hospital, Third Military Medical University, ChongQing, China, 2 Department of Biomedical Sciences, Baylor College of Dentistry, Texas Agriculture and Mechanic Health Science Center, Dallas, Texas, United States of America Abstract Funding: This study was supported by Foundation of State Key Laboratory of Trauma, Burns and combined injury (SKLZZ201124), National Science Foundation of China (81271935), and Special Fund for National Key Project “973” for Development of Basic Research (2011CB964701). The funders’ websites are http://202.202.224.96/; http://www.nsfc.gov.cn; and http://www.nsfc.gov.cn, respectively. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Abstract Funding: This study was supported by Foundation of State Key Laboratory of Trauma, Burns and combined injury (SKLZZ201124), National Science Foundation of China (81271935), and Special Fund for National Key Project “973” for Development of Basic Research (2011CB964701). The funders’ websites are http://202.202.224.96/; http://www.nsfc.gov.cn; and http://www.nsfc.gov.cn, respectively. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. E-mail: zongzhaowen@163.com (ZZ); JFeng@bcd.tamhsc.edu (JQF) ☯ These authors contributed equally to this work. Citation: Jia M, Chen S, Zhang B, Liang H, Feng J, et al. (2013) Effects of Constitutive β-Catenin Activation on Vertebral Bone Growth and Remodeling at Different Postnatal Stages in Mice. PLoS ONE 8(9): e74093. doi:10.1371/journal.pone.0074093 Editor: Dominique Heymann, Faculté de médecine de Nantes, France Received May 7, 2013; Accepted July 26, 2013; Published September 16, 2013 Copyright: © 2013 Jia 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. Funding: This study was supported by Foundation of State Key Laboratory of Trauma, Burns and combined injury (SKLZZ201124), National Science Foundation of China (81271935), and Special Fund for National Key Project “973” for Development of Basic Research (2011CB964701). The funders’ websites are http://202.202.224.96/; http://www.nsfc.gov.cn; and http://www.nsfc.gov.cn, respectively. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. * E-mail: zongzhaowen@163.com (ZZ); JFeng@bcd.tamhsc.edu (JQF) ☯ These authors contributed equally to this work. Received May 7, 2013; Accepted July 26, 2013; Published September 16, 2013 Jia et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits tribution, and reproduction in any medium, provided the original author and source are credited. Copyright: © 2013 Jia et al. This is an open-access article distributed under the terms of the Creative Commons Attribution Lic unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Abstract Background and Objective: The Wnt/β-catenin signaling pathway is essential for controlling bone mass; however, little is known about the variable effects of the constitutive activation of β-catenin (CA-β-catenin) on bone growth and remodeling at different postnatal stages. The goal of the present study was to observe the effects of CA-β-catenin on vertebral bone growth and remodeling in mice at different postnatal stages. In particular, special attention was paid to whether CA-β-catenin has detrimental effects on these processes. Methods: Catnblox(ex 3) mice were crossed with mice expressing the TM-inducible Cre fusion protein, which could be activated at designated time points via injection of tamoxifen. β-catenin was stabilized by tamoxifen injection 3 days, and 2, 4, 5, and 7 months after birth, and the effects lasted for one month. Radiographic imaging, micro- computed tomography, immunohistochemistry, and safranin O and tartrate-resistant acid phosphatase staining were employed to observe the effects of CA-β-catenin on vertebral bone growth and remodeling. Results: CA-β-catenin in both early (3 days after birth) and late stages (2, 4, 5, and 7 months after birth) increased bone formation and decreased bone resorption, which together increased vertebral bone volume. However, when β- catenin was stabilized in the early stage, vertebral linear growth was retarded, and the mice demonstrated shorter statures. In addition, the newly formed bone was mainly immature and located close to the growth plate. In contrast, when β-catenin was stabilized in the late stage, vertebral linear growth was unaffected, and the newly formed bone was mainly mature and evenly distributed throughout the vertebral body. Conclusions: CA-β-catenin in both early and late stages of growth can increase vertebral bone catenin has differential effects on vertebral growth and remodeling when activated at different postna Citation: Jia M, Chen S, Zhang B, Liang H, Feng J, et al. (2013) Effects of Constitutive β-Catenin Activation on Vertebral Bone Growth and Remodeling at Different Postnatal Stages in Mice. PLoS ONE 8(9): e74093. doi:10.1371/journal.pone.0074093 Editor: Dominique Heymann, Faculté de médecine de Nantes, France Received May 7, 2013; Accepted July 26, 2013; Published September 16, 2013 Copyright: © 2013 Jia 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. Introduction related protein 5 (LRP5), a Wnt co-receptor, induce high bone mass phenotype, whereas inactivating mutations cause osteoporosis-pseudoglioma syndrome, characterized by osteoporosis and blindness [9,10]. Studies in animal models with genetically modified LRP5 protein support these findings [11]. Overexpression of the Wnt ligand inhibitor Dickkopf-1, induced low bone mass and decreased bone formation [12]. In contrast, sclerostin (Wnt ligand inhibitor) knockout mice exhibited increased bone formation and high bone mass [13]. These results suggest that the canonical Wnt signaling The evolutionarily conserved Wnt signaling pathway mainly includes the canonical Wnt/β-catenin and non-canonical pathways that can be further subdivided into planar cell polarity and Wnt/Ca2+ pathways. Both canonical and non-canonical signaling pathways participate in many physiological and pathological processes [1-4]. Several lines of evidence suggest that Wnt/β-catenin signaling is essential for bone formation and resorption [4-8]. Activating mutations in lipoprotein receptor- PLOS ONE \| www.plosone.org September 2013 \| Volume 8 \| Issue 9 \| e74093 1 Effects of CA-β-Catenin on Vertebral Growth with TM (75 mg/kg) twice a week for one month. The start times for TM injection were day 3, and months 2, 4, 5, and 7. pathway plays an important role in controlling bone formation and resorption, thus presenting a potential target for the development of novel bone-building drugs. with TM (75 mg/kg) twice a week for one month. The start times for TM injection were day 3, and months 2, 4, 5, and 7. Site-specificity of Col1-Cre ERTM was examined by crossing Col1-Cre ERTM mice with ROSA26 reporter mice. Wild-type mice were also crossed with ROSA26 mice to serve as controls. TM was injected as aforementioned. The specificity of Col1-CreERTM expression was confirmed by β-gal staining. Site-specificity of Col1-Cre ERTM was examined by crossing Col1-Cre ERTM mice with ROSA26 reporter mice. Wild-type mice were also crossed with ROSA26 mice to serve as controls. TM was injected as aforementioned. The specificity of Col1-CreERTM expression was confirmed by β-gal staining. Wnt/β-catenin signaling involves multiple steps that may be considered for pharmacological intervention [5,14]. Enhanced Wnt/β-catenin signaling or deficiency/neutralization of Wnt antagonists is associated with increased bone formation and/or decreased bone resorption, suggesting potential therapeutic application in low bone volume conditions [15-22]. However, controversies exist with regard to several issues, including safety, dosage, duration of treatment, and mechanisms of intervention on osteoblastogenesis and osteoclastogenesis [22-26]. Materials and Methods The experimental protocol was reviewed and approved by the Ethical Committee of the Daping Hospital, Third Military Medical University (China). Radiographic imaging and tissue preparation At the designated time-points (1, 3, 5, 6 and 8 months after birth), mice were deeply anesthetized and radiographic images of entire skeletons obtained using a small animal X-ray machine (Model 8050-020, Field Emission Corporation, Inc., McMinnville, OR, USA) [29]. After whole-body X-ray was taken, 5 ml of whole blood was removed from the mice by heart puncture. The blood was incubated at room temperature for 30 min, and centrifuged for 5 min at 5000 rpm. The serum was then collected and enzyme-linked immunoassay (ELISA) or radioimmunoassay (RIA) was performed to determine the concentration of intact N-terminal propeptide of type I procollagen (P1NP), tartrate resistant acid phosphatase 5b (TRAP5b), osteocalcin, C-terminal telopeptides of type I collagen (CTX-I), according to the manufacturer’s instructions (Immunodiagnostic Systems Ltd, Boldon, UK). The large number of Wnt proteins, receptors, coreceptors, and soluble inhibitors of Wnt signaling makes it difficult to define the specific functions of Wnt /β-catenin signaling. Fortunately, a molecular node of canonical Wnt signaling involves β-catenin with only one encoding gene, which offers the possibility of genetic manipulation. Constitutive activation of β-catenin (CA-β-catenin) can thus be induced in mice expressing a β-catenin mutant allele in which all the serine and threonine residues of exon 3 are phosphorylated by GSK-3β [23]. In the present study, we investigated the effects of CA-β- catenin on vertebral bone growth and remodeling in mice at different stages using pro-collagen I Cre-ERTM as the promoter, which could be activated at designated time-points via injection of tamoxifen (TM) [27,28]. After whole-body x-ray was performed and whole blood was taken, the mice were sacrificed by overdose of anesthesia and the spines were removed and fixed in 4% paraformaldehyde at 4°C overnight. The fifth lumbar vertebra was cut and saved for micro-computed tomography (microCT) examination and real- time PCR. The samples that were left were decalcified in 10% ethylenediaminetetraacetic acid at 4°C for 7 days. X-ray imaging was used to determine if the samples were adequately decalcified. Then, the third and fourth lumbar vertebrae were tissue-processed, embedded in paraffin, and sectioned at a thickness of 5 µm horizontally or coronally. The sections were de-paraffined and rehydrated, and were used for hematoxylin and eosin (H&E) staining, immunohistochemistry, TRAP staining, and safranin O staining. Introduction To date, there have been no data demonstrating the effects of Wnt/β-catenin signaling pathway on bone growth and remodeling at different postnatal stages. In particular, no data are available on whether manipulation of this signaling pathway might have detrimental effects on bone growth. TRAP staining Two coplin jars (A and B) were pre-heated to 37°C with 50 ml stock basic incubation medium (sodium acetate anhydrous, 9.2 g; sodium tartrate dibasic dehydrate, 11.4 g; and glacial acetic acid, 2.8 ml dissolved in 1000 ml of distilled water and the pH was adjusted to 4.7-5.0 with 5M sodium Hydroxide). Then, 50 µl of 2% naphthol AS-BI phosphate substrate in ethylene glycol monoethyl ether was added to jar A, into which the slides were added, and incubated at 37°C for 45 min. A few minutes before the 45 min elapsed, 1 ml of 5% pararosaniline chloride and 1 ml of 4% sodium nitrite was mixed for 30 sec, incubated at room temperature for 2 min without mixing, and then transferred into jar B and mixed well. The slides from jar A were then moved into jar B without rinsing. Incubation at room temperature was done for 1-3 min until the color developed, followed by rinsing and counterstaining with methyl green for 5 min, dehydration, and covering with Permount. All chemicals were purchased from the Chuandong Corporation (Chongqing, China). MicroCT examination marker protein, pictures of each section were taken by an Olympus microscope and digital camera under a 200x magnification. The number of specific antigen-positive cells was counted in five random fields. The average percentage of positive cells and standard deviation were calculated in each group and used for statistics. The fifth lumbar vertebrae and tibia of mice from each group were dissected and scanned in a micro-CT imaging system (Scanco Medical, Bassersdorf, Switzerland), as described previously [29]. The scanning medium was ethanol, the x-ray tube potential was 45 kVp, and the voxel size was 10 µm3. Images were reconstructed and analyzed with EVS Beam software using a global threshold of 1400 Hounsfield units. Total vertebral bone volume fraction was calculated as the percentage of bone volume to total volume (BV/TV) of fifth lumbar vertebra. The mean trabecular thickness (Tb.Th), trabecular separation (Tb. Sp), and trabecular number (Tb.N) were also calculated [30]. As for the tibia, quantitative morphometry data were based on region of interest as following: trabecular bone region starting from growth plate reference level extending 44 slices (0.8 mm) distally, while cortical bone starting from mid-diaphysis extending 22 slices (0.4 mm) proximally. BV/TV, Tb.Th, Tb. Sp and Tb.N were qualified in trabecular bone region of tibia, while average cortical thickness (Ct.Th), total cross-sectional area (Tt.Ar), cortical bone area (Ct.Ar), cortical area fraction (Ct.Ar/Tt.Ar) were qualified in cortical bone region of tibia. Safranin O Staining Protocol Rehydrated sections were stained with Weigert’s iron hematoxylin for 1 min and rinsed in distilled water until clear. Sections were stained sequentially with 0.02% Fast Green for 5 min, 1% acetic acid for 30 s, and 0.1% Safranin O for 20 min. Slides were not rinsed between steps. Subsequently, slides were rinsed in 95% alcohol for 2 min, 100% alcohol for 3 min (twice), xylene for 2 min (twice), and coverslipped with Permount. All chemicals were purchased from Chuandong Corporation (Chongqing, China). Real-time PCR Genes Primers BSP F: 5’-CAATCCGTGCCACTCACT-3’, R: 5’-CAAACTCCAAGCCAAAGC-3’ GAPDH F: 5’-TCACTGCCACCCAGAAGA -3’, R: 5’-AAGTCGCAGGAGACAACC -3’ OPG F: 5’- GCATTATGACCCAGAAACT -3’, R: 5’- ACCTGAGAAGAACCCATC-3’ RANKL F: 5’- AACCAAGATGGCTTCTATTACC-3’, R: 5’- AAGGGTTGGACACCTGAATG-3’ RunX2 F: 5’-AGTCCCAACTTCCTGTGCT-3’, R: 5’-GGTGAAACTCTTGCCTCGTC-3’ TNAP F: 5’- ACGAGATGCCACCAGAGG-3’, R: 5’- AGTTCAGTGCGGTTCCAG -3’ TRAP F: 5’- GCCCTTACTACCGTTTGC-3’, R: 5’-TCTCGTCCTGAAGATACTGC-3’ doi: 10.1371/journal.pone.0074093.t001 Real-time PCR Mice expressing the TM-inducible Cre fusion protein, Cre- ERTM, under the control of a 3.2 kb mouse pro-collagen 1 promoter (3.2 kb Col1-Cre ERTM) active in osteoblasts, odontoblasts and tendon fibroblasts [27,28] were generated by TM injection and crossed with Catnb+/lox(exon 3) mice [23]. Genotyping was performed with a routine method. Briefly, DNA was extracted from the toe of each mouse using a standard protocol, and subjected to PCR for genotyping. The PCR primers for Cre were 5'-CCCGCAGAACCTGAAGATG-3' (sense) and 5'-GACCCGGCAAAACAGGTAG-3' (anti-sense), and the PCR primers for Catnb+/lox(exon 3) mice were 5'- AGGGTACCTGAAGCTCAGCG-3' (sense) and 5'- CAGTGGCTGACAGCAGCTTT-3' (anti-sense). Total RNA was extracted from 100 mg of the vertebral trabecular bone using TripureTM (Promega Corp, Madison, WI) according to the manufacturer’s instructions. The amount of RNA was quantified by spectrophotometry, after which cDNA was prepared from 1.0 µg RNA. The RNA was digested with DNase to eliminate any contaminating genomic DNA before quantitative real-time RT-PCR was performed. Then SYBR green detection method was used to examine the expression of receptor activator of nuclear factor kappa B ligand (RANKL), osteoprotegerin (OPG), TRCAP, tissue non-specific alkaline phosphatase (TNAP), RunX2, and bone sialoprotein (BSP). Glyceraldehyde-3-phosphatedehydrogenase (GAPDH) served as a control and the expression of a given gene was expressed as proportion relative to the average value of GAPDH. The relative ratio of RANKL:OPG expression was calculated and used for statistics. The primers used were sythesized by Sangon Biotech (Shang Hai, China) and are shown in Table 1. TM (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in a small volume of ethanol and diluted with corn oil at a concentration of 10 mg/ml [27,28]. CA-β-catenin and wild-type mice (n=6 at each time point) were intraperitoneally injected PLOS ONE \| www.plosone.org September 2013 \| Volume 8 \| Issue 9 \| e74093 2 Effects of CA-β-Catenin on Vertebral Growth Table 1. Primers used for real-time PCR. Genes Primers BSP F: 5’-CAATCCGTGCCACTCACT-3’, R: 5’-CAAACTCCAAGCCAAAGC-3’ GAPDH F: 5’-TCACTGCCACCCAGAAGA -3’, R: 5’-AAGTCGCAGGAGACAACC -3’ OPG F: 5’- GCATTATGACCCAGAAACT -3’, R: 5’- ACCTGAGAAGAACCCATC-3’ RANKL F: 5’- AACCAAGATGGCTTCTATTACC-3’, R: 5’- AAGGGTTGGACACCTGAATG-3’ RunX2 F: 5’-AGTCCCAACTTCCTGTGCT-3’, R: 5’-GGTGAAACTCTTGCCTCGTC-3’ TNAP F: 5’- ACGAGATGCCACCAGAGG-3’, R: 5’- AGTTCAGTGCGGTTCCAG -3’ TRAP F: 5’- GCCCTTACTACCGTTTGC-3’, R: 5’-TCTCGTCCTGAAGATACTGC-3’ doi: 10.1371/journal.pone.0074093.t001 Table 1. Primers used for real-time PCR. Table 1. Primers used for real-time PCR. Effect of CA-β-catenin on vertebral bone growth at different stages after birth The site specificity of Col1-CreERTM expression was confirmed by β-gal staining using ROSA26 mice. Positive staining was mainly found in osteoblasts, but not in the chondrocytes of the growth plate (Figure S1). Safranin O is commonly employed for proteoglycan staining. During rapid bone formation, immature bone stains positive for Safranin O [32,33]. Our data showed that the majority of cancellous bone in vertebral bodies of CA-β-catenin mice at month 1 was positive for Safranin O, indicating that excessive formed bone in vertebrae is immature. In contrast, at later stages (2, 4, 5 and 7 months after birth), excessive formed bone was negative for Safranin O staining (Figure 3). During the period that we performed our experiments, wild- type mice, Col-Cre mice, and β-catenin exon 3 fx +/+ mice (without Col I Cre) were injected with tamoxifen. The phenotypes of Col-Cre and β-catenin exon 3 fx +/+ mice (without Col I Cre) were the same as wild-type mice (data not shown). We feared that this data would make the manuscript too complicated, so only data from wild-type mice were presented as control in the current study. When β-catenin was stabilized with Col1-CreERTM as a promoter from day 3 of life, mice displayed shorter stature in their first month of life, compared with wild-type mice. In the group with β-catenin stabilization from month 2 to 3, mouse height was still slightly shorter than that of wild-type, but not to a significant extent. Interestingly, when β-catenin was stabilized from months 4, 5 and 7 of life, statures were similar to that in wild-type mice (Figure 1). Similar findings were observed in the long bone (Figure S3). Specifically, CA-β-catenin in both the early (3 days after birth) and late stages (2, 4, 5, and 7 months after birth) increased trabecular bone volume in the long bone (Figure S3). However, when β-catenin was stabilized from day 3 of life, most cancellous bone in the long bone were Safranin-O positive. In contrast, at later stages (2, 4, 5 and 7 months), most of the excessive formed bone were negative for Safranin O staining (Figure S4). CA-β-catenin had no obvious effect on cortical bone (Figure S5). Increased bone formation and decreased bone resorption contribute to excessive bone volume in CA- β-catenin mice Growth plate is the structure responsible for linear growth in vertebrates. H&E and Safranin O staining showed that growth plates in mice with β-catenin stabilization from day 3 were longer than that in wild-type mice (Figures 2 and 3). The hypertrophic layer of the growth plate was the most enlarged part. In contrast, β-catenin stabilization at months 4, 5, and 7 of life had no evident effect on growth plate length (Figures 2 and 3). These results clearly indicate that stabilization of β-catenin at early stages after birth hinders vertebral linear growth in mice. The balance of bone formation and bone resorption controls bone volume. TRAP staining showed that the number of osteoclasts in CA-β-catenin mice was lower than that in wild- type mice in both early and late postnatal stages (p<0.05, Figure 5A-I, Figure S6). RANKL was an important factor during the maturation of osteoclasts, and OPG was its decoy receptor. Real-time PCR showed that the RANKL:OPG ratio significantly decreased in CA-β-catenin mice at each time point compared with control mice (Figure 5J), which contributed to the decreased number of osteoclasts in CA-β-catenin mice. TNAP served as a useful marker of bone formation. IHC analyses showed higher expression of TNAP protein in ca-β-catenin than wild-type mice, indicative of increased bone formation ability (p<0.05, Figure 6A-E). Runx2 and BSP are markers for osteoblasts at the early and late stages, respectively [29,33]. Notably, the numbers of both Runx2 and BSP-positive cells in CA-β-catenin mice were greater than those in wild-type mice in both early and late postnatal stages (p<0.05, Figure 6F-J and Figure 6K-O). Similar findings were observed in the long bone. When β- catenin was stabilized from day 3 of life, the lengths of the tibia and femur were shorter than those in wild-type mice, and the hypertrophic layer of the growth plate was enlarged (Figure 1, Figure S2). In contrast, when β-catenin was stabilized 4, 5, and 7 months after birth, the lengths of the long bone and growth plate were not affected (Figure 1, Figure S2). Immunohistochemical analysis Immunohistochemistry was performed according to a previous report [31]. Briefly, 5-µm sections were blocked with 10% H2O2 at room temperature for 10 min, followed by treatment with 3% bovine serum albumin (BSA) at room temperature for 1 h. Sections were incubated with primary antibodies diluted at the appropriate concentrations in 2% BSA/0.1 M phosphate buffered saline (PBS) overnight at 4°C, washed in PBS, and further incubated with a biotinylated IgG antibody at room temperature for 1 h. Next, slides were treated with ABC reagents (Boster, Wuhan, China), and signals for antibody binding visualized with diaminobenzidine (Boster) substrate. Counterstaining was performed with Methyl Green. Slides were fully rinsed with 0.1 M PBS or distilled water between individual steps. The primary antibodies used included goat TNAP (1:200), goat polyclonal Runx2 (1:300) and rabbit anti- BSP (1:300). All primary antibodies were purchased from Santa Cruz Biotechnology, Inc (CA, USA), and biotinylated goat anti- mouse, rabbit anti-goat, and goat anti-rabbit IgG acquired from Boster. To obtain the percentage of cells expressing a given Pictures of each section were taken under a magnification of 200×, and the number of TRAP-positive cells was counted in five random fields. The average number and standard deviation of TRAP-positive cells were calculated and used for statistical analyses September 2013 \| Volume 8 \| Issue 9 \| e74093 PLOS ONE \| www.plosone.org 3 Effects of CA-β-Catenin on Vertebral Growth catenin mice (p<0.05, Figure 4). Except for the Tb.Th, Tb.N and Tb. Sp showed similar change in tibia of CA-β-catenin mice (Figure S3). Statistical analysis All data were expressed as means ± standard error. Statistical significance was evaluated using ANOVA with SPSS11.0 software. Data were considered significant at P<0.05. However, the location and extent of maturation of excessive formed bone varied with different stages of β-catenin stabilization. Upon induction of β-catenin stabilization from day 3 to month 1 of life, excessive formed bone was located in close proximity to the growth plate, while mice with β-catenin stabilization at the late stages (2, 4, 5 and 7 months after birth) displayed even distribution of excessive bone in the vertebral body (Figure 2). Effects of CA-β-Catenin on Vertebral Bone Volume at Different Stages after Birth The Wnt/β-catenin signaling pathway plays an important role in controlling bone volume. Our experiments disclosed significantly increased trabecular bone volumes in mice with constitutive activation of β-catenin at both early (day 3) and late stages (months 2, 4, 5 and 7), compared to wild-type mice (P <0.05, Figure 4, Figure S3). In addition, Tb.N and Tb.Th significantly increased in the vertebrae of CA-β-catenin mice, and Tb. Sp significantly decreased in the vertebrae of CA-β- The results of real-time PCR and ELISA further supported the histology findings. As shown in Figure 7A-C, the mRNA expression levels of Runx2, BSP, and TNAP significantly increased in CA-β-catenin mice compared with that in wild-type mice. PINP and osteocalcin are markers of bone formation, while CTX-I and TRAcP5b are markers of bone resorption. It PLOS ONE \| www.plosone.org September 2013 \| Volume 8 \| Issue 9 \| e74093 4 Effects of CA-β-Catenin on Vertebral Growth 1. Representative x-ray images of the spine. a-e X-ray images of the spine in wild-type and CA-β-catenin mice 1 month months (b), 5 months (c), 6 months (d) and 8 months (e) after birth. 71/journal.pone.0074093.g001 1. Representative x-ray images of the spine. a-e X-ray images of the spine in wild-type and CA-β-catenin mice 1 month months (b), 5 months (c), 6 months (d) and 8 months (e) after birth. 371/journal.pone.0074093.g001 Figure 1. Representative x-ray images of the spine. a-e X-ray images of the spine in wild-type and CA-β-catenin mice 1 month (a), 3 months (b), 5 months (c), 6 months (d) and 8 months (e) after birth. doi: 10.1371/journal.pone.0074093.g001 was found that PINP and osteocalcin serum concentrations significantly increased in CA-β-catenin mice, whereas CTX-I and TRAcP5b concentrations significantly decreased (p<0.05, Figure 7). Discussion Accumulating evidence supports the significance of Wnt/β- catenin signaling in bone formation and bone resorption. However, no data are available about the effect of the Wnt/β- catenin signaling pathway on bone growth and remodeling at different stages after birth. In the current study, we showed that constitutive β-catenin activation at both the early and late stages of growth results in increased bone formation and decreased bone resorption, which concertedly increase vertebral bone volume. However, the stabilization of β-catenin Taken together, these results indicate that stabilization of β- catenin in both early and late postnatal stages increases vertebral bone formation and decreases bone resorption. PLOS ONE \| www.plosone.org 5 September 2013 \| Volume 8 \| Issue 9 \| e74093 5 Effects of CA-β-Catenin on Vertebral Growth Figure 2. H&E staining of coronal sections of the fourth lumbar vertebra in wild-type and CA-β-catenin mice. Boxes in a, c, e, g, i, k, m, o, q and s are magnified in b, d, f, h, j, l, n, p, r and t, respectively. Bars: 200 µm (a, c, e, g, I, k, m, o, q and s) and 100 µm (b, d, f, h, j, l, n, p, r and t). doi: 10.1371/journal.pone.0074093.g002 Figure 2. H&E staining of coronal sections of the fourth lumbar vertebra in wild-type and CA-β-catenin mice. Boxes in a e, g, i, k, m, o, q and s are magnified in b, d, f, h, j, l, n, p, r and t, respectively. Bars: 200 µm (a, c, e, g, I, k, m, o, q and s) and µm (b, d, f, h, j, l, n, p, r and t). doi: 10.1371/journal.pone.0074093.g002 Figure 2. H&E staining of coronal sections of the fourth lumbar vertebra in wild-type and CA-β-catenin mice. Boxes in a, c, e, g, i, k, m, o, q and s are magnified in b, d, f, h, j, l, n, p, r and t, respectively. Bars: 200 µm (a, c, e, g, I, k, m, o, q and s) and 100 µm (b, d, f, h, j, l, n, p, r and t). doi: 10.1371/journal.pone.0074093.g002 at early and late stages had differential effects on vertebral growth. low bone mass, owing to accelerated bone resorption [11]. Similar results were observed with CA-β-catenin mice [23]. However, oral administration of AZD2858, a GSK-3 inhibitor, had no influence on the number of osteoclasts in mice [22]. Discussion Glass and colleagues suggested this difference might be caused by the fact that Wnt/β-catenin signaling in osteoblasts at different stages has variable effects on osteoclastogenesis [23]. Specifically, the Wnt/β-catenin pathway in mature osteoblasts negatively regulates osteoclastogenesis, but exerts The current study found that CA-β-catenin in both the early and late stages of growth increases bone volume by increasing bone formation and decreasing bone resorption. However, controversies exist regarding the role of the Wnt/β-catenin signaling pathway in osteoblastogenesis and osteoclastogenesis [22-26]. Using a mouse model with a Lrp6rs/rs mutation, Kubota et al. showed that mutant mice have PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org September 2013 \| Volume 8 \| Issue 9 \| e74093 6 Effects of CA-β-Catenin on Vertebral Growth Figure 3. Safranin O staining of lumbar vertebra. a-i Safranin O staining of coronal sections of the fourth lumbar vertebra in wild-type and CA-β-catenin mice at 1 month (a, b, e and f), 3 months (i and j), 5 months (c and d), 6 months (g and h) and 8 months (k and l) of age. Boxes in a and b are magnified in e and f, respectively. Bars: 200 µm (a, b, c, d, g, h, i, j, k and l) and 100 µm (e and f). doi: 10.1371/journal.pone.0074093.g003 Figure 3. Safranin O staining of lumbar vertebra. a-i Safranin O staining of coronal sections of the fourth lumbar vertebra in wild-type and CA-β-catenin mice at 1 month (a, b, e and f), 3 months (i and j), 5 months (c and d), 6 months (g and h) and 8 months (k and l) of age. Boxes in a and b are magnified in e and f, respectively. Bars: 200 µm (a, b, c, d, g, h, i, j, k and l) and 100 µm (e and f). d i 10 1371/j l 0074093 003 doi: 10.1371/journal.pone.0074093.g003 activation mainly occurs close to the growth plate and involves immature bone (staining positive for Safranin O), while excessive bone formed in the late stages is evenly distributed in the vertebral body and mature (staining negative for Safranin O). Until now, the precise mechanisms underlying the differential effects on vertebral growth by CA-β-catenin at different postnatal stages have remained unknown. Since the promoter used in the current study, pro-collagen I Cre-ERTM, is mainly expressed in osteoblasts, but not in chondrocytes, the effect of CA-β-catenin on the growth plate must be indirect. Discussion Our preliminary study showed that CA-β-catenin mice demonstrated hypophosphatemia, and a high phosphate diet could partially rescue the phenotype of enlarged growth plate in mice when β-catenin was stabilized at early stage (data not shown). These data indicated that hypophosphatemia might be the reason for retarded linear growth in CA-β-catenin mice. Future studies are needed to investigate the reason for these differential effects. However, regardless of the underlying mechanisms, these discrepancies should be addressed before therapeutics are used that target the Wnt/β-catenin signaling pathway. different effects in early osteoblasts and mesenchyaml stem cells. In the current study, a 3.2 kb Col I Cre that mainly activates Wnt/β-catenin in mature osteoblasts was employed [27,28], resulting in suppression of osteoclastogenesis. Although controversy exists, the effects of Wnt/β-catenin on bone formation are generally considered positive [19,25,34,35]. Wnt6, Wnt10a and Wnt10b inhibit adipogenesis and stimulate osteoblastogenesis through a β-catenin-dependent mechanism [19]. Moreover, oral administration of different types of GSK-3 inhibitors leads to increased bone formation [20,22]. Recently it was shown by Chen and colleagues that deletion of β-catenin in osteoblasts greatly reduced bone formation activity and indirectly increased osteoclast number and activity [36]. Consistent with that finding, the current study found that CA-β- catenin in both the early and late stages increased bone formation, which worked together with decreased bone resorption to increase bone volume. Another finding of the present study is that stabilization of β- catenin at early and late stages has differential effects on vertebral growth. When β-catenin was activated at the early stages (from day 3 after birth up to 1 month), maturation of vertebral growth plate was retarded in addition to increased bone formation, and mice displayed shortened statures. In contrast, when β-catenin was activated at the late stages (2, 4, 5 and 7 months after birth), growth plates were not significantly affected. Our results indicate that constitutive activation of β- catenin at the early stages has a detrimental effect on vertebral growth. Furthermore, excessive formation of bone in early In conclusion, the present study shows that CA-β-catenin at different stages after birth can increase vertebral bone volume. This increased bone volume was the result of increased bone formation and decreased bone resorption caused by CA-β- catenin. However, CA-β-catenin induced at different stages has differential effects on vertebral linear growth, as well as distribution and maturation of the newly formed bone. Discussion September 2013 \| Volume 8 \| Issue 9 \| e74093 PLOS ONE \| www.plosone.org 7 Effects of CA-β-Catenin on Vertebral Growth gure 4. MicroCT examination of lumbar vertebrae in each group. a-j representative, transverse MicroCT images of the fifth mbar vertebra in wild-type and CA-β-catenin mice at 1 month (a and b), 3 months (c and d), 5 months (e and f), 6 months (g and and 8 months (i and j) of age. k MicroCT analysis of BVF (BV/TV, %) of the fifth lumbar vertebra in CA-β-catenin mice and wild- pe mice. BV trabecular bone volume (mm3); TV total volume selected for analysis (mm3). l Trabecular number (Tb.N) of the fifth mbar vertebra in CA-β-catenin and wild-type mice. m Mean trabecular thickness (Tb.Th) of the fifth lumbar vertebra in CA-β- atenin and wild-type mice. n Trabecular separation (Tb. Sp) of the fifth lumbar vertebra in CA-β-catenin and wild-type mice. Bars present the mean ± SEM (n=6 for each group). p<0.05. i: 10.1371/journal.pone.0074093.g004 Figure 4. MicroCT examination of lumbar vertebrae in each group. a-j representative, transverse MicroCT images of the fifth lumbar vertebra in wild-type and CA-β-catenin mice at 1 month (a and b), 3 months (c and d), 5 months (e and f), 6 months (g and h) and 8 months (i and j) of age. k MicroCT analysis of BVF (BV/TV, %) of the fifth lumbar vertebra in CA-β-catenin mice and wild- type mice. BV trabecular bone volume (mm3); TV total volume selected for analysis (mm3). l Trabecular number (Tb.N) of the fifth lumbar vertebra in CA-β-catenin and wild-type mice. m Mean trabecular thickness (Tb.Th) of the fifth lumbar vertebra in CA-β- catenin and wild-type mice. n Trabecular separation (Tb. Sp) of the fifth lumbar vertebra in CA-β-catenin and wild-type mice. Bars represent the mean ± SEM (n=6 for each group). p<0.05. doi: 10.1371/journal.pone.0074093.g004 PLOS ONE \| www.plosone.org September 2013 \| Volume 8 \| Issue 9 \| e74093 8 Effects of CA-β-Catenin on Vertebral Growth Figure 5. TRAP staining of lumbar vertebra in each group. a-j TRAP staining of coronal sections of the fourth lumbar vertebra in wild-type and CA-β-catenin mice at 1 month (a and b), 3 months (c and d), 6 months (e and f), and 8 months (g and h) of age. Bars: 200 µm. i: The number of TRAP-positive cells in each group. j: The ratio of RANKL:OPG expression levels in each group. Discussion Bars represent the mean ± SEM (n=6 for each group). p<0.05. doi: 10.1371/journal.pone.0074093.g005 Figure 5. TRAP staining of lumbar vertebra in each group. a-j TRAP staining of coronal sections of the fourth lumbar vertebra in wild-type and CA-β-catenin mice at 1 month (a and b), 3 months (c and d), 6 months (e and f), and 8 months (g and h) of age. Bars: 200 µm. i: The number of TRAP-positive cells in each group. j: The ratio of RANKL:OPG expression levels in each group. Bars represent the mean ± SEM (n=6 for each group). p<0.05. doi: 10.1371/journal.pone.0074093.g005 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org September 2013 \| Volume 8 \| Issue 9 \| e74093 9 Effects of CA-β-Catenin on Vertebral Growth Figure 6. Examination of markers for bone formation and maturation by immunohistochemistry (IHC). a-d IHC against TNAP in wild-type and CA-β-catenin mice at 1 month (a, b) and 6 months (c, d) of age. e: Number of TNAP-positive cells in each group. f-i IHC against RunX2 in wild-type and CA-β-catenin mice at 1 month (f, g) and 6 months (h, i) of age. j: Number of RunX2- positive cells in each group. k-n IHC against BSP in WT and CA-β-catenin mice at 1 month (k, l) and 6 months (m, n) of age. o Number of BSP-positive cells in each group. Bars: 200 µm. Bars represent the mean ± SEM (n=6 for each group). p<0.05. doi: 10.1371/journal.pone.0074093.g006 Figure 6. Examination of markers for bone formation and maturation by immunohistochemistry (IHC). a-d IHC against TNAP in wild-type and CA-β-catenin mice at 1 month (a, b) and 6 months (c, d) of age. e: Number of TNAP-positive cells in each group. f-i IHC against RunX2 in wild-type and CA-β-catenin mice at 1 month (f, g) and 6 months (h, i) of age. j: Number of RunX2- positive cells in each group. k-n IHC against BSP in WT and CA-β-catenin mice at 1 month (k, l) and 6 months (m, n) of age. o Number of BSP-positive cells in each group. Bars: 200 µm. Bars represent the mean ± SEM (n=6 for each group). p<0.05. doi: 10.1371/journal.pone.0074093.g006 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 10 September 2013 \| Volume 8 \| Issue 9 \| e74093 Effects of CA-β-Catenin on Vertebral Growth xamination of markers for bone formation and bone resorption by Real-time PCR and EILISA. Supporting Information Figure S4. Safranin O staining of tibia in each group. a-h Safranin O staining of coronal sections of tibia in wild-type and CA-β-catenin mice at 1 month (a and b) 3 months (c,d g and h) 5 months (e and f), 6 months (i and j) and 8 months (k and l) of age. Boxes in c and d are magnified in g and h, respectively. Bars: 200 µm (a, b, c, d, e, f, i, j, k and l) and 100 µm (g and h). (TIF) Figure S1. Site-specificity of Col1-CreERTM. a Col 1 Cre expression was not found in growth plate in wild-type mice crossed with ROSA26 mice. b Col 1 Cre expression was not found in growth plate in Col1-CreERTM mice crossed with ROSA26 mice. c Col 1 Cre expression was not found in osteoblasts in wild-type mice crossed with ROSA26 mice. d Col 1 Cre expression was found in osteoblasts in Col1-CreERTM mice crossed with ROSA26 mice. Bars: 200 µm. Figure S5. MicroCT examination of cortical bone of tibia in each group. a-j representative, transverse MicroCT images of the cortical bone in CA-β-catenin and wild-type mice at 1 month (a and b), 3 months (c and d), 5 months (e and f), 6 months (g and h) and 8 months (i and j) of age. k Total cross-sectional area (Tt.Ar) in CA-β-catenin and wild-type mice. l Cortical bone area (Ct.Ar) in CA-β-catenin and wild-type mice. m Cortical area fraction (Ct.Ar/Tt.Ar) in CA-β-catenin and wild-type mice. n Average cortical thickness (Ct.Th) in CA-β-catenin and wild- type mice. Bars represent the mean ± SEM (n=6 for each group). p<0.05. (TIF) Figure S2. H&E staining of coronal sections of the tibia in wild-type and CA-β-catenin mice. Boxes in a, c, e, g, I, k, m, o, q and s are magnified in b, d, f, h, j, l, n, p, r and t, respectively. Bars: 200 µm (a, c, e, g, I, k, m, o, q and s) and 100 µm (b, d, f, h, j, l, n, p, r and t). (TIF) Figure S3. MicroCT examination of trabecular bone of proximal tibia in each group. a-j representative, transverse MicroCT images of the trabecular bone in CA-β-catenin and wild-type mice at 1 month (a and b), 3 months (c and d), 5 months (e and f), 6 months (g and h) and 8 months (i and j) of age. References 1. Clevers H, Nusse R (2012) Wnt/β-catenin signaling and disease. Cell 149: 1192-1205. doi:10.1016/j.cell.2012.05.012. PubMed: 22682243. promoter methylation is associated with enhanced Wnt signaling in advanced multiple myeloma. PLOS ONE 7: e30359. doi:10.1371/ journal.pone.0030359. PubMed: 22363428. promoter methylation is associated with enhanced Wnt signaling in advanced multiple myeloma. PLOS ONE 7: e30359. doi:10.1371/ journal.pone.0030359. PubMed: 22363428. j 2. Huang S, Zhu X, Liu Y, Tao Y, Feng G et al. (2012) Wls is expressed in the epidermis and regulates embryonic hair follicle induction in mice. PLOS ONE 27: e45904. doi:10.1371/journal.pone.0045904. PubMed: 23029304. 9. Williams BO, Insogna KL (2009) Where Wnts went: the exploding field of Lrp5 and Lrp6 signaling in bone. J Bone Miner Res 24: 171-178. doi: 10.1359/jbmr.081235. PubMed: 19072724. 3. Zhang Z, Wlodarczyk BJ, Niederreither K, Venugopalan S, Florez S et al. (2011) Fuz regulates craniofacial development through tissue specific responses to signaling factors. PLOS ONE 6: e24608. doi: 10.1371/journal.pone.0024608. PubMed: 21935430. 10. Joeng KS, Schumacher CA, Zylstra-Diegel CR, Long F, Williams BO (2011) Lrp5 and Lrp6 redundantly control skeletal development in the mouse embryo. Dev Biol 359: 222-229. PubMed: 21924256. 11. Kubota T, Michigami T, Sakaguchi N, Kokubu C, Suzuki A et al. (2008) Lrp6 hypomorphic mutation affects bone mass through bone resorption in mice and impairs interaction with Mesd. J Bone Miner Res 23: 1661-1671. doi:10.1359/jbmr.080512. PubMed: 18505367. 4. Yu HM, Jerchow B, Sheu TJ, Liu B, Costantini F et al. (2005) The role of Axin2 in calvarial morphogenesis and craniosynostosis. Development 132: 1995-2005. doi:10.1242/dev.01786. PubMed: 15790973. 12. Dela Cruz F, Terry M, Matushansky I (2012) A transgenic, mesodermal specific, Dkk1 mouse model recapitulates a spectrum of human congenital limb reduction defects. Differentiation 83: 220-230. doi: 10.1016/j.diff.2012.01.001. PubMed: 22406973. 5. Krishnan V, Bryant HU, Macdougald OA (2006) Regulation of bone mass by Wnt signaling. J Clin Invest 116: 1202-1209. doi:10.1172/ JCI28551. PubMed: 16670761. 6. Holmen SL, Zylstra CR, Mukherjee A, Sigler RE, Faugere MC et al. (2005) Essential role of beta-catenin in postnatal bone acquisition. J Biol Chem 280: 21162-21168. doi:10.1074/jbc.M501900200. PubMed: 15802266. 13. Li X, Ominsky MS, Niu QT, Sun N, Daugherty B et al. (2008) Targeted deletion of the sclerostin gene in mice results in increased bone formation and bone strength. J Bone Miner Res 23: 860-869. doi: 10.1359/jbmr.080216. PubMed: 18269310. 7. Gong Y, Bourhis E, Chiu C, Stawicki S, DeAlmeida VI et al. Author Contributions Conceived and designed the experiments: ZWZ. Performed the experiments: MJ SXC HL. Analyzed the data: BZ. Contributed reagents/materials/analysis tools: JQF. Wrote the manuscript: ZWZ JQF. Got funding: ZWZ. Supporting Information k MicroCT analysis of BVF (BV/TV, %) of the proximal tibia in CA-β-catenin mice and wild-type mice. BV trabecular bone volume (mm3); TV total volume selected for analysis (mm3). l Trabecular number (Tb.N) of the proximal tibia in CA- β-catenin and wild-type mice. m Mean trabecular thickness (Tb.Th) of the proximal tibia in CA-β-catenin and wild-type mice. n Trabecular separation (Tb. Sp) of the proximal tibia in CA-β-catenin and wild-type mice. Bars represent the mean ± SEM (n=6 for each group). p<0.05. Figure S6. TRAP staining of tibia in each group. a-h TRAP staining of coronal sections of tibia in wild-type and CA-β- catenin mice at 1 month (a and b), 3 months (c and d), 6 months (e and f), and 8 months (g and h) of age. (TIF) Discussion a Th vel of tissue non-specific alkaline phosphatase (TNAP) in each group. b The mRNA expression level of Runx mRNA expression level of bone sialoprotein (BSP) in each group. d The mRNA expression level of tartrate ase (TRAP) in each group. e The serum concentration of intact N-terminal propeptide of type I procollagen The serum concentration of osteocalcin in each group. g The serum concentration of C-terminal telopeptides X-I) in each group. h The serum concentration of tartrate resistant acid phosphatase 5b (TRAP5b) in each gr mean ± SEM (n=6 for each group). p<0.05. l 0074093 007 Figure 7. Examination of markers for bone formation and bone resorption by Real-time PCR and EILISA. a The mRNA expression level of tissue non-specific alkaline phosphatase (TNAP) in each group. b The mRNA expression level of Runx2 in each group. c The mRNA expression level of bone sialoprotein (BSP) in each group. d The mRNA expression level of tartrate resistant acid phosphatase (TRAP) in each group. e The serum concentration of intact N-terminal propeptide of type I procollagen (P1NP) in each group. f The serum concentration of osteocalcin in each group. g The serum concentration of C-terminal telopeptides of type I collagen (CTX-I) in each group. h The serum concentration of tartrate resistant acid phosphatase 5b (TRAP5b) in each group. Bars represent the mean ± SEM (n=6 for each group). p<0.05. doi: 10.1371/journal.pone.0074093.g007 September 2013 \| Volume 8 \| Issue 9 \| e74093 11 PLOS ONE \| www.plosone.org 11 Effects of CA-β-Catenin on Vertebral Growth (TIF) References (2010) Wnt isoform-specific interactions with coreceptor specify inhibition or potentiation of signaling by LRP6 antibodies. PLOS ONE 5: e12682. doi:10.1371/journal.pone.0012682. PubMed: 20856934. 14. Hoeppner LH, Secreto FJ, Westendorf JJ (2009) Wnt signaling as a therapeutic target for bone diseases. Expert Opin Ther Targets 13: 485-496. doi:10.1517/14728220902841961. PubMed: 19335070. 15. Kulkarni NH, Onyia JE, Zeng Q, Tian X, Liu M et al. (2006) Orally bioavailable GSK-3alpha/beta dual inhibitor increases markers of 8. Kocemba KA, Groen RW, van Andel H, Kersten MJ, Mahtouk K et al. (2012) Transcriptional silencing of the Wnt-antagonist DKK1 by PLOS ONE \| www.plosone.org 12 September 2013 \| Volume 8 \| Issue 9 \| e74093 Effects of CA-β-Catenin on Vertebral Growth Effects of CA-β-Catenin on Vertebral Growth fibroblasts. Mol Cells 30: 449-454. doi:10.1007/s10059-010-0139-3. PubMed: 20848229. cellular differentiation in vitro and bone mass in vivo. J Bone Miner Res 21: 910-920. doi:10.1359/jbmr.060316. PubMed: 16753022. 16. Hoeppner LH, Secreto FJ, Razidlo DF, Whitney TJ, Westendorf JJ (2011) Lef1DeltaN binds beta-catenin and increases osteoblast activity and trabecular bone mass. J Biol Chem 286: 10950-10959. doi: 10.1074/jbc.M110.165100. PubMed: 21270130. 27. Kamiya N, Kobayashi T, Mochida Y, Yu PB, Yamauchi M et al. (2010) Wnt inhibitors Dkk1 and Sost are downstream targets of BMP signaling through the type IA receptor (BMPRIA) in osteoblasts. J Bone Miner Res 25: 200-210. doi:10.1359/jbmr.090806. PubMed: 19874086. g yp p ( ) Res 25: 200-210. doi:10.1359/jbmr.090806. PubMed: 1987408 17. Padhi D, Jang G, Stouch B, Fang L, Posvar E (2011) Single-dose, placebo-controlled, randomized study of AMG 785, a sclerostin monoclonal antibody. J Bone Miner Res 26: 19-26. doi:10.1002/jbmr. 173. PubMed: 20593411. 28. Kamiya N, Ye L, Kobayashi T, Lucas DJ, Mochida Y et al. (2008) Disruption of BMP signaling in osteoblasts through type IA receptor (BMPRIA) increases bone mass. J Bone Miner Res 23: 2007-2017. doi: 10.1359/jbmr.080809. PubMed: 18684091. 29. Feng JQ, Ward LM, Liu S, Lu Y, Xie Y et al. (2006) Loss of DMP1 causes rickets and osteomalacia and identifies a role for osteocytes in mineral metabolism. Nat Genet 38: 1310-1315. doi:10.1038/ng1905. PubMed: 17033621. 18. Guo J, Liu M, Yang D, Bouxsein ML, Saito H et al. (2010) Suppression of Wnt signaling by Dkk1 attenuates PTH-mediated stromal cell response and new bone formation. Cell Metab 11: 161-171. doi: 10.1016/j.cmet.2009.12.007. PubMed: 20142103. 30. Bouxsein ML, Boyd SK, Christiansen BA, Guldberg RE, Jepsen KJ et al. (2010) Guidelines for assessment of bone microstructure in rodents using micro-computed tomography. References J Bone Miner Res 25:1468-1486. PubMed: 20533309. 19. Cawthorn WP, Bree AJ, Yao Y, Du B, Hemati N et al. (2012) Wnt6, Wnt10a and Wnt10b inhibit adipogenesis and stimulate osteoblastogenesis through a β-catenin-dependent mechanism. Bone 50: 477-489. doi:10.1016/j.bone.2011.08.010. PubMed: 21872687. 31. Zong ZW, Cheng TM, Su YP, Ran XZ, Shen Y et al. (2008) Recruitment of transplanted dermal multipotent stem cells to sites of injury in rats with combined radiation and wound injury by interaction of SDF-1 and CXCR4. Radiat Res 170: 444-450. doi:10.1667/RR0744.1. PubMed: 19024651. 20. Clément-Lacroix P, Ai M, Morvan F, Roman-Roman S, Vayssière B et al. (2005) Lrp5-independent activation of Wnt signaling by lithium chloride increases bone formation and bone mass in mice. Proc Natl Acad Sci U S A 102: 17406–17411. doi:10.1073/pnas.0505259102. PubMed: 16293698. 21. Verkaar F, van der Stelt M, Blankesteijn WM, van der Doelen AA, Zaman GJ (2011) Discovery of novel small molecule activators of β- catenin signaling. PLOS ONE 6: e19185. doi:10.1371/journal.pone. 0019185. PubMed: 21559429. 32. Baek WY, Lee MA, Jung JW, Kim SY, Akiyama H et al. (2009) Positive regulation of adult bone formation by osteoblast-specific transcription factor osterix. J Bone Miner Res 24: 1055-1065. doi:10.1359/jbmr. 081248. PubMed: 19113927. 22. Marsell R, Sisask G, Nilsson Y, Sundgren-Andersson AK, Andersson U et al. (2012) GSK-3 inhibition by an orally active small molecule increases bone mass in rats. Bone 50: 619-627. doi:10.1016/j.bone. 2011.11.007. PubMed: 22142634. 33. Baek WY, de Crombrugghe B, Kim JE (2010) Postnatally induced inactivation of Osterix in osteoblasts results in the reduction of bone formation and maintenance. Bone 46: 920-928. doi:10.1016/j.bone. 2009.12.007. PubMed: 20026264. 34. Gaur T, Lengner CJ, Hovhannisyan H, Bhat RA, Bodine PV, Komm BS et al. (2005) Canonical WNT signaling promotes osteogenesis by directly stimulating Runx2 gene expression. J Biol Chem 280: 33132-33140. doi:10.1074/jbc.M500608200. PubMed: 16043491. 23. Glass DA 2nd, Bialek P, Ahn JD, Starbuck M, Patel MS et al. (2005) Canonical Wnt signaling in differentiated osteoblasts controls osteoclast differentiation. Dev Cell 8: 751-764. doi:10.1016/j.devcel. 2005.02.017. PubMed: 15866165. 24. Boland GM, Perkins G, Hall DJ, Tuan RS (2004) Wnt 3a promotes proliferation and suppresses osteogenic differentiation of adult human mesenchymal stem cells. J Cell Biochem 93: 1210-1230. doi:10.1002/ jcb.20284. PubMed: 15486964. 35. Day TF, Guo X, Garrett-Beal L, Yang Y (2005) Wnt/beta-catenin signaling in mesenchymal progenitors controls osteoblast and chondrocyte differentiation during vertebrate skeletogenesis. Dev Cell 8: 739-750. doi:10.1016/j.devcel.2005.03.016. PubMed: 15866164. 25. Effects of CA-β-Catenin on Vertebral Growth References Chen Y, Whetstone HC, Lin AC, Nadesan P, Wei Q et al. (2007) Beta- catenin signaling plays a disparate role in different phases of fracture repair: implications for therapy to improve bone healing. PLOS Med 4: e249. doi:10.1371/journal.pmed.0040249. PubMed: 17676991. 36. Chen J, Long F (2013) β-catenin promotes bone formation and suppresses bone resorption in postnatal growing mice. J Bone Miner Res 28: 1160-1169. doi:10.1002/jbmr.1834. PubMed: 23188722. 26. Heo JS, Lee SY, Lee JC (2010) Wnt/β-catenin signaling enhances osteoblastogenic differentiation from human periodontal ligament PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 13 September 2013 \| Volume 8 \| Issue 9 \| e74093
https://openalex.org/W2529309802	https://nnp.ima-press.net/nnp/article/download/613/563	Russian	null	Effect of cerebrolysin on motor function restoration during medical rehabilitation	Nevrologiâ, nejropsihiatriâ, psihosomatika	2,016	cc-by	6,225	Кустова М.А., Толмачев А.П., Шамалов Н.А. НИИ цереброваскулярной патологии и инсульта ГБОУ ВПО «Российский национальный исследовательский медицинский университет им. Н.И. Пирогова» Минздрава России, Москва, Россия 117997, Москва, ул. Островитянова, 1 Влияние церебролизина на восстановление двигательной функции в процессе медицинской реабилитации Ишемический инсульт (ИИ) характеризуется высокими распространенностью, а также смертностью и инвалидизацией больных Терапия, направленная на коррекцию одного биохимического или молекулярного этапа ишемического повреждения клеток, не при- водит к успеху при инсульте, что указывает на необходимость изучения мультимодальной терапии, действующей на несколько связанных патофизиологических звеньев. Представлены опубликованные в январе 2016 г. результаты рандомизированного плацебоконтролируемого многоцентрового иссле- дования CARS, в котором продемонстрирован положительный эффект церебролизина по сравнению с плацебо по первичному кри- терию эффективности, шкале ARAT (тест оценки функции руки) и общему исходу спустя 90 дней после начала заболевания. В ис- следование включали преимущественно пациентов с умеренным или тяжелым ИИ (средний исходный балл по National Institutes of Health Stroke Scale, NIHSS – 9). Особенностями исследования CARS по сравнению с другими клиническими испытаниями нейропротекторов явились исходное пла- нирование более узких конечных критериев эффективности (восстановление двигательной функции руки, тогда как во многих ис- следованиях главной целью было снижение летальности), а также стандартизированная программа реабилитации в обеих груп- пах лечения. Ранее в подобных исследованиях характер и объем реабилитационных мероприятий не учитывались, хотя такие ме- роприятия могут оказать существенное влияние на исход инсульта. Исследование CARS является первым среди ранее проведенных клинических испытаний нейропротекторов, в котором достигнута первичная цель (восстановление двигательной функции), что открывает новые возможности для медикаментозной поддержки ре- абилитационных мероприятий у больных ИИ. Ключевые слова: ишемический инсульт; реабилитация; нейропротекторы; церебролизин; исследование CARS. Контакты: Николай Анатольевич Шамалов; shamalovn@gmail.com Для ссылки: Кустова МА, Толмачев АП, Шамалов НА. Влияние церебролизина на восстановление двигательной функции в процес- се медицинской реабилитации. Неврология, нейропсихиатрия, психосоматика. 2016;8(2):80–86. Effect of cerebrolysin on motor function restoration during medical rehabilitation Effect of cerebrolysin on motor function restoration during medical rehabilitation Kustova M.A., Tolmachev A.P., Shamalov N.A. Research Institute of Cerebrovascular Pathology and Stroke, N.I. Pirogov Russian National Research Medical University, Ministry of Health of Russia, Moscow, Russia 1, Ostrovityanov St., Moscow 117997 Effect of cerebrolysin on motor function restoration during medical rehabilitation Kustova M.A., Tolmachev A.P., Shamalov N.A. Research Institute of Cerebrovascular Pathology and Stroke, N.I. Pirogov Russian National Research Medical University, Ministry of Health of Russia, Moscow, Russia 1, Ostrovityanov St., Moscow 117997 Effect of cerebrolysin on motor function restoration during medical rehabilitation Kustova M.A., Tolmachev A.P., Shamalov N.A. Research Institute of Cerebrovascular Pathology and Stroke, N.I. Pirogov Russian National Research Medical University, Ministry of Health of Russia, Moscow, Russia 1, Ostrovityanov St., Moscow 117997 Ischemic stroke (IS) is characterized by high prevalence, mortality, and disability rates. О Б З О Р Ы О Б З О Р Ы DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 Ишемический инсульт (ИИ) является проблемой чрезвычайной медицинской и социальной значимости вследствие его высокой распространенности, а также смертности и инвалидизации таких больных [1]. Внутри- венный тромболизис с использованием рекомбинантного тканевого активатора плазминогена (rt-PA), а также метод тромбэкстракции являются наиболее эффективными в ле- чении ИИ в первые часы после его развития [2], однако ре- перфузионная терапия проводится не более чем 20–30% па- циентов [3–5]. Несмотря на положительные результаты экспериментальных исследований нейропротекторов [6], эффективность данной группы препаратов в клинических исследованиях не подтверждена [7–9], что может быть свя- зано с неадекватными моделями ишемии у животных [10], а также с дизайном исследований с участием человека. Те- рапия, направленная на коррекцию одного биохимического или молекулярного этапа патофизиологического каскада ишемического повреждения клеток, не приводит к успеху при инсульте, что говорит о необходимости изучения муль- тимодальной терапии, включающей лекарственные соеди- нения, действующие на несколько связанных патофизиоло- гических звеньев. Одним из таких мультимодальных препа- ратов является церебролизин, представляющий собой ком- плекс низкомолекулярных нейропептидов (<10 кДа) и сво- бодных аминокислот, получаемых из головного мозга сви- ньи с помощью стандартного производственного процесса. В ряде экспериментальных исследований было показано, что церебролизин эффективно влияет на эксайтотоксич- ность, подавляет образование свободных радикалов, акти- вацию микроглии/нейровоспаление и активацию калпаи- на/апоптоз и, кроме того, проявляет нейротрофическую ак- тивность: способствует спраутингу нейронов, повышает вы- живаемость клеток и стимулируют нейрогенез [11–15]. вления после инсульта на фоне проведения реабилитацион- ных мероприятий [21]. В исследовании CARS церебролизин или плацебо в су- точной дозе 30 мл вводили 1 раз в сутки на протяжении 21 дня инфузионно в течение 20 мин, в период от 24 до 72 ч с мо- мента развития инсульта. Каждому пациенту, включенному в исследование, проводилась стандартизированная про- грамма реабилитации длительностью в 21 день (5 дней в не- делю по 2 ч в сутки), состоявшая из массажа и пассивных и ак- тивных движений верхними и нижними конечностями. Пос- ле выписки пациенты продолжали занятия 2 раза по 15 мин 3 дня в неделю. Продолжительность исследования для каж- дого пациента составляла 90 дней. В исследование включали пациентов 18–80 лет с полу- шарным ИИ (верифицированным при помощи компьютер- ной томографии или магнитно-резонансной томографии) объемом >4 см3. DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 У пациентов, включенных в исследование, до инсульта не наблюдалось значимых нарушений (балл до инсульта по модифицированной шкале Рэнкина [МШР] со- ставлял 0–1), в течение предыдущих 3 мес они не переноси- ли инсульта, балл теста оценки функции руки [22] (ARAT) составлял <50 (оценка в диапазоне от 0 [функциональная активность отсутствует] до 57 [нарушения отсутствуют]), а балл по коммуникационной шкале Гудгласса и Каплана [23, 24] был >2 (из расчета от 0 [тяжелая афазия] до 5 [мини- мальная афазия]). Пациентов исключали из исследования по следующим причинам: прогрессирующий или неста- бильный инсульт; наличие в анамнезе или во время иссле- дования активного неврологического или психического заболевания; значимая алкогольная или наркотическая зависимость в течение предыдущих 3 лет; заболевание пече- ни, почек, сердца или легких в поздних стадиях; ожидаемая выживаемость <1 года; значительное снижение сознания на момент рандомизации; любое состояние, которое представ- ляет собой противопоказание для введения церебролизина, в том числе аллергия; беременность или период грудного вскармливания; участие в другом исследовании инсульта или восстановления после инсульта. Церебролизин изучали в нескольких клинических ис- следованиях в остром периоде ИИ [16–19], однако эти ис- следования проводились на небольшой выборке – преиму- щественно от 50 до 200 рандомизированных пациентов. Ре- троспективный подгрупповой анализ (n=252), основанный на данных более масштабного рандомизированного двой- ного слепого плацебоконтролируемого исследования CASTA выявил тенденцию к снижению летальности и улуч- шению состояния у пациентов с более тяжелым инсультом, получавших церебролизин (при поступлении балл по National Institutes of Health Stroke Scale, NIHSS >12) [20]. Те- рапию в указанных клинических исследованиях начинали в первые 12–48 ч после развития инсульта и обычно проводи- ли в течение 10 дней, более длительное применение препа- рата ранее не изучалось. В январе 2016 г. были опубликова- ны результаты проспективного рандомизированного двой- ного слепого плацебоконтролируемого многоцентрового исследования CARS («Церебролизин и восстановление по- сле инсульта»), целью которого явилось изучение эффек- тивности и безопасности церебролизина в период восстано- Первичным критерием эффективности в исследова- нии CARS являлось изменение балла по шкале ARAT [22]. Данную шкалу использовали для оценки двигательной ак- тивности верхних конечностей от начала терапии (исход- ный уровень) до 90-го дня. Влияние церебролизина на восстановление двигательной функции в процессе медицинской реабилитации Therapy aimed to correct one biochemical or molecular stage of ischemic cell injury fails to treat stroke, suggesting that it is necessary to study multimodality therapy affecting several relat- ed pathophysiological components. The paper gives the January 2016 results of the randomized placebo-controlled multicenter study CARS that demonstrates the positive effect of cerebrolysin versus placebo according to the primary efficiency criterion, the Action Research Arm Test (ARAT) scale, and total outcome 90 days after disease onset. The investigation enrolled mainly patients with moderate or severe IS (the mean National Institutes of Health Stroke Scale score was 9 at baseline). The specific features of the CARS study versus those of other clinical trials of neuroprotectors were the initial planning of narrower end criteria of efficiency (arm motor function recovery whereas the major goal of many investigations was to reduce mortality rates), as well as a standardized rehabilitation program in both treatment groups. Such investigations did not previously take into account the nature and volume of rehabilitation measures although the latter may have a substantial impact on the outcome of stroke. The CARS study is the first among the previously conducted clinical trials of neuroprotectors, which has attained the primary objective (to restore motor function), which opens up fresh opportunities for the medical support of rehabilitation measures in patients with IS. 80 О Б З О Р Ы Keywords: ischemic stroke; rehabilitation; neuroprotectors; cerebrolysin; CARS study. Keywords: ischemic stroke; rehabilitation; neuroprotectors; cerebrolysin; CARS study. Contact: Nikolai Anatolyevich Shamalov; shamalov@gmail.com For reference: Kustova MA, Tolmachev AP, Shamalov NA. Effect of cerebrolysin on motor function restoration during medical rehabilitation. Nevrologiya, neiropsikhiatriya, psikhosomatika = Neurology, neuropsychiatry, psychosomatics. 2016;8(2):80–86. For reference: Kustova MA, Tolmachev AP, Shamalov NA. Effect of cerebrolysin on motor function restoration during medical rehabilitation. Nevrologiya, neiropsikhiatriya, psikhosomatika = Neurology, neuropsychiatry, psychosomatics. 2016;8(2):80–86. evrologiya, neiropsikhiatriya, psikhosomatika = Neurology, neuropsychiatry, psychosomatics. 2016;8(2):80–86 Nevrologiya, neiropsikhiatriya, psikhosomatika = Neurology, neuropsychiatry, psychosomatics. 2016;8(2):80 DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 2); * – СО, рассчитанное с момента начала инсульта; АГ – артериальная гипертензия (здесь и в табл. 3); СД – сахарный диабет; ИБС – ишемическая болезнь сердца. ческая шкала депрессии), начиная от исходного уровня до 21-го дня (пос- ледний день, когда вводили исследуе- мый препарат) и 90-го дня. ческая шкала депрессии), начиная от исходного уровня до 21-го дня (пос- ледний день, когда вводили исследуе- мый препарат) и 90-го дня. Таблица 2. И с х о д н ы е п о к а з а т е л и к р и т е р и е в э ф ф е к т и в н о с т и Критерий эффективности Церебролизин Плацебо (n=104) (n=101) ARAT (паретичная сторона): среднее±СО 10,1±15,9 10,7±16,5 медиана (МКР) 0,0 (21,5) 2,0 (18,0) NIHSS: среднее±СО 9,1±3,2 9,2±3,2 медиана (МКР) 8,0 (4,0) 8,0 (5,0) Индекс Бартел: среднее±СО 35,5±24,9 35,4±24,6 медиана (МКР) 30,0 (40,0) 30,0 (40,0) МШР: среднее±СО 3,9±0,8 3,9±0,8 медиана (МКР) 4,0 (0,0) 4,0 (1,0) Примечание. МКР – межквартильный размах. Таблица 2. И с х о д н ы е п о к а з а т е л и к р и т е р и е в э ф ф е к т и в н о с т и С апреля 2008 г. по сентябрь 2010 г. в исследование было включено 208 па- циентов. Все пациенты получили ми- нимум 1 дозу исследуемого препарата или плацебо (104 – церебролизин и 104 – плацебо), 12 пациентов прежде- временно прекратили участие в иссле- довании: из-за наличия нежелательных явлений (НЯ), связанных с цереброли- зином (n=2) и плацебо (n=5); по собст- венному желанию (церебролизин, n=2; плацебо, n=2) или по административ- ным причинам (плацебо, n=1). Исход- ные характеристики пациентов в груп- пах церебролизина и плацебо были со- поставимы (табл. 1 и 2). Средний воз- раст больных составил 64 года, мужчин было 63 балл по шкале NIHSS при поступлении – 9,2 (м AR N И М П Анализ вторичных критериев эффективности позво- лил также выявить существенные различия между группа- ми, получавшими церебролизин и плацебо, по другим шка- лам. Хорошая степень функционального восстановления (от 0 до 1 балла по МШР) зарегистрирована у 42,3% пациен- тов в группе церебролизина по сравнению с 14,9% в группе плацебо, аналогичные результаты получены по МШР (от 0 до 2 баллов; рис. 2). DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 Вторичными критериями эффе- ктивности были: изменения скорости ходьбы (тест на ско- рость ходьбы), функции тонкой моторики (тест с девятью отверстиями и стержнями), общего неврологического де- фицита (NIHSS), уровня нарушений или зависимости в по- вседневной жизни (индекс Бартел, МШР), степени афазии (коммуникационная шкала Гудгласса и Каплана) [23, 24], степени игнорирования (тест вычеркивания линий, тест на выявление пропусков), качества жизни (опросник SF-36), сводный балл физического компонента и сводный балл психического компонента) и степени депрессии (гериатри- 81 О Б З О Р Ы Таблица 1. И с х о д н ы е д е м о г р а ф и ч е с к и е х а р а к т е р и с т и к и п а ц и е н т о в Таблица 1. И с х о д н ы е д е м о г р а ф и ч е с к и е х а р а к т е р и с т и к и п а ц и е н т о в Параметр Всего Церебролизин Плацебо (n=208) (n=104) (n=104) Примечание. ИМТ – индекс массы тела; CO – стандартное отклонение (здесь и в табл. 2); * – СО, рассчитанное с момента начала инсульта; АГ – артериальная гипертензия (здесь и в табл. 3); СД – сахарный диабет; ИБС – ишемическая болезнь сердца. Мужчины, n (%) 133 (63,9) 70 (67,3) 63 (60,6) Правши, n (%) 199 (95,7) 99 (95,2) 100 (96,2) Средний возраст, годы (СО) 64,0 (10,2) 64,9 (9,8) 63,0 (10,6) Средний ИМТ, кг/м2 (СО) 27,4 (4,2) 27,2 (4,1) 27,6 (4,3) Среднее время до начала лечения, ч (СО)* 53,2 (12,3) 51,9 (12,7) 54,6 (11,7) Тромболитическая терапия, n (%) 4 (1,9) 2 (1,9) 2 (1,9) Наличие факторов риска, n (%): АГ 173 (83,2) 86 (82,7) 87 (83,7) гиперлипидемия 105 (50,5) 55 (52,9) 50 (48,1) СД 39 (18,8) 19 (18,3) 20 (19,2) аритмия 54 (26,0) 26 (25,0) 28 (26,9) ИБС 83 (39,9) 38 (36,5) 45 (43,3) Курение в прошлом/в настоящее время 67 (32,2) 33 (31,8) 34 (32,7) Примечание. ИМТ – индекс массы тела; CO – стандартное отклонение (здесь и в табл. 2); * – СО, рассчитанное с момента начала инсульта; АГ – артериальная гипертензия (здесь и в табл. 3); СД – сахарный диабет; ИБС – ишемическая болезнь сердца. Примечание. ИМТ – индекс массы тела; CO – стандартное отклонение (здесь и в табл. DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 Величина изменений по шкале ARAT относительно исходного показателя в популяции больных. Анализ проведены с помощью критерия Вилкоксона–Манна–Уитни (б) Рис. 1. Динамика показателей по шкале ARAT при применении церебролизина (30 мл/сут) и плацебо, представленная в виде ко- робчатых диаграмм для 7-го (V3), 14-го (V4) и 21-го (V5) дня от начала терапии (начальная точка) и 42-го (V6) и 90-го (V7) дня после инсульта (а). Величина изменений по шкале ARAT относительно исходного показателя в популяции больных. Анализ проведены с помощью критерия Вилкоксона–Манна–Уитни (б) группы церебролизина и у 7 (6,7%) из группы плацебо за- регистрированы серьезные нежелательные явления (СНЯ), ни одно из которых, не было расценено как свя- занное с исследуемыми препаратами (табл. 4). СНЯ в группе церебролизина включали: тяжелую перифериче- скую ишемию, почечные колики (умеренной интенсив- ности) и острый инфаркт миокарда. Все эти СНЯ разре- шились в течение исследования. В группе плацебо умерли 4 (3,8%) пациента. Причинами летальных исходов явились сепсис с острой почечной недостаточностью и комой, сеп- сис с полиорганной недостаточностью, ишемия кишечни- ка, субдуральная и внутримозговая гематома. В группе, получавшей церебролизин, летальных исходов не было. У 69,2% пациентов, получавших церебролизин, заре- гистрировано по меньшей мере одно НЯ по сравнению с 71,2% пациентов в группе плацебо. Большинство НЯ име- ли легкую степень (в группе церебролизина – 76,1%, пла- цебо – 69,8%). Наиболее часто возникавшие во время ле- чения НЯ, описанные не менее чем у 5% пациентов в лю- бой группе, приведены в табл. 3. У 3 (2,9%) пациентов из Рис. 2. Распределение баллов по МШР. Кумулятивный процент (церебролизин по сравнению с плацебо): 8,65 по сравнению с 2,97 (0), 42,31 по сравнению с 14,85 (1), 65,38 по сравнению с 33,66 (2), 88,46 по сравнению с 75,25 (3), 98,08 по сравнению с 96,04 (4) и 100,0 по сравнению с 100,0 (5) Церебролизин (n=104) Плацебо (n=101) Баллы 0 1 2 3 4 5 6 0 20 40 60 80 100 Количество пациентов, % Церебролизин (n=104) Плацебо (n=101) Баллы 0 1 2 3 4 5 6 0 20 40 60 80 100 Количество пациентов, % Рис. 2. Распределение баллов по МШР. DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 Анализ первичного критерия эффективности показал увеличение балла по шкале ARAT с 10,1±15,9 (0,0; 21,5) при включении в исследование (среднее±СО, медиана, МКР) до 40,7±20,2 (51,0; 28,0) на 90-й день в группе, получавшей церебролизин, и с 10,7±16,5 (2,0; 18,0) до 26,5±21,0 (27,0; 44,0) в группе плацебо (рис. 1, а). Средние абсолютные из- менения баллов по шкале ARAT на 90-й день после инсуль- та по сравнению с исходными значениями составили 30,7±19,9 (32,0; 36,5) в группе церебролизина и 15,9±16,8 (11,0; 22,0) в группе плацебо. Увеличение балла по шкале ARAT наблюдалось у 96 (92,3%) из 104 пациентов из группы церебролизина по сравнению с 85 (84,2%) из 101 пациента из группы плацебо. Непараметрический анализ продемон- стрировал превосходство церебролизина над плацебо на 90-й день (рис. 1, б). Умеренное превосходство церебролизина наблюда- лось по 6 из 12 критериев эффективности, включая ARAT, NIHSS, индекс Бартел, МРШ, SF-36 (общий физический компонент) и шкалу депрессии (гериатрическая шкала депрессии; рис. 3). Незначительное превосходство церебро- лизина было продемонстрировано с помощью теста скоро- сти ходьбы, теста с девятью отверстиями и стержнями, ком- муникационной шкалы Гудгласса и Каплана и SF-36 (общий психический компонент). 82 О Б З О Р Ы Рис. 1. Динамика показателей по шкале ARAT при применении церебролизина (30 мл/сут) и плацебо, представленная в виде ко- робчатых диаграмм для 7-го (V3), 14-го (V4) и 21-го (V5) дня от начала терапии (начальная точка) и 42-го (V6) и 90-го (V7) дня после инсульта (а). Величина изменений по шкале ARAT относительно исходного показателя в популяции больных. Анализ проведены с помощью критерия Вилкоксона–Манна–Уитни (б) ARA-LCB3 MW=0,5280 LB=0,4540 UB=0,6019 P=0,4610 T:102/R:101 ARA-LCB4 MW=0,6191 LB=0,5407 UB=0,6975 P=0,0030 T:102/R:101 ARA-LCB5 MW=0,6391 LB=0,5598 UB=0,7185 P=0,0006 T:102/R:101 ARA-LCB6 MW=0,6778 LB=0,5976 UB=0,7580 P=0,0000 T:102/R:101 ARA-LCB7 MW=0,7118 LB=0,6307 UB=0,7928 P=0,0000 T:104/R:101 Значительное Умеренное Незначительное Равенство Незначительное Умеренное Значительное Преимущество церебролизина Преимущество плацебо Величина эффекта (критерий Манна–Уитни) 1,0 0,8 0,6 0,4 0,2 0 Баллы 50 40 30 20 10 0 50 40 30 20 10 0 Исходно V3 V4 V5 V6 V7 Исходно V3 V4 V5 V6 V7 Церебролизин Плацебо а б Баллы 50 40 30 20 10 0 50 40 30 20 10 0 Исходно V3 V4 V5 V6 V7 Исходно V3 V4 V5 V6 V7 Церебролизин Плацебо а Плацебо Рис. 1. Динамика показателей по шкале ARAT при применении церебролизина (30 мл/сут) и плацебо, представленная в виде ко- робчатых диаграмм для 7-го (V3), 14-го (V4) и 21-го (V5) дня от начала терапии (начальная точка) и 42-го (V6) и 90-го (V7) дня после инсульта (а). DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 Н а и б о л е е ч а с т о р е г и с т р и р у е м ы е Н Я ( ≥5 % п а ц и е н т о в ; п о п у л я ц и я д л я о ц е н к и б е з о п а с н о с т и ) ; n ( % ) (средний исходный балл по шкале NIHSS – 9), так как гипотезообразую- щий анализ в подгруппах в исследова- нии CASTA [20] выявил тенденцию к более высоким результатам после те- рапии церебролизином у пациентов с баллом по шкале NIHSS >12 (n=246). Данный подгрупповой анализ пока- зал, что после терапии церебролизи- ном у пациентов наблюдается улучше- ние балла по шкале NIHSS на 3 пунк- та на 90-й день по сравнению с груп- пой плацебо, и выявил величину эф- фекта, указывающую на умеренное превосходство церебролизина по сравнению с плацебо для всех доменов комбинированной конечной точки (NIHSS, индекс Бартел и МШР). (средний исходный балл по шкале NIHSS – 9), так как гипотезообразую- щий анализ в подгруппах в исследова- нии CASTA [20] выявил тенденцию к более высоким результатам после те- рапии церебролизином у пациентов с баллом по шкале NIHSS >12 (n=246). Данный подгрупповой анализ пока- зал, что после терапии церебролизи- ном у пациентов наблюдается улучше- ние балла по шкале NIHSS на 3 пунк- та на 90-й день по сравнению с груп- пой плацебо, и выявил величину эф- фекта, указывающую на умеренное превосходство церебролизина по сравнению с плацебо для всех доменов комбинированной конечной точки (NIHSS, индекс Бартел и МШР). Исследование CARS также под- тверждает результаты предыдущего Таблица 4. DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 Кумулятивный процент (церебролизин по сравнению с плацебо): 8,65 по сравнению с 2,97 (0), 42,31 по сравнению с 14,85 (1), 65,38 по сравнению с 33,66 (2), 88,46 по сравнению с 75,25 (3), 98,08 по сравнению с 96,04 (4) и 100,0 по сравнению с 100,0 (5) Таким образом, результаты рандомизированного плацебоконтролируемого многоцентрового исследования CARS продемонстрировали положительный эффект це- ребролизина по сравнению с плацебо по первичному критерию эффективности, шкале ARAT и общему исходу спустя 90 дней. У 69,2% пациентов, получавших церебролизин, заре- гистрировано по меньшей мере одно НЯ по сравнению с 71,2% пациентов в группе плацебо. Большинство НЯ име- ли легкую степень (в группе церебролизина – 76,1%, пла- цебо – 69,8%). Наиболее часто возникавшие во время ле- чения НЯ, описанные не менее чем у 5% пациентов в лю- бой группе, приведены в табл. 3. У 3 (2,9%) пациентов из В данном исследовании принимали участие преиму- щественно пациенты с умеренным или тяжелым инсультом 83 О Б З О Р Ы О Б З О Р Ы Оценка по шкалам на 90-й день ARAT GAIT VELOCITI 9 HOLE PEG NIHSS BARTEL MRS GOODGLASS LINE CANG. GAP SF 36 PCS SF 36 MCS DEPRESSION Общее состояние на 90-й день Величина эффекта (критерий Манна–Уитни) 0,29 0,36 0,44 0,5 0,56 0,64 0,71 В пользу плацебо В пользу церебролизина MW 0,7118 0,5937 0,5612 0,6754 0,6720 0,7339 0,5614 0,4696 0,4981 0,6727 0,5602 0,6805 0,6159 95% ДИ (0,6307–0,7928) (0,4585–0,7289) (0,4777–0,6448) (0,5977–0,7530) (0,5922–0,7518) (0,6612–0,8065) (0,4938–0,6290) (0,4041–0,5351) (0,4281–0,5681) (0,5900–0,7553) (0,4795–0,6409) (0,6007–0,7603) (0,5799–0,6518) N1/N2 104/101 34/35 90/93 104/101 104/101 104/101 104/101 98/100 97/100 102/95 102/95 102/95 104/101 р 0,0000 0,1743 0,1509 0,0000 0,0000 0,0000 0,75 0,3627 0,9574 0,0000 0,1438 0,0000 0,0000 Оценка по шкалам на 90-й день ARAT GAIT VELOCITI 9 HOLE PEG NIHSS BARTEL MRS GOODGLASS LINE CANG. GAP SF 36 PCS SF 36 MCS DEPRESSION Общее состояние на 90-й день Величина эффекта (критерий Манна–Уитни) 0,29 0,36 0,44 0,5 0,56 0,64 0,71 В пользу плацебо В пользу церебролизина MW 0,7118 0,5937 0,5612 0,6754 0,6720 0,7339 0,5614 0,4696 0,4981 0,6727 0,5602 0,6805 0,6159 95% ДИ (0,6307–0,7928) (0,4585–0,7289) (0,4777–0,6448) (0,5977–0,7530) (0,5922–0,7518) (0,6612–0,8065) (0,4938–0,6290) (0,4041–0,5351) (0,4281–0,5681) (0,5900–0,7553) (0,4795–0,6409) (0,6007–0,7603) (0,5799–0,6518) N1/N2 104/101 34/35 90/93 104/101 104/101 104/101 104/101 98/100 97/100 102/95 102/95 102/95 104/101 р 0,0000 0,1743 0,1509 0,0000 0,0000 0,0000 0,75 0,3627 0,9574 0,0000 0,1438 0,0000 0,0000 Оценка по шкалам на 90-й день ARAT GAIT VELOCITI 9 HOLE PEG NIHSS BARTEL MRS GOODGLASS LINE CANG. DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 GAP SF 36 PCS SF 36 MCS DEPRESSION Общее состояние на 90-й день Величина эффекта (критерий Манна–Уитни) 0,29 0,36 0,44 0,5 0,56 0,64 0,71 В пользу плацебо В пользу церебролизина MW 0,7118 0,5937 0,5612 0,6754 0,6720 0,7339 0,5614 0,4696 0,4981 0,6727 0,5602 0,6805 0,6159 95% ДИ (0,6307–0,7928) (0,4585–0,7289) (0,4777–0,6448) (0,5977–0,7530) (0,5922–0,7518) (0,6612–0,8065) (0,4938–0,6290) (0,4041–0,5351) (0,4281–0,5681) (0,5900–0,7553) (0,4795–0,6409) (0,6007–0,7603) (0,5799–0,6518) N1/N2 104/101 34/35 90/93 104/101 104/101 104/101 104/101 98/100 97/100 102/95 102/95 102/95 104/101 р 0,0000 0,1743 0,1509 0,0000 0,0000 0,0000 0,75 0,3627 0,9574 0,0000 0,1438 0,0000 0,0000 Величина эффекта (критерий Манна–Уитни) Рис. 3. Оценка по шкалам на 90-й день. Величина эффекта (критерий Манна–Уитни, MW) для отдельных и комбинированных параметров эффективности (процедура Вэй–Лачин) отражает изменения относительно исходного уровня в модифицированной популяции в соответствии с рандомизацией (перенос данных последнего наблюдения, n=205). Анализы проведены с использовани- ем многофакторного теста Вилкоксона. MRS – модифицированная шкала Рэнкина; MCS – сводный компонент психического здоровья; PCS – сводный компонент физического здоровья (качества жизни). ДИ – доверительный интервал Рис. 3. Оценка по шкалам на 90-й день. Величина эффекта (критерий Манна–Уитни, MW) для отдельных и комбинированных параметров эффективности (процедура Вэй–Лачин) отражает изменения относительно исходного уровня в модифицированной популяции в соответствии с рандомизацией (перенос данных последнего наблюдения, n=205). Анализы проведены с использовани- ем многофакторного теста Вилкоксона. MRS – модифицированная шкала Рэнкина; MCS – сводный компонент психического здоровья; PCS – сводный компонент физического здоровья (качества жизни). ДИ – доверительный интервал Таблица 3. Н а и б о л е е ч а с т о р е г и с т р и р у е м ы е Н Я ( ≥5 % п а ц и е н т о в ; п о п у л я ц и я д л я о ц е н к и б е з о п а с н о с т и ) ; n ( % ) НЯ Церебролизин Плацебо (n=104) (n=104) Инфекция мочевыводящих путей 13 (12,5) 17 (16,3) Депрессия 11 (10,6) 10 (9,6) Бессонница 6 (5,8) 4 (3,8) Атеросклероз сонной артерии 5 (4,8) 5 (4,8) Головная боль 6 (5,8) 3 (2,9) Стеноз сонной артерии 6 (5,8) 2 (1,9) АГ 9 (8,7) 12 (11,5) Гепатит с явлениями цитолиза 10 (9,6) 8 (7,7) Боль в верхней части живота 6 (5,8) 4 (3,8) Таблица 3. 15. Darsalia V, Heldmann U, Lindvall O, Kokaia Z. Stroke-induced neurogenesis in aged brain. Stroke. 2005 Aug;36(8):1790-5. Epub 2005 Jul 7. doi: 10.1161/01. STR.0000173151.36031.be. 16. Скворцова ВИ, Стаховская ЛВ, Губский ЛВ и др. Рандомизированное двой- ное слепое плацебоконтролируемое иссле- дование безопасности и эффективности це- ребролизина для лечения острого ишемиче- ского инсульта. Журнал неврологии и пси- хиатрии им. С.С. Корсакова (Прил. Инсульт). 2004;(S11):51-5. [Skvortsova VI, Stakhovskaya LV, Gubskii LV, et al. A randomized, double-blind, placebo- controlled trial of safety and efficacy of Cerebrolysin for treating acute ischaemic stroke. Zhurnal nevrologii i psikhiatrii im. S.S. Korsakova (Suppl. Stroke). 2004;(S11):51-5. (In Russ.)]. 17. Ladurner G, Kalvach P, Moessler H; Cerebrolysin Study Group. Neuroprotective treatment with cerebrolysin in patients with acute stroke: a randomised controlled trial. J Neural Transm (Vienna). 2005 Mar;112(3):415-28. Epub 2004 Dec 7. doi: 10.1007/s00702-004-0248-2. 18. Ziganshina LE, Abakumova T, Kuchaeva A. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2010 Apr 14;(4): CD007026. doi: 10.1002/14651858. CD007026.pub2. DOI: http://dx.doi.org/10.14412/2074-2711-2016-2-80-86 П о к а з а т е л и б е з о п а с н о с т и ( п о п у л я ц и я д л я а н а л и з а б е з о п а с н о с т и ) Параметр безопасности Всего (n=208) Церебролизин (n=104) Плацебо (n=104) Средняя продолжительность применения, сут 20,4 20,5 20,3 Пациенты с НЯ, n (%) 146 (70,2) 72 (69,2) 74 (71,2) в том числе: связанными с терапией 44 (21,2) 22 (21,2) 22 (21,2) приведшими к отмене терапии 7 (3,4) 2 (1,9) 5 (4,8) Количество НЯ, n 400 201 199 Пациенты с СНЯ, n (%) 10 (4,8) 3 (2,9) 7 (6,7) в том числе: связанными с терапией 0 0 0 приведшими к отмене терапии 6 (2,9) 1 (1,0) 5 (4,8) Количество СНЯ, n 16 3 13 Случаи смерти пациентов, n (%) 4 (1,9) 0 4 (3,8) Примечание. НЯ – нежелательное явление, которое возникло впервые или усилилось после начала терапии. П о к а з а т е л и б е з о п а с н о с т и ( п о п у л я ц и я д л я а н а л и з а б е з о п а с н о с т и ) лица 4. П о к а з а т е л и б е з о п а с н о с т и ( п о п у л я ц и я д л я а н а л и з а б е з о п а с н о с т и ) раметр безопасности Всего (n=208) Церебролизин (n=104) Плац 84 in stroke: from preclinical to clinical studies. Pharmacology. 2013;92(5-6):324-34. doi: 10.1159/000356320. Epub 2013 Dec 12. 8. Xu SY, Pan SY. The failure of animal models of neuroprotection in acute ischemic stroke to translate to clinical efficacy. Med Sci Monit Basic Res. 2013 Jan 28;19:37-45. 9. Tymianski M. Novel approaches to neuropro- tection trials in acute ischemic stroke. Stroke. 2013 Oct;44(10):2942-50. doi: 10.1161/ STROKEAHA.113.000731. Epub 2013 Sep 10. 10. Hossmann KA. The two pathophysiologies of focal brain ischemia: implications for transla- tional stroke research. J Cereb Blood Flow Metab. 2012 Jul;32(7):1310-6. doi: 10.1038/jcbfm.2011.186. Epub 2012 Jan 11. 11. Masliah E, Diez-Tejedor E. The pharmacology of neurotrophic treatment with Cerebrolysin: brain protection and repair to counteract pathologies of acute and chronic neurological disorders. Drugs Today (Barc). 2012 Apr;48 Suppl A:3-24. doi: 10.1358/ dot.2012.48(Suppl.A).1739716. 12. Hartbauer M, Hutter-Paier B, Skofitsch G, Windisch M. Antiapoptotic effects of the pep- tidergic drug cerebrolysin on primary cultures of embryonic chick cortical neurons. J Neural Transm (Vienna). 2001;108(4):459-73. 13. Zhang L, Chopp M, Meier DH, et al. Sonic hedgehog signaling pathway mediates cerebrolysin-improved neurological function after stroke. Stroke. 2013 Jul;44(7):1965-72. doi: 10.1161/STROKEAHA.111.000831. Epub 2013 May 21. 14. Gutmann B, Hutter-Paier B, Skofitsch G, et al. In vitro models of brain ischemia: the peptidergic drug cerebrolysin protects cultured chick cortical neurons from cell death. Neurotox Res. 2002 Feb;4(1):59-65. doi: 10.1080/ 10298420290007637. in stroke: from preclinical to clinical studies. Pharmacology. 2013;92(5-6):324-34. О Б З О Р Ы О Б З О Р Ы исследования W. Lang и соавт. [19], в котором цереброли- зин вводили в течение 10 дней после системного тромболи- зиса, после чего отмечалось выраженное улучшение в пер- вые 30 дней после начала заболевания. Однако с течением времени различия между двумя группами исчезали и стано- вились незначимыми через 90 дней после начала инсульта (0 баллов по МШР имели 30,4% пациентов из группы цере- бролизина по сравнению с 23,7% пациентов из группы пла- цебо), отсутствие существенных нарушений жизнедеятель- ности (1 балл по МШР) наблюдалось у 21,4% пациентов из группы, получавшей церебролизин, и у 28,8% пациентов, получавших плацебо. Положительный эффект цереброли- зина в исследовании CARS наблюдался в течение более длительного периода лечения, составившего 21 день, кроме того, выявлены более низкие темпы полного восстановле- ния пациентов из группы плацебо. Низкий уровень спон- танного восстановления в группе плацебо может быть свя- зан с включением пациентов с умеренным или тяжелым инсультом (средний исходный балл по шкале NIHSS – 9). тивно проводились реабилитационные мероприятия. Кроме того, раннее начало восстановительной терапии, возможно, оказало влияние на наблюдавшиеся исходы, что проявилось в более быстром начальном улучшении клинического состояния. В целом особенностями исследования CARS по срав- нению с другими клиническими испытаниями нейропроте- кторов явились исходное планирование более узких конечных критериев эффективности (восстановление дви- гательной функции руки, тогда как во многих исследовани- ях главной целью было снижение летальности), а также стандартизированная программа реабилитации в обеих группах лечения (как правило, ранее в подобных исследова- ниях характер и объем реабилитационных мероприятий не учитывались, хотя такие мероприятия могут оказать суще- ственное влияние на исход инсульта). Таким образом, исследование CARS является первым среди ранее проведенных клинических испытаний нейро- протекторов, в котором была достигнута первичная цель (восстановление двигательной функции), что открывает но- вые возможности для медикаментозной поддержки реаби- литационных мероприятий у больных с ИИ. Сложно напрямую сравнивать результаты исследо- вания CARS с предыдущими исследованиями цереброли- зина, так как в данном исследовании в обеих группах ак- Л И Т Е Р А Т У Р А 1. Murray CJ, Lopez AD. Measuring the global burden of disease. N Engl J Med. 2013 Aug 1;369(5):448-57. doi: 10.1056/NEJMra1201534. 2. Powers WJ, Derdeyn CP, Biller J, et al., on behalf of the American Heart Association Stroke Council. 2015 AHA/ASA Focused Update of the 2013 Guidelines for the Early Management of Patients With Acute Ischemic Stroke Regarding Endovascular Treatment. A Guideline for Healthcare Professionals from the American Heart Association/American Stroke Association. Stroke. 2015; 46:3020-3035. Stroke. 2015;46:3020-3035 Published online before print June 29, 2015, doi: 10.1161/ STR.0000000000000074 1. Murray CJ, Lopez AD. Measuring the global burden of disease. N Engl J Med. 2013 Aug 1;369(5):448-57. doi: 10.1056/NEJMra1201534. 2. Powers WJ, Derdeyn CP, Biller J, et al., on behalf of the American Heart Association Stroke Council. 2015 AHA/ASA Focused Update of the 2013 Guidelines for the Early Management of Patients With Acute Ischemic Stroke Regarding Endovascular Treatment. A Guideline for Healthcare Professionals from the American Heart Association/American Stroke Association. Stroke. 2015; 46:3020-3035. Stroke. 2015;46:3020-3035 Published online before print June 29, 2015, doi: 10.1161/ STR.0000000000000074 3. Saver JL, Smith EE, Fonarow GC, et al. GWTG- Stroke Steering Committee and Investigators. The «golden hour» and acute brain ischemia: presenting features and lytic therapy in >30,000 patients arriving within 60 minutes of stroke onset. Stroke. 2010 Jul;41(7):1431-9. doi: 10.1161/ STROKEAHA.110.583815. Epub 2010 Jun 3. 4. Zahuranec DB, Majersik JJ. Percentage of acute stroke patients eligible for endovascular treatment. Neurology. 2012 Sep 25;79(13 Suppl 1):S22-5. 5. Шамалов НА. Проблемы и перспективы реперфузионной терапии при ишемическом инсульте в России. Фарматека. 2015;302(9):14-9. [Shamalov NA. Problems and prospects of reperfusion therapy in ischemic stroke in Russia. Farmateka. 2015;302(9):14-9. (In Russ.)]. 6. O’Collins VE, Macleod MR, Donnan GA, et al. 1,026 experimental treatments in acute stroke. Ann Neurol. 2006 Mar;59(3):467-77. doi: 10.1002/ana.20741 7 Kaur H Prakash A Medhi B Drug therapy Л И Т Е Р А Т У Р А 1. Murray CJ, Lopez AD. Measuring the global burden of disease. N Engl J Med. 2013 Aug 1;369(5):448-57. doi: 10.1056/NEJMra1201534. 2. Powers WJ, Derdeyn CP, Biller J, et al., on behalf of the American Heart Association Stroke Council. 2015 AHA/ASA Focused Update of the 2013 Guidelines for the Early Management of Patients With Acute Ischemic Stroke Regarding Endovascular Treatment. A Guideline for Healthcare Professionals from the American Heart Association/American Stroke Association. Stroke. 2015; 46:3020-3035. Stroke. 2015;46:3020-3035 Published online before print June 29, 2015, doi: 10.1161/ STR.0000000000000074 1. Murray CJ, Lopez AD. Measuring the global burden of disease. N Engl J Med. 2013 Aug 1;369(5):448-57. doi: 10.1056/NEJMra1201534. 2. Powers WJ, Derdeyn CP, Biller J, et al., on behalf of the American Heart Association Stroke Council. 2015 AHA/ASA Focused Update of the 2013 Guidelines for the Early Management of Patients With Acute Ischemic Stroke Regarding Endovascular Treatment. A Guideline for Healthcare Professionals from the American Heart Association/American Stroke Association. Stroke. 2015; 46:3020-3035. Stroke. 2015;46:3020-3035 Published online before print June 29, 2015, doi: 10.1161/ STR.0000000000000074 15. Darsalia V, Heldmann U, Lindvall O, Kokaia Z. Stroke-induced neurogenesis in aged brain. Stroke. 2005 Aug;36(8):1790-5. Epub 2005 Jul 7. doi: 10.1161/01. STR.0000173151.36031.be. 16. Скворцова ВИ, Стаховская ЛВ, Губский ЛВ и др. Рандомизированное двой- ное слепое плацебоконтролируемое иссле- дование безопасности и эффективности це- ребролизина для лечения острого ишемиче- ского инсульта. Журнал неврологии и пси- хиатрии им. С.С. Корсакова (Прил. Инсульт). 2004;(S11):51-5. [Skvortsova VI, Stakhovskaya LV, Gubskii LV, et al. A randomized, double-blind, placebo- controlled trial of safety and efficacy of Cerebrolysin for treating acute ischaemic stroke. Zhurnal nevrologii i psikhiatrii im. S.S. Korsakova (Suppl. Stroke). 2004;(S11):51-5. (In Russ.)]. 17. Ladurner G, Kalvach P, Moessler H; Cerebrolysin Study Group. Neuroprotective treatment with cerebrolysin in patients with acute stroke: a randomised controlled trial. J Neural Transm (Vienna). 2005 Mar;112(3):415-28. Epub 2004 Dec 7. doi: 10.1007/s00702-004-0248-2. 18. Ziganshina LE, Abakumova T, Kuchaeva A. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2010 Apr 14;(4): CD007026. doi: 10.1002/14651858. CD007026.pub2. 17. Ladurner G, Kalvach P, Moessler H; Cerebrolysin Study Group. Neuroprotective treatment with cerebrolysin in patients with acute stroke: a randomised controlled trial. J Neural Transm (Vienna). 2005 12. Hartbauer M, Hutter-Paier B, Skofitsch G, Windisch M. Antiapoptotic effects of the pep- tidergic drug cerebrolysin on primary cultures of embryonic chick cortical neurons. J Neural Transm (Vienna). 2001;108(4):459-73. 4. Zahuranec DB, Majersik JJ. Hoemberg V, et al. Cerebrolysin and Recovery After Stroke (CARS). A Randomized, Placebo- Controlled, Double-Blind, Multicenter Trial. Stroke. 2016 Jan;47(1):151-9. doi: 10.1161/STROKEAHA.115.009416. Epub 2015 Nov 12. 22. Lyle RC. A performance test for assessment of upper limb function in physical rehabilitation treatment and research. Int J Rehabil Res. Л И Т Е Р А Т У Р А Percentage of acute stroke patients eligible for endovascular treatment. Neurology. 2012 Sep 25;79(13 Suppl 1):S22-5. 2012 Sep 25;79(13 Suppl 1):S22-5. 18. Ziganshina LE, Abakumova T, Kuchaeva A. Cerebrolysin for acute ischaemic stroke. 2012 Sep 25;79(13 Suppl 1):S22 5. 5. Шамалов НА. Проблемы и перспективы реперфузионной терапии при ишемическом инсульте в России. Фарматека. 2015;302(9):14-9. [Shamalov NA. Problems and prospects of reperfusion therapy in ischemic stroke in Russia. Farmateka. 2015;302(9):14-9. (In Russ.)]. 6. O’Collins VE, Macleod MR, Donnan GA, et al. 1,026 experimental treatments in acute stroke. Ann Neurol. 2006 Mar;59(3):467-77. doi: 10.1002/ana.20741 7 Kaur H Prakash A Medhi B Drug therapy 5. Шамалов НА. Проблемы и перспективы реперфузионной терапии при ишемическом инсульте в России. Фарматека. 2015;302(9):14-9. [Shamalov NA. Problems and prospects of reperfusion therapy in ischemic stroke in Russia. Farmateka. 2015;302(9):14-9. (In Russ.)]. 13. Zhang L, Chopp M, Meier DH, et al. Sonic hedgehog signaling pathway mediates cerebrolysin-improved neurological function after stroke. Stroke. 2013 Jul;44(7):1965-72. doi: 10.1161/STROKEAHA.111.000831. Epub 2013 May 21. 2015;302(9):14-9. [Shamalov NA. Problems and prospects of reperfusion therapy in ischemic stroke in Russia. Farmateka. 2015;302(9):14-9. (In Russ.)]. 19. Lang W, Stadler CH, Poljakovic Z, Fleet D; Lyse Study Group. A prospective, randomized, placebo-controlled, double-blind trial about safety and efficacy of combined treatment with alteplase (rt-PA) and Cerebrolysin in acute ischaemic hemispheric stroke. Int J Stroke. 2013 Feb;8(2):95-104. doi: 10.1111/j.1747- 4949.2012.00901.x. Epub 2012 Sep 26. 14. Gutmann B, Hutter-Paier B, Skofitsch G, et al. In vitro models of brain ischemia: the peptidergic drug cerebrolysin protects cultured chick cortical neurons from cell death. Neurotox Res. 2002 Feb;4(1):59-65. doi: 10.1080/ 10298420290007637. 6. O’Collins VE, Macleod MR, Donnan GA, et al. 1,026 experimental treatments in acute stroke. Ann Neurol. 2006 Mar;59(3):467-77. doi: 10.1002/ana.20741 7. Kaur H, Prakash A, Medhi B. Drug therapy 85 20. Heiss WD, Brainin M, Bornstein NM, et al; Cerebrolysin Acute Stroke Treatment in Asia (CASTA) Investigators. Cerebrolysin in patients with acute ischemic stroke in Asia: results of a double-blind, placebo-controlled randomized trial. Stroke. 2012 Mar;43(3):630-6. doi: 10.1161/STROKEAHA.111.628537. Epub 2012 Jan 26. 21. Muresanu DF, Heiss WD, О Б З О Р Ы О Б З О Р Ы 1981;4(4):483-92. doi: 10.1097/ 00004356-198112000-00001. 23. Posteraro L, Formis A, Grassi E, et al. Quality of life and aphasia. Multicentric stan- dardization of a questionnaire. Eura Medicophys. 2006 Sep;42(3):227-30. 24. Goodglass H, Kaplan E. The Assessment of Aphasia and Related Disorders. 2nd ed. Philadelphia, PA: Lea & Febiger; 1983. 21. Muresanu DF, Heiss WD, Исследование не имело спонсорской поддержки. Авторы несут полную ответственность за предоставление окончатель- ной версии рукописи в печать. Все авторы принимали участие в разработке концепции статьи и написании рукописи. Окон- чательная версия рукописи была одобрена всеми авторами. 86 86
https://openalex.org/W2135960481	https://sykepleien.no/sites/default/files/documents/forsknings/1145744.pdf	Norwegian	null	Forskrivning, bruk og dokumentasjon av effekt	Sykepleien forskning	2,013	cc-by	5,538	nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning ORIGINALARTIKKEL > Behovsmedisinering i sykehjem: Forskrivning, bruk og dokumentasjon av effekt nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning ORIGINALARTIKKEL > Behovsmedisinering i sykehjem: Forskrivning, bruk og dokumentasjon av effekt nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning NALARTIKKEL > Behovsmedisinering i sykehjem: Forskrivning, bruk og dokumentasjon av effekt Illustrasjonsfoto: Erik M. Sundt Illustrasjonsfoto: Erik M. Sundt mens studien pågikk. Dokumenta- sjon av effekt for utleverte behovsle- gemidler var mangelfull, både når det gjaldt kvantitet og kvalitet. Effekten ble dokumentert i 1/3 av tilfellene. Metode: Vi undersøkte 15 ukers-pre- valensen av foreskrevne og utleverte behovslegemidler, samt dokumenta- sjon av effekt for disse gjennom inn- samling av data fra legemiddelkort, elektronisk pasientjournal og kvit- teringsark fra 108 pasienter ved ett sykehjem i Bergen. Deskriptive ana- lyser ble utført i SPSS. sammendrag Bakgrunn: For å sikre sykehjems- pasienter effektiv og forsvarlig lege- middelbehandling, kreves hensikts- messige rutiner for forskrivning, hånd- tering og effektdokumentasjon både av legemidler gitt fast og ved behov. Vår kunnskap om forholdene rundt behovsmedisinering er begrenset. Konklusjon: Det er behov for hyppi- gere gjennomgang av legemiddelkor- tene med tanke på å endre, seponere eller forskrive behovslegemidler slik at ordinasjonene bedre gjenspei- ler pasientenes faktiske behov. God effektdokumentasjon er en forutset- ning for å kunne «skreddersy» lege- middelbehandlingen. Hensikt: Få kunnskap om behovsme- disinering ved å kartlegge forskrivning og utlevering av behovslegemidler ved ett utvalgt sykehjem, samt undersøke hvor ofte og hvordan effekten av gitte behovslegemidler ble dokumentert. Resultater: Pasientene fikk i gjen- nomsnitt forskrevet fire behovslege- midler hver i løpet av studieperioden. To av tre foreskrevne legemidler ble imidlertid aldri gitt til pasientene English Summary English Summary The documentation of effects for the administered P.R.N. medication was inadequate, both regarding quantity and quality. The drug effects were documented in 1/3 of the cases. The documentation of effects for the administered P.R.N. medication was inadequate, both regarding quantity and quality. The drug effects were documented in 1/3 of the cases. Medication Needs in Nursing Homes: Prescribing, Use and Documentation of Effect Medication Needs in Nursing Homes: Prescribing, Use and Documentation of Effect Method: We investigated the pre- valence of prescribed and adminis- tered P.R.N. Behovsmedisinering i sykehjem: Forskrivning, bruk og dokumentasjon av effekt Forfattere: Stine Wang Rønningen, Kjersti Bakken, Anne Gerd Granås Forfattere: Stine Wang Rønningen, Kjersti Bakken, Anne Gerd Granås eldre på sykehjem. Her finner vi de sykeste eldre og omfattende legemiddelbruk. Tilstrekkelig kompetanse hos de ansatte til å håndtere legemidler og delta i legemiddelbehandling av pasi- ent er en viktig forutsetning for pasientsikkerheten. Flere studier har i senere tid fokusert på lege- middelbehandling i sykehjem. Disse viser at gjennomsnittspa- sienten får forskrevet fra fem til 11,5 legemidler, avhengig av om både faste og behovslegemidler er inkludert og hvilken metode som er benyttet (4–7). Samtidig bruk av flere legemidler, såkalt poly- farmasi, øker risikoen for lege- middelrelaterte problemer (LRP), og tre av fire sykehjemspasienter har minst ett legemiddelrelatert problem (6). LRP defineres som «hendelser eller forhold som skjer i forbindelse med legemid- delbehandling og som faktisk eller potensielt interfererer med ønsket helseeffekt» (8, 9). De fleste studier gjort på dette feltet tar imidlertid for seg pasiente- nes faste legemidler, legemidler pasienten skal ta til forhåndsbe- stemte tidspunkter. Få studier har sett nærmere på forskrivning og reell bruk av legemidler som tas ved behov (også kalt «eventu- eltmedisin») (10–13). Behovsme- disinering representerer flere utfordringer for helsepersonell i sykehjem. For det første må de som observerer pasienten ha god kjennskap til sykdommer og symptomer og tilstrekkelig lege- middelkunnskap. Det kan dessu- ten forekomme forskrivninger av flere ulike behovslegemidler med samme indikasjon på legemiddel- kortene og de generelle skriftlige direktivene, samt uklar infor- masjon om dose og styrke (13). Dette legger et stort ansvar på sykepleier som tar beslutninger knyttet til utlevering. Samtidig bør pleierne kjenne pasientene godt for å kunne tolke pasien- tenes reelle behov, noe som kan være særlig utfordrende ved demens. Behovslegemidler deles • Medisinhåndtering • Kvalitet • Sykehjem • Legemidler • Dokumentasjon Nøkkelord Mer om forfatterne: S i W R i Mer om forfatterne: Stine Wang Rønningen har mas- ter i farmasi fra UiB og er ansatt ved Boots apotek, Kolbotn. Kjersti Bakken er førsteamanuensis ved institutt for samfunnsmedisin, Universitetet i Tromsø. Anne Gerd Granås er førsteamanuensis ved institutt for farmasi og bioingeniør- fag, Høgskolen i Oslo og Akershus. Kontakt: stiwaro@hotmail.com. Bakgrunn Legemidler er en viktig innsats- faktor i helsevesenet, og den sta- dig voksende gruppen av eldre står for størstedelen av legemid- delbruken (1). I de fleste vestlige land defineres gruppen eldre som personer som er 65 år eller eldre (2). I 2011 fikk ni av ti eldre her i landet ett eller flere legemidler på resept, og 57 prosent av de eldre legemiddelbrukerne fikk utlevert flere enn fem ulike lege- midler (1). Samme år ble det omsatt legemidler for nærmere 19 milliarder kroner. Økende legemiddelbruk medfører ikke bare kostnader, men også uhel- dige hendelser som kan gi ekstra liggedøgn på sykehus, føre til skade på pasient og i verste til- felle død (3). Det er derfor av stor betydning at legemiddelbehand- ling og -håndtering foregår på best mulig måte. Først og fremst av hensyn til pasientene det gjel- der, men også for samfunnet som helhet. En særlig utsatt gruppe er Noe for deg? nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning medications during 15 weeks, and observed the documen- tation of effects for these medicati- ons by collecting information from drug charts, the electronic patient record and medication administra- tion records for 108 patients in one nursing home in Bergen, Norway. We used SPSS to analyse the col- lected data (descriptive statistics). Method: We investigated the pre- valence of prescribed and adminis- tered P.R.N. medications during 15 weeks, and observed the documen- tation of effects for these medicati- ons by collecting information from drug charts, the electronic patient record and medication administra- tion records for 108 patients in one nursing home in Bergen, Norway. We used SPSS to analyse the col- lected data (descriptive statistics). Background: To ensure safe and appropriate use of medicine at all hours for nursing home patients, good routines for prescribing handling and documentation of effects for P.R.N. medication must be established. We have little know- ledge of how this is practiced today. Conclusion: There is a need for more regular review of the prescri- bed medication to change, stop or add P.R.N. medication so that the drug charts reflect the actual needs of the patients. Improved documen- tation of the effects of administered medication is a prerequisite for safe and tailormade drug treatment. Conclusion: There is a need for more regular review of the prescri- bed medication to change, stop or add P.R.N. medication so that the drug charts reflect the actual needs of the patients. Improved documen- tation of the effects of administered medication is a prerequisite for safe and tailormade drug treatment. Results: Within the study period each patient had on average four different of P.R.N. medications prescribed on their drug chart. Two thirds of the prescribed P.R.N. medication was never administered to the patients during the 15 weeks. Objective: To document prescription and administration of P.R.N. medi- cation in one nursing home, and to investigate how often, and how, the effects of P.R.N. medication are documented. Objective: To document prescription and administration of P.R.N. medi- cation in one nursing home, and to investigate how often, and how, the effects of P.R.N. medication are documented. Objective: To document prescription and administration of P.R.N. medi- cation in one nursing home, and to investigate how often, and how, the effects of P.R.N. medication are documented. Key Words: Medicine management, Quality, Nursing home, P.R.N. medi- cation, Documentation. 16 Resultater l Til sammen 108 pasienter, hvorav 60,2 prosent var kvinner, ble inkludert i analysene (tabell 1). Gjennomsnittsalderen var 84,5 år, og hver pasient hadde om lag fem diagnoser (median). Etikk og personvern j k ikk f Prosjektet var ikke fremleg- gingspliktig for REK, da de klassifiserte studien som et kva- litetssikringsprosjekt. Prosjektet ble meldt til Datatilsynet etter gjeldende regler og godkjent av Bergen kommune før det ble gjennomført. Materiale og metode Vi h f k l Vi har foretatt en kartlegging av forskrivning og bruk av behovs- legemidler, samt dokumentasjo- nen av disse legemidlenes effekter hos pasientene ved ett sykehjem. Vi benyttet pasientdata fra Løvå- sen undervisningssykehjem i Bergen. Sykehjemmet ble valgt av Bergen kommune i samar- beid med sykehjemsledelsen. Vi inkluderte journaldata fra 108 av totalt 125 inneliggende pasi- enter. 17 pasienter ble eksklu- dert da disse enten hadde vært på sykehjemmet i mindre enn 14 dager eller innlagt på korttidsav- delinger (n = 8), eller ikke hadde g God skriftlig og munt- lig kommunikasjon er av stor betydning ved de fleste aspek- ter av pasientbehandling. En klar og entydig forskrivning må danne utgangspunktet for all legemiddelbehandling. Enty- dig dokumentasjon av utdelte legemidler, og effekten av disse hos den enkelte pasient, er også viktig. Dette gjelder ikke minst ved behovsmedisinering. For å sikre effektiv og individuelt til- passet legemiddelbehandling må behandlingsapparatet få tilbake- melding på om legemidlene som Hva tilfører artikkelen? Artikkelforfatterne fant at dokumentasjonen av effekt av behovsmedisineringen ved syke- hjemmet, der undersøkelsen ble gjennomført, var mangelfull både når det gjaldt kvantitet og kvalitet. 17 ORIGINALARTIKKEL > Behovsmedisinering i sykehjem: Forskrivning, bruk og dokumentasjon av effekt nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning blir gitt har tiltenkt effekt eller ikke. Hvem som skal gjøre dette, eller hvordan, er ikke spesifisert ytterligere. Vi har ikke funnet forskning som omtaler effektdo- kumentasjon av behovslegemid- ler på sykehjem. Helsetilsynet har imidlertid gjennom flere til- synsrapporter påpekt mangelfull journalføring av legemiddelbe- handling, uvisshet om hvordan legemidlers virkninger og bivirk- ninger skal dokumenteres, samt uklarheter om hvor man kan finne viktige opplysninger i kom- fått forskrevet behovslegemidler i løpet av perioden (n = 9). ut til alle tider på døgnet, og da vil bemanning og kompetanse hos dem som er på jobb ha stor betydning for beslutningstaking om å gi behovslegemidler, for utlevering og for observasjon av effekt hos pasientene etter inntak. Ofte er mange personer involvert i behovsmedisinerin- gen; legen som har forskrevet legemidlet på legemiddelkortet eller i generelle skriftlige direk- tiv, helsefagarbeideren som observerer behovet for ekstra medisinering hos pasienten, Pasientenes journaler, i form av elektronisk pasientjournal (ved Løvåsen sykehjem brukes Geriatric Basis Dataset (GBD)) samt legemiddelkort og kvitte- ringsark for gitte legemidler, ble gjennomgått 15 uker tilbake i tid fra undersøkelsestidspunktet (1. oktober 2010 til 11. januar 2011). Følgende informasjon ble hentet inn: demografiske data, foreskrevne og utleverte behovslegemidler (inkludert preparatnavn, legemiddelsub- stans, styrke og dose), dato og tidspunkt for foreskrevne og utleverte behovslegemidler samt dokumentasjon av effekt (det vil si hva pleiepersonellet noterte og hvor dokumentasjonen var ned- tegnet). Både leger og pleiere har tilgang til å dokumentere i GBD. Nødvendige data ble lagt inn i SPSS versjon 18.0 for deskriptiv analyse. Legemidlene ble klas- sifisert etter ATC-systemet (17). En klar og entydig forskrivning må danne utgangspunktet for all legemiddelbehandling En klar og entydig forskrivning må danne utgangspunktet for all legemiddelbehandling binerte elektroniske og papir- baserte journaler (15). «Riktig legemiddelbruk i sykehjem» er da også valgt som ett av innsatsom- rådene i den pågående pasient- sikkerhetskampanjen «I trygge hender» (16). Hensikten med vår studie var å få kunnskap om for- hold rundt behovsmedisinering, ved å kartlegge forskrivning og utlevering av behovslegemidler ved ett utvalgt sykehjem, samt undersøke hvor ofte og hvordan effekten av behovslegemidlene blir dokumentert. sykepleieren som tar beslutnin- gene rundt utlevering, og helse- fagarbeider eller sykepleier som deler ut legemidlet og i etterkant skal observere effekten av lege- midlet. Vaktskifter kan bidra til enda flere involverte. Nødven- digheten av behovsmedisin til den enkelte pasient bør vurderes fortløpende av legen i samarbeid med sykepleier. Sykepleier skal vurdere når utdeling kan skje, og personell med fullmakt kan forestå utdelingen etter først å ha konferert med sykepleier. Det skal kvitteres på eget skjema ved utlevering av behovsmedisin, og effekten skal registreres (14). Hvordan? b For å beskrive effekten pasi- entene fikk av behovslegemid- lene benyttet personalet totalt 59 forskjellige ord og uttrykk. Disse kunne inndeles i fem ulike kategorier; god effekt, middels effekt, dårlig/ingen effekt, spe- sifikke beskrivelser og uviss/ tvetydig effektbeskrivelse (tabell 3). Ulike termer for «god effekt» utgjorde 71 prosent av det totale antallet utleveringer der effek- ten var dokumentert (N=285). For utleveringene der effekten var dokumentert to steder fant vi imidlertid seks tilfeller hvor betydningen av notatene ikke samsvarte mellom den elektro- niske pasientjournalen og kvit- teringsarkene. Mer spesifikke beskrivelser av effekt fantes i 5 prosent av utleveringene med effektdokumentasjon, ofte i for- bindelse med utdeling av soveme- disin og beroligende legemidler. b k l d Hvilke og hvor mye behovs- legemidler forskrives? Til de 108 pasientene var det totalt forskrevet 445 legemidler «ved behov» på legemiddelkor- tene, gjennomsnittlig 4,1 fore- 18 Tabell 1: Pasientkarakteristika n (%) Alder, gjennomsnitt (SD) Antall behovslegemidler, gjennomsnitt (SD) Antall diagnoser, median (variasjons- bredde) Menn 43 (39,8) 80,0 (9,4) 4,4 (2,5) 5 (2–11) Kvinner 65 (60,2) 87,5 (8,2) 3,9 (3,0) 6 (1–12) Totalt 108 (100) 84,5 (9,4) 4,1 (2,9) 5 (1–12) Tabell 2: Hyppigst forskrevne behovslegemidler og utleveringsfrekvens Forskrevet Utlevert ATC-nummer Legemiddel- substans Preparatnavn Antall ganger behovslege- midlene er for- skrevet, N=445 n Andel pasienter, N=108 % Antall ganger behovslege- midlene er gitt, N=839 n Andel pasient- er, N=108 % N02BE01 Paracetamol Paracet, Paracetamol, Pamol, Panodil, Pinex 67 52,8* 152 39,8 N05BA04 Oksazepam Alopam, Sobril 46 38,9* 82 23,1 N05CM02 Klometiazol Heminevrin 33 26,9* 75 13,9 N05BA01 Diazepam Valium, Stesolid, Vival 25 20,4* 120 6,5 N02AX02 Tramadol Tramadol, Nobligan, Tramagetic 23 21,3 20 10,2 N05CF01 Zopiklon Imovane, Zopiclone 21 19,4 55 7,4 C01DA02 Glyseroltrinitrat Nitroglycerin, Nitromex, Ni- trolingual, Minitran, Nitro-dur, Nitroven, Transiderm-nitro 19 17,6 11 4,6 N02AA01 Morfin Dolcontin, Morfin, Morfinsulfat 19 15,7* 54 4,6 N02AA05 Oksykodon Oxycodone, OxyContin, OxyNorm 16 13,9* 113 7,4 A03FA01 Metoklopramid Afipran 15 13,9 41 6,5 Andre 161 – 116 – Samme aktive substans er forskrevet flere ganger til samme pasient, i form av ulike preparatnavn eller ulike legemiddelformer, eller ved at legemi- dlet er startet, seponert og startet igjen i løpet av studieperioden. Samme aktive substans er forskrevet flere ganger til samme pasient, i form av ulike preparatnavn eller ulike legemiddelformer, eller ved at legemi- dlet er startet, seponert og startet igjen i løpet av studieperioden. 19 ORIGINALARTIKKEL > Behovsmedisinering i sykehjem: Forskrivning, bruk og dokumentasjon av effekt nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning skrevne behovslegemidler per pasient (SD=2,9) (tabell 1 og 2). Medianen var tre behovslegemid- ler per pasient (variasjonsbredde 1–19). På legemiddelkortene fantes det 57 ulike legemid- delsubstanser blant behovsfor- skrivningene. De ti hyppigst foreskrevne substansene finnes i tabell 2. Fra tabellen ser vi at enkelte legemiddelsubstanser er forskrevet til de samme pasien- flere detaljer benyttet perso- nalet henholdsvis 13 og 12 forskjellige ord/uttrykk i hver kategori. Mange av uttrykkene er tilnærmet like med omtrent samme betydning, men er for- mulert ulikt. Eksempler kan være «God» og «Bra», eller «Ingen effekt» og «Lite hjelp». Variasjonen var stor; fra null til hele 143 utleveringer per pasient over 15 uker. Oppfølging av legemiddelkortene d Et av de mest interessante fun- nene var at en såpass høy andel (70 prosent) av de foreskrevne legemidlene faktisk aldri ble utlevert til pasientene i løpet av de 15 ukene. Det var imidlertid stor variasjon mellom pasientene når det gjaldt antall utleveringer av behovslegemidler. Noen pasi- enter fikk aldri behovslegemid- del i løpet av perioden, mens én pasient fikk utlevert ulike behovslegemidler hele 143 gan- ger, hvilket i praksis innebærer flere ganger daglig. Samlet kan dette tyde på at legemiddelkor- tene revideres for sjelden. Jevnlig gjennomgang av legemiddelkor- tene, for å sikre at eksisterende legemiddelforskrivninger er rele- vante og adekvate, er nødvendig også når det gjelder behovslege- midler. Enkelte ganger må man vurdere om et behovslegemiddel i stedet bør oppføres som fast medisin. På den måten sikrer man pasienten jevnlig dosering av for eksempel smertestillende, slik at smertegjennombrudd kan unngås og medisineringen ikke blir tilfeldig og dermed mindre Dokumentasjon Hvor ofte? Det ble dokumentert effekt for 285 (34 prosent) av de totalt 839 utdelingene av behovslege- midler i perioden. I 58 tilfeller var det overlappende dokumen- tasjon, hvilket vil si at effekten Det ble oftere nedtegnet spe- sifikke og utfyllende beskrivel- ser av effekt i det elektroniske pasientjournalsystemet (GBD) sammenliknet med kvitterings- arkene, der effekten vanligvis var beskrevet med ett eller to ord, eksempelvis «God», «God effekt», «Liten», og «Ingen». 70 prosent av de foreskrevne legemidlene ble faktisk aldri utlevert til pasientene i løpet av de 15 ukene. tene flere ganger, for eksempel er paracetamol forskrevet 67 gan- ger til 57 pasienter. Dette betyr at samme behovslegemiddel enten er forskrevet flere ganger (startet, seponert og startet igjen) til samme pasient, eller at pasi- enten har fått oppført to ulike merkenavn med samme substans på legemiddelkortet, eksempel- vis både Pinex® og Paracet®. I enkelte tilfeller observerte vi også forskrivning av ulike legemiddel- former med samme aktive sub- stans, for eksempel både tabletter og stikkpiller. ble dokumentert både i elektro- nisk pasientjournal og på kvit- teringsarkene. Samlet sett fant vi at kvitteringsarkene ble hyp- pigst benyttet, 65,6 prosent av alle notater var gjort her. Hvor mye behovslegemidler utleveres? 135 (30 prosent) av de 445 behovslegemidlene som var for- skrevet på legemiddelkortene ble utlevert til pasientene i løpet av perioden på 15 uker. Med andre ord ble 310 (70 prosent) av de foreskrevne behovslege- midlene aldri benyttet i løpet av studieperioden. Samlet ble det utlevert behovslegemidler 839 ganger (tabell 2). Totalt fikk 79 (73 prosent) av de 108 pasientene utlevert ett eller flere behovsle- gemidler i perioden, og gjen- nomsnittlig antall utleveringer per pasient var 7,8 (median = 2). For å beskrive generelt god eller generelt dårlig effekt uten 20 Tabell 3: Kategorisering av effektdokumentasjon Gruppe 1 Gruppe 2 Gruppe 3 Gruppe 4 Gruppe 5 GOD EFFEKT n MIDDELS EFFEKT n DÅRLIG EFFEKT n SPESIFIKK BESKRIVELSE n UAVKLART n God 129 Middels 10 Ingen effekt 13 Effekt etter ca. 1,5 time 1 – 4 God effekt 69 Noe 4 Liten 5 Sovet godt i hele natt 1 Bra/ middels 1 Bra 14 Delvis 2 - effekt 5 Tilsynelatende god effekt. Oppleves for eksempel mindre rastløs og svingende i humøret 1 o 2 Bra effekt 11 Litt 3 Ingen 4 God i 3 timer 1 Usikker 1 Tydelig bedring 2 Noe effekt 5 Responderte ikke 1 Kastet opp og fikk frostanfall 1 Effekt bra 1 Delvis effekt 1 Lite hjelp 1 Pasienten sov frem til 2.30 1 God effekt 8 God - 1 Veldig liten effekt 1 Pasienten sov til ca. kl 3 1 «Uleselige» 2 Meget god effekt 1 Moderat effekt 1 Dårlig 2 Pasienten sovnet godt 1 God effekt 1 Noe et- terhvert 1 Overhodet ingen effekt 1 Roet seg 2 Effekt observert 1 Noenlunde bra effekt 1 Uten effekt 4 Roet seg litt etter 1 time 2 Tilsynela- tende god 2 Middels effekt 2 Pasienten sa det ikke hjalp 1 Fortsatt urolig 2 Bedring 1 Lite effekt 1 Sovnet 1 Effekt 1 Sovnet etter 1/2 time 1 Synkende blodsukker 1 Tok over 1 time før effekt 1 Pasienten roet seg ned 1 God effekt, lettere i pusten 1 Roet seg etter hvert 1 Somnolent og utslått som en kombinasjon av somatisk syk- dom/dehydrering og bivirkning av stesolid 1 TOTALT SUM 241 SUM 31 SUM 39 SUM 22 SUM 10 343 21 behovslegemidler ble utlevert, var effekt ikke dokumentert. I 66 prosent av tilfellene der behovslegemid- ler ble utlevert, var effekt ikke dokumentert. til tungpustethet eller angina. Da er det avgjørende at man vet om pasienten tidligere har respondert godt på det aktuelle legemidlet. Ved å notere effekten legger man grunnlag for nødven- dige endringer i medisineringen av den enkelte pasient og dermed individualisert behandling. Fordi mange personer er involvert ved behovsmedisinering er god skriftlig kommunikasjon ekstra viktig for å ivareta pasientsikker- heten. Vaktskifter understreker ytterligere behovet for nøyaktig og oppdatert skriftlig informa- sjon om pasientenes respons på gitte legemidler. til tungpustethet eller angina. Da er det avgjørende at man vet om pasienten tidligere har respondert godt på det aktuelle legemidlet. Ved å notere effekten legger man grunnlag for nødven- dige endringer i medisineringen av den enkelte pasient og dermed individualisert behandling. Fordi mange personer er involvert ved behovsmedisinering er god skriftlig kommunikasjon ekstra viktig for å ivareta pasientsikker- heten. Vaktskifter understreker ytterligere behovet for nøyaktig og oppdatert skriftlig informa- sjon om pasientenes respons på gitte legemidler. gemidler er oppført på legemid- delkortene kan dessuten være fordelaktig, framfor at pasien- tene får utlevert behovslegemid- ler fra såkalte generelle skriftlige direktiv (13). Dette er skriftlige tilleggsprosedyrer utarbeidet på hvert sykehjem, for generell ordi- nasjon av enkelte legemidler når legen ikke er til stede (14). Ordi- nasjonene her er dermed ikke individuelt tilpasset den enkelte pasient. Resultatene viste at de hyp- pigst utleverte legemidlene stort sett samsvarte med dem som var oftest forskrevet. Paracetamol var mest frekvent i bruk, etter- fulgt av diazepam. At diazepam kom så høyt opp var noe over- raskende, da diazepam anses som et uhensiktsmessig valg til eldre grunnet lang halveringstid og uheldige bivirkninger (22). Vi observerte imidlertid at dia- zepam kun ble gitt til et fåtall pasienter, og det er vanskelig å avgjøre hensiktsmessighet i enkelttilfeller med vårt begren- sete datagrunnlag. Hvor mye behovslegemidler utleveres? Tidligere studier har vist at mangel på tid, ansattes holdnin- ger, ledelsens fokus på legemid- delkunnskap hos de ansatte og opplæring innen dokumenta- sjonsarbeid, påvirker dokumen- tasjonshyppighet og -kvalitet (23–26). Ved ikke å dokumentere effekten av legemidler pasienten tar, kan pasientsikkerheten trues. Mange behovslegemidler gis ved akutte behov hos pasienten, det være seg alt fra sterke smerter effektiv (18). Samtidig påpeker forfattere av andre artikler en tendens til at legemidler blir stå- ende oppført på legemiddelkor- tene over lengre perioder, uten at de blir endret eller seponert i tråd med pasientens behov eller anbefalte retningslinjer (19–21). I mange tilfeller er det bedre at et legemiddel står oppført ved behov, for å hindre at pasienten må innta flere legemidler enn strengt nødvendig på fast basis. Et eksempel på dette kan være benzodiazepiner. At behovsle- være ulik for en og samme utde- ling. Det ville forenkle rutinene om dokumentasjonen konsekvent ble ført ett bestemt sted. Den største utfordringen når det gjel- der dokumentasjon av effekt var likevel, at det i 66 prosent av til- fellene hvor behovslegemidler er gitt, faktisk ikke er dokumentert om legemidlet har effekt eller ei. Hvordan forbedre dokumentasjonspraksis? k d f De fleste notater vedrørende lege- midlenes effekt fant vi på kvit- teringsarkene. Det var imidlertid interessant å se at det i 58 tilfeller ble dokumentert effekt av lege- midlet både på kvitteringsarket og i det elektroniske pasientjour- nalsystemet. Slik dobbeltføring gjør det vanskeligere å vite hvor man kan forvente å finne den aktuelle informasjonen. Det kan også, som vi så, bety at doku- mentasjonen de to stedene kan Ustrukturert terminologi b ff k Den observerte effekten av behovslegemidlene ble gjengitt på svært ulike måter. Nesten 60 forskjellige ord og uttrykk ble brukt til i hovedsak å beskrive god, middels, dårlig eller ingen effekt. Det var med andre ord liten eller ingen grad av standar- disering eller strukturering av innholdet i notatene. Dette kan åpne for ulike tolkninger av den beskrevne effekten av legemid- lene. Mer spesifikke og detal- jerte beskrivelser av legemidlenes effekter fant vi i kun 5 prosent av utleveringene av behovslegemid- del der effekten var dokumentert (N=285). I mange tilfeller er det nødvendig at effekten beskrives mer detaljrikt og spesifikt, for å kunne evaluere den individuelle legemiddelresponsen i ettertid. De mer spesifikke beskrivelsene fant vi oftest i den elektroniske pasientjournalen, antakelig grunnet begrenset plass på kvit- teringsarkene. I 66 prosent av tilfellene der behovslegemid- ler ble utlevert, var effekt ikke dokumentert. Hvordan forbedre dokumentasjonspraksis? k d f En tysk studie fra 2007 viste at antallet pleienotater økte etter implementeringen av et elektro- nisk system, og at kvaliteten på dokumentasjonen økte ved tre av fire sykehusavdelinger (27). En annen studie undersøkte hvor ofte legene leste pleiejour- nalene, og her oppga 75 prosent av legene at de oftere leste den elektroniske journalen sammen- Paracetamol var mest frekvent i bruk, etter- fulgt av diazepam. behovsmedisinering (10,11). At forhold knyttet til behovsmedi- sinering på Løvåsen sykehjem derfor kan være «bedre» enn ved andre sykehjem, er en rime- lig antakelse. Det gir oss sånn sett grunn til å frykte at rutiner for behovsmedisinering kan være dårligere i andre sykehjem. Det er rom for forbedring knyttet til behovsmedisineringen i syke- hjem. Våre resultater indikerer at forholdene, slik de er i dag, kan føre til at pasientene ikke får en tilpasset legemiddelbehandling og oppfølging av behovsmedi- sineringen på en adekvat måte. Behovslegemidler skal benyttes ved akutte behov hos pasientene, og det er av stor betydning at de får riktig medisin, i riktig dose, til rett tid. Dette kan man oppnå ved å kontinuerlig oppfølge pasi- entenes behovslegemidler basert på blant annet observasjon av legemidlenes effekt. I tråd med den pågående pasientsikkerhets- kampanjen vil dette være et steg i riktig retning mot bedre lege- middelbruk i sykehjem, og det er videre behov for flere og større studier på dette området. Takk til Løvåsen sykehjem, Sju- kehusapoteka Vest og Eli Tver- borgvik for godt samarbeid, og til Kjell Krüger og Magne Rekdal for god faglig og praktisk hjelp under datainnsamlingen. Et viktig incitament til at per- sonalet skal dokumentere effekt er at nytteverdien oppleves som stor. Dersom ikke notatene blir brukt i ettertid, blant annet til å foreta justeringer i pasientenes legemiddelbehandling, vil moti- vasjonen for å notere effekt anta- kelig være begrenset. Første steg i riktig retning er kanskje derfor å bevisstgjøre alle involverte slik at de i større grad benytter denne verdifulle informasjonen, for eksempel ved legemiddelgjen- nomganger. Slike systematiske legemiddelgjennomganger fører ofte til at legemidler seponeres (4, 32), og den høye andelen av behovslegemidler som ikke ble gitt i vår studieperiode under- støtter dette. dokumentasjonsnotater I 66 prosent av tilfellene der 22 liknet med den papirbaserte (28). En av årsakene legene oppga som viktig var at lesbarheten av notatene ble bedre ved bruk av et elektronisk system. Kanskje vil et fullelektronisk system kunne øke antallet dokumentasjoner og heve kvaliteten av disse? Enkelte mener dessuten at en struktu- rert elektronisk pasientjournal kan være et nyttig verktøy som underletter dokumentasjonen og anvendeligheten av denne (29,30). Ehrenberg og medar- beidere har konkludert med at ansatte som har fått opplæring innen strukturert dokumentasjon gjengir mer fullstendig og nøyak- tig informasjon (31). liknet med den papirbaserte (28). En av årsakene legene oppga som viktig var at lesbarheten av notatene ble bedre ved bruk av et elektronisk system. Kanskje vil et fullelektronisk system kunne øke antallet dokumentasjoner og heve kvaliteten av disse? Enkelte mener dessuten at en struktu- rert elektronisk pasientjournal kan være et nyttig verktøy som underletter dokumentasjonen og anvendeligheten av denne (29,30). Ehrenberg og medar- beidere har konkludert med at ansatte som har fått opplæring innen strukturert dokumentasjon gjengir mer fullstendig og nøyak- tig informasjon (31). opplæring og forskning, de har flere fast ansatte leger og et elek- tronisk pasientjournalsystem (GBD) som er velegnet til blant annet oppfølging av legemid- delbehandling. Studier har vist at sykehjemskultur i seg selv kan være avgjørende for blant annet organisering og bruk av tilpasset legemiddelbehandling. Vi fant at dokumentasjonen av effekt av behovsmedisineringen var mangelfull, både når det gjaldt kvantitet og kvalitet. God journalføring av effekt er med på å legge grunnlaget for en bedre legemiddelbehandling av pasien- tene. Referanser: 1. Folkehelseinstituttet. Legemid- delstatistikk 2012:2. Reseptregistret 2007–2011.i 2. WHO. Definition of an older or elderly person. Tilgjengelig fra: http://www.who. int/healthinfo/survey/ageingdefnolder/ en/. (Nedlastet 03.11.2012.) 2. WHO. Definition of an older or elderly person. Tilgjengelig fra: http://www.who. int/healthinfo/survey/ageingdefnolder/ en/. (Nedlastet 03.11.2012.) 3. Hjort PF. Uheldige hendelser i helse- tjenesten: en lære-, tenke- og faktabok. Gyldendal akademisk, Oslo. 2007. 4. Halvorsen KH, Ruths S, Granås AG, Viktil KK. Multidisciplinary intervention to identify and resolve drug-related problems in Norwegian nursing homes. Scand J Prim Health Care. 2010;28:82–8. 5. Rytter E, Nakken KO, Mørch-Reiersen LT, Efjestad A, Selvig K. Bruk av antiepi- leptika hos sykehjemsbeboere. Tidsskr Nor Lægeforen. 2007;127:1185–7. 6. Ruths S, Straand J, Nygaard HA. Mul- tidisciplinary medication review in nur- sing home residents: what are the most significant drug-related problems? The Bergen District Nursing Home (BED- NURS) study. Qual Saf Health Care. 2003;12:176–80. Avslutningsvis er det viktig å understreke at resultatene fra denne studien er basert på data fra ett enkelt sykehjem, som i tillegg er et undervisningssyke- hjem. Vi kan derfor ikke anta at funnene er representative for norske sykehjem generelt. Data- grunnlaget er begrenset, og vi har ikke hatt tilgang til karak- teristika for pasientene som ble ekskludert fra studien. Løvåsen sykehjem har ekstra fokus på 7. Kirkevold Ø, Engedal K. Is Covert Med- ication in Norwegian Nursing Homes Still a Problem?: A Cross-Sectional Study. Drug Aging. 2009;26:333–44. 8. Pharmaceutical Care Network Europe. PCNE Classification for Drug Related Problems V6.2. Tilgjengelig fra: http://www.pcne.org/sig/drp/docu- ments/drp/PCNE%20classification%20 V6-2.pdf. (Nedlastet 06.10.2010)i 10. Stokes JA, Purdie DM, Roberts MS. Konklusjon S di i Studien viser at det er behov for jevnlig gjennomgang og even- tuell revidering av foreskrevne behovslegemidler til pasienter på sykehjem for å sikre pasi- entene rasjonell og individuelt Studien viser at det er behov for jevnlig gjennomgang og even- tuell revidering av foreskrevne behovslegemidler til pasienter på sykehjem for å sikre pasi- entene rasjonell og individuelt 9. Ruths S, Viktil KK, Blix HS. Klassifise- ring av legemiddelrelaterte problemer. Tidsskr Nor Lægeforen. 2007;127:3073– 6. 10. Stokes JA, Purdie DM, Roberts MS. 23 ORIGINALARTIKKEL > Behovsmedisinering i sykehjem: Forskrivning, bruk og dokumentasjon av effekt nr 1, 2013; 8: 14-24 doi: 10.4220/sykepleienf.2013.0005 forskning tions/atcddd/en/. (Nedlastet 26.05.2012) 18. Lellan KM. A chart audit reviewing the prescription and administration trends of analgesia and the documenta- tion of pain, after surgery. J Adv Nurs. 1997;26:345–50. tions/atcddd/en/. (Nedlastet 26.05.2012) 18. Lellan KM. A chart audit reviewing the prescription and administration trends of analgesia and the documenta- tion of pain, after surgery. J Adv Nurs. 1997;26:345–50. Factors influencing PRN medication use in nursing homes. Pharm World Sci. 2004;26:148–54. 26. Björvell C, Wredling R, Thorell- Ekstrand I. Long-term increase in qua- lity of nursing documentation: effects of a comprehensive intervention. Scand J Caring Sci. 2002;16:34–42. 11. Roberts MS, King M, Stokes JA, Lynne TA, Bonner CJ, McCarthy S, et al. Medication prescribing and admi- nistration in nursing homes. Age Ageing. 1998;27:385–92. 27. Mahler C, Ammenwerth E, Wagner A, Tautz A, Happek T, Hoppe B, et al. Effects of a computer-based nursing documentation system on the quality of nursing documentation. J Med Syst. 2007;31:274–82. 19. Nygaard HA. Varighet av legemiddel- bruk i sykehjem. Tidsskr Nor Lægeforen. 2001;121:1469–72. 12. Curtis J, Capp K. Administration of ‘as needed’ psychotropic medication: A retrospective study. Int J Ment Health Nurs. 2003;12:229–34. 20. Avorn J, Gurwitz JH. Drug Use in the Nursing Home. Ann Intern Med. 1995;123:195–204. 28. Ammenwerth E, Eichstadter R, Haux R, Pohl U, Rebel S, Ziegler S. A rando- mized evaluation of a computer-based nursing documentation system. Methods Inf Med. 2001;40:61–8. 13. Andreassen LM, Halvorsen KH, Granås AG. Store ulikskapar i innhold, å 13. Andreassen LM, Halvorsen KH, Granås AG. Store ulikskapar i innhold, utforming og kvalitet på dei generelle skriftlige skriftlege direktivene. Syke- pleien forskning. 2011;6:46–52. 21. Ruths S, Straand J, Nygaard HA. Psy- chotropic drug use in nursing homes – diagnostic indications and variations between institutions. Eur J Clin Phar- macol. 2001;57:523–8. 13. Andreassen LM, Halvorsen KH, Granås AG. Konklusjon S di i Store ulikskapar i innhold, utforming og kvalitet på dei generelle skriftlige skriftlege direktivene. Syke- pleien forskning. 2011;6:46–52. 14. Bergen kommune. Retningslinjer for legemiddelhåndtering i institusjon i Ber- gen kommune. Versjon 2.1. Bergen 2010. 15. Helsetilsynet. Sårbare pasienter – utrygg tilrettelegging Tilgjengelig fra: http://www.helsetilsynet.no/upload/ Publikasjoner/rapporter2010/helse- tilsynetrapport7_2010.pdf. (Nedlastet 16.01.2011) 16 N j l h f i ikk 29. Krüger K. Elektroniske pasientjour- naler bør være strukturerte. Tidsskr Nor Lægeforen. 2007;127:2090–3. 14. Bergen kommune. Retningslinjer for legemiddelhåndtering i institusjon i Ber- gen kommune. Versjon 2.1. Bergen 2010. 15. Helsetilsynet. Sårbare pasienter – utrygg tilrettelegging Tilgjengelig fra: http://www.helsetilsynet.no/upload/ Publikasjoner/rapporter2010/helse- tilsynetrapport7_2010.pdf. (Nedlastet 16.01.2011) 22. Ruths S, Straand J. Legemidler og eldre. I: Norsk Legemiddelhåndbok for helsepersonell. Foreningen for utgivelse av Norsk legemiddelhåndbok, Oslo. 2010. (s. 1785–90). 30. Sahlstedt S, Adolfsson H, Ehnfors M, Kallstrom B. Nursing process docu- mentation – effects on workload and quality when using a computer program and a key word model for nursing docu- mentation. Stud Health Technol Inform. 1997;46:330–6. 23. Sjöberg I, Öhman C. Faktorer som påverkar sjuksköterskors omvårdnads- dokumentation: Luleå tekniska Univer- sitet; 2009. 16. Nasjonal enhet for pasientsikker- het. Pasientsikkerhetskampanjen I trygge hender 2011–2013. Tilgjengelig fra: http://www.pasientsikkerhetskam- panjen.no/. (Nedlastet 08.05.2012) 17. WHO. The Anatomical Therapeutic Chemical Classification System with Defined Daily Doses (ATC/DDD). Tilgjen- gelig fra: http://www.who.int/classifica- 31. Ehrenberg A, Ehnfors M. The accuracy of patient records in Swe- dish nursing homes: congruence of record content and nurses’ and pati- ents’ descriptions. Scand J Caring Sci. 2001;15:303–10. 16. Nasjonal enhet for pasientsikker- het. Pasientsikkerhetskampanjen I trygge hender 2011–2013. Tilgjengelig fra: http://www.pasientsikkerhetskam- panjen.no/. (Nedlastet 08.05.2012) 24. Törnvall EVA, Wahren LK, Wilhelms- son S. Impact of primary care manage- ment on nursing documentation. J Nurs Manage. 2007;15:634–42. 25. Hansebo G, Kihlgren M, Ljunggren G. Review of nursing documentation in nursing home wards–changes after intervention for individualized care. J Adv Nurs. 1999;29:1462–73. 32. Kersten H, Ruths S, Wyller TB. Farmakoterapi i sykehjem. Tidsskr Nor Lægeforen. 2009;129:1732–5. Les kommentaren på side 25 Les kommentaren på side 25 24
https://openalex.org/W4243567404	https://europepmc.org/articles/pmc8481305?pdf=render	English	null	Tanshinone IIA Affects the Malignant Growth of Cholangiocarcinoma Cells by Inhibiting the PI3K-Akt-mTOR Pathway	Research Square (Research Square)	2,021	cc-by	6,508	Tanshinone IIA affects the malignant growth of Cholangiocarcinoma cells by inhibiting the PI3K‑Akt‑mTOR pathway In the present study, we aimed to find the target of Tanshinone IIA (Tan-IIA) in Cholangiocarcinoma by network pharmacology-based prediction and investigate the possible mechanism through experimental verification. In this study, we combined Tan-IIA-specific and Cholangiocarcinoma-specific targets with protein–protein interactions (PPI) to construct a Tan-IIA targets-Cholangiocarcinoma network, and network pharmacology approach was applied to identify potential targets and mechanisms of Tan-IIA in the treatment of Cholangiocarcinoma. The anti-cancer effects of Tan- IIA were investigated by using subcutaneous tumorigenic model in nude mice and in the human Cholangiocarcinoma cell lines in vitro. Our results showed that Tan-IIA treatment considerably suppressed the proliferation and migration of Cholangiocarcinoma cells while inducing apoptosis of Cholangiocarcinoma cells. Western blot results demonstrated that the expression of PI3K, p-Akt, p-mTOR, and mTOR were inhibited by Tan-IIA. Meanwhile, After treatment with Tan-IIA, the level of Bcl2 was downregulated and cleaved caspase-3 expression increased. Further studies revealed that the anticancer effects of Tan-IIA were severely mitigated by pretreatment with a PI3K agonist. Our research provides a new anticancer strategy and strengthens support for the use of Tan-IIA as an anticancer drug for the treatment of CCA. Cholangiocarcinoma (CCA) is an aggressive malignant tumor of the biliary tract that is often challenging to diag- nose and treat1. It is the second most common primary malignancy, and its incidence has increased significantly in recent decades2,3. Although our understanding of the molecular biology of CCA has greatly increased and treatment options for CCA have improved in the last few decades, the prognosis of CCA is poor. Most patients have advanced disease and recurrence after resection4. Thus, increasing attention has been focused on traditional Chinese herbal medicines to explore new treatments for patients with CCA. Danshen (Salvia miltiorrhiza) is a traditional Chinese medicine. Tanshinone IIA (Tan-IIA) is extracted from Danshen and has been reported to exhibit a number of pharmacological activities against cancer5,6. Numerous studies have shown that the antican- cer potential of Tan-IIA has also been implicated including hepatocellular carcinoma7 and pancreatic cancer8,9. However, the specific mechanism of the anti-tumor effect of Tan-IIA on CCA cells remains to be elucidated. if Cyberpharmacology is a relatively new discipline that is based on the theory of systems biology, which allows for a more accurate study of the relevant therapeutic targets of a drug or human disease10. www.nature.com/scientificreports www.nature.com/scientificreports Scientific Reports \| (2021) 11:19268 Materials and methodsh Ethical approval. The protocols of all animal experiments were reviewed and approved by the Research Ethics Committee of Qingdao Municipal Hospital (The ethic approval ID: 028), and all animal experimental studies were conducted in accordance with the ARRIVE guidelines. All experiments in the text were carried out in compliance with the relevant rules and regulations and under the supervision and guidance of the Ethics Committee of Qingdao Municipal Hospital. Chemicals and reagents. Tan-IIA (Sigma-Aldrich, purity > 97%). The PI3K agonist 740y-p was purchased from MCE (Shanghai, China), Annexin V-FITC/PI staining kit (absin, abs50001, China), Cell Counting Kit-8 (CCK-8) reagent (APExBIO, K1018, USA), Matrigel glue (BD Biosciences, NJ, USA), and BCA protein anal- ysis kit (Beyotime, Shanghai, China), and ECL reagent (Millipore, Massachusetts, USA). Primary antibodies included anti-PI3K (20584-1-AP, Proteintech, China) and anti-Akt antibody (10176-2-AP, Proteintech, China), anti-mTOR (2983, CST, USA), anti-p-Akt (4060, CST, USA), anti-p-mTOR (5536, CST, USA), anti-Bax (2774, CST, USA), anti- Bcl2 (3498, CST, USA), and anti-Caspase-3 (9662, CST, USA) and anti-Cleaved-Caspase-3 (9661, CST, USA). Cell lines and culture conditions. Human Cholangiocarcinoma cell lines, HuCCT-1 and RBE, were pur- chased from the Chinese Academy of Sciences (Shanghai) Cell Bank. The cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM, Hyclone, USA) containing 10% fetal bovine serum (FBS, Excellbio, USA) and 1% penicillin–streptomycin (HyClone, UT, USA). The cells were cultured at 37 °C in an incubator containing 5% CO2. Cell viability. The cells were incubated overnight in 96 well plates at a density of 5 × 103 cells per well. Then the cells were treated with different concentrations (0, 5, 10, 20, and 30 µg/mL) of Tan-IIA at different times (12, 24, 48, and 72 h), and a 10 µL CCK8 reagent was added (APExBIO, K1018, USA) to each well. After incubation at 37 °C for 1 h, the optical density (OD) was measured at 450 nm. IC50 was calculated using GraphPad Prism 7.0 software. Plate cloning. Both cell lines were cultured in a well plate at a density of 300 cells per well. The cells were gently rotated to disperse the cells evenly. After 6 h, the Cholangiocarcinoma cells were treated with Tan-IIA (0, 5, 10, 20, and 30 µg/mL) and incubated in a cell culture incubator at 37 °C with 5% CO2 for two weeks. Next, the cells were washed with PBS three times. Subsequently, 4% paraformaldehyde was used to fix the cells for 15 min. Materials and methodsh The cells were then stained with crystal violet for 10 min. Subsequently, the staining solution was washed off with PBS. The six well plates were then inverted and an overlay of transparency sheet with a grid was performed and the clones was manually counted directly with the naked eye: clone formation rate = (number of clones/number of inoculated cells) × 100%. Scratch‑induced wound healing assay. Plated in six well culture dishes were 4 × 105 of Cholangiocar- cinoma cells. After 24 h, a 200 μL tip was used to wound confluent cells. The detached cells were washed with PBS three times, and the cells were incubated with Tan-IIA (HuCCT-1: 27 µg/mL, RBE: 49 µg/mL) for 24 h. Cell migration images were recorded using an inverted microscope. The results of the scratch experiment were obtained using the formula wound closure rate = post-healing area/initial wound area. All experiments were repeated three times. Transwell invasion assay. Cell invasion was detected in a 24-well Transwell chamber. The upper chamber was precoated with 50 µL Matrigel. Cholangiocarcinoma cells (5 × 104) were suspended in 100 µL serum-free medium supplemented with or without Tan-IIA (HuCCT-1: 27 µg/mL, RBE: 49 µg/mL) and were seeded into the upper chamber. At the same time, 600 µL DMEM containing 10% FBS was placed into the lower chamber. After 24 h incubation, Cholangiocarcinoma cells on the upper side of the membrane were wiped with a clean swab. Then, the cells on the underside of the membrane were fixed with methanol and stained with crystalline violet. Nine regions were randomly selected to count the number of invading cells using an inverted microscope. Apoptosis determined by the Annexin‑V‑FITC/PI. Cholangiocarcinoma cells were cultured in six well plates at a density of 2 × 105 cells per well. The cells were then treated with or without Tan-IIA (HuCCT-1: 27 µg/ mL, RBE: 49 µg/mL) for 24 h. Then, cells were digested by trypsin without EDTA and harvested and then washed twice with cold PBS. Subsequently, the cells were suspended in binding buffer and stained with 5 μL Annexin V-FITC for 30 min. Next, PI (10 µL) was added for 15 min at room temperature in darkness. Finally, the samples were analyzed by Accuri C6 flow cytometry (BD Biosciences, CA, USA). Identification of common targets for Tan‑IIA and Cholangiocarcinoma. Tanshinone IIA affects the malignant growth of Cholangiocarcinoma cells by inhibiting the PI3K‑Akt‑mTOR pathway In this study, a network pharmacological analysis identified common targets for Tan-IIA and Cholangiocarcinoma disease, and KEGG and GO analyses revealed that these common targets were enriched in pathways associated with PI3K-Akt. It is well known that in human cancers, phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway is the most common aberrant kinase cascade signaling pathway that promotes tumor cell proliferation through the activation of growth factor receptors such as insulin-like growth 1Department of Medicine, Qingdao University, Qingdao, China. 2Department of Hepatobiliary Surgery, The Affiliated Qingdao Municipal Hospital of Qingdao University, Qingdao, China. 3Department of Physiology, School of Basic Medicine, Qingdao University, Qingdao, China. 4Graduate School of Dalian Medical University, Dalian, China. 5Department of Gynecology, The Affiliated Qingdao Municipal Hospital of Qingdao University, Qingdao, China. 6These authors contributed equally: Huayuan Liu and Caiyun Liu. email: sgjzp@hotmail.com \| https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 www.nature.com/scientificreports/ factor 1 receptor (IGF1R), vascular endothelial growth factor receptor (VEGFR), and epidermal growth factor receptor (EGFR)11. Therefore, we selected the PI3K/Akt/mTOR signaling pathway for further study. The present study suggested that Tan-IIA inhibited the proliferation, invasion, and migration of Cholangiocarcinoma cells by inhibiting the PI3K/Akt/mTOR pathway. Furthermore, Tan-IIA upregulated the level of cleaved caspase-3 and suppressed the expression of Bcl2, which ultimately induced apoptosis of Cholangiocarcinoma cells. Materials and methodsh The targets of Tan-IIA were identified using the TCMSP database (https://tcmspw.com/tcmsp.php), and then gene annotation of the Scientific Reports \| (2021) 11:19268 \| https://doi.org/10.1038/s41598-021-98948-z www.nature.com/scientificreports/ TCM targets was conducted on the Uniprot website (https://www.uniprot.org/). From OMIM (https://omim. org/), GeneCards (https://www.genecards.org/), PharmGKB (https://www.pharmgkb.org/), and TTD (http://db. idrblab.net/ttd/) databases were used to obtain relevant targets of Cholangiocarcinoma, and the VennDiagram package12 in R software (version number: 4. 0. 0) was used to analyze the common target genes of both. TCM targets was conducted on the Uniprot website (https://www.uniprot.org/). From OMIM (https://omim. org/), GeneCards (https://www.genecards.org/), PharmGKB (https://www.pharmgkb.org/), and TTD (http://db. idrblab.net/ttd/) databases were used to obtain relevant targets of Cholangiocarcinoma, and the VennDiagram package12 in R software (version number: 4. 0. 0) was used to analyze the common target genes of both. Protein–protein interaction (PPI) network. For building a protein–protein interaction (PPI) network, the common targets of Tan-IIA and Cholangiocarcinoma were entered into the string website (https://string-db. org/) to construct PPI plots, and then Cytoscape software was used to construct network pharmacograms. GO and KEGG pathway analysis. Gene Ontology (GO) is an international standard classification system for gene function, consisting of three major components: cellular components, molecular function, and biologi- cal processes13. To investigate the gene functions involved in common targets, we ran GO functional enrichment using the clusterProfiler package14 in R software (version number: 4. 0. 0). The analysis of common target- associated pathways between Tan-IIA and Cholangiocarcinoma was performed using Kyoto Encyclopedia of Genes and Genome (KEGG)15,16. Western blotting. Cholangiocarcinoma cells were lysed in RIPA for 30 min at 4 °C and total protein was extracted by centrifugation for 20 min. Then, BCA protein analysis reagent was used to determine protein con- centrations. Protein extracts were boiled for 5 min. After resolution using SDS/PAGE (10%), proteins were trans- ferred to PVDF membranes (Millipore, Bedford, MA, USA). At room temperature, the membranes were sealed with 5% bovine serum albumin (BSA) for 2 h. Diluted primary antibody: Akt (1: 1000), phospho (p)-Akt (1: 2000), PI3K (1:1000), mTOR (1: 1000), phospho (p)-mTOR (1: 2000) (CST, CA, USA), and β-actin(1:5000). Subsequently, the target membranes were incubated with the specific primary antibodies overnight at 4 °C. Membranes were then washed with TBST three times and then diluted HRP-coupled secondary antibodies were added and incubated for 2 h at room temperature. Finally, immunocomplexes were detected using an ECL detec- tion reagent (Millipore, MA, USA). Materials and methodsh The measurement dates were obtained from three separate experiments. The intensity of each band was determined using ImageJ software. In vivo experiments. NOD-SCID (NOD CB17-Prkdcscid/NcrCrl, male, 5 weeks of age) mice were pro- vided by Beijing Vital River Lab-oratory Animal Technology Co., Ltd (Beijing, China). All mice were placed in a 12-h light/dark cycle at 25 + −1 °C and 56% humidity with free access to food and water. The initial body weights of these mice ranged from 20 to 23 g. Following subcutaneous injection of 2 × 106 HuCCT-1 Cholangiocarcinoma cells into the back of 15 NOD-SCID mice, the mice were divided into three groups: the control group (n = 5), the Tan-IIA (50 mg/kg) treatment group (n = 5), and the Tan-IIA (50 mg/kg) combined with 740y-p (10 mg/kg) treatment group. Tan-IIA was diluted with DMSO: Methanol: Hydroxypropyl-β-cydodextrin (HP-β-CD) = 1: 1: 1. 740y-p was dissolved in the same way. Seven days after the injection of HuCCT-1 Cholangiocarcinoma cells, drugs were injected intraperitoneally into the experimental groups of mice on every other day, and normal saline was given to the control group. Mice were killed at day 21 of inoculation with tumor cells. All mice were executed by dislocation of the cervical vertebrae. Tumor volumes were measured every 3 days before execution. Statistical analysis. The statistical software GraphPad 7 was used for data analysis. The experimental results listed in the article represent the dates of at least three separate replicate experiments. Dates are shown as mean and standard deviation. Student’s t-test was used for differences between two groups. Multiple group comparisons were made using one-way ANOVA. Differences were considered statistically significant at P values of less than 0.05. Resultsf Effects of Tan‑IIA on Cholangiocarcinoma cells proliferation, migration, colony formation, and invasion. To explore the effect of Tan-IIA on Cholangiocarcinoma cells, we incubated Cholangiocarcinoma cells at various concentrations (0, 5, 10, 20, and 30 µg/mL) of Tan-IIA for 12, 24, 48, and 72 h. The effect of Tan- IIA on the proliferation of Cholangiocarcinoma cells was then detected by CCK8 (Fig. 1A,B). This indicated that compared with the control group, Tan-IIA inhibited the proliferation of Cholangiocarcinoma cells in a time- and dose-dependent manner (Fig. 1A,B). At the same time, the plate cloning experiment demonstrated that Tan-IIA significantly suppressed Cholangiocarcinoma cells growth compared with the control group (Fig. 1C–F). Scratch wound assay showed that Tan-IIA treatment significantly suppressed the motility of Cholangiocarcinoma cells, as determined by the migration area (Fig. 1G–J). We further assessed the role of Tan-IIA in invasion by a tran- swell assay. The results showed that Tan-IIA significantly decreased the invasion capacity of Cholangiocarci- noma cells (Fig. 1K–N). In addition, the apoptotic rate detected by flow cytometry analysis showed that Tan-IIA could promote the apoptosis of Cholangiocarcinoma cells (Fig. 1O–R). Next, western blot analysis revealed that compared with the control group, the expression of Bcl-2 was significantly decreased. However, the protein level of BAX was not altered by Tan- IIA. In addition, the results showed that the expression of caspase-3 was mark- edly decreased, while the cleaved caspase-3 was markedly upregulated induced by Tan-IIA (Fig. 1S–U). Common targets of Tan‑IIA and Cholangiocarcinoma effects. Relevant targets for Cholangiocarci- noma were obtained from OMIM, GeneCards, PharmGKB, and TTD databases, and the target of Tan-IIA was found from the TCMSP database. After cross-analysis, 17 common drug-disease-related targets were identi- fied (Fig. 2A). The screened targets were then used to construct the PPI network through the string website https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 \| www.nature.com/scientificreports/ Figure 1. Cytotoxic effect of Tan-IIA on Cholangiocarcinoma cells. (A, B) Different concentrations (0, 5, 10, 20, and 30 µg/ml) of Tan-IIA were co-cultured with Cholangiocarcinoma cells for 12, 24, 48, and 72 h, and then the cytotoxic effect of Tan-IIA on Cholangiocarcinoma cells were detected by the CCK8 method. Cell viability histogram was made according to the formula (%) = [(experimental group OD value) − (blank group OD value)]/[(control group OD value) − (blank group OD value)] × 100%, and IC50 values were calculated for each time period. Resultsf (C–F) Plate clone formation experiments were performed, Cholangiocarcinoma cells were co-cultured with Tan-IIA at concentrations of (0, 5, 10, 20, and 30 µg/mL) for two weeks and the number of clones was counted. Tan-IIA inhibits the migration and invasion of Cholangiocarcinoma cells. (G–J) Tan-IIA (24 h IC50 concentration) was co-cultured with Cholangiocarcinoma cells, then after scratching the cell layer for 24 h, cell migration by wound healing assay was determined. (K–N) The effect of Tan-IIA on the invasive ability of Cholangiocarcinoma cells was measured by the Transwell method. After crystalline violet staining, cell images were taken with an inverted microscope. Tan-IIA induced apoptosis of Cholangiocarcinoma cells. (O–R) The IC50 concentration for 24 h of Tan-IIA was selected and co-cultured with Cholangiocarcinoma cells for 24 h. Annexin V-FITC/PI double staining was used to detect apoptosis. (S–U) Western blotting to detect the effect of Tan-IIA on the expression of Bax, Bcl-2, as well as caspase-3 and cleaved caspase-3. Compared with control, p < 0.05; p < 0.01; p < 0.001; **p < 0.0001. https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 \| /scientificreports/ Figure 2. Identifying common targets for Tan-IIA and Cholangiocarcinoma. (A) Venn diagrams were created by using the VennDiagram package12 in R software (version number: 4. 0. 0), depicting the intersection of Tan-IIA and Cholangiocarcinoma targets. (B) PPI network was constructed using 17 common targets of Cholangiocarcinoma and Tan-IIA. (C) Cytoscape plotted the network pharmacology, using red and yellow lines to indicate the interactions between Tan-IIA and targets and target-to-target interactions. GO and KEGG analysis of common targets. (D) GO functional enrichment was ran by using the clusterProfiler package14 in R software (version number: 4. 0. 0) and GO analysis of the enriched biological functions of the common target genes was performed (counts ≥ 30). (E) KEGG15,16 analysis of signaling pathways enriched by common target genes (counts ≥ 30). (F) Map of PI3K/AKT pathway obtained from KEGG analysis. www.nature.com/scientificreports/ Figure 2. Identifying common targets for Tan-IIA and Cholangiocarcinoma. (A) Venn diagrams were created by using the VennDiagram package12 in R software (version number: 4. 0. 0), depicting the intersection of Tan-IIA and Cholangiocarcinoma targets. (B) PPI network was constructed using 17 common targets of Cholangiocarcinoma and Tan-IIA. (C) Cytoscape plotted the network pharmacology, using red and yellow lines to indicate the interactions between Tan-IIA and targets and target-to-target interactions. GO and KEGG analysis of common targets. Discussion Tan-IIA was one of the main components of Danshen. Accumulated evidence from preclinical and clinical studies has confirmed that Tan-IIA has good anti-tumor properties17. Several studies have shown that it can inhibit the growth of various tumor cell lines, including liver cancer, pancreatic cancer, and colorectal cancer7,8,18. Neverthe- less, the inhibitory effect of Tan-IIA on CCA cells and its underlying mechanisms are unknown. Consistent with the previous findings, this study showed that Tan-IIA could inhibit malignant growth, migration, and invasion and promote apoptosis of Cholangiocarcinoma cells, providing a new understanding of the role of Tan-IIA in the treatment of CCA. To investigate the mechanism by which Tan-IIA inhibits Cholangiocarcinoma, we analyzed the possible pathways that Tan-IIA inhibited the malignant proliferation of Cholangiocarcinoma cells through network pharmacology. KEGG analysis showed that Tan-IIA affected the expression of the PI3K/Akt pathway in Cholan- giocarcinoma cells. Several studies have shown that elevated expression of PI3K- associated proteins is considered a hallmark of cancer19. The PI3K/Akt pathway is closely related to cancer progression in many types of human cancers, including lung cancer, stomach cancer, liver cancer, and pancreatic cancer. Previous studies have shown that the PI3K/Akt pathway plays a significant role in inhibiting tumor proliferation, invasion, migration, and apoptosis20. It has been reported that the PI3K/Akt signaling pathway is extremely important in CCA develop- ment and progression21. In the present study, we detected the expression of PI3K/Akt proteins in Cholangio- carcinoma cells. Consistent with KEGG analysis, we observed that Tan-IIA significantly inhibited the levels of PI3K and p-Akt compared with those in the control group in Cholangiocarcinoma cells. p p g p g Recent studies have shown that the activation of the PI3K/Akt pathway can activate or inhibit a variety of downstream target proteins, such as mTOR, Bad, Caspase9 and GSK-3. And the PI3K/Akt/mTOR signaling cas- cade has a major impact on cell proliferation and survival as well as cell cycle regulation22,23. In this pathway, PI3K is activated and further activates Akt proteins located on the plasma membrane, which eventually phosphorylate Akt, and then p-Akt activates various regulators downstream of the pathway, including mTOR24. mTOR is the main downstream effector of the PI3K/Akt pathway. Activated mTOR is associated with cell proliferation and survival25. Resultsf (D) GO functional enrichment was ran by using the clusterProfiler package14 in R software (version number: 4. 0. 0) and GO analysis of the enriched biological functions of the common target genes was performed (counts ≥ 30). (E) KEGG15,16 analysis of signaling pathways enriched by common target genes (counts ≥ 30). (F) Map of PI3K/AKT pathway obtained from KEGG analysis. (Fig. 2B). Using the data tables obtained from the PPI network, the network pharmacology map was entered into Cytoscape software (Fig. 2C). The relationship between the common target genes and the relationship between the genes and Tan-IIA can be clearly seen in the figure. Then, we analyzed the GO and KEGG results of the common targets of Tan-IIA and Cholangiocarcinoma. Based on the 17 common targets of Tan-IIA and Cholan- giocarcinoma, we used R software to analyze the top ten terms of the three major categories of BP, CC, and MF enriched by GO (Fig. 2D), and the results of GO analysis showed that the common targets of Tan- IIA and Chol- angiocarcinoma were closely related to cell proliferation. Among the top 30 terms in KEGG analysis (Fig. 2E), we found that the PI3K-AKT pathway was more significantly associated with proliferation (Fig. 2F). https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 \| www.nature.com/scientificreports/ Tan‑IIA inhibited activation of PI3K/AKT/mTOR. Western blotting was used to detect the expression of PI3K, Akt, p-Akt, mTOR, and p-mTOR. The results indicated that Tan-IIA inhibited the expression of PI3K, p-Akt, p-mTOR, and mTOR compared to control group in a dose-dependent manner (Fig. 3A–C). Furthermore, pretreatment of Cholangiocarcinoma cells with the 740 y-p (PI3K agonist) abolished the effects of Tan-IIA on the restrain PI3K, p-Akt, mTOR, and p-mTOR in Cholangiocarcinoma cells (Fig. 3D–F). Tan‑IIA inhibited activation of PI3K/AKT/mTOR. Western blotting was used to detect the expression of PI3K, Akt, p-Akt, mTOR, and p-mTOR. The results indicated that Tan-IIA inhibited the expression of PI3K, p-Akt, p-mTOR, and mTOR compared to control group in a dose-dependent manner (Fig. 3A–C). Furthermore, pretreatment of Cholangiocarcinoma cells with the 740 y-p (PI3K agonist) abolished the effects of Tan-IIA on the restrain PI3K, p-Akt, mTOR, and p-mTOR in Cholangiocarcinoma cells (Fig. 3D–F). PI3K agonists abolished the effects of Tan‑IIA. To further demonstrate that Tan-IIA affects the growth of Cholangiocarcinoma cells by suppressing the PI3K/Akt/mTOR signaling pathway, Cholangiocarcinoma cells were pretreated with 740 y-p (PI3K agonist) or without it in the presence of Tan-IIA. Resultsf These results confirm that PI3K agonists could attenuate the tumor suppressive effect of Tan-IIA. The data showed that the effect of Tan- IIA inhibited cell proliferation, migration, and invasiveness, and promoted apoptosis was reversed by 740 y-p treatment (Fig. 4A–O). Tan‑IIA suppresses tumor growth in vivo. We used HuCCT-1 cell xenograft model to investigate the anti-tumor effect of Tan-IIA and the role of PI3K/Akt/mTOR signaling pathway in the anti-tumor effect of Tan- IIA. After 3 weeks of treatment with 50 mg/kg Tan-IIA or 50 mg/kg Tan-IIA combined with 10 mg/kg 740y-p. The results showed that Tan-IIA also had a significant anti-tumor effect in vivo. However, the anti-tumor effect of Tan-IIA was largely attenuated by 740y-p (Fig. 5A–C). Discussion Phosphorylation of ribosomal protein S6 kinase (S6K) and eukaryotic translation initiation factor 4E binding protein 1 (4E-BP1) can be mediated by activated mTORC1, which leads to the re-release of eukaryotic translation initiation factor 4E (EIF4E). These factors directly lead to protein translation and cell cycle progres- sion. In addition, mTORC2 can act as an activator of Akt, forming a positive feedback loop between Akt and mTOR, further promoting cell proliferation and survival26. It is well documented that Tan-IIA plays an anti-tumor role by regulating the PI3K/Akt/mTOR signaling pathway. Tan-IIA can block the Ras/Raf/MEK/ERK and PI3K/ Akt/mTOR pathways, thereby inhibiting the malignant growth of human pancreatic cancer cells8. Tan-IIA has also been reported to induce apoptosis and autophagy in acute monocytic leukemia by inhibiting the PI3K/Akt/ mTOR signaling pathway27. In the present study, the results revealed that Tan-IIA significantly inhibited the level of PI3K, p-Akt, mTOR and p-mTOR protein in Cholangiocarcinoma cells. Following the addition of the PI3K agonist (740 y-p), the levels of p-Akt, mTOR and p-mTOR increased, and 740 y-p eliminated the inhibitory effect of Tan-IIA on Cholangiocarcinoma cells. These results indicated that Tan-IIA inhibited the proliferation, migration, and invasion of Cholangiocarcinoma cells by inhibiting the PI3K/AKT/mTOR pathway. Apoptosis is an important manifestation of cell death, and promoting tumor cell apoptosis plays a crucial role in inhibiting the development and progression of tumors28. Previous studies have shown that the PI3K/Akt/ mTOR signaling pathway plays an essential role in regulating cell apoptosis29. The novel finding in this study is that inhibition of PI3K by 740 y-p eliminated the effect of Tan-IIA on the proapoptosis of Cholangiocarcinoma cells. Our findings indicated that Tan-IIA induced Cholangiocarcinoma cells apoptosis via inhibition of the Akt/ mTOR pathway. In addition, the Bcl-2 family and the caspase family have been reported to participate in the mitochondria-mediated pathway of apoptosis. Caspase-3 belongs to the cysteine-aspartic acid protease (caspase) family, which regulates apoptosis by interacting with caspase-8 and caspase-930. Whereas Bax and Bcl-2 are mem- bers of the Bcl-2 gene family, Bcl-2 mainly plays a role in inhibiting apoptosis and can maintain cell survival31. Bax, on the other hand, is a pro-apoptotic factor in the Bcl-2 family that can affect the mitochondrial membrane, leading to the release of cytochrome C and the production of reactive oxygen species, in addition, Bax can form https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 \| www.nature.com/scientificreports/ Figure 3. Tan-IIA can inhibit the PI3K-Akt-mTOR pathway. Discussion (A–C) Tan-IIA at concentrations of 0, 10, 20, and 30 µg/mL was added to fresh medium and co-cultured with Cholangiocarcinoma cells for 24 h. Next, the protein expression levels of PI3K, Akt, mTOR, p-Akt, and p-mTOR were measured using western blots. (D–F) After the addition of 740y-p (10 µg/mL) to the system in which Tan-IIA (24 h IC50 concentration) and Cholangiocarcinoma cells were co-cultured. The inhibitory effect of Tan-IIA on PI3K-Akt-mTOR pathway was diminished under the influence of PI3K agonists. Compared with control, p < 0.05; p < 0.01; p < 0.001; **p < 0.0001. Figure 3. Tan-IIA can inhibit the PI3K-Akt-mTOR pathway. (A–C) Tan-IIA at concentrations of 0, 10, 20, and 30 µg/mL was added to fresh medium and co-cultured with Cholangiocarcinoma cells for 24 h. Next, the protein expression levels of PI3K, Akt, mTOR, p-Akt, and p-mTOR were measured using western blots. (D–F) After the addition of 740y-p (10 µg/mL) to the system in which Tan-IIA (24 h IC50 concentration) and Cholangiocarcinoma cells were co-cultured. The inhibitory effect of Tan-IIA on PI3K-Akt-mTOR pathway was diminished under the influence of PI3K agonists. Compared with control, p < 0.05; p < 0.01; p < 0.001; **p < 0.0001. Figure 3. Tan-IIA can inhibit the PI3K-Akt-mTOR pathway. (A–C) Tan-IIA at concentrations of 0, 10, 20, and 30 µg/mL was added to fresh medium and co-cultured with Cholangiocarcinoma cells for 24 h. Next, the protein expression levels of PI3K, Akt, mTOR, p-Akt, and p-mTOR were measured using western blots. (D–F) After the addition of 740y-p (10 µg/mL) to the system in which Tan-IIA (24 h IC50 concentration) and Cholangiocarcinoma cells were co-cultured. The inhibitory effect of Tan-IIA on PI3K-Akt-mTOR pathway was diminished under the influence of PI3K agonists. Compared with control, p < 0.05; p < 0.01; p < 0.001; **p < 0.0001. heterodimer with Bcl-2 and activate endogenous apoptosis. Therefore, Bax is a major regulator involved in the process of apoptosis32. Moreover, compared with the expression of Bax and Bcl-2, respectively, the ratio of Bax/ Bcl-2 is more important for apoptosis. In this study, western blotting and flow cytometry assays revealed that heterodimer with Bcl-2 and activate endogenous apoptosis. Therefore, Bax is a major regulator involved in the process of apoptosis32. Moreover, compared with the expression of Bax and Bcl-2, respectively, the ratio of Bax/ Bcl-2 is more important for apoptosis. Discussion In this study, western blotting and flow cytometry assays revealed that https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 \| m/scientificreports/ Figure 4. PI3K agonists attenuate the tumor suppressive effect of Tan-IIA. After the addition of 740y-p (10 µg /mL) to the system in which Tan-IIA (24 h IC50 concentration) and Cholangiocarcinoma cells were co-cultured, the tumor suppressive effect of Tan-IIA was attenuated. (A–E) The inhibitory effect of Tan-IIA on the proliferation of Cholangiocarcinoma cells decreased after the addition of 740y-p. (F–L) The ability of Tan- IIA to inhibit the invasion and migration of Cholangiocarcinoma cells was reduced after the addition of 740y-p. (M–O) The apoptosis-inducing effect of Tan-IIA on Cholangiocarcinoma cells was diminished after 740y-p was added. Compared with control, p < 0.05; p < 0.01; p < 0.001. www.nature.com/scientificreports/ Figure 4. PI3K agonists attenuate the tumor suppressive effect of Tan-IIA. After the addition of 740y-p (10 µg /mL) to the system in which Tan-IIA (24 h IC50 concentration) and Cholangiocarcinoma cells were co-cultured, the tumor suppressive effect of Tan-IIA was attenuated. (A–E) The inhibitory effect of Tan-IIA on the proliferation of Cholangiocarcinoma cells decreased after the addition of 740y-p. (F–L) The ability of Tan- IIA to inhibit the invasion and migration of Cholangiocarcinoma cells was reduced after the addition of 740y-p. (M–O) The apoptosis-inducing effect of Tan-IIA on Cholangiocarcinoma cells was diminished after 740y-p was added. Compared with control, p < 0.05; p < 0.01; p < 0.001. Scientific Reports \| (2021) 11:19268 \| https://doi.org/10.1038/s41598-021-98948-z www.nature.com/scientificreports/ Figure 5. Tan-IIA suppresses tumor growth in vivo. Mice were treated with normal saline, Tan-IIA (50 mg/ kg) and Tan-IIA (50 mg/kg) combined with 740y-p (10 mg/kg) for 3 weeks, tumor volume was measured every 3 days (n = 5). (A) After treatment with normal saline, Tan-IIA and Tan-IIA combined with 740y-p for 3 weeks, tumors in these groups were removed, tumor pictures were captured (B) and tumors were weighed (C). Compared with control, p < 0.05; p < 0.01; p < 0.001. Figure 5. Tan-IIA suppresses tumor growth in vivo. Mice were treated with normal saline, Tan-IIA (50 mg/ kg) and Tan-IIA (50 mg/kg) combined with 740y-p (10 mg/kg) for 3 weeks, tumor volume was measured every 3 days (n = 5). (A) After treatment with normal saline, Tan-IIA and Tan-IIA combined with 740y-p for 3 weeks, tumors in these groups were removed, tumor pictures were captured (B) and tumors were weighed (C). Discussion Compared with control, p < 0.05; p < 0.01; *p < 0.001. Tan-IIA treatment promoted apoptosis and upregulated the level of cleaved caspase-3 and the ratio of Bax/Bcl-2. The results demonstrated that caspase3 and Bax/Bcl2 play critical regulatory roles in the progression of Tan-IIA promoted apoptosis in Cholangiocarcinoma cells. Tan-IIA treatment promoted apoptosis and upregulated the level of cleaved caspase-3 and the ratio of Bax/Bcl-2. The results demonstrated that caspase3 and Bax/Bcl2 play critical regulatory roles in the progression of Tan-IIA promoted apoptosis in Cholangiocarcinoma cells. References Tanshinone IIA can inhibit MiaPaCa2 human pancreatic cancer cells by dual blockade of the Ras/Raf/MEK/ERK and PI3K/AKT/mTOR pathways. Oncol. Rep. 40, 3102–3111. https://doi.org/10.3892/or.2018.6670 (2018). 8. Su, C. C. Tanshinone IIA can inhibit MiaPaCa2 human pancreatic cancer cells by dual blockade of the Ras/Raf/MEK/ERK and PI3K/AKT/mTOR pathways. Oncol. Rep. 40, 3102–3111. https://doi.org/10.3892/or.2018.6670 (2018). p y p p g 9. Huang, C. Y. et al. Tanshinone IIA inhibits the growth of pancreatic cancer BxPC3 cells by decreasing protein expression of TCTP MCL1 and BclxL. Mol. Med. Rep. 7, 1045–1049. https://doi.org/10.3892/mmr.2013.1290 (2013).i 0. Yuan, H. et al. How can synergism of traditional medicines benefit from network pharmacology?. Molecules https://doi.org/10 3390/molecules22071135 (2017). 1. Liu, P., Cheng, H., Roberts, T. M. & Zhao, J. J. Targeting the phosphoinositide 3-kinase pathway in cancer. Nat. Rev. Drug Discov 8, 627–644. https://doi.org/10.1038/nrd2926 (2009). p g 2. Chen, H. & Boutros, P. C. VennDiagram: A package for the generation of highly-customizable Venn and Euler diagrams in R BMC Bioinform. 12, 35. https://doi.org/10.1186/1471-2105-12-35 (2011).ih f p g 3. Ashburner, M. et al. Gene ontology: Tool for the unification of biology. The gene ontology consortium. Nat. Genet. 25, 25–29 https://doi.org/10.1038/75556 (2000).i 14. Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: An R package for comparing biological themes among gene clusters. OMICS 16, 284–287. https://doi.org/10.1089/omi.2011.0118 (2012). p g 5. Kanehisa, M. & Goto, S. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28, 27–30. https://doi.org/10.1093/ nar/28.1.27 (2000). 16. Kanehisa, M., Goto, S., Furumichi, M., Tanabe, M. & Hirakawa, M. KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res. 38, D355-360. https://doi.org/10.1093/nar/gkp896 (2010). 17. Ansari, M. A. et al. Prospective therapeutic potential of Tanshinone IIA: An updated overview. Pharmacol. Res. https://doi.org/ 10.1016/j.phrs.2020.105364 (2020). 18. Qian, J. et al. Tanshinone IIA promotes IL2-mediated SW480 colorectal cancer cell apoptosis by triggering INF2-related mitochon- drial fission and activating the Mst1-Hippo pathway. Biomed. Pharmacother. 108, 1658–1669. https://doi.org/10.1016/j.biopha. 2018.09.170 (2018).h ( ) 9. Fruman, D. A. et al. The PI3K pathway in human disease. Cell 170, 605–635. https://doi.org/10.1016/j.cell.2017.07.029 (2017). h 0. Zhang, J., Yu, X. H., Yan, Y. G., Wang, C. & Wang, W. J. PI3K/Akt signaling in osteosarcoma. Clin. Chim. Acta 444, 182–192. https:/ doi.org/10.1016/j.cca.2014.12.041 (2015). g j 1. Corti, F. et al. Targeting the PI3K/AKT/mTOR pathway in biliary tract cancers: A review of current evidences and future perspec- tives. Cancer Treat. Rev. 72, 45–55. References References 1. El-Diwany, R., Pawlik, T. M. & Ejaz, A. Intrahepatic Cholangiocarcinoma. Surg. Oncol. Clin. N Am. 28, 587–599. https://doi.org/ 10.1016/j.soc.2019.06.002 (2019). 1. El-Diwany, R., Pawlik, T. M. & Ejaz, A. Intrahepatic Cholangiocarcinoma. Surg. Oncol. Clin. N Am. 28, 587–599. https://doi.org/ 10.1016/j.soc.2019.06.002 (2019). j ( ) 2. Rizvi, S., Khan, S. A., Hallemeier, C. L., Kelley, R. K. & Gores, G. J. Cholangiocarcinoma—Evolving concepts and therapeutic strategies. Nat. Rev. Clin. Oncol. 15, 95–111. https://doi.org/10.1038/nrclinonc.2017.157 (2018). j 2. Rizvi, S., Khan, S. A., Hallemeier, C. L., Kelley, R. K. & Gores, G. J. Cholangiocarcinoma—Evolving concepts and therapeutic strategies. Nat. Rev. Clin. Oncol. 15, 95–111. https://doi.org/10.1038/nrclinonc.2017.157 (2018). g g 3. Patel, N. & Benipal, B. Incidence of Cholangiocarcinoma in the USA from 2001 to 2015: A US cancer statistics analysis of 50 states Cureus 11, e3962. https://doi.org/10.7759/cureus.3962 (2019). p g 4. Weigt, J. & Malfertheiner, P. Cisplatin plus gemcitabine versus gemcitabine for biliary tract cancer. Expert Rev. Gastroenterol Hepatol. 4, 395–397. https://doi.org/10.1586/egh.10.45 (2010). Hepatol. 4, 395–397. https://doi.org/10.1586/egh.10.45 (2010). p p g g 5. Che, A. J. et al. Separation and determination of active components in Radix Salviae miltiorrhizae and its medicinal preparations by nonaqueous capillary electrophoresis. J. Sep. Sci. 27, 569–575. https://doi.org/10.1002/jssc.200301710 (2004). y q p y p p p g j 6. Zhou, L., Zuo, Z. & Chow, M. S. Danshen: An overview of its chemistry, pharmacology, pharmacokinetics, and clinical use. J. Clin Pharmacol. 45, 1345–1359. https://doi.org/10.1177/0091270005282630 (2005). 6. Zhou, L., Zuo, Z. & Chow, M. S. Danshen: An overview of its chemistry, pharmaco Pharmacol. 45, 1345–1359. https://doi.org/10.1177/0091270005282630 (2005). p g ( ) 7. Ma, L. et al. Tanshinone IIA mediates SMAD7-YAP interaction to inhibit liver cancer growth by inactivating the transforming h f b l h ( lb ) h d ( ) p g 7. Ma, L. et al. Tanshinone IIA mediates SMAD7-YAP interaction to inhibit liver cancer growth by inactivating the transforming growth factor beta signaling pathway. Aging (Albany NY) 11, 9719–9737. https://doi.org/10.18632/aging.102420 (2019). g g g p y g g ( y ) , p g g g ( ) 8. Su, C. C. Tanshinone IIA can inhibit MiaPaCa2 human pancreatic cancer cells by dual blockade of the Ras/Raf/MEK/ERK PI3K/AKT/ TOR th O l R p 40 3102 3111 htt //d i /10 3892/ 2018 6670 (2018) g g g p y g g y p g g g 8. Su, C. C. Conclusion In conclusion, our study revealed the potential anti-tumor effects of Tan-IIA in Cholangiocarcinoma cells. Further mechanism studies demonstrated that Tan-IIA promoted apoptosis and suppressed malignant growth, invasion, and migration of Cholangiocarcinoma cells through inhibited PI3K/Akt/mTOR signaling pathway. Furthermore, we suggested that Tan-IIA could be a potent agent for the treatment of CCA. Received: 19 May 2021; Accepted: 17 September 2021 Acknowledgements g We would like to thank Editage [http://www.editage.cn] for English language editing. g We would like to thank Editage [http://www Author contributions H.L.: Investigation, formal analysis, writing original draft. C.L.: Investigation, methodology, funding acquisition. Writing review and editing. M.W.: Resources. D.S.: Resources. P.Zhu.: Resources. P.Zhang.: Supervision. X.T.: Writing—review and editing, funding acquisition. G.S.: Conceptualization, project administration, Writing— review and editing, funding acquisition. References https://doi.org/10.1016/j.ctrv.2018.11.001 (2019). https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 \| www.nature.com/scientificreports/ www.nature.com/scientificreports/ 22. Polo, M. L. et al. Activation of PI3K/Akt/mTOR signaling in the tumor stroma drives endocrine therapy-dependent breast tumor regression. Oncotarget 6, 22081–22097. https://doi.org/10.18632/oncotarget.4203 (2015). g g g g 23. Xia, P. & Xu, X. Y. PI3K/Akt/mTOR signaling pathway in cancer stem cells: From basic research to clinical application. A Cancer Res. 5, 1602–1609 (2015).h ( ) 4. Ersahin, T., Tuncbag, N. & Cetin-Atalay, R. The PI3K/AKT/mTOR interactive pathway. Mol. Biosyst. 11, 1946–1954. https://doi org/10.1039/c5mb00101c (2015). 5. Tian, T., Li, X. & Zhang, J. mTOR signaling in cancer and mTOR inhibitors in solid tumor targeting therapy. Int. J. Mol. Sci. https:// doi.org/10.3390/ijms20030755 (2019).i g j 6. Chang, W. et al. A critical role for the mTORC2 pathway in lung fibrosis. PLoS ONE 9, e106155. https://doi.org/10.1371/journal pone.0106155 (2014). p 27. Zhang, Y. et al. Tanshinone IIA induces apoptosis and autophagy in acute monocytic leukemia via downregulation of PI3K pathway. Am. J. Transl. Res. 11, 2995–3006 (2019). p y 28. Bold, R. J., Termuhlen, P. M. & McConkey, D. J. Apoptosis, cancer and cancer therapy. Surg. Oncol. 6, 133–142. https://doi.or 1016/s0960-7404(97)00015-7 (1997). ( ) ( ) 29. Manfredi, G. I. et al. PI3K/Akt/mTOR signaling in medullary thyroid cancer: A promising molecular target for cancer therapy. Endocrine 48, 363–370. https://doi.org/10.1007/s12020-014-0380-1 (2015). p g 0. Alnemri, E. S. et al. Human ICE/CED-3 protease nomenclature. Cell 87, 171. https://doi.org/10.1016/s0092-8674(00)81334-3 (1996). ( ) 1. Hardwick, J. M. & Soane, L. Multiple functions of BCL-2 family proteins. Cold Spring Harb. Perspect. Biol. https://doi.org/10.1101/ cshperspect.a008722 (2013). p p 32. Gross, A., Jockel, J., Wei, M. C. & Korsmeyer, S. J. Enforced dimerization of BAX results in its translocation, mitochondrial dys- function and apoptosis. EMBO J. 17, 3878–3885. https://doi.org/10.1093/emboj/17.14.3878 (1998). Acknowledgements © The Author(s) 2021 Fundingh g This work was supported by research grants from the National Natural Science Foundation of China (No. 81601617); Shandong Province Key Research and Development Project (No. 2018G SF118057). Competing interests h p g The authors declare no competing interests. Additional informationh Supplementary Information The online version contains supplementary material available at https://doi.org/ 10.1038/s41598-021-98948-z. Correspondence and requests for materials should be addressed to G.S. Reprints and permissions information is available at www.nature.com/reprints. 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 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 Author(s) 2021 https://doi.org/10.1038/s41598-021-98948-z Scientific Reports \| (2021) 11:19268 \|
https://openalex.org/W4390221767	https://jurnal.stituwjombang.ac.id/index.php/irsyaduna/article/download/1429/625	Indonesian	null	Membimbing Siswa Bermasalah Melalui Refleksi (Analisis Peran Bu Prani Sebagai Guru Bk Dalam Film “Budi Pekerti” Karya Wregas)	Irsyaduna	2,023	cc-by-sa	2,913	DOI: https://doi.org/10.54437/irsyaduna MEMBIMBING SISWA BERMASALAH MELALUI REFLEKSI (Analisis Peran Bu Prani Sebagai Guru Bk Dalam Film “Budi Pekerti” Karya Wregas) Nur Hikmah Okti Sania Putri oktisania303@gmail.com Universitas Hasyim Asyari Tebuireng Jombang Nur 'Azah azahnur31@gmail.com Universitas Hasyim Asyari Tebuireng Jombang Miftahul Jannah jannahmiftahul2@icloud.com Universitas Hasyim Asyari Tebuireng Jombang Hilmah Zahrotun Nahdliyah hilmahzahrotunn@gmail.com Universitas Hasyim Asyari Tebuireng Jombang Hilmah Zahrotun Nahdliyah hilmahzahrotunn@gmail.com Universitas Hasyim Asyari Tebuireng Jombang IRSYADUNA: Jurnal Studi Kemahasiswaan Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490; E-ISSN : 2776-5393 https://jurnal.stituwjombang.ac.id/index.php/irsyaduna DOI: https://doi.org/10.54437/irsyaduna MEMBIMBING SISWA BERMASALAH MELALUI REFLEKSI (Analisis Peran Bu Prani Sebagai Guru Bk Dalam Film “Budi Pekerti” Karya Wregas) Nur Hikmah Okti Sania Putri oktisania303@gmail.com Universitas Hasyim Asyari Tebuireng Jombang Nur 'Azah azahnur31@gmail.com Universitas Hasyim Asyari Tebuireng Jombang Miftahul Jannah jannahmiftahul2@icloud.com Universitas Hasyim Asyari Tebuireng Jombang Hilmah Zahrotun Nahdliyah hilmahzahrotunn@gmail.com Universitas Hasyim Asyari Tebuireng Jombang Abstract IRSYADUNA: Jurnal Studi Kemahasiswaan Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490; E-ISSN : 2776-5393 https://jurnal.stituwjombang.ac.id/index.php/irsyaduna DOI: https://doi.org/10.54437/irsyaduna Abstrak: Film "Budi Pekerti" karya Wregas Bhanot menampilkan kisah inspirasif tentang peran seorang guru BK dalam membimbing siswa bermasalah. Penelitian ini bertujuan menganalisis representasi peran guru BK dalam film tersebut. Metode penelitian kualitatif dengan pendekatan analisis wacana model Teun A. van Dijk diterapkan. Hasil analisis menunjukkan bahwa tokoh Bu Prani direpresentasikan sebagai guru BK ideal yang bijaksana dan penuh pengertian dalam membimbing siswa bermasalah. Ia tidak memberikan hukuman, namun selalu memberikan "refleksi" yang mendorong introspeksi diri dan perubahan positif siswa. Representasi ini relevan dengan kebutuhan pendekatan baru dalam pendidikan saat ini, di mana hukuman dinilai tidak efektif mengubah perilaku siswa jangka panjang. Guru BK dipandang sebagai figur kunci untuk mewujudkan perubahan paradigma ini dengan pendekatan yang dialogis dan humanis. Penelitian ini menyimpulkan bahwa sosok Bu Prani dapat menginspirasi guru BK Indonesia untuk lebih optimal dan kreatif dalam membimbing siswa bermasalah melalui refleksi, bukan hukuman. Film ini berpotensi menjadi media edukasi yang powerfull untuk membangun sudut pandang baru dalam penanganan siswa bermasalah. ata Kunci: Hukuman, Refleksi, guru BK, Film Budi Pekerti. Kata Kunci: Hukuman, Refleksi, guru BK, Film Budi Pekerti. eywords: Punishment, reflection, school counselor dan Budi Pekerti film Keywords: Punishment, reflection, school counselor dan Budi Pekerti film Abstract The film "Budi Pekerti" by Wregas Bhanot presents an inspirational story about the role of a school counselor in guiding troubled students. This study aims to analyze the representation of the school counselor's role in the film. A qualitative research method with Teun A. van Dijk's discourse analysis approach was applied. The results of the analysis show that the character Bu Prani is represented as an ideal school counselor who is wise and compassionate in guiding troubled students. She does not punish, but always provides "reflection" which encourages students' self-introspection and positive change. This representation is relevant to the need for a new approach in education today, where punishment is considered ineffective in changing long- term student behavior. School counselors are seen as key figures in realizing this paradigm shift through dialogical and humanistic approaches. This study concludes that the figure of Bu Prani can inspire Indonesian school counselors to be more optimal and creative in guiding troubled students through reflection, not punishment. This film has the potential to become a powerful Irsyaduna, Jurnal Studi Kemahasiswaan. https://jurnal.stituwjombang.ac.id/index.php/irsyaduna Irsyaduna, Jurnal Studi Kemahasiswaan. https://jurnal.stituwjombang.ac.id/index.php/irsyaduna Vol. 3, No. 3, Dessember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 330 Nur Hikmah Okti Sania Putri, dkk. , Membimbing Siswa Bermasalah Melalui Refleksi… Pendahuluan Film "Budi Pekerti" karya Wregas Bhanot menampilkan kisah seorang siswa SMA bernama Budi yang bermasalah dalam pergaulan dan perilakunya. Ia sering terlibat perkelahian dan melakukan tindakan-tindakan negatif lainnya. Bu Prani, guru BK di sekolah Budi, berupaya membimbing Budi agar menyadari dan memperbaiki perilakunya. Interaksi antara Bu Prani dan Budi menunjukkan peran penting guru BK dalam memandu siswa yang bermasalah agar tidak hanya dihukum, tetapi juga melakukan refleksi dan introspeksi diri (antaranews.com, 2023). Peran guru BK dalam penanganan siswa bermasalah menjadi isu penting dalam pendidikan. Berbagai penelitian menunjukkan bahwa hukuman tidak efektif dalam mengubah perilaku jangka panjang siswa (Mulyati & Kamaruddin, 2020). Sebaliknya, dibutuhkan pendekatan yang lebih manusiawi, melibatkan dialog dan refleksi bersama siswa (Brown, 2007; Osher et al., 2010). Guru BK dipandang sebagai figur kunci dalam pendekatan ini karena kompetensinya dalam konseling dan membimbing siswa (Assingkily & Mahidin, 2022). Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna 331 Nur Hikmah Okti Sania Putri, dkk. , Membimbing Siswa Bermasalah Melalui Refleksi… Dalam konteks pendidikan di Indonesia, penelitian (Irmayanti & Yuliani, 2020) menemukan bahwa seringkali guru BK hanya berperan administratif dan kurang optimal dalam membimbing siswa berkebutuhan khusus. Peran guru BK seharusnya diarahkan pada upaya pemahaman karakteristik siswa secara mendalam dan pemberian layanan yang tepat. Senada dengan itu, Aprilia (2021) menekankan perlunya peningkatan kompetensi guru BK Indonesia dalam layanan konseling individual. Kajian yang lebih spesifik tentang peran guru BK dalam menangani siswa bermasalah juga masih terbatas. Beberapa penelitian terkait menyimpulkan perlunya penguatan fungsi guru BK sebagai konselor dan fasilitator siswa dalam memahami dan mengatasi masalahnya (Handayani, 2020; Maryani & Fatimah, 2021). Interaksi dialogis antara guru BK dan siswa berperan besar dalam membangun kesadaran dan motivasi perubahan pada diri siswa (Astiti dkk., 2018). Dengan demikian, studi ini bertujuan menganalisis peran Bu Prani sebagai guru BK dalam membimbing Budi melakukan refleksi dan perubahan perilaku dalam film "Budi Pekerti". Film ini dipilih karena menampilkan dinamika yang mendalam antara guru BK dan siswa bermasalah, yang jarang ditampilkan dalam film Indonesia. Analisis dilakukan dengan pendekatan analisis wacana Teun A. van Dijk. Kerangka van Dijk (2009) menitikberatkan pada bagaimana wacana/teks media merepresentasikan aktor sosial dan relasi sosial di dalamnya. Analisis mencakup tiga dimensi: teks, kognisi sosial, dan konteks sosial. Dimensi teks mengkaji struktur wacana verbal maupun visual. Dimensi kognisi sosial mengungkap representasi mental, pengetahuan, dan ideologi yang melandasi wacana. Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna Pendahuluan Dimensi konteks sosial mempertimbangkan konteks situasi dan struktur sosial yang melingkupi wacana (Fauzan, 2014). Beberapa penelitian terdahulu tentang film telah menerapkan kerangka van Dijk. Mislanya, Pratama (2018) menemukan stereotip negatif terhadap kelompok minoritas dan marjinal dalam film Indonesia populer. Sementara itu, kajian Alfadian (2021) pada film komedi romantis Hollywood menunjukkan representasi bias gender tentang peran perempuan. Kedua studi ini menekankan bagaimana analisis wacana kritis penting untuk mengungkap ideologi di balik representasi teks media. Dengan menerapkan kerangka van Dijk, penelitian ini diharapkan mampu secara komprehensif mengungkapkan representasi peran guru BK dan relasinya dengan siswa bermasalah dalam film "Budi Pekerti". Secara teoretis, hasil analisis diharapkan memperkaya kajian media dan konseling, khususnya representasi peran konselor sekolah dalam film Indonesia. Secara praktis, hasil penelitian dapat menjadi masukan bagi para guru BK dan calon guru BK dalam memahami dan meningkatkan peran mereka membimbing siswa bermasalah. Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna 332 Nur Hikmah Okti Sania Putri, dkk. Membimbing Siswa Bermasalah Melalui Refleksi… Membimbing Siswa Bermasalah Melalui Refleksi… Metode Penelitian Penelitian ini menggunakan metode analisis isi kualitatif dengan pendekatan analisis wacana Teun A. van Dijk (Iskandar, 2022). Data penelitian adalah scene- scene dalam film "Budi Pekerti" yang menampilkan interaksi antara Bu Prani sebagai guru BK dengan Budi sebagai siswa bermasalah. Pengumpulan data dilakukan dengan teknik dokumentasi, yaitu mengunduh scene-scene relevan dari film "Budi Pekerti" yang diunggah di platform media sosial YouTube. Kriteria inklusi datanya adalah scene yang menampilkan verbal maupun nonverbal interaksi antara Bu Prani dan Budi terkait permasalahan perilaku Budi. Analisis data mengacu pada model (Huberman & Miles, 2002) melalui tahapan: 1) reduksi data, memilih scene yang relevan dengan fokus penelitian; 2) penyajian data dalam bentuk transkrip dialog dan deskripsi visual; 3) interpretasi data menggunakan kerangka analisis wacana van Dijk (Susilo dkk., 2021) Pengecekan keabsahan temuan dilakukan dengan triangulasi sumber, yaitu mendiskusikan hasil analisis dengan dua rater lain yang kompeten. Rater lainnya adalah dosen Bimbingan dan Konseling dan psikolog. Diskusi dilakukan untuk memastikan validitas interpretasi data yang dilakukan peneliti. Hasil Penelitian Penelitian ini menganalisis representasi peran Bu Prani sebagai guru BK dalam film "Budi Pekerti" karya Wregas Bhanot. Data penelitian berupa scene-scene dalam film yang menampilkan interaksi antara Bu Prani dan siswa-siswa bermasalah, khususnya Budi. Hasil analisis menunjukkan bahwa film "Budi Pekerti" secara keseluruhan merepresentasikan peran guru BK sebagai fasilitator bagi siswa dalam melakukan refleksi dan introspeksi diri. Representasi ini dibangun melalui tiga dimensi analisis wacana Teun A. van Dijk, yaitu teks, kognisi sosial, dan konteks sosial. Pada level teks, guru BK (Bu Prani) digambarkan memiliki relasi yang dekat dan kekeluargaan dengan para siswa. Ia diposisikan sebagai figur sentral dalam membimbing siswa melakukan refleksi atas perilaku mereka. Verbal maupun nonverbal, Bu Prani ditampilkan penuh empati dan bijaksana dalam berinteraksi dengan siswa. Ia tidak pernah menghukum atau memarahi siswa, melainkan memberi "refleksi" yang mendorong introspeksi dan perubahan positif pada diri siswa. Secara kognisi sosial, film ini merepresentasikan pemikiran bahwa hukuman tidak efektif dalam mengubah perilaku siswa jangka panjang. Sebaliknya, pendekatan humanis dengan melibatkan dialog dan refleksi bersama siswa dipandang lebih tepat. Bu Prani digambarkan mewujudkan pendekatan ini. Ia Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna 333 Nur Hikmah Okti Sania Putri, dkk. Membimbing Siswa Bermasalah Melalui Refleksi… Nur Hikmah Okti Sania Putri, dkk. Membimbing Siswa Bermasalah Melalui Refleksi… Nur Hikmah Okti Sania Putri, dkk. , Membimbing Siswa Bermasalah Melalui Refleksi… , Membimbing Siswa Bermasalah Melalui Refleksi… memahami karakter dan kebutuhan masing-masing siswa, kemudian memberikan bimbingan yang sesuai. Dalam konteks sosial, profesi guru BK masih belum optimal perannya dalam membimbing siswa berkebutuhan khusus. Film ini hadir untuk menunjukkan potensi dan peran penting guru BK, khususnya dalam pendekatan terhadap siswa bermasalah. Sosok Bu Prani merepresentasikan guru BK ideal yang bijaksana dan penuh pengertian terhadap siswa. Dengan demikian, film "Budi Pekerti" secara keseluruhan merepresentasikan guru BK (Bu Prani) sebagai figur sentral dan teladan dalam membimbing siswa bermasalah melalui pendekatan yang dialogis dan humanis, yaitu dengan memberikan "refleksi" bukan hukuman. Representasi ini sekaligus menunjukkan potensi dan harapan akan optimalisasi peran guru BK dalam pendidikan di Indonesia. Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna Pembahasan Representasi peran guru BK dalam film "Budi Pekerti" sangat relevan dengan konteks pendidikan di Indonesia saat ini. Berbagai penelitian menunjukkan bahwa hukuman tidak efektif dalam mengubah perilaku siswa jangka panjang. Sebaliknya, pendekatan yang melibatkan dialog dan refleksi bersama siswa terbukti lebih manjur dalam membimbing siswa bermasalah (Brown, 2007; Osher et al., 2010). Sayangnya, praktik pemberian hukuman atau punishment masih lazim ditemukan di sekolah-sekolah Indonesia. Penelitian Ndaru (2017) menemukan kecenderungan guru menggunakan hukuman fisik dan psikis untuk mengendalikan perilaku siswa. Seringkali hukuman diberikan tanpa mempertimbangkan dampak jangka panjangnya bagi perkembangan siswa (Rochimi & Suismanto, 2018). Kondisi ini menunjukkan urgensi untuk mengubah cara pandang dan pendekatan guru dalam menangani siswa bermasalah. Di sini, guru BK dipandang sebagai figur kunci untuk mewujudkan perubahan pendekatan tersebut. Seperti digambarkan dalam film, guru BK seperti Bu Prani dapat menjadi fasilitator bagi siswa bermasalah dalam melakukan refleksi dan introspeksi diri. Interaksinya yang penuh empati dan bijaksana mampu menyentuh siswa untuk sadar dan termotivasi mengubah perilakunya (Masri, 2020). Sayangnya, peran guru BK dalam hal ini masih belum optimal. Penelitian Dwijayanti (2017) mengkritisi guru BK Indonesia masih terfokus pada urusan administratif dan belum maksimal membimbing siswa berkebutuhan khusus. Oleh karena itu, sosok Bu Prani dalam film "Budi Pekerti" dapat menjadi teladan untuk menginspirasi dan memotivasi guru BK Indonesia agar lebih aktif membimbing siswa bermasalah melalui pendekatan humanis. Beberapa strategi yang dapat dilakukan guru BK untuk mengubah hukuman menjadi refleksi bagi siswa bermasalah, antara lain: Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 334 , Membimbing Siswa Bermasalah Melalui Refleksi… 1. Melakukan pendekatan secara personal kepada siswa untuk memahami latar belakang dan karakteristiknya. Ini penting sebagai dasar dalam memberikan bimbingan yang sesuai (Mufrihah, 2014). 2. Mendiskusikan masalah siswa secara terbuka dan memberi kesempatan siswa mengemukakan sudut pandangnya. Dialog dan diskusi tanpa menyalahkan dapat membangun kesadaran siswa (Rahmawati dkk., 2021). 3. Memberikan tugas reflektif yang relevan, seperti menulis essai tentang pelajaran hidup dari kesalahannya, melakukan kegiatan sosial, atau membuat karya seni yang merepresentasikan perasaannya. 4. Mengintegrasikan nilai-nilai positif dalam setiap bimbingan, misalnya empati, kerja sama, dan tanggung jawab. Ini dapat menanamkan karakter positif pada diri siswa (Rohmah dkk., 2023). kerja sama, dan tanggung jawab. Ini dapat menanamkan karakter positif pada diri siswa (Rohmah dkk., 2023). 5. Melakukan konseling dan diskusi berkala untuk memantau perkembangan siswa dan memberikan dukungan positif bagi siswa (Purwaningrum dkk., 2023). Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna Pembahasan Pendekatan-pendekatan tersebut dapat menjadikan hukuman sebagai momentum bagi siswa untuk belajar dari kesalahan dan berubah menjadi pribadi yang lebih baik. Guru BK dituntut kreativitas dan kearifan dalam merancang "refleksi" yang sesuai bagi setiap siswa bermasalah. Kolaborasi dengan orang tua siswa juga penting untuk mendukung efektivitas bimbingan yang diberikan. Berdasarkan hasil penelitian ini, beberapa rekomendasi yang dapat diberikan adalah: Bagi guru BK, sosok Bu Prani dalam film ini dapat menjadi teladan untuk lebih aktif dan kreatif dalam membimbing siswa bermasalah melalui pendekatan yang dialogis dan humanis, yaitu dengan memberikan refleksi bukan hukuman. Perlu adanya pelatihan dan workshop untuk meningkatkan kompetensi guru BK dalam hal ini. Bagi sekolah, perlu kebijakan yang mendukung pengembangan peran guru BK sebagai konselor dan fasilitator siswa berkebutuhan khusus. Beban kerja guru BK perlu disesuaikan agar mereka dapat optimal dalam membimbing siswa. Bagi peneliti selanjutnya, dapat mengembangkan penelitian dengan menganalisis respons dan tanggapan siswa/guru BK yang menonton film ini, misalnya melalui survei atau wawancara. Penelitian lanjutan juga dapat mengembangkan model/panduan bagi guru BK dalam menerapkan pendekatan reflektif untuk siswa bermasalah. Bagi sineas Indonesia, film-film bertema pendidikan seperti ini sangat positif dan perlu digalakkan. Sekaligus dapat terus mengeksplorasi potensi film sebagai media edukasi dan pembentukan cara pandang positif terhadap dunia pendidikan di Indonesia. Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna 335 Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna 335 Nur Hikmah Okti Sania Putri, dkk. , Membimbing Siswa Bermasalah Melalui Refleksi… Dengan mengubah cara pandang dan pendekatan terhadap siswa bermasalah, diharapkan dunia pendidikan Indonesia dapat memberikan solusi positif dalam mengentaskan masalah kenakalan remaja. Bukan hukuman, melainkan refleksi yang dapat menumbuhkan kesadaran dan perubahan perilaku siswa ke arah yang lebih baik. Film "Budi Pekerti" telah menunjukkan potensi guru BK untuk mewujudkan hal tersebut. Kesimpulan Film "Budi Pekerti" karya Wregas Bhanot merupakan sebuah karya yang brilian dalam merepresentasikan peran seorang guru BK, khususnya dalam menangani siswa bermasalah. Melalui tokoh Bu Prani, film ini menggambarkan bagaimana seharusnya guru BK membimbing siswa yang bermasalah, yaitu dengan pendekatan yang dialogis dan humanis, bukan dengan memberikan hukuman. Pendekatan Bu Prani yang unik dan luar biasa, yaitu dengan memberikan "refleksi" alih-alih hukuman, terbukti sangat efektif untuk membangun kesadaran dan mendorong perubahan positif pada diri para siswa. Sosok dan pendekatan Bu Prani sangat relevan dengan kebutuhan dunia pendidikan kita saat ini yang membutuhkan model teladan untuk menggantikan paradigma lama yang mengandalkan hukuman dalam menangani siswa bermasalah. Film ini menginspirasi para guru BK Indonesia untuk meningkatkan peran dan kreativitasnya dalam membimbing siswa bermasalah dengan pendekatan yang lebih manusiawi dan mendidik. Dengan demikian, film "Budi Pekerti" mampu menjadi media edukasi yang powerful untuk membangun cara pandang baru dalam penanganan siswa bermasalah di Indonesia. Daftar Pustaka antaranews.com. (2023, Oktober 31). Dedikasi sang guru dalam film “Budi Pekerti.” Antara News. https://www.antaranews.com/berita/3800337/dedikasi-sang- guru-dalam-film-budi-pekerti Assingkily, R., & Mahidin, M. (2022). Upaya Guru Bimbingan dan Konseling Mengatasi Perilaku Prokrastinasi Akademik Siswa Pasca Pandemi Covid-19. Hikmah, 19(2), Article 2. https://doi.org/10.53802/hikmah.v19i2.167 Astiti, P., Suminar, J. R., & Rahmat, A. (2018). Konstruksi Identitas Guru Bimbingan Konseling sebagai Komunikator Pendidikan. Jurnal Kajian Komunikasi, 6(1), Article 1. https://doi.org/10.24198/jkk.v6i1.7738 Fauzan, U. (2014). Analisis wacana kritis dari model Fairclough hingga Mills. Jurnal Pendidik, 6(1). uberman, M., & Miles, M. B. (2002). The qualitative researcher’s companion. sage. Irmayanti, R., & Yuliani, W. (2020). Peran bimbingan dan konseling di sekolah inklusif. JPK (Jurnal Pendidikan Khusus), 16(2), 87–93. Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna 336 Nur Hikmah Okti Sania Putri, dkk. Nur Hikmah Okti Sania Putri, dkk. Nur Hikmah Okti Sania Putri, dkk. Membimbing Siswa Bermasalah Melalui Refleksi… Iskandar, D. (2022). METODOLOGI PENELITIAN KUALITATIF: Petunjuk Praktis Untuk Penelitian Lapangan, Analisis Teks Media, Dan Kajian Budaya. Maghza Pustaka. Masri, S. (2020). Multicultural Awareness, Teknik Cinemeducation, Dan Bibliotherapy. Penerbit Aksara Timur. Mufrihah, A. (2014). Implikasi prinsip bimbingan dan konseling terhadap kompetensi multikultural konselor. Jurnal Pelopor Pendidikan, 7(1), 73–85. Mulyati, S., & Kamaruddin, K. (2020). Peran Guru dalam Pelaksanaan Bimbingan Konseling. Al-Liqo: Jurnal Pendidikan Islam, 5(02), 172–184. https://doi.org/10.46963/alliqo.v5i02.241 Purwaningrum, R., Surur, N., & Asrowi, A. (2023). Harmonisasi Hubungan Guru Bimbingan dan Konseling dengan Orang Tua melalui Strategi Kolaborasi: Systematic Literature Review. Indonesian Journal of Guidance and Counseling: Theory and Application, 12(1), Article 1. https://doi.org/10.15294/ijgc.v12i1.74559 Rahmawati, R., Evi, A., & Bangun, Y. W. (2021). Bimbingan dan Konseling Multibudaya: Vol. (R. Rahmawati, A. Evi, & Y. W. Bangun, Ed.; Nomor). Media Edukasi Indonesia. https://drive.google.com/file/d/14lu5j73kszNDiOXCC0t18sXVeSpMXgIq/v iew?usp=sharing Rochimi, I. F., & Suismanto, S. (2018). Upaya Guru Menanamkan Nilai-nilai Kedisplinan pada Anak Usia Dini. Golden Age: Jurnal Ilmiah Tumbuh Kembang Anak Usia Dini, 3(4), 231–246. Rohmah, N. N. S., Markhamah, Narimo, S., & Widyasari, C. (2023). Strategi Penguatan Profil Pelajar Pancasila Dimensi Berkebhinekaan Global Di Sekolah Dasar. Jurnal Elementaria Edukasia, 6(3), Article 3. https://doi.org/10.31949/jee.v6i3.6124 Susilo, D., Kom, S. I., & Kom, M. I. (2021). Analisis wacana kritis van dijk: Sebuah model dan tinjauan kritis pada media daring. Unitomo Press. Vol. 3, No. 3, Desember 2023 P-ISSN : 2777-1490 E-ISSN : 2776-5393 Irsyaduna, Jurnal Studi Kemahasiswaan. DOI: https://doi.org/10.54437/irsyaduna 337
https://openalex.org/W4387158620	https://academic.oup.com/grurint/advance-article-pdf/doi/10.1093/grurint/ikad098/51800813/ikad098.pdf	English	null	Overcoming Barriers to Text and Data Mining in the Era of ChatGPT: the Proposed Data Act as a Game-Changer	GRUR international	2,023	cc-by	13,131	GRUR International, 73(1), 2024, 34–44 GRUR International, 73(1), 2024, 34–44 https://doi.org/10.1093/grurint/ikad098 Advance Access Publication Date: 29 September 2023 Article 1 For a general overview of TDM techniques and methods see Jiawei Han, Micheline Kamber and Jian Pei, Data Mining: Concepts and Techniques (3rd edn, Elsevier 2012). 2 Directive on Copyright in the Digital Single Market (CDSM), art 2(2) defines TDM as ‘any automated analytical technique aimed at analyzing text and data in digital form in order to generate information which includes but is not limited to patterns, trends and correlations.’ Maryna Manteghi* Article 35 of the proposed Data Act intends to free databases comprised of machine-generated data from the constraints of sui generis database protection. The provision may provide some remedy for researchers facing chal- lenges under the text and data mining (TDM) exceptions of the Directive on Copyright in the Digital Single Market. However, the wording of Art. 35 may also raise questions and create legal uncertainties for scholars, as scientific research is not the main aim of the proposed legislation. This article argues that, even after the proposed Art. 35 of the Data Act, some databases comprised of machine-generated data may still fall within the scope of the sui generis database protection under certain circumstances. The discussion revolves, inter alia, around the concept of recorded data in the context of the ‘obtaining-creating dichotomy’, the use of mixed databases and the notion of derived or inferred data, and the issue of researchers’ access to machine-generated data. The findings of this article are intended to offer guidance to researchers using Artificial Intelligence tools, such as ChatGPT, to mine databases on how they may effectively avoid a potential infringement of the sui generis database right. The paper hopes to encourage new changes in the EU regulation on TDM that could create a more balanced and research-friendly framework. helping to quickly identify potential vaccine candidates.4 Furthermore, TDM lies at the core of Artificial Intelligence (AI) technologies.5 The tool is employed to develop new types of information-based services and applications. One of the recent examples of AI technologies whose develop- ment and functioning depend on TDM is a cutting-edge chat called ChatGPT. © Published by OUP on behalf of GRUR e.V. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/ by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. * Affiliated Doctoral Researcher, Faculty of Law, University of Turku, Finland. 3 Sean Flynn and Lokesh Vyas, ‘Examples of Text and Data Mining Research Using Copyrighted Materials’ (Kluwer Copyright Blog, 6 March 2023) <https://copyrightblog.kluweriplaw.com/2023/03/06/ examples-of-text-and-data-mining-research-using-copyrighted-materi- als/> accessed 20 March 2023. 4 Hao Lv and others, ‘Application of Artificial Intelligence and Machine Learning for COVID-19 Drug Discovery and Vaccine Design’ (2021) 22 Briefings in Bioinformatics 1. 5 Harry Surden, ‘Machine Learning and Law: An Overview’ in Roland Vogl (ed), Research Handbook on Big Data Law (Edward Elgar Publishing 2021) 171. 8 Jürgen Rudolph, Samson Tan and Shannon Tan, ‘ChatGPT: Bullshit Spewer or the End of Traditional Assessments in Higher Education?’ (2023) 6 Journal of Applied Learning and Teaching 1, 3. 6 Jianyang Deng and Yijia Lin, ‘The Benefits and Challenges of ChatGPT: An Overview’ (2022) 2 FCIS 81, 82. 4 Hao Lv and others, ‘Application of Artificial Intelligence and Machine Learning for COVID-19 Drug Discovery and Vaccine Design’ (2021) 22 Briefings in Bioinformatics 1. 5 Harry Surden, ‘Machine Learning and Law: An Overview’ in Roland Vogl (ed), Research Handbook on Big Data Law (Edward Elgar Publishing 2021) 171. 7 Rehan Ahmed Khan and others, ‘ChatGPT-Reshaping Medical Education and Clinical Management’ (2023) 39 Pak J Med Sci 605, 605. 8 Jürgen Rudolph, Samson Tan and Shannon Tan, ‘ChatGPT: Bullshit Spewer or the End of Traditional Assessments in Higher Education?’ (2023) 6 Journal of Applied Learning and Teaching 1, 3. 9 Dan Milmo, ‘ChatGPT reaches 100 million users two months after 6 Jianyang Deng and Yijia Lin, ‘The Benefits and Challenges of ChatGPT: An Overview’ (2022) 2 FCIS 81, 82. 7 Rehan Ahmed Khan and others, ‘ChatGPT-Reshaping Medical Education and Clinical Management’ (2023) 39 Pak J Med Sci 605, 605. 6 Jianyang Deng and Yijia Lin, ‘The Benefits and Challenges of ChatGPT: An Overview’ (2022) 2 FCIS 81, 82. 7 Rehan Ahmed Khan and others, ‘ChatGPT-Reshaping Medical Education and Clinical Management’ (2023) 39 Pak J Med Sci 605, 605. 4 Hao Lv and others, ‘Application of Artificial Intelligence and Machine Learning for COVID-19 Drug Discovery and Vaccine Design’ (2021) 22 Briefings in Bioinformatics 1. 5 Harry Surden, ‘Machine Learning and Law: An Overview’ in Roland Vogl (ed), Research Handbook on Big Data Law (Edward Elgar Publishing 2021) 171 9 Dan Milmo, ‘ChatGPT reaches 100 million users two months after launch’ The Guardian (London, 2 February 2023) <https://www. theguardian.com/technology/2023/feb/02/chatgpt-100-million-users- open-ai-fastest-growing-app> accessed 31 March 2023. I. Introduction 3, benefits research organizations and cultural heritage institutions carrying out TDM for the purpose of scientific research.16 Other types of users are covered by Art. 4, the second excep- tion, permitting TDM for any purpose.17 Although these provisions provide more legal certainty compared to the situation that existed before the adoption of the CDSM Directive, the wording of these provisions has attracted heavy criticism. For instance, Geiger and Jütte argue that the exclusion of private actors from the scope of Art. 3 would deprive the EU of a crucial source of innovation and research, especially in the fields of medicine and AI.18 Moreover, Ducato and Strowel emphasize that it could be challenging for users to prove that their research proj- ects fall within scientific research as scientific character- istics and methods vary.19 The exception under Art. 4 has also been subject to criticism, especially for including a so-called ‘opt-out’ mechanism, which allows rightholders to reserve the use of their works.20 chat has the potential to revolutionize all areas affected by its applications. Particularly, ChatGPT may have a signif- icant impact on academia by helping to improve research and learning. For instance, this advanced tool could assist researchers in finding relevant literature among a huge number of scientific publications, generating patterns and correlations from such data, translating and under- standing the content written in a foreign language, finding answers to specific questions and summarizing research materials.10 Although the use of AI technologies, such as ChatGPT, which are based on TDM, may provide lots of benefits for the public, it may also raise legal issues. In order to mine, a computer typically needs to collect and copy large volumes of digital data from various sources such as websites, social media platforms, databases and other digital repositories. Such data could be protected by copyright or database rights. Therefore, if original works or databases are copied without authorization, the act may infringe the rightholders’ exclusive right to reproduction, and result in protracted expensive litiga- tion in court. However, an infringement may also occur when TDM is performed on non-original compilations of data. In this case, the sui generis database right could be at risk. I. Introduction Text and data mining (TDM) is a significant development in research and innovation. This automated analytical technique enables users to obtain new knowledge from huge masses of raw digital data.1 The outcome of this data analytics usually reveals new patterns, trends and insights which could be beneficial for the public. The research tool works by retrieving new information from pre-existing digital data and recombining it into targeted output.2 Nowadays, many private and public actors employ TDM to improve their operations and meet higher standards. For instance, pharmaceutical companies use data analyt- ics to reveal drug interactions, online movie platforms such as Netflix employ these tools to analyze ratings and predict users’ preferences, and analytical journals like The Wall Street Journal leverage TDM to analyze reports and predict stock market prices.3 Moreover, TDM has nota- ble implications in the field of medicine. For instance, it was employed to analyze a significant number of sci- entific publications regarding the coronavirus family, ChatGPT is a large-scale language model designed by OpenAI to generate content based on a huge number of pre-existing texts obtained from the internet within the setting of human-like conversations.6 This AI-based soft- ware application can perform different processing tasks such as translation, summarization and creation of texts.7 The chat has been trained on a huge amount of text data to promptly generate responses to users’ requests.8 ChatGPT was released on 30 November 2022 and it reached 100 million users within just a few months of its launch.9 The 6 Jianyang Deng and Yijia Lin, ‘The Benefits and Challenges of ChatGPT: An Overview’ (2022) 2 FCIS 81, 82. 3 Sean Flynn and Lokesh Vyas, ‘Examples of Text and Data Mining Research Using Copyrighted Materials’ (Kluwer Copyright Blog, 6 March 2023) <https://copyrightblog.kluweriplaw.com/2023/03/06/ examples-of-text-and-data-mining-research-using-copyrighted-materi- als/> accessed 20 March 2023. 9 Dan Milmo, ‘ChatGPT reaches 100 million users two months after launch’ The Guardian (London, 2 February 2023) <https://www. theguardian.com/technology/2023/feb/02/chatgpt-100-million-users- open-ai-fastest-growing-app> accessed 31 March 2023. 9 Dan Milmo, ‘ChatGPT reaches 100 million users two months after launch’ The Guardian (London, 2 February 2023) <https://www. theguardian.com/technology/2023/feb/02/chatgpt-100-million-users- open-ai-fastest-growing-app> accessed 31 March 2023. 9 Dan Milmo, ‘ChatGPT reaches 100 million users two months after launch’ The Guardian (London, 2 February 2023) <https://www. theguardian.com/technology/2023/feb/02/chatgpt-100-million-users- open-ai-fastest-growing-app> accessed 31 March 2023. 34 Overcoming Barriers to Text and Data Mining in the Era of ChatGPT 35 Directive)15 could provide a solution for researchers in this situation. The first exception, in Art. 20 Andrew Tyner, ‘The EU Copyright Directive: ‘Fit for the Digital Age or Finishing It?’’ (2020) 26 J. INTELL. PROP. L. 275, 281; P Bernt Hugenholtz, ‘The New Copyright Directive: Text and Data Mining (Articles 3 and 4)’ (Kluwer Copyright Blog, 24 July 2019) <http://copy- rightblog.kluweriplaw.com/2019/07/24/the-new-copyright-directive- text-and-data-mining-articles-3-and-4/> accessed 12 April 2023. 21 Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act), Brussels [2022] COM (2022) 68 final (proposed Data Act). 22 Commission Staff Working Document, Impact Assessment Report Accompanying the document Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act) SWD (2022) 34 final, 6. 19 Rossana Ducato and Alain M Strowel, ‘Ensuring Text and Data Mining: Remaining Issues with the EU Copyright Exceptions and Possible Ways Out’ (2021) 43 E.I.P.R. 322, 333. I. Introduction The sui generis protection functions inde- pendently of the copyright protection of databases that protects databases that, because of the selection or arrangement of their contents, constitute the authorʼs own intellectual creation.11 Unlike the latter, the sui generis database right, provided under Art. 7 of the Directive on the legal protection of databases (Database Directive),12 may apply irrespective of whether the data- base or the contents of that database meet the original- ity requirement and could be protected by copyright.13 The provision protects the maker of a database who has made qualitatively or quantitatively a substantial invest- ment in either the obtaining, verification or presentation of the contents of the database.14 Against this background, if the CDSM Directive’s TDM exceptions prove to be ineffective in addressing the needs of researchers, scientists may have to explore alter- native solutions to lawfully conduct TDM on sui gener- is-protected databases. For instance, the situation can be improved under the proposed Regulation on harmonised rules on fair access to and use of data (the proposed Data Act),21 which in its current form limits the scope of sui generis protection to a certain extent. This legislative ini- tiative was introduced on 23 February 2022 as a part of the European data strategy. The proposed Data Act addresses problems related to the extension of the sui generis right to machine-generated databases.22 In partic- ular, Art. 35 of this legislation clarifies that the sui generis protection should not apply to databases ‘containing data obtained from or generated by the use of a prod- uct or a related service’.23 The provision could provide a legal ground for scientists to lawfully conduct TDM on databases consisting of data automatically generated Therefore, if researchers intend to analyze a large cor- pus of materials protected under the sui generis regime, they must either obtain permission from the righthold- ers through contractual agreements or rely on applicable exceptions or limitations to avoid infringement. Although the Database Directive provides some exceptions to the sui generis protection under Arts. 8(1) and 9(b), the pro- visions are neither practical nor effective in the case of TDM. Researchers could encounter legal uncertainty and challenges in complying with specific conditions and requirements stipulated under these exceptions. Two spe- cific TDM exceptions to copyright and the sui generis right introduced under Arts. 23 Data Act, art 35. See for similar provisions Data Governance Act, art 5(7) and the 2019 Open Data Directive, art 1(6). 11 Directive 96/9 on the legal protection of databases [1996] OJ L 77/20 (Database Directive), art 3(1). 12 Directive 96/9 on the legal protection of databases [1996] OJ L 77/20. 13 Database Directive, art 7(4). See also Matthias Leistner and Lucie Antoine, ‘Attention, Here Comes the EU Data Act! A Critical In-depth Analysis of the Commission’s 2022 Proposal’ (2022) 13 JIPITEC 339, 342. 14 Database Directive, art 7(1). 10 Brady D Lund and Yi-Shun Wang, ‘Chatting about ChatGPT: How May AI and GPT Impact Academia and Libraries?’ (2023) 40 Library Hi Tech News 26, 27. 15 Directive 2019/790 on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC [2019] OJ L 130/92 (CDSM Directive). 18 Christophe Geiger and Bernd Justin Jütte, ‘Conceptualizing a ‘Right to Research’ and Its Implications for Copyright Law: An International and European Perspective’ (2022) Joint PIJIP/TLS Research Paper Series 7/2022, 1, 13 <https://digitalcommons.wcl.american.edu/cgi/viewcon- tent.cgi?article=1079&context=research> accessed 27 February 2023. 13 Database Directive, art 7(4). See also Matthias Leistner and Lucie Antoine, ‘Attention, Here Comes the EU Data Act! A Critical In-depth Analysis of the Commission’s 2022 Proposal’ (2022) 13 JIPITEC 339, 342. 11 Directive 96/9 on the legal protection of databases [1996] OJ L 77/20 (Database Directive), art 3(1). 17 CDSM Directive, art 4. 14 Database Directive, art 7(1). 16 CDSM Directive, art 3. 10 Brady D Lund and Yi-Shun Wang, ‘Chatting about ChatGPT: How May AI and GPT Impact Academia and Libraries?’ (2023) 40 Library Hi Tech News 26, 27. 10 Brady D Lund and Yi-Shun Wang, ‘Chatting about ChatGPT: How May AI and GPT Impact Academia and Libraries?’ (2023) 40 Library Hi Tech News 26, 27. I. Introduction 3 and 4 of the Directive on Copyright in the Digital Single Market (the CDSM 18 Christophe Geiger and Bernd Justin Jütte, ‘Conceptualizing a ‘Right to Research’ and Its Implications for Copyright Law: An International and European Perspective’ (2022) Joint PIJIP/TLS Research Paper Series 7/2022, 1, 13 <https://digitalcommons.wcl.american.edu/cgi/viewcon- tent.cgi?article=1079&context=research> accessed 27 February 2023. 21 Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act), Brussels [2022] COM (2022) 68 final (proposed Data Act). 22 Commission Staff Working Document, Impact Assessment Report Accompanying the document Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act) SWD (2022) 34 final, 6. 23 Data Act, art 35. See for similar provisions Data Governance Act, art 5(7) and the 2019 Open Data Directive, art 1(6). Maryna Manteghi 36 by Internet of Things (IoT)24 technologies. Even though this may not remedy the situation, researchers could ben- efit from more legal certainty regarding the mining of non-original databases. from a quantitative or qualitative perspective.29 The for- mer refers to quantifiable resources (e.g. time and money) and the latter to resources which are difficult or impos- sible to measure using numerical values (e.g. intellectual effort or energy).30 The Court of Justice of the European Union (CJEU) has clarified that a ‘substantial investment’ should be assessed by taking into account only investments made in the creation of the database as such.31 Therefore, the resources utilized for the creation of data which con- stitutes the contents of a database would be excluded from the assessment in terms of Art. 7(1) of the Database Directive.32 This derives from the purpose of sui generis protection, which is to promote the establishment of storage and processing systems for existing data and not the creation of works that can be accumulated and sub- sequently organized in a database.33 In other words, the protection of databases under the sui generis regime relates to the protection of the database as a whole, rather than protecting individual data within it.34 This article aims to contribute to the ongoing discus- sions on TDM by assessing the perspectives of Art. 35 of the proposed Data Act to alleviate tension between sui generis database holders and researchers. 32 The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) paras 31-49; Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) paras 28-42; Fixtures Marketing Ltd v Svenska Spel AB (n 29) paras 23-37 and Fixtures Marketing Ltd v Organismos prog- nostikon agonon podosfairou AE (n 29) paras 37-53. 31 Case C-203/02 The British Horseracing Board Ltd and Others v William Hill Organization Ltd ECLI:EU:C:2004:695, paras 31-49; Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) paras 28-42; Fixtures Marketing Ltd v Svenska Spel AB (n 29) paras 23-37; and Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (n 29) paras 37-53. 29 See Case C-444/02 Fixtures Marketing Ltd v Organismos Prognostikon Agnon Podosfairou ECLI:EU:C:2004:697, para 44; Case C-338/02 Fixtures Marketing Ltd v Svenska Spel AB ECLI:EU:C:2004:696, para 28; Case C-46/02 Fixtures Marketing Ltd v Oy Veikkaus Ab ECLI:EU:C:2004:694, para 38; Database Directive recitals 7, 39 and 40; Kuschel and Dolling (n 27) 253. 30 See Fixtures Marketing Ltd v Organismos Prognostikon Agnon Podosfairou (n 29) para 44; Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 28; Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) para 38. 25 ‘Database’ can be defined as a ‘collection of independent works, data or other materials arranged in a systematic or methodical way and indi- vidually accessible by electronic or other means’; Database Directive, art 1(2). 24 IoT is ‘an emerging paradigm that enables the communication between electronic devices and sensors through the internet in order to facilitate our lives’; Sachin Kumar, Prayag Tiwari and Mikhail Zymbler, ‘Internet of Things is a Revolutionary Approach for Future Technology Enhancement: A Review’ (2019) 6 Journal of Big Data 1, 1. 33 Database Directive, Recital 12; Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) paras 33-34; Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 24; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 40. 27 Database Directive, art 7(4). See also Linda Kuschel and Jasmin Dolling, ‘Access to Research Data and EU Copyright Law’ (2022) 13 JIPITEC 247. I. Introduction This objective was moti- vated by the argument that the current regulation of TDM tends to prioritize the rights and interests of rightholders at the expense of TDM users, creating an imbalance that can hinder scientific research and innovation in the EU. Therefore, the structure of this paper is as follows. After the introduction (Part I), the article will explore how sui generis database protection could restrict TDM research in the EU. In addition, this section will examine how the exceptions provided under the Database Directive and the CDSM Directive could potentially mitigate these restric- tions (Part II). Next, the article will analyze the proposed Data Act within the framework of TDM regulation. The paper will examine how Art. 35 of the proposed Data Act could impact the application of TDM in research (Part III). Finally, the article will summarize general findings and crit- icisms of the current legal framework on TDM, highlight- ing potential solutions for researchers intending to analyze sui generis-protected databases by using AI tools (Part IV). Further, the CJEU has also clarified the terms ‘obtain- ing’, ‘verification’ or ‘presentation’ of the contents of the database. The first concept addresses the users’ intention ‘to seek out existing independent materials and collect them in the database …’.35 The term ‘verifying’ is under- stood as ‘ensuring the reliability of the information con- tained in that database, to monitor the accuracy of the materials collected when the database was created and during its operation.’36 And finally, the term ‘presenting’ refers to ‘the resources used for the purpose of giving the database its function of processing information, that is to say those used for the systematic or methodical arrange- ment of the materials contained in that database and the organisation of their individual accessibility.’37 26 See eg, Marny Lopez, ‘ChatGPT: Different Uses and Examples’ (Devlane, February 2023) <https://www.devlane.com/blog/chatgpt-dif- ferent-uses-and-examples> accessed 20 April 2023. 35 Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 24; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 40. 37 Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) para 37; Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 27; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 43. 34 Kuschel and Dolling (n 27) 255. 36 Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) para 37; Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 27; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 43. 28 Database Directive, art 7(1). 1. The impact of sui generis database protection on TDM Overcoming Barriers to Text and Data Mining in the Era of ChatGPT 37 37 The Database Directive confers two transferable rights to sui generis database rightholders: the right to extraction and the right to re-utilization.38 The former refers to the transfer of the contents of a database to another medium while the latter relates to the act of making the contents of a database available to the public.39 The sui generis database rightholder is entitled to prohibit ‘extraction or re-utiliza- tion of the whole or of a substantial part, evaluated quali- tatively or quantitatively, of the contents of that database.’40 The CJEU has clarified the concept of a ‘substantial part’ by relying on the scale and volume of the investment made in the part of a database that has been extracted or re-utilized, irrespective of whether that part constitutes a substantial portion of the general contents of such database.41 The assessment must refer to the harm that the acts of extract- ing or re-utilizing cause to that investment.42 In addition, the sui generis database rightholder can prohibit the use of insubstantial parts of the contents of the database when it is performed in a repeated and systematic manner.43 Such use is infringing when it conflicts with a normal exploitation of that database, or when it unreasonably prejudices the legit- imate interests of the creator of the database.44 The CJEU clarifies that these conditions refer to unauthorized acts for the purpose of reconstituting, through the cumulative effect of acts of extraction, the entire or a substantial part of the contents of a database, which could seriously prejudice the investment made by the creator of the database.45 Therefore, regardless of whether the act of transfer is for the purpose of creating another database in the new medium or merely analyzing the contents of the original database, as in the case of TDM, both actions may consti- tute the act of extraction.48 However, if TDM is deemed to infringe the sui generis right to extraction, it should be regarded as an unau- thorized appropriation of mere facts and data from a database.49 As Ducato and Strowel reasonably claim, researchers ‘consult’ databases for informational pur- poses only, and any reconstitution of a database which may occur during the copying stage of the TDM process is necessary solely for accessing incorporated data and extracting new insights.50 The CJEU has already clar- ified that the sui generis right to extraction should not apply to the right to consult a database, which is a fun- damental right that must be weighed against the interests of the database owner.51 In this sense, Recital 46 of the Database Directive also clarifies that the right to prevent unauthorized extraction should not lead to the creation of a property right on the information itself that could limit access to data included in the database.52 Against this background, TDM should be viewed only as an act of ‘consultation’ rather than the act of ‘extraction’ of the contents of sui generis-protected databases. 1. The impact of sui generis database protection on TDM Researchers may employ TDM to extract new knowledge from existing large databases.25 For instance, ChatGPT can be used to analyze a corpus of financial data to predict stock prices, a corpus of legal cases to predict the outcomes of future cases, a corpus of legal texts to identify patterns and biases in legal language or a corpus of network traf- fic data to identify unusual patterns that may indicate a cyber-attack or other security threat.26 Even if these data- bases fail to meet the originality requirement for copyright protection, researchers could potentially be held liable for mining them under the sui generis database protection.27 The sui generis database right provides protection for the maker of a database, who has made a substantial invest- ment in either the obtaining, verification or presentation of the contents of the database.28 The investment could be financial, human or technical, and it must be substantial 33 Database Directive, Recital 12; Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) paras 33-34; Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 24; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 40. 37 Fixtures Marketing Ltd v Oy Veikkaus Ab (n 29) para 37; Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 27; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 43. Veikkaus Ab (n 29) para 35; The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) paras 32, 45, 46 and 51; Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 25; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 41; Case C-545/07 Apis-Hristovich EOOD v Lakorda AD ECLI:EU:C:2009:132, para 40; Case C-202/12 Innoweb BV v Wegener ICT Media BV and Wegener Mediaventions BV ECLI:EU:C:2013:850, paras 33-34, 38. 41 The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) paras 68-71. 42 Database Directive, recital 42. See also The British Horseracing Board Ltd and Others v William Hill Organization Ltd. (n 31) para 69. 43 Database Directive, art 7(5). See also The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) paras 83-95 and Case C-304/07 Directmedia Publishing GmbH v Albert-Ludwigs- Universität Freiburg ECLI:EU:C:2008:552, paras 43-44. Case C-762/19 SIA ‘CV-Online Latvia’ v SIA ‘Melons’ ECLI:EU:C:2021:434, para 47. 38 Database Directive, art 7(2)(a), (b). 39 Database Directive, art 7(2)(a), (b). 40 Database Directive, art 7(1). 52 Database Directive, recital 46. Directmedia Publishing GmbH v Albert-Ludwigs-Universität Freiburg (n 43) paras 51-54; The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) para 55; Ducato and Strowel, ‘Ensuring Text and Data Mining: Remaining Issues with the EU Copyright Exceptions and Possible Ways Out’ (n 19) 351. 45 The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) para 89. 51 See The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) para 54; Directmedia Publishing GmbH v Albert-Ludwigs-Universität Freiburg (n 43) paras 51-54. 48 See Directmedia Publishing GmbH v Albert-Ludwigs-Universität Freiburg (n 43) paras 39, 46, 47 (referring to Recital 44 of the Database Directive and to The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) paras 47-48). 47 Directmedia Publishing GmbH v Albert-Ludwigs-Universität Freiburg (n 43) paras 31-33 relying on Fixtures Marketing Ltd v Oy 44 Database Directive, art 7(5). See also The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) paras 83-95; Directmedia Publishing GmbH v Albert-Ludwigs-Universität Freiburg (n 43) paras 43-44 and SIA ‘CV-Online Latvia’ v SIA ‘Melons’ (n 43) para 47. 49 Ducato and Strowel, ‘Ensuring Text and Data Mining: Remaining Issues with the EU Copyright Exceptions and Possible Ways Out’ (n 19) 351. 46 Database Directive, art 7(2)(a). 50 ibid 350. 53 Database Directive, art 8(1). 2. Applying the CDSM Directive’s TDM exceptions to ChatGPT-based research 2. Applying the CDSM Directive’s TDM exceptions to ChatGPT-based research The requirement of a ‘lawful user’ is also a condition of another potential exception provided under Art. 9(b) of the Database Directive that could apply to TDM. The provision allows extraction or re-utilization of a sub- stantial part of the contents of a database for the pur- pose of illustration for teaching or scientific research.57 However, the exception consists of requirements which are difficult to meet in the case of TDM. First, the pro- vision requires that users indicate the source of a work when they extract the contents of a database. However, it could be difficult in practice to attribute huge amounts of digital data utilized for mining. The output of TDM research does not contain any parts (samples) of mined works, but only new patterns and correlations. Therefore, it is unlikely that the maker of a processed database would obtain any significant benefits from being listed among thousands of sources. Second, Art. 9(b) requires that the extraction of a substantial part of a database is justified by a non-commercial purpose. This require- ment would deprive private actors of the possibility to conduct TDM research under this exception as they usually pursue, directly or indirectly, commercial pur- poses.58 The condition would also affect many research organizations conducting research in the framework of public-private partnerships, as it is in practice, difficult to distinguish between commercial and non-commercial activities within these collaborations.59 Moreover, the optional nature of this provision has led to a fragmented or divergent implementation of the exception within the EU Member States.60 Researchers engaging in TDM research on databases, including those using AI language models, such as ChatGPT, would probably benefit from specific TDM exceptions provided under Arts. 3 and 4 of the CDSM Directive. The former provision allows research organiza- tions and cultural heritage institutions to reproduce and extract works for the purpose of text and data mining, providing that they have lawful access to the works and the mining is for scientific research purposes.61 The lat- ter provision is intended to compensate for the narrow scope of Art. 55 The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) para 58. See for more discussions on the con- cept of ‘lawful user’ eg, Tatiana Eleni Synodinou, ‘Lawfulness for Users in European Copyright Law: Acquis and Perspectives’ (2019) 10 JIPITEC 20. 61 CDSM Directive, art 3(1). 62 CDSM Directive, art 4(1). 63 Documentation for Python’s standard library, along with tutorials and guides, is available at <https://www.python.org/> accessed 12 March 2023. 56 Christophe Geiger, Giancarlo Frosio and Oleksandr Bulayenko, ‘The Exception for Text and Data Mining (TDM) in the Proposed Directive on Copyright in the Digital Single Market ‒ Legal Aspects’ (2018) Centre for International Intellectual Property Studies Research Paper No 2018-02, 1, 16 <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3160586> accessed 20 April 2023. 54 Maryna Manteghi, ‘The Insufficiency of the EU’s Text and Data Mining Exceptions for Using Artificial Intelligence’ (2022) 44 E.I.P.R. 657. 1. The impact of sui generis database protection on TDM However, as long as a research process involves copying, the risk of infringement exists. Therefore, it could reasonably be argued that the sui generis protection restricts TDM by intensifying the control over access to data. The concept of ‘re-utilization’ is less relevant to TDM compared to the term ‘extraction’. This is because the findings of TDM research, which are made publicly available through research reports, scientific articles, or conference papers, do not usually include any parts of the mined works. The right to extraction involves ‘the permanent or temporary transfer of all or a substantial part of the contents of a database to another medium by any means or in any form’.46 The CJEU has held that the concept of ‘extraction’ should be broadly interpreted to protect the makers of a database from the unauthorized appropriation of the results of their investment by acts of the reconstitution of that database or a substantial part of it.47 The purpose of the transfer should be irrelevant when determining whether an act constitutes ‘extraction’ within the meaning of Art. 7(2)(a) of the Database Directive. Therefore, if researchers intend to analyze a large cor- pus of materials protected under the sui generis regime, they must either obtain permission from the righthold- ers through contractual agreements or rely on applica- ble exceptions or limitations to avoid infringement. The Database Directive provides a few exceptions to the sui generis right which could potentially apply to TDM. The first possible candidate is the exception for a ‘lawful user’ provided under Art. 8(1) of the Directive. The pro- vision permits extraction or re-utilization of insubstantial parts of the contents of a database, evaluated qualita- tively or quantitatively, for any purpose.53 However, this 53 Database Directive, art 8(1). Maryna Manteghi 38 exception could be regarded as ineffective and worthless for TDM users as the successful training of algorithms during a mining process requires the copying of a whole or substantial part of accessed databases.54 Moreover, the exception applies only to ‘lawful users’ of a database who can access the database either through contractual agree- ments or through relevant exceptions.55 In this regard, the rightholders have control over the contents of a database, as they would be able to restrict or even prohibit TDM through licensing for the sake of their own interests.56 facing legal uncertainty under these provisions may find a remedy in Arts. 58 Rossana Ducato and Alain M Strowel, ‘Limitations to Text and Data Mining and Consumer Empowerment: Making the Case for a Right to ‘Machine Legibility’’ (2019) 50 IIC 649, 661. 59 Maria Bottis and others, ‘Text and Data Mining in the EU Acquis Communautaire Tinkering with TDM & Digital Legal Deposit’ (2019) 2 Erasmus Law Review 190, 193. 1. The impact of sui generis database protection on TDM 3 and 4 of the CDSM Directive, which introduce specific TDM exceptions to copyright and the sui generis database right. The next section of this article elaborates on the potential application of such exceptions to ChatGPT-based research. 57 Database Directive, art 9(b). 60 Geiger, Frosio and Bulayenko (n 56) 12. 2. Applying the CDSM Directive’s TDM exceptions to ChatGPT-based research ChatGPT, when integrated with Python, could help researchers create a more sophisticated and intelli- gent application of the medical research in question.64 the human sciences’.68 The research in question would likely fall within this categorization of the fields of sci- ence, as medical research is part of both branches and would also contribute to the current state of science.69 However, in order to qualify as scientific research, the research project in question should employ scientific methods which should satisfy many characteristics such as replicability, precision, falsifiability, objectivity, reli- ability, parsimony and others.70 Therefore, the research in question would not be considered scientific if it is unable to explore medical databases using appropriate research methods.71 To lawfully mine medical databases, presumably pro- tected under the sui generis database regime, research- ers should obtain permission from rightholders through contractual agreements or rely on the TDM exception in Art. 3 of the CDSM Directive. However, not all researchers may conduct research under this exception, even though the provision was specifically designed for researchers. Only research organizations and cul- tural heritage institutions can be the beneficiaries of this exception. This may include universities, research institutions, libraries, museums, or any other entity con- ducting scientific research on a nonprofit basis or that reinvests all the profits in its scientific research or pur- suant to a public interest mission.65 Therefore, since pri- vate actors, such as individual researchers, journalists, small and medium-sized enterprises (SMEs), and start- ups, are usually profit-oriented, TDM research could only be carried out by universities or similar institutions under this provision. Even if research organizations prove that the research in question is scientific and their access to databases is lawful, they could still be deprived of the possibility to mine due to the possible application of technological protection measures (TPMs). Article 3(3) of the CDSM Directive allows rightholders to employ such measures only to ensure the security and integrity of their systems and databases. 2. Applying the CDSM Directive’s TDM exceptions to ChatGPT-based research 3, and as a result it benefits any type of ‘law- ful’ user and covers any purposes provided that the use of works has not been reserved by their rightholders.62 Therefore, researchers are required to comply with cer- tain conditions and limitations to enjoy the TDM excep- tions under the CDSM Directive. Considering real-life scenarios can provide more tan- gible insight into how these exceptions might function in practice and the effects they could have on researchers. As an example, we could consider TDM research that would employ ChatGPT to analyze large databases of various types of medical data to uncover new insights and patterns with the aim of improving the accuracy and efficiency of medical diagnosis and treatment. This research project would include several essential stages. Initially, research- ers need to obtain databases of health-related data either from publicly available sources or through collaboration with medical institutions. Once medical data is collected, researchers need to employ TDM tools to pre-process, clean and transform it into a suitable format for data analysis. Python, a commonly used programming lan- guage, is one of the popular tools that can be employed for such tasks. Therefore, researchers could upload the obtained databases into Python through its libraries, and pre-process them in a format that would meet the input requirements of the ChatGPT model. Once the data is pre-processed, researchers need to load ChatGPT into Python by using the Hugging Face Transformers library to perform data analysis.63 The ChatGPT model can be trained on databases for a variety of natural language processing applications including text classification, clus- tering, summarization, information retrieval and others. This would help researchers uncover, for instance, recur- ring medical conditions or symptoms, or identify patterns in the language used in the medical discharge summaries, which could improve medical diagnosis and treatment. The determination of the specific data analysis technique To sum up, even though the Database Directive pro- vides some exceptions to the sui generis protection under Arts. 8(1) and 9(b), the provisions are neither practical nor effective in the case of TDM. In this regard, researchers 60 Geiger, Frosio and Bulayenko (n 56) 12. Overcoming Barriers to Text and Data Mining in the Era of ChatGPT 39 depends on the type of processed data and the research purpose. 75 Case C-360/10 Belgische Vereniging van Auteurs, Componisten en Uitgevers CVBA (SABAM) v Netlog NV ECLI:EU:C:2012:85, paras 50 and 52; Case C-70/10 Scarlet Extended SA v Société belge des auteurs, compositeurs et éditeurs SCRL (SABAM) ECLI:EU:C:2011:771, para 50; Case C-401/19 Republic of Poland v European Parliament and Council of the European Union ECLI:EU:C:2022:297, para 86; Ahmet Yıldırım v Turkey App No 3111/10 (ECtHR, 18 December 2012), paras 66-68; Kharitonov v Russia App No 10795/14 (ECtHR, 23 June 2020), paras 38-40; See also Julia Reda, Joschka Selinger and Michael Servatius, ‘Article 17 of the Directive on Copyright in the Digital Single Market: A Fundamental Rights Assessment’ 26-27 <https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=3732223> accessed 1 May 2023. 64 I Gencay, ‘ChatGPT and AI Merged in Data Science with Python’ (DataDrivenInvestor, 13 April 2023) <https://medium.datadriven- investor.com/chatgpt-and-ai-merged-in-data-science-with-python- 6ca8c2cb387> accessed 2 May 2023. 65 CDSM Directive, art 2(1) and (3). 66 Geiger and Jütte (n 18) 44. Ducato and Strowel, ‘Limitations to Text and Data Mining and Consumer Empowerment: Making the Case for a Right to ‘Machine Legibility’’ (n 58) 651. 67 For more discussions on what can be considered ‘science’ see Anol Bhattacherjee, ‘Social Science Research: Principles, Methods, and Practices’ (Textbooks Collection 3, University of South Florida 2012) <https://digitalcommons.usf.edu/cgi/viewcontent.cgi?article=1002&con- text=oa_textbooks> accessed 2 March 2023. 76 LIBER, ‘Europe’s TDM Exception for Research: Will It Be Undermined By Technical Blocking From Publishers?’ (Libereurope, 10 March 2020) <https://libereurope.eu/article/tdm-technical-protec- tion-measures/> accessed 30 April 2023. Republic of Poland v European Parliament and Council of the European Union (n 75) para 60. 74 Tatiana-Eleni Synodinou, ‘‘Intermediaries’ Liability for Copyright Infringement in the EU: Evolutions and Confusions’ (2015) 31 Computer Law & Security Review 57, 62. 68 CDSM Directive, recital 12. 69 Brian Kennett, Planning and Managing Scientific Research: A guide for the beginning researcher (ANU Press 2014) 2. 70 Bhattacherjee (n 67). 71 Kennett (n 69) 2. 72 CDSM Directive, art 3(3) and recital 16. 73 CDSM Directive, recital 16. Case C-484/14 Tobias Mc Fadden v Sony Music Entertainment Germany GmbH ECLI:EU:C:2016:689, paras 94, 98-99. 74 Tatiana-Eleni Synodinou, ‘‘Intermediaries’ Liability for Copyright Infringement in the EU: Evolutions and Confusions’ (2015) 31 Computer Law & Security Review 57, 62. 75 Case C-360/10 Belgische Vereniging van Auteurs, Componisten en Uitgevers CVBA (SABAM) v Netlog NV ECLI:EU:C:2012:85, paras 50 and 52; Case C-70/10 Scarlet Extended SA v Société belge des auteurs, compositeurs et éditeurs SCRL (SABAM) ECLI:EU:C:2011:771, para 50; Case C-401/19 Republic of Poland v European Parliament and Council of the European Union ECLI:EU:C:2022:297, para 86; Ahmet Yıldırım v Turkey App No 3111/10 (ECtHR, 18 December 2012), paras 66-68; Kharitonov v Russia App No 10795/14 (ECtHR, 23 June 2020), paras 38-40; See also Julia Reda, Joschka Selinger and Michael Servatius, ‘Article 17 of the Directive on Copyright in the Digital Single Market: A Fundamental Rights Assessment’ 26-27 <https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=3732223> accessed 1 May 2023. 76 LIBER, ‘Europe’s TDM Exception for Research: Will It Be Undermined By Technical Blocking From Publishers?’ (Libereurope, 10 March 2020) <https://libereurope.eu/article/tdm-technical-protec- tion-measures/> accessed 30 April 2023. Republic of Poland v European Parliament and Council of the European Union (n 75) para 60. 73 CDSM Directive, recital 16. Case C-484/14 Tobias Mc Fadden v Sony Music Entertainment Germany GmbH ECLI:EU:C:2016:689, paras 94, 98-99. 82 Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the dig- ital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (Digital Markets Act) (Text with EEA relevance) [2022] OJ L 265, 1-66. 83 Regulation (EU) 2022/868 of the European Parliament and of the Council of 30 May 2022 on European data governance and amending Regulation (EU) 2018/1724 (Data Governance Act) (Text with EEA rel- evance) [2022] OJ L 152, 1-44. 77 CDSM Directive, art 4(3) and recital 18. 78 Bernt Hugenholtz, ‘The New Copyright Directive: Text and Data Mining (Articles 3 and 4)’ (Kluwer Copyright Blog, 24 July 2019) <http:// copyrightblog.kluweriplaw.com/2019/07/24/the-new-copyright-direc- tive-text-and-data-mining-articles-3-and-4/> accessed 12 April 2023. 79 CDSM Directive, art 7(1). 81 Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions A European Strategy for Data, COM/2020/66 final. 84 Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market For Digital Services and amending Directive 2000/31/EC (Digital Services Act) (Text with EEA relevance) [2022] OJ L 277, 1-102. 2. Applying the CDSM Directive’s TDM exceptions to ChatGPT-based research 4 of the CDSM Directive, which was designed to benefit all users seeking to engage in mining. However, they may benefit from this exception unless rightholders reserve the use of their works, either through contractual agreements or through the terms and conditions of their websites, or by implementing TPMs.77 As Hugenholtz argues, the ‘opt-out’ mechanism of Art. 4 gives rightholders exces- sive power to limit or completely prohibit TDM.78 This may impede research and innovation in different fields, including healthcare and medicine. Moreover, even if rightholders do not reserve the use of their works in an appropriate manner, they may still prevent TDM through licensing terms since, unlike Art. 3, this excep- tion can be overridden by a contract.79 To sum up, the vague and narrow nature of the CDSM Directive’s TDM exceptions could force users to con- sider alternative solutions to avoid potential infringement claims. One of these solutions could be Art. 35 of the proposed Data Act, which clarifies the scope of applica- tion of the sui generis right provided for by the Database Directive. In particular, the provision excludes certain types of databases from the protection under this regime. Therefore, the next chapter will explore the extent to which Art. 35 of the proposed regulation can remedy the situation in question. 2. Applying the CDSM Directive’s TDM exceptions to ChatGPT-based research Such measures could, for instance, be applied to ensure that only users with lawful access to mined data can access them.72 However, if TPMs such as IP address validation or user authentication could be viewed as necessary for controlling access to protected databases,73 technological measures such as IP address blocking or domain name server blocking may lead to overblocking (blocking of lawful access).74 As auto- mated filtering technologies cannot properly distinguish between lawful and unlawful access, the ‘lock-outs’ may occur frequently and take lots of time to resolve.75 This may affect the research in question and lead to unde- sirable consequences. If researchers exploring medical databases encounter blockages, they may have to sus- pend research for weeks or even months until they are reconnected to the digital repositories. The delays could require additional costs and time (e.g. researchers would need to continue receiving their salaries) and could make research outdated and irrelevant.76 In theory, researchers wishing to conduct the research in question Therefore, private clinics would not be able to mine databases containing health-related data for the purpose of the research in question, without permission from rightholders. Would it be appropriate to prevent such users from conducting TDM research on databases to which they have obtained lawful access through con- tractual agreements or other lawful means? Why should they pay additional fees to merely analyze such data- bases? Both commercial and non-commercial research organizations could produce valuable, informative outputs, which are crucial for economic and techno- logical developments.66 Society will often benefit from profit-driven research. For instance, private clinics may develop predictive models for diagnosing certain con- ditions that would improve treatment outcomes, and potentially reduce healthcare costs (e.g. excluding some tests and procedures) and also the time that patients spend in the hospital. Further, researchers relying on the exception in Art. 3 should take into account that the provision allows researchers to employ TDM tools only for the purpose of scientific research. Therefore, researchers would need to prove that their research on medical databases qualifies as scientific research. The process is complex because sci- entific research is characterized by different approaches, categories and styles.67 The CDSM Directive does not provide much clarification of this concept, merely stat- ing that it should cover ‘both the natural sciences and Maryna Manteghi 40 may rely on another TDM exception in Art. III. The proposed data act in the context of TDM Regulation Against this background, both research organizations and private clinics seeking to mine databases that con- tain health-related data could face challenges and legal uncertainty when relying on the CDSM Directive’s TDM exceptions. The passive and unclear nature of these pro- visions could potentially reduce the research power of the EU. In this regard, there is a need to expand the research TDM exception to commercial research, or even introduce a wider TDM exception which would allow TDM for any purpose without a rightholders ‘opt-out’ mechanism, to improve the situation. A broad TDM exception would ensure greater access to data by researchers, journalists, commercial companies, AI developers and other stakeholders. The improved legal framework for TDM would stimulate new discoveries and advances in various fields that could contribute to economic and technological development, and thus foster social progress and wider innovation. However, broadening the existing TDM exceptions may raise concerns that copyright and database owners could become less motivated to produce new content since they would have less control over their works. The con- cerns appear to be groundless, as rightholders would still rely on the ‘lawful access’ requirement, and thus be able to receive adequate compensation for the use of their works through licensing, subscriptions or other lawful means. Moreover, copyright and database own- ers would still have the opportunity to apply TPMs to ensure the security and integrity of their systems and databases. The said safeguards, therefore, would help rightholders obtain financial gain and also protect their content. Furthermore, the introduction of a wider TDM exception would encourage many investors and business companies to set up and run their businesses within the digital single market (DSM). As a result, the expanding application of advanced analytical tools such as TDM and ChatGPT would significantly increase the demand for data access. This would have to encourage content creators to produce vast amounts of original and quality data as they would still gain licensing revenue for allow- ing access to such content. 85 Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) and Amending Certain Union Legislative Acts, COM/2021/206 final. 80 Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act), Brussels [2022] COM (2022) 68 final (proposed Data Act). 80 Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act), Brussels [2022] COM (2022) 68 final (proposed Data Act). 87 Proposed Data Act, cc I-IV and 88 Proposed Data Act, c V. 89 Proposed Data Act, c VI. 90 Proposed Data Act, c VIII. 91 Proposed Data Act, c IV. 92 Proposed Data Act, art 5. 105 European Commission, ‘Impact Assessment Report’ Accompanying the document Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act), SWD (2022) 34 final, 15; Graef and Husovec (n 104) 4. In this sense, see also Federal Supreme Court (BGH), 25 March 2010, I ZR 47/08 ‒ Autobahnmaut. and Jukka Mähönen (eds), Promoting Sustainable Innovation and the Circular Economy: Legal and Economic Aspects (Routledge 2022) 26; Derclaye and Husovec (n 86) 2; Inge Graef and Martin Husovec, ‘Seven Things to Improve in the Data Act’ 1, 4 <https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=4051793> accessed 12 March 2023. 2. The application of Art. 35 to TDM research 2. The application of Art. 35 to TDM research Researchers using ChatGPT to conduct TDM research on medical databases are likely to process different types of databases comprised of data generated by humans,107 machines,108 or a combination of both.109 Article 35 pro- vides a safe harbor for users needing to access, use or share only databases made of data generated by the use of a product or related service (IoT data).110 This would allow researchers to freely access and analyze databases which consist of data generated by, for instance, wearable devices for healthcare monitoring such as fitness trackers, smartwatches, blood glucose monitors, smart technol- ogy clothing and others. This would enable researchers to identify fresh insights and patterns to develop new treatments, improve diagnosis, and optimize preventive care.111 Even though the proposed Data Act intends to free machine-generated data from the constraints of sui generis protection, some aspects of the proposed regula- tion may require further clarification to ensure a greater and fairer flow of data in all sectors, including research and related fields.112 Although Art. 35 intends to provide more clarity on the protection of machine-generated raw data, the Therefore, the proposed regulation aims, inter alia, to clarify what is not protected under the sui generis regime without expressly amending the Database Directive.104 95 European Commission, ‘Study to Support an Impact Assessment for the Review of the Database Directive Final Report’ 1 <https://copen- hageneconomics.com/wp-content/uploads/2022/02/study-to-support- an-impact-assessment-for-the-review-of-the-database-directive.pdf> accessed 12 March 2023. 96 Giuseppe Colangelo, ‘European Proposal for a Data Act-A First Assessment’ (20 July 2022) CERRE Evaluation Paper 2022, 1, 6 <https:// papers.ssrn.com/sol3/papers.cfm?abstract_id=4199565> accessed 10 April 2023; European Commission, ‘Study to Support an Impact Assessment for the Review of the Database Directive Final Report’ (n 95) 1. 97 ibid 1 and 2. 98 ibid 2. 99 Proposed Data Act, art 35. 100 Proposed Data Act, art 35. 101 Proposed Data Act, art 2(2). 102 Proposed Data Act, art 2(2). 103 Proposed Data Act, recital 84. 104 Guido Noto La Diega and Estelle Derclaye, ‘Opening Up Big Data for Sustainability: What Role for Database Rights in the Fourth Industrial Revolution?’ in Ole-Andreas Rognstad, Taina Pihlajarinne 93 Proposed Data Act, art 14. 94 Proposed Data Act, art 15. 1. Background h d for private companies to make data available to public sector bodies to satisfy exceptional needs,93 including the prevention of a public emergency.94 Access to data and the ability to use it lie at the core of the EU data policy for promoting innovation and fostering economic growth in the time of increasing use of digital technologies and the internet.95 The proposal comes with a special emphasis on broadening access to and use of data collected by sen- sors and machines based on cutting-edge technologies, such as IoT.96 The regulation reviews the Database Directive as a part of its broader policy initiative to ensure that the Directive remains relevant in the data- driven society, without impeding access to and use of data within the EU.97 Particularly, the proposed Data Act aims to get rid of legal uncertainty regarding the relation of the sui generis right to machine-generated data.98 However, only one provision, namely Art. 35, addresses this issue.99 The provision provides that such protection ‘does not apply to databases containing data obtained from or generated by the use of a product or a related service’.100 The proposed Data Act defines a ‘product’ as ‘a tangible, movable item, including where incorporated in an immovable item, that obtains, gener- ates or collects, data concerning its use or environment, and that is able to communicate data via a publicly available electronic communications service and whose primary function is not the storing and processing of data.’101 Further, the proposed regulation describes the term ‘related service’ as ‘a digital service, including soft- ware, which is incorporated in or inter-connected with a product in such a way that its absence would prevent the product from performing one of its functions’.102 The proposed Data Act includes only one Recital that relates to Art. 35, namely Recital 84 which further emphasizes that the ‘regulation should clarify that the sui generis right does not apply to such databases as the require- ments for protection would not be fulfilled.’103 96 Giuseppe Colangelo, ‘European Proposal for a Data Act-A First Assessment’ (20 July 2022) CERRE Evaluation Paper 2022, 1, 6 <https:// papers.ssrn.com/sol3/papers.cfm?abstract_id=4199565> accessed 10 April 2023; European Commission, ‘Study to Support an Impact Assessment for the Review of the Database Directive Final Report’ (n 95) 1. 1. Background h d The proposed Data Act80 is one of the essential elements of the European strategy for data81 together with the Digital Markets Act,82 the Data Governance Act,83 the Digital Services Act84 and the AI Act.85 The regulation aims to unlock huge amounts of digital data which is con- centrated in the hands of relatively few actors, and ensure fairness in access to and use of this data.86 This would be achieved by facilitating access to and use of data by con- sumers and businesses including clarifying the Database Directive,87 enabling public sector bodies to use data held by enterprises in exceptional situations,88 facilitating switching between cloud and edge services,89 establishing safeguards against unlawful data transfer without notifi- cation by cloud service providers and developing interop- erability standards for data.90 The proposed Data Act introduces the right of users to access and use data gener- ated by the use of products or related services,91 the right to share such data with third parties,92 and an obligation 85 Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) and Amending Certain Union Legislative Acts, COM/2021/206 final. 86 Proposed Data Act. See also Estelle Derclaye and Martin Husovec, ‘Why the sui generis database clause in the Data Act is counter-produc- tive and how to improve it?’ 1 <https://papers.ssrn.com/sol3/papers. cfm?abstract_id=4052390> accessed 12 May 2023. 87 Proposed Data Act, cc I-IV and X. 88 Proposed Data Act, c V. 89 Proposed Data Act, c VI. 90 Proposed Data Act, c VIII. 91 Proposed Data Act, c IV. 92 Proposed Data Act, art 5. Overcoming Barriers to Text and Data Mining in the Era of ChatGPT 41 Article 35 of the proposed Data Act is welcomed as it aims to reduce the availability of excessive IP protection over at least some types of databases.105 The legal cer- tainty regarding the use of machine-generated data could stimulate innovation and promote research activities within the DSM.106 Against this background, the follow- ing section of this article will examine the application of Art. 35 of the proposed Data Act to TDM, covering not only the positive aspects but also potential limitations or restrictions related to the use of this provision in the con- text of TDM research. 104 Guido Noto La Diega and Estelle Derclaye, ‘Opening Up Big Data for Sustainability: What Role for Database Rights in the Fourth Industrial Revolution?’ in Ole-Andreas Rognstad, Taina Pihlajarinne 106 European Commission, ‘Impact Assessment Report’ Accompanying the document Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act), SWD (2022) 34 final, 16, 70 and 139. 97 ibid 1 and 2. 98 ibid 2. 94 Proposed Data Act, art 15. 95 European Commission, ‘Study to Support an Impact Assessment for the Review of the Database Directive Final Report’ 1 <https://copen- hageneconomics.com/wp-content/uploads/2022/02/study-to-support- an-impact-assessment-for-the-review-of-the-database-directive.pdf> accessed 12 March 2023. 96 Giuseppe Colangelo, ‘European Proposal for a Data Act-A First Assessment’ (20 July 2022) CERRE Evaluation Paper 2022, 1, 6 <https:// papers.ssrn.com/sol3/papers.cfm?abstract_id=4199565> accessed 10 April 2023; European Commission, ‘Study to Support an Impact Assessment for the Review of the Database Directive Final Report’ (n 95) 1. 97 ibid 1 and 2. 98 ibid 2. 99 Proposed Data Act, art 35. 100 Proposed Data Act, art 35. 101 Proposed Data Act, art 2(2). 102 Proposed Data Act, art 2(2). 103 Proposed Data Act, recital 84. 104 Guido Noto La Diega and Estelle Derclaye, ‘Opening Up Big Data for Sustainability: What Role for Database Rights in the Fourth Industrial Revolution?’ in Ole-Andreas Rognstad, Taina Pihlajarinne 93 Proposed Data Act, art 14. 94 Proposed Data Act, art 15. 99 Proposed Data Act, art 35. 95 European Commission, ‘Study to Support an Impact Assessment for the Review of the Database Directive Final Report’ 1 <https://copen- hageneconomics.com/wp-content/uploads/2022/02/study-to-support- an-impact-assessment-for-the-review-of-the-database-directive.pdf> accessed 12 March 2023. 2. The application of Art. 35 to TDM research 7 of the Database Directive ‘where such data- bases do not qualify for the sui generis right’ as ‘the requirements for protection would not be fulfilled.’114 These two excerpts from Recital 84 have caused some controversy among scholars. Some claim that the for- mer element of this provision indicates that databases of machine-generated data could still fall within the scope of this right under certain circumstances.115 Others see this wording as a ‘confirmatory statement’ claiming that the latter element of Recital 84 suggests that databases of machine-generated data were not eligible for protection under the sui generis regime from the outset.116 However, both perspectives fail to fully elucidate the scope of the sui generis right in relation to databases outlined in Art. 35 of the proposed legislation. data in machine-generated databases.120 For instance, data generated by wearable health monitoring devices should be viewed as ‘collected’ rather than ‘created’ since such data already exists. Researchers who obtain measurements, such as sugar levels, heart rate, blood pressure and body temperature, do not create the data itself but rather create a record of that data through the use of monitoring devices.121 g However, the concept of recorded data122 raises intense debates among scholars in the context of the ‘obtaining-creating dichotomy’. Some academics claim that this data is created because it constitutes ‘represen- tations of natural phenomena, not the phenomena them- selves,’123 but others insist that it is both created and obtained since the two activities cannot be separated when recording data that already exists in nature.124 However, certain scholars, including the author of this article, assume that recorded data refers to the process of obtaining only.125 In this regard, data generated by wearable health monitoring devices should be con- sidered as obtained rather than created because such information exists regardless of its recording.126 Health trackers only process pre-existing data to make it ‘intel- ligible’ and accessible.127 In this sense, the sui generis database protection could certainly extend to some sce- narios involving, for instance, the installation of sensors or expensive software in wearable health monitoring devices as such actions could be considered as a separa- ble relevant investment in obtaining data rather than in its creation.128 Moreover, there could be a lot of intelli- gence in the form of algorithms in the sensors that could already select and arrange the data. 118 Noto La Diega and Derclaye (n 104) 26; Colangelo (n 96) 20; European Commission, ‘Study to Support an Impact Assessment for the Review of the Database Directive Final Report’ (n 95). See also Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29), Fixtures Marketing Ltd v Svenska Spel AB (n 29), Fixtures Marketing Ltd v Oy Veikkaus AB (n 29) and The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31). 125 See for instance Matthias Leistner, ‘Big Data and the EU Database Directive 96/9/EC: Current Law and Potential Reform’ in Sebastian Lohsse, Reiner Schulze and Dirk Staudenmyer (eds), Trading Data in the Digital Economy: Legal Concepts and Tools (Nomos 2017) 27, 28. 113 Proposed Data Act, art 35. 114 Proposed Data Act, recital 84. 115 In this sense see eg, Derclaye and others (n 112) 3; Graef and Husovec (n 104) 4; Derclaye and Husovec (n 86) 2. 116 See for instance Martin Senftleben, ‘Study on EU Copyright and Related Rights and Access to and Reuse of Data’ (European Commission. Research and Innovation. Independent Expert Report 2022) 50 <https:// pure.uva.nl/ws/files/85957820/KI0822205ENN.en.pdf%20p.50> accessed 20 March 2023. 122 Estelle Derclaye defines recorded data as ‘data [that] occur in nature or in time and are generally recorded by instruments of measure in order for them to be intelligible by man. Examples include results of sport com- petitions, meteorological, astronomical data and genomic data’ (Estelle Derclaye, The legal Protection of Databases: a Comparative Analysis (Edward Elgar Publishing 2008) 99). 126 In this sense, see for instance Case C-490/14 Freistaat Bayern v Verlag Esterbauer GmbH ECLI:EU: C:2015:735; Autobahnmaut (n 105). 128 In this sense, see Derclaye and Husovec (n 86) 2; Graef and Husovec (n 104) 4; Robbert Fisher and others, ‘Study in Support of the Evaluation of Directive 96/9/EC on the Legal Protection of Databases’ Brussels: European Commission 2018, 1, 30-31 <http://publications.europa.eu/ resource/cellar/244f227a-597d-11e8-ab41-01aa75ed71a1.0001.01/ DOC_1> accessed 12 March 2023. 127 See for a similar conclusion Freistaat Bayern v Verlag Esterbauer GmbH (n 126). 119 Fixtures Marketing Ltd v Svenska Spel AB (n 29) para 24; Fixtures Marketing Ltd v Organismos prognostikon agonon podosfairou AE (OPAP) (n 29) para 40; The British Horseracing Board Ltd and Others v William Hill Organization Ltd (n 31) para 80. 2. The application of Art. 35 to TDM research 106 European Commission, ‘Impact Assessment Report’ Accompanying the document Proposal for a Regulation of the European Parliament and of the Council on harmonised rules on fair access to and use of data (Data Act), SWD (2022) 34 final, 16, 70 and 139. 107 For instance, an electronic health record is a database containing patients’ health information (eg, diagnoses, allergies, prescribed medica- tion etc,), which is created and maintained by medical personnel. 108 For instance, wearable health devices (eg, fitness trackers, smart- watches, smart inhalers, surgical robots and others). 109 For instance, biobanks are collections of biological samples, which may consist of human-generated data (eg, patient’s medical history) and machine-generated data (eg, genetic sequencing). 110 Proposed Data Act, arts 4-5 and art 35. 111 Roberta De Michele and Marco Furini, ‘IoT Healthcare: Benefits, Issues and Challenges’ (GoodTechs ‘19: EAI International Conference on 111 Roberta De Michele and Marco Furini, ‘IoT Healthcare: Benefits, Issues and Challenges’ (GoodTechs ‘19: EAI International Conference on Smart Objects and Technologies for Social Good, Valencia, September 2019) 111 Roberta De Michele and Marco Furini, ‘IoT Healthcare: Benefits, Issues and Challenges’ (GoodTechs ‘19: EAI International Conference on Smart Objects and Technologies for Social Good, Valencia, September 2019). Smart Objects and Technologies for Social Good, Valencia, September 2019). 101 Proposed Data Act, art 2(2). 112 Estelle Derclaye and others, ‘Opinion of the European Copyright Society on selected aspects of the proposed Data Act’ (European Copyright Society, 12 May 2022) 5 <https://europeancopyrightsociety- dotorg.files.wordpress.com/2022/05/opinion-of-the-ecs-on-selected-as- pects-of-the-data-act-1.pdf> accessed 20 March 2023. 103 Proposed Data Act, recital 84. 104 Guido Noto La Diega and Estelle Derclaye, ‘Opening Up Big Data for Sustainability: What Role for Database Rights in the Fourth Industrial Revolution?’ in Ole-Andreas Rognstad, Taina Pihlajarinne Maryna Manteghi 42 wording of this provision may still raise questions and create uncertainties. The provision excludes databases made of machine-generated data from the sui generis database protection to safeguard the rights of users to access, use and share such data specified in Arts. 4 and 5 of the proposed Data Act.113 The question is whether such exclusion is absolute or only applicable if it inter- feres with outlined users’ rights. A possible clarifica- tion could be found in Recital 84 of the proposed Data Act, which indicates that databases comprised of data obtained by connected products should not be protected under Art. 123 Mark J Davison and P Bernt Hugenholtz, ‘Football Fixtures, Horse Races and Spin-offs: the ECJ Domesticates the Database Right’ (2005) 3 E.I.P.R. 113, 118. 120 Matthias Leistner and Lucie Antoine, ‘IPR and the Use of Open Data and Data Sharing Initiatives by Public and Private Actors’ (Study com- missioned by the European Parliament’s Policy Department for Citizens’ Rights and Constitutional Affairs at the request of the Committee on Legal Affairs 2022) 50 <https://www.europarl.europa.eu/thinktank/ en/document/IPOL_STU(2022)732266> accessed 12 April 2023; Commission Staff Working Document, ʻEvaluation of Directive 96/9/EC on the legal protection of databasesʼ, SWD (2018) 146 final, 36; Noto La Diega and Derclaye (n 104) 15; Innoweb v Wegener (n 47) para 39. 121 I hi I b W ( 47) 39 d 57 124 Derclaye (n 122) 99. 120 Matthias Leistner and Lucie Antoine, ‘IPR and the Use of Open Data and Data Sharing Initiatives by Public and Private Actors’ (Study com- missioned by the European Parliament’s Policy Department for Citizens’ Rights and Constitutional Affairs at the request of the Committee on Legal Affairs 2022) 50 <https://www.europarl.europa.eu/thinktank/ en/document/IPOL_STU(2022)732266> accessed 12 April 2023; Commission Staff Working Document, ʻEvaluation of Directive 96/9/EC on the legal protection of databasesʼ, SWD (2018) 146 final, 36; Noto La Diega and Derclaye (n 104) 15; Innoweb v Wegener (n 47) para 39. 121 I hi I b W ( 47) 39 d 57 117 ibid 49. 2. The application of Art. 35 to TDM research Therefore, under the proposed Data Act, the presumption that researchers may automatically acquire the right to access databases As Senftleben observed, the proposed Data Act sug- gests, inter alia, that the ‘mere putting into operation of an automated process that constantly creates data is not sufficient to acquire sui generis database rights’.117 But what for example if the installation of sensors or soft- ware in wearable health monitoring devices requires a substantial investment? Can it be considered an invest- ment in obtaining data rather than in its creation and thus protected under the sui generis database regime? The answer to these questions revolves around the ‘obtaining-creating dichotomy’, which is an essential element in the CJEU’s legal test on the scope of the sui generis right.118 As mentioned above, the CJEU inter- preted the term ‘obtaining’ as referring to resources used to search existing materials and collect them in the database, excluding resources used for creating data itself.119 However, it could be difficult to draw a clear line between the ‘collection’ and the mere ‘creation’ of 113 Proposed Data Act, art 35. 114 Proposed Data Act, recital 84. 115 In this sense see eg, Derclaye and others (n 112) 3; Graef and Husovec (n 104) 4; Derclaye and Husovec (n 86) 2. 116 See for instance Martin Senftleben, ‘Study on EU Copyright and Related Rights and Access to and Reuse of Data’ (European Commission. Research and Innovation. Independent Expert Report 2022) 50 <https:// pure.uva.nl/ws/files/85957820/KI0822205ENN.en.pdf%20p.50> accessed 20 March 2023. Overcoming Barriers to Text and Data Mining in the Era of ChatGPT 43 technological mechanisms at the expense of users’ rights, Art. 35 should be constructed in a way that would clarify that the provision cannot be overridden by a contract or TPMs. This could ensure better access and utilization of data. made of machine-generated data may not be absolute in all cases. Such databases could still fall within the scope of the sui generis database protection under cer- tain circumstances. Against this background, it would be appropriate to refine the wording of Art. 35 to clarify that databases made of machine-generated data could never enjoy protection under the sui generis regime. This minor legislative revision, if implemented, would result in greater data openness. 2. The application of Art. 35 to TDM research Moreover, the main aim of the proposed Data Act is not focused on facilitating or ensuring access to data for the purpose of scientific research.135 The pro- posed legislation intends to facilitate the accessibility of machine-generated data by users, trade and busi- ness persons and, where there is an exceptional need to access such data, by public sector bodies.136 Therefore, researchers should affiliate themselves with either of these groups to benefit from data access specified under this new legislation. In this regard, researchers may benefit from the provisions allowing users to share machine-generated data with third parties as ‘third party’ also covers research organizations or not-for- profit organizations.137 The data holders should make available such data to these beneficiaries ‘without undue delay, free of charge to the user, of the same quality as is available to the data holder’.138 Moreover, researchers may potentially rely on Art. 14(1) of the proposed Data Act which obliges data holders, in cases of exceptional need, to make machine-generated data available to pub- lic sector bodies.139 In this regard, Recital 56 further clarifies that ‘research-performing organizations and research-funding organizations could also be organized as public sector bodies or bodies governed by public law.’140 Therefore, publicly funded medical research institutions could be able to benefit from this provision when carrying out TDM on databases comprised of data generated by health monitoring devices. Moreover, they are also entitled to share this data with ‘individu- als or organizations in view of carrying out scientific research’141 providing that such actors ‘act either on a not-for-profit basis or in the context of a public-in- terest mission recognized by the State.’142 This implies that independent individual researchers and private research institutions would not be able to acquire even indirect access, based on a collaboration with public actors, to databases made of machine-generated data. Furthermore, the provisions of the proposed Data Act allowing public bodies to access such data and share it with other actors could be only applicable in the case of exceptional needs (e.g. 2. The application of Art. 35 to TDM research public emergency situations).143 Another point of uncertainty may emerge when researchers use mixed databases involving data falling within the scope of the proposed Data Act and so-called derived or inferred data excluded from it.129 Recital 14 clarifies that the latter type of data when lawfully held should not be considered within the scope of the pro- posed legislation.130 This would imply that researchers intending to perform TDM on databases containing information (e.g. statistical data such as average heart rate during different activities) derived from data col- lected by health monitoring devices (e.g. heart rate col- lected by a fitness tracker) would have to acquire lawful access to it through licensing or other lawful means. In other words, such databases may be covered by the sui generis protection. However, it is unclear how research- ers would recognize which data is subject to the proposed Data Act and which is not.131 Moreover, the exclusion of derived or inferred data from the scope of Art. 35 is not well-justified. Recital 14 indicates that to qualify as machine-generated, data should ‘represent the digitali- zation of user actions and events’ and be ‘valuable to the user and support innovation and the development of digital and other services protecting the environment, health and the circular economy’.132 But what if derived or inferred data meets these requirements? What if such data constitutes valuable information and insights that may contribute to research? The inclusion of inferen- tial data in the scope of the proposed Data Act would enhance data availability and its integrity for research purposes. Further, it should be highlighted that the exclusion of databases made of machine-generated data from the sui generis protection in Art. 35 of the proposed Data Act would not automatically guarantee researchers the right to access and use such databases. Database holders may still employ TPMs (e.g. password-protected login to a website with data repositories) or establish contrac- tual agreements to control such activities (e.g. block or restrict access to data collections).133 Although some pro- visions of the proposed regulation clearly indicate that TPMs or contracts should not be applicable to the detri- ment of the user’s rights and obligations, such provisions refer only to third parties’ rights and data sharing agree- ments.134 Moreover, the relevance of these limitations to the sui generis database right remains unclear. 131 Noto La Diega and Derclaye (n 104) 28. 132 Proposed Data Act, recital 14. 133 Derclaye and others (n 112) 5; Noto La Diega and Derclaye (n 104) 28. 134 Proposed Data Act, arts 11(1) and 12(2). 129 Noto La Diega and Derclaye (n 104) 28. 130 Proposed Data Act, recital 14. 135 Proposed Data Act, art 1. 136 Proposed Data Act, art 1. 137 Proposed Data Act, art 51(1) and recital 29. 138 Proposed Data Act, art 51(1). 139 Proposed Data Act, art 14(1). 140 Proposed Data Act, recital 56. 141 Proposed Data Act, art 21(1). 142 Proposed Data Act, recital 68 and art 21(2). 143 Proposed Data Act, arts 14(1), 15 and 21(1) and recitals 56 and 62. 144 Derclaye and others (n 112) 6. 2. The application of Art. 35 to TDM research To prevent a risk of data appropriation by means of contractual or To sum up, the proposed Data Act does not prioritize the needs of researchers, resulting in a lack of ‘robust access and use guarantees for scientific research’ under this legislation.144 The scope of data access outlined in Art. 14(1) and 21(1) of the proposed Data Act is narrow and thus offers limited benefits to researchers conducting 135 Proposed Data Act, art 1. 136 Proposed Data Act, art 1. 137 Proposed Data Act, art 51(1) and recital 29. 138 Proposed Data Act, art 51(1). 139 Proposed Data Act, art 14(1). 140 Proposed Data Act, recital 56. 141 Proposed Data Act, art 21(1). 142 Proposed Data Act, recital 68 and art 21(2). 143 Proposed Data Act, arts 14(1), 15 and 21(1) and recitals 56 and 62. 144 Derclaye and others (n 112) 6. Maryna Manteghi 44 scientific research based on TDM tools. Against this back- ground, Art. 35 should be amended to enable efficient and broad access to and use of machine-generated raw data collections for research purposes. In this regard, it is cru- cial to include researchers among beneficiaries of this pro- vision and address their specific needs in the framework of the proposed regulation. This would facilitate analyt- ical work based on data-driven research activities and ensure a research-friendly regime in a time of the growing relevance of data. constraints of sui generis protection. However, researchers may also encounter legal uncertainty when relying on this provision as many aspects remain unclear. The ‘refined’ wording of Art. 35 and Recital 84 of the proposed Data Act could address many problematic issues discussed in detail in the last chapter of this article. In particular, it needs to be clarified that the databases in question could never fulfil the requirements for sui generis database pro- tection. Moreover, it seems reasonable to include derived or inferred data in the scope of the proposed Data Act, as it may constitute valuable information and insights, and also refine Art. 35 in a manner that would clarify that the provision cannot be overridden by a contract or TPMs at the expense of users’ rights. And finally, there is a need to improve the wording of Art. 35, or even introduce new provisions, to explicitly address researchers’ specific needs in the framework of the proposed regulation. 2. The application of Art. 35 to TDM research This would facilitate analytical work based on data-driven research activities and ensure a research-friendly regime in a time of the growing relevance of data. To sum up, all the proposed solutions, if implemented, would definitely stimulate cut- ting-edge scientific research based on TDM and advanced AI tools, such as ChatGPT, within the DSM. This could strengthen the research and innovative power of the EU at a global level. scientific research based on TDM tools. Against this back- ground, Art. 35 should be amended to enable efficient and broad access to and use of machine-generated raw data collections for research purposes. In this regard, it is cru- cial to include researchers among beneficiaries of this pro- vision and address their specific needs in the framework of the proposed regulation. This would facilitate analyt- ical work based on data-driven research activities and ensure a research-friendly regime in a time of the growing relevance of data. IV. Conclusion Researchers engaging in TDM research on databases, including those using AI language models such as ChatGPT, might not find safe harbors from sui generis database infringement under the CDSM Directive’s TDM exceptions. The construction of these provisions perfectly demonstrates how strong copyright and database protec- tion, on the one hand, and the limited exceptions, on the other hand, could affect research in the EU. Therefore, researchers may need to seek alternative solutions to avoid potential infringement claims. One of these solutions could be Art. 35 of the proposed Data Act, which intends to free databases comprised of machine-generated data from the
https://openalex.org/W3004602545	https://ruc.udc.es/dspace/bitstream/2183/25090/3/SanjurjoSanchez_Jorge_2020_Covering_layers_granite_buildings_northwestern_Iberian_Peninsula.pdf	English	null	Covering Layers on Granite Buildings of Northwestern Iberian Peninsula: When Observable Characteristics and Lab Characterization Do Not Match	Coatings	2,020	cc-by	4,902	Received: 2 January 2020; Accepted: 30 January 2020; Published: 4 February 2020 Abstract: Illustrated glossaries on stone pathologies help to describe deterioration forms in built heritage without resorting to any laboratory analyses. In this way, terms such as crust, deposit, and soiling which according to ICOMOS-ISCS: Illustrated Glossary on Stone Deterioration Patterns may include exogenic material, a patina which results from ageing of the material in an endogenous process, and a ﬁlm included under the broad term of a coating layer in the glossary, can be macroscopically identiﬁed on site. However, a deﬁnition on the basis of characteristics only observable with the naked eye (without further analysis in the laboratory) is certainly complicated, and if in addition, the case studies are on granitic rock (a major building stone used across Europe), the picture becomes even more complicated. The intention of this brief report is to engender an open, constructive debate about the casuistry of the covering layers on granite (a poorly reactive and less porous rock) and the diﬃculty of using the ICOMOS nomenclature on them. Keywords: decay phenomena and processes; granite buildings; urban areas; temperate weather; NW Spain; ICOMOS-ISCS glossary; stone deterioration Brief Report Covering Layers on Granite Buildings of Northwestern Iberian Peninsula: When Observable Characteristics and Lab Characterization Do Not Match Patricia Sanmartín 1,* , Jorge Sanjurjo-Sánchez 2 and Beatriz Prieto 1 Patricia Sanmartín 1,* , Jorge Sanjurjo-Sánchez 2 and Beatriz Prieto 1 1 Departamento de Edafoloxía e Química Agrícola. Facultade de Farmacia, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain; beatriz.prieto@usc.es 2 Instituto Universitario de Xeoloxía, Universidade da Coruña, Campus de Elviña, 15071 A Coruña, Spain; jorge.sanjurjo.sanchez@udc.es 1 Departamento de Edafoloxía e Química Agrícola. Facultade de Farmacia, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain; beatriz.prieto@usc.es 2 Instituto Universitario de Xeoloxía, Universidade da Coruña, Campus de Elviña, 15071 A Coruña, Spain; jorge.sanjurjo.sanchez@udc.es j g j j * Correspondence: patricia.sanmartin@usc.es; Tel.: +34-881-814-984 www.mdpi.com/journal/coatings Coatings 2020, 10, 137; doi:10.3390/coatings10020137 coatings coatings 1. Introduction Some examples of deposits: splashes of paint or mortar, sea salt aerosols, atmospheric particles such as soot or dust, remains of conservation materials such as cellulose poultices, blast materials etc.”; Film is used to describe a “Thin covering or coating layer generally of organic nature, generally homogeneous, follows the stone surface. A ﬁlm may be opaque or translucent”; Film is used to describe a “Thin covering or coating layer generally of organic nature, generally homogeneous, follows the stone surface. A ﬁlm may be opaque or translucent”; Patina is used to describe a “Chromatic modiﬁcation of the material, generally resulting from natural or artiﬁcial ageing and not involving in most cases visible surface deterioration” without perceivable thickness to the naked eye and often having a favorable connotation; Patina is used to describe a “Chromatic modiﬁcation of the material, generally resulting from natural or artiﬁcial ageing and not involving in most cases visible surface deterioration” without perceivable thickness to the naked eye and often having a favorable connotation; Soiling is used to describe a “Deposit of a very thin layer of exogenous particles (e.g., soot) giving a dirty appearance to the stone surface”. Soiling is used to describe a “Deposit of a very thin layer of exogenous particles (e.g., soot) giving a dirty appearance to the stone surface”. Both sources can be useful to describe covering layers on building materials. However, since sedimentary rocks such as limestone and sandstone are by far the most common type of building stone used, and because when dealing with porous rocks such as these, some exogenic materials can penetrate the rock; the result may be better described with the ICOMOS-ISCS glossary than with Dorn’s nomenclature. Facing this situation, it is worth wondering what happens with less porous (and poorly reactive) rocks such as granite, which is also a major building stone used across Europe [3], being very common in Galicia (NW Spain) and some surrounding regions (N Portugal), where mostly Paleozoic granite is the principal building material. Thus, the aim of the present study was to analyze the adequacy of the ICOMOS-ISCS glossary terms relating to covering layers in granitic rocks, based on the properties of this rock type. 1. Introduction We illustrate this with six case studies taken in granite buildings of urban areas of the Galician cities of Santiago de Compostela and A Coruña, which are characterized by a temperate climate and no signiﬁcantly high air pollution. 1. Introduction The study of covering layers on stony materials is not new. As Dorn [1] remarked, there has long been historical interest in them, which deepened during the last century. Indeed, in the last 50 years, many studies of coatings on natural outcrops and historical buildings have been published. Because of this, in some cases, a broad but poorly structured and consensual terminology has been proposed to refer to diﬀerent surface layers. Dorn’s proposal of terms is based on microscopic and geochemical properties: the origin of materials, aspect, thickness, and composition. Furthermore, this author proposed the use of the term coating to group all of the coverings together and deﬁnes the term as a “mixture of materials accreted to the rock by physicochemical processes that are often biologically mediated”, at the same time noting that the term was not standardized. While recognizing that the compilation should not be considered a systematic classiﬁcation reference, Dorn suggested that it could be a ﬁrst step towards unifying the proposed deﬁnitions [1]. On the other hand, and closer in context to the cultural heritage conservation researcher than to the geologist, there is the ICOMOS-ISCS: Illustrated Glossary on Stone Deterioration Patterns [2], a very useful tool for identifying and comparing decay phenomena on a visual basis. It appeared as a result of decades of work, after extensive analyses and discussions of terminologies and glossaries Coatings 2020, 10, 137; doi:10.3390/coatings10020137 www.mdpi.com/journal/coatings www.mdpi.com/journal/coatings 2 of 9 Coatings 2020, 10, 137 from multiple origins, such as in the Italian, French, German, and English language. One of the six categories included in the glossary (“Discoloration and Deposit”) encompasses a series of terms relating to covering layers: Crust is used to describe a “Generally coherent accumulation of materials on the surface. A crust may include exogenic deposits in combination with materials derived from the stone. A crust is frequently dark coloured (black crust) but light colours can also be found. Crusts may have an homogeneous thickness, and thus replicate the stone surface, or have irregular thickness and disturb the reading of the stone surface details”; Deposit is used to describe an “Accumulation of exogenic material of variable thickness. Some examples of deposits: splashes of paint or mortar, sea salt aerosols, atmospheric particles such as soot or dust, remains of conservation materials such as cellulose poultices, blast materials etc.”; Deposit is used to describe an “Accumulation of exogenic material of variable thickness. 2. Case Studies Six coatings formed on the surfaces of historical granite buildings of diﬀerent ages, placed in diﬀerent sites and exposed to diﬀerent environmental conditions, were considered. The buildings are located in the cities of Santiago de Compostela and A Coruña (Galicia, NW Spain), both with an oceanic climate (a humid subtropical climate according to the Food and Agriculture Organization of the United Nations’ agro-ecological zoning) with high atmospheric humidity caused by high rainfall and mild temperatures throughout the year [4], as well as no signiﬁcantly high air pollution. Sampling of coatings on buildings was performed in a similar manner to that described in Prieto et al. [5] and Sanjurjo-Sánchez et al. [6]. Because covering layers on building surfaces and façades are highly variable, analytical studies and techniques cannot be standardized. Consequently, diﬀerent methods are used for the characterization of the covering layer in each case. For the present study, both surface and polished cross-sections of coating samples, which included the granite underneath the coating layer, were prepared and observed using scanning electron microscopy (SEM) by secondary and backscattered electrons, and X-ray maps in a JEOL JSM 6400 scanning electron microscope at 16 keV coupled to a electron dispersive spectroscopic analyzer (EDS) Oxford 200 Inca Energy EDS at the University of A Coruña and a LEO-435VP at University of Santiago de Compostela. Diﬀerences in the mineralogical composition of the coating layer and the granite underlying the coating were analyzed by X-ray diﬀraction (XRD) of the powdered samples, using a D5000 SIEMENS Coatings 2020, 10, 137 3 of 9 X-Ray Diﬀractometer at University of A Coruña and a PW1710 Philips diﬀractometer at University of Santiago de Compostela. A Bruker IFS66 infrared spectrometer with an FRA 106 Raman module and Raman microscope attachment were also used. IR/Raman Spectra were obtained using an Nd:Yag laser of 350mW and operating at 1064 nm. In all cases, a term relating to the covering layers was chosen from the ICOMOS-ISCS glossary based on an on-site visual assessment using the naked eye. Subsequently, that term was compared with the term that appeared according to the characterization made in the laboratory, using the aforementioned techniques. q Three case studies in the old town of Santiago de Compostela [7] were considered. 2. Case Studies The ﬁrst concerns the exterior wall of a house (Number 2) in Toural Square (8◦32′42.0′′ W, 42◦52′39.6′′ N, WGS84) on which a dark layer that is diﬃcult to separate from the granite substrate had formed (Figure 1). It was identiﬁed on site as a patina according the ICOMOS-ISCS glossary. Further examination of samples of the material by electron microscopy revealed two layers of variable thicknesses: a superﬁcial layer of thickness between 82 and 260 µm and an inner layer of thickness between 50 and 470 µm (Figure 1C). The inner layer was interspersed between the mineral grains and the diﬀuse boundary, making it very diﬃcult to distinguish the material from the underlying rock. Furthermore, the covering layer was almost exclusively formed by inorganic particles (including gypsum crystals in their characteristic habit. Given that gypsum was present in the joint mortar, the most likely hypothesis is that it moved to the ashlar forming part of the layer). Element distribution mapping (Figure 1D) identiﬁed S and Ca as the main components (components of calcium sulphate, i.e., gypsum), with Al and Si appearing in the underlying granitic material [5]. Biological characterization did not show any evidence of the presence of organisms [8]. Thus, because this layer actually consisted of two diﬀerent layers (a case not included in the ICOMOS-ISCS glossary), it could be described as a ﬁlm, a deposit (the most superﬁcial layer), or a crust (the inner layer), but in no case a patina since exogenic materials are involved. Whatever the term used, none of the present ICOMOS glossary allows for a suitable description of this covering layer. Figure 1. Covering layer on the exterior wall of Number 2 Toural Square (Santiago de Compostela). (A) Entrance to the Toural Square, (B) detail of the sampled area, (C) SEM picture of a cross-section of the covering layer, (D) EDS of the cross-section, (E) XRD spectra of the covering layer (green line) and of the granite underneath the covering layer (brown line), and (F) IR/Raman spectra of the covering layer (green line) and of the granite underneath the covering layer (brown line). Figure 1. Covering layer on the exterior wall of Number 2 Toural Square (Santiago de Compostela). 2. Case Studies (A) Entrance to the Toural Square, (B) detail of the sampled area, (C) SEM picture of a cross-section of the covering layer, (D) EDS of the cross-section, (E) XRD spectra of the covering layer (green line) and of the granite underneath the covering layer (brown line), and (F) IR/Raman spectra of the covering layer (green line) and of the granite underneath the covering layer (brown line). Coatings 2020, 10, 137 4 of 9 In the second case (Figure 2), the covering layer from the Monastery of San Martiño Pinario (8◦32′40.4′′ W, 42◦52′56.1′′ N, WGS84) also seemed to be a patina after on-site identiﬁcation (Figure 2B). In the second case (Figure 2), the covering layer from the Monastery of San Martiño Pinario (8◦32′40.4′′ W, 42◦52′56.1′′ N, WGS84) also seemed to be a patina after on-site identiﬁcation (Figure 2B). In the lab characterization (Figure 2C,D), it presented S (forming gypsum crystals in their characteristic habit) and P, although it was not possible to determine in what form. In other related studies, the presence of phosphorus was associated with accumulations of apatite, carbonate apatite, and hydroxyapatite [7], and its accumulation was referred to as the deposition of atmospheric particles, the residues left by soaps used in cleaning treatments, the residues of stone conservation treatments based on mixtures of soluble phosphates and silicates, or the mineralization of organic preservatives used in preservation containing phosphorus compounds, especially milk derivatives and casein salts [7]. This may also have a biological origin, as some bacterial colonies increase the phosphate in the solution present in the pores of the rock, causing the local precipitation of apatite, and also, phosphates can come from animal excrements, which is very rich in these compounds. However, apatite can also be present in granite rocks, and P can also be found in low concentrations in K- and alkali feldspars, and some accessory minerals of granite rocks such as monacite and xenotime, making it diﬃcult to assess if its origin is exogenic or endogenic. In any case, while the observed gypsum and calcite could be exogenous, there are signiﬁcant doubts about the other components such as P, making it diﬃcult to use the glossary again for this case. Figure 2. Covering layer on Monastery of San Martiño Pinario (Santiago de Compostela). 2. Case Studies (A) Area of the wall on the main entrance where the covering layer was taken, (B) detail of the sampled area, (C) SEM pictures of the covering layer, and (D) SEM picture (left) and EDS pictures (middle and right) of the cross-section of the coating. Figure 2. Covering layer on Monastery of San Martiño Pinario (Santiago de Compostela). (A) Area of the wall on the main entrance where the covering layer was taken, (B) detail of the sampled area, (C) SEM pictures of the covering layer, and (D) SEM picture (left) and EDS pictures (middle and right) of the cross-section of the coating. The third case (Figure 3) concerns a covering layer on the façade of Fonseca Palace (8◦32′43.3′′ W, 42◦52′47.3′′ N, WGS84) taken from a sculptural element (Figure 3A). The layer was dark-grey and of a compact, waxen aspect, identiﬁed as a patina on site. Examination of samples of the material by electron microscopy revealed two layers of a very diﬀerent composition to the underlying rock, which were clearly separate from the rock (marked with yellow and orange line segments, Figure 3B). The most superﬁcial layer was of a thickness of about 100 µm and the main element was carbon, although some highly electron-reﬂective particles, mainly composed of Pb, were also detected [5]. Analysis of the material by IR-Raman spectroscopy [7] allowed us to assign the origin to beeswax, which was commonly used in the early 1960s in Galicia to treat the surface of valuable monuments with the aim of preventing the erosive action of atmospheric agents [9]. The inner layer also contained C, along 5 of 9 Coatings 2020, 10, 137 with S, Ca, Pb, and Cl (Figure 3C). Examination of samples of the layer by XRD analysis revealed the presence of gypsum, calcite, portlandite, goethite, and pseudocotumnite (a compound consisting of potassium, lead, and chlorine), as is shown in Figure 3D. Figure 3. Covering layer on the façade of Fonseca Palace (Santiago de Compostela). (A) Full view of the sculpture where the covering layer was taken, (B) SEM picture of a cross-section of the covering layer, (C) EDS from diﬀerent zones of the inner layer surface, and (D) XRD spectra of the covering layer (green line) and of the granite underneath the covering layer (brown line). Figure 3. Covering layer on the façade of Fonseca Palace (Santiago de Compostela). 2. Case Studies (A) Full view of the sculpture where the covering layer was taken, (B) SEM picture of a cross-section of the covering layer, (C) EDS from diﬀerent zones of the inner layer surface, and (D) XRD spectra of the covering layer (green line) and of the granite underneath the covering layer (brown line). Taking all of these ﬁndings into account, the outer part of this covering layer (but not the inner part) could be described as a ﬁlm. The inner part could be described as a deposit, crust, or even as soiling, or as none of these. Taking all of these ﬁndings into account, the outer part of this covering layer (but not the inner part) could be described as a ﬁlm. The inner part could be described as a deposit, crust, or even as soiling, or as none of these. Case studies of two buildings in the center of the city of A Coruña are also considered: the Church of the Capuchinas (8◦23′57.22′′ W, 43◦22′21.45′′ N, WGS84) and the Colegiata de Santa Maria del Campo. (8◦23′33.26′′ W, 43◦22′14.8′′ N, WGS84). Both buildings are made from granite. The layer in question covered most of the main façade of the Church of the Capuchinas (Figure 4A), which overlooks a road with heavy traﬃc. On site this could be classiﬁed as a patina or a ﬁlm. The layer was dark-grey, porous, and 100-200 µm thick (Figure 4B–D). It was composed of gypsum crystals and airborne soil particles (organic particles and silicate minerals), including particulate air pollution (C-rich spheres, Fe- and –Ba rich aggregates) (Figure 4E,F). It was very diﬃcult to determine whether or not the materials are of completely exogenic origin, and therefore the terms crust, ﬁlm (as it contains organic carbon), and soiling (the layer has a dirty appearance and contains soot) could all be applied to this covering layer, but not the term patina. 6 of 9 Coatings 2020, 10, 137 Figure 4. Covering layer on a wall of the Church of the Capuchinas (A Coruña). (A) Area of the wall where the covering layer was taken, (B) surface SEM picture, (C) EDS of the surface, (D) SEM picture of a cross-section of the covering layer, (E) Fe-rich particle, and (F) EDS of the particle. Figure 4. Covering layer on a wall of the Church of the Capuchinas (A Coruña). 2. Case Studies (A) Area of the wall where the covering layer was taken, (B) surface SEM picture, (C) EDS of the surface, (D) SEM picture of a cross-section of the covering layer, (E) Fe-rich particle, and (F) EDS of the particle. The covering layer on the south façade of Colegiata de Santa Maria del Campo (Figure 5A) seemed to be a ﬁlm. It was a dark-grey layer of thickness 100–150 µm (Figure 5B–D). It was composed of airborne soil particles (organic particles and silicate minerals) and gypsum crystals and also contained halite crystals and a Ca-S-C-rich amorphous material. The building is located 200 m from the coastline and the façade overlooks a road with almost no traﬃc. The location explains the presence of exogenic halite, but the silicate minerals identiﬁed are the same as those found in the underlying granite rock, as observed in other patinas on the same building [10]. Therefore, this cannot be considered to be a ﬁlm. Figure 5. Covering layer on the façade of the Colegiata de Santa Maria del Campo (A Coruña). (A) Detail of the area on the wall where the covering layer was taken, (B,C) SEM pictures of the covering layer surface, and (D) cross-section and X-ray map of the covering layer. Figure 5. Covering layer on the façade of the Colegiata de Santa Maria del Campo (A Coruña). (A) Detail of the area on the wall where the covering layer was taken, (B,C) SEM pictures of the covering layer surface, and (D) cross-section and X-ray map of the covering layer. 7 of 9 Coatings 2020, 10, 137 The ﬁnal case considered was a black layer, 100–150 µm thick, on the east façade of the same building (Figure 6A,B). This seemed to be a patina to the naked eye. However, the layer was composed of a porous matrix (Figure 6C) of airborne soil particles (organic particles and silicate minerals), gypsum crystals, halite, and Ba- and Fe-rich aggregates, with traces of P (Figure 6D,E). The building is located 1200 m from the coastline and the façade overlooks a road with little traﬃc. Again, it is not easy to assign any of the terms in the ICOMOS-ISCS glossary to this covering layer. This could be a crust, patina, ﬁlm, or soiling, and, according to a single ﬁeld observation, it could be classiﬁed as a patina. Figure 6. 2. Case Studies Covering layer on the façade of the Colegiata de Santa Maria del Campo (A Coruña). (A) Full view of the façade where the covering layer was taken, (B) SEM picture of the cross-section of the layer, (C) SEM picture of the surface of the layer, and (D,E) EDS of the layer surface. Figure 6. Covering layer on the façade of the Colegiata de Santa Maria del Campo (A Coruña). (A) Full view of the façade where the covering layer was taken, (B) SEM picture of the cross-section of the layer, (C) SEM picture of the surface of the layer, and (D,E) EDS of the layer surface. 3. Discussion and Final Remarks ICOMOS-ISCS illustrated glossary terms are descriptive, therefore notions such as the origin and composition should be avoided as far as possible, since they are not objectively observable characteristics of the deterioration form to the naked eye. ICOMOS-ISCS illustrated glossary terms are descriptive, therefore notions such as the origin and composition should be avoided as far as possible, since they are not objectively observable characteristics of the deterioration form to the naked eye. From the characterization of the above covering layers and our observations [5–8,10–12], we can state that the components of most of the covering layers on the granite surfaces are partly related to the surrounding environment and clearly formed from exogenic materials, but also include some of the underlying granite minerals [6,11,12]. Thus, they are not necessarily related to the natural ageing of the rock surfaces and the present term patina is not applicable. Although the term crust could be used, its deﬁnition can be used for a variety of light/dark colors and thicknesses of layers, which only sometimes follow the stone surface and that do not contain organic carbon. In the ICOMOS-ISCS glossary, the term ﬁlm refers to organic carbon, but our observations for granite are not consistent with the other properties cited in the glossary for ﬁlms, i.e., “homogeneous” and “follows the stone surface”. Soiling and deposit could also be applied, as the samples contain exogenous materials. The very purpose of an international glossary is to have access to a dedicated tool-kit of precise terms that are accepted, understood, and used by professionals in diﬀerent contexts. This includes researchers and practitioners of conservation. In that sense, it is important that such a glossary is regularly discussed and updated in order to be adapted to the widest possible range of situations. It is important that the terms can be used on several types of stone, preventing deﬁnition overlap. This is particularly problematic in the case of granite rocks, a major construction material in European 8 of 9 Coatings 2020, 10, 137 buildings historically [3]. Unlike other porous and reactive rocks (e.g., limestone, marble), granite rocks are not very porous, highly impermeable, and mostly composed of silicate minerals, which react poorly with certain molecules that reach the stone surface (through air or water transport) in urban environments in temperate climates. 3. Discussion and Final Remarks Many studies on the coatings found on granite surfaces of historical buildings show a mixture of exogenous and endogenous particles, such as carbonaceous or S- and P- rich particles. In such cases, the term patina from the glossary cannot be applied. The use of the other terms is also problematic: carbonaceous particles are not necessary organic, as the deﬁnition of ﬁlm requires, or not necessarily all components of the coating are exogenous, as the deﬁnition of deposit shows. Sometimes other terms, such as soiling, cannot be applied because dust coatings combine endogenous and exogenous particles. Thus, in granite rocks only the term crust is applicable for the characterization of most coatings In conclusion, we suggest a review of the ﬁve terms in the ICOMOS-ISCS glossary, thus opening a debate with the the possibility of redeﬁning some of the terms, or at least some aspects of these terms. The addition of new terms would help make the glossary more clear and useful in the future for all stone types, including granite rocks. The purpose of this revision would be to compile a glossary applicable to all study cases. If not, there is the risk that the glossary will fall out of use, being useless for both researchers and practitioners when applied to certain rock types, such as granite. A new glossary should consider that some materials and surfaces are more prone to react with exogenous particles, and so all the terms are not applicable in the same manner to all materials and study cases. Of course, any future review of the terms should be based on a complete overview of the covering layers described on granite rocks, which is beyond the scope of this paper. Author Contributions: Conceptualization, P.S., J.S.-S., B.P.; writing—original draft preparation, P.S, J.S.-S.; writing—review and editing, P.S., J.S.-S., B.P. All authors have read and agreed to the published version of the manuscript. Funding: The authors are grateful for funding from the Xunta de Galicia within the projects “Consolidación y estructuración de unidades de investigación competitivas—Grupos de referencia competitiva (GRC)” (Ref. ED431C 2018/32) and “Consolidación y estructuración de unidades de investigación competitivas—Grupos de potencial crecimiento (GPC)” (Ref. GPC2015/024). Conﬂicts of Interest: The authors declare no conﬂict of interest. References 1. Dorn, R.I. Rock Coatings; Elsevier: Amsterdam, The Netherlands, 1998. 1. Dorn, R.I. Rock Coatings; Elsevier: Amsterdam, The Netherlands, 1998. 1. Dorn, R.I. Rock Coatings; Elsevier: Amsterdam, The Netherlands, 1998. 2. Vergès-Belmin, V. (Ed.) Illustrated Glossary on Stone Deterioration Patterns; Monuments and Sites XV; IC Paris, France, 2008. 3. Vicente, M.A.; Delgado Rodrigues, J.; Acevedo, J. Degradation and Conservation of Granitic Rocks in Monument; Protection and Conservation of the European Cultural Heritage Research Report No. 5; European Commission: Brussels, Belgium, 1996. 3. Vicente, M.A.; Delgado Rodrigues, J.; Acevedo, J. Degradation and Conservation of Granitic Rocks in Monument; Protection and Conservation of the European Cultural Heritage Research Report No. 5; European Commission: Brussels, Belgium, 1996. 4. Martínez-Cortizas, A.; Pérez-Alberti, A. (Eds.) Atlas Climático de Galicia; Consellería de Medioambiente: Xunta de Galicia, Spain, 1999. 4. Martínez-Cortizas, A.; Pérez-Alberti, A. (Eds.) Atlas Climático de Galicia; Consellería de Medioambiente: Xunta de Galicia, Spain, 1999. 5. Prieto, B.; Aira, N.; Silva, B. Comparative study of dark patinas on granitic outcrops and buildings. Sci. Total Environ. 2007, 381, 280–289. [CrossRef] 5. Prieto, B.; Aira, N.; Silva, B. Comparative study of dark patinas on granitic outcrops and buildings. Sci. Total Environ. 2007, 381, 280–289. [CrossRef] 6. Sanjurjo-Sánchez, J.; Romaní, J.R.V.; Alves, C. Comparative analysis of coatings on granitic substrates from urban and natural setings (NW Spain). Geomorphology 2012, 138, 231–242. [CrossRef] 7. Aira, N. Pátinas Oscuras Sobre Rocas Graníticas: Génesis y Composicioén. Ph.D. Thesis, University of Santiago de Compostela, Santiago de Compostela, Spain, 2007. . Aira, N.; Jurado, V.; Prieto, B.; Silva, B. Gas chromatography applied to cultural heritage. Analysis of d patinas on granite surfaces. J. Chromatogr. A 2007, 1147, 79–84. [CrossRef] [PubMed] 9. Pan, A.; Chiussi, S.; Serra, J.; González, P.; León, B. Excimer laser removal of beeswax from galician granite monuments. J. Cult. Herit. 2009, 10, 48–52. [CrossRef] 9 of 9 Coatings 2020, 10, 137 10. Sanjurjo Sánchez, J.; Vidal Romani, J.R.; Fernández Mosquera, D.; Alves, C.A. Study of origin and composition of coatings in a monument built with granitic rocks, by SEM, XRD, XRF and DTA-TGA. X-ray Spectrom. 2008, 37, 346–354. [CrossRef] 11. Sanjurjo-Sánchez, J.; Romaní, J.R.V.; Alves, C. Deposition of particles on gypsum-rich coatings of historic buildings in urban and rural environments. Constr. Build. Mater. 2011, 25, 813–822. [CrossRef] 12. Sanjurjo-Sánchez, J. La Secuencia de Alteración de Superﬁcies Graníticas: Inﬂuencia de Factores Litológicos, Geodinámicos, Climáticos y Biológicos. Ph.D. References Thesis, University of A Coruña, A Coruña, Spain, 2005. © 2020 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/). 10. Sanjurjo Sánchez, J.; Vidal Romani, J.R.; Fernández Mosquera, D.; Alves, C.A. Study of origin and composition of coatings in a monument built with granitic rocks, by SEM, XRD, XRF and DTA-TGA. X-ray Spectrom. 2008, 37, 346–354. [CrossRef] 10. Sanjurjo Sánchez, J.; Vidal Romani, J.R.; Fernández Mosquera, D.; Alves, C.A. Study of origin and composition of coatings in a monument built with granitic rocks, by SEM, XRD, XRF and DTA-TGA. X-ray Spectrom. 2008, 37, 346–354. [CrossRef] Sanjurjo-Sánchez, J.; Romaní, J.R.V.; Alves, C. Deposition of particles on gypsum-rich coatings of historic buildings in urban and rural environments. Constr. Build. Mater. 2011, 25, 813–822. [CrossRef] 12. Sanjurjo-Sánchez, J. La Secuencia de Alteración de Superﬁcies Graníticas: Inﬂuencia de Factores Litológicos, Geodinámicos, Climáticos y Biológicos. Ph.D. Thesis, University of A Coruña, A Coruña, Spain, 2005. © 2020 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/W2594648906	https://www.life-science-alliance.org/content/lsa/1/3/e201800092.full.pdf	English	null	DHX15 regulates CMTR1-dependent gene expression and cell proliferation	bioRxiv (Cold Spring Harbor Laboratory)	2,017	cc-by	14,214	DHX15 regulates CMTR1-dependent gene expression and cell proliferation CMTR1 contributes to mRNA cap formation by methylating the ﬁrst transcribed nucleotide ribose at the O-2 position. mRNA cap O-2 methylation has roles in mRNA stabilisation and translation, and self-RNA tolerance in innate immunity. We report that CMTR1 is recruited to serine-5–phosphorylated RNA Pol II C-terminal domain, early in transcription. We isolated CMTR1 in a complex with DHX15, an RNA helicase functioning in splicing and ribosome biogenesis, and characterised it as a regulator of CMTR1. When DHX15 is bound, CMTR1 activity is repressed and the methyl- transferase does not bind to RNA pol II. Conversely, CMTR1 ac- tivates DHX15 helicase activity, which is likely to impact several nuclear functions. In HCC1806 breast carcinoma cell line, the DHX15–CMTR1 interaction controls ribosome loading of a subset of mRNAs and regulates cell proliferation. The impact of the CMTR1–DHX15 interaction is complex and will depend on the relative expression of these enzymes and their interactors, and the cellular dependency on different RNA processing pathways. A series of enzymes catalyse mRNA cap formation, which have different conﬁgurations in different species (Shuman, 2002). In mammals, RNGTT/capping enzyme catalyses guanosine cap ad- dition and RNA guanine-7 methyltransferase (RNMT)-RNMT-acti- vating miniprotein (RAM) catalyses guanosine cap N-7 methylation. RNGTT/capping enzyme and RNMT-RAM are recruited to RNA pol II at the initiation of transcription (Buratowski, 2009). CMTR1 and CMTR2 methylate the O-2 position of ﬁrst and second transcribed nucleotide riboses, respectively (Belanger et al, 2010; Werner et al, 2011; Inesta-Vaquera & Cowling, 2017). CMTR1 (ISG95, FTSJD2, KIAA0082) was ﬁrst identiﬁed as a human-interferon–regulated gene (Su et al, 2002; Geiss et al, 2003; Guerra et al, 2003; Kato et al, 2003). It was rec- ognised to have several functional domains including a methyl- transferase domain (Haline-Vaz et al, 2008). Subsequently, CMTR1 was biochemically characterised as the O-2 ribose methyltransferase of the ﬁrst transcribed nucleotide and the catalytic domain was crystalized with S-adenosyl methionine (SAM) and a capped oli- gonucleotide (Belanger et al, 2010; Smietanski et al, 2014). CMTR1 and ﬁrst nucleotide O-2 methylation have functions in gene expression and innate immunity. O-2 methylation is likely to inﬂuence recruitment of cap-binding protein complexes and promote ribosomal subunit binding (Muthukrishnan et al, 1976b). Following fertilization of sea urchin eggs, N-7 and O-2 cap meth- ylation is associated with translational up-regulation of a subset of maternal transcripts (Caldwell & Emerson, 1985). 1Centre for Gene Regulation and Expression, School of Life Sciences, University of Dundee, Dundee, UK 2Institute of Molecular, Cell and Systems Biology, School of Life Sciences, University of Glasgow, Glasgow, UK 3Nufﬁeld Department of Clinical Neurosciences, Medical Sciences Division, University of Oxford, Oxford, UK 4Center for Interdisciplinary Studies on the Nervous System and Institute of Medicine, Universidad Austral de Chile, Valdivia, Chile 5Division of Signal Transduction Therapies, School of Life Sciences, University of Dundee, Dundee, UK DHX15 regulates CMTR1-dependent gene expression and cell proliferation During Xenopus laevis oocyte maturation, ﬁrst nucleotide O-2 methylation sig- niﬁcantly increases translation efﬁciency and is required for the translation of maternal mRNA (Kuge & Richter, 1995; Kuge et al, 1998). Recently, cap O-2 methylation was demonstrated to be critical for preventing decapping exoribonuclease-mediated decapping, which leads to RNA degradation (Picard-Jean et al, 2018). In mice, a sig- niﬁcant proportion of the ﬁrst nucleotides were found to be O-2 methylated on the ribose, although the relative proportion of this methylation varied between organs, indicating a regulated event (Kruse et al, 2011). on 23 October, 2024 life-science-alliance.org Downloaded from http://doi.org/10.26508/lsa.201800092 Published Online: 18 June, 2018 \| Supp Info: on 23 October, 2024 life-science-alliance.org Downloaded from http://doi.org/10.26508/lsa.201800092 Published Online: 18 June, 2018 \| Supp Info: on 23 October, 2024 life-science-alliance.org Downloaded from http://doi.org/10.26508/lsa.201800092 Published Online: 18 June, 2018 \| Supp Info: DOI 10.26508/lsa.201800092 \| Received 16 May 2018 \| Revised 4 June 2018 \| Accepted 5 June 2018 \| Published online 18 June 2018 Correspondence: v.h.cowling@dundee.ac.uk DHX15 regulates CMTR1-dependent gene expression and cell proliferation Francisco Inesta-Vaquera1, Viduth K Chaugule2, Alison Galloway1, Laurel Chandler3, Alejandro Rojas-Fernandez4, Simone Weidlich5, Mark Peggie5, Victoria H Cowling1 position of the riboses of the ﬁrst and second transcribed nucleo- tides are sites of abundant methylation (Langberg & Moss, 1981). position of the riboses of the ﬁrst and second transcribed nucleo- tides are sites of abundant methylation (Langberg & Moss, 1981). A series of enzymes catalyse mRNA cap formation, which have different conﬁgurations in different species (Shuman, 2002). In mammals, RNGTT/capping enzyme catalyses guanosine cap ad- dition and RNA guanine-7 methyltransferase (RNMT)-RNMT-acti- vating miniprotein (RAM) catalyses guanosine cap N-7 methylation. RNGTT/capping enzyme and RNMT-RAM are recruited to RNA pol II at the initiation of transcription (Buratowski, 2009). CMTR1 and CMTR2 methylate the O-2 position of ﬁrst and second transcribed nucleotide riboses, respectively (Belanger et al, 2010; Werner et al, 2011; Inesta-Vaquera & Cowling, 2017). CMTR1 (ISG95, FTSJD2, KIAA0082) was ﬁrst identiﬁed as a human-interferon–regulated gene (Su et al, 2002; Geiss et al, 2003; Guerra et al, 2003; Kato et al, 2003). It was rec- ognised to have several functional domains including a methyl- transferase domain (Haline-Vaz et al, 2008). Subsequently, CMTR1 was biochemically characterised as the O-2 ribose methyltransferase of the ﬁrst transcribed nucleotide and the catalytic domain was crystalized with S-adenosyl methionine (SAM) and a capped oli- gonucleotide (Belanger et al, 2010; Smietanski et al, 2014). CMTR1 contributes to mRNA cap formation by methylating the ﬁrst transcribed nucleotide ribose at the O-2 position. mRNA cap O-2 methylation has roles in mRNA stabilisation and translation, and self-RNA tolerance in innate immunity. We report that CMTR1 is recruited to serine-5–phosphorylated RNA Pol II C-terminal domain, early in transcription. We isolated CMTR1 in a complex with DHX15, an RNA helicase functioning in splicing and ribosome biogenesis, and characterised it as a regulator of CMTR1. When DHX15 is bound, CMTR1 activity is repressed and the methyl- transferase does not bind to RNA pol II. Conversely, CMTR1 ac- tivates DHX15 helicase activity, which is likely to impact several nuclear functions. In HCC1806 breast carcinoma cell line, the DHX15–CMTR1 interaction controls ribosome loading of a subset of mRNAs and regulates cell proliferation. The impact of the CMTR1–DHX15 interaction is complex and will depend on the relative expression of these enzymes and their interactors, and the cellular dependency on different RNA processing pathways. © 2018 Inesta-Vaquera et al. CMTR1 interacts directly with DHX15 To investigate the regulation and function of CMTR1, we identiﬁed CMTR1-interacting proteins. HA-CMTR1 was immunoprecipitated from HeLa cell extracts and resolved by SDS–PAGE, and co-puriﬁed proteins were identiﬁed by mass spectrometry (Fig 1A). DHX15 (O43143), a 95-kD DEAH-box RNA helicase, was the only protein identiﬁed with signiﬁcant mascot scores and coverage in HA-CMTR1 immunoprecipitates (IP) (Fig S1) (Imamura et al, 1997). Conversely, CMTR1 was identiﬁed in HA-DHX15 IPs using mass spectrometry (Figs 1B and S1). To verify their interaction, GFP-CMTR1 and FLAG- DHX15 were expressed in HeLa cells and co-immunoprecipitated from cell extracts (Fig 1C). To investigate endogenous CMTR1, an antibody was raised against recombinant human CMTR1, which detected CMTR1 in HeLa cell extracts using Western blot and im- munoprecipitation (IP) (Fig S2A and B). The interaction of endog- enous CMTR1 and DHX15 was conﬁrmed using IP in HeLa, HCC1809, U2OS, and HEK293 cell extracts (Fig 1D). In CMTR1 null HeLa cells, the anti-CMTR1 antibody was unable to purify DHX15, discounting non- speciﬁc interactions of DHX15 with resin or antibody (Fig 1E). Demonstrating the speciﬁcity of the CMTR1–DHX15 interaction, DHX16, another DEAH RNA helicase, was not detected in CMTR1 IPs, and RNMT, another cap methyltransferase, was not detected in DHX15 IPs (Fig S2C and D). Furthermore, the CMTR1–DHX15 in- teraction was sustained on RNaseA treatment and therefore is not maintained by an RNA connector (Fig 1F). In HeLa cells, DHX15 and CMTR1 are expressed at equivalent molar ratios (Nagaraj et al, 2011). Equimolar recombinant human CMTR1 and His6-DHX15 co- immunoprecipitated, conﬁrming their direct interaction (Figs 1G and S2E). Neither CMTR1 or DHX15 homodimerise (Fig S2F and G). DHX15 is a prototypic member of the DEAH family of RNA heli- cases (Jankowsky, 2011). It contains an N-terminal domain of un- known function (N-term, residues 1–146), two Rec-A tandem repeats (Rec A1–A2, residues 147–518), a WH domain (residues 519–572), a ratchet domain (residues 572–671), and an OB-fold (residues 671–795) (Fig 2C). The OB-fold is the predominant site of interaction with RNA and proteins, although the Rec-A domains can also es- tablish functional interactions with RNA and protein (Lebaron et al, 2009). Other G-patch–containing proteins have been demonstrated to interact with DHX15 via the OB-fold (Fig S3) (Lin et al, 2009; Niu et al, 2012; Chen et al, 2014; Memet et al, 2017). CMTR1 and DHX15 co-eluted after the 158 kD marker consistent with DHX15–CMTR1 heterodimers. As described above, no other inter- acting proteins were identiﬁed in mass spectrometry analysis of CMTR1 IPs (Figs 1A and S1). Conversely, DHX15, was also observed in larger, CMTR1-independent complexes by gel ﬁltration (Fig 1H, 11.5 ml). Mass spectrometry analysis of DHX15 IPs identiﬁed a series of other proteins, including some with a G-patch domain. Novel and all previously reported DHX15 interactors were identiﬁed, including those involved in ribosomal biogenesis and splicing (Figs 1B and S3). This result reﬂects the well-documented function of DHX15 and its homologues in these processes (Robert-Paganin et al, 2015; Memet et al, 2017). Although these other DHX15 complexes are biologically interesting, for the remainder of this study we focus on the relationship between DHX15 and CMTR1. The composition of the 59 cap is also an important determinant of self- (host) versus non–self-RNA during viral infection (Leung & Amarasinghe, 2016). The absence of O-2 methylation in viral tran- scripts results in enhanced sensitivity to the interferon-induced IFIT proteins; ﬁrst nucleotide O-2 methylation distinguishes self from non–self-RNA (Dafﬁs et al, 2010). CMTR1-dependent O-2 methylation abrogates the activation of retinoic acid inducible gene I, a helicase that initiates immune responses on interaction with uncapped or aberrantly capped transcripts (Schuberth-Wagner et al, 2015). Here, we report the ﬁrst regulator of CMTR1 function. We dem- onstrate that CMTR1 and the DEAH (Asp-Glu-Ala-His)-box RNA heli- case, DHX15, form a stable complex in cells and reciprocally inﬂuence activity and action. DHX15 reduces CMTR1 methyltransferase activity. CMTR1 activates DHX15 helicase activity and inﬂuences nuclear localisation. Disruption of the CMTR1–DHX15 interaction leads to increased ribosome loading of a subset of mRNAs involved in key metabolic functions and impacts on cell proliferation. CMTR1 G-patch domain interacts with the DHX15 oligonucleotide/ oligosaccharide binding (OB)–fold CMTR1 is an 835 residue protein containing an NLS (residues 2–19), a G-patch domain (residues 85–133), an RrmJ-type SAM-dependent O-2 methyltransferase domain (RFM, residues 170–550), a guany- lyltransferase-like domain (GT-like, residues 560–729), and a WW domain (755–790) (Fig 2A) (Aravind & Koonin, 1999; Haline-Vaz et al, 2008; Belanger et al, 2010; Smietanski et al, 2014). To map the interacting regions of CMTR1 and DHX15, HeLa cells were transfected with GFP-CMTR1 wild-type (WT) or mutants ΔN (25–835), ΔG (143–835), 1–143, and Gp (85–143), or GFP alone (Fig 2A). Endogenous DHX15 co-immunoprecipitated with GFP-CMTR1 WT and mutants except GFP-ΔG, the G-patch deletion mutant (Fig 2B). Furthermore, DHX15 co-immunoprecipitated with GFP-Gp, the G-patch domain of CMTR1, indicating that this domain mediates the interaction. Introduction Formation of the mRNA cap initiates the maturation of RNA pol II transcripts into translation-competent mRNA (Furuichi, 2015). The mRNA cap protects transcripts from degradation and recruits protein complexes involved in nuclear export, splicing, 39 processing, and translation initiation (Topisirovic et al, 2011; Ramanathan et al, 2016). mRNA cap formation initiates with the addition of an inverted guanosine group, via a tri-phosphate bridge, to the ﬁrst transcribed nucleotide of nascent RNA pol II transcripts. Subsequently, this guanosine cap is methylated on the N-7 position to create the cap 0 structure, which binds efﬁciently to CBC, eIF4F, and other com- plexes involved in RNA processing and translation initiation. The initial transcribed nucleotides are further methylated at several other positions in a species-speciﬁc manner. In mammals, the O-2 https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 1 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. CMTR1 interacts directly with DHX15 To determine whether the DHX15 OB-fold is required for CMTR1 binding, HA-DHX15 WT or C-terminal deletion mutant (ΔC, 1–635) were expressed in HeLa cells. Endogenous CMTR1 immunoprecipitated with HA-DHX15 WT but not ΔC (Fig 2D). G-patch domains have the consensus sequence, hhxxxGaxxGxGhGxxxxG; “G” is glycine, “h” is a bulky, hydrophobic residue (I, L, V, M), “a” is an aromatic residue (F, Y, W), and “x” is any residue (Aravind & Koonin, 1999). The CMTR1 G-patch has this consensus and in addition leucines at residues 94, 106, and 128 which are conserved in the G-patch domains of the established DHX15 interactors, PINX1, NKRF, GPATCH2, and RBM5 (Fig 2E) (Lin et al, 2009; Niu et al, 2012; Chen et al, 2014; Memet et al, 2017). Because these conserved leucines in other DHX15-binding proteins are required for the interaction, the impact of mutating CMTR1 L94, L106, and L128 to alanine was investigated using the 2L/A mutant (L94A and L106A), To gain insight into CMTR1 and DHX15 complexes in HeLa cells, gel ﬁltration analysis was performed. Used as markers, recombinant CMTR1 and His6-DHX15 monomers migrated as expected for ~100 kD monomers (Fig 1H, lower panels, 14.5 ml). HeLa extracts were treated with RNases I and A before gel ﬁltration to prevent protein–protein interaction via a RNA linker (Figs 1H, upper panels, and S2H). Cellular https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 2 of 2 of 18 Figure 1. DHX15 interacts with CMTR1. (A) HA-CMTR1 was doxycyline-induced (Dox) in a HeLa cell line. Anti-HA antibody immunoprecipitation (IP) was performed on cell extracts. IPs were resolved by SDS–PAGE and Coomassie Blue-stained. HA-CMTR1, DHX15, antibody heavy chain (HC), and light chain (LC) are indicated. (B) pcDNA5 HA-DHX15 WT, 1-635 (ΔC), and vector control (V) were transfected into HeLa cells. Anti-HA IP was performed and analysed as in (A). (C) pcDNA5 GFP-CMTR1 and FLAG-DHX15 were expressed in HeLa cells. IPs were performed on cell extracts using anti-GFP and FLAG antibodies. (D) Endogenous CMTR1 (left panels) or DHX15 (right panels) was immunoprecipitated from extracts of the cell lines indicated; Western blot analysis. Sheep IgG was used for IP control. (E) In HeLa cells, transfection of guide RNAs and Cas9 resulted in a CMTR1 null cell line. Endogenous CMTR1 IP was performed on extracts of HeLa cells (WT) or CMTR1 null HeLa cells (null). CMTR1 interacts directly with DHX15 (F) CMTR1 IPs from HeLa cells were untreated or incubated with RNAase A prior to analysis. (G) 100 ng recombinant CMTR1 and 100 ng His6-DHX15 were mixed and subjected to IP with indicated antibodies. denotes non-speciﬁc band. (H) HeLa cell extracts were treated with RNaseI and RNaseA. Extracts, recombinant CMTR1, or recombinant His6-DHX15 were resolved on a Superdex s200 10/30 column. Fractions were analysed by Western blot Elution of standards indicated Figure 1. DHX15 interacts with CMTR1. (A) HA-CMTR1 was doxycyline-induced (Dox) in a HeLa cell line. Anti-HA antibody immunoprecipitation (IP) was performed on cell extracts. IPs were resolved by SDS–PAGE and Coomassie Blue-stained. HA-CMTR1, DHX15, antibody heavy chain (HC), and light chain (LC) are indicated. (B) pcDNA5 HA-DHX15 WT, 1-635 (ΔC), and vector control (V) were transfected into HeLa cells. Anti-HA IP was performed and analysed as in (A). (C) pcDNA5 GFP-CMTR1 and FLAG-DHX15 were expressed in HeLa cells. IPs were performed on cell extracts using anti-GFP and FLAG antibodies. (D) Endogenous CMTR1 (left panels) or DHX15 (right panels) was immunoprecipitated from extracts of the cell lines indicated; Western blot analysis. Sheep IgG was used for IP control. (E) In HeLa cells, transfection of guide RNAs and Cas9 resulted in a CMTR1 null cell line. Endogenous CMTR1 IP was performed on extracts of HeLa cells (WT) or CMTR1 null HeLa cells (null). (F) CMTR1 IPs from HeLa cells were untreated or incubated with RNAase A prior to analysis. (G) 100 ng recombinant CMTR1 and 100 ng His6-DHX15 were mixed and subjected to IP with indicated antibodies. denotes non-speciﬁc band. (H) HeLa cell extracts were treated with RNaseI and RNaseA. Extracts, recombinant CMTR1, or recombinant His6-DHX15 were resolved on a Superdex s200 10/30 column. Fractions were analysed by Western blot. Elution of standards indicated. Figure 1. DHX15 interacts with CMTR1. Figure 1. DHX15 interacts with CMTR1. Figure 1. DHX15 interacts with CMTR1. (A) HA-CMTR1 was doxycyline-induced (Dox) in a HeLa cell line. Anti-HA antibody immunoprecipitation (IP) was performed on cell extracts. IPs were resolved by SDS–PAGE and Coomassie Blue-stained. HA-CMTR1, DHX15, antibody heavy chain (HC), and light chain (LC) are indicated. (B) pcDNA5 HA-DHX15 WT, 1-635 (ΔC), and vector control (V) were transfected into HeLa cells. Anti-HA IP was performed and analysed as in (A). (C) pcDNA5 GFP-CMTR1 and FLAG-DHX15 were expressed in HeLa cells. IPs were performed on cell extracts using anti-GFP and FLAG antibodies. CMTR1 interacts directly with DHX15 (D) Endogenous CMTR1 (left panels) or DHX15 (right panels) was immunoprecipitated from extracts of the cell lines indicated; Western blot analysis. Sheep IgG was used for IP control. (E) In HeLa cells, transfection of guide RNAs and Cas9 resulted in a CMTR1 null cell line. Endogenous CMTR1 IP was performed on extracts of HeLa cells (WT) or CMTR1 null HeLa cells (null). (F) CMTR1 IPs from HeLa cells were untreated or incubated with RNAase A prior to analysis. (G) 100 ng recombinant CMTR1 and 100 ng His6-DHX15 were mixed and subjected to IP with indicated antibodies. denotes non-speciﬁc band. (H) HeLa cell extracts were treated with RNaseI and RNaseA. Extracts, recombinant CMTR1, or recombinant His6-DHX15 were resolved on a Superdex s200 10/30 column. Fractions were analysed by Western blot. Elution of standards indicated. Figure 1. DHX15 interacts with CMTR1. (A) HA-CMTR1 was doxycyline-induced (Dox) in a HeLa cell line. Anti-HA antibody immunoprecipitation (IP) was performed on cell extracts. IPs were resolved by SDS–PAGE and Coomassie Blue-stained. HA-CMTR1, DHX15, antibody heavy chain (HC), and light chain (LC) are indicated. (B) pcDNA5 HA-DHX15 WT, 1-635 (ΔC), and vector control (V) were transfected into HeLa cells. Anti-HA IP was performed and analysed as in (A). (C) pcDNA5 GFP-CMTR1 and FLAG-DHX15 were expressed in HeLa cells. IPs were performed on cell extracts using anti-GFP and FLAG antibodies. (D) Endogenous CMTR1 (left panels) or DHX15 (right panels) was immunoprecipitated from extracts of the cell lines indicated; Western blot analysis. Sheep IgG was used for IP control. (E) In HeLa cells, transfection of guide RNAs and Cas9 resulted in a CMTR1 null cell line. Endogenous CMTR1 IP was performed on extracts of HeLa cells (WT) or CMTR1 null HeLa cells (null). (F) CMTR1 IPs from HeLa cells were untreated or incubated with RNAase A prior to analysis. (G) 100 ng recombinant CMTR1 and 100 ng His6-DHX15 were mixed and subjected to IP with indicated antibodies. denotes non-speciﬁc band. (H) HeLa cell extracts were treated with RNaseI and RNaseA. Extracts, recombinant CMTR1, or recombinant His6-DHX15 were resolved on a Superdex s200 10/30 column. Fractions were analysed by Western blot. Elution of standards indicated. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 3 of 18 https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 3 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. Figure 2. CMTR1 interacts directly with DHX15 CMTR1 G-patch domain binds to DHX15 OB-fold domain. (A) Diagram of CMTR1 domains and mutants made. (B) HeLa cells were transfected with pcDNA5 (V), or with pcDNA5 GFP, GFP-CMTR1 WT, and mutants. Anti-GFP-antibod IPs (GFP-IP) were performed and analysed by Western blot. Inputs were also analysed (in). (C) Diagram of DHX15 domains and mutants made. (D) HeLa cells were transfected with pcDNA5 (V), pcDNA5 HA-DHX15, and mutants. Anti-HA antibody IPs (HA-IP) were performed and analysed by Western blot. (E) Alignment of G-patch domains in DHX15 interactors generated by Clustal Omega Alignment tool. Conserved leucines in CMTR1 in red. Residues: “.,” weakly similar; “:,” strongly similar; “,” conserved. (F) HeLa cells were transfected with pcDNA5 HA-CMTR1 WT and mutants. HA-IPs were performed and analysed by Western blot. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 4 of 18 Figure 2. CMTR1 G-patch domain binds to DHX15 OB-fold domain. (A) Diagram of CMTR1 domains and mutants made (B) HeLa cells were transfected with pcDNA5 (V) or with pcDNA5 GFP GFP CMTR1 WT and mutants Anti GFP a Figure 2. CMTR1 G-patch domain binds to DHX15 OB-fold domain. Figure 2. CMTR1 G patch domain binds to DHX15 OB fold domain. (A) Diagram of CMTR1 domains and mutants made. (B) HeLa cells were transfected with pcDNA5 (V), or with pcDNA5 GFP, GFP-CMTR1 WT, and mutants. Anti-GFP-antibody IPs (GFP-IP) were performed and analysed by Western blot. Inputs were also analysed (in). (C) Diagram of DHX15 domains and mutants made. (D) HeLa cells were transfected with pcDNA5 (V), pcDNA5 HA-DHX15, and mutants. Anti-HA antibody IPs (HA-IP) were performed and analysed by Western blot. (E) Alignment of G-patch domains in DHX15 interactors generated by Clustal Omega Alignment tool. Conserved leucines in CMTR1 in red. Residues: “.,” weakly similar; “:,” strongly similar; “,” conserved. (F) HeLa cells were transfected with pcDNA5 HA-CMTR1 WT and mutants. HA-IPs were performed and analysed by Western blot. p agram of CMTR1 domains and mutants made. (B) HeLa cells were https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 4 of 1 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 4 of 18 Figure 3. CMTR1 methyltransferase activity is repressed by DHX15. (A) GpppG capped RNA was incubated with 3 nM CMTR1 or 3 nM CMTR1 and 3 nM His6 DHX15 for the time indicated Caps throughout the ﬁgure are 32P labelled on Figure 3. Percentage conversion of GpppG to GpppGm (% GpppG O-2 meth) is plotted. Average and standard deviation for three independent experiments are plotted here and throughout the ﬁgure. (C) m7GpppG-capped RNA was incubated with 0.5 nM CMTR1 and indicated nM His6-DHX15 for 30 min. Generated of m7GpppGm detected. (D) Average percentage of m7GpppG O-2 meth and standard deviation are plotted. (E) HeLa cells were transfected with 5 μg pcDNA GFP or GFP-DHX15 and 0, 0.4, or 2 μg pcDNA5 HA-CMTR1, as indicated. 1 d after transfection, cells were harvested and 5 μg cell extract was used in O-2 methyltransferase assay. Percentage of GpppG O-2 meth and standard deviation of duplicates are reported. (F) 100 ng recombinant GST-CMTR1ΔG and 100 ng His6-DHX15 were mixed and subjected to IP with indicated antibodies; Western blot analysis. denotes non-speciﬁc band. (G) GpppG-capped RNA was incubated with 5 nM GST-CMTR1ΔG or 5 nM GST-CMTR1ΔG and 5 nM His6-DHX15. Average percentage of GpppG O-2 meth and standard deviation are plotted. (H) GpppG-capped RNA was incubated with 3 nM CMTR1 and 3 nM His6-DHX15 WT or E261Q. Average percentage of GpppG O-2 meth and standard deviation are plotted. Where relevant, t test was performed. P-value < 0.05; P-value < 0.01; P-value < 0.005. (I) 55 nt 32P GpppG-RNA incubated with 100 ng CMTR1 and indicated ng His6-DHX15. Complexes were resolved in a 4%–20% tris-borate-EDTA buffer non-denaturing acrylamide gel and detected using a phosphorimager (GE healthcare Life Sciences). Migration of 32P GpppG-RNA probe and complexes are indicated. CMTR1 increases DHX15 helicase activity To determine the impact of the CMTR1–DHX15 interaction, we also investigated whether CMTR1 inﬂuences DHX15 ATPase and helicase activity (Fig 4). Recombinant His6-DHX15 was incubated with α32P- ATP and hydrolysis visualised by the detection of α32P-ADP (Fig 4A and B). As established, addition of RNA signiﬁcantly increased ATP hydrolysis (Fig 4A and B, and S5) (Walbott et al, 2010; Memet et al, 2017). Previously characterised G-patch–containing interactors of Prp43 or DHX15 have variable impact on ATPase activity (Tanaka et al, 2007; Lebaron et al, 2009; Niu et al, 2012). To determine whether CMTR1 inﬂuences DHX15-dependent ATP hydrolysis, a titration of recombinant CMTR1 was included in the ATPase assay. CMTR1 did not activate or repress basal DHX15-dependent ATP hydrolysis (Fig 4C and D). Furthermore, CMTR1 did not activate or repress RNA- stimulated DHX15-dependent ATP hydrolysis (Fig 4E). Addition of SAM had no impact on DHX15-dependent ATP hydrolysis, including in the presence of CMTR1 or RNA (Fig 4F). To investigate whether the interaction of DHX15 and CMTR1 is required for inhibition of methyltransferase activity, CMTR1ΔG mutant, which does not interact with DHX15, was utilised (Figs 2A and B, and 3F). Recombinant CMTR1ΔG had slightly reduced activity compared with WT (Fig S4F); however, its methyltransferase activity was not inhibited by DHX15 (Fig 3G). This indicates that direct binding of DHX15 is required for the inhibition of CMTR1 methyl- transferase activity. Because DHX15 is a helicase, we determined the requirement for its catalytic activity to inhibit CMTR1 methyl- transferase activity. Incubation of recombinant CMTR1 with equi- molar recombinant His6-DHX15 WT or E261Q (Memet et al, 2017), a catalytically inactive mutant, resulted in an equivalent reduction in methyltransferase activity (Fig 3H). Therefore, inhibition of CMTR1 activity is a non-catalytic function of DHX15. Because interactors of Prp43, the yeast DHX15 homologue, can regulate helicase activity independently of ATPase activity, we in- vestigated whether CMTR1 could regulate DHX15 helicase activity (Tanaka et al, 2007). A 32P-DNA-RNA duplex was incubated with 1,000 nM His6-DHX15 and helicase activity was conﬁrmed by the loss of the duplex (Fig 4G, compare lanes 2 and 7). When 100 nM His6-DHX15 was used in this assay, reduced helicase activity was observed, as expected (lane 3). Addition of CMTR1 increased helicase activity in a dose-dependent manner (Fig 4G, lanes 4–6, and H). CMTR1 alone had no helicase activity (Fig 4G, lane 8). DHX15 inhibits CMTR1 methyltransferase activity To investigate the biochemical impact of the CMTR1–DHX15 in- teraction, we ﬁrst analysed CMTR1 methyltransferase activity. A 32P- labelled, guanosine-capped transcript (GpppG) was incubated with recombinant CMTR1 and a methyl donor, SAM (Fig 3A). First tran- scribed nucleotide O-2 methylation (GpppG to GpppGm conversion) was observed by resolution on thin-layer chromatography (Belanger et al, 2010). Incubation of recombinant CMTR1 with equimolar recombinant His6-DHX15 resulted in an approximately 50% reduction in O-2 methylation (Fig 3A and B). Similarly, His6-DHX15 signiﬁcantly inhibited O-2 methylation of a 7-methylguanosine-capped tran- script (m7GpppG to m7GpppGm conversion) (Figs 3C and D, and S4A and B). His6-DHX15 inhibited CMTR1-dependent methylation in a dose-dependent manner (Fig 3C and D). The impact of DHX15 on CMTR1-methyltransferase activity was also observed in cells. Trans- fection of HeLa cells with a titration of pCDNA5 HA-CMTR1 resulted in dose-dependent O-2 methyltransferase activity in cell extracts (Fig 3E). Transfection with pCDNA5 GFP-DHX15 inhibited methyl- transferase activity in cell extracts. As controls, ATP did not alter DHX15-dependent repression of CMTR1, and BSA and recombinant protein storage buffer did not affect CMTR1-dependent methylation (Fig S4C–E). and the 3L/A mutant (L94A, L106A, and L128A) (Fig 2E). The 2L/A mutation was sufﬁcient to abrogate the interaction of CMTR1 with DHX15, conﬁrming that CMTR1 interacts with DHX15 through the G-patch domain (Fig 2F). and the 3L/A mutant (L94A, L106A, and L128A) (Fig 2E). The 2L/A mutation was sufﬁcient to abrogate the interaction of CMTR1 with DHX15, conﬁrming that CMTR1 interacts with DHX15 through the G-patch domain (Fig 2F). CMTR1–RNA interactions in cells. However, recombinant CMTR1 could be detected interacting with GpppG-RNA, visualised as mobility shifts on a non-denaturing acrylamide gel (Fig 3I). Two major mobility CMTR1-RNA shifts were observed. Because the recombinant CMTR1 used in these studies was highly puriﬁed (Fig S2E), two mobility shifts is likely to indicate conformational isomers in CMTR1-RNA complexes. An interaction between His6-DHX15 alone and GpppG-RNA was not observed. In the presence of increasing amounts of DHX15, the CMTR1 GpppG-RNA complex was not reduced and further bands were ob- served, indicating GpppG-RNA-CMTR1-DHX15 complexes. Although in vitro, these RNA-protein interaction studies indicate that DHX15 does not reduce the afﬁnity of CMTR1 for the GpppG-RNA substrate. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 CMTR1 interacts directly with DHX15 CMTR1 methyltransferase activity is repressed by DHX15. Figure 3. CMTR1 methyltransferase activity is repressed by DHX15. (A) GpppG-capped RNA was incubated with 3 nM CMTR1 or 3 nM CMTR1 and 3 nM His6-DHX15 for the time indicated. Caps throughout the ﬁgure are 32P-labelled on α-phosphate. Generated GpppGm (ﬁrst transcribed nucleotide ribose 0–2 methylated) resolved by thin-layer chromatography and detected by autoradiography. (B) Figure 3. CMTR1 methyltransferase activity is repressed by DHX15. (A) GpppG-capped RNA was incubated with 3 nM CMTR1 or 3 nM CMTR1 and 3 nM His6-DHX15 for the time indicated. Caps throughout the ﬁgure are 32P-labelled on α-phosphate. Generated GpppGm (ﬁrst transcribed nucleotide ribose 0–2 methylated) resolved by thin-layer chromatography and detected by autoradiography. (B) https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 5 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. CMTR1 interaction with DHX15 and RNA pol II CTD are mutually exclusive To gain insight into the mechanism of CMTR1 inhibition by DHX15, we investigated the impact on RNA binding. Unfortunately, we were unable to purify CMTR1 bound to RNA in cells using cross-linking methodologies which however detected RNMT–RNA interactions (Varshney et al, 2018). This may reﬂect the dynamic nature of CMTR1 interacts with RNA pol II, which may permit efﬁcient co- transcriptional ﬁrst nucleotide O-2 methylation (Haline-Vaz et al, https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 6 of 18 2008). We investigated the impact of DHX15 on CMTR1 recruitment to RNA pol II. The interaction of recombinant CMTR1 and DHX15 with a biotinylated peptide consisting of three RNA pol II C-terminal domain heptad repeats, (YSPTSPS)3, unphosphorylated (CTD), or phosphorylated on serine-5 (pCTD), was investigated (Fig 5A–C). As a control, RNGTT was demonstrated to interact with pCTD but not CTD (Fig S6A) (Ho & Shuman, 1999). The CMTR1 monomer bound to pCTD but not to CTD (Fig 5A), whereas the His6-DHX15 monomer did not bind to either peptide (Fig 5B). To investigate whether CMTR1 can recruit DHX15 to pCTD, the recombinant proteins were mixed before peptide pulldown. Although CMTR1 and DHX15 interact in vitro (Fig 1G), CMTR1 was recruited to pCTD, whereas DHX15 was not, revealing that DHX15–CMTR1 complexes are not recruited to pCTD (Fig 5C). Because we had observed that CMTR1 stimulates DHX15 helicase activity in vitro, we investigated whether it also inﬂuences DHX15 localisation in cells. Prp43, the yeast homologue of DHX15, occupies several cellular locations to execute different biological functions (Heininger et al, 2016). The different Prp43 locations are achieved by the competition of G-patch proteins, which recruit the helicase to different parts of the cell. In HeLa cells, DHX15 exhibited a diffuse nuclear localisation (Fig 6B and C). However, in approximately a third of cells, DHX15 also exhibited speckled nuclear foci, a common feature of splicing factors (Tannukit et al, 2009). DHX15 foci partially co-localized with the splicing factor SC35 (Pawellek et al, 2014) (Fig S7A). The number of cells exhibiting DHX15 nuclear foci increased when CMTR1 was knocked down (Figs 6C and S7B) or in the cmtr1 null HeLa cell line (Fig S7C and D). The potential impact of CMTR1 on DHX15 localisation is explored in the Discussion section. The interaction of CMTR1 and DHX15 with RNA pol II was in- vestigated in cells. II (Fig 5E and F). Initially we attempted to express CMTR1 WT and 2L/A mutant in CMTR1 null HeLa cells (Fig 1E). CMTR1 WT and 2L/A were expressed equivalently in the ﬁrst 3 d after initial transduction. However, after selection, cells expressing the WT protein survived, whereas cells expressing the 2L/A mutant did not, indicating some growth dis- advantage of the mutant. Therefore, CMTR1 2L/A was expressed in the presence of endogenous CMTR1. HCC1806 cells, a mammary epithelial tumour cell line, were selected for these experiments because we had previously established that these cells are highly sensitive to changes in the CMTR1 expression level. Cells were transfected with pcDNA5 HA-CMTR1 WT, 2L/A, or vector control and polysome proﬁling analysis was performed in which free ribosomes and ribosomal subunits are separated from translating ribosomes in a sucrose gradient (Fig 7A). Expression of HA-CMTR1 WT and 2/LA were equivalent (Fig 7B). Although expression of CMTR1 WT and 2L/A had a mild impact on the polysome proﬁles, in multiple experi- ments, the ratios between polysomes and monosomes were not signiﬁcantly different, indicating that the CMTR1–DHX15 interaction was not having a broad impact on translational control (Fig 7C). To investigate gene-speciﬁc effects, RNA sequencing (RNAseq) anal- ysis was used to quantify total cellular transcripts and the tran- scripts associated with most dense ribosome binding (Fig 7A, shaded area, and Tables S1 and S2). Out of the 12,238 gene tran- scripts that passed quality thresholds, none exhibited a signiﬁcant difference in expression level in cells expressing CMTR1 WT and 2L/A (Table S1 and Fig 7D). This indicates that in HCC1806 cells, the CMTR1–DHX15 interaction is unlikely to have a signiﬁcant impact on Although DHX15 is not required for CMTR1 recruitment to pSer5- CTD (Fig 5A, C, E, and F), it was important to determine whether DHX15 binds to RNA pol II in cells, either independently of CMTR1 or in a complex with it. HA-DHX15 was expressed in HeLa cells and immunoprecipitated via the HA tag (Fig 5G). As expected, HA-DHX15 bound to CMTR1 but not to pSer5-CTD or unphosphorylated RNA pol II. This conﬁrms that in cells, as in vitro, DHX15 does not ap- preciably interact with RNA pol II (Fig 5B, C, and G) and, furthermore, that the CMTR1-DHX15 complex does not interact with RNA pol II (Fig 5C and G). CMTR1 interaction with DHX15 and RNA pol II CTD are mutually exclusive In HeLa cell extracts, endogenous CMTR1 co- immunoprecipitated with RNA pol II phospho-Ser5-CTD (pSer5-CTD), but not phospho-Ser2-CTD (pSer2-CTD) or unphosphorylated RNA pol II (Fig 5D). The interaction of CMTR1 and pSer5-CTD was RNase A–insensitive and, therefore, RNA-independent (Fig S6B). To map the interaction of CMTR1 with RNA pol II, GFP-CMTR1 WT; 1–143, ΔG, and ΔC (1–173); and HA-CMTR1, 2L/A, and ΔC were expressed in HeLa cells and immunoprecipitated via their tags, and pSer5-CTD binding was determined (Fig 5E and F). GFP-CMTR1 and HA-CMTR1 interacted with pSer5-CTD, whereas GFP-CMTR1ΔC and HA-CMTR1ΔC did not, indicating that the WW domain mediates the interaction with RNA pol II (Fig 5E and F). GFP-CMTR1ΔG and HA-CMTR1 2L/A (both de- fective for DHX15 binding) interacted with pSer5-CTD, conﬁrming that in cells DHX15 is not required for CMTR1 recruitment to RNA pol II (Fig 5E and F). DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. DHX15 represses CMTR1-dependent translation We had observed that DHX15 inhibits CMTR1-dependent O-2 methylation (Fig 3), CMTR1 activates DHX15-dependent helicase activity (Fig 4), and DHX15–CMTR1 complexes are not present on RNA pol II (Fig 5). The net outcome of the DHX15 and CMTR1 relationships is therefore complex and will depend on many factors (see the Discussion section). Here, we focussed our investigation on the impact of DHX15 on CMTR1 function using the CMTR1 2L/A mutant, which does not bind to DHX15 (Fig 2F), but is recruited to RNA pol II (Fig 5F). Because ﬁrst nucleotide ribose O-2 methylation is asso- ciated with enhanced translation, the impact of the DHX15–CMTR1 interaction on this process was investigated (Kuge et al, 1998). https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 7 of 18 CMTR1 inﬂuences DHX15 localisation Following biochemical analysis of the CMTR1-DHX15 interaction, we investigated whether DHX15 inﬂuences CMTR1 localisation. We conﬁrmed that GFP-CMTR1 and endogenous CMTR1 are predomi- nantly nuclear in HeLa cells (Fig 6A–C) (Haline-Vaz et al, 2008). A potential CMTR1 nuclear localisation signal was identiﬁed, 14KKQKK (Lange et al, 2007). Mutation of CMTR1 14KKQKK to 14EEQEE (4K/E) or removal of the ﬁrst 25 residues (ΔN) resulted in partial cytoplasmic localisation, conﬁrming that 14KKQKK contributes to nuclear localisation. To determine the impact of DHX15 on CMTR1 local- isation, the localisation of GFP-CMTR1 3L/A (does not bind DHX15) was investigated. GFP-CMTR1 3L/A was predominantly nuclear, indicating that DHX15 does not inﬂuence CMTR1 localisation (Fig 6A). Furthermore, suppression of DHX15 expression did not alter CMTR1 nuclear localisation (Fig 6B). https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 7 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 7 of 18 Figure 4. CMTR1 methyltransferase stimulates DHX15 helicase activity. (A) 100 nM His6-DHX15 was incubated with α32P-ATP in the absence or presence of 1 μg HeLa cell RNA for the time indicated. The ADP generated was resolved by TLC and detected by phosphoimager throughout the ﬁgure unless indicated. (B) Average percentage of ATP hydrolysis and standard deviation are plotted for two experiments. (C) 100 nM His6-DHX15 was incubated with α32P-ATP and 1 μg RNA or nM CMTR1 indicated for 30 min. ATP hydrolysis was detected. (D) Average percentage ATP hydrolysis and standard deviation are plotted for three independent experiments. t test was performed relative to DHX15 alone. P-value < 0.005. (E) 100 nM His6-DHX15 alone or 100 nM His6-DHX15, 1 μg HeLa RNA, and indicated concentration of CMTR1 were incubated with α32P-ATP for 30 min. Percentage of ATP hydrolysis is plotted over time. (F) 100 nM DHX15 was incubated with 10 μM SAM, 1 μg HeLa RNA, and 100 nM CMTR1 as indicated for 30 min. Percentage of ATP hydrolysis is plotted. (G) RNA-32P-DNA duplex was incubated with with1mM ATP and indicated nM His6-DHX15 and CMTR1. After 30 min, samples were resolved by native PAGE and visualized using a phosphorimager. “Unwound” indicates single stranded 32P-DNA oligonucleotide. ΔT is the 95°C denatured substrate. (H) Average percentages of unwound substrate and standard deviations are presented for three independent experiments. Figure 4. CMTR1 methyltransferase stimulates DHX15 helicase activity. Figure 4. CMTR1 methyltransferase stimulates DHX15 helicase activity. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. CMTR1 inﬂuences DHX15 localisation CMTR1 interaction with DHX15 and RNA pol II CTD are mutually exclusive. ( ) bi ( ) bi i 6 ( ) i d d i 6 Figure 5. CMTR1 interaction with DHX15 and RNA pol II CTD are mutually exclusive. (A) Recombinant CMTR1, (B) recombinant His6-DHX15, (C) pre-mixed CMTR1, and His6-DHX15, were incubated with biointinylated peptides of three RNA pol II CTD heptad repeats, either unphosphorylated (CTD) or phosphorylated on S5 (pCTD). Peptides were enriched on streptavidin beads and associated proteins analysed by Western blot. (D) CMTR1 was immunoprecipitated from extracts of HeLa cells (WT) or HeLa CMTR1 null cells. Western blot analysis was performed. (E) pcDNA5 GFP-CMTR1 or indicated mutants were expressed in HeLa cells. IPs were performed with anti-GFP antibodies and analysed by Western blot. (F) pcDNA 5 HA-CMTR1 or indicated mutants were expressed in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. (G) pcDNA5 HA-CMTR1 and mutants were expressed in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. Aff, afﬁnity puriﬁed fraction; In, input; FT, ﬂow through. g p y (A) Recombinant CMTR1, (B) recombinant His6-DHX15, (C) pre-mixed CMTR1, and His6-DHX15, were incubated with biointinylated heptad repeats, either unphosphorylated (CTD) or phosphorylated on S5 (pCTD). Peptides were enriched on streptavidin beads Western blot. (D) CMTR1 was immunoprecipitated from extracts of HeLa cells (WT) or HeLa CMTR1 null cells. Western blot analysis or indicated mutants were expressed in HeLa cells. IPs were performed with anti-GFP antibodies and analysed by Western blo mutants were expressed in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. (G) pcDNA5 H in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. Aff, afﬁnity puriﬁed fraction; In, input processes (Table S3). Notably, there was enrichment for gene encoding factors involved with the cell cycle and DNA damage responses including ATR, UBR4, UBR2, CENPE, PRKDC, and BIRC6; factors involved in RNA processing including GCN1, TPR, RANBP2, and NUP205; metabolic enzymes including FASN, EPRS, and CAD; and focal adhesion-associated molecules HSPG2, FLNB, MYH9, FAT1, IGF2R, and CLTC. The polysome loading of selected genes was con- ﬁrmed by RT–PCR and enhanced protein expression conﬁrmed by Western blot (Fig 7G and H). This analysis indicates that DHX15 processes (Table S3). DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. CMTR1 inﬂuences DHX15 localisation (A) 100 nM His6-DHX15 was incubated with α32P-ATP in the absence or presence of 1 μg HeLa cell RNA for the time indicated. The ADP generated was resolved by TLC and detected by phosphoimager throughout the ﬁgure unless indicated. (B) Average percentage of ATP hydrolysis and standard deviation are plotted for two experiments. (C) 100 nM His6-DHX15 was incubated with α32P-ATP and 1 μg RNA or nM CMTR1 indicated for 30 min. ATP hydrolysis was detected. (D) Average percentage ATP hydrolysis and standard deviation are plotted for three independent experiments. t test was performed relative to DHX15 alone. P-value < 0.005. (E) 100 nM His6-DHX15 alone or 100 nM His6-DHX15, 1 μg HeLa RNA, and indicated concentration of CMTR1 were incubated with α32P-ATP for 30 min. Percentage of ATP hydrolysis is plotted over time. (F) 100 nM DHX15 was incubated with 10 μM SAM, 1 μg HeLa RNA, and 100 nM CMTR1 as indicated for 30 min. Percentage of ATP hydrolysis is plotted. (G) RNA-32P-DNA duplex was incubated with with1mM ATP and indicated nM His6-DHX15 and CMTR1. After 30 min, samples were resolved by native PAGE and visualized using a phosphorimager. “Unwound” indicates single stranded 32P-DNA oligonucleotide. ΔT is the 95°C denatured substrate. (H) Average percentages of unwound substrate and standard deviations are presented for three independent experiments. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 8 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. Figure 5. CMTR1 interaction with DHX15 and RNA pol II CTD are mutually exclusive. (A) Recombinant CMTR1, (B) recombinant His6-DHX15, (C) pre-mixed CMTR1, and His6-DHX15, were incubated with biointinylated peptides of three RNA pol II CTD heptad repeats, either unphosphorylated (CTD) or phosphorylated on S5 (pCTD). Peptides were enriched on streptavidin beads and associated proteins analysed by Western blot. (D) CMTR1 was immunoprecipitated from extracts of HeLa cells (WT) or HeLa CMTR1 null cells. Western blot analysis was performed. (E) pcDNA5 GFP-CMTR1 or indicated mutants were expressed in HeLa cells. IPs were performed with anti-GFP antibodies and analysed by Western blot. (F) pcDNA 5 HA-CMTR1 or indicated mutants were expressed in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. (G) pcDNA5 HA-CMTR1 and mutants were expressed in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. Aff, afﬁnity puriﬁed fraction; In, input; FT, ﬂow through. Figure 5. Figure 5. CMTR1 interaction with DHX15 and RNA pol II CTD are mutually exclusive. (A) Recombinant CMTR1, (B) recombinant His6-DHX15, (C) pre-mixed CMTR1, and His6-DHX15, were incubated with biointinylated peptides of three RNA pol II CTD heptad repeats, either unphosphorylated (CTD) or phosphorylated on S5 (pCTD). Peptides were enriched on streptavidin beads and associated proteins analysed by Western blot. (D) CMTR1 was immunoprecipitated from extracts of HeLa cells (WT) or HeLa CMTR1 null cells. Western blot analysis was performed. (E) pcDNA5 GFP-CMTR1 or indicated mutants were expressed in HeLa cells. IPs were performed with anti-GFP antibodies and analysed by Western blot. (F) pcDNA 5 HA-CMTR1 or indicated mutants were expressed in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. (G) pcDNA5 HA-CMTR1 and mutants were expressed in HeLa cells. IPs were performed with anti-HA antibodies and analysed by Western blot. Aff, afﬁnity puriﬁed fraction; In, input; FT, ﬂow through. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 9 of 18 CMTR1 inﬂuences DHX15 localisation (A) Representative ﬂuorescence images of GFP-CMTR1 WT or mutants expressed in HeLa cells. Cells were DAPI-stained to visualise DNA. (B) HeLa cells were transfected with DHX15 siRNA or non-targeting control. Representative immunoﬂuorescence (IF) images of endogenous CMTR1 and DHX15 presented. Images show CMTR1 (red), DHX15 (green), and DAPI (blue). (C) As in (B), except cells were transfected with CMTR1 siRNA or control. Yellow asterisks indicate cells with DHX15 accumulation in foci. 4.5× magniﬁed areas marked. Data representative of three independent experiments. Bar indicates 20 μm. represses the positive effect of CMTR1 on cell growth and proliferation. represses the positive effect of CMTR1 on cell growth and proliferation. resulted in increased polysome loading of genes involved in metabolism and cell cycle, we investigated its impact on cell pro- liferation. When HA-CMTR1 was transiently expressed in sparsely plated HCC1806 cells, a signiﬁcant increase in cell number was observed after 24 h (Fig 8D). Transient expression of HA-CMTR1 2L/A resulted in a further increase in cell number, supporting the hy- pothesis that DHX15 suppresses the translation of a subset of pro- growth genes. To investigate if the increased expression of these genes may contribute to the enhanced proliferation observed on expression of CMTR1 2L/A, two genes with increased polysome CMTR1 inﬂuences DHX15 localisation Notably, there was enrichment for gene encoding factors involved with the cell cycle and DNA damage responses including ATR, UBR4, UBR2, CENPE, PRKDC, and BIRC6; factors involved in RNA processing including GCN1, TPR, RANBP2, and NUP205; metabolic enzymes including FASN, EPRS, and CAD; and focal adhesion-associated molecules HSPG2, FLNB, MYH9, FAT1, IGF2R, and CLTC. The polysome loading of selected genes was con- ﬁrmed by RT–PCR and enhanced protein expression conﬁrmed by Western blot (Fig 7G and H). This analysis indicates that DHX15 transcription or RNA stability. However, 59 gene transcripts were transcription or RNA stability. However, 59 gene transcripts were signiﬁcantly enriched in polysomes in cells expressing CMTR1 2L/A compared with WT. This indicates that the DHX15–CMTR1 interaction restricts the translation of a subset of mRNAs (Fig 7E and F and Table S2). Conversely, no genes were signiﬁcantly depleted from polysomes in cells expressing CMTR1 2L/A compared with WT. The genes that exhibited enhanced translation in cells expressing CMTR1 2L/A were enriched for gene ontology terms associated with important metabolic functions and cell cycle https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 9 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 9 of 18 Figure 6. DHX15 localisation is inﬂuenced by CMTR1. (A) Representative ﬂuorescence images of GFP-CMTR1 WT or mutants expressed in HeLa cells. Cells were DAPI-stained to visualise DNA. (B) HeLa cells were transfected with DHX15 siRNA or non-targeting control. Representative immunoﬂuorescence (IF) images of endogenous CMTR1 and DHX15 presented. Images show CMTR1 (red), DHX15 (green), and DAPI (blue). (C) As in (B), except cells were transfected with CMTR1 siRNA or control. Yellow asterisks indicate cells with DHX15 accumulation in foci. 4.5× magniﬁed areas marked. Data representative of three independent experiments. Bar indicates 20 μm. Figure 6. DHX15 localisation is inﬂuenced by CM (A) Representative ﬂuorescence images of GFP-CM WT or mutants expressed in HeLa cells. Cells wer DAPI-stained to visualise DNA. (B) HeLa cells wer transfected with DHX15 siRNA or non-targeting con Representative immunoﬂuorescence (IF) images o endogenous CMTR1 and DHX15 presented. Images show CMTR1 (red), DHX15 (green), and DAPI (blue) As in (B), except cells were transfected with CMTR siRNA or control. Yellow asterisks indicate cells w DHX15 accumulation in foci. 4.5× magniﬁed areas marked. Data representative of three independen experiments. Bar indicates 20 μm. Figure 6. DHX15 localisation is inﬂuenced by CMTR1. Figure 6. DHX15 localisation is inﬂuenced by CMTR1. (A) Representative ﬂuorescence images of GFP-CMTR1 WT or mutants expressed in HeLa cells. Cells were DAPI-stained to visualise DNA. (B) HeLa cells were transfected with DHX15 siRNA or non-targeting control. Representative immunoﬂuorescence (IF) images of endogenous CMTR1 and DHX15 presented. Images show CMTR1 (red), DHX15 (green), and DAPI (blue). (C) As in (B), except cells were transfected with CMTR1 siRNA or control. Yellow asterisks indicate cells with DHX15 accumulation in foci. 4.5× magniﬁed areas marked. Data representative of three independent experiments. Bar indicates 20 μm. The interaction of DHX15 and CMTR1 inhibits proliferation Average and standard deviation presented for 3–5 independent wells. (E) As in (D) except cells were also transfected with and 50 nM CAD, GCN1 or non-targeting control siRNA. t test was performed. P-value < 0.05; P-value < 0.01; P-value < 0.005. 1998; Schuberth-Wagner et al, 2015; Leung & Amarasinghe, 2016). We investigated cellular regulation of CMTR1, the ﬁrst nucleotide ribose O-2 methyltransferase (Belanger et al, 2010). CMTR1 was isolated from mammalian cells in a complex with DHX15 (human orthologue of yeast Prp43), a DEAH-box helicase involved in RNA processing, splicing, and ribosome biogenesis (Koodathingal & Staley, 2013; Robert-Paganin et al, 2015). DHX15 has an OB-fold domain through which it binds to the G-patch domain in a series of proteins, fa- cilitating contribution to several nuclear functions (Niu et al, 2012; Chen et al, 2014; Robert-Paganin et al, 2015; Memet et al, 2017; Tauchert et al, 2017). Previous studies indicated that CMTR1 and DHX15 participate in the same mRNA processing events (Yoshimoto loading in CMTR1 2L/A cells, CAD and GCN1, were suppressed by siRNA transfection (Fig 8E). Suppression of CAD and GCN1 resulted in reduced proliferation of HCC1806 cells, supporting the hypoth- esis that DHX15 controls CMTR1-dependent translation of a subset of pro-growth genes. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 The interaction of DHX15 and CMTR1 inhibits proliferation The impact of CMTR1 expression on cell proliferation was in- vestigated in HCC1806 cells and another mammary epithelial tu- mour line, MCF7. Transfection of two independent CMTR1 siRNAs resulted in the suppression of CMTR1 expression and a reduction in cell proliferation (Fig 8A–C). Given that expression of CMTR1 2L/A https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 10 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 11 of 18 https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 11 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. Figure 8. The DHX15–CMTR1 interaction inhibits cell proliferation. (A) MCF7 cells were transfected with two independent CMTR1 siRNAs or non-targeting control. Cell extracts were analysed by Western blot. (B) MCF7, (C) HCC1806 cells were transfected with two independent CMTR1 siRNAs or non-targeting control. Cells were counted every day. HCC1806 were transfected with (D) pcDNA5 CMTR1 WT or 2L/A or vector control. After 48 h cells were counted. Average and standard deviation presented for 3–5 independent wells. (E) As in (D) except cells were also transfected with and 50 nM CAD, GCN1 or non-targeting control siRNA. t test was performed. P-value < 0.05; P-value < 0.01; P-value < 0.005. Figure 8. The DHX15–CMTR1 interaction inhibits cell proliferation. ( ) Figure 8. The DHX15 CMTR1 interaction inhibits cell proliferation. (A) MCF7 cells were transfected with two independent CMTR1 siRNAs or non-targeting control. Cell extracts were analysed by Western blot. (B) MCF7, (C) HCC1806 cells were transfected with two independent CMTR1 siRNAs or non-targeting control. Cells were counted every day. HCC1806 were transfected with (D) pcDNA5 CMTR1 WT or 2L/A or vector control. After 48 h cells were counted. Average and standard deviation presented for 3–5 independent wells. (E) As in (D) except cells were also transfected with and 50 nM CAD, GCN1 or non-targeting control siRNA. t test was performed. P-value < 0.05; P-value < 0.01; P-value < 0.005. Figure 8. The DHX15 CMTR1 interaction inhibits cell proliferation. (A) MCF7 cells were transfected with two independent CMTR1 siRNAs or non-targeting control. Cell extracts were analysed by Western blot. (B) MCF7, (C) HCC1806 cells were transfected with two independent CMTR1 siRNAs or non-targeting control. Cells were counted every day. HCC1806 were transfected with (D) pcDNA5 CMTR1 WT or 2L/A or vector control. After 48 h cells were counted. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. DHX15 constrains CMTR1 function We observe that CMTR1 activates DHX15 helicase activity. Similar to several other G-patch proteins, CMTR1 activates helicase activity without regulating ATPase activity (Memet et al, 2017). A similar ob- servation was made with the yeast splicing factor Ntr1, which does not alter Prp43 ATPase activity but does activate the helicase (Tanaka et al, 2007). Furthermore, there are now several examples of G-patch proteins impacting differentially on Prp43 activity and function (Aravind & Koonin, 1999; Banerjee et al, 2015; Tauchert et al, 2017). mRNA cap formation initiates early during transcription, when RNGTT (RNA triphosphatase/guanylyltransferase) and RNMT (N-7 cap guanosine methyltransferase) are recruited to Ser-5 phos- phorylated RNA pol II CTD (Ramanathan et al, 2016). We report that CMTR1 also interacts directly with Ser-5 phosphorylated RNA pol II CTD. DHX15 reduces CMTR1 methyltransferase activity in a dose- dependent manner and DHX15–CMTR1 complexes are not recruited to RNA pol II. Thus, the interaction of DHX15 with CMTR1 is likely to constrain ﬁrst nucleotide O-2 methylation to the initial stages of transcription, and restrain post-transcriptional, aberrant methylation. We demonstrate that CMTR1 interaction with DHX15 requires the G-patch domain, whereas the CMTR1 interaction with RNA pol II CTD requires the WW domain. The CMTR1-DHX15 complex may be unable to be recruited to RNA pol II either because DHX15 changes the conformation of CMTR1 or sterically hinders it from binding to the CTD. The competition of cofactors for Prp43 was previously observed to regulate its distribution between different pathways (Heininger et al, 2016). Consistent with a previous publication, we observed DHX15 co- localising with a splicing factor, SC35, in nuclear speckles (Tannukit et al, 2009). Suppression of CMTR1 expression resulted in an increased number of cells with DHX15 in nuclear speckles. CMTR1 may compete with other G-patch proteins for DHX15 binding and/or the impact that CMTR1 has on the cell cycle may indirectly inﬂuence the number of nuclear speckles. The indirect impact of CMTR1 suppression can be observed by a change in nuclear morphology from spheroid to lobed (Fig 6), and by an impact on cell proliferation (Fig 8). To our knowledge, CMTR1 is the only annotated G-patch–containing protein with a catalytic domain and this is the ﬁrst example of DHX15 regulating a catalytic activity. Interaction with DHX15 is required for inhibition of CMTR1 methyltransferase activity; however, the helicase/ATPase activity of DHX15 is not. Discussion Higher eukaryotes carry unique mRNA cap modiﬁcations, including ﬁrst transcribed nucleotide ribose O-2 methylation, which is as- sociated with translation and self-RNA identiﬁcation (Kuge et al, Figure 7. The DHX15–CMTR1 interaction inhibits translation of a subset of genes. ( ) Figure 7. The DHX15–CMTR1 interaction inhibits translation of a subset of genes. (A) Extracts from HCC1806 cells transfected with pCDNA5 HA-CMTR1, 2L/A, or vector control (V) were centrifuged through 10%–50% sucrose gradients. 259 nm absorbance was determined across the gradient in four independent experiments. Representative experiment presented. The fraction of the gradients analysed by RNAseq is indicated by shaded boxes. The identities of the peaks in the polysome proﬁles is indicated. (B) Cell extracts were analysed by Western blot. (C) Average and standard deviation of ratio of polysome to mononosome absorbance for four independent experiments. RNAseq analysis was used to determine uniquely aligned reads per gene (transcript isoforms collapsed) in RNA and polysome samples from HCC1806 cells expressing HA-CMTR1 WT and 2L/A (n = 2). (D) Scatter plot of log2 transformed total counts per million (log2CPM) in total RNA. (E) Scatter plot of log2-fold change (FC) in 2L/A/WT for total RNA and polysomal RNA. Signiﬁcantly regulated genes determined by a negative binomial rest in EdgeR in red. (F) Scatter plot of log2-transformed counts per million (log2CPM) in polysomal RNA. (G) Polysomal mRNA levels quantiﬁed by RT–PCR. Average and standard deviation for three independent experiments. t test was performed **P-value < 0.005. (H) Western blot analysis of proteins in HCC1806 cells expressing pCDNA5 HA-CMTR1, 2L/A, or vector control (V). Charts represented average protein signal in Western blots and standard deviation for three independent experiments. Quantitation performed in ImageJ software. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 12 of 18 et al, 2014; Gebhardt et al, 2015). Here, we demonstrate that the direct interaction of CMTR1 and DHX15 regulates the catalytic ac- tivity of both enzymes, which impacts gene expression and cell proliferation. et al, 2014; Gebhardt et al, 2015). Here, we demonstrate that the direct interaction of CMTR1 and DHX15 regulates the catalytic ac- tivity of both enzymes, which impacts gene expression and cell proliferation. OB-fold, they adopt a stable open secondary structure (Christian et al, 2014). The co-ordinated impact of DHX15 and CMTR1 on gene expression Ultimately, we want to understand how the relationship between DHX15 and CMTR1 impacts on cell function. This is complex because of the multifunctional nature of both enzymes. The impact of the DHX15–CMTR1 interaction will depend on the relative expression of the enzymes and their other interacting partners, which also inﬂu- ence enzyme activity and localisation. The impact of the DHX15– CMTR1 interaction will also depend on the underlying cell physiology, including relative dependency on O-2 methylation, splicing, rRNA processing, and translation. Of note, CMTR1 is an interferon-regulated gene, whereas DHX15 is not, and therefore, interferon signalling alters the ratio of these factors. In this study, we focus on the biological impact of DHX15 repression of CMTR1 methyltransferase activity and function. We stress that, in certain contexts, the impact of CMTR1 activation of DHX15 helicase activity may be equivalently or more biologically important, given the role of DHX15 in RNA processing. First nucleotide ribose O-2 methylation is associated with trans- lational efﬁciency (Muthukrishnan et al, 1976a; Kuge & Richter, 1995; Kuge et al, 1998). The impact of DHX15 on CMTR1-dependent trans- lation was investigated using the CMTR1 2L/A mutant, which does not bind to DHX15 but is recruited to RNA pol II. Expression of CMTR1 2L/A resulted in increased ribosome loading of a subset of transcripts, including those involved in metabolic pathways and cell cycle control. The genes sensitive to the DHX15–CMTR1 interaction may be partic- ularly dependent on O-2 methylation for polysome loading or require high levels of CMTR1 activity to be O-2 methylated. Expression of CMTR1 2L/A also resulted in increased cell proliferation. In summary, we now recognise that the mRNA capping enzymes, RNGTT, RNMT, and CMTR1 are regulated by different co-factors and posttranslational modiﬁcations, with multiple impacts on gene expression and cell physiology. The challenge now is to understand the complex integration of these regulatory events during devel- opment and in the adult. Discussion G-patch domains can inﬂuence both the stacking of the adenosine base and interactions of N- and C-terminal domains, thus regulating Prp43 catalytic activity (Robert-Paganin et al, 2017; Tauchert et al, 2017). DHX15 constrains CMTR1 function The catalytic domain lies in the centre of the CMTR1 polypeptide and the non-catalytic C- and N-terminal domains inﬂuence methyltransferase activity (Smietanski et al, 2014). Our data are consistent with DHX15 inducing a confor- mational change in CMTR1 to reduce methyltransferase activity. CMTR1 inﬂuences DHX15 function The mechanism by which G-patch proteins regulate the activity and localisation of Prp43 (yeast DHX15 orthologue), are well charac- terised. Prp43 uses the energy generated by ATP hydrolysis to power helicase activity (Walbott et al, 2010). The Prp43 RecA domains interact with the C-terminal domain, rendering the enzyme in a closed conformation. Disruption of these interactions by ATP binding promotes changes in the helicase structure, leading to the open conformation required for unwinding complex RNA stretches (Tauchert et al, 2017). In addition, the stacking of the adenosine base with the RecA1 R and RecA2 F motifs is important for the activity and regulation of the helicase (Robert-Paganin et al, 2017). When intrinsically unstructured G-patch domains bind to the Prp43 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. Mass spectrometry analysis IPs were resolved using SDS–PAGE and stained with Novex Colloidal Blue (Invitrogen). Bands were excised and washed sequentially on a vibrating platform with 0.5 ml water, 1:1 (vol/vol) water and acetonitrile (AcN), 0.1 M ammonium bicarbonate, and 1:1 (vol/vol) 0.1 M ammonium bicarbonate and AcN. Samples were reduced in 10 mM dithiothreitol (20 min, 37°C), alkylated in 50 mM iodoa- cetamide/0.1 M ammonium bicarbonate (20 min, dark), and washed in 50 mM ammonium bicarbonate and 50 mM ammonium bicarbonate/50% AcN. When gel pieces were colourless they were washed with AcN for 15 min and dried. Gel pieces were incubated (16 h, 30°C) in 5 μg/ml trypsin/25 mM triethylammonium bicarbon- ate. Tryptic digests were analysed by liquid chromatography–mass spectrometry on an Applied Biosystems 4000 QTRAP system with precursor ion scanning. The resulting MS/MS data were analysed by Mascot search algorithm (http://www.matrixscience.com). Generation of a CMTR1 knockout cell line (CMTR1 del) using CRISPR/Cas9 CMTR1 gene (ensembl ENSG00000137200) optimal scoring guide target site GGGAGGTTCATCATCGGACG[TGG] was cloned into pBabeD pU6 and sequence-veriﬁed (http://crispr.mit.edu/). HeLa cells stably expressing Tet-regulated Cas9 were transfected with 3 μg pBabeD pU6 CMTR1 using Lipofectamine 2000. Cells were incubated in DMEM, 10% FBS, 2 mM L-glutamine, and 100 μg/ml Normocin (InvivoGene). After 12 h, 4 μg/ml puromycin was added in fresh medium. After 48-h selection, 2 μg/ml doxycycline was added to induce Cas9 expression. After 72 h, single cells were FACS-sorted into a 96-well plate containing DMEM/20% FBS, 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 100 μg/ml Nor- mocin (InvivoGene). CMTR1 protein expression was screened using Western blot in 40 clones. For clones displaying an absence of protein, genomic DNA was isolated and ampliﬁed by PCR, using CMTR1 genomic forward: CTGTGATCCCAGTGGCTGT and CMTR1 ge- nomic reverse: CCAAGGGGCAGTGGACTAT primers. PCR products from control and CMTR1 del clones were sequenced. CMTR1 del clone 5 had no region homology to WT cells around Cas9 targeting sequence and selected for experiments. Bacterial expression and puriﬁcation of DHX15 and CMTR1 proteins Bacterial expression plasmids pET15b-His6-DHX15 WT or E261Q and pGEX6P1-GST-3C-CMTR1 WT or ΔG (residues 143–835) were trans- formed into Rosetta 2 (DE3) cells (Novagen) and plated on LB agar/ 100 μg/ml ampicillin and 35 μg/ml chloramphenicol. Starter cul- tures were incubated at 37°C overnight. After 16 h, cultures were diluted 1/50 into fresh 6-9L LB cultures and incubated at 37°C with agitation. At OD600 ~ 0.3–0.4, the temperature was reduced to 16°C and at OD600 ~ 0.7–0.8 protein expression was induced with 50 μM Isopropyl β-D-1-thiogalactopyranoside for 14–16 h. Cultures were harvested by centrifugation and pellets resuspended in ice-cold re- suspension buffer (Tris 50 mM, pH 8, 200 mM NaCl, 1 mM DTT, 25 mM imidazole, 10% glycerol [vol/vol], and Complete-EDTA free protease inhibitor purchased from Roche). Resuspended cells were ﬂash- frozen in liquid nitrogen, thawed in a lukewarm (25°C) water bath, and treated with lysozyme (1 mg/ml) and DNase I (1:10,000 dilution) for 30 min at 4°C. Cell debris and insoluble material were clariﬁed by centrifugation at 38,000 relative centrifugal force at 4°C for 45 min. Soluble lysates were passed through poly-prep columns packed with either Ni-NTA agarose (His6-DHX15) or GSH-sepharose (GST-3C-CMTR1) pre-equilibrated in wash buffer (50 mM Tris, pH 8, 500 mM NaCl, 1 mM DTT, 25 mM imidazole, and 5% glycerol [vol/vol]). After binding, resins were extensively washed with wash buffer followed by a ﬁnal wash step in low-salt (200 mM NaCl) wash buffer. For His6-DHX15, Ni-NTA bound material was eluted with an elution buffer (50 mM Tris, pH 8, 200 mM NaCl, 1 mM DTT, 5% glycerol, and 300 mM imidazole), whereas GSH-sepharose–bound GST-CMTR1 medium. All cells were cultured with 10% (vol/vol) FBS, 100 U/ml penicillin, 0.1 mg/ml streptomycin, and 2 mM L-glutamine. For plasmid transfection, cells in 10-cm dishes were transfected with 1–5 μg plasmid using 20 μg polyethylenimine (Polysciences). Cells were cultured for 36 h before lysis in 0.45 ml ice-cold lysis buffer (50 mM Tris–HCl, pH 7.5, 1 mM EGTA, 1 mM EDTA, 1 mM sodium orthovanadate, 10 mM β-glycerophosphate, 50 mM NaF, 5 mM sodium pyrophosphate, and 0.27 M sucrose) supplemented with 1% (vol/vol) Triton X-100, 2 mM DTT, 1% (vol/vol) aprotonin, 10 μM leupeptin, and 1 μM pepstatin. Extracts were centrifuged at 16,200 g at 4°C for 15 min and supernatant retained. For siRNA transfections, cells were transfected with 50 nM siRNA (Dharmacon siGenome range; non-targeting or CMTR1), for 48 h unless stated otherwise. UAGAUGAUGUUCGGGAUUA, 01 CMTR1 siRNA; GUAAGAGCGUGUUU- GAUGU, 02 CMTR1 siRNA. For plasmid and siRNA co-transfection, lipofectamine 2000 (Thermo Fisher Scientiﬁc) was used. 2–6 × 104 cells were plated in a six-well plate, 0.5 μg pcDNA5 GFP, and 0.5 μg pcDNA5 HA-CMTR1 WT or 2L/A mixed with siRNA before transfection. Countess cell counter was used for cell counting. were eluted in SDS-sample buffer. Western blots were performed by standard protocols. CMTR1 antibody was raised in sheep by Orygen Antibodies Limited, UK and afﬁnity puriﬁed against the human recombinant protein. Second bleed was used at 1 μg/ml for Western blotting and the ﬁrst bleed was used at 1 μg/IP. Other antibodies: RNGTT (in-house), HA (Sigma), DHX15 (ab70454; Abcam), GAPDH (Cell Signaling Technologies), RNA pol II (N20; Santa Cruz), PSer5-PolII (3E8; ChromoTek), and pSer2-PolII (3E10; ChromoTek). Secondary antibodies were from Pierce. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. Cell culture HeLa, HEK293, MCF7, and U2OS cells were cultured in DMEM and HCC1806 cells were cultured in Roswell Park Memorial Institute 1640 https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 13 of 18 Unwinding assay Unwinding assay was performed as in Tanaka et al (2007). The 126-nucleotide (nt) RNA strand (59-GGGCGAAUUGGGCCCGACGUCGC- AUGCUCCCGGCCGCCAUGGCGGCCGCGGGAAUUCGAUUAUCACUAGU- GAAUUCGCGGCCGCCUGCAGGUCGACCAUAUGGGAGAGCUCCCAACGC- GUUGGAUG-39) was synthesized by in vitro transcription, and annealed at a 1:3 M ratio to 59 32P-labelled DNA (59-GACGTCGGGCCCAATTCGCCC-39) to yield 59-tailed 30-bp duplex. The annealed substrate was gel- puriﬁed on native 6% PAGE. 10 μl reaction mixtures containing indicated concentrations of His6-DHX15 and CMTR1, and 2.5 nM RNA/DNA substrate were incubated in 40 mM Tris–HCl, pH 7.0, 2 mM DTT, and 1 mM ATP-Mg2+ at 37°C for 45 min. Reactions were halted by transfer to ice and addition of 5 μl loading buffer (100 mM Tris–HCl, pH 7.4, 5 mM EDTA, 0.5% SDS, 50% glycerol, 0.1% [wt/vol] bromophenol blue, and xylene cyanol dyes). Samples were re- solved on 16% polyacrylamide gel in 40 mM Tris–borate, 0.5 mM EDTA, and 0.1% SDS. 32P-labeled substrate and products were vi- sualized by autoradiography. times with a binding buffer and bound fraction eluted with 30 μl SDS–PAGE loading buffer at 95°C for 5 min. was treated with GST-3C protease overnight at 4°C to cleave GST tag. Eluted His6-DHX15 and CMTR1 proteins were concentrated using an Amicon 30 kD- molecular weight cut off concentrator and subjected to size-exclusion chromatography (Superdex 200) in 50 mM Tris, pH 8, 200 mM NaCl, 1 mM DTT, and 5% glycerol (vol/vol) buffer on an AKTA puriﬁer FPLC system (Amersham). Proteins were further concentrated and ﬂash-frozen as single use aliquots, stored at −80°C. was treated with GST-3C protease overnight at 4°C to cleave GST tag. Eluted His6-DHX15 and CMTR1 proteins were concentrated using an Amicon 30 kD- molecular weight cut off concentrator and subjected to size-exclusion chromatography (Superdex 200) in 50 mM Tris, pH 8, 200 mM NaCl, 1 mM DTT, and 5% glycerol (vol/vol) buffer on an AKTA puriﬁer FPLC system (Amersham). Proteins were further concentrated and ﬂash-frozen as single use aliquots, stored at −80°C. First nucleotide O-2 methylation assay A guanosine-capped substrate 32P-labelled on α-phosphate (GpppG-RNA) was produced. 200 ng 55-base in vitro–transcribed RNA was incubated with 100 ng RNA guanylyltransferase and 59-phosphatase (RNGTT), 2 μl (10 μCi) [α32P]GTP, and 1 μl RNAsin (Promega) in 10 μl reaction buffer (0.05 M Tris–HCl, pH 8.0, 6 mM KCl, and 1.25 mM MgCl2) at 37°C for 60 min. When used, a guanosine- capped transcript was incubated with 100 ng RNMT and 0.2 μM SAM for 20 min at 30°C to produce N-7 methylated guanosine cap (m7GpppG-RNA). RNA was puriﬁed by ammonium acetate pre- cipitation. Methyltransferase assay performed in 10 μl reaction buffer, with 3 nM CMTR1 (or indicated nM), 2 ng 32P-labeled GpppG- RNA or m7GpppG-RNA, and 10 μM SAM at 30°C for 30 min (unless otherwise stated). Following the reaction, RNA was precipitated and resuspended in 4 μl 50-mM sodium acetate and 0.25 U P1 nuclease for 30 min at 37°C. Cap structures were resolved in 0.4 M ammonium sulphate on polyethyleneimine–cellulose plates, and visualized and quantiﬁed by autoradiography. In Fig 3E, when cell extracts were used in the assay, cells were lysed in a lysis buffer (as above) and 5 μg cell extract was used in the methyltransferase reaction. ATPase activity assay ATPase reactions performed as in Lebaron et al (2009). 100 nM His6- DHX15, 0.6 μCi/μl [α32P] ATP, and 100–200 μM ATP were incubated in 5 μl buffer (25 mM Tris–acetate, pH 8.0, 2.5 mM Mg(CH3-COO)2, 100 mM KCl, 0.2 mM DTT, 100 μg/ml BSA [Sigma]) at 30°C for the time indicated. 1 μl reaction mix was resolved on polyethyleneimin– cellulose plates using 0.75 M KH2PO4. Spots were visualised and quantiﬁed on a phosphorimager. When indicated, 1 μg HeLa cell RNA was added to the reaction. Gel ﬁltration 0.5 ml of 2 mg/ml cell extract was 0.22-μm ﬁltered and resolved on Superdex 200 10/300 GL preparative grade column (GE Healthcare) in 50 mM Tris–HCl, pH 7.4, 1 mM EDTA, 0.1 M sodium chloride, 0.03% Brij-35, and 2 mM DTT at 0.4 ml/min ﬂow. 0.5 ml fractions were collected. Molecular weight markers (Bio-Rad) were used: bovine thyroglobulin (670 kD), bovine gamma globulin (158 kD), and chicken ovalbumin (44 kD). Cap-binding assays Cap-binding assays were performed as in Rio (2014) with minor modiﬁcations. Reactions were performed with 32P-labeled GpppG- RNA as in O-2 methylation assays with indicated amount of en- zymes. After 20 min incubation, reactions were stopped with 5 μl of loading buffer (10 mM Tris, pH 7.5, 60 mM KCl, 5% glycerol, 0.01% xylene cyanol, and bromophenol blue), loaded, and resolved in 4%–12% tris borate EDTA buffer non-denaturing acrylamide gel (Invitrogen) for 75 min at constant 50 V and 4°C. Gel were ﬁxed (10% acetic acid and 10% methanol) for 5 min, washed in 10% glycerol for 10 min, and detected using a phosphorimager. Molecular biology cDNA cloning and mutagenesis were performed by standard pro- tocols. Constructs were sequence-veriﬁed. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. IP and Western blot For IP of GFP or HA-tagged proteins, anti–GFP antibody–conjugated agarose (ChromoTek) and anti-HA antibody–conjugated agarose (Sigma) were used. For endogenous proteins, 1 μg relevant antibody was pre-incubated with 10 μl protein-G Sepharose packed beads and washed to remove non-bound antibody. 0.5–2.5 mg lysates were precleared with 10–30 μl protein-G Sepharose (GE Healthcare), and then incubated with an antibody–resin conjugate for 2 h at 4°C under gentle agitation. IPs were washed three times with lysis buffer containing 0.1 M NaCl. Typically, 1%–2% input and 30% of IP eluates were loaded for Western blot, except in Figs 1F and 4A–C in which 15% of eluate was loaded. For RNAse treatment, 40 μg RNAse A was incubated with IP or 2 μg HeLa RNA for 60 min at 4°C. Proteins https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 14 of 18 Supplementary Information is available at https://doi.org/10.26508/lsa. 201800092. Supplementary Information is available at https://doi.org/10.26508/lsa. 201800092. Acknowledgements Cells were incubated in 100 μg/ml cycloheximide for 10 min and washed in ice-cold PBS supplemented with 100 μg/ml cyclohexi- mide, and extracts were collected in polysome lysis buffer (15 mM Tris, pH 7.5, 15 mM MgCl2, 0.3 M NaCl, 1 mM DTT, 1% Triton X-100, 100 mg/ml cycloheximide, and 100 U/ml RNasin). 10% extracts were retained as input and 90% resolved by centrifugation through 10 ml 10%–50% sucrose gradient at 18,000 g for 2 h at 4°C. Fractions were collected on a FoxyR1 fractionator (Teledyne ISCO) with OD259 nm monitoring. We thank the Cowling laboratory for assistance and Janusz Bujnicki labo- ratory, Frances Fuller-Pace, and Sara Ten-Have for discussions. Research was funded by Medical Research Council Senior Fellowship (VH Cowling) MR/ K024213/1, Lister Institute Prize Fellowship (VH Cowling), Wellcome Trust Technology Platform award 097945/B/11/Z, and Medical Research Council, UK, Next Generation Optical Microscopy award MR/K015869/1. RNA-sequencing and analysis F Inesta-Vaquera: conceptualization, data curation, formal analysis, validation, investigation, methodology, and writing—original draft, review, and editing. Extraction of polysomal RNA was performed as described pre- viously (Grasso et al, 2016). Brieﬂy, polysomal fractions 16–20 were puriﬁed using Phenol:Chloroform:Isoamyl alcohol (25:24:1). RNA was precipitated overnight with 2 M of LiCl, 20 mM Tris, pH 7.5, and EDTA 10 mM. Input RNA was puriﬁed using Trizol Reagent. RNA was se- quenced at the Tayside Centre for Genomic Analysis. RNA was quality checked using TapeStation (Agilent Technologies). RNAseq libraries were prepared with TruSeq Stranded Total RNA with Ribo- Zero-Gold kit (Illumina). Sequencing was performed using NextSeq Series High Output kit 2 × 75 bp (Illumina). Reads were quality controlled using FastQC, then mapped to (GRCh38/hg25) assembly of human genome and reads per gene quantiﬁed using STAR 2.5.2b (Dobin et al, 2013). Differentially expressed genes were identiﬁed using edgeR package (Robinson et al, 2010). Genes with at least 1 count per million in all samples were analysed for differential expression. Pairwise comparisons were made between input RNA and between polysomal RNA for samples transfected with pcDNA5, pcDNA 5 HA-CMTR1 WT or 2L/A. Plots comparing differential input or polysome mRNA abundance were drawn using the ggplot2 R package. V Chaugule: investigation and methodology. A Galloway: data curation and formal analysis. L Chandler: investigation. A Rojas-Fernandez: investigation. S Weidlich: investigation. M Peggie: investigation. VH Cowling: conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, method- ology, writing—original draft, project administration, and wri- ting—review, and editing. Immunoﬂuorescence RNA pol II CTD peptide chromatography was performed as in Ho & Shuman (1999). 0.5 nmol biotinylated CTD peptides were absorbed on 0.2 mg Streptavidin Dynabeads M-280 (Invitrogen) in 300 μl binding buffer (25 mM Tris–HCl, pH 8, 50 mM NaCl, 1 mM DTT, 5% Glycerol, and 0.03% Triton X-100). After 45 min at 4°C, beads were magnet-concentrated and washed in binding buffer. 0.2 μg in- dicated protein was mixed with peptide-beads in 50 μl binding buffer. After incubation for 45 min at 4°C, beads were washed three Immunoﬂuorescence was performed in 2% BSA/PBS at room temperature. Cells were ﬁxed in 4% paraformaldehyde for 10 min, permeabilized with 1% NP-40/PBS for 3 min, blocked with 10% donkey serum for 20 min, and incubated in 1.4 μg/ml polyclonal sheep anti-CMTR1 or DHX15 antibodies for 1.5 h, washed, and incubated with 1.4 μg/ml Alexa Fluor 594– or 488–conjugated donkey anti-sheep or rabbit antibodies for https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 15 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 15 of 18 45 min. Cells were counterstained with 1 μg/ml DAPI, mounted in 2.5% 1, 4-diazobicyclo-(2,2,2-octane), and visualized by ﬂuo- rescence microscopy (LSM 700; Zeiss). RT–PCR Banerjee D, McDaniel PM, Rymond BC (2015) Limited portability of G-patch domains in regulators of the Prp43 RNA helicase required for pre- mRNA splicing and ribosomal RNA maturation in Saccharomyces cerevisiae. Genetics 200: 135–147. doi:10.1534/genetics.115.176461 RNA was extracted using Trizol Reagent (Invitrogen). cDNA was synthesised using iScript cDNA Synthesis Kit (Bio-Rad). RT–PCR was performed using Quanta Biosciences SYBR Green. Primers: PRKDC_Fwd GAGAAGGCGGCTTACCTGAG, PRKDC_Rvr CGAAGGCCCGC- TTTAAGAGA, IGF2R_Fwd AGCGAGAGCCAAGTGAACTC, IGF2R_Rvr TCG- CTGTAAGCAGCTGTGAA, CAD_Fwd AGGTTTGCCAGCTGAGGAG, CAD_Rvr TAATGAGTGCAGCAGGGGTG, ATR_Fwd GGAGGAGTTTTGGCCTCCAC, A- TR_Rvr TGTGGCACTGCCCAGCTC, GCN1_Fwd CTTGTGCCCAAGCTGACAAC, GCN1_Rvr GCCCTGTGTCATCCTCTACG, MTOR_Fwd GAAGCCGCGCGAACCT, MTOR_Rvr CTGGTTTCCTCATTCCGGCT, CMTR1_Fwd CATTGCCCCATTTCA- CATTTGC, CMTR1_Rvr TCTTAGGCCCTGTGCATCTG. RNA was extracted using Trizol Reagent (Invitrogen). cDNA was synthesised using iScript cDNA Synthesis Kit (Bio-Rad). RT–PCR was performed using Quanta Biosciences SYBR Green. Primers: PRKDC_Fwd GAGAAGGCGGCTTACCTGAG, PRKDC_Rvr CGAAGGCCCGC- TTTAAGAGA, IGF2R_Fwd AGCGAGAGCCAAGTGAACTC, IGF2R_Rvr TCG- CTGTAAGCAGCTGTGAA, CAD_Fwd AGGTTTGCCAGCTGAGGAG, CAD_Rvr TAATGAGTGCAGCAGGGGTG, ATR_Fwd GGAGGAGTTTTGGCCTCCAC, A- TR_Rvr TGTGGCACTGCCCAGCTC, GCN1_Fwd CTTGTGCCCAAGCTGACAAC, GCN1_Rvr GCCCTGTGTCATCCTCTACG, MTOR_Fwd GAAGCCGCGCGAACCT, MTOR_Rvr CTGGTTTCCTCATTCCGGCT, CMTR1_Fwd CATTGCCCCATTTCA- CATTTGC, CMTR1_Rvr TCTTAGGCCCTGTGCATCTG. Belanger F, Stepinski J, Darzynkiewicz E, Pelletier J (2010) Characterization of hMTr1, a human Cap1 29-O-ribose methyltransferase. J Biol Chem 285: 33037–33044. doi:10.1074/jbc.m110.155283 Buratowski S (2009) Progression through the RNA polymerase II CTD cycle. Mol Cell 36: 541–546. doi:10.1016/j.molcel.2009.10.019 Caldwell DC, Emerson CP (1985) The role of cap methylation in the translational activation of stored maternal histone messenger-RNA in sea-urchin embryos. Cell 42: 691–700. doi:10.1016/0092-8674(85) 90126-6 Chen YL, Capeyrou R, Humbert O, Mouffok S, Kadri YA, Lebaron S, Henras AK, Henry Y (2014) The telomerase inhibitor Gno1p/PINX1 activates the helicase Prp43p during ribosome biogenesis. Nucleic Acids Res 42: 7330–7345. doi:10.1093/nar/gku357 References Aravind L, Koonin EV (1999) G-patch: A new conserved domain in eukaryotic RNA-processing proteins and type D retroviral polyproteins. Trends Biochem Sci 24: 342–344. doi:10.1016/s0968-0004(99)01437-1 Supplementary Information Supplementary Information is available at https://doi.org/10.26508/lsa. 201800092. Conﬂict of Interest Statement The authors declare that they have no conﬂict of interest. doi:10.1016/j.sbi.2016.02.001 Lin ML, Fukukawa C, Park JH, Naito K, Kijima K, Shimo A, Ajiro M, Nishidate T, Nakamura Y, Katagiri T (2009) Involvement of G-patch domain containing 2 overexpression in breast carcinogenesis. Cancer Sci 100: 1443–1450. doi:10.1111/j.1349-7006.2009.01185.x Grasso L, Suska O, Davidson L, Gonatopoulos-Pournatzis T, Williamson R, Wasmus L, Wiedlich S, Peggie M, Stavridis MP, Cowling VH (2016) mRNA cap methylation in pluripotency and differentiation. Cell Rep 16: 1352–1365. doi:10.1016/j.celrep.2016.06.089 Memet I, Doebele C, Sloan KE, Bohnsack MT (2017) The G-patch protein NF- kappaB-repressing factor mediates the recruitment of the exonuclease XRN2 and activation of the RNA helicase DHX15 in human ribosome biogenesis. Nucleic Acids Res. doi:10.1093/nar/gkx013 Guerra S, Lopez-Fernandez LA, Pascual-Montano A, Munoz M, Harshman K, Esteban M (2003) Cellular gene expression survey of vaccinia virus infection of human HeLa cells. J Virol 77: 6493–6506. doi:10.1128/ jvi.77.11.6493-6506.2003 Muthukrishnan S, Furuichi Y, Both GW, Shatkin AJ (1976a) Effect of 5’-terminal structures on the binding of ribopolymers to eukaryotic ribosomes. Progress Nucleic Acid Res Mol Biol 19: 473–476. Haline-Vaz T, Silva TC, Zanchin NI (2008) The human interferon-regulated ISG95 protein interacts with RNA polymerase II and shows methyltransferase activity. Biochem Biophys Res Commun 372: 719–724. doi:10.1016/j.bbrc.2008.05.137 Muthukrishnan S, Morgan M, Banerjee AK, Shatkin AJ (1976b) Inﬂuence of 5’- terminal m7G and 29-O-methylated residues on messenger ribonucleic acid binding to ribosomes. Biochemistry 15: 5761–5768. doi:10.1021/bi00671a012 Heininger AU, Hackert P, Andreou AZ, Boon KL, Memet I, Prior M, Clancy A, Schmidt B, Urlaub H, Schleiff E, et al (2016) Protein cofactor competition regulates the action of a multifunctional RNA helicase in different pathways. RNA Biol 13: 320–330. doi:10.1080/ 15476286.2016.1142038 Nagaraj N, Wisniewski JR, Geiger T, Cox J, Kircher M, Kelso J, Paabo S, Mann M (2011) Deep proteome and transcriptome mapping of a human cancer cell line. Mol Syst Biol 7: 548. doi:10.1038/msb.2011.81 Ho CK, Shuman S (1999) Distinct roles for CTD Ser-2 and Ser-5 phosphorylation in the recruitment and allosteric activation of mammalian mRNA capping enzyme. Mol Cell 3: 405–411. doi:10.1016/ s1097-2765(00)80468-2 Niu Z, Jin W, Zhang L, Li X (2012) Tumor suppressor RBM5 directly interacts with the DExD/H-box protein DHX15 and stimulates its helicase activity. FEBS Lett 586: 977–983. doi:10.1016/j.febslet.2012.02.052 Pawellek A, McElroy S, Samatov T, Mitchell L, Woodland A, Ryder U, Gray D, Luhrmann R, Lamond AI (2014) Identiﬁcation of small molecule inhibitors of pre-mRNA splicing. J Biol Chem 289: 34683–34698. Kuge H, Brownlee GG, Gershon PD, Richter JD (1998) Cap ribose methylation of c-mos mRNA stimulates translation and oocyte maturation in Xenopus laevis. Nucleic Acids Res 26: 3208–3214. doi:10.1093/nar/ 26.13.3208 Christian H, Hofele RV, Urlaub H, Ficner R (2014) Insights into the activation of the helicase Prp43 by biochemical studies and structural mass spectrometry. Nucleic Acids Res 42: 1162–1179. doi:10.1093/nar/ gkt985 Kuge H, Richter JD (1995) Cytoplasmic 3’ poly(a) addition induces 5’ cap ribose methylation: Implications for translational control of maternal mRNA. EMBO J 14: 6301–6310. Dafﬁs S, Szretter KJ, Schriewer J, Li JQ, Youn S, Errett J, Lin TY, Schneller S, Zust R, Dong HP, et al (2010) 29-O methylation of the viral mRNA cap evades host restriction by IFIT family members. Nature 468: 452–456. doi:10.1038/nature09489 Langberg SR, Moss B (1981) Post-transcriptional modiﬁcations of messenger- RNA. Puriﬁcation and characterization of cap-I and cap-Ii RNA (Nucleoside-29-)-methyltransferases from HeLa-cells. J Biol Chem 256: 54–60. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR (2013) STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29: 15–21. doi:10.1093/bioinformatics/bts635 Furuichi Y (2015) Discovery of m(7)G-cap in eukaryotic mRNAs. Proc Jpn Acad Ser B Phys Biol Sci 91: 394–409. doi:10.2183/pjab.91.394 Lange A, Mills RE, Lange CJ, Stewart M, Devine SE, Corbett AH (2007) Classical nuclear localization signals: Deﬁnition, function, and interaction with importin alpha. J Biol Chem 282: 5101–5105. doi:10.1074/jbc.r600026200 Gebhardt A, Habjan M, Benda C, Meiler A, Haas DA, Hein MY, Mann A, Mann M, Habermann B, Pichlmair A (2015) mRNA export through an additional cap-binding complex consisting of NCBP1 and NCBP3. Nat Commun 6: 8192. doi:10.1038/ncomms9192 Lebaron S, Papin C, Capeyrou R, Chen YL, Froment C, Monsarrat B, Caizergues- Ferrer M, Grigoriev M, Henry Y (2009) The ATPase and helicase activities of Prp43p are stimulated by the G-patch protein Pfa1p during yeast ribosome biogenesis. EMBO J 28: 3808–3819. doi:10.1038/ emboj.2009.335 Geiss GK, Carter VS, He Y, Kwieciszewski BK, Holzman T, Korth MJ, Lazaro CA, Fausto N, Bumgarner RE, Katze MG (2003) Gene expression proﬁling of the cellular transcriptional network regulated by alpha/beta interferon and its partial attenuation by the hepatitis C virus nonstructural 5A protein. J Virol 77: 6367–6375. doi:10.1128/ jvi.77.11.6367-6375.2003 Leung DW, Amarasinghe GK (2016) When your cap matters: Structural insights into self vs non-self recognition of 5’ RNA by immunomodulatory host proteins. Curr Opin Struct Biol 36: 133–141. RNAseq data RNAseq data can be found in the GEO database record GSE113573. https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 16 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 16 of 18 Robinson MD, McCarthy DJ, Smyth GK (2010) edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26: 139–140. doi:10.1093/ bioinformatics/btp616 Tauchert MJ, Fourmann JB, Luhrmann R, Ficner R (2017) Structural insights into the mechanism of the DEAH-box RNA helicase Prp43. eLife 6: e21510. doi:10.7554/eLife.21510 Topisirovic I, Svitkin YV, Sonenberg N, Shatkin AJ (2011) Cap and cap-binding proteins in the control of gene expression. Wiley Interdiscip Rev RNA 2: 277–298. doi:10.1002/wrna.52 Schuberth-Wagner C, Ludwig J, Bruder AK, Herzner AM, Zillinger T, Goldeck M, Schmidt T, Schmid-Burgk JL, Kerber R, Wolter S, et al (2015) A conserved histidine in the RNA sensor RIG-I controls immune tolerance to N1-29O-methylated self RNA. Immunity 43: 41–51. doi:10.1016/j.immuni.2015.06.015 Varshney D, Lombardi O, Schweikert G, Dunn S, Suska O, Cowling VH (2018) mRNA cap methyltransferase, RNMT-RAM, promotes RNA pol II- dependent transcription. Cell Rep 23: 1530–1542. doi:10.1016/j. celrep.2018.04.004 Shuman S (2002) What messenger RNA capping tells us about eukaryotic evolution. Nat Rev Mol Cell Biol 3: 619–625. doi:10.1038/ nrm880 Walbott H, Mouffok S, Capeyrou R, Lebaron S, Humbert O, van Tilbeurgh H, Henry Y, Leulliot N (2010) Prp43p contains a processive helicase structural architecture with a speciﬁc regulatory domain. EMBO J 29: 2194–2204. doi:10.1038/emboj.2010.102 Smietanski M, Werner M, Purta E, Kaminska KH, Stepinski J, Darzynkiewicz E, Nowotny M, Bujnicki JM (2014) Structural analysis of human 2’-O-ribose methyltransferases involved in mRNA cap structure formation. Nat Commun 5: 3004. doi:10.1038/ ncomms4004 Werner M, Purta E, Kaminska KH, Cymerman IA, Campbell DA, Mittra B, Zamudio JR, Sturm NR, Jaworski J, Bujnicki JM (2011) 2’-O-ribose methylation of cap2 in human: Function and evolution in a horizontally mobile family. Nucleic Acids Res 39: 4756–4768. doi:10.1093/nar/gkr038 Su AI, Pezacki JP, Wodicka L, Brideau AD, Supekova L, Thimme R, Wieland S, Bukh J, Purcell RH, Schultz PG, et al (2002) Genomic analysis of the host response to hepatitis C virus infection. Proc Natl Acad Sci U S A 99: 15669–15674. doi:10.1073/pnas.202608199 Yoshimoto R, Okawa K, Yoshida M, Ohno M, Kataoka N (2014) Identiﬁcation of a novel component C2ORF3 in the lariat-intron complex: Lack of C2ORF3 interferes with pre-mRNA splicing via intron turnover pathway. Genes Cells 19: 78–87. doi:10.1111/gtc.12114 Tanaka N, Aronova A, Schwer B (2007) Ntr1 activates the Prp43 helicase to trigger release of lariat-intron from the spliceosome. Genes Dev 21: 2312–2325. doi:10.1074/jbc.m114.590976 Imamura O, Sugawara M, Furuichi Y (1997) Cloning and characterization of a putative human RNA helicase gene of the DEAH-box protein family. Biochem Biophys Res Commun 240: 335–340. doi:10.1006/ bbrc.1997.7585 Picard-Jean F, Brand C, Tremblay-Letourneau M, Allaire A, Beaudoin MC, Boudreault S, Duval C, Rainville-Sirois J, Robert F, Pelletier J, et al (2018) 2’-O-methylation of the mRNA cap protects RNAs from decapping and degradation by DXO. PLoS One 13: e0193804. doi:10.1371/journal.pone.0193804 Inesta-Vaquera F, Cowling VH (2017) Regulation and function of CMTR1- dependent mRNA cap methylation. WIRE RNA Accepted. doi:10.1002/ wrna.1450 Jankowsky E (2011) RNA helicases at work: Binding and rearranging. Trends Biochem Sci 36: 19–29. doi:10.1016/j.tibs.2010.07.008 RamanathanA,RobbGB, ChanSH (2016)mRNA capping: Biologicalfunctionsand applications. Nucleic Acids Res 44: 7511–7526. doi:10.1093/nar/gkw551 Kato A, Homma T, Batchelor J, Hashimoto N, Imai S, Wakiguchi H, Saito H, Matsumoto K (2003) Interferon-alpha/beta receptor-mediated selective induction of a gene cluster by CpG oligodeoxynucleotide 2006. BMC Immunol 4: 8. doi:10.1186/1471-2172-4-8 Rio DC (2014) Electrophoretic mobility shift assays for RNA-protein complexes. Cold Spring Harbor Protoc 2014: 435–440. doi:10.1101/pdb.prot080721 Koodathingal P, Staley JP (2013) Splicing ﬁdelity: DEAD/H-box ATPases as molecular clocks. RNA Biol 10: 1073–1079. doi:10.4161/rna.25245 Robert-Paganin J, Halladjian M, Blaud M, Lebaron S, Delbos L, Chardon F, Capeyrou R, Humbert O, Henry Y, Henras AK, et al (2017) Functional link between DEAH/RHA helicase Prp43 activation and ATP base binding. Nucleic Acids Res 45: 1539–1552. doi:10.1093/nar/gkw1233 Kruse S, Zhong S, Bodi Z, Button J, Alcocer MJ, Hayes CJ, Fray R (2011) A novel synthesis and detection method for cap-associated adenosine modiﬁcations in mouse mRNA. Scientiﬁc Rep 1: 126. doi:10.1038/ srep00126 Robert-Paganin J, Rety S, Leulliot N (2015) Regulation of DEAH/RHA helicases by G-patch proteins. Biomed Res Int 2015: 931857. doi:10.1155/2015/931857 Robert-Paganin J, Rety S, Leulliot N (2015) Regulation of DEAH/RHA helicases by G-patch proteins. Biomed Res Int 2015: 931857. doi:10.1155/2015/931857 https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 17 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. 17 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al. doi:10.1101/gad.1580507 Tannukit S, Crabb TL, Hertel KJ, Wen X, Jans DA, Paine ML (2009) Identiﬁcation of a novel nuclear localization signal and speckle-targeting sequence of tuftelin-interacting protein 11, a splicing factor involved in spliceosome disassembly. Biochem Biophys Res Commun 390: 1044–1050. doi:10.1016/j.bbrc.2009.10.111 License: This article is available under a Creative Commons License (Attribution 4.0 International, as described at https://creativecommons.org/ licenses/by/4.0/). License: This article is available under a Creative Commons License (Attribution 4.0 International, as described at https://creativecommons.org/ licenses/by/4.0/). https://doi.org/10.26508/lsa.201800092 vol 1 \| no 3 \| e201800092 18 of 18 DHX15 and CMTR1 co-regulation Inesta-Vaquera et al.
https://openalex.org/W2979812977	https://pubs.rsc.org/en/content/articlepdf/2019/sm/c9sm01802f	English	null	Microgels as viscosity modifiers influence lubrication performance of continuum	Soft matter	2,019	cc-by	11,504	Volume 15 Number 47 21 December 2019 Pages 9599–9818 Volume 15 Number 47 21 December 2019 Pages 9599–9818 Volume 15 Number 47 21 December 2019 Pages 9599–9818 a Food Colloids and Bioprocessing Group, School of Food Science and Nutrition, University of Leeds, UK. E-mail: A.Sarkar@leeds.ac.uk b Molecular and Nanoscale Physics Group, School of Physics and Astronomy, University of Leeds, UK † Electronic supplementary information (ESI) available. See DOI: 10.1039/c9sm01802f Microgels as viscosity modifiers influence lubrication performance of continuum† Cite this: Soft Matter, 2019, 15, 9614 Efren Andablo-Reyes,a Demetra Yerani,a Ming Fu,a Evangelos Liamas, a Simon Connell,b Ophelie Torresa and Anwesha Sarkar a Biocompatible microgels have been demonstrated to act as excellent lubricants, however, the influence of the continuum on their overall mechanical performance has been neglected so far. In this work, the mechanical performance of colloidal whey protein microgels (hydrodynamic diameter B100 nm measured using dynamic light scattering and atomic force microscopy) of diﬀerent rigidity dispersed in Newtonian (buﬀer and corn syrup) or complex non-Newtonian fluids (xanthan gum) is investigated for the first time via rheology and soft tribology. Dispersions of both soft microgels (G0 B 100.0 Pa) and hard microgels (G0 B 10.0 kPa) were observed to act as thickeners in buﬀer as well as in low viscosity corn syrup and correspondingly reduced the friction, latter decreased as a function of the increased rigidity of the microgels. Diﬀerently, in high viscosity continuum, the microgels acted as thinning agents and increased the friction. In the lubrication limit, microgels in buﬀer or corn syrup behaved as Newtonian fluids with eﬀective viscosity corresponding to their second Newtonian plateau value (ZN). However, the lubrication performance of the microgels dispersed in the complex fluid (xanthan gum) could not be described quantitatively by ZN. For the low viscosity xanthan gum, the microgels had no influence on friction. Nevertheless, for the high viscosity counterparts, the soft microgels acted as thinning agents whilst the hard microgels accelerated the onset of elastohydrodynamic regime. This study demonstrates that microgels act as viscosity modifiers directly influencing the tribological performance, depending upon a subtle interplay of rheological properties of the particles and continuum. Received 6th September 2019, Accepted 29th September 2019 DOI: 10.1039/c9sm01802f rsc.li/soft-matter-journal Soft Matter ISSN 1744-6848 PAPER Anwesha Sarkar et al . Microgels as viscosity modifiers influence lubrication performance of continuum rsc.li/soft-matter-journal rsc.li/soft-matter-journal PAPER Anwesha Sarkar et al . Microgels as viscosity modifiers influence lubrication performance of continuum This journal is ©The Royal Society of Chemistry 2019 Soft Matter Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Microgels as viscosity modifiers influence lubrication performance of continuum† Efren Andablo-Reyes,a Demetra Yerani,a Ming Fu,a Evangelos Liamas, a Simon Connell,b Ophelie Torresa and Anwesha Sarkar a 2.2 Methods Preparation of whey protein gels. Whey protein solution was prepared by dissolving exact quantities of whey protein isolate (10.0 wt% or 15.0 wt%) in 20 mM phosphate buﬀer at pH 7.0 and stirred for 2 hours at room temperature to ensure complete dissolution. The disulphide crosslinking of whey protein was achieved by heating the whey protein solutions at 90 1C for 30 minutes. Thermally-crosslinked whey protein gels were allowed to cool down at room temperature before placing them at 4 1C for 12 hours. Although both concentrations of protein produce soft deformable gels, in this study, 10.0 wt% is referred as soft gels whilst 15.0 wt% is referred as hard gel. Preparation of whey protein microgel (WPM) dispersions in Newtonian continuum. Phosphate buﬀer or two diﬀerent concentrations of corn syrup solutions, 50.0 wt% and 75.0 wt%, were used as the Newtonian continuum for the preparation of microgel dispersions. Solutions were prepared by mixing corn syrup with phosphate buﬀer at room temperature and stirred at 40 1C for an hour until a homogeneous solution was obtained. The previously crosslinked whey protein gels (10.0 wt% or 15.0 wt%) were mixed with either phosphate buﬀer or corn syrup solutions at a 1 : 1 w/w ratio and subsequently sheared with a hand blender (HB711M, Kenwood, UK) for 1 minute. The whey protein gel dispersed in either the phosphate buﬀer or the corn syrup was then placed under vacuum for 30 minutes in order to remove the excess air introduced during blending. The system was finally homogenized via the Leeds Jet Homogenizer (University of Leeds, UK), with two passes at 300 20 bars.14,27 In this work, colloidal whey protein particles are investi- gated as models of soft microgels dispersed in continuums of diﬀerent rheological behaviours. Whey protein microgels have been recently regarded as important biocompatible lubricants.14,15 Two diﬀerent types of well-characterised fluids are used as continuum in this study. On the one hand, corn syrup solutions with diﬀerent viscosities are used as Newtonian continuum to serve as a comparison with various literature. On the other hand, xanthan gum solutions are investigated as examples of complex non-Newtonian conti- nuum. 2.1 Materials Whey protein isolate (WPI) powder containing 96.3 wt% protein was kindly donated by Fonterra Limited (Auckland, New-Zealand). Xanthan gum was purchased from Sigma Aldrich, Dorset, UK and used as received. Corn syrup was purchased from Special Ingredients Ltd (Chesterfield, UK). Monosodium phosphate (NaH2PO4) and disodium phosphate (Na2HPO4) were purchased from VWR Chemicals, UK. Milli-Q water (resistivity of 18 M O cm by Milli-Q apparatus, Millipore Corp., Bedford, MA, USA) was used for preparation of 20 mM phosphate buﬀer at pH 7.0. 2. Experimental 2.1 Materials An important aspect which has been as-of-yet neglected in the literature, whilst possibly having an exceptional relevance in the mechanical performance of the colloidal dispersions, is the role of the continuous phase i.e., the continuum in which the microgels are dispersed.19 Farres et al. carried out an elegant study where the role of co-solutes, such as glucose and glycerol on the lubrication properties of agar fluid gels was investigated.20 The authors proposed that altering both the rate of ordering transition and the degree of solvation of the gel particles by the co-solutes present in the continuum could influence the lubricating behaviour of fluid gels. Nevertheless, many, if not most studies conducted on microgel or fluid gel lubrication have focused on simple aqueous dispersions. Complex fluids are, in general terms, a combination of a variety of microscopic components with diﬀerent microstructures, such as colloidal particles and polymers. The self and collective dynamics and configuration of these components contribute in diﬀerent manner to the macroscopic behaviour of the material.21 For instance, the response of complex fluids to mechanical deformation is dominated by a wide spectrum of relaxation times associated with diﬀerent levels of component level reorganisation, such as chain segments decorrelation22 or colloidal mesostructures breakage under shear.23 The under- standing of the mechanical behaviour of microgel dispersions in complex continuum might also be relevant as simple aqueous media are often altered with the addition of rheology modifiers in real-life applications. Particularly such funda- mental knowledge will open up possibilities for templating new soft microgel + complex continuum combinations with optimized performance, such as shear-thinning with ultra- low friction coeﬃcients, and will pave the way towards the development of novel functional biomedical, personal care and nutritional applications. 1. Introduction surfaces like eyelids.7 Sensory properties related to skin or oral processing are also examples where lubrication is of paramount importance. For instance, recent studies have demonstrated correlation between the tribological performance of edible hydrogels in soft silicon contacts with oral perception attributes, such as pastiness and slipperiness8 and eﬀects of such unique lubrication on satiety responses.9 Microgels are soft colloids made of cross-linked polymers that are capable of entrapping significant proportions of solvent or water. Their composition and size, which ranges from tens of nanometers to hundreds of microns, can be tailored in order to provide desirable chemical and physical characteristics to target diﬀerent applications.1 Recently, biocompatible microgels, such as those made up of proteins have shown remarkable properties in highly demanding biological applications, such as drug delivery,2,3 altering interfacial digestion kinetics4 and lubrication of biological contacts.5 For instance, submicron- sized whey protein microgels have been used to stabilise Pickering emulsion, providing a transient barrier against enzy- matic diﬀusion by lipase.6 Another example of microgel per- formance is lubrication, which is essential in the development of physiological processes involved in mechanical contacts, such as highly loaded synovial joints as well as delicate rubbing Aqueous lubrication has been recognized as one of the main mechanisms for bio-lubrication due to the abundance of water in complex living organisms.10 Aqueous lubrication is attri- buted to the capacity of biopolymers to electrostatically bind water molecules while being tethered to the bio-surfaces. Hydrated layers formed near the surfaces by this mechanism are capable of supporting high pressure such as those developed in synovial joints.11 The aqueous lubrication performance of colloidal microgels has recently attracted research attention, such as in the case of particles made of thermo-responsive hydrogels12 or biopolymers including starch,13 whey protein14,15 as well as non-starch polysaccharides (alginate,16 agarose,17 k-carrageenan18). The capacity of these soft particles to reduce friction in soft mechanical contacts has been attributed to their action as surface separators in combination to a roll bearing mechanism, the latter phenomenon is still under debate.13 This journal is ©The Royal Society of Chemistry 2019 9614 \| Soft Matter, 2019, 15, 9614--9624 Soft Matter View Article Online Soft Matter View Article Online View Article Online Paper This journal is ©The Royal Society of Chemistry 2019 Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 This article is licensed under a Creative Commons Attribution 3.0 Unported although the protein content in the microgels was either 10.0 or 15.0 wt%. Tribology. Tribological measurements were performed using a Mini Traction Machine (MTM2, PCS instruments, UK) equipped with a tribopair consisting of polydimethylsiloxane (PDMS, Young modulus B2.6 MPa) ball-on-disk. A constant load of 2 N was used in all experiments. Therefore, in this study with the use of a 19.0 mm diameter ball, the Hertzian pressure corresponded approximately to 200.0 kPa29 with a contact radius of 2.0 mm. All measurements were performed at 37 1C in order to mimic physiological conditions. Friction coeﬃcient (m) as a function of the entrainment speed (U) covering a range between 0.001 to 0.3 m s1 with U = (uD + uS)/2, where uD and uS were the disc and ball speeds, respectively. Sliding rolling ratio defined as SRR = \|uD + uS\|/U was kept constant at 0.5. After each measurement, the surfaces were cleaned by 3 cycles of sonica- tion in a 3.0 wt% solution of Decon 90 for 10 minutes each, followed by 2 cycles of sonication in distilled water. A new set of surfaces was used for diﬀerent lubricants. Curves of friction coeﬃcients versus entrainment speed presented here are obtained as the average of three measurements on freshly prepared samples. In order to compare the tribological behaviour of microgels dispersions with that of Newtonian fluids, a master curve for the latter is constructed from the Stribeck curves obtained for buﬀer and corn syrup solutions. For this purpose, the Stribeck curves of the Newtonian fluids with diﬀerent viscosities are overlapped and an average was found using a smoothing algorithm (i.e., neighbour average of 5 points). The curves are presented like friction as a function of entrainment speed or the product of viscosity and entrainment speed. In order to ease the discussion of results, labels are given to the diﬀerent samples according to the content of protein in microgels and the type of continuum (i.e., buﬀer, corn syrup or xanthan gum solution) as described in Table 1. Dynamic light scattering. The particle size of WPM (10.0 or 15.0 wt% protein) was measured using a Zetasizer (Nano ZS series, Malvern Instruments, Worcestershire, UK) by means of dynamic light scattering. Colloidal dispersions containing 50.0 vol% of microgels were diluted in a 20 mM phosphate buﬀer solution at a 1 : 500 v/v ratio and placed in a disposable micro plastic cuvette (ZEN 0040). 2.2 Methods While the lubrication performance of corn syrup has been found to be controlled by the viscosity24 of the New- tonian fluid, the lubrication capacity of xanthan gum has also been largely attributed to aqueous lubrication.25,26 In addition to the changes in the continuum properties, the role of the particle softness was also investigated by synthesising microgels of diﬀerent whey protein content to influence the corresponding protein-crosslinking densities. Variation of mechanical charac- teristics of microgels and continuum allowed establishing a relationship between the high shear rheology of the dispersions and their lubrication properties, which to the best of our knowledge has never been reported for this kind of complex continuum before. The final concentration of microgels in either the buﬀer or the two corn syrup solutions was 50.0 vol%, although the protein content in the microgels was either 10.0 or 15.0 wt%. Volume fraction calculation was based on gravimetric measurement based on previous literature,14 however future work should consider calculating volume fraction of the particles based on theoretical models on soft-hairy colloidal spheres.28 Preparation of whey protein microgel (WPM) dispersions in the non-Newtonian continuum. Xanthan gum powder (0.5 or 1.0 wt%) was added to the whey protein microgel dispersion (10.0 wt% or 15.0 wt% protein) prepared in buﬀer (50 vol% microgels) and stirred at room temperature for 24 h. A final homogenization step was conducted in a water bath at 50 1C and stirred at 65 rpm for 4 hours. The final concen- tration of microgels in xanthan gum solutions was 50.0 vol%, Soft Matter, 2019, 15, 9614--9624 \| 9615 This journal is ©The Royal Society of Chemistry 2019 View Article Online Soft Matter Paper Table 1 Sample nomenclature according to the whey protein content in microgels and continuum composition (i.e., xanthan gum or corn syrup content) Table 1 Sample nomenclature according to the whey protein content in microgels and continuum composition (i.e., xanthan gum or corn syrup content) temperature, using a fluid cell loaded with 20 mM phosphate buffer solution (pH = 7). Images were acquired at 768 pixel resolution and processed using Bruker Nanoscope Analysis v1.9. To our knowledge, this is the first study that has carried out AFM imaging of whey protein microgels. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 This article is licensed under a Creative Commons Attribution 3.0 Unported Measurements were performed using a detection angle of 1731. Time dependent correlation functions were fitted using a cumulant analysis satisfying stan- dard criteria to obtain representative particle size distributions. Analysis was performed assuming refractive index of 1.54 and 1.33 for particles and fluid, respectively. The dispersant viscosity was assumed to be the viscosity of water (i.e., 8.9 104 Pa s). Atomic force microscopy (AFM). Whey protein microgels (10.0 or 15.0 wt% protein) were diluted by a factor of 1000 using 20 mM phosphate buﬀer solution (pH = 7), deposited (100 mL) on clean silicon wafers, and were allowed to adsorb for 10 minutes. Subsequently, to remove the non-adsorbed WPM particles that could adhere onto the AFM tip, the solution was exchanged with 20 mM phosphate buﬀer using a pipette while ensuring that the sample was kept constantly hydrated. Finally, the samples were transferred to an atomic force microscope (AFM) for imaging. Topographic images were acquired using a Bruker Multimode 8 AFM equipped with a Bruker Nanoscope V controller. Silicon nitride AFM cantilevers (model MLCT-DC-BIO) with a nominal spring constant of 0.01 N m1 were purchased from Bruker AFM probes (Camarillo, CA). These cantilevers are thermally stable despite the low spring constant, so it was possible to maintain the force set-point at extremely small values, typically o100 pN producing the most defined images, whilst at 150 pN sample deformation was clearly observed. Slow line rates of 0.5–0.8 Hz and gains at the very upper limits of stability were necessary to track the microgels with minimum disturbance. The measurements were performed at room Statistics analysis. Significant diﬀerences in the experimental data were tested using a one-way ANOVA test with Bonferroni correction. Threshold for significant diﬀerence was established at p o 0.05 using Tukey’s Multiple Comparison Test. This journal is ©The Royal Society of Chemistry 2019 2.2 Methods although the protein content in the microgels was either 10.0 or Sample name Microgel protein content (wt%) Corn syrup content (wt%) Xanthan gum content (wt%) Buﬀer No particles 0.0 0.0 Buﬀer10 10.0 0.0 0.0 Buﬀer15 15.0 0.0 0.0 50CS No particles 50.0 0.0 50CS10 10.0 50.0 0.0 50CS15 15.0 50.0 0.0 75CS No particles 75.0 0.0 75CS10 10.0 75.0 0.0 75CS15 15.0 75.0 0.0 0p5XG No particles 0.0 0.5 0p5XG10 10.0 0.0 0.5 0p5XG15 15.0 0.0 0.5 1XG No particles 0.0 1.0 1XG10 10.0 0.0 1.0 1XG15 15.0 0.0 1.0 Rheology. All rheological measurements were performed using a Kinexus rheometer (Malvern Instruments Ltd, Worcestershire, UK) equipped with a 60 mm diameter cone and plate geometry. The experimental temperature was fixed at 37 1C to mimic oral physiological conditions. The shear viscosity of the continuum and microgel dispersions was obtained for shear rates ranging from 0.1 to 1000.0 s1. For each point, the measured stress stability was ensured within a 0.5% of variability. An adiabatic hood was placed in order to reduce the probability of evaporation occurring during the tests. Viscosity curves presented here are obtained as the average of three measurements on freshly prepared samples. Rheology. All rheological measurements were performed using a Kinexus rheometer (Malvern Instruments Ltd, Worcestershire, UK) equipped with a 60 mm diameter cone and plate geometry. The experimental temperature was fixed at 37 1C to mimic oral physiological conditions. The shear viscosity of the continuum and microgel dispersions was obtained for shear rates ranging from 0.1 to 1000.0 s1. For each point, the measured stress stability was ensured within a 0.5% of variability. An adiabatic hood was placed in order to reduce the probability of evaporation occurring during the tests. Viscosity curves presented here are obtained as the average of three measurements on freshly prepared samples. 3.1 Particle size Atomic force microscopy (AFM) allowed structural characterisation of the submicron-sized whey protein microgels. The extremely soft 9616 \| Soft Matter, 2019, 15, 9614--9624 This journal is ©The Royal Society of Chemistry 2019 Fig. 2 Histograms showing the particle size distributions obtained by means of atomic force microscopy (AFM) for (a) Buﬀer10 and (b) Buﬀer15. Solid green lines show particle size distribution of the same suspensions obtained by means of dynamic light scattering (DLS). Soft Matter View Article Online View Article Online Paper and adherent nature of the microgels led to large instabilities when using Peak Force Tapping mode (even with its fine and absolute force control) and standard Tapping Mode, hence contact mode was used. Topographical images of the microgels Buﬀer10 (Fig. 1) deposited on a silicon surface revealed a homogenous distribution of single spherical particles of microgels as well as small clusters that contained between two to six individual particles. The surface of the microgels appeared to be relatively smooth also seen in other polymeric microgels.30 Using dynamic light scattering, the average hydrodynamic diameter (Dh) of the microgels produced at 10.0 wt% or 15.0 wt% protein content was estimated to be 86.0 (Fig. 2a) and 92.0 nm (Fig. 2b), respectively. Such values are in close range of previous protein microgels reported in literature that are designed using top down approach of heat-set crosslinking and controlled shearing.27,31 Fig. 2 Histograms showing the particle size distributions obtained by means of atomic force microscopy (AFM) for (a) Buﬀer10 and (b) Buﬀer15. Solid green lines show particle size distribution of the same suspensions obtained by means of dynamic light scattering (DLS). Particle size distribution was also calculated by processing the topographic AFM image with Nanoscope Analysis v1.9 and represented as histograms for microgels containing 10.0 (Fig. 2a) or 15.0 wt% (Fig. 2b) protein. Size distributions obtained by AFM are in good agreement with DLS measurements regarding the hydrodynamic radii of the microgels. The particle size distri- butions of microgels with 10.0 wt% and 15.0 wt% protein content dispersed either in buﬀer or 75.0 wt% corn syrup solution are presented in the Fig. S1 (ESI†). The viscosity of the continuum that has been used during microgel formation did not influence the particle size significantly. Fig. 1 Three-dimensional topographic images of microgels Buﬀer10 deposited on silicon substrate, as obtained by contact mode AFM at diﬀerent magnifications. 3.1 Particle size Most of the particles are distributed homo- geneously on the surface, while some of them form aggregates. This journal is ©The Royal Society of Chemistry 2019 Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Fig. 4 Relative shear viscosity of microgel dispersions with Newtonian continuum as function of Pe. Buﬀer10 (close squares), Buﬀer15 (open squares), 50CS10 (close circles), 50CS15 (open circles), 75CS10 (close triangles) and 75CS15 (open triangles). The particle volume fraction of microgels is fixed at 50 vol% in all systems. Error bars represent standard deviations. Zð_gÞ ¼ Z0 Z1 1 þ ðK _gÞm þ Z1 (1) (1) Here, Z0 represents zero shear viscosity, ZN, m and K are the fitting parameters. Fittings to shear viscosity curves on xanthan Here, Z0 represents zero shear viscosity, ZN, m and K are the fitting parameters. Fittings to shear viscosity curves on xanthan gum experimental data are shown as continuous lines in Fig. 3. Extrapolated values of ZN for xanthan gum solutions 0p5XG and 1XG are 8.0 mPa s and 34.0 mPa s, respectively. As anticipated, the estimated values of ZN were close to the Newtonian viscosities of the corn syrup solutions, Table 2. In order to appreciate the impact of microgel inclusions on the overall properties of the dispersions, their relative viscosity (Zr) was plotted as function of the Peclet number (Pe) as defined in eqn (2) and (3), respectively. Fig. 4 Relative shear viscosity of microgel dispersions with Newtonian continuum as function of Pe. Buﬀer10 (close squares), Buﬀer15 (open squares), 50CS10 (close circles), 50CS15 (open circles), 75CS10 (close triangles) and 75CS15 (open triangles). The particle volume fraction of microgels is fixed at 50 vol% in all systems. Error bars represent standard deviations. Zr ¼ Z Zc (2) (2) Pe = 6p_gZca3/KBT (3) (3) Pe = 6p_gZca3/KBT where, Z and Zc are the dispersion and continuum viscosities, a, kB and T are the microgel radius, Boltzmann constant and temperature, respectively. Pe represents the ratio between the characteristic flow time (_g1) and Brownian diﬀusion time of the microgels. Fig. 4 shows the Zr for microgel dispersions in Newtonian continuum. In this representation, dispersions con- taining microgels with 10.0 wt% protein content (Buﬀer10, 50CS10 and 75CS10) show similar shear thinning behaviour, however, this region extends to higher values of Pe as the particle Brownian diﬀusion is slowed down on increasing the continuum viscosity. Although influence of particle size is not explored here, in agreement to eqn (3), increasing particle size should also extend the shear thinning region. This journal is ©The Royal Society of Chemistry 2019 Paper Article Online Soft Matter 3.2 Rheology In this study, microgels were suspended in Newtonian and non- Newtonian fluids, referred as continuum hereafter. Fig. 3 shows the shear viscosity of the diﬀerent continuums, i.e., phosphate buﬀer at pH 7.0, corn syrup and xanthan gum solutions without any added microgels. Phosphate buﬀer (Buﬀer), 50.0 wt% (50CS) and 75.0 wt% (75CS) corn syrup solutions were observed to be Newtonian fluids with viscosities of 0.7, 8.0 and 60.0 mPa s, respectively. Viscosity values found here for corn syrup solution are in agreement with values previously reported in literature.25 Xanthan gum solutions are well known shear-thinning fluids,25,32 Fig. 3 Shear viscosity of Newtonian and non-Newtonian fluids used as continuum in microgel dispersions. Buﬀer (dashed line), 50CS (filled square) and 75CS (filled triangle) are Newtonian fluids with shear viscosity independent to the shear rate _g. 0p5XG (open square) and 1XG (open triangle) are shear thinning non-Newtonian fluids. The continuous lines correspond to fittings using eqn (1). Fig. 3 Shear viscosity of Newtonian and non-Newtonian fluids used as continuum in microgel dispersions. Buﬀer (dashed line), 50CS (filled square) and 75CS (filled triangle) are Newtonian fluids with shear viscosity independent to the shear rate _g. 0p5XG (open square) and 1XG (open triangle) are shear thinning non-Newtonian fluids. The continuous lines correspond to fittings using eqn (1). Fig. 1 Three-dimensional topographic images of microgels Buﬀer10 deposited on silicon substrate, as obtained by contact mode AFM at diﬀerent magnifications. Most of the particles are distributed homo- geneously on the surface, while some of them form aggregates. 3.2 Rheology Soft Matter, 2019, 15, 9614--9624 \| 9617 This journal is ©The Royal Society of Chemistry 2019 Table 2 High shear rate viscosity and friction coeﬃcient for U = 0.005 m s1 (boundary/mixed lubrication regime) and mixed U = 0.04 m s1 (mixed lubrication regime) for all microgels studied here Sample ZN (Pa) m (U = 0.005 m s1, boundary/mixed lubrication) m (U = 0.04 m s1, mixed lubrication) Buﬀer 0.0007 0.67 0.08 0.41 0.06 Buﬀer10 0.004 0.48 0.1* 0.20 0.05 Buﬀer15 0.010 0.23 0.06 0.09 0.01 50CS 0.008 0.89 0.18 0.18 0.03 50CS10 0.010 0.36 0.09* 0.09 0.02 50CS15 0.022 0.14 0.07 0.023 0.005 75CS 0.070 0.07 0.05 0.0049 0.006 75CS10 0.022 0.26 0.05 0.04 0.02 75CS15 0.041 0.07 0.02 0.009 0.001 0p5XG 0.008 0.22 0.08* 0.09 0.03 0p5XG10 No data 0.31 0.04 0.14 0.03 0p5XG15 0.12 0.05 0.05 0.01 1XG 0.034 0.08 0.03 0.020 0.008 1XG10 No data 0.17 0.03 0.080 0.004 1XG15 0.06 0.01 0.032 0.004 Fig. 4 Relative shear viscosity of microgel dispersions with Newtonian continuum as function of Pe. Buﬀer10 (close squares), Buﬀer15 (open squares), 50CS10 (close circles), 50CS15 (open circles), 75CS10 (close triangles) and 75CS15 (open triangles). The particle volume fraction of microgels is fixed at 50 vol% in all systems. Error bars represent standard deviations. Paper View Article Online View Article Online Paper Table 2 High shear rate viscosity and friction coeﬃcient for U = 0.005 m s1 (boundary/mixed lubrication regime) and mixed U = 0.04 m s1 (mixed lubrication regime) for all microgels studied here Table 2 High shear rate viscosity and friction coeﬃcient for U = 0.005 m s1 (boundary/mixed lubrication regime) and mixed U = 0.04 m s1 (mixed lubrication regime) for all microgels studied here quality that can be appreciated in Fig. 3 where viscosities of 0p5XG and 1XG decrease at least two orders of magnitude in the shear rate experimental window covering a range from 0.1 to 1000 s1. At 0.3 s1, 0p5XG and 1XG had values of shear viscosity of approximately 1.5 and 30.0 Pa s, respectively. However, at the highest shear rate (1000 s1), viscosity values observed for 0p5XG and 1XG approximated the Newtonian viscosities of 50CS and 75CS, respectively. One of the aims of this study was to establish the mechanisms dominating the lubrication capacity of soft colloidal microgel dispersions in complex media using a macro- scopic approach. The lubrication performance of some poly- saccharide solutions has been successfully explained in terms of hydrodynamics.25 In this approach, the strength of hydrodynamic forces are approximated using the fluid viscosity in the high shear rate limit. Hence, in order to obtain an estimate for the high shear rate viscosity plateau ZN, experimental data were fitted using the cross model33 presented in eqn (1). Sample ZN (Pa) m (U = 0.005 m s1, boundary/mixed lubrication) m (U = 0.04 m s1, mixed lubrication) Buﬀer 0.0007 0.67 0.08 0.41 0.06 Buﬀer10 0.004 0.48 0.1* 0.20 0.05 Buﬀer15 0.010 0.23 0.06 0.09 0.01 50CS 0.008 0.89 0.18 0.18 0.03 50CS10 0.010 0.36 0.09* 0.09 0.02 50CS15 0.022 0.14 0.07 0.023 0.005 75CS 0.070 0.07 0.05 0.0049 0.006 75CS10 0.022 0.26 0.05 0.04 0.02 75CS15 0.041 0.07 0.02 0.009 0.001 0p5XG 0.008 0.22 0.08* 0.09 0.03 0p5XG10 No data 0.31 0.04 0.14 0.03 0p5XG15 0.12 0.05 0.05 0.01 1XG 0.034 0.08 0.03 0.020 0.008 1XG10 No data 0.17 0.03 0.080 0.004 1XG15 0.06 0.01 0.032 0.004 Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. For instance, shear viscosity reported elsewhere14 for dispersions of microgels with diameter of 300.0 nm (three times larger than reported here) showed a shear thinning region that extended beyond the shear rate (larger than 100.0 s1). In Fig. 4, at the end of the shear thinning region, the second Newtonian plateau is observed, where the values of Zr decrease on increasing the continuum viscosity. For instance, microgels in Buﬀer10 and 50CS10 have an incremental eﬀect on the high shear rate viscosity showing values of relative viscosity of about 5.7 and 1.3, respectively. However, it is clear that increasing continuum viscosity has a detrimental eﬀect on the second plateau viscosity. This is more evident in the case of 75CS10, having the highest viscosity continuum (0.07 Pa s), where the relative plateau viscosity drops about 0.4, i.e., absolute viscosity of the microgel disper- sion is less than half in comparison to the continuum. The role of the microgel properties is further explored by increasing their whey protein content to 15.0 wt%. Viscoelasti- city of the parent whey protein gels was studied before shearing them into microgels in order to have an estimate of the mechanical properties of microgels. The Fig. S2 (ESI†) shows the linear elasticity of the parent whey protein gels. The elastic shear modulus of the whey protein gels at 10.0 wt% and 15.0 wt% protein content were about 0.1 and 10.0 kPa, respectively. Therefore, the microgels containing 10.0 and 15.0 wt% protein in this study have been referred to as soft and hard microgels, This journal is ©The Royal Society of Chemistry 2019 9618 \| Soft Matter, 2019, 15, 9614--9624 View Article Online Paper Soft Matter Soft Matter respectively. Dispersions of microgels with 15.0 wt% protein content show diﬀerences in the shear thinning region. Sample Buﬀer15 exhibits similar thinning in comparison to Buﬀer10 with second Newtonian plateau starting at similar value of Pe B 0.5. Interestingly, relative viscosity of Buﬀer15 shows a twofold increase in comparison to the Buﬀer10 in the plateau region. The increase in viscosity can be associated to relatively higher elastic modulus of microgels with higher protein content. Here again microgels in Buﬀer15 appears to act as thickening agents increasing the viscosity relative to the con- tinuum. Samples 50CS15 and 75CS15 show a similar thinning region between them extending towards Pe values higher than the experimental window. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. It is important to mention that all rheological discussion has been carried out under the assumption of stick conditions where the macroscopic deformation (imposed at the sample- rheometer surface interface) is representative of the deforma- tion at any point of the fluid bulk. However, wall slip is known to be an artefact present when studying highly concentrated dispersions of Brownian solid and soft particles.34 Noteworthy, that microgel dispersions studied here exhibit a fluid-like behaviour, where transmission of strain/stress from the rheo- logical surfaces to the fluid bulk is mostly performed through the continuum making the chances of slippage to be highly unlikely. Furthermore, stress curves are shown in the Fig. S4 and S5 (ESI†) for microgel dispersions in Newtonian and non-Newtonian continuum, respectively. As it can be observed, shear stress increases monotonically showing no signs of discontinuities suggesting no obvious wall slippage. Overall, the rheological analyses are intended to provide an explanation to the tribological performance of the microgels presented in the next section. Fig. 5 shows the relative shear viscosity of microgel disper- sions in xanthan gum as a function of shear rate. Due to the non-Newtonian nature of the continuum, it is not possible to define a single Brownian diﬀusion time, thus the use of Pe was not suitable to describe the balance between the hydrodynamic and Brownian relaxation. Fig. 5 shows a detrimental eﬀect of microgel with 10.0 wt% (0p5XG10 and 1XG10) protein content on the shear viscosity relative to the continuum in the whole range of shear rates studied here. The decrease in viscosity is higher in sample 1XG10 in comparison to 0p5XG10 due to the higher viscosity of the continuum in the former. Both samples started to show an onset of a plateau region at shear rates above 200.0 s1 with Zr about 0.4 and 0.6 for 0p5XG10 and 1XG10, respectively. Diﬀerently, microgels with 15.0 wt% protein This journal is ©The Royal Society of Chemistry 2019 Soft Matter, 2019, 15, 9614--9624 \| 9619 Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. However, values of relative viscosity exhibited by these two fluids are considerably higher in com- parison to the rest of the experimental fluids in the shear thinning region with Pe o 2.0. Thus, their response in the shear thinning region at low Pe numbers splits the microgel dispersions studied here into two groups depending on both, particle rigidity and continuum viscosity. This might be because the definition of Pe does not consider particle deform- ability during flow. Fitted curves to the experimental data are shown in the Fig. S3 (ESI†), where absolute viscosity is presented as function of shear rate. content exhibit either neutral or active role in the dispersion viscosity depending on the shear conditions and continuum properties. Relative viscosity of 0p5XG15 is about 1.0 up to shear rates of 1.0 s1, increasing afterwards to reach value of about 2.0 at 100.0 s1. Interestingly, 1XG15 shows values of Zr as high as 10.0 for shear rates up to 100 s1. After this point, Zr decreases monotonically reaching a value of about 4.0 at the shear rate of 1000.0 s1. In comparison to the rest of the dispersions, the behaviour of 1XG15 is counterintuitive since the higher viscosity of the continuum should contribute to decrease Zr values. This behaviour is evidence of a synergistic eﬀect between microgels with 15.0 wt% protein content and the 1.0 wt% xanthan gum mesh. Further studies are necessary to understand this inversion. In the case of dispersion in xanthan gum continuum, shear viscosity do not show the onset of the second Newtonian plateau making it not possible to fit eqn (1) to the experimental data. 3.3 Soft tribology Fig. 5 Relative shear viscosity of microgel dispersions in non-Newtonian continuum. Namely 0p5XG10 (close squares), 0p5XG15 (open squares), 1XG10 (close triangles) and 1XG15 (open triangles). The particle volume fraction of microgels is fixed at 50.0 vol% in all systems. Error bars represent standard deviations. Dispersions of whey protein microgels in aqueous media have demonstrated the capability to reduce friction in soft contacts. For example, Sarkar et al. studied the lubrication properties of 300 nm diameter whey protein (10.0 wt%) microgels dispersed in aqueous media.14 They demonstrated the capability of colloidal whey protein microgels to reduce friction in compliant sliding/rolling contacts. In this study, the microgels were three- times smaller (Fig. 1 and 2) and also prepared with diﬀerent protein content resulting in diﬀerent degree of protein cross- linking and elasticity (Fig. S2, ESI†) and thus expected to have diﬀerent lubrication performance in absence or presence of the complex continuum. Fig. 6 shows the lubrication performance of microgel dispersions in comparison to their respective continuum (black squares) in the form of Stribeck curves, where the friction coeﬃcients are shown as a function of entrainment speed. Stribeck curves in Fig. 6a show the boundary and mixed lubrication regimes. The boundary regime is identified as the region where friction coeﬃcient is independent of the Fig. 5 Relative shear viscosity of microgel dispersions in non-Newtonian continuum. Namely 0p5XG10 (close squares), 0p5XG15 (open squares), 1XG10 (close triangles) and 1XG15 (open triangles). The particle volume fraction of microgels is fixed at 50.0 vol% in all systems. Error bars represent standard deviations. Fig. 5 Relative shear viscosity of microgel dispersions in non-Newtonian continuum. Namely 0p5XG10 (close squares), 0p5XG15 (open squares), 1XG10 (close triangles) and 1XG15 (open triangles). The particle volume fraction of microgels is fixed at 50.0 vol% in all systems. Error bars represent standard deviations. This journal is ©The Royal Society of Chemistry 2019 Soft Matter, 2019, 15, 9614--9624 \| 9619 Fig. 6 Friction coeﬃcient as a function of U (a, c and e) or ZNU (b, d and f) for microgel dispersions in Newtonian continuum. (a and b) Buﬀer (squares), Buﬀer10 (circles) and Buﬀer15 (triangles). (c and d) 50CS (squares), 50CS10 (circles) and 50CS15 (triangles). (e and f) 75CS (squares), 75CS10 (circles) and 75CS15 (triangles). The particle volume fraction of microgels is fixed at 50.0 vol% in all systems. 3.3 Soft tribology 6 Friction coeﬃcient as a function of U (a, c and e) or ZNU (b, d and f) for microgel dispersions in Newtonian continuum. (a and b) Buﬀer (squares), Buﬀer10 (circles) and Buﬀer15 (triangles). (c and d) 50CS (squares), 50CS10 (circles) and 50CS15 (triangles). (e and f) 75CS (squares), 75CS10 (circles) and 75CS15 (triangles). The particle volume fraction of microgels is fixed at 50.0 vol% in all systems. Continuous line in b, d and e represents an average Stribeck curve of Newtonian fluids (buﬀer and corn syrups) used in this study. Continuous lines in c and d are fittings to the hydrodynamic lubrication regime using eqn (4). Error bars represent standard deviations. performed in the mixed lubrication regime. While values of friction coeﬃcients for 50CS and 50CS10 are not significantly diﬀerent (p 4 0.05), the friction coeﬃcient measured for 50CS15 was significantly lower in comparison to both, 50CS and 50CS10. Moreover, at approximately U = 0.1 m s1, the onset of the hydrodynamic lubrication can be clearly appre- ciated only in the curve corresponding to 50CS15. Fig. 6c presents the Stribeck curves obtained for microgels in 50 wt% syrup solutions (50CS10 and 50CS15) as well as the continuum alone (50CS). Stribeck curves for 50CS and 50CS10 showed the boundary and mixed lubrication regimes. Sample 50CS15 showed the mixed lubrication regime and the onset of the elastohydrodynamic lubrication regime. The latter is observed as an upturn of the Stribeck curve with friction coeﬃcients increasing with the entrainment speeds above 0.1 m s1. At U = 0.005 m s1, friction coeﬃcient obtained for 50CS10 is about 0.36, twice as low in comparison to CS50 (plain continuum) at the same entrainment speed. While at U = 0.005 m s1, 50CS is found to be working in the boundary regime (friction coeﬃcient independent of speed), 50CS10 is working in the mixed lubrication regime showing a subtle decrease in friction on increasing entrainment speed. 50CS75 is also found to be working in the mixed lubrication regime with friction coeﬃcient about 0.14, approximately four and two times lower in comparison to 50CS and 50CS10, respectively. This confirms the positive influence of increasing whey protein in microgels, decreasing the friction coeﬃcient as observed also for microgels dispersed in buﬀer (Fig. 6a). In Fig. 3.3 Soft tribology Continuous line in b, d and e represents an average Stribeck curve of Newtonian fluids (buﬀer and corn syrups) used in this study. Continuous lines in c and d are fittings to the hydrodynamic lubrication regime using eqn (4). Error bars represent standard deviations. Paper View Article Online View Article Online Soft Matter Paper entrainment speed. On the other hand, in the mixed lubrica- tion regime, the friction coeﬃcient decreases monotonically on increasing the speed. The presence of microgels helped decreasing the contact friction in the mixed lubrication regime, in comparison to the continuous phase alone (Buﬀer), however no significant diﬀerences are found in the boundary region. For example, at entrainment speed U = 0.005 m s1 corresponding to the boundary regime, friction coeﬃcient show no signifi- cant diﬀerence between Buﬀer and Buﬀer10 ( p 4 0.05). At the same speed, sample Buﬀer15 was found to be performing in the mixed lubrication regime as evident in the decrease of friction coeﬃcient on increasing entrainment speed. Thus, at U = 0.005 m s1, friction coeﬃcient for Buﬀer15 was about 0.23, significantly lower ( p o 0.05) than values obtained for Buﬀer and Buﬀer10. At U = 0.04 m s1, Buﬀer, Buﬀer10 and Buﬀer15 were found to be working in the mixed lubrication regime. Friction coeﬃcient of buﬀer was about three to five times higher with respect to the friction coeﬃcient values for Buﬀer10 and Buﬀer15, respectively ( p o 0.05). These observations show the influence of the microgel mechanical properties (i.e., elastic modulus) on their lubrication performance, the higher the particle elastic modulus (i.e., higher protein content) the lower was the contact friction, at least in the mixed lubrication regime. Friction coeﬃcient Buﬀer10 is similar to friction mea- surements (using the same set-up) reported elsewhere14 for similar whey protein microgels (10.0 wt%) with average diameter three times larger. Thus, increasing particle size of the microgels might have no influence on the tribological performance of the final dispersion. A systematic study on impact of average microgel size on lubrication properties is required; however, this is out of scope of the present study. All values of friction are summarised in Table 2 with the corresponding values of ZN obtained using eqn (1) except for microgels in xanthan gum solutions. Fig. This journal is ©The Royal Society of Chemistry 2019 9620 \| Soft Matter, 2019, 15, 9614--9624 Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 This article is licensed under a Creative Commons Attribution 3.0 U m = 1.46U¯ 0.65 %W 0.70 + SRR(3.8U¯ 0.71 %W 0.76 + 0.96U¯ 0.36 %W 0.11) (4) ) (4) Here, SRR is the sliding/rolling ratio, taking values from 0.0 in pure rolling conditions to 2.0 in the pure sliding conditions. Here, SRR is the sliding/rolling ratio, taking values from 0.0 in pure rolling conditions to 2.0 in the pure sliding conditions. In eqn (4), U ¼ UZ ER0 and W ¼ W ER02, where E0 and R0 are the reduced Young modulus and the ball radius, respectively. Eqn (4) was successfully fitted to the experimental data using viscosity as the only fitting parameter. Fittings are presented as lines in Fig. 6. Values of viscosity estimated for 50CS15, 75CS and 75CS15 using eqn (4) are 0.017, 0.05 and 0.025 Pa s, respectively. Although in comparisons to values presented in Table 2, values obtained using eqn (4) are slightly lower, the trend obtained in rheology is replicated. Thus, when dispersed in the highest viscosity continuum CS75 (0.07 Pa s), microgels have a detrimental eﬀect on both, rheological and lubrication properties. Here, SRR is the sliding/rolling ratio, taking values from 0.0 in pure rolling conditions to 2.0 in the pure sliding conditions. In eqn (4), U ¼ UZ ER0 and W ¼ W ER02, where E0 and R0 are the reduced Young modulus and the ball radius, respectively. Eqn (4) was successfully fitted to the experimental data using viscosity as the only fitting parameter. Fittings are presented as lines in Fig. 6. Values of viscosity estimated for 50CS15, 75CS and 75CS15 using eqn (4) are 0.017, 0.05 and 0.025 Pa s, respectively. Although in comparisons to values presented in Table 2, values obtained using eqn (4) are slightly lower, the trend obtained in rheology is replicated. Thus, when dispersed in the highest viscosity continuum CS75 (0.07 Pa s), microgels have a detrimental eﬀect on both, rheological and lubrication properties. Having established the influence of microgels in Newtonian continuum of diﬀerent viscosities in the rheological and tribo- logical limit, a similar study was also conducted for the non- Newtonian continuum, i.e., xanthan gum solutions. Fig. 7a shows the Stribeck curves obtained for microgels dispersions 0p5XG10 and 0p5XG15 in comparison to their continuum alone (0p5XG). Stribeck curves for all lubricants shown in Fig. 7a show only the mixed lubrication regime. In Fig. 7a, only 0p5XG15 shows significant diﬀerences. 3.3 Soft tribology 6b, at entrainment speed U = 0.04 m s1, 50CS, 50CS10 and 50CS15 In the case of simple Newtonian fluids, increasing lubricant viscosity decreases the speed at which the elastohydrodynamic lubrication regime starts.24 This can be clearly observed com- paring 50CS (Fig. 6c) and 75CS (Fig. 6e), both Newtonian fluids having viscosities of 0.008 and 0.07 Pa s, respectively. Due to its larger viscosity, 75CS displays the elastohydrodynamic lubrica- tion, while in the same range of speeds 50CS shows only the mixed regime. Applying this reasoning to dispersions 50CS10 and 50CS15 in Fig. 6c, the appearance of an elastohydro- dynamic regime for the latter indicates larger viscosity in comparison to the former. In Fig. 6e, Stribeck curves correspond to 75CS and microgels dispersions 75CS10 and 75CS15. Opposite to the observations in dispersion in lower viscosity continuum (i.e., Buﬀer10, 50CS10), microgels in 75CS10 increased friction in the mixed lubrication regime in comparison to the continuum alone 9620 \| Soft Matter, 2019, 15, 9614--9624 This journal is ©The Royal Society of Chemistry 2019 9620 View Article Online Paper Soft Matter (75CS). For instance, at U = 0.005 m s1, friction coeﬃcient found for 75CS10 is about 0.26, representing more than a three- fold increase with respect to 75CS. 75CS and 75CS15 show similar lubrication, however the onset of elastohydrodynamic lubrication reveals diﬀerences in viscosity. Onset of elasto- hydrodynamic lubrication of 75CS appears at the lowest speed at about 0.03 m s1, followed by 75CS15 at 0.07 m s1 and finally 75CS10 shows the end of mixed lubrication (and start of elastohydrodynamic regime) at a higher speed of about 0.2 m s1. Thus, in agreement to discussion presented above, viscosity values of the lubricants follow the same order from highest to lowest, i.e., 75CS 4 75CS15 4 75CS10. This is also in line with values of viscosity obtained by means of steady shear rheological experiments as shown in Table 2. The appearance of the hydrodynamic lubrication regime in friction curves for samples 50CS15, 75CS and 75CS15 makes it possible to have an estimation of eﬀective viscosity from a tribological perspective. de Vicente et al.,35 provided an arithmetic expression for the soft-hydrodynamic lubrication regime of a rolling/sliding ball on plate contact as introduced in eqn (4): representing the friction coeﬃcient as function of the product of entrainment speed and ZN. This journal is ©The Royal Society of Chemistry 2019 Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. 3.3 Soft tribology In fact, the successful fitting of eqn (4) in the elastohydrodynamic lubrication regime is a clear indication that this procedure is adequate at least for microgels in Newtonian continuum. In the case of lubricants 75CS10 and 75CS15, the plateau was not observed in the experimental window and values for ZN were estimated by extrapolation using eqn (1). Fig. 6b, d and e show friction coeﬃcient as function of the product UZN for microgel dispersion with Buﬀer, 50CS and 75CS as continuum, respectively. A Stribeck curve representing the average lubrication performance of Newtonian lubricants is also presented for comparison purposes. There is a reasonable overlap of Stribeck curves for dispersions regardless of their continuum and protein content in the microgels. Hence, lubrication performance of colloidal microgel dispersions in media with diﬀerent viscosity (covering three orders of magnitude) can be approximated by considering them as Newtonian fluids having viscosity ZN. In other words, tribological deformation corresponds to the rheological shear rates in the second Newtonian plateau of the microgel dispersions, which has been seldom reported in literature. It is worth noticing that lubricants 50CS15 in Fig. 6d, 75CS10 and 75CS15 in Fig. 6f, show a clear overlap in the onset of the transition between the mixed and hydrodynamic lubrication regime. This is an important indication that hydrodynamic forces acting on the microgel dispersions in the soft tribo- logical limit are still dominated by the rheological response in the equivalent shear rate range, i.e., high shear rate limit. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 This article is licensed under a Creative Commons Attribution 3.0 U For instance, at entrainment speed 0.005 and 0.04 m s1, friction coeﬃcient values for 0p5XG15 are about half the values found for 0p5XG and 0p5XG10 at the same speeds. Fig. 7b shows the Stribeck curves for microgels dispersed in the highest viscosity non- Newtonian continuum (1XG). While Stribeck curves for 1XG and 1XG15 show no relative diﬀerences, friction coeﬃcients obtained for 1XG10 are significantly higher in comparison to the other two curves up to U = 0.1 m s1, where a minimum for the friction coeﬃcient is reached. The appearance of a hydro- dynamic lubrication regime is only evident in the case of 1XG15, where fitting of eqn 4 generates a value of viscosity of about 0.2 Pa s. Using this calculation, a comparison between rheological and tribological performance can be carried out for 1XG and 1XG15. Eﬀective viscosity of 1XG15 in the elasto- hydrodynamic regime (eqn (4)) represents a six-fold increase in comparison to the high shear viscosity value extrapolated for 1XG (Table 2). However, in the mixed lubrication regime, no significant diﬀerences are found between friction coeﬃcients of 1XG and 1XG15. Thus, it is clear that a single value of A common procedure to compare the tribological perfor- mance of diﬀerent fluids is accomplished by using their rheological properties representing friction coeﬃcient as func- tion of the product of viscosity and entrainment speed. In the case of Newtonian fluids, this procedure is well known to deliver master curves (overlapping of curves representing diﬀerent lubricants) that covers at least the mixed and hydro- dynamic lubrication regimes, where fluid dynamics plays an important role in the lubrication.35 Although for most non- Newtonian lubricants the viscosity is a function of working conditions (i.e., shear rate), a similar representation is commonly found where the Newtonian viscosity is replaced by a value of viscosity at a high shear rate (above 1000.0 s1).25 This approxi- mation relies on the fact that the tribological limit imposes high shear rates on the lubricant due to the proximity between contact surfaces involving submicron separations. Here, the relation between lubrication performance of non-Newtonian microgel dispersions and their rheology is carried out by Soft Matter, 2019, 15, 9614--9624 \| 9621 View Article Online Soft Matter Paper Soft Matter Paper Fig. 7 Friction coeﬃcient as a function of U for the microgel dispersions non-Newtonian continuum. (a) 0p5XG (squares), 0p5XG10 (circles) and 0p5XG15 (triangles). 4. Discussion Ca = _gZc/G (5) (5) with G as the particle elastic modulus. They demonstrated that a value of Ca = 0.88 represented a threshold for the rheological performance of dispersions of soft elastic particles. Below this value, relative viscosity of the dispersion was greater than 1.0 regardless of particle concentration, however, above this value, relative viscosity drops below 1.0. In order to understand the role of particle elasticity on the rheological response of our microgel dispersions, eqn (5) is evaluated using the elasticity of the parent gels as estimates for the mechanical properties of microgels in Newtonian continuum. The highest value for Ca obtained for a dispersion in the thinning region of Fig. 8 is 0.07 corresponding to 75CS10 when eqn (5) is evaluated at a shear rate of 100.0 s1. This value of shear rate is chosen since it is the starting of the second Newtonian plateau for 75CS10 as shown in Fig. S3 (ESI†). The fact that Ca value calculated for 75CS10 is an order of magnitude lower than the threshold value found by Avazmohammadi and Castaneda36 (Ca = 0.88) indicates that these soft microgels with low elastic modulus (high deformability) are not solely responsible for their beha- viour as thinning agents. An alternative explanation for relative viscosity lower than one is draining of the buﬀer from particles into the continuum decreasing the eﬀective concentration of the corn syrup, hence decreasing the eﬀective viscosity of the continuum and dispersion. The use of eqn (5) is restricted to Fig. 8 summarizes the role of microgels on the high shear rheology and lubrication performance of all the microgel dis- persions studied. To do so, the relative friction coeﬃcient, with G as the particle elastic modulus. They demonstrated that a value of Ca = 0.88 represented a threshold for the rheological performance of dispersions of soft elastic particles. Below this value, relative viscosity of the dispersion was greater than 1.0 regardless of particle concentration, however, above this value, relative viscosity drops below 1.0. In order to understand the role of particle elasticity on the rheological response of our microgel dispersions, eqn (5) is evaluated using the elasticity of the parent gels as estimates for the mechanical properties of microgels in Newtonian continuum. The highest value for Ca obtained for a dispersion in the thinning region of Fig. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 This article is licensed under a Creative Commons Attribution 3.0 Unported Fig. 7 Friction coeﬃcient as a function of U for the microgel dispersions non-Newtonian continuum. (a) 0p5XG (squares), 0p5XG10 (circles) and 0p5XG15 (triangles). (b) 1XG (squares), 1XG10 (circles) and 1XG15 (triangles). The particle volume fraction of microgels is fixed at 50 vol% in all fluids. Continuous lines in b represent fitting to the hydrodynamic lubrication regime using eqn (4). Error bars represent standard deviations. Using the definition of relative viscosity, data presented in Fig. 8 is divided in two regions representing the influence of microgels as thickening or thinning agents. In the thinning region, relative high shear rate viscosity is lower than 1.0, with microgels decreasing the viscosity with respect to the conti- nuum. Whether a dispersions falls in either regime depends on both, continuum viscosity and the mechanical properties of the microgel particles. viscosity ZN cannot be used to explain the lubrication perfor- mance of microgels in non-Newtonian continuum (xanthan gum). In addition, in the non-Newtonian continuum, microgels in the lowest viscosity continuum (0p5XG) can improve the lubrication with respect to the continuum and the same microgels can cause a detrimental eﬀect when dispersed in the higher viscosity continuum (1XG). Regardless of the con- tinuum properties, lubrication performance of the microgel dispersion is better for the higher protein content particles. This is true for both, Newtonian and non-Newtonian con- tinuum. Based on a balance between stored (in soft elastic particles) and dissipated (by Newtonian continuum) energy during shear flow, Avazmohammadi and Castaneda36 esti- mated the viscosity of soft particles dispersions in Newtonian continuum as function of the dimensionless parameter Ca defined in eqn (5), as shown below: This journal is ©The Royal Society of Chemistry 2019 4. Discussion 8 is 0.07 corresponding to 75CS10 when eqn (5) is evaluated at a shear rate of 100.0 s1. This value of shear rate is chosen since it is the starting of the second Newtonian plateau for 75CS10 as shown in Fig. S3 (ESI†). The fact that Ca value calculated for 75CS10 is an order of magnitude lower than the threshold value found by Avazmohammadi and Castaneda36 (Ca = 0.88) indicates that these soft microgels with low elastic modulus (high deformability) are not solely responsible for their beha- viour as thinning agents. An alternative explanation for relative viscosity lower than one is draining of the buﬀer from particles into the continuum decreasing the eﬀective concentration of the corn syrup, hence decreasing the eﬀective viscosity of the continuum and dispersion. The use of eqn (5) is restricted to Fig. 8 Relative friction coeﬃcient as function of high shear rate relative viscosity for all fluids studied here. Square and triangle shaped symbols represent microgels with 10.0 and 15.0 wt% protein content, respectively. Relative friction coeﬃcients were calculated using absolute values of friction coeﬃcient corresponding to that of the mixed lubrication regime values obtained at U = 0.04 m s1. Closed symbols correspond to Newtonian continuum, whilst open symbols are for non-Newtonian continuum. Increasing symbol size represent higher continuum viscosity. Fig. 8 Relative friction coeﬃcient as function of high shear rate relative viscosity for all fluids studied here. Square and triangle shaped symbols represent microgels with 10.0 and 15.0 wt% protein content, respectively. Relative friction coeﬃcients were calculated using absolute values of friction coeﬃcient corresponding to that of the mixed lubrication regime values obtained at U = 0.04 m s1. Closed symbols correspond to Newtonian continuum, whilst open symbols are for non-Newtonian continuum. Increasing symbol size represent higher continuum viscosity. Fig. 8 Relative friction coeﬃcient as function of high shear rate relative viscosity for all fluids studied here. Square and triangle shaped symbols represent microgels with 10.0 and 15.0 wt% protein content, respectively. Relative friction coeﬃcients were calculated using absolute values of friction coeﬃcient corresponding to that of the mixed lubrication regime values obtained at U = 0.04 m s1. Closed symbols correspond to Newtonian continuum, whilst open symbols are for non-Newtonian continuum. Increasing symbol size represent higher continuum viscosity. Fig. 8 Relative friction coeﬃcient as function of high shear rate relative viscosity for all fluids studied here. Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 This article is licensed under a Creative Commons Attribution 3.0 U (b) 1XG (squares), 1XG10 (circles) and 1XG15 (triangles). The particle volume fraction of microgels is fixed at 50 vol% in all fluids. Continuous lines in b represent fitting to the hydrodynamic lubrication regime using eqn (4). Error bars represent standard deviations. Soft Matter defined as the ratio of the friction coeﬃcient of the dispersions to the value of the plain continuum, is represented as function relative viscosity in the high shear rate limit. Friction coeﬃcient values used here correspond to an entrainment speed of 0.04 m s1, which lies in the mixed lubrication regime for all dispersions. Choice of relative viscosity diﬀers between dispersions in Newtonian and non-Newtonian continuum. For dispersions in buﬀer or corn syrup solutions, the viscosity corresponds to the extrapolated values ZN using eqn (1). Since extrapolation was not suitable for dispersions in non-Newtonian continuum, viscosity used in Fig. 8 corresponds to values at the highest shear rate accessible (1000.0 s1). It is evident that lubrication and high shear rate viscosity are closely related, with friction coeﬃcients decreasing as the dispersion viscosity increases for both, Newtonian and non-Newtonian continuum. References 1 J. K. Oh, R. Drumright, D. J. Siegwart and K. Matyjaszewski, Prog. Polym. Sci., 2008, 33, 448–477. 2 N. Murthy, M. C. Xu, S. Schuck, J. Kunisawa, N. Shastri and J. M. J. Frechet, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 4995–5000. Acknowledgements This study has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement no 757993). (6) The shear rate is then approximated as _g = SRRU/hc. Samples showing elastohydrodynamic regime in the range of entrainment speeds studied here are 50CS15, 75CS, 75CS15 and 1XG15. Table 3 shows estimation of central film thickness and shear rate together with values of viscosity and speed used to evaluate eqn (6). Entrainment speed values used for calculations correspond to the onset of elastohydrodynamic lubrication. Viscosity corre- sponds to values obtained from fitting eqn (4) to the elasto- hydrodynamic lubrication regime of the diﬀerent dispersions. Values of hc for all microgels in Newtonian continuum are about 7.0 106 m, whilst for dispersion in xanthan gum solution thickness is one order of magnitude higher i.e., about 3.0 105 m. Nevertheless, in all cases, shear rate is above 1000.0 s1. This provides further justification to the use of second Newtonian plateau viscosity to represent the tribo- logical performance of microgels in Newtonian continuum. In the case of microgels in non-Newtonian continuum (xanthan gum solution), their tribological performance is not simply represented by one viscosity value. However, their elastohydro- dynamic lubrication regime might be well approximated by their bulk viscosity. This journal is ©The Royal Society of Chemistry 2019 Open Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Microgels enhanced the rheological and tribological performance when dispersed in continuum with relatively low viscosity, increasing the viscosity i.e., acting as thickeners, thus lowering the friction coeﬃcient. However in high viscosity continuum, microgels caused a detrimental eﬀect in mechanical performance, decreasing the viscosity and increasing the friction coeﬃcients. Regardless of being dispersed in Newtonian or non-Newtonian continuum, increasing the elastic modulus of the microgels (higher protein content) showed benefits by increasing the high shear rate viscosity of the final dispersion. This consequently improved their tribological performance, i.e., decreased the friction coeﬃcients. Thus, in order to obtain a benefit from microgels on the tribology performance of a dispersion, it is necessary to have harder particles when the continuum viscosity is increased. However, further studies are necessary to determine when microgels become too hard that they can increase the abrasiveness in biological contacts. Future work should look into a more thorough calculation of the eﬀective volume fraction of the soft microgel particles in order to compare with theoretical models. Newtonian continuum and thus not suitable to describe micro- gels in xanthan gum solutions. However, since high shear rate viscosity of Newtonian and non-Newtonian continuums is similar, it can be argued that even in the case of microgels in xanthan gum solutions, low elastic modulus of microgels is not responsible for the drop in relative viscosity below 1.0. Since the hydrodynamic-elastic balance is not capable to explain the reduction of viscosity, other mechanisms possible include particle draining and disruption of the xanthan gum network by microgels due to possible polymer–particle interactions. Finally, a rough estimation of the shear rates in the elasto- hydrodynamic lubrication regime is possible by using the expression provided by de Vicente et al.35 for the central film thickness as shown in eqn (6). hc = 3.3R0U¯ 0.6 %W 0.14 (6) hc = 3.3R0U¯ 0.6 %W 0.14 4. Discussion Square and triangle shaped symbols represent microgels with 10.0 and 15.0 wt% protein content, respectively. Relative friction coeﬃcients were calculated using absolute values of friction coeﬃcient corresponding to that of the mixed lubrication regime values obtained at U = 0.04 m s1. Closed symbols correspond to Newtonian continuum, whilst open symbols are for non-Newtonian continuum. Increasing symbol size represent higher continuum viscosity. Fig. 8 Relative friction coeﬃcient as function of high shear rate relative viscosity for all fluids studied here. Square and triangle shaped symbols represent microgels with 10.0 and 15.0 wt% protein content, respectively. Relative friction coeﬃcients were calculated using absolute values of friction coeﬃcient corresponding to that of the mixed lubrication regime values obtained at U = 0.04 m s1. Closed symbols correspond to Newtonian continuum, whilst open symbols are for non-Newtonian continuum. Increasing symbol size represent higher continuum viscosity. 9622 \| Soft Matter, 2019, 15, 9614--9624 This journal is ©The Royal Society of Chemistry 2019 View Article Online Paper Soft Matter Table 3 Film thickness hc and shear rates calculated for dispersions showing elastohydrodynamic lubrication regime in tribological experi- ments. Values of viscosity and entrainment speed used for calculations are also shown of microgels dispersed in either Newtonian or non-Newtonian continuum (xanthan gum solution) is related to their hydro- dynamics. On the one hand, the tribological performance of microgels dispersed in Newtonian continuum can be quantita- tively explained in terms of their second Newtonian plateau bulk viscosity. On the other hand, for microgels dispersed in xanthan gum solutions, the relation between bulk rheology and soft tribological performance is qualitative. Dispersion Viscosity (Pa) U (m s1) hc (m) Shear rate (s1) 50CS10 0.017 0.10 7.4 106 6764.0 75CS 0.050 0.03 6.9 106 2187.0 75CS15 0.025 0.07 7.5 106 4653.0 1XG15 0.200 0.10 3.2 105 1541.0 Dispersion Viscosity (Pa) U (m s1) hc (m) Shear rate (s1) n Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. 25 J. de Vicente, J. R. Stokes and H. A. Spikes, Food Hydro- colloids, 2006, 20, 483–491. 10 S. Jahn and J. Klein, Macromolecules, 2015, 48, 5059–5075. 26 J. R. Stokes, L. Macakova, A. Chojnicka-Paszun, C. G. de Kruif and H. H. J. de Jongh, Langmuir, 2011, 27, 3474–3484. 11 M. Wathier, B. A. Lakin, P. N. Bansal, S. S. Stoddart, B. D. Snyder and M. W. Grinstaﬀ, J. Am. Chem. Soc., 2013, 135, 4930–4933. 27 A. Araiza-Calahorra and A. Sarkar, Food Struct., 2019, 21, 100113. 12 G. Q. Liu, X. L. Wang, F. Zhou and W. M. Liu, ACS Appl. Mater. Interfaces, 2013, 5, 10842–10852. 28 W. C. K. Poon, E. R. Weeks and C. P. Royall, Soft Matter, 2012, 8, 21–30. 13 O. Torres, E. Andablo-Reyes, B. S. Murray and A. Sarkar, ACS Appl. Mater. Interfaces, 2018, 10, 26893–26905. 29 A. Sarkar, E. Andablo-Reyes, M. Bryant, D. Dowson and A. Neville, Curr. Opin. Colloid Interface Sci., 2019, 39, 61–75. 14 A. Sarkar, F. Kanti, A. Gulotta, B. S. Murray and S. Y. Zhang, Langmuir, 2017, 33, 14699–14708. 30 A. Aufderhorst-Roberts, D. Baker, R. J. Foster, O. Cayre, J. Mattsson and S. D. Connell, Nanoscale, 2018, 10, 16050–16061. 15 M. L. Olivares, K. Shahrivar and J. de Vicente, J. Food Eng., 2019, 245, 157–165. 31 A. Sarkar, V. Ademuyiwa, S. Stubley, N. H. Esa, F. M. Goycoolea, X. F. Qin, F. Gonzalez and C. Olvera, Food Hydrocolloids, 2018, 84, 282–291. 16 I. F. Farres, M. Douaire and I. T. Norton, Food Hydrocolloids, 2013, 32, 115–122. 32 O. Torres, A. Yamadaa, N. M. Rigby, T. Hanawab, Y. Kawano and A. Sarkar, Biotribology, 2019, 17, 8–18. 17 A. Gabriele, F. Spyropoulos and I. T. Norton, Soft Matter, 2010, 6, 4205–4213. 18 D. A. Garrec and I. T. Norton, Food Hydrocolloids, 2013, 33, 160–167. 33 J. Xie and Y. C. Jin, Eng. Appl. Comp. Fluid, 2016, 10, 111–129. 34 S. P. Meeker, R. T. Bonnecaze and M. Cloitre, J. Rheol., 2004, 48, 1295–1320. 19 J. de Vicente, H. A. Spikes and J. R. Stokes, J. Tribol., 2006, 128, 795–800. 20 I. F. Farres and I. T. Norton, Food Hydrocolloids, 2015, 45, 186–195. 35 J. de Vicente, J. R. Stokes and H. A. Spikes, Tribol. Lett., 2005, 20, 273–286. 20 I. F. Farres and I. T. Norton, Food Hydrocolloids, 2015, 45, 186–195. 35 J. de Vicente, J. R. Stokes and H. A. This journal is ©The Royal Society of Chemistry 2019 9624 \| Soft Matter, 2019, 15, 9614--9624 5. Conclusions 3 S. V. Vinogradov, Curr. Pharm. Des., 2006, 12, 4703–4712. 4 A. Sarkar, S. N. Zhang, M. Holmes and R. Ettelaie, Adv. Colloid Interface Sci., 2019, 263, 195–211. Soft whey protein microgels with average size of about 100.0 nm (determined by AFM and DLS) dispersed in Newtonian and non-Newtonian continuum were investigated by means of rheology and soft tribology. Steady shear rheology and soft tribology experiments showed that the lubrication performance 5 G. Q. Liu, Z. L. Liu, N. Li, X. L. Wang, F. Zhou and W. M. Liu, ACS Appl. Mater. Interfaces, 2014, 6, 20452–20463. 6 A. Sarkar, B. Murray, M. Holmes, R. Ettelaie, A. Abdalla and X. Y. Yang, Soft Matter, 2016, 12, 3558–3569. Soft Matter, 2019, 15, 9614--9624 \| 9623 This journal is ©The Royal Society of Chemistry 2019 View Article Online Soft Matter Paper Paper 22 R. G. Larson and P. S. Desai, Annu. Rev. Fluid Mech., 2015, 47, 47–65. 7 O. Sterner, C. Karageorgaki, M. Zurcher, S. Zurcher, C. W. Scales, Z. Fadli, N. D. Spencer and S. G. P. Tosatti, ACS Appl. Mater. Interfaces, 2017, 9, 20150–20160. 23 S. Adams, W. J. Frith and J. R. Stokes, J. Rheol., 2004, 48, 1195–1213. 8 E. M. Krop, M. M. Hetherington, M. Holmes, S. Miquel and A. Sarkar, Food Hydrocolloids, 2019, 88, 101–113. 24 J. de Vicente, J. R. Stokes and H. A. Spikes, Proc. Inst. Mech. Eng., Part J, 2006, 220, 55–63. 9 E. M. Krop, M. M. Hetherington, S. Miquel and A. Sarkar, Food Qual. Prefer., 2019, 74, 118–124. n Access Article. Published on 11 October 2019. Downloaded on 10/24/2024 6:56:49 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Spikes, Tribol. Lett., 2005, 20, 273–286. 21 T. Kairn, P. J. Daivis, I. Ivanov and S. N. Bhattacharya, J. Chem. Phys., 2005, 123, 194905. 36 R. Avazmohammadi and P. P. Castaneda, J. Fluid Mech., 2015, 763, 386–432. 36 R. Avazmohammadi and P. P. Castaneda, J. Fluid Mech., 2015, 763, 386–432. 21 T. Kairn, P. J. Daivis, I. Ivanov and S. N. Bhattacharya, J. Chem. Phys., 2005, 123, 194905. 9624 \| Soft Matter, 2019, 15, 9614--9624
https://openalex.org/W4318219079	https://link.springer.com/content/pdf/10.1007/s40618-023-02013-7.pdf	English	null	Is transperitoneal laparoscopic adrenalectomy for pheochromocytoma really more challenging? A propensity score-matched analysis	Journal of endocrinological investigation	2,023	cc-by	5,961	* D. Corallino diletta.corallino1989@gmail.com Abstract Purpose Minimally invasive surgery is the gold standard treatment for adrenal masses, but it may be a challenging procedure in the case of pheochromocytoma (PHEO). The aim of the present study is to report the results of transperitoneal laparoscopic adrenalectomy (TLA) in cases of PHEO in comparison to other types of adrenal lesions.i Methods From 1994 to 2021, 629 patients underwent adrenalectomy. Twenty-two and thirty-five patients, respectively, were excluded because they underwent bilateral and open adrenalectomy, leaving 572 patients for inclusion. Of these, 114 patients had PHEO (Group A), and 458 had other types of lesions (Group B). To adjust for potential baseline confounders, a propensity score matching (PSM) analysis was conducted. p p y g y Results After PSM, 114 matched pairs of patients were identified from each group. Statistically significant differences were not observed when comparing the median operative time (85 and 90 min in Groups A and B, respectively, p = 0.627), conversion rate [6 (5.3%) in each group, p = 1.000], transfusion rate [4 (3.5%) and 3 (2.6%) in Groups A and B, respectively, p = 1.000], complication rate [7 (6.1%) and 9 (7.9%) in Groups A and B, respectively, p = 0.796), median postoperative hospi- tal stay (3.9 and 3.6 days in Groups A and B, respectively, p = 0.110), and mortality rate [1 (0.9%) in each group, p = 1.000]. Conclusions Based on this analysis, the results of TLA for PHEO are equivalent to those of TLA for other types of adrenal lesions, but the fundamental requirements are multidisciplinary patient management and adequate surgeon experience. Further prospective studies are required to draw definitive conclusions. Keywords Pheochromocytoma (PHEO) · Adrenal lesions · Transperitoneal laparoscopic adrenalectomy (TLA) · Open adrenalectomy (OA) · Propensity score matching (PSM) Is transperitoneal laparoscopic adrenalectomy for pheochromocytoma really more challenging? A propensity score‑matched analysis D. Corallino1 · A. Balla2 · L. Palmieri1 · I. Sperduti3 · M. Ortenzi4 · M. Guerrieri4 · A. M. Paganini1 D. Corallino1 · A. Balla2 · L. Palmieri1 · I. Sperduti3 · M. Ortenzi4 · M. Guerrieri4 · A. M. Pagan Received: 13 September 2022 / Accepted: 12 January 2023 / Published online: 27 January 2023 © The Author(s) 2023 Journal of Endocrinological Investigation (2023) 46:1589–1596 https://doi.org/10.1007/s40618-023-02013-7 Journal of Endocrinological Investigation (2023) 46:1589–1596 https://doi.org/10.1007/s40618-023-02013-7 ORIGINAL ARTICLE ORIGINAL ARTICLE 1 Department of General Surgery and Surgical Specialties, Sapienza University of Rome, Viale del Policlinico 155, 00161 Rome, Italy * D. Corallino diletta.corallino1989@gmail.com 1 Department of General Surgery and Surgical Specialties, Sapienza University of Rome, Viale del Policlinico 155, 00161 Rome, Italy 2 UOC of General and Minimally Invasive Surgery, Hospital “San Paolo”, Largo Donatori del Sangue 1, 00053 Civitavecchia, Rome, Italy 3 Department of Biostatistics, IRCCS Regina Elena National Cancer Institute, Via Elio Chianesi 53, 00144 Rome, Italy 4 Department of General Surgery, Università Politecnica Delle Marche, Piazza Roma 22, 60121 Ancona, Italy Introduction condition due to excessive catecholamine production and is characterized by clinical symptoms, including sweating, headaches, palpitations, perspiration, tremors, and facial pal- lor [4, 5]. These symptoms are often paroxysmal or can be induced by a variety of events, such as strenuous physical exertion, delivery, trauma, anaesthesia, and surgery [4, 5]. Pheochromocytoma (PHEO) is a rare catecholamine-pro- ducing neuroendocrine tumour arising from chromaffin cells of the adrenal medulla [1–3]. PHEO is a life-threatening In patients with PHEO, adrenalectomy is the gold stand- ard treatment, although resection may be challenging for surgeons due to a high risk of intraoperative haemodynamic instability from excessive catecholamine release during the induction of anaesthesia and with intraoperative surgical manipulation of the gland [6–9].i 1 Department of General Surgery and Surgical Specialties, Sapienza University of Rome, Viale del Policlinico 155, 00161 Rome, Italy 2 UOC of General and Minimally Invasive Surgery, Hospital “San Paolo”, Largo Donatori del Sangue 1, 00053 Civitavecchia, Rome, Italy Since the first report of transperitoneal laparoscopic adre- nalectomy (TLA) in 1992 [10], minimally invasive surgery (MIS) has largely replaced open adrenalectomy (OA) as the standard procedure, but its role in the case of large PHEOs and cancer is still debated [6, 11–14]. 3 Department of Biostatistics, IRCCS Regina Elena National Cancer Institute, Via Elio Chianesi 53, 00144 Rome, Italy 4 Department of General Surgery, Università Politecnica Delle Marche, Piazza Roma 22, 60121 Ancona, Italy (0121 3456789) 3 3456789) 3 1590 Journal of Endocrinological Investigation (2023) 46:1589–1596 Although laparoscopic adrenalectomy (LA) and robotic adrenalectomy (RA) versus OA for the management of PHEO have been reported to be associated with better post- operative outcomes in terms of lower estimated blood loss, a shorter hospital stay, and less haemodynamic instability [15, 16], the effectiveness of MIS compared to OA for PHEO is still controversial [11, 17]. Furthermore, the influence of tumour type and activity on postoperative outcomes has not yet been well established [18]. From 1994 to 2021, 629 patients underwent adrenalec- tomy in two centres (Department of General Surgery and Surgical Specialties, Policlinico Umberto I, Sapienza Uni- versity of Rome, and Department of General Surgery, Uni- versità Politecnica delle Marche, Ancona, Italy). Patients who underwent bilateral [29] and OA were excluded (22 and 35 patients, respectively), leaving 572 patients for inclusion in the study (Fig. 1). Introduction Out of the 572 patients, 114 (19.9%) and 458 (80.1%) patients underwent LA for PHEO and for other types of lesions, respectively (Fig. 1). y Several studies have compared the postoperative out- comes after LA for PHEO versus other types of adrenal lesions [18, 19]. Some authors reported that PHEO and cor- tisol-producing adenoma are associated with more demand- ing surgery and poor postoperative outcomes [19–22]. Other authors reported no differences in postoperative compli- cations based on the type of adrenal lesions [18, 23–25]. However, most of these comparative studies include a rela- tively small number of patients and a paucity of postopera- tive complications [18–21, 23–26]. The aim of the present propensity score-matched (PSM) analysis is to compare the 30-day surgical outcomes of TLA for PHEO compared to other types of adrenal lesions. In all patients, the same medical treatment protocol was followed (which has changed over the years in accord- ance with the guidelines), and they underwent an identi- cal surgical approach (which, on the contrary, has remained unchanged over the years), as previously reported [27–32]. The patients were evaluated together with the endo- crinologist and anaesthesiologist consultants by physical examination and hormonal assessment, which currently includes plasma free or urinary fractionated metanephrines to exclude PHEO, 1 mg overnight dexamethasone suppres- sion test to exclude cortisol excess, aldosterone/renin ratio to exclude primary aldosteronism in case of concomitant hypertension or unexplained hypokalaemia and measure- ment of sex hormones and steroid precursors in case of clinical or imaging features suggestive of adrenocortical carcinoma, as recommended [1, 17]. Magnetic resonance imaging (MRI) and computerized tomography (CT) scans were performed for each patient [30–32]. Scintigraphy with 131I-metabenzilguanidine or 6-18F-fluoro-l-dopa positron Fig. 1 Patient selection Results The patient characteristics and surgical outcomes of the entire series after PSM comparing Group A versus Group B are reported in Table 1. The preoperative treatment also includes administration of normal saline 2000 cc the evening before surgery (500 or 1000 cc in patients with heart or renal failure) to reverse catecholamine-induced blood volume contraction and to pre- vent severe hypotension after tumour removal, as previously reported and as recommended [17, 30–32]. After PSM, two homogeneous groups were obtained, without any statistically significant differences regarding the preoperative characteristics. Blood pressure, heart rate, and blood glucose level mon- itoring with adjustment of associated therapies were per- formed in the immediate postoperative period, as recom- mended [17, 30–32], and these were administered based on close collaboration with the anaesthesiology team. Comparing the postoperative outcomes, no statisti- cally significant differences were observed in median operative time (85 and 90 min in Groups A and B, respec- tively, p = 0.627), conversion rate [6 (5.3%) in each group, p = 1.000], transfusion rate [4 (3.5%) and 3 (2.6%) in Groups A and B, respectively, p = 1.000], complication rate [7 (6.1%) and 9 (7.9%) in Groups A and B, respectively, p = 0.796], median postoperative hospital stay (3.9 and 3.6 days in Groups A and B, respectively, p = 0.110), and mortality rate [1 (0.9%) in each group, p = 1.000]. Statistical analysis emission tomography was performed in any case of equivo- cal CT and MRI or discrepancy between imaging and bio- chemical tests [30–32]. Continuous data are expressed as medians and 95% con- fidence intervals (CIs), while categorical variables are expressed as frequencies and percentages. The Mann–Whit- ney U test and Fisher’s exact test were used for the com- parisons between the groups for continuous and categori- cal variables, respectively. A p value lower than 0.05 was considered statistically significant. Statistical analyses were carried out with SPSS software 22.0 (SPSS Inc., Chicago, Illinois, USA). A total body CT scan was performed for all patients who had PHEO. Last, in all the patients with secreting adrenal lesions, arterial hypertension, chronic heart disease and electrocar- diographic abnormalities, echocardiography was performed to rule out any underlying cardiac disease, according to our endocrinological and anaesthesiological protocols. In the case of PHEO, the goals of perioperative manage- ment have always been blood pressure control and normali- zation of intravascular volume, although in this study, this has been achieved with different protocols over the years [30–32]. To minimize the selection bias and potential confound- ing factors in this retrospective cohort study, a propensity score matching (PSM) analysis 1:1 was performed to match the Group A patients with the Group B patients. A logis- tic regression model was applied to evaluate the propensity score of each patient according to the following covariables: sex, age, BMI, ASA grade, CCI score, lesion side, lesion size, previous abdominal surgery, type of TLA, and asso- ciated procedures. Two well-matched patient cohorts were obtained using a 1:1 nearest neighbour matching algorithm that pairs patients with the closest PS. Currently, all patients with PHEO start preoperative phar- macological preparation with alpha-blockers (doxazosin: starting dose 2 mg/day) 14 days before surgery. If the patient is not affected by hypertension, the patient still starts with a low dose of an alpha-blocker, which is then modulated by the endocrinologists based on the blood pressure levels, and the final dose can be as high as 32 mg/day, according to the current guidelines [17, 30]. Beta-blockers (atenolol: starting dose 25 mg/day, final dose 50 mg/day) are added in case of tachycardia episodes, after at least three days of alpha blockade administration, to avoid hypertensive crises due to the beta blockade without alpha blockade [17, 30]. Materials and methods This study is a retrospective analysis of prospectively col- lected data. Institutional review board approval (number: 1/2022) and informed consent were obtained from all participants. ig. 1 Patient selection 1 3 Journal of Endocrinological Investigation (2023) 46:1589–1596 1591 Study design Patient characteristics [including sex, age and body mass index (BMI)], American Society of Anaesthesiologists (ASA) grade, Charlson comorbidity index (CCI) score [33], previous abdominal surgery, lesion side, type of TLA (anterior or flank approach on the right, anterior, flank or submesocolic retropancreatic approach on the left), associ- ated procedures, conversion rate, operative time, intra- and 30-day postoperative complications (graded according to the Clavien–Dindo classification [34]), adrenal lesion size and histology, hospital stay and 30-day mortality were stored in a Microsoft Excel program (Microsoft Corporation, Redmond, Washington, USA). Concerning the type of TLA, the anterior approach was mostly used in both groups [69 cases (61%) versus 80 cases (70%), p = 0.164], followed by the left submesocolic retro- pancreatic approach [45 cases (40%) versus 33 cases (29%), p = 0.0124], while the flank approach was the least adopted in both groups [0 cases versus 1 case (1%), p = 0.122]. Comparing the incidence of metastatic versus benign lesions at the time of surgery, statistically significant differ- ences were not observed between the two groups. A negative resection margin was achieved in all cases. Journal of Endocrinological Investigation (2023) 46:1589–1596 1593 ASA American Society of Anaesthesiologists, CCI Charlson comorbidity index, TLA transperitoneal lapa- roscopic adrenalectomy, CI confidence intervals Statistically significant differences in bold Group A Group B p value n = 114 (50%) n = 114 (50%) - Non-secreting adenoma – 30 (26.3) < 0.001 - Hyperplasia – 18 (15.8) < 0.001 - Myelolipoma – 10 (8.8) 0.002 - Adrenal cyst – 2 (1.8) 0.450 - Angiomyolipoma – – 1.000 Malignant lesions 13 (11.4) 14 (12.3) 1.000 - Pheochromocytoma 13 (11.4) – < 0.001 - Adrenal carcinoma – 10 (8.8) 0.002 - Adrenal metastases – 4 (3.5) 0.123 Median hospital stay, days (CI 95%) 4 (4.1–5) 4.1 (3.6–4.5) 0.110 Mortality, n (%) 1 (0.9) 1 (0.9) 1.000 ASA American Society of Anaesthesiologists, CCI Charlson comorbidity index, TLA transperitoneal lapa- roscopic adrenalectomy, CI confidence intervals Statistically significant differences in bold i Statistically significant differences in bold An associated surgical procedure was performed in four patients (3.5%) in Group A [cholecystectomy in 3 (2.6%) and pedunculated uterine fibroid resection in 1 (0.9%) patients] and in three patients (2.6%) in Group B [cholecystectomy in 1 (0.9%) and umbilical hernia repair in 2 (1.8%) patients].i caution, and our results are in line with those reported by other authors [7, 11, 24, 38]. In a recent study by Arolfo et al., including 114 patients, MIS has showed to be safe and effective even in cases of PHEOs greater than 5 cm in diameter [38]. A recent meta-analysis by Li et al., includ- ing more than 700 patients, also reported the superiority of LA compared to OA in the case of PHEO in terms of esti- mated blood loss, transfusion rate, haemodynamic instabil- ity, postoperative complication rate, and length of hospital stay [11]. Regardless, this meta-analysis included only one randomized controlled trial, and it compared heterogeneous patient groups, as the tumour size was smaller [11]. How- ever, the BMI was higher in the LA group than in the OA group [11]. Moreover, some preoperative variables, such as previous abdominal surgery and associated procedures, are missing, which might influence the surgical outcomes [11]. In Group A, one (0.9%) human immunodeficiency virus (HIV)-positive, ASA III, male patient, who had undergone a previous coronary artery bypass graft died from an acute myocardial infarction on postoperative Day 4. In Group B, 1 (0.9%) female patient with ASA III died from acute respira- tory failure on postoperative Day 4. Study design 1 3 Journal of Endocrinological Investigation (2023) 46:1589–1596 1592 Group A Group B p value n = 114 (50%) n = 114 (50%) Sex ratio, n (%) - Men 49 (43) 35 (30.7) 0.074 - Women 65 (57) 79 (69.3) Median age, years (CI 95%) 52.5 (49.3–55.1) 53 (50.7–56.4) 0.706 Median body mass index, kg/m2 (CI 95%) 25.5 (24.7–26.4) 26 (26.2–28.2) 0.714 ASA grade, n (%) - I–II 93 (81.6) 90 (78.9) 0.618 - III–IV 21 (18.4) 24 (21.1) CCI score, n (%) - 0–1 61 (53.5) 60 (52.6) 0.894 - ≥ 2 53 (46.5) 54 (47.4) Previous abdominal surgery, n (%) 8 (7) 6 (5.3) 0.784 Lesion side, n (%) - Right 63 (55.3) 70 (61.4) 0.420 - Left 51 (44.7) 44 (38.6) TLA, n (%) - Anterior 69 (60.5) 80 (70.2) 0.164 - Left anterior submesocolic retropancreatic 45 (39.5) 33 (28.9) 0.124 - Flank – 1 (0.9) 0.122 Associated procedures, n (%) 4 (3.5) 5 (4.4) 1.000 Conversion rate, n (%) 6 (5.3) 6 (5.3) 1.000 - Adhesions from previous surgery 1 (0.9) 1 (0.9) 1.000 - Adhesion to pancreas – – 1.000 - Adhesion to liver 1 (0.9) 1 (0.9) 1.000 - Bleeding 2 (1.8) 2 (1.8) 1.000 - Respiratory failure from pneumoperitoneum 1 (0.9) 1 (0.9) 1.000 - Retrocaval mass growth 1 (0.9) 1 (0.9) 1.000 Median operative time, minutes (CI 95%) 85 (88.6–105.9) 90 (92–112.5) 0.627 Postoperative complications, n (%, Clavien–Dindo clas- sification grade) 10 (8.8) 10 (8.8) 1.000 Surgical complications 7 (6.1) 6 (5.3) 1.000 - Acute urinary retention – – 1.000 - Ileus – – 1.000 - Abdominal abscess – – 1.000 - Anaemia 4 (3.5, II) 3 (2.6, II) 1.000 - Fever 1 (0.9, II) 2 (1.8, II) 1.000 - Wound infection 1 (0.9, II) 1 (0.9, II) 1.000 - Colonic fistula – – 1.000 - Chylous ascites 1 (0.9, III-a) – 1.000 - Hemoperitoneum – – 1.000 Medical complications 3 (2.6) 4 (3.5) 1.000 - Pleural effusion 1 (0.9, I) 1 (0.9, I) 1.000 - Pneumonia 2 (1.8, II) 2 (1.8, I) 1.000 - Atrial fibrillation – – 1.000 - Acute myocardial infarction – 1 (0.9, II) 1.000 Median lesion size at definitive histology, cm (CI 95%) 4.3 (4–4.7) 4.1 (3.9–4.8) 0.151 Definitive histology (benign: malignant), n % Benign lesions 101 (88.6) 100 (87.7) 1.000 - Pheochromocytoma 101 (88.6) – < 0.001 - Secreting adenoma – 40 (35.1) < 0.001 Table 1 Results after propensity score matching - Pneumonia - Pheochromocytoma - Secreting adenoma 1 3 Journal of Endocrinological Investigation (2023) 46:1589–1596 Discussion The present PSM analysis demonstrates that TLA for PHEO is associated with the same 30-day surgical outcomes as for other types of lesions, with no statistically significant differences. l It should be emphasized that in this study the most com- monly used approach of TLA was the anterior approach. As reported in the literature [20, 25, 26], early identifica- tion and ligation of the adrenal vein performed prior to any gland manipulation and careful dissection with no adrenal capsule disruption are the recommended strategies to reduce the intraoperative catecholamine release, and this represents one of the main advantages of the anterior and submesocolic TLA [22, 27, 30–32]. f As reported in the literature some preoperative charac- teristics (high BMI, previous abdominal surgery, and large adrenal masses) are related to a higher risk of intra- and postoperative complications [12, 14, 35]. Therefore, a PSM analysis was conducted to balance potential confounders and to obtain two homogeneous groups. As reported in the European Society for Medical Oncol- ogy (ESMO) guidelines, MIS is the gold standard treatment in cases of PHEO up to 6 cm. However, OA is still recom- mended if complete resection cannot be achieved by MIS [1, 17]. Some authors report that LA for PHEO might result in higher intra- and postoperative complication rates, more blood loss, a longer operative time and a longer hospital stay, but these issues are still debated [19, 26, 36, 37]. The present analysis, however, does not support these words of Since there is no clear superiority of one approach over another [39–42], the recommendations of the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) suggest that the surgeon should employ the sur- gical approach for LA that he or she is most familiar with [43]. For this reason, even if TLA by the anterior approach is an uncommon procedure worldwide, in our experience we have continued to use it. For us, it is more familiar in comparison to other approaches, and our results are similar 1 3 1 3 1594 Journal of Endocrinological Investigation (2023) 46:1589–1596 to those reported in the literature by other authors using the lateral or posterior approaches in terms of operative time, conversion rates and morbidity rates [22, 27, 30–32, 39–41]. approval for publication. LP: data collection, data interpretation, criti- cal revision, approval for publication. IS: data analysis, data interpreta- tion, critical revision, approval for publication. Discussion MO: data collection, data interpretation, critical revision, approval for publication. MG: data interpretation, critical revision, approval for publication. AMP: concept and design, data interpretation, manuscript writing, critical revision, overall supervision, approval for publication. These data suggest that, regardless of the adopted approach, the fundamental requirements are adequate sur- gical experience in advanced MIS and multidisciplinary patient management in high-volume centres [1, 44–50]. Funding Open access funding provided by Università degli Studi di Roma La Sapienza within the CRUI-CARE Agreement. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Our study suggests that TLA for PHEO using the trans- peritoneal anterior and submesocolic approach is as safe and effective as TLA performed for other types of tumours. This further supports the concept that other preoperative parameters are probably useful to predict the difficulty of laparoscopic adrenalectomy, such as sex, previous surgery, BMI, site and size of the lesion, associated procedures, and comorbidities [12, 35, 37, 51–53]. Data availability The datasets generated analysed during the cur- rent study are available from the corresponding author on reasonable request. Declarations The main limitations of the present study are its retro- spective nature and the lack of data regarding intraoperative monitoring values for blood pressure, heart rate and glu- cose in all patients. The long study period also represents a limitation as it involves a certain heterogeneity of the study population, since the recommendations in the management of patients with adrenal lesions have changed considerably over the years [17, 53–56]. The diagnosis and management of patients with adrenal lesions differed between the patients treated in the first and last years of the present study (e.g. the diagnosis of PHEO was initially based on catechola- mines and vanillylmandelic acid, while currently, it is based on urinary or plasma metanephrines [17, 55, 56]. Also, our perioperative medical management has undergone changes over the years [13, 22, 30–32]). However, this heterogeneity concerns both the study and the control group; moreover, we did not find differences with regard to the surgical approach over the years which is the real purpose of the study. Fur- thermore, to the best of our knowledge, the present study is the first and largest one comparing the surgical outcomes of LA performed for PHEO versus other types of lesions, and this study used PSM analysis to account for potential confounders. Conflict of interest Diletta Corallino, Andrea Balla, Livia Palmieri, Isabella Sperduti, Monica Ortenzi, Mario Guerrieri and Alessandro M. Paganini declare that they have no conflict of interest or financial ties to disclose. Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the insti- tutional national research committee and with the 1964 Helsinki Dec- laration. Informed consent Informed consent was obtained from all participants. Open Access This article is licensed under a Creative Commons Attri- bution 4.0 International License, which permits use, sharing, adapta- tion, 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/. In conclusion, based on the present study, the surgical outcomes of TLA for PHEO are equivalent to those of TLA for other types of tumours with the same preoperative char- acteristics. Adequate surgical experience in advanced MIS and disease-specific LA, coupled with a multidisciplinary patient management including consultations with the endo- crinologist and the anaesthesiologist, are essential require- ments. Further prospective randomized controlled trials with a larger number of patients are required to draw definitive conclusions. Author contributions DC: concept and design, data collection, data analysis, data interpretation, manuscript writing, critical revision, over- all supervision, approval for publication. AB: concept and design, data collection, data interpretation, manuscript writing, critical revision, Research involving human participants and/or animals Not applicable. Informed consent Informed consent was obtained from all participants. References 1. Fassnacht M, Assie G, Baudin E, Eisenhofer G, de la Fouchardiere C, Haak HR, de Krijger R, Porpiglia F, Terzolo M, Berruti A, ESMO Guidelines Committee (2020) Adrenocortical carcinomas and malignant phaeochromocytomas: ESMO-EURACAN Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 31(11):1476–1490. https://doi.org/10.1016/j.annonc.2020. 08.2099 2. Ma W, Mao Y, Zhuo R, Dai J, Fang C, Wang C, Zhao J, He W, Zhu Y, Xu D, Sun F (2020) Surgical outcomes of a randomized controlled trial compared robotic versus laparoscopic adrenalec- tomy for pheochromocytoma. Eur J Surg Oncol 46(10 Pt A):1843– 1847. https://doi.org/10.1016/j.ejso.2020.04.001 Author contributions DC: concept and design, data collection, data analysis, data interpretation, manuscript writing, critical revision, over- all supervision, approval for publication. AB: concept and design, data collection, data interpretation, manuscript writing, critical revision, 3. Picchetto A, Paganini AM, Balla A, Quaresima S, Canti- sani V, D’Ambrosio G, Lezoche E (2014) Rare extra-adrenal 1 3 1595 Journal of Endocrinological Investigation (2023) 46:1589–1596 paraganglioma mimicking a painful Schwannoma: case report. Ann Ital Chir 85(ePub):S2239253X14022907 18. Limberg J, Stefanova D, Ullmann TM, Thiesmeyer JW, Buicko JL, Finnerty BM, Zarnegar R, Fahey TJ 3rd, Beninato T (2021) Not all laparoscopic adrenalectomies are equal: analysis of post- operative outcomes based on tumor functionality. Surg Endosc 35(6):2601–2606. https://doi.org/10.1007/s00464-020-07678-2 4. Bihain F, Klein M, Nomine-Criqui C, Brunaud L (2020) Robotic adrenalectomy in patients with pheochromocytoma: a system- atic review. Gland Surg 9(3):844–848. https://doi.org/10.21037/ gs-2019-ra-05 19. Gotoh M, Ono Y, Hattori R, Kinukawa T, Ohshima S (2002) Laparoscopic adrenalectomy for pheochromocytoma: morbidity compared with adrenalectomy for tumors of other pathology. J Endourol 16(4):245–9. https://doi.org/10.1089/089277902753752 223 5. Bai S, Wu B, Yao Z, Zhu X, Jiang Y, Wang H (2020) Development and validation of a clinical model to predict intraoperative hemo- dynamic instability in patients with pheochromocytomas surgery. Endocr J 67(1):81–89. https://doi.org/10.1507/endocrj.EJ19-0278 20. Kalady MF, McKinlay R, Olson JA Jr, Pinheiro J, Lagoo S, Park A, Eubanks WS (2004) Laparoscopic adrenalectomy for pheochromocytoma. A comparison to aldosteronoma and inci- dentaloma. Surg Endosc 18(4):621–5. https://doi.org/10.1007/ s00464-003-8827-0 p g j 6. Lenders JW, Eisenhofer G, Mannelli M, Pacak K (2005) Phaeo- chromocytoma. Lancet 366(9486):665–75. https://doi.org/10. 1016/S0140-6736(05)67139-5 7. Fu SQ, Wang SY, Chen Q, Liu YT, Li ZL, Sun T (2020) Laparoscopic versus open surgery for pheochromocytoma: a meta-analysis. BMC Surg 20(1):167. https://doi.org/10.1186/ s12893-020-00824-6 21. References Chen Y, Scholten A, Chomsky-Higgins K, Nwaogu I, Gosnell JE, Seib C, Shen WT, Suh I, Duh QY (2018) Risk factors associated with perioperative complications and prolonged length of stay after laparoscopic adrenalectomy. JAMA Surg 153(11):1036– 1041. https://doi.org/10.1001/jamasurg.2018.2648 8. Fang AM, Rosen J, Saidian A, Bae S, Tanno FY, Chambo JL, Bloom J, Gordetsky J, Srougi V, Phillips J, Rais-Bahrami S (2020) Perioperative outcomes of laparoscopic, robotic, and open approaches to pheochromocytoma. J Robot Surg 14(6):849–854. https://doi.org/10.1007/s11701-020-01056-9 22. Balla A, Quaresima S, Palmieri L, Ortenzi M, Sbardella E, Puliani G, Isidori AM, Guerrieri M, Paganini AM (2019) Is laparoscopic left adrenalectomy with the anterior submesocolic approach for Conn’s or Cushing’s syndrome equally safe and effective as the lateral and anterior ones? Surg Endosc 33(9):3026–3033. https:// doi.org/10.1007/s00464-018-6601-6 p g 9. Kiernan CM, Du L, Chen X, Broome JT, Shi C, Peters MF, Solor- zano CC (2014) Predictors of hemodynamic instability during surgery for pheochromocytoma. Ann Surg Oncol 21(12):3865–71. https://doi.org/10.1245/s10434-014-3847-7 p g 10. Gagner M, Lacroix A, Bolté E (1992) Laparoscopic adrenalec- tomy in Cushing’s syndrome and pheochromocytoma. N Engl J Med 327(14):1033. https://doi.org/10.1056/NEJM19921001327 1417 23. Kiernan CM, Shinall MC Jr, Mendez W, Peters MF, Broome JT, Solorzano CC (2014) Influence of adrenal pathology on perioperative outcomes: a multi-institutional analysis. Am J Surg 208(4):619–25. https://doi.org/10.1016/j.amjsurg.2014.06.002 24. Toniato A, Boschin I, Bernante P, Opocher G, Guolo AM, Pelizzo MR, Mantero F (2007) Laparoscopic adrenalectomy for pheochro- mocytoma: is it really more difficult? Surg Endosc 21(8):1323–6. https://doi.org/10.1007/s00464-006-9190-8 11. Li J, Wang Y, Chang X, Han Z (2020) Laparoscopic adrenalec- tomy (LA) vs open adrenalectomy (OA) for pheochromocytoma (PHEO): a systematic review and meta-analysis. Eur J Surg Oncol 46(6):991–998. https://doi.org/10.1016/j.ejso.2020.02.009 p g j j 12. Alberici L, Paganini AM, Ricci C, Balla A, Ballarini Z, Ortenzi M, Casole G, Quaresima S, Di Dalmazi G, Ursi P, Alfano MS, Selva S, Casadei R, Ingaldi C, Lezoche G, Guerrieri M, Minni F, Tiberio GAM (2022) Development and validation of a pre- operative “difficulty score” for laparoscopic transabdominal adrenalectomy: a multicenter retrospective study. Surg Endosc 36(5):3549–3557. https://doi.org/10.1007/s00464-021-08678-6 25. Mellon MJ, Sundaram CP (2008) Laparoscopic adrenalectomy for pheochromocytoma versus other surgical indications. JSLS 12(4):380–384 26. Natkaniec M, Pędziwiatr M, Wierdak M, Białas M, Major P, Matłok M, Budzyński P, Dworak J, Buziak-Bereza M, Budzyński A (2015) Laparoscopic adrenalectomy for pheochromocytoma is more difficult compared to other adrenal tumors. Wideochir Inne Tech Maloinwazyjne 10(3):466–71. https://doi.org/10.5114/wiitm. 2015.52869 13. References Balla A, Palmieri L, Meoli F, Corallino D, Ortenzi M, Ursi P, Guerrieri M, Quaresima S, Paganini AM (2020) Are adrenal lesions of 6 cm or more in diameter a contraindication to lapa- roscopic adrenalectomy? A case-control study. World J Surg 44(3):810–818. https://doi.org/10.1007/s00268-019-05287-2 27. Scoglio D, Balla A, Paci M, Guerrieri M, Lezoche G, D’Ambrosio G, Fabiani B, Ursi P, Paganini AM (2013) Laparoscopic transperi- toneal anterior adrenalectomy. Ann Ital Chir 84(4):411–416 14. Balla A, Corallino D, Ortenzi M, Palmieri L, Meoli F, Guerri- eri M, Paganini AM (2022) Cancer risk in adrenalectomy: are adrenal lesions equal or more than 4 cm a contraindication for laparoscopy? Surg Endosc 36(2):1131–1142. https://doi.org/10. 1007/s00464-021-08380-7 28. Ortenzi M, Balla A, Ghiselli R, Vergari R, Silecchia G, Guerri- eri E, Maria Paganini A, Guerrieri M (2019) Minimally invasive approach to the adrenal gland in obese patients with Cushing’s syndrome. Minim Invasive Ther Allied Technol 28(5):285–291. https://doi.org/10.1080/13645706.2018.1536669 15. Bai S, Yao Z, Zhu X, Li Z, Jiang Y, Wang R, Wu B (2019) Com- parison of transperitoneal laparoscopic versus open adrenalec- tomy for large pheochromocytoma: a retrospective propensity score-matched cohort study. Int J Surg 61:26–32. https://doi.org/ 10.1016/j.ijsu.2018.11.018 29. Balla A, Ortenzi M, Palmieri L, Corallino D, Meoli F, Ursi P, Puliani G, Sbardella E, Isidori AM, Guerrieri M, Quaresima S, Paganini AM (2019) Laparoscopic bilateral anterior trans- peritoneal adrenalectomy: 24 years experience. Surg Endosc 33(11):3718–3724. https://doi.org/10.1007/s00464-019-06665-6 30. Paganini AM, Balla A, Guerrieri M, Lezoche G, Campagnacci R, D’Ambrosio G, Quaresima S, Antonica MV, Lezoche E (2014) Laparoscopic transperitoneal anterior adrenalectomy in pheochro- mocytoma: experience in 62 patients. Surg Endosc 28(9):2683–9. https://doi.org/10.1007/s00464-014-3528-4 16. Zhu W, Wang S, Du G, Liu H, Lu J, Yang W (2019) Compari- son of retroperitoneal laparoscopic versus open adrenalectomy for large pheochromocytoma: a single-center retrospective study. World J Surg Oncol 17(1):111. https://doi.org/10.1186/ s12957-019-1649-x 17. Lenders JW, Duh QY, Eisenhofer G, Gimenez-Roqueplo AP, Grebe SK, Murad MH, Naruse M, Pacak K, Young WF Jr; Endo- crine Society (2014) Pheochromocytoma and paraganglioma: an endocrine society clinical practice guideline. J Clin Endocrinol Metab 99(6):1915–42. https://doi.org/10.1210/jc.2014-1498 31. Balla A, Quaresima S, Ortenzi M, Palmieri L, Meoli F, Coral- lino D, Guerrieri M, Ursi P, Paganini AM (2019) Results after laparoscopic left anterior transperitoneal submesocolic adrenal- ectomy for the treatment of pheochromocytoma. Ann Ital Chir 90:220–224 1 3 3 1596 Journal of Endocrinological Investigation (2023) 46:1589–1596 volume or specialty? Arch Surg 144(11):1060–7. https://doi.org/ 10.1001/archsurg.2009.191 32. References Paganini AM, Guerrieri M, Balla A, Quaresima S, Isidori AM, Iafrate F, D’Ambrosio G, Lezoche G, Lezoche E (2016) Manage- ment of adrenal incidentaloma by laparoscopic transperitoneal anterior and submesocolic approach. Langenbecks Arch Surg 401(1):71–9. https://doi.org/10.1007/s00423-015-1367-y 45. Bergamini C, Martellucci J, Tozzi F, Valeri A (2011) Compli- cations in laparoscopic adrenalectomy: the value of experi- ence. Surg Endosc 25(12):3845–51. https://doi.org/10.1007/ s00464-011-1804-0 p g y 33. Charlson ME, Pompei P, Ales KL, MacKenzie CR (1987) A new method of classifying prognostic comorbidity in longitudinal stud- ies: development and validation. J Chronic Dis 40(5):373–383. https://doi.org/10.1016/0021-9681(87)90171-8 46. Kazaure HS, Sosa JA (2019) Volume-outcome relationship in adrenal surgery: a review of existing literature. Best Pract Res Clin Endocrinol Metab 33(5):101296. https://doi.org/10.1016/j. beem.2019.101296 p g 34. Clavien PA, Barkun J, de Oliveira ML, Vauthey JN, Dindo D, Schulick RD, de Santibañes E, Pekolj J, Slankamenac K, Bassi C, Graf R, Vonlanthen R, Padbury R, Cameron JL, Makuuchi M (2009) The Clavien–Dindo classification of surgical complica- tions: five-year experience. Ann Surg 250(2):187–96. https://doi. org/10.1097/SLA.0b013e3181b13ca2 47. Kercher KW, Novitsky YW, Park A, Matthews BD, Litwin DE, Heniford BT (2005) Laparoscopic curative resection of pheochro- mocytomas. Ann Surg 241(6):919–926. https://doi.org/10.1097/ 01.sla.0000164175.26785.06 48. Yip L, Duh QY, Wachtel H, Jimenez C, Sturgeon C, Lee C, Velázquez-Fernández D, Berber E, Hammer GD, Bancos I, Lee JA, Marko J, Morris-Wiseman LF, Hughes MS, Livhits MJ, Han MA, Smith PW, Wilhelm S, Asa SL, Fahey TJ 3rd, McKenzie TJ, Strong VE, Perrier ND (2022) American association of endo- crine surgeons guidelines for adrenalectomy: executive summary. JAMA Surg. https://doi.org/10.1001/jamasurg.2022.3544 35. Srougi V, Barbosa JAB, Massaud I, Cavalcante IP, Tanno FY, Almeida MQ, Srougi M, Fragoso MC, Chambô JL (2019) Pre- dictors of complication after adrenalectomy. Int Braz J Urol 45(3):514–522. https://doi.org/10.1590/S1677-5538.IBJU.2018. 0482 36. Parikh PP, Rubio GA, Farra JC, Lew JI (2017) Nationwide review of hormonally active adrenal tumors highlights high morbidity in pheochromocytoma. J Surg Res 215:204–210. https://doi.org/10. 1016/j.jss.2017.04.011 49. Adrenal (MTG1) Endo-ERN. https://endo-ern.eu/it/competenze- specifiche/ghiandola-surrenale-mtg1/. Accessed 4 Nov 2022 i 50. ENS@T European network for the study of adrenal tumors. http:// www.ensat.org/ens@t-centers-of-excellence. Accessed 4 Nov 2022 j j 37. Tiberio GA, Solaini L, Arru L, Merigo G, Baiocchi GL, Giulini SM (2013) Factors influencing outcomes in laparoscopic adrenal surgery. Langenbecks Arch Surg 398(5):735–43. https://doi.org/ 10.1007/s00423-013-1082-5 51. Wang H, Wu B, Yao Z, Zhu X, Jiang Y, Bai S (2020) Nomo- gram for predicting severe morbidity after pheochromocytoma surgery. Endocr Connect 9(4):309–317. References https://doi.org/10.1530/ EC-20-0004 38. Arolfo S, Giraudo G, Franco C, Parasiliti Caprino M, Seno E, Morino M (2022) Minimally invasive adrenalectomy for large pheochromocytoma: not recommendable yet? Results from a sin- gle institution case series. Langenbecks Arch Surg 407(1):277– 283. https://doi.org/10.1007/s00423-021-02312-8 52. Inversini D, Manfredini L, Galli F, Zhang D, Dionigi G, Rausei S (2020) Risk factors for complications after robotic adrenalectomy: a review. Gland Surg 9(3):826–830. https://doi.org/10.21037/gs. 2020.04.10 39. Nigri G, Rosman AS, Petrucciani N, Fancellu A, Pisano M, Zor- colo L, Ramacciato G, Melis M (2013) Meta-analysis of trials comparing laparoscopic transperitoneal and retroperitoneal adre- nalectomy. Surgery 153(1):111–119. https://doi.org/10.1016/j. surg.2012.05.042 53. Rodríguez-Hermosa JI, Planellas-Giné P, Cornejo L, Gironès J, Recasens M, Ortega FJ, Moreno-Navarrete JM, Latorre J, Fernan- dez-Real JM, Codina-Cazador A (2021) Comparison of outcomes between obese and nonobese patients in laparoscopic adrenalec- tomy: a cohort study. Dig Surg 38(3):237–246. https://doi.org/10. 1159/000515589 g 40. Jiang YL, Qian LJ, Li Z, Wang KE, Zhou XL, Zhou J, Ye CH (2020) Comparison of the retroperitoneal versus transperitoneal laparoscopic adrenalectomy perioperative outcomes and safety for pheochromocytoma: a meta-analysis. BMC Surg 20(1):12. https:// doi.org/10.1186/s12893-020-0676-4 54. Chen H, Sippel RS, O’Dorisio MS, Vinik AI, Lloyd RV, Pacak K, North American Neuroendocrine Tumor Society (NANETS) (2010) The North American Neuroendocrine Tumor Society con- sensus guideline for the diagnosis and management of neuroendo- crine tumors: pheochromocytoma, paraganglioma, and medullary thyroid cancer. Pancreas 39(6):775–783. https://doi.org/10.1097/ MPA.0b013e3181ebb4f0f g 41. Gavriilidis P, Camenzuli C, Paspala A, Di Marco AN, Palazzo FF (2021) Posterior retroperitoneoscopic versus laparoscopic transperitoneal adrenalectomy: a systematic review by an updated meta-analysis. World J Surg 45(1):168–179. https://doi.org/10. 1007/s00268-020-05759-w 55. Manger WM, Gifford RW Jr (1993) Pheochromocytoma: current diagnosis and management. Clevel Clin J Med 60(5):365–378. https://doi.org/10.3949/ccjm.60.5.365 42. Cardinali L, Skrami E, Catani E, Carle F, Ortenzi M, Balla A, Guerrieri M (2021) Laparoscopic transperitoneal adrenalec- tomy: a comparative study of different techniques for vessel sealing. Surg Endosc 35(2):673–683. https://doi.org/10.1007/ s00464-020-07432-8 56. Manu P, Runge LA (1984) Biochemical screening for pheochro- mocytoma. Superiority of urinary metanephrines measurements. Am J Epidemiol 120(5):788–790. https://doi.org/10.1093/oxfor djournals.aje.a113947 43. Stefanidis D, Goldfarb M, Kercher KW, Hope WW, Richardson W, Fanelli RD; Society of Gastrointestinal and Endoscopic Sur- geons (2013) SAGES guidelines for minimally invasive treatment of adrenal pathology. Surg Endosc 27(11):3960–80. https://doi. org/10.1007/s00464-013-3169-z 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. References Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 44. Park HS, Roman SA, Sosa JA (2009) Outcomes from 3144 adre- nalectomies in the United States: which matters more, surgeon 1 3
https://openalex.org/W4382792037	https://link.springer.com/content/pdf/10.1007/s44154-023-00092-3.pdf	English	null	Novel discovery in roles of structural variations and RWP-RK transcription factors in heat tolerance for pearl millet	Stress biology	2,023	cc-by	3,594	© 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/. Introduction Global warming with elevated temperatures poses a major threat to food security worldwide (Zhao et al. 2017). Developing heat-tolerant crops is crucial to ensure agricultural production and meet the increasing food demands of a growing population. A thorough under- standing of plant heat tolerance mechanisms is essential to improve stress adaptation and mitigate yield loss due to increasing temperatures. Correspondence: Bingru Huang huang@sebs.rutgers.edu 1 Department of Plant Biology, Rutgers University, New Brunswick, NJ 08901, USA 2 Department of Genetics, University of Georgia, Athens, GA, USA 3 College of Grassland Science and Technology, Sichuan Agricultural University, Chengdu, China Correspondence: Bingru Huang huang@sebs.rutgers.edu 1 Department of Plant Biology, Rutgers University, New Brunswick, NJ 08901, USA 2 Department of Genetics, University of Georgia, Athens, GA, USA 3 College of Grassland Science and Technology, Sichuan Agricultural University, Chengdu, China Abstract Global warming adversely affects crop production worldwide. Massive efforts have been undertaken to study mecha- nisms regulating heat tolerance in plants. However, the roles of structural variations (SVs) in heat stress tolerance remain unclear. In a recent article, Yan et al. (Nat Genet 1–12, 2023) constructed the first pan-genome of pearl millet (Pennisetum glaucum) and identified key SVs linked to genes involved in regulating plant tolerance to heat stress for an important crop with a superior ability to thrive in extremely hot and arid climates. Through multi-omics analy- ses integrating by pan-genomics, comparative genomics, transcriptomics, population genetics and and molecular biological technologies, they found RWP-RK transcription factors cooperating with endoplasmic reticulum-related genes play key roles in heat tolerance in pearl millet. The results in this paper provided novel insights to advance the understanding of the genetic and genomic basis of heat tolerance and an exceptional resource for molecular breed- ing to improve heat tolerance in pearl millet and other crops. Keywords Heat stress, Pearl millet, Structural variations, RWP-RK Approaches to unraveling stress tolerance mechanisms include analyzing the natural variation of a broader range of plants for sequence variations showing rapid changes in response to heat stress (Verslues et al. 2023) and inte- grating transcriptomics, comparative genomics, popula- tion genetics, and molecular biological technology (Yan et al. 2023). Structural variations (SVs) indicate large genomic alternations, which contribute to the plant’s phenotypic changes and shape the responses to environ- ments. Recent studies have been carried out to reveal the roles of SVs in the responsiveness of stressful conditions in plants. Fuentes et al. (2019) found the stress response genes were enriched on the rice genomic regions with frequent SVs. Bayer et al. (2019) and Dolatabadian et al. (2020) revealed that disease-resistance genes show diverse SV patterns among different Brassica accessions which seems to be a common feature of plant pange- nomes. Pan-genomes are becoming necessary to identify the SVs through constructing multiple reference-quality genome assemblies via identifying loci with variants Open Access Open Access Novel discovery in roles of structural variations and RWP‑RK transcription factors in heat tolerance for pearl millet Bingru Huang1, Haidong Yan2,3, Min Sun3 and Yarong Jin3 Bingru Huang1, Haidong Yan2,3, Min Sun3 and Yarong Jin3 Stress Biology Stress Biology Stress Biology Huang et al. Stress Biology (2023) 3:12 https://doi.org/10.1007/s44154-023-00092-3 Page 2 of 5 Huang et al. Stress Biology (2023) 3:12 Huang et al. Stress Biology were repetitive sequences (Varshney et al. 2017), which increases difficulties to assemble genomes. Yan et al. (2023) utilized a recent technology called PacBio HiFi sequencing with accuracy of 99.9% similar to accuracy via short read and sanger sequencing and have a long read comparable to traditional long read sequencing (Fig. 1b). The authors utilized different approaches to evaluate the assemble qualities per genome on basis of different evaluation scores to prove each assembly having a good quality. They also evaluated the representative- ness of the pan-genome, and compared the distribution of SNPs between the 11 accessions and the published 394 core lines. The results showed a similar pattern across the genome and showed strong significant correlations in SNP density, nucleotide diversity and synonymous and nonsynonymous substitution rates. These results indicate the constructed pan-genome are representative of the diversities of the pearl millet population, providing high- quality pan-genome resources for downstream analyses. represented by alternative sequences. Recent studies have constructed pan-genomes in major crops, such as maize (Zea mays), rice (Oryza sativa), tomato (Solanum lycopersicum), and soybean (Glycine max) (Wang et al. 2018, 2023; Liu et al. 2020), and provided a number of candidate SVs and their surrounding genes linking to important agricultural traits, but the roles of SVs in heat tolerance are un-characterized.hi This article highlights the novel findings reported in a recent publication by Yan et al. (2023) which constructed the graph-based pan-genome in pearl millet and revealed the critical roles of SVs contributing to heat tolerance in pearl millet. They used a graph-based pan-genome anal- ysis of pearl millet to identify genetic variations associ- ated with heat stress adaptation and domestication. The information from this article is very useful for genetic improvement and molecular breeding to develop heat- tolerant germplasms of other crops, particularly grass species. Pearl millet is the sixth most important cereal crop after rice, wheat (Triticum aestivum), maize, barley (Hordeum vulgare) and sorghum (Sorghum bicolor). It is grown in 30 million ha area in the semi-arid and arid tropical regions of Asia and Africa and is a staple food for more than 90 million poor people. Pearl millet is consid- ered as an ideal model crop to study abiotic stress due to its unique characteristics, such as the following: 1) It is a C4 plant species with high photosynthetic efficiency; 2) It can thrive and maintain high and productivity in a wide range of adverse environments, particularly hot and dry conditions. This species can endure high temperature up to 42℃ during reproductive stage (Satyavathi et al. 2021). Recent study have revealed potential molecular mecha- nisms underlying abiotic stresses (Sun et al. 2020; Khan et al. 2023); however, gene resources of pearl millet are still lacking, and majority of excellent genes or regulatory elements responsive to abiotic stress are unexploited. Pan-genomes were shown to be applied to iden- tify genetic variations, gene expression patterns, and gene functions. For instance, a pan-genome of 26 soy- bean accessions was constructed to identify the genetic basis of domestication and adaptation (Liu et al. 2020). Another study on maize pan-genome demonstrated that SVs accounted for a significant proportion of the genetic variation, and identified genes related to flowering time, kernel weight, and starch synthesis (Wang et al. 2023). Yan et al. (2023) identified 4,411 frequency-differentiated structural variations (fdSVs) between tropical and tem- perate accessions associated with 269 selection sweep regions and 1,471 genes. They also identified 3,952 fdSVs and 1,285 genes associated with domestication. The researchers next conducted a genome-wide association study to examine the associations of the SVs and SNPs with grain number and identified 142 SVs associated with 20 traits (Fig. 1c). Moreover, they analyzed the tiller trait (tiller number/plant) and found four significantly asso- ciated SVs near six genes that were not identified based on the SNP data, revealing additional hidden genetic variations that may not be represented by SNPs. These results used several examples to show potential utiliza- tions of the pan-genome that could be used to advance our understanding of contributions of SVs to plant stress adaption and domestication.i Yan et al. (2023) successfully constructed a pan-genome of pearl millet, including 10 new high-quality genomes and one previously released genome (Fig. 1a, b), and identified 744,364 SVs, most of which were presence and absence variations (PAVs; 622,584), following by 91,852 copy number variations (CNVs), 27,751 translocations (TRANSs), and 2,177 inversions (INVs) (Fig. 1c). The SVs were further categorized into private SVs that were pre- sent in only one accession, and the PAVs accounted for 74.70% of private SVs but constituted relatively high pro- portion (87.51%) of the non-private SVs. Similar trends were detected in CNVs and TRANSs. The quality of pan-genome relies on assembly of each genome as the low quality-assemblies with incorrect orders of contigs of repeats would introduce false positive SVs. Pearl mil- let has a complex genome with more than 70% sequences Yan et al. (2023) further clarified molecular mecha- nisms underlying the heat tolerance by integrating tran- scriptomics, comparative genomics, population genetics, and molecular biological technology. They evaluated heat tolerance of pearl millet and performed RNA sequencing (RNA-seq) under heat stress on two organs at eight-time points and selected six accessions to perform the RNA- seq under two-time points (Fig. 1d, e). They found that Page 3 of 5 Huang et al. Stress Biology (2023) 3:12 Huang et al. Stress Biology Fig. 1 A schematic model exhibiting main results in Yan et al (2023) study. a Ten pearl millet accessions are derived from geographically representative regions. b PacBio HiFi sequencing was applied to build high quality assemblies that were further used to build a graph-based pan-genome. c Five types of SVs were identified in the pan-genome and they could be used as markers linking important agricultural traits. d Evaluate heat tolerance of pearl millet based on phenotype and physiology indicators. e Transcriptome analysis revealed differential expressed genes (DEGs) involved in endoplasmic reticulum (ER) pathways. f RWP-RK TF family was identified to be expanded compared to other species and a transgenic analysis was applied to reveal contribution of RWP-RK to heat tolerance. g SVs were found to contribute to heat tolerance adapation and domestication through selection sweep analysis between two accessions groups located in tropical and temperature zones and between cultivar and wild accessions h A pipeline was developed to filter focal SVs potentially affecting nearby gene expression which was further confirmed Fig. 1 A schematic model exhibiting main results in Yan et al (2023) study. a Ten pearl millet accessions are derived from geographically representative regions. b PacBio HiFi sequencing was applied to build high quality assemblies that were further used to build a graph-based pan-genome. c Five types of SVs were identified in the pan-genome and they could be used as markers linking important agricultural traits. d Evaluate heat tolerance of pearl millet based on phenotype and physiology indicators. e Transcriptome analysis revealed differential expressed genes (DEGs) involved in endoplasmic reticulum (ER) pathways. f RWP-RK TF family was identified to be expanded compared to other species and a transgenic analysis was applied to reveal contribution of RWP-RK to heat tolerance. g SVs were found to contribute to heat tolerance adapation and domestication through selection sweep analysis between two accessions groups located in tropical and temperature zones and between cultivar and wild accessions. h A pipeline was developed to filter focal SVs potentially affecting nearby gene expression which was further confirmed through a transient gene expression experiment in tobacco i d i d f hi ll Fig. 1 A schematic model exhibiting main results in Yan et al (2023) study. a Ten pearl millet accessions are derived from geographically representative regions. b PacBio HiFi sequencing was applied to build high quality assemblies that were further used to build a graph-based pan-genome. c Five types of SVs were identified in the pan-genome and they could be used as markers linking important agricultural traits. d Evaluate heat tolerance of pearl millet based on phenotype and physiology indicators. e Transcriptome analysis revealed differential expressed genes (DEGs) involved in endoplasmic reticulum (ER) pathways. f RWP-RK TF family was identified to be expanded compared to other species and a transgenic analysis was applied to reveal contribution of RWP-RK to heat tolerance. g SVs were found to contribute to heat tolerance adapation and domestication through selection sweep analysis between two accessions groups located in tropical and temperature zones and between cultivar and wild accessions. h A pipeline was developed to filter focal SVs potentially affecting nearby gene expression which was further confirmed through a transient gene expression experiment in tobacco elimination of misfolded proteins (Zhu 2016). The protein folding process in the ER is a necessary stress response pathway that can repair uncorrected protein folding differentially expressed genes from the two transcrip- tome data set were mainly enriched in endoplasmic retic- ulum (ER) related pathways involved in the repair and Huang et al. Stress Biology (2023) 3:12 Huang et al. Stress Biology (2 Page 4 of 5 revealed the SVs potentially particicpate in important processes in the ER system, extending our understanding of roles of SVs underlying heat tolerance.i caused by high temperature stress (Liu et al. 2012). High temperature induced misfolding peptide chains are rec- ognized and processed by three calnexin and calreticulin, and the genes encoding them are upregulated and differ- entially expressed in the transcriptome data reported in Yan et al. (2023). These results highlight an essential role of ER system in defense of the heat resistance. In summary, Yan et al. (2023) developed the first pan- genome and identified key SVs affecting expression of their nearby heat-responsive genes linked to heat toler- ance and domestication for pearl millet. Through an integrated approach, they uncovered a novel TF family RWP-RK that plays an essential role in plant tolerance to heat stress, and functionally validated one member in the major crop rice, enabling improved heat tolerance of the transgenic plant. The results of this study emphasized roles of SVs in abiotic stresses and represent a major advancement in understanding SV-environment interac- tions and shed light on the molecular mechanisms under- lying heat tolerance in pearl millet. The novel findings of RWP-RK provides insights and genetic and genomic basis of genetic modification and molecular breeding for heat-tolerant crops that are better able to withstand high temperatures, which are becoming increasingly impor- tant as the climate changes. y Through comparative genomic analyses, Yan et al. (2023) identified an expanding RWP-RK transcription factor family (Fig. 1f), and found that early LTR expan- sion may have contributed to the expansion of this tran- scription factor family in pearl millet. To confirm the role of RWP-RK genes in heat tolerance, they overex- pressed one of these genes (PMF0G00024.1) in rice and found that the resulting transgenic plants were more tolerant to high temperatures, as manifested by higher levels of antioxidant enzymes (peroxidase and superox- ide dismutase) and lower levels of damage-indicating molecules (malondialdehyde) in the transgenic plants than wild-type plants. The researchers further revealed potential associations between the RWP-RKs and the ER- related genes and used transient co-expression to prove the PMF0G00024.1 could transactivate two important ER-related genes. These findings provide a new clue to indicate potential roles of RWP-RK during heat stress responsiveness, and experimentally confirmed they could interact ER-related genes. The RWP-RKs are also reported mainly involved in nitrate starvation responses (Amin et al. 2023), indicating the RWP-RKs might be a bridge to connect heat tolerance and nitrate signal trans- duction pathways, and further studies need to be done to uncovered this relationship. Perspectiveh The availability of the pearl millet genome resource pro- vides a valuable tool for advancing crop improvement efforts in the face of increasing climate change-induced heat stress. This study has provided massive candi- date heat stress-related genes and potential nearby SVs affecting these genes, which assist community to clone key genes in pearl millet or contribute to other major crop improvement through genetic engineering. Also, the genome resource is applicable to be a reference for mutant library in pearl millet. By comparing the genomes or re-sequencing data of the mutant plants with desirable traits to the pan-genome, researchers can identify spe- cific mutations particularly for SVs that are responsible for the observed phenotypes. This application is achiev- able in any pearl millet panel through mapping their whole-genome sequences to the pan-genome to identify SV-phenotype links. In summary, the availablility of the pearl millet genome sequence opens up new avenues for research aimed at improving heat tolerance in this important crop. Yan et al. (2023) further examined whether the SVs were involved in stress repair in the ER system. Through selection sweep analysis, this study genotyped SVs by mapping all of the re-sequences to the graph-based pan- genome and focused on SVs with population frequency differences (fdSVs) between accessions from tropical and temperate zones by applying a sliding window methodol- ogy. In total, 269 selection sweep regions harboring 4,411 fdSVs surrounding 1,471 genes were identified. And 27 of these genes were significantly and functionally annotated as belonging to ER-related pathways (Fig. 1g). In addition, this study developed a pipeline to systematically identify and validated several focal SVs associated with heat- related gene expression (Fig. 1h). These abovementioned results indicate SVs indeed were involved the expression of some genes in the ER-related pathways contributing to heat tolerance. Although SVs have been identified to be responsive to environmental stress and defence response (Fuentes et al. 2019; Bayer et al. 2019 and Dolatabadian et al. 2020), the SVs underlying heat tolerance were not detaily characterized in these studies. Yan et al. (2023) Abbreviations ER Endoplasmic reticulum fdSVs Population frequency differences RWP-RK Transcription factors TF Transcriotion factor SNP Single nucleotide polymorphism SVs Structural variations Acknowledgements The author thanks Dr. Linkai Huan from Sichuan Agricultural University for assistance in writing this manuscript. g The author thanks Dr. Linkai Huan from Sichuan Agricultural University for assistance in writing this manuscript. Authors’ contributions Yan H, Sun M, Zhang Z et al. (2023) Pangenomic analysis identifies structural variation associated with heat tolerance in pearl millet. Nat Genet 55;1–12. link: https://www.nature.com/articles/s41588-023-01302-4. Authors’ contributions B.H, H.Y, M.S., & Y.J. wrote the manuscript. The authors read and approved the manuscript. Zhao C, Liu B, Piao S et al. (2017) Temperature increase reduces global yields of major crops in four independent estimates. Proc Natl Acad Sci 114:9326– 9331. link: https://www.pnas.org/doi/abs/10.1073/pnas.1701762114. Zhu J-K. (2016) Abiotic stress signaling and responses in plants. Cell 167:313– 324. link: https://www.sciencedirect.com/science/article/pii/S009286741 6310807. Availability of data and materials Not applicable. Availability of data and materials Not applicable. Competing interests Not applicable. Received: 4 April 2023 Accepted: 3 May 2023 Received: 4 April 2023 Accepted: 3 May 2023 Acknowledgements The author thanks Dr. Linkai Huan from Sichuan Agricultural University for assistance in writing this manuscript. Page 5 of 5 Huang et al. Stress Biology (2023) 3:12 Huang et al. Stress Biology (2023) 3:12 Huang et al. Stress Biology Yan H, Sun M, Zhang Z et al. (2023) Pangenomic analysis identifies structural variation associated with heat tolerance in pearl millet. Nat Genet 55;1–12. link: https://www.nature.com/articles/s41588-023-01302-4. Zhao C, Liu B, Piao S et al. (2017) Temperature increase reduces global yields of major crops in four independent estimates. Proc Natl Acad Sci 114:9326– 9331. link: https://www.pnas.org/doi/abs/10.1073/pnas.1701762114. Zhu J-K. (2016) Abiotic stress signaling and responses in plants. Cell 167:313– 324. link: https://www.sciencedirect.com/science/article/pii/S009286741 6310807. Declarations Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. References Amin N, Ahmad N, Khalifa MA et al. (2023) Identification and molecular charac- terization of RWP-RK transcription factors in soybean. Genes 14:369. link: https://www.mdpi.com/2073-4425/14/2/369. Bayer PE, Golicz AA, Tirnaz S et al. (2019) Variation in abundance of predicted resistance genes in the Brassica oleracea pangenome. Plant Biotechnol J 17:789–800. link: https://onlinelibrary.wiley.com/doi/full/10.1111/pbi. 13015. Dolatabadian A, Bayer PE, Tirnaz S et al. (2020) Characterization of disease resistance genes in the Brassica napus pangenome reveals significant structural variation. Plant Biotechnol J 18:969–982. link: https://onlinelibr ary.wiley.com/doi/full/10.1111/pbi.13262. Fuentes RR, Chebotarov D, Duitama J et al. (2019) Structural variants in 3000 rice genomes. Genome Res 29:870–880. link: https://genome.cshlp.org/ content/29/5/870.short. Khan I, Awan SA, Rizwan M et al. (2023) Physiological and transcriptome analyses demonstrate the silver nanoparticles mediated alleviation of salt stress in pearl millet (Pennisetum glaucum L). Environ Pollut 318:120863. link: https://www.sciencedirect.com/science/article/pii/S026974912 2020784?casa_token=SyXgVWBT8GYAAAAA:WqzzyyVUe9NYdW87Y4bJ RrJQomZaCBhIxnKYMsRCWrzmMn-XvcjsibIEXsIOfMqhy9W1oDu_Lg. Liu Y, Burgos JS, Deng Y et al. (2012) Degradation of the endoplasmic reticu- lum by autophagy during endoplasmic reticulum stress in Arabidopsis. Plant Cell 24:4635–4651. link: https://academic.oup.com/plcell/article/24/ 11/4635/6098127. Liu Y, Du H, Li P et al. (2020) Pan-genome of wild and cultivated soybeans. Cell 182: 162–176. e113. link: https://www.sciencedirect.com/science/article/ pii/S0092867420306188. Satyavathi CT, Ambawat S, Khandelwal V et al. (2021) Pearl millet: a climate- resilient nutricereal for mitigating hidden hunger and provide nutritional security. Front Plant Sci 12:659938. link: https://www.frontiersin.org/artic les/10.3389/fpls.2021.659938/full. Sun M, Huang D, Zhang A et al. (2020) Transcriptome analysis of heat stress and drought stress in pearl millet based on Pacbio full-length transcrip- tome sequencing. BMC Plant Biol 20:1–15. link: https://bmcplantbiol. biomedcentral.com/articles/10.1186/s12870-020-02530-0. Varshney RK, Shi C, Thudi M et al. (2017) Pearl millet genome sequence pro- vides a resource to improve agronomic traits in arid environments. Nat Biotechnol 35:969–976. link: https://www.nature.com/articles/nbt.3943. Verslues PE, Bailey-Serres J, Brodersen C et al. (2023) Burning questions for a warming and changing world: 15 unknowns in plant abiotic stress. Plant Cell 35:67–108. link: https://academic.oup.com/plcell/article/35/1/67/ 6677234. Wang W, Mauleon R, Hu Z et al. (2018) Genomic variation in 3,010 diverse accessions of Asian cultivated rice. Nature 557:43–49. link: https://www. nature.com/articles/s41586-018-0063-9. Wang B, Hou M, Shi J et al. (2023) De novo genome assembly and analyses of 12 founder inbred lines provide insights into maize heterosis. Nat Genet 55;1–12. link: https://www.nature.com/articles/s41588-022-01283-w.
https://openalex.org/W4224921468	https://nottingham-repository.worktribe.com/preview/7834831/guiney-2022-parole-parole-boards-and-the-institutional-dilemmas-of-contemporary-prison-release.pdf	English	null	Parole, parole boards and the institutional dilemmas of contemporary prison release	Punishment & society	2,022	cc-by	9,509	Article Punishment & Society 2023, Vol. 25(3) 621–640 © The Author(s) 2022 Keywords parole, prison release, sentencing, penal policy, institutions Parole, parole boards and the institutional dilemmas of contemporary prison release Article reuse guidelines: sagepub.com/journals-permissions DOI: 10.1177/14624745221097371 journals.sagepub.com/home/pun Thomas C Guiney University of Nottingham, UK Abstract The decision to release is a deﬁning feature of the carceral experience: at once a neces- sary function of a dynamic penal system, and a highly contested form of symbolic com- munication where the anxieties and contradictions of contemporary penality begin to coalesce. In this paper I argue that the institutions we rely upon to make these determi- nations in a fair, consistent and efﬁcient manner are under increasing strain. Drawing upon insights from historical institutionalism, I seek to show that the parole board model of discretionary decision-making that ﬁrst emerged during the highwater mark of mid-twentieth century penal modernism has proved remarkably resilient to reform, but is slowly fracturing into a more complex, multi-layered prison release landscape. I explore the implications of this gradual historical transformation and conclude this paper with a call for new ways of thinking about prison release as an increasingly inter- connected sphere of penal governance. Corresponding author: Thomas C Guiney, School of Sociology and Social Policy, University of Nottingham, University Park, Nottingham, NG7 2RD, UK. Email: thomas.guiney@nottingham.ac.uk Introduction The decision to release is a deﬁning feature of the carceral experience: at once a necessary function of a dynamic penal system, and a highly contested form of symbolic Corresponding author: 622 Punishment & Society 25(3) communication where the anxieties and contradictions of contemporary penality begin to coalesce. Over time, a myriad of penal practices have developed to manage this decision- making discretion in ways that are consistent with broader political, penological and administrative aims (van Zyl Smit and Corda, 2018). However, as I seek to show here, the institutions we have relied upon for more than a generation are under increasing strain. The decision to release is now highly contested and a bewildering array of penal developments continue to destabilise existing decision-making structures with sig- niﬁcant implications for the humanity, fairness and effectiveness of the penal system (see Freiberg et al., 2018; Padﬁeld, 2020; Reitz and Rhine, 2020) g While by no means exhaustive, this general picture suggests that an ascendant ‘parole populism’ (Moffa et al., 2019) has eroded the rights of prisoners and relegated the careful work of rehabilitation to the imperatives of an all-encompassing public protection. Moments of crisis, such as the Worboys case in England and Wales, or the Dutroux scandal in Belgium, continue to undermine public conﬁdence in the parole system (see Fitzgerald et al., 2021), while a host of previously excluded policy actors are now ‘press- ing in’ on prison release decision-making with growing assuredness (Annison, 2020: 11). Release from prison is subject to a plethora of licence conditions (Padﬁeld, 2019) and parole boards are now ﬁrmly established within a broader apparatus of risk management, community supervision and control (Barry, 2021; Hannah-Moffatt and Yule, 2011). The extraordinary growth of indeterminate sentencing in some jurisdictions has seen the locus of sentencing discretion move ‘downstream’ from the criminal courts (Rhine et al., 2017) and an expanding cohort of recall prisoners has renewed longstanding concerns about executive overreach and the limits of the liberal democratic state (Padﬁeld and Maruna, 2006). In this context, the decision to release has become the subject of intense public scru- tiny and marked differences can now be observed in the recent developmental trajectories of English-speaking and European penal systems (see van Zyl Smit and Corda, 2018). The United States has experienced a troubling increase in the number of prisoners serving life sentences without parole (Seeds, 2021). Introduction While in Australia a series of legis- lative initiatives have restricted the discretion afforded to state-level parole boards, pro- moted the use of ‘no body no release’ clauses and, in some cases, resulted in the use of emergency legislation to prevent the release of certain named individuals (Moffa et al., 2019). In many European jurisdictions the direction of travel has been altogether less punitive, but no less consequential. In the early 2000s, Denmark, Finland and Sweden acted upon recommendations from the Council of Europe by introducing judicial systems of release for life-sentence prisoners (Schartmueller, 2019: 392). More recently, the Republic of Ireland has passed legislation that will place the parole board on a statu- tory footing and, it is hoped, signify a long-term shift towards a more continental rights-based approach (Grifﬁn, 2018). g pp In this paper I want to argue that these examples of contemporary institution building should be of considerable interest to academic researchers and must be understood as important objects of study in and of themselves, rather than the simple corollary of legal powers, individual actions or political calculation. That greater analytical sensitivity to these meso-level dynamics can open up new vistas for the study of prison release and 623 Guiney the institutional dilemmas that shape policy and practice. How do we account for the complex picture of continuity and change that characterise the recent histories of so many local, state level and national parole boards? How do we explain the patchwork of decision-making structures that now deﬁne the day-to-day administration of prison release in many jurisdictions? What are the likely consequences of recent attempts at institution building and how is the role of the state changing with respect to prison release decision-making? In seeking to answer this broad constellation of questions, I draw upon recent the- oretical insights from historical institutionalism, an established political science research tradition, that starts from the basic proposition that it is more enlightening to study human interactions in the context of: (a) rule-based structures that are them- selves human creations, and (b) sequentially, as life is lived, rather than by taking ana- lytical snapshots of these interactions at any one point in time (Sanders, 2006: 39). Introduction My central claim is that the parole board model of discretionary decision-making that ﬁrst emerged during the highwater mark of mid-twentieth century penal modern- ism has proved remarkably resilient to reform, but is slowly fracturing into a more complex, multi-layered prison release landscape deﬁned by the interactions between four institutionalised decision-making styles I characterise here as: administrative-bureaucratic, non-executive committee, legislative pre-commitment and judicial body. That over time, this gradual transformation is producing new and poorly understood ‘institutional mosaics’ of decision-making authority (Orren and Skowronek, 1994) that occupy an uncertain position at the intersection between citizen and state, the prison and the community, the past, present and possible futures of the indi- vidual life course (see Dagan, 2022) g This line of argument will proceed as follows: First, I begin to sketch out a theor- etical framework for the study of prison release in historical institutional perspective. Second, I locate the decision to release within a broader historical context associated with the decline of penal modernism, and a series of crises that have destabilized the penological assumptions upon which many parole boards were founded. Third, I survey the emerging contours of a more ﬂuid and multi-layered prison release land- scape before turning in the fourth section to examine the implications of this gradual historical transformation. I conclude that prison release can no longer be understood in conventional terms as a problem to be solved by government but must now be viewed instead as an increasingly complex site of governance that invites us to ‘fracture’ the penal state (Rubin and Phelps, 2017) and reﬂect upon the interconnectivity of the prison release system as a whole. Methodological note Data for this study was drawn from the authors detailed historical study of early release in England and Wales (see Guiney, 2018), supplemented by an extensive literature review of recent published work relating to parole, parole boards, prison release and prison recall. Where comparative research exists, international examples are used to illustrate key theoretical claims and provide a richer account of how the institutional dilemmas 624 Punishment & Society 25(3) of contemporary prison release achieve contingent expression at a local, national and supranational level (see e.g. Rubin and Phelps, 2017; Seeds, 2021). This research strategy helps to ground the paper and situate key claims within a broader international context. However, there are limitations to this approach. The uneven coverage of existing (English language?) research means that it has not been possible to systematically analyse recent developments in Eastern Europe, Asia, Latin America, Africa, or the Caribbean. This necessarily limits the generalisability of this work which, at this stage, is presented with particular reference to the distinct institutional mosaics, and layered structures of decision-making authority, that are slowly taking shape in many English-speaking jurisdictions. Why institutions matter While these intersecting research agendas have greatly advanced our knowledge of prison release law, policy and practice, they are not without limitations. Burke (1984: 49) once observed that a ‘way of seeing is always a way of not seeing’ and these varied literatures often pay insufﬁcient attention to the meso-level dynamics that help shape the explana- tory puzzle of contemporary prison release. Liberal democratic systems of government depend, not only upon macro-structural dynamics and the motivations of individual actors, but the design of institutions. That in order to understand how governments really work we must look beyond the formal legal-administrative institutions of the state to consider the organisational settings that structure everyday social, political and economic life. As Hall and Taylor (1996: 947) remind us, institutions are ‘… not just formal rules, procedures or norms, but the symbol systems, cognitive scripts, and moral templates that provide the “frames of meaning” guiding human action’. Thinking institutionally can therefore help us to explain why it is that parole boards change markedly over time whilst at the same time presenting an outward appearance of stability and permanence with respect to their formalised rules, powers and operating procedures (see King, 2018; Padﬁeld, 2019). This way of seeing draws attention to the diversity of international decision-making practices that are characterised as ‘parole’ (van Zyl Smit and Corda, 2018), and invites us to look beyond the parole board as a monolithic actor to consider the proliferation of internal decision-making practices that now encompass expedited reviews ‘on the papers’, oral hearings and judicial appeal mechanisms (Padﬁeld, 2020). Moreover, institutional analysis can provide a more precise vocabulary for describing the spatial dimensions of criminal justice and the small but perceptible shifts in organisational tone and culture that prisoners, victims and practitioners experience as they progress through what Aviram (2020: 9) has described as a succession of ‘performative spaces’. As Herzog-Evans has observed (2019: 188), the decision to release must be understood as the ﬁnal step along an inter- connected ‘penal continuum’ that encompasses the ‘front-end’ decisions of a sentencing judge and the ‘back door’ determinations of parole boards, probation ofﬁcers and other street level penal bureaucrats (see also Dagan, 2022). Theorising the decision to release While the public contestation of crime continues to focus, to a signiﬁcant extent, upon the ‘front end’ crime control capabilities of the state, there is now growing recognition that the way prison release discretion is exercised can have very signiﬁcant implica- tions for the scale, scope and reach of the penal system (Fitzgerald et al., 2021; Rhine et al., 2017). In a recent contribution to the literature, Reitz and Rhine (2020) demon- strate that discretionary release is central to the story of American mass incarceration and state-level parole boards now wield considerable power within contemporary penal policymaking. A varied punishment and society literature has located prison release within a broader sociological context associated with the new penology (Feeley and Simon, 1992), penal populism and the social construction of dangerous- ness (Annison, 2020). While a series of careful empirical studies have illuminated the slow professionalization of paroling authorities (Paparozzi and Caplan, 2009), the evolving logics of parole supervision (Barry, 2021), and the subjective experience of captive populations as they confront the new pains of indeterminacy (Crewe et al., 2017). In many instances, academic lawyers have been at the forefront of contemporary research and policy debate (see e.g. Grifﬁn, 2018; Freiberg et al., 2018; Padﬁeld, 2020; Rhine et al., 2017). A varied public law literature continues to scrutinise the quality of prison release decision-making, and the extent to which parole boards possess the necessary capacities to discharge the responsibilities of a court. Legal commentators have reﬂected upon the implications of international human rights commitments and common law principles of due process and the right to a fair trial (Reitz, 2020). In parallel, researchers working within the socio-legal tradition have - on those rare occasions when research access has been granted – sought to illuminate the everyday routinized practices of parole board decision-making (see e.g. Hood and Shute, 2000), and the central importance of parole board culture in both catalysing and resisting global trends in the use of punishment (Dagan, 2022; Grifﬁn, 2018; Ruhland, 2020). 625 Guiney The building blocks of historical institutionalism Layering: New rules, ideas and narratives are introduced on top of, or alongside, exist- ing institutional practices. 1. Layering: New rules, ideas and narratives are introduced on top of, or alongside, exist- ing institutional practices. 2. Drift: The formal rules of an institution remain stable, but the external operating envir- onment shifts in ways that alter their practical application. A lack of institutional main- tenance may result from ‘inertia’ or a deliberate process of political neglect. 2. Drift: The formal rules of an institution remain stable, but the external operating envir- onment shifts in ways that alter their practical application. A lack of institutional main- tenance may result from ‘inertia’ or a deliberate process of political neglect. y p p g 3. Displacement: New institutions replace an existing set of arrangements that were meant to perform the same task. y p p g 3. Displacement: New institutions replace an existing set of arrangements that were meant to perform the same task. 4. Conversion: New purposes become attached to old institutions which are ‘redirected’ or reinterpreted in novel ways. (Mahoney and Thelen, 2010: 14–15) 4. Conversion: New purposes become attached to old institutions which are ‘redirected’ or reinterpreted in novel ways. (Mahoney and Thelen, 2010: 14–15) Third, historical institutionalism draws upon inter-disciplinary insights from economics, sociology and organisational studies to explain how institutions constrain individual actors and shape policy outcomes (Hall and Taylor, 1996). In this respect, it can be helpful to think of institutions as ‘processes’, rather than ‘things’ (Goodin, 1996: 19) that exist to simplify human interaction. As Lowndes and Roberts (2013: 47–60) help- fully summarise: formal rules organise how power is constituted, exercised, legitimated and controlled. Informal practices and standard operating procedures reduce complexity and create uniform, routinized and predictable patterns of behaviour. Narratives and storytelling rituals support individuals to distil, mobilise and act upon a collective under- standing of the organisational settings they inhabit (see also Annison, 2021). Third, historical institutionalism draws upon inter-disciplinary insights from economics, sociology and organisational studies to explain how institutions constrain individual actors and shape policy outcomes (Hall and Taylor, 1996). In this respect, it can be helpful to think of institutions as ‘processes’, rather than ‘things’ (Goodin, 1996: 19) that exist to simplify human interaction. As Lowndes and Roberts (2013: 47–60) help- fully summarise: formal rules organise how power is constituted, exercised, legitimated and controlled. The building blocks of historical institutionalism To date, the phenomena outlined here have been explored from a broadly legal or socio- logical disciplinary standpoint but, as I seek to show in the analysis that follows, histor- ical institutionalism offers an additional, and complementary, analytical point of departure for scholars seeking to (a) understand the forces that drive institutional creation, adaptation, and longevity, and (b) to make sense of how parole boards shape, and are in turn shaped by, the evolving aims and techniques of penal policy. A comprehensive review of this literature is well beyond the scope of this paper (see e.g. Fioretos et al., 2016; Guiney, 2018; Sanders, 2006), however, it is possible to identify three conceptual building blocks of the historical institutionalist tradition that are particularly relevant in this context. 626 Punishment & Society 25(3) First, historical institutionalism is closely associated with the concept of path depend- ency. This basic idea suggests that what happens at an early stage in a causal sequence will affect possible outcomes at a later point in time (see Rubin, 2021). On the one hand, work in this ﬁeld has drawn attention to both the long-term inﬂuence of positive and negative feedback mechanisms that alter the ‘switching costs’ of moving from one developmental pathway to another (Pierson, 2004). While on the other, recent research has explored the creative potential of ‘critical junctures’, such as the 2008 global ﬁnancial crash, where these normal structuring conditions are loosened, and a broader range of policy options become imaginable (Karstedt et al., 2019). In this respect, path depend- ency offers a sophisticated theoretical repertoire that underscores the importance of timing and sequence, the disruptive effects of historical contingency, the potential for ‘large’ consequences to ﬂow from ‘small events’ and the ‘stickiness’ of institutional tra- jectories that, once established, can be very difﬁcult to reverse. Second, recent iterations of historical institutionalism have sought to overcome the structuralism of early theoretical perspectives that emphasised stability and stop-go periods of ‘punctuated equilibrium’ (Steinmo et al., 1992). As Mahoney and Thelen (2010) observe, institutions are, by their very nature resilient to external shocks and, in this context, change is likely to be deﬁned by a long-term process of ‘gradual transform- ation’ as the political coalitions of support that underpin existing institutional arrange- ments shift and reconﬁgure. In practice, these temporal processes can take a number of distinct forms: 1. Institutional change: parole, penal policy and the crises of penal modernism It is perhaps the deﬁning contribution of the sociology of punishment tradition that penal practices are not simply an unproblematic response to crime trends, but inexorably bound up with the transformation of broader social, economic and political structures (see e.g. Simon, 1993: 5). While supporting the broad thrust of this argument, the theoretical tools assembled here offer a number of important clues for thinking about how institutions interrupt and mediate this process across the broad sweep of historical time. In this section, I draw upon insights from historical institutionalism to explore how the parole board model of discretionary decision-making locked in many of the ideational assump- tions of penal modernism and in so doing projected these ideas into the future in ways that have become increasingly at odds with the evolving aims and techniques of penal policy in many English-speaking jurisdictions. The building blocks of historical institutionalism Informal practices and standard operating procedures reduce complexity and create uniform, routinized and predictable patterns of behaviour. Narratives and storytelling rituals support individuals to distil, mobilise and act upon a collective under- standing of the organisational settings they inhabit (see also Annison, 2021). 627 Guiney Parole boards and penal modernism While it is now generally accepted that these established power structures began to unravel in the ﬁnal decades of the twentieth century, the key point to note here is that the parole board model of decision-making has proved remarkably resilient to these broader macro- structural shifts, and it is this institutional legacy that is key to understanding the crises of contemporary prison release. The crisis of administration. The administrative demands placed upon the institutions of prison release increased signiﬁcantly in the second half of the twentieth century. In part, this reﬂects the basic challenge of administering a penal system at scale at a time when criminal justice budgets have failed to keep pace with signiﬁcant prison expansion- ism (Reitz and Rhine, 2020: 285). However, the crisis of administration also reﬂects the growing complexity of contemporary sentencing frameworks. In many English-speaking jurisdictions a general process of ‘violence realignment’ (Seeds, 2017: 600) has signiﬁ- cantly altered the offender / offence proﬁle of the general prison population, while a long- term shift towards greater conditionality in release arrangements has swelled the recall prison population and accelerated the ‘revolving door’ between prison and the commu- nity (Padﬁeld and Maruna, 2006). In many instances, these administrative challenges have been exacerbated by a series of victims’ rights initiatives and new forms of danger- ousness legislation. In California, the introduction of Proposition 9, also known as Marsy’s Law, has altered parole board proceedings since 2008 with the effect that lifers denied parole must now wait a presumptive period of ﬁfteen years before the parole board will reconsider their case, unless there is clear and convincing evidence that an extended deferral period is unjustiﬁed (Aviram, 2020: 26). The crisis of legitimacy. At a time when English-speaking societies have become less def- erential towards those in positions of authority and less trusting of penal expertise, parole boards have often struggled to resolve a longstanding and systemic legitimacy deﬁcit. There remains a lack of clarity about how decision-making discretion is organised and, by extension, what prisoners, prisoner families, victims and the general public can legit- imately expect from this process (see Fitzgerald et al., 2021). In this context, parole boards have often sought to legitimise their work by appealing to their decision-making performance. Parole boards and penal modernism It is testament to the power of institution building that the parole boards ﬁrst established more than half a century ago remain synonymous with the decision to release. We quickly become used to the way things are and this can confer a concreteness upon arrangements that are in fact contingent, ﬂeeting and unresolved. Throughout the late-nineteenth century a suite of penal practices, including the ticket of leave system, remission and the prerogative of mercy evolved to expedite the release of those sentenced to imprison- ment by the courts (Guiney, 2018: 55–61). Viewed in Weberian terms, this administrative apparatus began to place the decision to release upon a more systematic bureaucratic footing. However, it is clear that by the mid-twentieth century these practices were ill-suited to the emerging aims of penal modernism and the insoluble problem of how to exercise penal power ‘rationally’ (Simon, 1993). This question generated considerable policy debate in the decades following 1945 and it was in this context that the parole board model of decision-making emerged as the most promising institutional counter- point to a ‘modern’ system of parole premised upon the following ideational, epistemic and governmental assumptions: • Penological support for indeterminate sentences and the personalisation of punishment. • Deference to authority and those who offer ‘expertise’. • Optimism about the transformative potential of criminological research and knowledge. • A bureaucratic system predicated upon the assembly of detailed caseﬁles and decision- making on the papers. • A paternalistic system where prison release conceptualised as a ‘privilege’ rather than a ‘right’. g • Limited institutional access for prisoner families, victims or the general public. 628 Punishment & Society 25(3) Of course, these assumptions were by no means universally accepted. Discretionary parole was the subject of considerable contemporaneous criticism from sentencing judges, prison governors and politicians (Guiney, 2018). However, they were sustained, initially at least, by a powerful coalition of support that gave ofﬁcial expression to the policy ambitions of ‘penal welfarism’ (Garland, 2001) and a broad disposition towards the governance of crime Loader has described as ‘liberal elitism’ (2006: 568). Parole boards and penal modernism Senior parole board members frequently invoke statistics relating to the serious further offending rate (SFOs) and this is particularly common during periods of crisis and scandal when ethical considerations, such as the false positive rate or the treatment of indeterminate prisoners repeatedly denied parole, are subordinated to questions of public protection (Moffa et al., 2019). More generally, parole boards have been heavily inﬂuenced by the principles of new public management and even a cursory review of parole board annual reports reveals that legitimacy is often 629 Guiney framed in overwhelmingly managerialist terms with reference to key performance indicators, risk management and value for money considerations (Paparozzi and Caplan, 2009). In contrast, a series of hard-won legal, political and social movements have compelled parole boards, with varying degrees of success, to reframe their work in ways that better reﬂect the competing legitimacy claims of procedural justice. Penal policy-makers have often proved resistant to research ﬁndings that highlight the importance of voluntary rule compliance and, in this context, change has often been exogenous, with demands for reform emanating from outside the system (van Zyl Smit and Corda, 2018). For example, in England and Wales, a series of high-proﬁle legal cases have cast doubt upon the independence of the Parole Board1 and reafﬁrmed the right of prisoners to an oral hearing in all parole and recall cases.2 In the landmark case of re Lawrence3, the Californian Supreme Court placed important limits upon the decision-making discretion of both the parole board, and state governor, thereby creating justiciable legal standards that continue to be relied upon by prisoners in judicial review proceedings (see Aviram, 2020: 31–35). The crisis of ideas. As crime control has emerged as a less certain ﬁeld (Garland, 2001), penal policymakers have struggled to articulate a clear strategic vision outlining the pur- pose(s) and function of prison release. Parole, as originally conceived, was justiﬁed on the basis of a ‘recognisable peak’ in an individual’s rehabilitation where the interests of the community were better served by the careful reform and reintegration of the offender in the community rather than their continued incarceration (Guiney, 2018). Parole boards and penal modernism This grand policy narrative often diverted attention away from more prosaic penal policy concerns (Shute, 2003: 385), but the crises of prison release outlined above have stripped penal policymaking of its early optimism and inhibited the formulation of a clear strategic vision that connects everyday policy and practice with broader peno- logical ideas. In this less certain climate, the decision to release has become imbued with all the anxieties and contradictions of late-modern penality; the management of risk (Barry, 2021; Hannah-Moffat and Yule, 2011), fear of dangerous ‘others’ (Annison, 2020); the disruptive inﬂuence of penal populism (Moffa et al., 2019) and the challenge of reconciling parole with a retributivist sentencing philosophy premised upon moral ‘desert’, proportionality and truth in sentencing (Dagan, 2022). Intercurrence and the institutional mosaics of contemporary prison release Intercurrence and the institutional mosaics of contemporary prison release When taken as a whole, the crises of prison release have destabilised the institutions that were built during the high-water mark of mid-twentieth century penal modernism. However, this process did not result in the immediate abolition of parole boards nor the creation of radically different decision-making structures. Too much political, 630 Punishment & Society 25(3) ﬁnancial and human capital has been invested in existing institutional arrangements and, in many cases, the switching costs associated with wholesale reform were neither desir- able nor practicable. Instead, institutional change has been deﬁned by layering, drift, dis- placement and conversion (Mahoney and Thelen, 2010). New ideas, techniques and policies have been layered upon existing frameworks (see Rubin, 2016). Policy makers have disinvested from certain organisational capacities and re-directed resources else- where within the penal system. Administrative innovations from cognate spheres of public policy have been adapted through an ongoing process of bricolage and policy transfer. Orren and Skowronek (1994) characterise this gradual process of institutional cre- ation, reproduction and change as ‘intercurrence’ and call for greater analytical sensitivity to the distinct institutional ‘mosaics’ and layered structures of authority that accumulate within all mature systems of government over time. Since the 1970s the demands placed upon the institutions of prison release have changed markedly (van Zyl Smit and Corda, 2018). Successive phases of problem-solving activity have attempted to address the crises of prison release outlined above and, over time, this has contributed to the development of a more complex, multi-layered institutional landscape that is deﬁned by the interactions between four institutionalised decision-making styles I characterise as: administrative- bureaucratic, non-executive committee, legislative pre-commitment and judicial body (see Table 1). Each of these institutionalised decision-making styles is deﬁned by a unique history, internal logic and administrative rhythm (Seeds, 2021). In some instances, these decision- making styles have been incorporated within the existing formal legal-administrative structures of the parole board. While elsewhere new institutional practices have been grafted upon existing sentencing arrangements and now operate alongside parole board decision-making. For example, in Canada, prisoners denied parole by the National Parole Board (NPB) become eligible for statutory release after completing two-thirds of their sentence. Intercurrence and the institutional mosaics of contemporary prison release At this point in time the law requires that the offender be released and placed under community supervision unless board members can establish that the individual is highly likely to commit serious further offences (Hannah-Moffat and Yule, 2011: 169). Administrative-bureaucratic. A decision-making style with origins that can be traced back to the creation of the modern nation state. In complex systems of government, street-level bureaucrats wield considerable discretion on behalf of sovereign authorities that: (a) must out of necessity delegate some decision-making power to junior ofﬁcials and (b) are not privy to the ‘facts on the ground’ in every case (Lipsky, 1980). In keeping with the Weberian tradition of bureaucratic rationality, decision-making is largely rule-based, and heavily circumscribed by policy circulars, codes of conduct and statutory instru- ments. Bureaucracies are hierarchical in design and confer structured discretion upon those who hold speciﬁed positions of seniority, such as Prison Governors or Parole Ofﬁcers. For this reason, administrative-bureaucratic release mechanisms, such as the American system of ‘good time credits’ (see e.g. Aviram, 2020: 20), are often closely aligned with the requirements of prison discipline, incentives and earned privileges, 631 Guiney Table 1. Contemporary prison release and its institutional arrangements. Decision-making styles Administrative-bureaucratic Non-executive committee Legislative pre-commitment Judicial body Rules Policy circulars Legal terms of reference Penal code, sentencing Legislation The Constitution, public law Practices Weberian tradition of bureaucratic rationality Collective discussion and consensus building Automatic (where pre-agreed criteria are satisﬁed) Legal due process. Strengths Efﬁcient, ﬂexible, closely aligned with broader policy aims Defensible, evidence-based. Trusted to deliver a ‘defensible’ consensus Universalizing, suited to high-volume systems Transparent, high levels of due process and procedural justice Weaknesses Secretive, lack of transparency. Little or no independence from government Independence from the executive, lack of procedural safeguards Lack of instrumentality, difﬁcult to distinguish between categories of prisoner Cost, time-intensive, lack of alignment with broader penal policy aims and resources Key actors Street-level bureaucrats Parole Board members The legislature The judiciary Tribunal members Archetypal example(s) Good time credits, the marks system, prison recall Anglophone Parole Boards Automatic Statutory release Continental Sentence Implementation Courts Table 1. Contemporary prison release and its institutional arrangements. Decision-making styles Administrative-bureaucratic Non-executive committee Legislative pre-commitment Judicial body Rules Policy circulars Legal terms of reference Penal code, sentencing Legislation The Constitution, public law Practices Weberian tradition of bureaucratic rationality Collective discussion and consensus building Automatic (where pre-agreed criteria are satisﬁed) Legal due process. Intercurrence and the institutional mosaics of contemporary prison release Strengths Efﬁcient, ﬂexible, closely aligned with broader policy aims Defensible, evidence-based. Trusted to deliver a ‘defensible’ consensus Universalizing, suited to high-volume systems Transparent, high levels of due process and procedural justice Weaknesses Secretive, lack of transparency. Little or no independence from government Independence from the executive, lack of procedural safeguards Lack of instrumentality, difﬁcult to distinguish between categories of prisoner Cost, time-intensive, lack of alignment with broader penal policy aims and resources Key actors Street-level bureaucrats Parole Board members The legislature The judiciary Tribunal members Archetypal example(s) Good time credits, the marks system, prison recall Anglophone Parole Boards Automatic Statutory release Continental Sentence Implementation Courts 632 Punishment & Society 25(3) and the need for administrative ‘safety valves’ that can be activated during periods of emergency (Shute, 2003: 836). While there has been a long-term shift away from administrative-bureaucratic decision-making as the primary release mechanism for most prisoners, this decision-making style has arguably moved even further ‘down- stream’ of the courts and is now closely associated with community supervision and a growing number of prison recall cases (Padﬁeld and Maruna, 2006). Non-executive committees. The non-executive committee was perhaps the major institu- tional innovation of penal modernism (Hawkins, 1983). Non-executive committee struc- tures were layered upon existing administrative-bureaucratic systems of government where it was felt that the exercise of discretion required specialist expertise to arrive at a ‘right’ or ‘defensible’ consensus in complex cases. Decision-making is non-executive in so far as parole board members have no responsibility for the day-to-day administra- tion of the penal system and decision-making authority is exercised in accordance with formally prescribed terms of reference. In some cases, non-executive committees are purely advisory, as is currently the case in the Republic of Ireland (Grifﬁn, 2018), while in Canada the parole board is independent of the Canadian Correctional Service and government ministers have no statutory authority to give direction to the Chairperson (Hannah-Moffat and Yule, 2011: 152). Moreover, as Reitz (2020: 2754– 2762) observed in a recent review, there remains signiﬁcant state level variation in the proportion of a sentence that is subject to parole boards discretion within the American federal system of government. Intercurrence and the institutional mosaics of contemporary prison release While procedural justice and due process are likely to be important considerations for policy makers, who are required by law to comply with established public law principles and standards, non-executive committees are embedded within the administrative appar- atus of the state and their performance is therefore closely linked to the administrative rhythms of government and the delivery of broader strategic aims, such as public protec- tion or the rehabilitation of prisoners (Kemshall, 2013). The key point being that some executive control is traded for access to the perceived beneﬁts of arm’s length decision- making, and it is for this reason that non-executive committees do frequently arrive at release decisions that may be politically embarrassing for government. Legislative pre-commitment. Legislative pre-commitment grew in popularity in the 1970s as policy-makers began to wrestle with the administrative challenges of prison expan- sionism and the growing inﬂuence of the ‘just deserts’ movement (van Zyl Smit and Corda, 2018). Here, decision-making authority is orientated towards general populations, or prisoner sub-cohorts, rather than individuals. For this reason, legislative pre- commitment is closely associated with determinate sentencing frameworks where prison release decisions are made on the basis of transparent, consistent and universal cri- teria that are made known to prisoners at the point of sentence. Statutory release schemes can reduce the overall period of time spent in prison, but there is nothing in principle that prevents legislative pre-commitment from being used to impose whole life tariffs in the most serious cases (Seeds, 2021). 633 Guiney This is both a major strength and weakness of legislative pre-commitment. For while automatic release schemes are well calibrated to high volume systems where there is insufﬁcient time to review applications on a robust case-by-case basis, their universaliz- ing quality often means they are less well suited to forms of instrumental decision-making which (rightly or wrongly) seek to distinguish between prisoners on the basis of prevail- ing legal, actuarial or moral categories. Ultimately, the executive does reserve the right to amend existing legislative frameworks. However, this can be an onerous process and legislative pre-commitment will often tie the hands of political ministers for a generation or more, as is the case in England and Wales where prison release remains subject to the legal provisions set out in the Criminal Justice Act 2003 (Padﬁeld, 2020). Judicial body. Intercurrence and the institutional mosaics of contemporary prison release Calls for prison release decisions to be made by a judicial authority have been largely resisted in English-speaking jurisdictions (van Zyl Smit and Corda, 2018: 435–461). However, there has been renewed interest in judicial decision-making in recent decades given the growing inﬂuence of international human rights movements and the resurgence of indeterminate sentencing frameworks (Grifﬁn and O’Donnell, 2012; Reitz and Rhine, 2020). Here, the termination of prison sentences is viewed as a legal rather than an administrative matter. Typically, the decision to release is made by a court or specialist tribunal, in accordance with legally enshrined standards of decision- making independence and due process (Padﬁeld, 2020; Schartmueller, 2019). In both form and function, decision-making follows a legal rather than a bureaucratic rationality. Decision-makers have access to greater legal expertise and court mandated powers, but this constitutional separation of powers may result in a corresponding loss of executive inﬂuence and resources (Padﬁeld, 2020: 471). In many continental European systems, such as Belgium and France, this discretionary power is exercised by professional Sentence Implementation Courts. For example, the French tradition of execution des peines grants considerable latitude to the state appointed judge de l’appli- cation des peines (JAP) to determine a wide-range of sentence management outcomes including, parole licence conditions, remission for good conduct and recall (see Herzog-Evans, 2019). Institutional dilemmas: parole, parole boards and penal policy The demands now placed upon contemporary penal systems mean that a ‘one size ﬁts all’ approach to prison release decision-making is neither desirable nor effective. A more complex, multi-layered prison release landscape is emerging with very signiﬁcant impli- cations for the role of the state and the criminal justice institutions we rely upon to make these decisions on a day-to-day basis. High proﬁle terrorist attacks in England and Wales have prompted considerable public discussion of legislative pre-commitment and the appropriate dividing line between automatic and discretionary release arrangements (Guiney, 2019). In Australia, a series of policy reviews have cast doubt upon the desir- ability of broad parole board discretion (Freiberg et al., 2018; Moffa et al., 2019), while state-level parole boards in America continue to be pulled in different directions by the competing legitimacy claims of performance and procedure (Reitz and Rhine, 2020). 634 Punishment & Society 25(3) In this section I reﬂect upon these key sites of penal policy contestation before turning in the conclusion to consider the political implications of the analysis presented here. Constituting penal power: bifurcation and sentence stratiﬁcation Each of the institutionalised decision-making styles described in the preceding section confer both advantages and disadvantages upon the penal system. However, it is import- ant to note that there is nothing inherently punitive about these arrangements. As Reitz (2020: 2749) rightly notes, ‘the law-given authority to reduce prison sentence lengths always entails a corresponding power to increase prison stays’. In other words, institu- tions constrain policy actors, but they do so in ways that are constitutive of penal power, creating new possibilities for the classiﬁcation, management and supervision of prisoners (Kemshall, 2013). Two examples can be used to illustrate this point. First, the sentence of the court is now frequently stratiﬁed so that different phases of a prison sentence are subject to dif- ferent prison release mechanisms. This process of sentence stratiﬁcation allows policy actors to pursue a wide range of penal policy objectives within a broader statutory frame- work and may help to explain the proliferation of ancillary early release mechanisms, such as Home Detention Curfew (HDC) electronic tagging, Release on Temporary Licence (ROTL) and the short-lived End of Custodial Licence scheme that have all oper- ated in England and Wales in recent years. Second, as I have argued elsewhere (Guiney, 2019), the institutions of prison release are integral to the operation of a bifurcated penal strategy that pivots upon the discretionary release of ‘dangerous’ prisoners found guilty of violent and sexual offences, and automatic release mechanisms that can be used to expedite the safe release of nonviolent, high-volume offenders (see also Aviram, 2020; Beyens, 2019; Seeds, 2017). Crossing the rubicon: The non-executive committee or judicial decision-making? Indeterminate sentences radically alter the balance of power between sentencing judges and paroling authorities members and, in this context, the question of whether parole boards should be organised as non-executive committees or judicial bodies is perhaps the deﬁning institutional dilemma of contemporary prison release. While it is true to say that there have been incremental improvements to the decision-making procedures of many parole boards (King, 2018), it is also clear that many English-speaking jurisdic- tions have been unwilling to give up on the perceived beneﬁts of arm’s length non- executive decision-making and transfer control for high-proﬁle prisoners to the judicial branch of government (Padﬁeld, 2019; Reitz and Rhine, 2020; Schartmueller, 2019). These competing visions for the parole board as a non-executive or judicial body are of course deeply political, and the subject of considerable normative theorising (Grifﬁn and O’Donnell, 2012; Rhine et al., 2017), but there is also value in situating these trends within a broader historical context. Viewed through the lens of path dependency (see Rubin, 2021) it is signiﬁcant that in many liberal democratic systems of government the historical roots of prison release were not judicial, but closely associated with the 635 Guiney exercise of executive power and the royal prerogative of mercy (Guiney, 2018). From these initial starting conditions prison release has evolved within the executive’s sphere of inﬂuence, and, over time, this basic historical fact has altered the switching costs of transferring decision-making authority from the executive to the judicial branch of government. In a fascinating account of recent developments in Belgium, Beyens (2019) describes the partial transfer of decision-making responsibility for prison release from political min- isters to the court system. Following a series of politically damaging release decisions, including the infamous ‘Dutroux Case’ (2019: 170), the ‘Act of 2006 on the External Legal Position’ was passed in order to facilitate the transfer of all release decisions to independent Sentence Implementation Courts staffed by professional judges with at least ﬁve years judicial experience. However, in what might be considered a quintessen- tial example of institutional drift (Mahoney and Thelen, 2010: 14), successive govern- ments have delayed the activation of these statutory powers resulting in a de facto system of penal bifurcation. Crossing the rubicon: The non-executive committee or judicial decision-making? At the time of writing, Prison Governors retain responsibility for the release of prisoners serving sentences of three years, or under, while a smaller cohort of long-term prisoners are now subject to a robust system of judicial decision- making (Beyens, 2019: 170). Unlike in Belgium, which has so far managed only a partial transfer of prison release authority to the judiciary, the well documented ‘punitive turns’ experienced by many English-speaking jurisdictions in the ﬁnal decades of the twentieth century have made it progressively more difﬁcult for advocates of reform to build the necessary coalitions of political support required to cross the constitutional Rubicon and undertake such a major change in the machinery of government (Rhine et al., 2017: 282). In this context, a series of positive feedback mechanisms have generated signiﬁcant institutional incen- tives for penal policy-makers to continue working within existing institutional arrange- ments (Pierson, 2004). Politicians conditioned by the short-term demands of the electoral cycle are likely to discount the long-term beneﬁts of judicial decision-making, while the public contestation of crime continues to create electoral ‘winners’ and ‘losers’ with the implication that political incumbents will frequently use their position to institution- alise their advantages and impose punitive policy choices upon political outsiders. Narratives, storytelling and the curious case of ‘court-like’ bodies The push and pull of these path dependent processes means that many non-executive parole boards are now caught in a precarious institutional position betwixt and between two sym- bolically powerful decision-making styles: administrative-bureaucratic decision-making and the judicial body. In a recent retrospective, King (2018: 19) observes that the Parole Board of England and Wales has adopted a more ‘court like’ working model since its establishment in 1967. And yet, it remains the case that the board is legally constituted as an ‘independent executive Non-Departmental Public Body’ that is sponsored by the Ministry of Justice. The Secretary of State retains the power to issue policy directions, approve the appointment of parole board members and conduct an annual ‘performance review’ with the Parole Board chairperson (Padﬁeld, 2020: 474). 636 Punishment & Society 25(3) In this context, institutional narratives have played an important role in resolving the apparent disconnect between the formal rules and everyday working practices of the parole board. As Annison (2021: 5) has recently argued, storytelling and narrative are indispensable features of political action. They equip policy makers with a sense of moral purpose and help to deﬂect contestation by locating speciﬁc policies within a broader account of policy change. Through the regular repetition of organisational narra- tives that position the parole board as a ‘court-like body’, parole members seek to create a sense of shared identity and purpose. They provide a way to overcome the contradictions of existing rules and practices and, in so doing, present a more fully resolved image of the parole board to its many audiences that now include prisoners, their families, politicians, the media and the general public. Conclusion: From government to the governance of prison release? In this paper I have argued that institutions have played an important role in shaping the long-term developmental trajectory of prison release policy and practice in many English-speaking jurisdictions. Drawing upon insights from historical institutionalism I have argued that the parole board model of discretionary decision-making that ﬁrst emerged during the high-water mark of mid-twentieth century penal modernism has proved remarkably resilient to reform but is now slowly fracturing into a more complex and multi-layered prison release landscape deﬁned by the interactions between four institutionalised decision-making styles I have characterised as: administra- tive bureaucratic, non-executive committee, legislative pre-commitment and judicial. Over time, these institutional arrangements have constrained the choices available to policy actors, but as we have explored, they are also constitutive of penal power, creating new opportunities for the stratiﬁcation of prison sentences and the operationalisation of bifurcated penal strategies. If correct, the analysis presented here has far-reaching implications for how we think about the penal state and the role of government (see Rubin and Phelps, 2017). For while many politicians continue to talk tough on prison release, and promote an image of gov- ernment command and control, their ability to affect substantive policy change is in fact increasingly limited. As one leading commentator put it in a different context, ‘the centre has rubber levers; pulling the central policy lever does not necessarily mean something happens at the bottom’ (Rhodes, 1997: 1255). Left unchecked, this growing dissonance between what politicians say, and what they can actually do, can have a corrosive effect upon public trust in government and create the space for populist policies to thrive (Fitzgerald et al., 2021). g A more complex, multi-layered prison release landscape is slowly emerging and new ways of thinking about the role of state are urgently needed. In my view, prison release can no longer be understood in conventional terms as a problem to be solved by govern- ment but should instead be viewed as a complex sphere of governance that de-centres the state as monolith and invites greater analytical sensitivity to how policy is negotiated 637 Guiney within, and between, policy networks deﬁned by interdependence, coordination and plur- alism (Rhodes, 1997). Conclusion: From government to the governance of prison release? Increasingly then, the primary task of the state is no longer to cen- trally manage prison release but ought instead to be focused on creating the conditions for negotiation and compromise so that a mixed economy of institutional arrangements can ﬂourish. To facilitate public discussion of the principles that ought to underpin prison release and safeguard the integrity of this framework by promoting a reliable system of institutional checks and balances. Of course, there are no right answers to these questions and all nation states will arrive at different solutions to these intractable governance problems. Rather the key point here is that new institution building must be better attuned to the needs of the prison release system as a whole rather than focusing narrowly on the strengths and weaknesses of parole board decision-making alone. Changes in one part of the system can have dramatic and unintended consequences elsewhere. If we start from the basic principle that ‘the greater the authority concentrated in the prison release decision, the greater the need for procedural regularity’ (Rhine et al., 2017: 281) this has implications that reach far beyond the caseload of the parole board. If we assume the continued (but modest use) of indeterminate sentences for the most serious violent and sexual offences (Seeds, 2017), is it really so difﬁcult to envisage a balanced prison release framework where statutory release anchors the release date for the overwhelming majority of determinate sentence prisoners? Where street-level penal bureaucrats are encouraged to develop agile and light touch early release schemes that incentivise good conduct and rehabilitation while in prison. Where the re-release of indi- viduals recalled to prison is delegated to problem-solving panels with knowledge of desistance and public protection, and the parole board is reconstituted as a judicial body with the institutional resources to determine whether the continued incarceration of indeterminate sentence prisoners meets the necessary burden of proof and remains in the public interest? Acknowledgment A version of this paper was delivered at the University of Cambridge Institute of Criminology Seminar Series. The author would also like to thank Harry Annison, Ailbhe O’Loughlin and the anonymous peer reviews for their constructive feedback on this paper as it has developed. Declaration of Conﬂicting Interests The author(s) declared no potential conﬂicts of interest with respect to the research, authorship, and/ or publication of this article. Funding The author(s) received no ﬁnancial support for the research, authorship, and/or publication of this article. Notes 1. R (Brooke) v Parole Board [2008] EWCA Civ 29. 2. Osborn Booth and Reilly v. The Parole Board [2013] UKSC 61. h 3. Re Lawrence , 44 Cal. 4th (2008) ORCID iD Thomas C Guiney https://orcid.org/0000-0002-5601-4298 638 Punishment & Society 25(3) References Annison H (2020) Re-examining risk and blame in penal controversies: parole in England and Wales. In: Pratt J and Anderson J (eds) Criminal Justice, Risk and the Revolt Against Uncertainty. Basingstoke: Palgrave Macmillan, pp. 139–163. Annison H (2021) The role of storylines in penal policy change. Punishment & Society. Epub ahead of print 129 January 2021. https://doi.org/10.1177/1462474521989533 Aviram H (2020) Yesterday’s Monsters: The Manson Family and the Illusion of Parole. Oakland, CA: University of California Press. Barry M (2021) ‘Walking on ice’: the future of parole in a risk-obsessed society. Theoretical Criminology 25(2): 325–342. Beyens K (2019) Belgian Release policies, rationales and practices. European Journal of Probation 11(3): 169–187. Burke K (1984) Permanence and Change: An Anatomy of Purpose. Berkley: University of California Press. Crewe B, Hulley S and Wright S (2017) Swimming with the tide: adapting to long-term imprisonment. Justice Quarterly 34(3): 517–541. Dagan N (2022). Overcoming penal boundaries: exploring the evolution of retributive time through parole decision-making. British Journal of Criminology 62(1): 37–54. Feeley M and Simon J (1992) The new penology: notes on the emerging strategy of correc- tions and its implications. Criminology; An Interdisciplinary Journal 30(4): 449–474. Fioretos O, Falleti TG, Sheingate A (2016) Historical institutionalism in political science. In: Fioretos O, et al. (eds) The Oxford Handbook of Historical Institutionalism. Oxford: Oxford University Press, pp. 3–28. Fitzgerald R, Freiberg A, Dodd S, et al. (2021) Building public conﬁdence in parole boards: ﬁndings from a four-country study’. British Journal of Criminology. Epub ahead of print 2 November 2021. https://doi.org/10.1093/bjc/azab097 Freiberg A, Bartels L, Fitzgerald R, et al. (2018) Parole, politics and penal policy. QUT Law Review 18(1): 191–216. Garland D (2001) The Culture of Control: Crime and Social Order in Contemporary Society. Oxford: Oxford University Press. Goodin RE (1996) Institutions and their designs. In: Goodin RE (ed) The Theory of Institutional Design. Cambridge: Cambridge University Press, pp. 1–53. Grifﬁn D (2018) Killing Time: Life Imprisonment and Parole in Ireland. London: Palgrave Macmillan. Grifﬁn D and O’Donnell I (2012) The life sentence and parole. The British Journal of Criminology 52(3): 611–629. Guiney T (2018) Getting Out: Early Release in England and Wales, 1960–1995. Oxford: Oxford: University Press. Guiney T (2019) Marginal gains or diminishing returns? Penal bifurcation, policy change and the administration of prisoner release in England and Wales. European Journal of Probation 11(3): 139–152. References 639 Guiney Hall PA and Taylor RCR (1996) Political science and the three new institutionalisms. Political Studies 44(5): 936–957. Hannah-Moffat K and Yule C (2011) Gaining insight, changing attitudes and managing ‘risk’: parole release decisions for women convicted of violent crimes. Punishment & Society 13(2): 149–175. Hawkins K (1983) Assessing evil: decision behaviour and parole board justice. British Journal of Criminology 23(2): 101–127. Herzog-Evans M (2019) French Early release: McProcedures and McRe-entry. European Journal of Probation 11(3): 188–201. Hood R and Shute S (2000) Parole Decision-Making: Weighing the Risk to the Public. Home Ofﬁce Research Findings No. 114. London: Home Ofﬁce. Karstedt S, Bergin T and Koch M (2019) Critical junctures and conditions of change: explor- ing the fall of prison populations in US states. Social & Legal Studies 28(1): 58–80. Kemshall H (2013) Dangerous offenders: release and resettlement. In: Hucklesby A and Hagley-Dickinson L (eds) Prisoner Resettlement: Policy and Practice. Cullompton: Willan, pp. 270–288. pp King R (2018) Decisions, decisions, decisions: reﬂections on 50 years of parole. Prison Service Journal 237(May): 18–21. Lipsky M (1980) Street-Level Bureaucracy: Dilemmas of the Individual in Public Services. New York: Russel Sage Foundation. Loader I (2006) Fall of the ‘platonic guardians’: liberalism, criminology and political responses to crime in England and Wales. British Journal of Criminology 46(4): 561–586. responses to crime in England and Wales. British Journal of Criminology 46(4): 561–586. Lowndes V and Roberts M (2013) Why Institutions Matter: The New Institutionalism in Political Science. Basingstoke: Palgrave Macmillan. Lowndes V and Roberts M (2013) Why Institutions Matter: The New Institutionalism in Political Science. Basingstoke: Palgrave Macmillan. Mahoney J and Thelen K (2010) A theory of gradual institutional change. In: Mahoney J and Thelen K (eds) Explaining Institutional Change: Ambiguity, Agency, and Power. Cambridge: Cambridge University Press, pp. 1–37. Moffa M, Stratton G and Ruyters M (2019) Parole populism: the politicisation of parole in Victoria. Current Issues in Criminal Justice 31(1): 75–90. Orren K and Skowronek S (1994) Institutions and intercurrence: theory building in the full- ness of time. Political Order 38: 111–146. Padﬁeld N (2019) Giving and getting parole: the changing characteristics of parole in England and Wales. European Journal of Probation 11(3):153–168. Padﬁeld N (2020) The function of the parole board – avoiding failure or promoting success? Public Law July: 468–487. References Padﬁeld N and Maruna S (2006) The revolving door at the prison gate: exploring the dramatic increase in recalls to prison. Criminology & Criminal Justice 6(3): 329–352. Paparozzi MA and Caplan JM (2009) A proﬁle of paroling authorities in America: the strange bedfellows of politics and professionalism. The Prison Journal 89(4): 401–425. Pierson P (2004) Politics in Time. Princeton: Princeton University Press. Reitz KR (2020) Prison-release reform and American decarceration. Minnesota Law Review 104: 2741–2779. Reitz KR and Rhine EE (2020) Parole release and supervision: critical drivers of American prison policy. Annual Review of Criminology 3(1): 281–298. Rhine EE, Petersilia J and Reitz KR (2017) The future of parole release. Crime and Justice 46(1): 279–338. Rhodes RAW (1997) Understanding Governance: Policy Networks, Governance and Accountability. Buckingham: Open University Press. 640 Punishment & Society 25(3) Rubin A (2016) Penal change as penal layering: A case study of proto-prison adoption and capital punishment reduction, 1785–1822. Punishment & Society 18(4): 420–441. Rubin A (2021) The promises and pitfalls of path dependence frameworks for analyzing penal change. Punishment & Society. Epub ahead of print 12 October 2021. https://journals. sagepub.com/doi/abs/10.1177/14624745211043543 (accessed 19 April 2022). Rubin A and Phelps MS (2017) Fracturing the penal state: state actors and the role of conﬂict in penal change. Theoretical Criminology 21(4): 422–440. Ruhland EL (2020) Philosophies and decision making in parole board members. The Prison Journal 100(5): 640–661. Sanders E (2006) Historical institutionalism. In: Rhodes RAW, Binder SA and Rockman BA (eds) The Oxford Handbook of Political Institutions. Oxford: Oxford University Press, pp. 39–55. Schartmueller D (2019) How long is life? Comparing the processes of release for life- imprisoned offenders in Denmark, Finland, and Sweden. European Journal of Criminal Policy Research 25: 391–408. Seeds C (2017) Bifurcation nation: American penal policy in late mass incarceration. Punishment & Society 19(5): 590–610. Seeds C (2021) Life sentences and perpetual conﬁnement. Annual Review of Criminology 4: 287–309. Shute S (2003) The development of parole and the role of research in its reform. In: Zedner L and Ashworth A (eds) The Criminological Foundations of Penal Policy: Essays in Honour of Roger Hood. Oxford: Oxford University Press, pp. 377–442. Simon J (1993) Poor Discipline: Parole and the Social Control of the Underclass, 1890– 1990. Chicago: University of Chicago Press. Steinmo S, Thelen K and Longstreth F (eds) (1992) Structuring Politics: Historical Institutionalism in Comparative Analysis. New York: Cambridge University Press. References van Zyl Smit D and Corda A (2018) American Exceptionalism in parole release and supervi- sion: A European perspective. In Reitz KR (Ed) American Exceptionalism in Crime and Punishment. New York: Oxford University Press, pp. 410–486. Dr Thomas Guiney is an Assistant Professor of Criminology at the University of Nottingham. Tom’s most recent research has focused on questions concerning penal policy-making, the politics of law and order, and the historical development of parole and other ‘early release’ mechanisms.

Subsets and Splits

No community queries yet

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