pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
35883225 | PMC9508847 | pmc | 8,292 | {
"abstract": "Abstract The budding yeast Saccharomyces cerevisiae has been used extensively in fermentative industrial processes, including biofuel production from sustainable plant-based hydrolysates. Myriad toxins and stressors found in hydrolysates inhibit microbial metabolism and product formation. Overcoming these stresses requires mitigation strategies that include strain engineering. To identify shared and divergent mechanisms of toxicity and to implicate gene targets for genetic engineering, we used a chemical genomic approach to study fitness effects across a library of S. cerevisiae deletion mutants cultured anaerobically in dozens of individual compounds found in different types of hydrolysates. Relationships in chemical genomic profiles identified classes of toxins that provoked similar cellular responses, spanning inhibitor relationships that were not expected from chemical classification. Our results also revealed widespread antagonistic effects across inhibitors, such that the same gene deletions were beneficial for surviving some toxins but detrimental for others. This work presents a rich dataset relating gene function to chemical compounds, which both expands our understanding of plant-based hydrolysates and provides a useful resource to identify engineering targets.",
"conclusion": "Conclusion We anticipate that the rich dataset of yeast cellular responses to hydrolysate toxins presented here will aid in future biofuel studies including strain engineering. Selecting gene targets may be most effective by considering the response of engineered strains like the gene-deletion lines studied here to multiple toxins, including those that may be present in the same hydrolysate. This data set may also provide a useful backdrop for modeling toxins in complex mixtures, by comparing chemical genomic footprints in those mixtures to the single inhibitors studied here.",
"introduction": "Introduction A major goal for sustainable bioenergy is to use non-edible plant biomass to produce renewable energy fuels and other chemical products by microbial factories. The budding yeast Saccharomyces cerevisiae has been used extensively in fermentative industrial processes, including biofuel production from sustainable plant-based hydrolysates (Ekas et al. 2019 , Nielsen 2019 ). Although a promising alternative to fossil fuels, there is still the need to decrease costs through improved efficiency of biomass conversion to useful products (Kumar and Kumar 2017 , Ekas et al. 2019 ). Two main bottlenecks challenge this improvement. First, Saccharomyces cerevisiae cannot natively ferment pentoses and oligosaccharides released from deconstructed plant biomass, thereby underutilizing a significant carbon fraction (Kricka et al. 2015 , Zhao et al. 2020 ). Second, toxins found in processed plant biomass are stressful to biofuel microbes, and stress responses mounted by cells redirect resources away from bioproduct formation (Palmqvist and Hahn-Hägerdal 2000 , Almeida et al. 2007 , Liu 2011 , Piotrowski et al. 2014 , Cunha et al. 2019 , Fletcher and Baetz 2020 ). Thus, a major goal in sustainable biofuel research is to engineer pentose-consuming microbes that are resilient to toxins derived from lignocellulose and pretreatment processes. Microbial inhibitors found in plant-based hydrolysates are derived from a number of different sources. One source of inhibitors is the biomass pre-treatment method, which can employ chemically-transformative conditions using acid, heat, ammonia, or solvents (Lau et al. 2009 , Singh et al. 2015 , Baruah et al. 2018 ). Emerging technologies include solvents that produce purified sugar streams, including gamma-valerolactone (GVL) (Alonso et al. 2013a , Luterbacher et al. 2014 ) and imidazolium ionic liquids (IIL) such as [C 2 C 1 im]Cl (also known as 1-ethyl-3-methylimidazolium chloride or EMIM-Cl) (Swatloski et al. 2002 , Socha et al. 2014 , Hou et al. 2017 ). Despite solvent recovery methods, residual concentrations of these solvents remain in purified carbohydrate streams at levels that significantly inhibit microbial growth and metabolism (Ouellet et al. 2011 ). A second class of inhibitors is produced via chemical reactions with the plant biomass (Palmqvist and Hahn-Hägerdal 2000 , Klinke et al. 2004 , Almeida et al. 2007 , Jönsson et al. 2013 ). The largest set includes phenolic compounds that are released during the breakdown of hemicellulose and lignin and comprise a diverse group of molecules including acids, aldehydes, and ketones (Palmqvist and Hahn-Hägerdal 2000 , Klinke et al. 2004 , Almeida et al. 2007 ). In contrast, the furans furfural and 5-hydroxymethylfurfural (5-HMF) are generated during acid pretreatment from the dehydration of pentoses and hexoses, respectively (Palmqvist and Hahn-Hägerdal 2000 , Almeida et al. 2007 ). Furans are found in hydrolysates to varying levels, with high concentrations found in GVL-based hydrolysates (Alonso et al. 2013b , Alonso et al. 2013a ). A final class of microbial inhibitors is the metabolic end products themselves, such as ethanol (EtOH), isobutanol (IBA) and other commodity chemicals and biofuels that are toxic at high concentrations (Carmona-Gutierrez et al. 2012 , Zhang et al. 2015 , Kuroda et al. 2019 , Mota et al. 2021 ). Engineering efforts to increase production therefore require concomitant strategies that increase cellular tolerance to those products. An added challenge to rational engineering of microbes is that hydrolysate composition can vary substantially from batch to batch, because the suite and concentrations of toxins can be impacted by the pretreatment method but also feedstock growth conditions, seasonal effects, and harvesting characteristics (Klinke et al. 2004 , Lau et al. 2009 , Chundawat et al. 2010 , Bunnell et al. 2013 , Jönsson and Martín 2016 , Ong et al. 2016 , Wehrs et al. 2020 ). Hydrolysate toxins can also exert combinatorial effects due to interactions among inhibitors. All these features complicate engineering efforts to produce microbial factories customized for specific hydrolysates. A deeper understanding of resistance mechanisms to hydrolysate toxins individually and in combination will be critical to producing flexible sets of strains appropriate for handling a variety of complex hydrolysate stresses. Several studies have investigated the response to particular hydrolysate toxins, including phenolic compounds (Fletcher and Baetz 2020 ), ionic liquids (Kumari et al. 2020 ), or various other classes of compounds (Jönsson and Martín 2016 , Kim 2018 , Li et al. 2022 ). Yet much remains unknown about mechanisms of toxicity and how to overcome them, especially for sets of toxins variably found together and under anaerobic conditions, which is preferred for the industrial production of fermentative biofuels. One approach successfully applied to understanding mechanisms of pharmaceutical and other drugs is chemical genomics, in which genes and pathways required to survive specific chemicals are identified via high-throughput interrogation of gene-deletion libraries (Winzeler et al. 1999 , Giaever et al. 2002 , Hillenmeyer et al. 2008 , Enserink 2012 , Roemer et al. 2012 , Giaever and Nislow 2014 ). Past chemical genomic screens revealed mechanisms of action from many compounds, implicating molecules as potential therapeutic drugs (Delneri 2010 , Ho et al. 2011 , Lee et al. 2014 , Silberberg et al. 2016 ) and contributing to our understanding of chemical compounds that impact lignocellulosic biofuel production and other industrial processes (Skerker et al. 2013 , Pereira et al. 2014 , Dickinson et al. 2016 , Bottoms et al. 2018 , Xue et al. 2018 , Fletcher et al. 2019 , Kuroda et al. 2019 ). Here, we used chemical genomics in S. cerevisiae to understand genes and processes required to survive anaerobic treatment with each of 34 inhibitory chemicals, including solvents used in pre-treatment, toxins generated during hydrolysis of plant material, and biofuel products that stress cells at high levels. The results identified classes of toxins based on chemical genomic profiles, suggested mechanisms of cellular defense, and revealed widespread antagonistic gene-deletion effects. Our results raise important considerations for strain engineering to mitigate variable inhibitor concentrations in different types of hydrolysates.",
"discussion": "Discussion Chemical genomics has been successfully applied in previous biofuel studies, helping to uncover new insights that foster microbial engineering for increased efficiency. Many prior studies used such approaches to investigate cellular defense mechanisms to hydrolysates or to specific compounds studied in isolation and under aerobic conditions (Pereira et al. 2014 , Dickinson et al. 2016 , Ong et al. 2016 , Bottoms et al. 2018 , Fletcher et al. 2019 ). Here, we undertook a broader, comparative approach to study a variety of lignocellulose-derived inhibitors, solvents, and biofuel products relevant to multiple pretreatment methods and under anaerobic conditions. We used a drug-sensitive strain library that allows to uncover a large number of genes involved in the adaptive response, and consequently, to reveal specific mechanisms of toxicity of the toxins studied. Despite the high sensitivity of this strain background, our validation experiments with a less sensitive strain confirm findings under the conditions used here. However, we note that strains can vary significantly in their tolerance profiles due to genetic background; understanding the influence of genetic background on toxin tolerance is an important and active area of research (Sardi and Gasch 2017 , 2018 , Cámara et al. 2022 ). We used concentrations of each inhibitor for a consistent growth among them, but differences in composition of inhibitors in hydrolysates from batch to batch could cause a variance in the severity of the stress for the same inhibitors. This approach allows for both a deeper understanding of defense mechanisms while developing a rich dataset for further analysis, in certain industrially relevant fermentative conditions. Our results have important implications on both cellular defense and engineering strategies. First, comparative analysis of a larger set of inhibitors helped to distinguish shared and unique mechanisms of stress. Clustering gene fitness contributions across many inhibitors helped to distinguish subsets of genes with shared, or opposing, effects across compounds including those with known mechanisms. For example, although the molecular details remain to be worked out, the shared responses suggested that maintenance of the plasma membrane potential as well as secretion, membrane/cell surface stress responses, pH, mitochondrial function, and lipid biosynthesis play important roles in IIL and cation tolerance. However, many of these responses were inversely important for surviving phenolic compounds, suggesting that these inhibitor classes are provoking opposing effects to the same cellular physiology. We propose that a key aspect of that physiology is influenced by ion and pH homeostasis, in opposing ways for tolerance to cationic compounds, which may induce alkalinization, and some phenolic compounds that acidify the cell. Cation influx is predicted to raise internal pH due to corresponding efflux of H + , and the requirement of the Rim101 alkaline-response regulon for tolerating these compounds is consistent with this notion (Ariño et al. 2010 ). IILs can insert into lipid bilayers, and permeability of IIL can be affected by differences in lipid composition, fluidity, and other properties (Gal et al. 2012 , Cook et al. 2019 ). Perturbations to lipid bilayer asymmetry can directly activate Rim101, underscoring the intimate relationship between membrane status and pH (Ikeda et al. 2008 , Obara and Kamura 2020 ). The inverse relationships extended to gene deletions sensitized to phospholipid flipase perturbation, including Neo1; ergosterol biosynthesis; and vesicle trafficking, all of which also influence vacuolar pH homeostasis and membrane biology (Brett et al. 2011 ). In contrast, phenolic compounds (especially phenolic acids) and other stresses may decrease cellular pH leading to acid stress. Many of these compounds may become deprotonated inside the cell, in a similar manner to other organic acids. Ethanol stress is known to acidify the cytosol due to plasma membrane permeabilization and H + influx (Madeira et al. 2010 , Lam et al. 2014 , Charoenbhakdi et al. 2016 ), thereby producing inverse dependencies on pH-related genes. The widespread antagonistic effects for certain mutants exposed to different classes of inhibitors raise some considerations for strain engineering. While previous strategies have successfully identified engineering strategies that improve tolerance of single inhibitors (Dickinson et al. 2016 , Bottoms et al. 2018 , Higgins et al. 2018 , Fletcher et al. 2019 , Kuroda et al. 2019 ), our results show the importance of considering that those genetic changes could affect resistance to other toxins in the same hydrolysates, producing a strain that is, in the end, less fit for the complex mixture. Hydrolysate composition is influenced by different factors such as the type of biomass, plant growth conditions, or pre-treatment methods (Cunha et al. 2019 ). Phenolics are produced in higher concentrations in AFEX-treated biomass than acid hydrolysates, including specific phenolic compounds, such as amides (Keating et al. 2014 ). Our results showed some differences between amides and ketones when compared to other phenolics, which would also imply variable responses according to the pre-treatment method. The antagonistic analysis between couples of inhibitors also showed the importance of selecting the appropriate pre-treatment according to the desired product to decrease negative effects during the fermentation. An alternative strategy may be to engineer suites of strains with resistance to sets of inhibitors that co-fluctuate across hydrolysate types or batches, and to choose the best strain for each batch of hydrolysate based on their composition and the desired product."
} | 3,554 |
38107944 | PMC10719924 | pmc | 8,294 | {
"abstract": "Wide temperature tolerance and superior mechanical properties\nare\nhighly required for composite hydrogels in electronic applications\nsuch as electronic skins and soft robotics. In this work, a unique\npolyacrylamide-based and double-network hydrogel system is designed\nand fabricated by introducing graphene oxide and glycerol to improve\nmechanical properties as well as antifreezing and antiheating properties.\nMaximum stress of the graphene oxide-incorporated hydrogel increases\nrapidly to 500.0 kPa which is much higher than that of polymetric\nacrylamide/carboxymethylcellulose sodium hydrogel (281.7 kPa), probably\ndue to the inhibition from graphene oxide in generation and propagation\nof cracks. With constantly adding glycerol, total elongation and antifreezing\nand heating properties of the composite hydrogels increase gradually.\nEspecially, sample with 20 vol % of glycerol not only shows stable\nconductivity and wide temperature tolerance (−50 to 50 °C)\nbut also has ideal strength-toughness match (597.6 kPa and 1263.4%),\nsuggesting that synergistic effect of different layers in the asymmetric\nstructure plays an active role in improvement of mechanical properties.",
"conclusion": "4 Conclusions In this work, novel asymmetric\nand conductive PAAm-based hydrogels\nwere designed, and their structure and multiple properties were studied\nand discussed. First, PAAm/CMC hydrogels with 15% and 16% solid contents\nshow a DN structure and smooth surface with cracks after cold-drying.\nThe morphology of the hydrogels is rough with few cracks after introducing\nGO/glycerol. Second, the maximum stress of the PAAm/CMC-GO hydrogel\nincreases to 500 kPa, which is much higher than that of the PAAm/CMC\nhydrogel (281.7 kPa), while both are low in total elongation. The\nPAAm/CMC-GO-gly 20% not only has outstanding total elongation\n(1263.4%) after addition of 20 vol % glycerol, but also shows a higher\nstress (597.6 kPa), due to the asymmetric structure and suture function\nfrom GO. With further increase in the content of glycerol, the maximum\nstress of PAAm/CMC-GO-gly 40% decreases to 339.0 kPa. In\naddition, the crystallization temperature of the samples gradually\ndecreases from −21 °C to below −50 °C as the\nvolume fraction of glycerol increases from 10 to 20%. As for the PAAm/CMC-GO-gly 20% hydrogel, for one thing, the total elongation and maximum\nstress reach to 975% and 719.6 kPa at 50 °C, respectively. For\nanother, the hydrogel shows stable conductivity and outstanding sensitivity\non monitoring small changes in the process of tensile test and various\nhuman motions and small tensile stress. In summary, the PAAm/CMC hydrogels\nhave great capability to be used as a strain sensor for detecting\nsubtle strain changes over a wide temperature range.",
"introduction": "1 Introduction Electronic skins (e-skins) 1 − 4 have aroused great attention of many scientists and\nbeen used in different applications such as artificial intelligence\nsystems and wearable health care devices. 5 − 10 To simulate functions of real skins, it is essential for e-skins\nto have good mechanical properties, high conductivity, and antifreezing\nand antiheating properties. In recent years, hydrogels have been considered\nas one of the best candidates for preparation of e-skins owing to\ntheir tissue-mimicking features. 11 , 12 To endow hydrogels\nwith high conductivity, inorganic nanofillers (such as metal ions 13 , 14 and nanomaterials 15 ) have been employed\nto make e-skins respond to different stimuli rapidly. 16 , 17 This is the reason that e-skins have great potential in monitoring\nphysiological signals and developing interactive robots with multiple\nfunctions. 18 , 19 According to many reports, output\nelectronic signal of hydrogels can successfully change under tension\nor compression voice conditions. 20 , 21 To detect\nand collect weak signals, Bao’s group recently designed and\nprepared a stretchable all-polymer light-emitting diode with high\ncurrent efficiency, achieving signal transfer under low voltage and\nfurther promoting industrial application of e-skins. 22 To date, skin-mimicking hydrogels have been developed\nbased on\nboth synthetic polymers and natural polymers. Among them, polyacrylamide\n(PAAm)-based hydrogels have shown great advantages due to their low\nprice, ease of production, and good biocompatibility. 23 , 24 However, maximum stress of PAAm-based hydrogels were generally lower\nthan 100 kPa, 23 , 25 , 26 which makes them difficult to use in applications that demand good\nmechanical properties. Many efforts have been made to improve mechanical\nproperties of PAAm-based hydrogels. Gong’s group first proposed\nconstruction of a double-network (DN) structure to effectively improve\nthe maximum stress of hydrogels. 27 The\nDN structure consisted of a first network with good stiffness and\na second network with good toughness, providing effective energy dissipation.\nSubsequently, many hydrogels including PAAm-based ones with DN structures\nwere designed by using this mechanism and promoted to become good\ncandidates for application. 28 − 31 Zheng et al. successfully prepared a DN hydrogel\nbased on agar and PAAm by a one-pot method. 32 The DN agar/PAAm hydrogel showed a tensile strength of 1.0 MPa and\na fraction strain of 2000%, while these values were only 0.3 MPa and\n∼1800% for PAAm hydrogels. However, the DN structure for most\nhydrogels still had difficulty meeting requirements under high tension,\nwhich made multiple strengthen mechanism necessary. Introducing nanomaterials\nincluding carbon nanotubles (CNTs), silver nanowires (AgNWs), and\ngraphene oxides (GO) into hydrogels can not only increase tensile\nstress, but also obtain excellent properties, such as mechanical properties\nand conductivity. 4 , 15 , 33 Moreover, structures of previously reported e-skins and hydrogels\nwere almost homogeneous, 31 , 34 − 36 showing only single strain mechanism during deformation. Inspired\nby building hierarchical microstructure of alloys to improve their\ntensile performance, 37 the asymmetric structure\nwith different sublayers was designed and achieved collaboration between\nlayers to adapt to deformation. Significantly, asymmetric structure\nis consistent with the real skin. What we all know is that skins are\nroughly divided into three layers: epidermis, dermis and subcutaneous\ntissue, and different layers play different roles. For example, epidermis\nprovide good strength, while dermis has excellent toughness. In addition,\nreal skins can work within a wide temperature ranging from −10\nto 50 °C, which requests the e-skin to have antifreezing and\nantiheating properties. In this work, a high-performance strain\nsensor with temperature\ntolerance was developed to face the challenges of PAAm-based e-skin,\nincluding low stress and loss of water. This work is mainly engaged\nin the following sections: First, a novel DN PAAm/carboxymethylcellulose\n(CMC) hydrogel incorporated with nanoscale GO was designed and prepared\nto increase the strength of the PAAm-based hydrogel. Then, the hydrogel\nwas able to persistently work at both low and high temperatures after\nthe addition of glycerol, owing to the fact that hydrogen bonding\ncould improve the hydrogels’ ability to lock water. In addition,\nan asymmetric structure simulating real skins was obtained by tuning\nthe gradient distribution of GO under gravity, encouraging each layer\nto work in conjunction and improving the tensile performance of the\nhydrogel.",
"discussion": "3 Results and Discussion 3.1 Structure Analysis of the PAAm/CMC Hydrogels 3.1.1 Appearance and Swelling Property of the\nHydrogels A DN hydrogel with good temperature tolerance is\ndesigned to obtain an ideal strength-toughness match by tuning the\ndistribution of GO. APS and MBA were added to an AAm solution to initiate\npolymerization into PAAm and form a rigid network structure. The other\nflexible network structure was made of CMC + Ca 2+ chelate,\nwhich was connected to the PAAm network through hydrogen bonding.\nGO displayed a gradient distribution under gravity and had an asymmetric\nstructure in the hydrogel matrix, as shown in Figure 1 . Figure 1 Schematic evolution diagram from the PAAm/CMC\nhydrogel to PAAm/CMC-GO-glycerol\nhydrogels with an asymmetric structure. Table 3 shows the\nappearance of PAAm/CMC hydrogels with different solid contents (SC).\nThe appearance of samples with no more than 8 wt % SC exhibited instability,\nsuggesting the strength of chemical network was weak. With SC increased\nto 10–20 wt %, the appearance of hydrogels became more stable,\nleading to a gradual increase in strength and decrease in toughness.\nThe hydrogels with SC of 15 and 16 wt % displayed good strength and\ntoughness, explaining the reason for choosing the SC of 15 wt % for\nfurther research in this work. Table 3 Appearance of the PAAm/CMC Hydrogels\nwith 5, 8, 10, 13, 15, 18 and 20% Solid Content solid content 5, 8% 10, 13% 15, 16% 18, 20% appearance of hydrogel unstable lower strength stable low strength stable good toughness-strength match stable toughness\ndecreasing To clearly show the feasibility of the above design,\nthe PAAm/CMC\nhydrogel and the PAAm/CMC hydrogel containing GO and 20 vol % of glycerol\nwere prepared ( Figure 2 (a,b)). The former sample is transparent and homogeneous, while the\nlatter one with GO and glycerol displayed gradient distribution of\nGO in the direction of gravity and has asymmetric structure, tallying\nwith the predesign. Moreover, Figure 2 (c) displays the weight ratio of PAAm/CMC, PAAm/CMC-GO,\nPAAm/CMC-GO-gly 10% , and PAAm/CMC-GO-gly 20% hydrogels\nduring the swelling test. Values of Q rapidly increased\nin the beginning, and then became more and more slow until the slope\nwas close to 0. Images recorded on day 0, day 1, and day 17 are provided\nin Figure 2 (d–f),\nrespectively. The diameter of the four samples became bigger and bigger,\nand this phenomenon is evident on the first day, which is consistent\nwith the curves. Also, the weight ratio of the PAAm/CMC-GO hydrogel\nis lower than that of the PAAm/CMC hydrogel. However, with the increase\nof the volume fraction of glycerol, the change of the diameter and\nweight is more significant. The results suggest the cross-link indexes\nof PAAm/CMC and PAAm/CMC-GO hydrogels are higher than those of PAAm/CMC-GO-gly 10% and PAAm/CMC-GO-gly 20% hydrogels, due to the\ninfluence of hydrogen bonding. Figure 2 Images of (a) PAAm/CMC and (b) PAAm/CMC-GO-gly 20% hydrogels;\nSwelling curves (c) in 17 days of PAAm/CMC, PAAm/CMC-GO, PAAm/CMC-GO-gly 10% and PAAm/CMC-GO-gly 20% hydrogels and images\non (d) day 0, (e) day 1, and (f) day 17. PAAm/CMC, PAAm/CMC-GO, PAAm/CMC-GO-gly 10% and PAAm/CMC-GO-gly 20% hydrogels are shown from\nleft to right in (d–f). 3.1.2 Chemical Structure and Microstructure of\nthe PAAm/CMC Hydrogels To further determine the chemical\nstructure of the PAAm/CMC hydrogels, FTIR was used to measure the\nAAm powder, PAAm, CMC+CaCl 2 , and PAAm/CMC-GO-gly 20% hydrogels as shown in Figure 3 (a). The FTIR spectroscopy of AAm (purple curve) displays\ntypical characteristic absorption peaks of C=C (1680 and 1750\ncm –1 ) and –NH 2 (3201 and 3415\ncm –1 ). The C=C and –NH 2 peaks disappear in the cyan line, indicating that APS and MBA participated\nin the reaction with AAm. Moreover, the absorption peaks of PAAm/CMC-GO-gly 20% have no new complex peak except for a wide −OH (3,320\ncm –1 ) peak, suggesting that the PAAm/CMC hydrogels\nform a DN structure. Figure 3 (a) FTIR spectra of AAm, PAAm, CMC + CaCl 2 and\nPAAm/CMC-GO-gly 20% hydrogels; typical SEM images of the\ntop surface of (b)\nPAAm/CMC, (c) PAAm/CMC-GO, and (d) PAAm/CMC-GO-gly10% hydrogels after\nfreeze-drying. Figure 3 (b–d)\ndisplays the morphology of PAAm/CMC, PAAm/CMC-GO, and PAAm/CMC-GO-gly 10% hydrogels, respectively. Compared to the other micropore\nsurface in previous reports, 15 , 23 the surface of the\nPAAm/CMC hydrogel exhibits a smooth plane and a long fiber with cracks,\ndue to lower solid content. After addition of GO, a rough plane and\na few cracks are widely distributed on the hydrogel. Moreover, a rough\nplane and no obvious cracks are observed in PAAm/CMC-GO-gly 10% hydrogel under the combined effect of GO and glycerol. 3.2 Mechanical Properties of the PAAm/CMC Hydrogels\nat Room Temperature 3.2.1 Tensile Performances of the PAAm/CMC Hydrogels Figure 4 (a) shows\ntensile curves of the PAAm/CMC, PAAm/CMC-GO, PAAm/CMC-GO-gly 20% , and PAAm/CMC-GO-gly 40% hydrogels. Both PAAm/CMC and\nPAAm/CMC-GO hydrogels without glycerol have similar total elongation\n(782.3 and 807.7%), but obvious difference in maximum stress (281.7\nand 500.0 kPa). Compared with the mechanical properties of the PAAm/CMC-GO\nhydrogel, the total elongation and maximum stress of the PAAm/CMC-GO-gly 20% hydrogel rapidly go up to 1263.4 and 597.6 kPa. However,\nmany literature studies 38 − 40 have reported that the stress\nof hydrogels with homogeneous composition gradually decreased as glycerol\nwas constantly added. It is worth noting that the PAAm/CMC-GO-gly 20% hydrogel has an asymmetric structure that leads to different\nmechanical properties from those with homogeneous structure. To be\nspecific, the layer with more GO makes the hydrogel undertake a great\ntension, and the one with less GO is helpful for total elongation,\nrealizing that all layers work together and obtaining excellent strength\ntoughness in the progress of tensile test. As the volume fraction\nof glycerol increases to 40%, the hydrogel shows a higher total elongation\n(1845.6%) but significantly decreased stress (339.0 kPa), which is\nclosely related to the nondense network structure accepted by hydrogen\nbonding. Figure 4 (a) Tensile stress–strain curves of PAAm/CMC, PAAm/CMC-GO,\nPAAm/CMC-GO-gly 20% and PAAm/CMC-GO-gly 40% hydrogels\nat room temperature; (b) different dissipated energy and the hysteresis\ncurves of PAAm/CMC-GO-gly 20% hydrogel during loading and\nunloading cycles were studied (50, 100, 300, 500%); (c) six times\ncontinuing cyclic curves of PAAm/CMC-GO hydrogel at 100% strain and\ncorresponding dissipated energy; dynamic rheological curves of PAAm/CMC,\nPAAm/CMC-GO, PAAm/CMC-GO-gly 20% and PAAm/CMC-GO-gly 30% hydrogels under (d) strain and (e) frequency, respectively. The fatigue resistance of the PAAm/CMC-GO-gly 20% hydrogel\nwas studied by tensile cycle tests with different strains as shown\nin Figure 4 (b). With\nthe increase of tensile strain, hysteresis loops increase obviously,\nsuggesting that the cross-linking points are severely damaged. Figure 4 (c) displays the\ncurves of 6 times of tensile cycles without residence time. The hysteresis\ncurve and hysteresis energy of the following cycles are slightly lower\nthan that of the first cycle but tend toward stability, indicating\nthat the hydrogel containing GO and glycerol possesses fatigue resistance\nand rapid self-recovery. 3.2.2 Rheological Performances of the PAAm/CMC\nHydrogels at Room Temperature To reveal the microscopic mechanical\nperformance of hydrogels after introducing GO and glycerol, their\ndynamic storage modulus ( G ′) and loss modulus\n( G ″) are measured with different frequency\nsweeps in Figure 4 (d).\nBoth G ′ and G ″ values\nof the PAAm/CMC-GO hydrogel are higher than those of the PAAm/CMC\nhydrogel over the entire frequency range. Also, both hydrogels show\nthe same result that G ′ is much higher than G ″ over the frequency range, which is identical with\nthe solid-like, elastic nature of hydrogels. In addition, the values\nof G ′ and G ′/ G ″ dramatically increase for the PAAm/CMC-GO-gly 20% hydrogel, attributed to the strong interface interaction\nfrom asymmetric structure. G ′ and G ″ values decrease gradually as the volume fraction\nof glycerol rises, which is consistent with tensile loading–unloading\ntests. A significant strain-dependent viscoelastic response for all\nsamples could be observed in the curves ( Figure 4 (e)) for G ′ and G ″ as a function of shear strain (γ). In the\nlinear viscoelastic region of γ from 0.1 to 10%, the G ′ and G ″ of PAAm/CMC-GO-gly 20% , PAAm/CMC-GO-gly 30% , PAAm/CMC-GO, and PAAm/CMC\nhydrogels decrease in sequence, suggesting that an appropriate content\nof glycerol and GO is capable of effectively optimizing mechanical\nperformances. However, the hydrogels exhibit slop-droop nonlinear\nviscoelastic behavior at an else shear strain range from 10 to 1000%,\nand the modulus of the hydrogels without glycerol decreases slowly.\nThis phenomenon indicates that the DN incorporation is easily induced\nby the shear strain. 3.2.3 Microstructure of the PAAm/CMC-GO Hydrogel\nNear Cracks Generated by the Tensile Process Figure 5 (a,c) presents typical SEM\nimages showing the tensile fracture-surface morphology of the PAAm/CMC-GO\nhydrogel. Some whiskers are observed in cracks that are scattered\nin the plane and short fiber-shaped region. To reveal the clear morphology\nof whiskers, the higher-magnification SEM images taken from the region\nmarked in Figure 5 (a,c)\nare displayed in Figure 5 (b,d), respectively. The nanoscale GO that the orange arrows point\nout sew up the cracks like a thread, leading to improved strength\nof the hydrogels. Figure 5 (a, c) Typical SEM morphology of tensile fracture-surface\nof the\nPAAm/CMC-GO hydrogel; (b, d) higher-magnification SEM images of PAAm/CMC-GO\nnear cracks. 3.3 Temperature-Tolerance Performance and Mechanical\nProperties of PAAm/CMC Hydrogels 3.3.1 Antifreezing and Antiheating Performances\nof the PAAm/CMC, PAAm/CMC-GO, PAAm/CMC-GO-gly 10% , and PAAm/CMC-GO-gly 20% Hydrogels To preliminarily understand the antifreezing\nproperties of the hydrogels, PAAm/CMC, PAAm/CMC-GO, PAAm/CMC-GO-gly 10% , and PAAm/CMC-GO-gly 20% hydrogels are placed\nin a refrigerator at −25 °C for 8 days, as shown in Figure 6 (a,b). It is observed\nthat the PAAm/CMC-GO-gly 20% hydrogel still keeps its original\nstate, while others are frozen after 8 days. To further explore the\nantifreezing performance, the crystallization temperature of the hydrogels\nis determined by DSC analysis. As shown in Figure 6 (c), the crystallization peak temperatures\nof PAAm/CMC and PAAm/CMC-GO hydrogels are −1.5 and −4.3\n°C, respectively. The crystallization peak temperature of the\nhydrogel is −21.4 °C after addition of 10 vol % glycerol,\ndue to the addition of hydrogen bond lowering their crystallization\ntemperature. As the volume fraction of glycerol reaches 20 vol %,\nthe temperature drops to below −50 °C, and the result\nwas consistent with the phenomenon shown in Figure 6 (a,b). Figure 6 Antifreezing images of PAAm/CMC, PAAm/CMC-GO,\nPAAm/CMC-GO-gly 10% and PAAm/CMC-GO-gly 20% hydrogels\ntesting at\n−25 °C for (a) 0 day and (b) 8 day; (c) DSC measurements\nin the range from −50 to 20 °C; (d) antiheating curves\nat 50 °C in 8.5 h; tensile stress–strain curves of PAAm/CMC-GO-gly 20% and PAAm/CMC-GO-gly 40% hydrogels (e) at −20\n°C and (f) at 50 °C. It is difficult for most hydrogels to avoid water\nloss at room\ntemperature or elevated temperature, restricting further application\nof the materials. A universal and effective way to solve this problem\nis introducing hydrogen bonding, which can be damaged under continuously\nabsorbing energy and is finally helpful to slow down evaporation.\nThe curves of Q W at 50 °C are shown\nin Figure 6 (d) and\nused to display the antiheating performance of all hydrogels. Compared\nwith the hydrogels without glycerol, the Q W value of the PAAm/CMC-GO-gly 10% hydrogel is lower, suggesting\nthat glycerol has a positive effect on preventing evaporation. As\nglycerol increases to 20 vol %, the Q W value further decreases. 3.3.2 Tensile Performance of PAAm/CMC-GO-gly 20% and PAAm/CMC-GO-gly 40% Hydrogels at Low and\nHigh Temperatures According to the above results, PAAm/CMC-GO-gly 20% and PAAm/CMC-GO-gly 40% hydrogels are chosen\nto study antifreezing and antiheating properties since the other samples\nare difficult to be applied in lower or higher temperature. Furthermore, Figure 6 (e,f) displays the\ntensile testing curves of the two hydrogels at −20 and 50 °C,\nrespectively. Compared with the same samples at room temperature shown\nin Figure 4 (a), the\ntotal elongation and maximum stress of the PAAm/CMC-GO-gly 20% (1503.7%, 558.5 kPa) and PAAm/CMC-GO-gly 40% (1874.1%,\n342.3 kPa) hydrogels show no significant change at −20 °C,\nindirectly confirming that glycerol is helpful to reduce crystallization\ntemperature. When PAAm/CMC-GO-gly 20% and PAAm/CMC-GO-gly 40% hydrogels are tested at 50 °C, the total elongations\nof PAAm/CMC-GO-gly 20% and PAAm/CMC-GO-gly 40% hydrogels fall (975.0 and 1009.5%), while the maximum stress rises\n(719.6 and 403.6 kPa), due to the fact that losing water makes the\ndouble-network structure become denser, and the hydrogels still keep\ngood strength-toughness match and are still applied. 3.4 Sensing Properties of PAAm/CMC-GO-gly 20% Based on the overall performances of real skin\nwith good mechanical properties, stable conductivity, and outstanding\nsensitivity, the PAAm/CMC-GO-gly 20% hydrogel simulating\nreal skin has enormous potential for monitoring various human motions.\nAs shown in Figure 7 (a,b), the relative resistance changes (Δ R / R 0 ) of the hydrogel covering the knuckle\nand knee show the corresponding regular changes with a reciprocating\njoint bending angle, which is caused by the length and cross-section\narea fluctuation of the hydrogel during the joint movement. Furthermore,\nthe hydrogel transducer possesses the ability of speech recognition;\nfor instance, different current signal patterns could be detected\nwhen the tester uttered different voices such as “Jia”,\n“Xing”, “Xue”, “Yuan”, and\n“Asymmetric”; thus, the hydrogel sensor could be used\nfor language rehabilitation and recognition of language-disabled people\n( Figure 7 (c,d)). Figure 7 Resistance\nvariation of PAAm/CMC-GO-gly 20% hydrogel\nresponds to the different bending angles of (a) knuckle; (b) knee;\n(c, d) detection of speaking when the hydrogel mounts on throat. Relative\nresistance of PAAm/CMC-GO-gly 20% changes in response to\n(e) the tensile strain varied from 1 to 300% and (f) the tensile speed\nof 80, 150, and 300 mm/min. Figure 7 (e,f) shows\nthe Δ R / R 0 of PAAm/CMC-GO-gly 20% hydrogel at different tensile strains (1–300%) and\nspeed (80–300 mm/min), respectively. The repeatable and stable\nsignals could be monitored under different strains and speed. Especially,\nthe hydrogel is extremely sensitive to both small strain and fast\nspeed, and superb reliability is exhibited during the repeated stretching-produced\nstable output signals."
} | 5,639 |
28553264 | PMC5425589 | pmc | 8,300 | {
"abstract": "Our understanding of the diverse interactions between hosts and microbes has grown profoundly over the past two decades and, as a product, has revolutionized our knowledge of the life sciences. Through primarily laboratory experiments, the current framework for holobionts and their respective hologenomes aims to decipher the underpinnings and implications of symbioses between host and microbiome. However, the laboratory setting restricts the full spectrum of host-associated symbionts as compared to those found in nature; thus, limiting the potential for a holistic interpretation of the functional roles the microbiome plays in host biology. When holobionts are studied in nature, associated microbial communities vary considerably between conditions, resulting in more microbial associates as part of the “hologenome” across environments than in either environment alone. We review and synthesize empirical evidence suggesting that hosts may associate with a larger microbial network that, in part, corresponds to experiencing diverse environmental conditions. To conceptualize the interactions between host and microbiome in an ecological context, we suggest the “host-associated microbial repertoire,” which is the sum of microbial species a host may associate with over the course of its life-history under all encountered environmental circumstances. Furthermore, using examples from both terrestrial and marine ecosystems, we discuss how this concept may be used as a framework to compare the ability of the holobiont to acclimate and adapt to environmental variation, and propose three “signatures” of the concept."
} | 406 |
35822790 | PMC9264390 | pmc | 8,302 | {
"abstract": "Environmental problems associated with marine pollution and climate warming create favorable conditions for the penetration and survival of pathogenic bacteria in marine ecosystems. These microorganisms have interspecific competitive interactions with marine bacteria. Co-culture, as an important research strategy that mimics the natural environment of bacteria, can activate silent genes or clusters through interspecies interactions. The authors used modern biotechnology of co-cultivation to dynamically study intercellular interactions between different taxa of bacteria—pathogenic enterobacteria Yersinia pseudotuberculosis and Listeria monocytogenes and saprotrophic marine bacteria Bacillus sp. and Pseudomonas japonica isolated in summer from the coastal waters of the recreational areas of the Sea of Japan. The results of the experiments showed that during the formation of polycultural biofilms, horizontal transfer of genes encoding some pathogenicity factors from Y. pseudotuberculosis and L. monocytogenes to marine saprotrophic bacteria with different secretion systems is possible. It was previously thought that this was largely prevented by the type VI secretion system (T6SS) found in marine saprotrophic bacteria. The authors showed for the first time the ability of marine bacteria Bacillus sp. and P. japonica to biofilm formation with pathogenic enterobacteria Y. pseudotuberculosis and L. monocytogenes , saprophytic bacteria with type III secretion system (T3SS). For the first time, a marine saprotrophic strain of Bacillus sp. Revealed manifestations of hyaluronidase, proteolytic and hemolytic activity after cultivation in a polycultural biofilm with listeria. Saprotrophic marine bacteria that have acquired virulence factors from pathogenic enterobacteria, including antibiotic resistance genes, could potentially play a role in altering the biological properties of other members of the marine microbial community. In addition, given the possible interdomain nature of intercellular gene translocation, acquired virulence factors can be transferred to marine unicellular and multicellular eukaryotes. The results obtained contribute to the paradigm of the epidemiological significance and potential danger of anthropogenic pollution of marine ecosystems, which creates serious problems for public health and the development of marine culture as an important area of economic activity in coastal regions.",
"conclusion": "4. Conclusions Thus, during the experiment, we found: The strain of Bacillus sp. after co-cultivation with Listeria acquired hemolytic, plasma-coagulase and hyaluronidase activities. When this saprophyte was co-cultivated with Yersinia , these pathogenic properties did not appear in it. Yersinia, after co-cultivation with Bacillus sp., appeared in plasma coagulase activity. The adhesiveness index of Yersinia after cultivation in a mixed biofilm with Bacillus sp. has decreased, while cultivation with the P. japonica 3P9 strain, on the contrary, led to an increase in this index in Yersinia . For L. monocytogenes 870, a partial ability to lyse erythrocytes and the appearance of hyaluronidase activity was noted when co-cultivated in a biofilm with Bacillus sp. The adhesiveness of Listeria in a monofilm was 3.75 ± 0.05, but after cultivation with Bacillus sp., it reached 4.1 ± 0.13. The P. japonica strain did not show potential pathogenic properties both in the monofilm and after co-cultivation with pathogens. The results of the experiments showed the possibility of transferring virulent properties from enterobacteria to saprotrophic marine bacteria Bacillus sp., as well as a decrease in the severity of some pathogenicity factors in the tested pathogens (and a reduction of Yersinia adhesion after cultivation in a mixed biofilm with Bacillus sp.). A change in the pathogenic properties of Listeria and Yersinia was revealed—the appearance of the ability to lyse erythrocytes, plasma-coagulase activity and an increase in the adhesiveness of Listeria after cultivation with Bacillus sp. and in Yersinia after co-cultivation with P. japonica. Marine bacteria make up the largest part of the biomass in the oceans and play a key role not only in the microbial community, but also in the functioning of the entire ecosystem. Environmental problems associated with sea pollution and global warming create favorable conditions for pathogenic bacteria to enter and remain viable in marine ecosystems. In addition to the epidemiological threat, these microorganisms have competitive interactions with marine bacteria [ 35 ]. Our experimental studies have shown that the co-cultivation of pathogenic enterobacteria and nonpathogenic saprotrophic marine bacteria in a biofilm can lead to the horizontal transfer of genes encoding pathogenicity factors between non-closely related species of prokaryotes with different secretion systems. Previously, it was believed [ 36 ] that this is largely prevented by the use of the type VI secretion system (T6SS) in marine saprotrophic bacteria, which is one of the least studied competitive strategies among heterotrophic bacteria living in marine ecosystems. The authors have shown for the first time the ability of marine bacteria Bacillus sp. and P. japonica for biofilm formation with pathogenic enterobacteria Y. pseudotuberculosis and L. monocytogenes , saprophytic bacteria with type III secretion systems (T3SS). For the first time, using traditional microbiological methods, not only the possibility of co-cultivation of pathogenic and marine bacteria with different secretion systems in the biofilm, but also the possibility of intercellular horizontal transfer of some virulence properties from pathogenic enterobacteria to marine saprotrophic bacteria, was shown. Being part of a biofilm formed by various strains of microorganisms, bacteria can contribute to better adaptation and survival of pathogenic bacteria in the marine environment and influence the formation of pathogenic properties in saprophytic marine bacteria. To further elucidate the mechanism of horizontal transfer of pathogenicity factors to marine prokaryotes, it is necessary to use additional conditions and methodological approaches, for example, third-generation genome-wide sequestration technology and genetic editing based on CRISPR-Cas9, which will help confirm the fact of horizontal gene transfer and identify the involved mobile genetic elements. This understanding will allow us to confirm the importance of mobile elements in the evolution of marine microorganisms and a deeper understanding of the mechanisms of interaction, and to draw attention to the growing importance of the environmental and epidemiological problems of pollution of the World Ocean, which increases with climate warming on the planet.",
"introduction": "1. Introduction Bacteria co-cultivation technologies are widely used in microbial ecology for the dynamic study of intercellular interactions between different bacterial taxons, evolution and adaptation to changing environmental conditions. The possibility of horizontal transfer of not only antibiotic resistance genes but also those encoding certain pathogenicity factors in natural environments has attracted wide attention in recent years [ 1 , 2 ]. Recent studies have shown that horizontal gene transfer (HGT) and biofilm formation are interrelated processes in mixed cultures of bacteria. It has been established that the ability of microorganisms to form biofilms is one of the mechanisms of their adaptation to the effects of adverse environmental conditions. This mediates the preservation and maintenance of the population of pathogenic bacteria, including when it enters the marine environment with various domestic, industrial and other effluents [ 1 , 3 ]. In recent years, it has become increasingly clear that under natural conditions, biofilms are more often formed not by one but by several types of bacteria [ 4 ]. This increases the fundamental and practical significance of studying polycultural biofilms [ 5 ]. There is more and more evidence that the presence of microorganisms in a particular natural environment is determined not only by its conditions but also by the presence of control by other microorganisms. The role of signaling molecules produced by microorganisms acting as intra- or interspecies regulators of microbial interaction has been established. In addition, the significance of mobile genetic elements (MGE) in the intercellular HGT encoding of some bacterial functions was determined [ 2 , 3 , 6 ]. It has been established that microbial interactions, including horizontal genetic exchange, signaling, and metabolite exchange, occur between microorganisms in biofilm communities [ 1 , 7 , 8 , 9 ]. It has been found that within a biofilm, bacteria of different taxa form a single genetic system in the form of plasmids and other MGE that carry a behavioral code for members of the biofilm and determine their trophic, energy, and other connections between themselves and the outside world [ 2 , 4 , 9 , 10 ]. Microbial interaction in these consortia leads to complex relationships leading not only to growth suppression, but also to absorption, transfer/acquisition of DNA elements, leading to the emergence of new metabolic properties, properties in community members [ 1 , 3 , 4 , 6 ]. However, experimental studies of this plan are few, especially when pathogenic enterobacteria and marine bacteria are co-cultivated. In this regard, studies seem to be very significant, allowing us to establish: to what extent the marine environment is a “cemetery” of microorganisms and to what time it is a “bridgehead” for the formation of epidemic variants of pathogenic bacteria [ 7 , 10 ]. Yersinia pseudotuberculosis is a widespread pathogen of sapronous infections. Due to the two-phase paradigm of existence (in the body of warm-blooded organisms and the environment), the Listeria monocytogenes can lead parasitic and saprophytic lifestyles [ 8 , 11 ]. In the course of our previous studies [ 12 ], the possibility of preserving the pathogenic properties of L. monocytogenes and Y. pseudotuberculosis in marine ecosystems was experimentally shown, but also the ability of marine saprotrophs to modulate the growth activity of these enterobacteria [ 12 , 13 ]. In addition, an assumption was made about the possibility of HPG coding for pathogenicity factors from enteropathogens to marine bacteria (including factors that ensure the interaction of pathogens with the epithelium, persistence, and secretion of bacterial modulins and toxins). On the other hand, in recent years it has been established that marine bacteria have a type 6 secretion system (T6SS), which is actively used by them in interspecies competition. The labeling of bacteria containing T6SS implies their predatory behavior is associated not only with growth suppression, but also with the absorption of competing bacteria [ 3 ]. This work aimed to evaluate the possibility of transferring pathogenic characteristics from pathogenic enterobacteria ( Y. pseudotuberculosis and L. monocytogenes ) to marine saprotrophic bacteria after their co-cultivation.",
"discussion": "3. Research Results and Discussion 3.1. Selection of Strains for the Experiment, Their Identification In previous studies [ 12 ], we proved the ability of L. monocytogenes 870 and Y. pseudotuberculosis 3515 to form both monocultural and polycultural biofilms with marine saprotrophs. Bacillus sp. is a Gram-positive and while P. japonica is Gram-negative bacteria. Different types of bacteria (with different cell wall compositions and not closely related) were chosen for the experiment. Possible ways of transferring genes responsible for pathogenic properties and virulence (islands of pathogenicity, etc.) from pathogenic microorganisms to their related species are known in the literature. Changes in pathogenic properties after co-cultivation of pathogenic bacteria with saprotrophic microorganisms with cell walls of different structures were not found in the literature. In this regard, we decided to experimentally test the possibility of transferring pathogenic properties between such bacteria, so we used saprotrophic microorganisms of different species. We determined that the strains of marine microorganisms used in our work isolated from the waters of the Golden Horn Bay had the ability to biofilm formation (the average OD values and the standard deviation between the series of measurements in terms of the ability to biofilm formation were 0.530 ± 0.03 arb. units and 0.560 ± 0.02 arb. units, respectively). It allowed us to use them in our experiment. The strain was identified as Bacillus sp., and the strain was assigned to P. japonica . The assessment of the degree of formation of polycultural biofilms and the reproduction dynamics was carried out according to proven methods [ 27 , 28 ]. 3.2. The Study of Some Properties of Pathogenicity of the Studied Strains 3.2.1. Determination of Hemolytic Activity To determine the hemolytic activity, bacterial strains grown on each experiment day from a polycultural biofilm were used. Strains from monocultural biofilms served as a control for comparison. To differentiate marine saprotrophs from opportunistic bacteria, inoculation was carried out on differential diagnostic media indicated in the methods, and the individual formed colonies were microscopically examined. On Petri dishes with blood agar for the primary separation, all the studied microorganisms were seeded after cultivation in mono- or multicultural biofilm from each experiment day. After 24–48 h on the plates, the appearance of a transparent zone of hemolysis or the formation of a greenish-brown halo around the colonies was noted. In our experiment, erythrocyte lysis was characteristic of L. monocytogenes 870 from the biofilm with Bacillus sp. from the third day of cultivation in a joint biofilm ( Figure 3 B). Listeria itself in monoculture also did not have a visible zone of hemolysis on the third day of cultivation ( Figure 3 A). According to the literature, Listeria is characterized by β-hemolysis on blood agar [ 27 , 28 , 29 ]. Furthermore, cultivation with a saprotrophic marine strain was unfavorable for Listeria (for example, competition for nutrients, etc.) conditions. The strain of Bacillus sp. did not show any hemolytic properties in the monoculture ( Figure 3 C). However, for L. monocytogenes 870, a decrease in the ability to lyse erythrocytes was noted when co-cultivated in a biofilm with Bacillus sp. 3.2.2. Determination of Plasma-Coagulase Activity The next stage of the experiment was the determination of plasma-coagulase activity. According to the literature data [ 30 ], the plasma coagulase enzyme is present in Gram-negative bacteria, including Pseudomonas . However, in our experiment, the enzyme plasma coagulase was not detected in Pseudomonas either in mono- or polycultural biofilms. Yersinia showed plasma-coagulase activity after co-cultivation with Bacillus sp. In monoculture, Yersinia did not have plasma-coagulase activity. The plasma-coagulase activity was observed in both Gram-positive bacteria ( L. monocytogenes 870 and Bacillus sp.) after co-cultivation on the third day. 3.2.3. Determination of Hyaluronidase Activity The enzyme hyaluronidase can cleave mucopolysaccharides, which are part of marine plants and, therefore, are found in the marine environment [ 31 ]. Similar enzymes can be active in microorganisms in clean and polluted water bodies. The enzyme hyaluronidase, which destroys hyaluronic acid, was noted in Listeria both in a monofilm and after cultivation with Bacillus sp. on the third and fifth days. Hyaluronidase activity in Listeria was also recorded after co-cultivation with P. japonica. According to the literature, hyaluronidase activity is characteristic of Listeria (stable sign) [ 32 ]. The appearance of hyaluronidase activity in the saprotrophic strain of Bacillus sp. (3 days after co-cultivation with Listeria in biofilm). In monoculture, hyaluronidase activity in this microorganism was not observed. It once again confirms the possibility of the appearance of pathogenic properties in the saprotrophic strain of Bacillus sp. just after co-cultivation with Listeria . 3.2.4. Evaluation of the Adhesive Properties of the Studied Bacteria The assessment of the adhesive properties of the studied bacteria made it possible to establish that, according to the IAM indicators, the strain of Bacillus sp. was characterized as low adhesive. The IAM value was 1.8 ± 0.2 ( p < 0.05). After co-cultivation in a biofilm with L. monocytogenes 870, on the third day, the IAM was 8.3 ± 0.11, and on the fifth day, it was 8.0 ± 0.15, which significantly exceeds its value in the monofilm. According to the results of our study, the P. japonica strain turned out to be non-adhesive throughout the day of the experiment. The adhesiveness of Listeria in the monofilm was moderate (IAM 3.75 ± 0.05, at p < 0.05), but after cultivation with Bacillus sp. indicators of IAM exceeded 4.1 ± 0.13. So, on the third and fifth days of co-cultivation, IAM values were 5.12 ± 0.21 and 7.14 ± 0.12, respectively. In a consortium with P. japonica , the listeria adhesion index was 6.02 ± 0.16 on the third day of cultivation and 6.61 ± 0.22 on the fifth. The adhesion index of Yersinia in a monofilm was average (2.63 ± 0.05, p < 0.05). After cultivation in a mixed biofilm with Bacillus sp. strain became low-adhesive (IAM 1.81 ± 0.11). According to the literature, bacteria of the genus Yersinia , in particular Y. pseudotuberculosis , can lose or acquire pathogenicity islands in adaptation to unfavorable conditions [ 26 , 33 , 34 ]. Thus, cultivation with the P. japonica strain increased the adhesion index in Yersinia to an average value (3.12 ± 0.12) on the fifth day of cultivation [ 22 , 35 , 36 , 37 ]. Thus, cultivation with the P. japonica strain led to an increase in the adhesion index in Yersinia to an average value (3.12 ± 0.12) on the fifth day of cultivation. At the same time, during the experiment, a decrease in the abundance of the saprotrophic strain P. japonica was noted on the third and subsequent days after co-cultivation in a biofilm with the pathogens used in work. Coculture, as an important research strategy that mimics the natural environment of bacteria, can activate silent genes or clusters through interspecies interactions [ 38 , 39 ]. Unlike standard laboratory growing conditions, co-cultivation is not only an effective method for studying the biosynthesis of various secondary metabolites and enzymes of microorganisms but is important for revealing the mechanisms of interspecific competitive and symbiotic relationships between enteropathogens and marine bacteria, as well as new gene functions [ 40 , 41 , 42 ]. The results obtained are a clear illustration of the possibility of acquiring pathogenicity factors by marine saprotrophic bacteria after co-cultivation with pathogenic enterobacteria. These bacteria, due to pollution of the oceans and warming of coastal waters, inhabit marine ecosystems and interact with marine bacteria [ 38 , 39 ]. These interactions are mediated by a wide range of mechanisms and very often involve the secretion of various molecules from bacterial cells. The results of the experiments were obtained using routine microbiological studies, but they are quite convincing. Much has been written in recent years about the type VI secretion system (T6SS) found in marine bacteria as an effective weapon of interbacterial competition [ 43 , 44 , 45 ]. Over the past two decades, there has been a shift in microbiological research towards the use of systems approaches to study the interactions between diverse organisms and their communities in an ecological context. Our results show that the presence of T6SS in marine bacteria is not an obstacle to the horizontal transfer of virulence factors from pathogenic enterobacteria that have entered marine ecosystems. As a possible mechanism, first of all, it is necessary to assume the horizontal transfer of mobile genetic elements encoding the secretion of new proteins, enzymes that change the biological properties of marine bacteria and mediate the acquisition of virulence factors by them. Intercellular possible strategies for genetic exchange include plasmid-mediated conjugation, bacteriophage transduction, and transformation via bacterial uptake of extracellular DNA. Here, we will focus on this aspect of the acquisition of new biochemical properties by marine bacteria. Future studies on the implications of co-culture of enteropathogens and marine bacteria using whole-genome sequencing (WGS) molecular techniques will shed light on the transgenic mechanism."
} | 5,225 |
32972943 | PMC7516154 | pmc | 8,303 | {
"abstract": "The plant soil rhizobiome induces critical functions in the plant proximal environment. Linkages between soil microbiota and primary functional attributes are underexplored. Here, we present the metagenomes of maize soil rhizosphere organisms with functional diversity associated with farms at two different municipalities in North West and Gauteng provinces of South Africa. We describe a plenteous and diverse microbial community."
} | 108 |
29062721 | PMC5645169 | pmc | 8,304 | {
"abstract": "Graphical abstract",
"conclusion": "4 Conclusions The bacterial strains; T1CS3 D and T2BW3 1 , isolated from the fresh water milieu of the Raymond Mhlaba Municipality, Eastern Cape, South Africa which have shown novel ligninolytic activities were identified Raoultella ornithinolytica OKOH-1 and Ensifer adhaerens NWODO-2 with KX640917 and KX640918 as respective accession numbers. These Proteobacteria strains produced extracellular enzymes with the capacity to degrade dyes with ortho , meta and para arene substituent and as such, decolourize the model dyes. Consequently, Raoultella ornithinolytica OKOH-1 and Ensifer adhaerens NWODO-2 hold high potentials for industrial applications, particularly in the dye-wastewater management process. Besides their potential industrial relevance on industrial treatment, they may serve important purpose on the pretreatment of lignocellulosic biomass, a significant step in the bioconversion of lignocellulose to ethanol. Nevertheless, further study on the mechanism(s) of action of these novel bacterial strains for lignin-degradation is imperative as this is significant to their scalability and commercial application in the future.",
"introduction": "1 Introduction Lignin; the aromatic, non-carbohydrate, component of lignocellulose is recalcitrant to degradation. Thus, effective degradation of lignin is of prime importance to the industrial sectors utilizing lignocellulose as raw materials for various value-added products [1] . More so, the recalcitrance of lignin to degradation constitutes an undesirable barrier to the efficient and optimum utilization of the abundant lignocellulosic materials. On the same note, the large amount of lignin generated during industrial production of ethanol, pulp and paper making processes, accumulates and, thus, constitutes serious environmental challenge hence, the need for effective and eco-friendly lignin degradation techniques [2] . The biological means of lignin degradation involves microbial or/and microbial enzymes degradative activities. This technique is advocated over the physical and chemical methods which are generally expensive and saddled with lots of other limitations [3] , [4] . Fungal degradation of lignin, particularly, the white-rot basidiomycetes have been studied extensively [5] , [6] , [7] and, white-rot fungi have been reported as the most effective microbial lignin-degrader. Effectiveness in lignin degradation has been attributed to some extracellular enzymes produced by the white-rot fungi [8] . These extracellular enzymes include laccases (EC 1.10.3.2), some heme-peroxidases such as lignin peroxidase (EC 1.11.1.14), manganese peroxidase (EC 1.11.1.13), versatile peroxidase (EC 1.11.1.16) and dye-decolourizing peroxidase (EC 1.11.1.19). Nonetheless, industrialization of white-rot fungal bio-catalytic/extracellular enzyme process for the de-polymerization of lignin is yet to be achieved. Perhaps, the insufficiency in the maneuverability of the white-rot fungal genome for optimum extracellular enzyme yield, as a function of production cost to commercial value quotient may have constituted an important factor impeding industrialization of the process [9] , [10] . Bacteria, on the other hand, hold very strong potential considering their striking resilience in diverse environments and, as well, their biotechnological significance following, faster growth rate and high dexterity in genome maneuverability [10] , [11] . Hence, the imperativeness in the exploration of bacteria species for lignin depolymerization potentials. Besides, the evolving significance of bacteria in the degradation of lignin has been severally documented [9] . Bacteria species classed into the actinomycetes, α-proteobacteria and γ-proteobacteria have been reported to possess lignin degrading ability [9] , [12] , [13] . Documented ligninolytic bacteria includes Streptomyces viridosporus T7A, Rhodococcus sp , Nocardia autotrophica \n [14] , Microbacterium sp, Brucella melitensis , Ochrobactrum sp, Sphingomonas sp [15] , Streptomyces coelicolor, Arthrobacter globiformis, Rhodococcus jostii RHA1, Pseudomonas putida mt-2 \n [16] , Serratia sp. JHT01 , Serratia liquefacien PT01 , Pseudomonas chlororaphis PT02, Stenotrophomonas maltophilia PT03 and Mesorhizobium sp. PT04 [11] . Ligninolytic bacteria similarly produce extracellular oxidative enzymes including peroxidases which have been implicated in lignin degradation. Besides the association of these extracellular peroxidases in lignin degradation, they have applications in the removal of phenolic pollutants [17] , synthetic dye decolourization [18] , and the synthesis of natural aromatic flavours [19] , [20] . Other applications have likewise included deodourization of manure [21] , applications in peroxidase biosensors [22] , analysis and diagnostic kits [23] and development of skin lightening agents [24] , [25] . Given, the diverse applications of peroxidases in different industrial sectors, the exploration of bacteria species with novel ligninolytic abilities and high potentials for peroxidase production is of prime importance. Consequently, the reported study evaluated bacterial isolates from fresh water milieu of the Raymond Mhlaba Municipality, Eastern Cape, South Africa for peroxidase production potentials and ligninolytic activities.",
"discussion": "3 Results and discussion 3.1 Ligninolytic bacteria isolation and identification A total of thirty (30) potential ligninolytic bacteria were isolated from the samples collected from the Tyhume River ( Table 1 ) and, the 16S rDNA sequence analysis of the two bacteria strains with the best ligninolytic and peroxidase production potentials revealed T1CS3 D as having 99% similarity to Raoultella ornithinolytica strain G.W-CD.10 (KP418804) while T2BW3 1 had 99% similarity to Ensifer adhaerens strain S4-6 (KY496256). The respective nucleotide sequences of the organisms were deposited in a GenBank as Raoultella ornithinolytica OKOH-1 (accession number KX640917) and Ensifer adhaerens NWODO-2 (accession number KX640918). These ligninolytic bacteria are classified into the alpha-proteobacteria ( Ensifer adhaerens NWODO-2) and gamma-proteobacteria ( Raoultella ornithinolytica OKOH-1) respectively. This finding is consistent with earlier classification of ligninolytic bacteria into Actinomycetes, α-Proteobacteria and γ-Proteobacteria [9] . However, the ligninolytic potential of some Bacillus sp. has also been reported [29] , [36] . Some of the reported Proteobacteria with ligninolytic activity include but not limited to, Sphingobium sp. SYK-6 [37] , Pseudomonas putida mt-2, Acinetobacter sp. [16] and Raoultella ornithinolytica S12 [38] . Genome sequencing analysis of Raoultella ornithinolytica strain S12 (CP010557) isolated in China has revealed many genes involved in aromatic compound degradation and other pathways implicated in lignin degradation mechanism [38] , [39] . This further confirms the lignin degradation potential of Raoultella ornithinolytica OKOH-1 as claimed in this study. Furthermore, Fig. 1 showed the phylogenetic relationships between the ligninolytic bacteria in this study and some of those previously reported. The ligninolytic bacteria in this study (indicated with green tips) are, perhaps, more closely related. Fig. 1 Phylogenetic tree showing the relationship between ligninolytic bacteria in this study and some previously reported ligninolytic bacteria in the NCBI database. The tips shown in green represent the ligninolytic bacteria isolated and sequenced in this study while the tips with other colours (red, black and blue) indicate the previously reported ligninolytic bacteria. Red tips indicate Bacilli; blue tips represent Actinobacteria while the green and black tips indicate Proteobacteria. Fig. 1 3.2 Ligninolytic activities The utilization of lignin monomers; guaiacol (2-methoxyphenol) and veratryl alcohol (3, 4-Dimethoxybenzyl alcohol), was indicative of lignin utilization and degradation potentials of the isolates. Guaiacol and veratryl alcohol utilization serves as ligninolysis indicator and as well, lignin oxidation [29] . Only 17% (5 isolates) of the test isolates were able to degrade both guaiacol (phenolic substrate) and veratryl alcohol (non-phenolic substrate). However, all the test isolates grew either on guaiacol or on veratryl alcohol. Isolates substrates utilization intensity was determined by the zone of degradation ( Table 2 ) which became visual after flooding with Grams’ iodine. The reaction of hydrogen iodide (HI) with the substrates in the presence of oxygen resulted in a brown coloration of the un-degraded part of the medium while the degraded part was revealed as a clear halo zone around the colony. Isolate T2BW3 1 showed the highest halo zone on both substrates (32 mm against guaiacol and 34 mm against veratryl alcohol) while isolate T1B1W3 1 had the least (25 mm). However, about 80% of the positive isolates showed halo zones of over 25 mm. Table 2 Degradation of guaiacol and veratryl alcohol by bacterial isolates. Table 2 S/N Isolate code Diameter of halo zone for GA (mm) Diameter of halo zone for VA (mm) 1. T1B1S3 1 26.00 ± 2.00 a 27.00 ± 1.00 a 2. T1B1S3 4 27.00 ± 1.00 a 30.00 ± 0.00 b 3. T1B1W3 1 25.00 ± 3.00 a 25.00 ± 1.00 c 4. T1CS3 D 28.00 ± 0.00 a 31.00 ± 1.00 d 5. T2BW3 1 32.00 ± 0.00 b 34.00 ± 0.00 e GA:Guaiacol; VA:Veratryl Alcohol. Values represent mean ± standard deviation, number of replicate, n = 3. Values with the same superscript letter along the same column are not significantly different ( P > 0.05 ). 3.3 Decolourization of dyes with different arene substituents The structural complexity of dyes is somewhat similar to those of lignin and the recalcitrance of dyes to degradation has been variously documented [29] . The enzymatic decomposition of the phenolic compounds in lignin leads to effective degradation and this is only possible due to the hydrophilic attack at the arene substituents [40] , [41] . Consequently, application of such enzyme system in the decolourization of dye would only be effective if the arene substituents of the dye are susceptible to hydrolyzation [42] . Isolates showing ligninolytic activity on guaiacol and veratryl alcohol were evaluated for dye decolourization using Azure B (AZB), Remazol Brilliant Blue R (RBBR) and Congo Red (CR). Azure B, a thiazine dye, can only be decolourized by high redox potential agents, particularly lignin peroxidases [43] , [44] , [45] . On the other hand, manganese peroxidase and laccase alone cannot oxidize Azure B [43] , [46] . The inclusion of dyes with ortho and para arene (phenolic and non-phenolic) substituents ( Fig. 2 ), was motivated by the quest to ascertain the broad spectrum of activity and specificity of the oxidative enzyme systems produced by these organisms. Congo red has two azo groups (—N \n \n\n<svg xmlns=\"http://www.w3.org/2000/svg\" version=\"1.0\" width=\"20.666667pt\" height=\"16.000000pt\" viewBox=\"0 0 20.666667 16.000000\" preserveAspectRatio=\"xMidYMid meet\"><metadata>\nCreated by potrace 1.16, written by Peter Selinger 2001-2019\n</metadata><g transform=\"translate(1.000000,15.000000) scale(0.019444,-0.019444)\" fill=\"currentColor\" stroke=\"none\"><path d=\"M0 440 l0 -40 480 0 480 0 0 40 0 40 -480 0 -480 0 0 -40z M0 280 l0 -40 480 0 480 0 0 40 0 40 -480 0 -480 0 0 -40z\"/></g></svg>\n\n N—) which impacts the chromophore properties shown by the dye and, the azo groups are attached at the ortho position. Conversely, Remazol Brilliant Blue R is an anthraquinone dye with a para position arene substituent and this dye is recalcitrant to degradation. The carbonyl group (C O), which constitutes the structural backbone of the dye, has been shown to impact the chromophore properties of the dye. The arene substituents position on the aromatic rings is the factor impacting degradation recalcitrant to the dyes. As such, the effective cleavage of the arene substituents at the ortho, meta and para positions marks for novelty. Thus, the natures of the enzymes produced by the organisms are, perhaps, novel and the kinetics as well as the properties shall be further investigated. Fig. 2 Structures of dyes used in this study. Ortho positions are shown in red circles, Meta positions in black circles while Para position is indicated in blue circle. Fig. 2 Dye decolourization ( Table 3 ) showed 10% (3 bacteria strains) of the isolates as positive against AZB, 7% (2) against RBBR and 17% (5) against CR. Quite remarkable was isolates T1CS3 D and T2BW3 1 which showed competence in the decolourization of dyes with the representative ortho, meta and para positions arene substituent. Perhaps, the extracellular oxidative enzymes produced by these organisms are novel or, are, known enzymes with a blend of properties including peroxidases and laccases. Nonetheless, extracellular oxidative enzyme decolourization of Azure B has been associated with lignin peroxidase [47] , while decolourization of azo and anthraquinone dyes are linked with the activity of DyP-type peroxidases [48] , [49] . Table 3 Decolourization of dyes with different arene substituent. Table 3 S/N Isolate code AZB RBBR CR 1. T1B1S3 1 − − + 2. T1B1S3 4 − − + 3. T1B1W3 1 + − + 4. T1CS3 D + + + 5. T2BW3 1 + + + +: positive; −: negative; AZB: Azure B; RBBR: Remazol Brilliant Blue R; CR: Congo Red. 3.4 Peroxidase activity These isolates; T1CS3 D and T2BW3 1 , which respectively showed activity against representative dyes with ortho , meta and para substituents, similarly holds high potentials as peroxidase producers. These ligninolytic bacteria strains were qualitatively positive for peroxidase production ( Fig. 3 ), as was reflected in the appearance of yellowish-brown colouration of the bacterial colony after interaction with 0.4% v/v hydrogen peroxide (H 2 O 2 ) and 1% w/v pyrogallol [31] . Fig. 3 Qualitative peroxidase activity of ligninolytic bacteria. Fig. 3 Upon quantitation of peroxidase production ( Table 4 ), T1CS3 D showed activity of 5250 U/L and T2BW3 1 showed 5833 U/L activity at 48 h incubation timeline. Some related investigation reported similar result however, the peroxidase activity achieved with T1CS3 D and T2BW3 1 were significantly higher than what has been reported in previous studies; Streptomyces strain EC22 had an extracellular peroxidase activity of 270 U/L [50] and Streptomyces sp. F6616 showed peroxidase activity of 535 U/L [51] . The reason for the marked difference in the peroxidase activity observed with these isolates in comparison with documented report is unclear however; it is a motivation for further investigation. Perhaps, it would be pertinent to note that the peroxidase activity shown by T1CS3 D and T2BW3 1 are consistent with their ligninolytic activities as shown with model lignin compounds degradation ( Table 2 ) and decolourization of dyes with varied arene substituent ( Table 3 ). This, therefore, may be suggestive of the production of lignin modifying enzymes by the test bacterial strains including peroxidases. Table 4 Evaluation of ligninolytic bacteria for peroxidase production. Table 4 S/N Isolate Code Peroxidase Activity (U/L) 1. T1CS3 D 5250.00 ± 0.00 2. T2BW3 1 5833.00 ± 0.00 Values represent mean ± standard deviation, number of replicate, n = 3."
} | 3,845 |
28725266 | PMC5513056 | pmc | 8,305 | {
"abstract": "Background The fact that microalgae perform very efficiently photosynthetic conversion of sunlight into chemical energy has moved them into the focus of regenerative fuel research. Especially, biogas generation via anaerobic digestion is economically attractive due to the comparably simple apparative process technology and the theoretical possibility of converting the entire algal biomass to biogas/methane. In the last 60 years, intensive research on biogas production from microalgae biomass has revealed the microalgae as a rather challenging substrate for anaerobic digestion due to its high cell wall recalcitrance and unfavorable protein content, which requires additional pretreatment and co-fermentation strategies for sufficient fermentation. However, sustainable fuel generation requires the avoidance of cost/energy intensive biomass pretreatments to achieve positive net-energy process balance. Results Cultivation of microalgae in replete and limited nitrogen culture media conditions has led to the formation of protein-rich and low protein biomass, respectively, with the last being especially optimal for continuous fermentation. Anaerobic digestion of nitrogen limited biomass (low-N BM) was characterized by a stable process with low levels of inhibitory substances and resulted in extraordinary high biogas, and subsequently methane productivity [750 ± 15 and 462 ± 9 mL N g −1 volatile solids (VS) day −1 , respectively], thus corresponding to biomass-to-methane energy conversion efficiency of up to 84%. The microbial community structure within this highly efficient digester revealed a clear predominance of the phyla Bacteroidetes and the family Methanosaetaceae among the Bacteria and Archaea, respectively. The fermentation of replete nitrogen biomass (replete-N BM), on the contrary, was demonstrated to be less productive (131 ± 33 mL N CH 4 g −1 VS day −1 ) and failed completely due to acidosis, caused through high ammonia/ammonium concentrations. The organization of the microbial community of the failed (replete-N) digester differed greatly compared to the stable low-N digester, presenting a clear shift to the phyla Firmicutes and Thermotogae , and the archaeal population shifted from acetoclastic to hydrogenotrophic methanogenesis. Conclusions The present study underlines the importance of cultivation conditions and shows the practicability of microalgae biomass usage as mono-substrate for highly efficient continuous fermentation to methane without any pretreatment with almost maximum practically achievable energy conversion efficiency (biomass to methane). Graphical abstract Growth condition dependence of anaerobic conversion efficiency of microalgae biomass to methane \n Electronic supplementary material The online version of this article (doi:10.1186/s13068-017-0871-4) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions Biogas generation from microalgae biomass has been researched for approximately 60 years with the major outcome that microalgae represent a rather challenging substrate for anaerobic digestion due to high cell wall recalcitrance and unfavorable C/N ratio, owing to its high protein content [ 18 , 19 ]. The present study investigated the anaerobic digestion from microalgae biomass generated in replete-N as well as naturally occurring (nitrogen limitation, low-N) conditions. The use of algal biomass from replete nitrogen conditions, especially at OLR 4 have led to an inhibition of the digester, caused by high TAN/FAN and VFA concentrations, and thus to fermentation failure with very low methane productivity. In the failed reactor (replete-N biomass, OLR 4), a clear shift could be observed in the bacterial community to the phyla Firmicutes and Thermotogae and archaeal population changed from acetoclastic to hydrogenotrophic methanogenesis. In contrast to fermentation of replete-N biomass, the application of nitrogen limitation during the microalgae cultivation resulted in generation of biomass with significant changes in the composition (highly accessible biomass with two times lower protein content), and thus in an optimal mono-substrate for efficient AD process in continuous manner. The fermentation process was characterized by stable process parameters with very low levels of main inhibitory compounds. The investigation of the microbial communities revealed Bacteroidetes phyla as subsequently dominating the efficiently preforming digester, indicating that these members adapted most efficiently to the microalgae mono-substrate. Furthermore, among the methanogens, the family of Methanosaeta sp. was predominant, suggesting the acetoclastic methanogenesis to be the main pathway during the successful anaerobic degradation of microalgae. The productivity of methane was constantly on a high level (464 ± 9 and 462 ± 9 mL N g −1 VS day −1 at OLR of 2 and 4, respectively), thus corresponding to an energy conversion efficiency (biomass to methane) of 84%. Taken into account the amount of organic matter used to form new microbial cells and energy for cell metabolism was 12–15% [ 14 ], algae substrate conversion efficiency reached in this study almost the practically achievable maximum of 96–99%. According to these considerations, algae biomass can be used highly efficiently for AD without any energy or cost intensive pretreatments. Thus, the presented results of the efficient continuous fermentation of low-N biomass are moving the industrial application of biofuel generation from algal biomass in a more economically feasible direction, especially because the generation of algae biomass under these conditions saves significantly expensive fertilizers (e.g., nitrogen).",
"discussion": "Results and discussion Algae cultivation and resulting biomass properties In previous work, it was elucidated that the composition and the recalcitrance of microalgae biomass strongly depends on the growth conditions, in particular on nutrient availability and harvesting time [ 30 ]. To highlight the importance of nutrient availability, microalgae ( Chlamydomonas reinhardtii CC-1690) biomass for the continuous fermentation was generated using cultivation media with two different nitrogen concentrations (replete-N with 11.77 mM nitrogen and low-N with 3.56 mM nitrogen, supplied as NaNO 3 ). In addition, to avoid changes in biomass characteristics due to storage artifacts, e.g., freezing [ 32 ] or drying [ 33 ], algae biomass was cultured parallel to the fermentation experiments. The growth of the microalgae biomass in photobioreactors was periodically monitored by measuring organic biomass concentration (Fig. 1 ). According to the results from previous work [ 30 ], biomass harvesting was always performed after 6 days of cultivation for both conditions. Fig. 1 Photoautotrophic accumulation of algal biomass under replete-N and low-N culture conditions. Harvesting for fermentation experiments was performed at day 6 for both media conditions (indicated by arrow ). VS volatile solids \n The phototrophic algae, cultivated in culture media with low and replete nitrogen concentrations, showed no significant differences in biomass accumulation rates at the harvesting time (Fig. 1 , 6 days). After 7 days of cultivation, an obvious starvation of biomass accumulation could be monitored in low-N media, due to nitrogen depletion. In accordance with the expectation, biomass accumulation was observed in replete-N conditions up to day 10. Conclusively, no obvious disadvantages in biomass productivity (until day 6, harvesting time point) could be observed after the application of nitrogen limiting culturing conditions (Fig. 1 ). The biomass composition of C. reinhardtii cultivated under replete-N and low-N conditions revealed significant differences regarding the protein and almost no difference in lipid content (Table 1 ), which is consistent with earlier observations on the total lipid and carbohydrate (mainly starch) content in C. reinhardtii CC-1690 under nitrogen deprivation [ 34 , 35 ]. Consequently, carbohydrates represent the main carbon sink in nitrogen starved C. reinhardtii cells. Table 1 Microalgae biomass characteristics Replete-N BM Low-N BM Proteins (% DW) 61.0 ± 5.1 28.0 ± 3.1 Carbohydrates (% DW) 21.0 ± 3.8 52.9 ± 3.5 Lipids (% DW) 20.1 ± 0.8 21.4 ± 1.2 C (% DW) 50.3 ± 1.6 46.4 ± 1.7 N (% DW) 7.3 ± 0.7 2.9 ± 0.2 Volatile solids (% DW) 95.3 ± 1.0 95.6 ± 0.4 COD (g −g DW) 1.34 ± 0.11 1.31 ± 0.11 C/N ratio 6.9 ± 0.7 16.3 ± 1.1 Theoretical methane potential (mL N g −1 VS) ~551 ~549 After harvesting for fermentation, important parameters of C. reinhardtii biomass were determined and presented as mean values. Error bars represent standard error (SE, n = 8) \n BM biomass, DW dry weight, N nitrogen, C carbon, VS volatile solids, TMP theoretical methane potential, COD chemical oxygen demand \n Based on biomass composition, the theoretical methane potential was calculated using the Buswell equation [ 36 ] and empirical formula stated by Heaven et al. [ 37 ] and revealed no significant difference with approximately 551 and 549 mL N g −1 VS between replete-N and low-N biomass, respectively (Table 1 ). Furthermore, corresponding to 2.2-fold lower protein content, the concentration of elemental nitrogen in the low-N biomass was decreased to only 2.9 ± 0.2% of dry weight (DW), whereas the nitrogen amount in the replete-N conditions resulted in 7.3 ± 0.7% of DW. This finding has a direct impact on the C/N ratio in the biomass, which is one of the most critical factors for a continuous fermentation process (C/N ratio: replete-N = 6.9 ± 0.7, low-N = 16.3 ± 1.1, Table 1 ) [ 38 , 39 ]. In this particular case, the C/N ratio of the biomass, cultured under low-N conditions was within the range of 15–30, which is generally regarded as optimal for fermentation processes [ 15 , 39 , 40 ]. Anaerobic digestion of microalgae biomass as mono-substrate The continuous fermentation of algal biomass, generated under replete-N and low-N culture conditions was performed under a constant HRT of 20 days, and the organic loading rate (ORL) was subsequently increased from 1 g VS L −1 day −1 in the beginning, over 2 g VS L −1 day −1 and to 4 g VS L −1 day −1 at the end of the experiment (ORL 1, 2 and 4, respectively, Fig. 2 ). These loading rates were chosen since ORL 2 and 4 (2 and 4 g VS L −1 day −1 , respectively) are generally used in biogas plants for continuous wet fermentation processes on industrial scale [ 15 ]. Differences in the fermentation performance of these two types of biomass were already obvious in the beginning at OLR 1 (adaptation phase), where the gas productivity was not only lower in the replete-N reactor, but was also coupled to a slower adaptation process (defined by stable biogas production). During the whole OLR 2-period, biogas as well as methane productivities were lower and less constant in the replete-N reactor compared to the low-N reactor. With the start of OLR 4, the gas productivity of the replete-N reactor started to decrease and reached the minimum level of specific biogas productivity of 62 ± 2 mL N day −1 g −1 VS, at the end of the experiment. In contrast to replete-N biomass, the biogas as well as methane productivity of the low-N BM reactor remained constantly high (Fig. 2 ) during the whole experiment (exclusive adaptation period, OLR 1). Despite the significantly lower methane concentration in the biogas of low-N digester with 61 ± 0.4% compared to 65 ± 0.9% of replete-N digester (Additional file 1 : Figure S1), the overall methane productivity was higher from low-N biomass (Fig. 2 ) during the complete experimental time course. The overview of the mean biogas and methane productivities, presented in Table 2 , underlines that microalgae biomass from replete-N conditions can only efficiently be used at OLR 2 (2 g VS L −1 day −1 ). However, even this organic loading rate of replete-N biomass is already critical since the biogas productivity was not continually stable. The application of a higher loading rate (OLR 4) has a strongly negative effect on the biogas productivity from replete-N biomass (Fig. 2 ). On the other hand, fermentation of low-N biomass was observed to be stable over both periods OLR 2 and 4, with constantly high methane productivities of 464 ± 9 and 462 ± 9 mL N g −1 VS day −1 , respectively (Table 2 ). The overall achieved methane productivity of low-N algal biomass showed a 36% higher productivity in comparison to maize (Table 2 ) [ 41 ]. Fig. 2 Biogas and methane productivity via anaerobic fermentation of algal biomass in continuous mode. The biogas productivity was monitored online and methane content was measured weekly (left = replete-N BM, right = low-N BM). Organic loading rate (OLR) is indicated by shades of gray in the background, thereby following biomass concentrations were applied: OLR1 = 1 g VS L −1 day −1 , OLR2 = 2 g VS L −1 day −1 , OLR4 = 4 g VS L −1 day −1 . Error bars represent mean productivity of previous 7 days (SE, n = 7). N nitrogen, BM biomass, VS volatile solids \n Table 2 Overview of mean biogas and methane productivities for the low-N and replete-N reactors Specific biogas productivity Specific methane productivity (mL N g −1 VS day −1 ) (mL N g −1 VS day −1 ) Replete-N BM Low-N BM Maize silage Replete-N BM Low-N BM Maize silage OLR 2 g VS L −1 day −1 \n 634 ± 15 761 ± 12 740 a \n 416 ± 11 464 ± 9 404 a \n OLR 4 g VS L −1 day −1 \n 203 ± 50 750 ± 15 620 a \n 131 ± 33 462 ± 9 339 a \n The values were summarized by distinct OLR-phases (OLR 2 = 2 g VS L −1 day −1 , OLR 4 = 4 g VS L −1 day −1 ). Maize silage productivities were included for comparison as predominantly used renewable substrate for industrial scale fermentation. Error bars represent standard error (SE, n = 8) \n N nitrogen, VS volatile solids \n a Literature values for maize silage [ 41 ] \n Despite of the fact that the theoretical methane potential of replete-N and low-N biomass were quite similar, the specific methane productivity of low-N biomass was significantly higher compared to the biomass derived from replete-N conditions [464 ± 9 mL N g −1 VS day −1 vs. 416 ± 11 mL N g −1 VS day −1 at OLR 2 and 462 ± 9 mL N g −1 VS day −1 vs. 131 ± 33 mL N g −1 VS day −1 at OLR 4, respectively (Table 2 )]. However, this finding corresponds well to previous observations, where starved biomass showed a higher accessibility and biodegradability compared to biomass from the linear growth phase [ 30 ]. To evaluate the possible reasons for the productivity differences between replete-N and low-N biomass, some essential fermentation parameters were analyzed for both reactors (Fig. 3 ; Additional file 1 : Figures S2, S3, S4, Additional file 1 : Table S1). Fig. 3 Analysis of essential fermentation parameters during anaerobic digestion of algal biomass in continuous mode. Left = replete-N BM, Right = low-N BM. Organic loading rate (OLR) is indicated by shades of gray in the background: OLR 1 = 1 g VS L −1 day −1 , OLR 2 = 2 g VS L −1 day −1 , OLR 4 = 4 g VS L −1 day −1 . Error bars represent standard deviation (SD, n = 3). Detailed VFA concentration values in SI, Additional file 1 : Table S1. N = nitrogen, BM biomass, VS volatile solids, TAN total ammonium nitrogen, FAN free ammonia nitrogen, VFA volatile fatty acids \n One of the most crucial parameters for the fermentation of protein-rich biomass is nitrogen, which is released during anaerobic decomposition of biomass in form of ammonium into the reactor supernatant [ 26 ]. Monitoring of total ammonia nitrogen (TAN) concentration in the reactor revealed indeed a huge difference between the protein-rich (replete-N BM) and low protein (low-N BM) biomass (Fig. 3 ). The TAN concentrations in low-N reactor were observed to be constantly below 600 mg L −1 during the entire experiment. However, the TAN concentration in the replete-N reactor increased at OLR 2 to a value of nearly 1500 mg L −1 , which is already close to described inhibitory levels of 1700–1800 mg L −1 [ 26 , 42 , 43 ]. These inhibitory levels were exceeded directly after the loading rate of 4 g VS L −1 day −1 (OLR 4), reaching the maximum of 3507 ± 14 mg L −1 at day 140. Nevertheless, free ammonia is known to be a more efficient inhibitor than ammonium and to have a strong negative effect primarily to the methanogens already at low concentration of 50–100 mg L −1 [ 26 ]. Indeed, high free ammonia nitrogen (FAN) concentration was observed in the replete-N reactor already at OLR 2 (Fig. 3 ), which could have had an inhibitory effect on methanogens, indicated by simultaneous decline in methane productivity at days 45–60 (Fig. 2 ). Yet, despite further increase of FAN to 74 ± 0.06 mg L −1 at day 77, the methane productivity remained stable, which may be due to Bacteria or Archaea adaptation to these FAN concentrations, and then the FAN-levels decreased again to 32 ± 0.05 mg L −1 (Fig. 3 ). At the beginning of OLR 4 (day 105), the FAN concentration in replete-N reactor increased again and reached maximal levels at day 112 with 73 ± 0.11 mg L −1 comparable to the maximal levels at OLR 2. Additionally, this increase was accompanied by a simultaneous increase of TAN (starting at day 105 as well), followed by a subsequent accumulation of acetate (from day 112, Fig. 3 ). Nevertheless, the FAN concentration started to decrease after day 112 (Fig. 3 , replete-N BM, upper graph), mostly due to a drop of the pH which was caused by the constant increase of the volatile fatty acid (VFA) concentration. Especially, acetate (up to 170 mM) and other intermediate fermentation products (from day 120) such as propionate, n -butyrate, i -valerate, i -butyrate, n -caproate, n -valerate increased further during the time course of the experiment (Fig. 3 , replete-N BM, lower graph; detailed values in SI, Additional file 1 : Table S1). It can be assumed that an efficient adaptation of anaerobic microorganisms (especially methanogens) was not possible within the short time period, when the change of crucial factors such as FAN, TAN and VFA occurred. As a consequence, the process inhibition could not be surmounted, resulting in a drastic decrease of methane productivity and finally a complete failure of the fermentation process (Figs. 2 , 3 , replete-N BM). Similar observations were also made in other continuous fermentation approaches with protein-rich algal biomass as mono-substrate, where high TAN/FAN concentrations, and consequently increasing VFAs have led to decreased methane productivities [ 20 , 24 , 25 , 29 , 44 , 45 ]. On the other hand, the reactor, fed with low-N biomass, did not show any imbalances in fermentation parameters, being constantly low throughout the entire experiment (Fig. 3 , low-N BM). Especially, the FAN concentration showed values lower than 5 mg L −1 during the complete experimental time, far below inhibitory levels [ 26 ]. Furthermore, this observation is also reflected by constantly high methane productivity at different loading rates (Fig. 2 , low-N BM, Table 2 ). Since the fermentation of microalgae biomass, generated under nitrogen limited conditions was stable and produced constant amounts of methane, it was interesting to evaluate the conversion efficiency level of this process. For this purpose, the theoretical methane potential (TMP) of the biomass was compared to the specific methane productivity reached in the experiments [ 46 ]. According to our calculations, the conversion efficiency for low-N biomass to methane reached 84% [calculation specific methane productivity (Table 2 ) of TMP (Table 1 )] for both loading rates (OLR 2 and 4). Having in mind that approximately 12–15% of the organic matter is used for bacterial growth and maintenance requirements during the anaerobic digestion process [ 14 ], and therefore being not available for fermentation to methane. The fermentation of low-N biomass within this study reached almost the maximal capacity and represents the most efficient process so far described in the literature for algal biomass as a mono-substrate [ 19 ]. For instance, Samson and colleagues observed maximal methane productivity by digestion of Spirulina maxima of only 350 mL N g −1 VS day −1 , and thus a maximal conversion efficiency of 59%. These results, however, were achieved only under OLR 1 and HRT of 30 days, whereas the productivities decreased significantly when higher loading rates were applied, due to pronounced ammonia inhibition [ 45 ]. Even lower maximal productivities of only 267 mL N CH 4 g −1 VS day −1 (at OLR 4 and HRT of 20 days) were obtained in another recent study using Spirulina biomass [ 47 ]. Similar results could be achieved for green algae biomass in other studies, where only 160 mL N CH 4 g −1 VS day −1 could be reached for raw Chlorella vulgaris biomass, corresponding to 32% conversion efficiency. After thermal pretreatment of the biomass, the yield could be increased by 1.5-fold and still reached only 233 mL N CH 4 day −1 g −1 VS corresponding to only 49% of TMP (OLR 0.8, HRT 15) [ 25 ]. Very low methane productivities of only 70 mL N day −1 g −1 VS were published by Mahdy and co-workers for C. vulgaris , corresponding to only 15% conversion efficiency (OLR 1, HRT 15). Nevertheless, parallel digestion of enzymatically pretreated algae biomass was 2.2 times more efficiently digested and resulted in 196 mL N CH 4 day −1 g −1 VS corresponding again to only 49% of TMP (OLR 1, HRT 20) [ 20 ]. Moreover, in comparison to the fermentation performance with microalgae, the theoretical maximum achieved for macroalgae substrate was in the range of 25–45% [ 48 ]. Moreover, the methane productivity from macroalgae fermentation lies often in the range of less than 300 mL CH 4 g −1 VS day −1 [ 27 , 49 , 50 ], which is significantly lower compared to the productivity of 462 mL N CH 4 g −1 VS day −1 achieved in this work with microalgae. Apart from the finding that the methane yield from batch experiments with macroalgae biomass [ 27 ] is rather low compared to microalgae, the continuous fermentation under comparable conditions (regarding loading rate) seems also to be less efficient and sensible towards residual salt content in the biomass due to marine origin [ 49 ]. Thus, the biomass-to-methane conversion efficiency of 84% demonstrated within this work by the application of low-N algae biomass is not only significantly higher compared to other long-term fermentation trails with untreated biomass but also compared to the results achieved after successful pretreatment of microalgae biomass. Furthermore, this efficiency may represent the maximum practically achievable under the AD conditions [ 14 ]. Considering the energy consumption of microbial biomass, the practical efficiency of the fermentation process presented here is at 96–99%, and thus the process may be described as optimal. Based on these “proof of concept” results, this strategy can also be performed under more applied levels. So for instance, the cultivation of microalgae under non-axenic conditions was tested and revealed rather low/negligible contamination levels due to the nature of the photoautotropic culture media (especially low-N conditions) and no negative effect during the fermentation process of this biomass could be observed (unpublished observations). Additionally, other more industrially relevant microalgae species can also be tested in continuous fermentation mode, since our previous batch results for Parachlorella kessleri and Scenedesmus obliquus were quite promising, exhibiting very similar properties in terms of C/N ratios and methane yields such as C. reinhardtii [ 30 ]. Moreover, to reduce the cultivations costs of the microalgae and to include more positive environmental aspects to the process, wastewater could be used as nutrition source and flue gas (e.g., biogas after combustion) could be integrated as CO 2 source in the process [ 27 ]. Consequence of the fermentation parameter on the microbial community High-throughput 16S rRNA gene amplicon sequencing was accomplished to investigate how this suboptimal and optimal performance of the replete-N and low-N biomass digesters is reflected on the microbial community. For the comparison of the dynamics of the bacterial community in the different conditions, samples of inoculum (local waste water treatment plant) and the biogas fermenter, fed with replete-N and low-N biomass on the end of OLR 2 (after 100 days) and OLR 4 (after 160 days) were chosen. In all investigated samples, no evidence of eukaryotic plastid 16S rRNA could be found, suggesting that the algal DNA was completely disintegrated during the anaerobic fermentation. Based on the 16S rRNA gene amplicon database (RDP) [ 51 ], the biogas producing microbial community was dominated by Bacteria with 99%, and the Archaea was only represented with approximately 1% (Fig. 4 ). These findings have previously been reported [ 52 – 55 ], and are in agreement with the fact that bacteria are involved in the first three steps of biomass transformation with a high variety of substrate preferences, and Archaea are restricted to a very narrow substrate spectrum in terms of acetate, methyl-group containing compounds as well as CO 2 and H 2 . Fig. 4 Bacterial diversity dynamic as assessed by high-throughput 16S rDNA amplicon sequencing. The data is represented at the phyla level for Bacteria ( a ) and family level for Archaea ( b ). The reactors fed with biomass cultivated with replete and low nitrogen content (replete-N BM and low-N BM) were exposed to increasing organic loading rates OLR 2 (2 g VS L −1 day −1 ) and OLR 4 (4 g VS L −1 day −1 ). The inoculum and the sampling periods at the end of each OLR were chosen for microbial community monitoring \n According to The prokaryots [ 56 ] and Bergey’s Manual of Systematic Bacteriology [ 57 , 58 ], all identified bacterial community members within the investigated samples are typically involved in the anaerobic degradation of the supplied feedstock as they are described to have cellulolytic, saccharolytic, glycolytic, lipolytic, proteolytic and/or acido-/acetogenic capacities. However, many of the bacterial 16S rRNA amplicon reads (27.26 ± 2.75% for inoculum, 28.94 ± 1.37 and 10.39 ± 0.43% for replete-N BM OLR 2 and OLR 4, as well as 48.01 ± 1.77 and 40.58 ± 1.59% for low-N BM OLR 2 and OLR 4, respectively, Fig. 4 a) could not be classified at the phylum levels, respectively, confirming that largely bacterial communities in AD reactors remain unknown [ 59 ]. The active sludge (inoculum) revealed very high species diversity comprised 603 ± 52 OTUs (Additional file 1 : Table S2). Overall, 73% of the identified sequence reads could be assigned to 18 different phyla, with the most abundant among them the members of the phyla Chloroflexi (26.78%), Actinobacteria (17.96%), Verrucomicrobia (7.80%) and Firmicutes (7.01%), whereas all other phyla were found only to a minor portion (Fig. 4 a). The bacterial diversity dropped significantly during the anaerobic fermentation of algal biomass as mono-substrate, cultivated under replete-N and low-N culture conditions and revealed 178 ± 34 and 111 ± 7 OTUs, as well as 269 ± 20 and 177 ± 2 OTUs for OLR2 and OLR4, respectively (Additional file 1 : Table S2). This development indicates that distinct bacteria species begun to dominate due to the selection pressure based on the certain substrate type and amount and other species were extinct. Similar observations were obtained in other studies [ 60 , 61 ]. Furthermore, in the reactors with no obvious inhibition, the members of the phyla Bacteroidetes became dominant in the AD process, followed by Chlorobi in the digester with replete-N biomass at OLR 2 or Spirochaetes with low-N biomass at OLR 2 and 4 (Fig. 4 a). Interestingly, within the phylum Bacteroidetes , mainly three different main OTUs were identified (OTU_2, 3 and 26; Additional file 1 : Figure S5). OTU_26 is representing the genus Paludibacter of the family Porphyromonadaceae, which was described to ferment various sugars to acetate and propionate as the major fermentation products [ 58 ], and is mostly abundant in the low-N BM digester with high amount of carbohydrates (Table 1 ). The phyla Chlorobi is represented by only one member of the genus Ignavibacterium (OTU_36, Additional file 1 : Figure S5), which was also described to utilize various carbohydrates [ 56 ]. The phyla Spirochaetes mainly consists of two OTUs of the order Spirochaetales (OTU_8 and 18, Additional file 1 : Figure S4), of which OTU_18 could be classified to the genus Treponema that utilizes carbohydrates and/or amino acids as carbon and energy source [ 58 ]. Interestingly, the digester (replete-N BM, OLR 4), which experienced acidosis because of the high FAN/TAN and VFA concentrations (Fig. 3 replete-N BM), showed a completely different bacterial population, with the members of phyla Firmicutes and Thermotogae being the most abundant in this samples (Fig. 4 a). Thereby, the Firmicutes were to 70% represented by the genus Sporanaerobacter (OTU_108), and the Thermotogae to 99.9% by the species (OTU_125, Additional file 1 : Figure S5) similar to Defluviitoga tunisiensis [ 62 ]. Sporanaerobacter was described to be able to utilize some sugars, peptides and various single amino acids into acetate [ 57 , 63 ]. Moreover, members of Thermotogae have been characterized for complex polysaccharide fermentation and hydrogen production [ 62 , 64 ], what might promote beneficial associations with hydrogenotrophic methanogens [ 65 ]. The phyla Bacteroidetes is also present in these samples, however, it is in contrast to the well-performing digesters, mainly represented by other members of the family Porphyromonadaceae (OTU_78 and 111, Additional file 1 : Figure S5). The most members of the family Porphyromonadaceae are primarily described to be weakly saccharolytic in contrast to Paludibacter observed in well-performing digester, since the bacterial growth was not observed to be significantly affected by carbohydrates, but is enhanced by protein hydrolysates [ 58 ], which is also in agreement with the fact that this digester was fed with protein-rich biomass. In general, archaeal communities were much less diverse than bacterial ones (Fig. 4 a, b), with Methanomicrobiaceae , Methanobacteriaceae and Methanosaetaceae being the dominant families. The members of Euryarchaeota in the inoculum (active sludge of the local waste water treatment plant) are present to 1.18% ± 0.13 and are consistent on the genus level of Methanobrevibacter , Methanolinea and Methanospirillum and Methanosaeta , with the last being the most abundant of the methanogenic community. This finding is also in agreement with the general consideration of the acetoclastic activity being the dominant methanogenic pathway [ 66 , 67 ]. Distribution, similar to the inoculum, could be observed in the well-performing (replete-N BM OLR 2 and low-N BM OLR 2, 4) digesters, with Methanosaeta sp. representing the most abundant Archaea in the methanogenic community, followed by Methanoculleus sp. and Methanospirillum sp. and Methanolinea sp (Fig. 4 b). On the other hand, the archaeal community in replete-N BM digester OLR 4 is dominated by Methanoculleus sp. and to lesser extent by Methanosaeta sp., suggesting an apparent redirection from the acetoclastic towards hydrogenotrophic methanogenesis. The increased abundance of Methanoculleus sp. could possibly be attributable to the sensitivity of acetoclastic Archaea towards volatile fatty acid intoxication (acidosis) and/or higher availability of hydrogen provided by certain bacterial species [ 68 ] like the members of the phyla Thermotogae . Similar behavior could be also observed in other studies, whereby the authors suggested that the replacement of the dominant Methanosaeta sp. by Methanoculleus sp. might be a potential warning indicator of acidosis within the fermenter [ 60 , 61 , 69 ]."
} | 8,072 |
30335785 | PMC6207324 | pmc | 8,308 | {
"abstract": "RAVEN is a commonly used MATLAB toolbox for genome-scale metabolic model (GEM) reconstruction, curation and constraint-based modelling and simulation. Here we present RAVEN Toolbox 2.0 with major enhancements, including: (i) de novo reconstruction of GEMs based on the MetaCyc pathway database; (ii) a redesigned KEGG-based reconstruction pipeline; (iii) convergence of reconstructions from various sources; (iv) improved performance, usability, and compatibility with the COBRA Toolbox. Capabilities of RAVEN 2.0 are here illustrated through de novo reconstruction of GEMs for the antibiotic-producing bacterium Streptomyces coelicolor . Comparison of the automated de novo reconstructions with the iMK1208 model, a previously published high-quality S . coelicolor GEM, exemplifies that RAVEN 2.0 can capture most of the manually curated model. The generated de novo reconstruction is subsequently used to curate iMK1208 resulting in Sco4, the most comprehensive GEM of S . coelicolor , with increased coverage of both primary and secondary metabolism. This increased coverage allows the use of Sco4 to predict novel genome editing targets for optimized secondary metabolites production. As such, we demonstrate that RAVEN 2.0 can be used not only for de novo GEM reconstruction, but also for curating existing models based on up-to-date databases. Both RAVEN 2.0 and Sco4 are distributed through GitHub to facilitate usage and further development by the community ( https://github.com/SysBioChalmers/RAVEN and https://github.com/SysBioChalmers/Streptomyces_coelicolor-GEM ).",
"introduction": "Introduction Genome-scale metabolic models (GEMs) are comprehensive in silico representations of the complete set of metabolic reactions that take place in a cell [ 1 ]. GEMs can be used to understand and predict how organisms react to variations on genetic and environmental parameters [ 2 ]. Recent studies demonstrated the extensive applications of GEMs in discovering novel metabolic engineering strategies [ 3 ]; studying microbial communities [ 4 ]; finding biomarkers for human diseases and personalized and precision medicines [ 5 , 6 ]; and improving antibiotic production [ 7 ]. With the increasing ease of obtaining whole-genome sequences, significant challenges remain to translate this knowledge to high-quality GEMs [ 8 ]. To meet the increasing demand of metabolic network modelling, the original RAVEN ( R econstruction, A nalysis and V isualization of M e tabolic N etworks) toolbox was developed to facilitate GEM reconstruction, curation, and simulation [ 9 ]. In addition to facilitating the analysis and visualization of existing GEMs, RAVEN particularly aimed to assist semi-automated draft model reconstruction, utilizing existing template GEMs and the KEGG database [ 10 ]. Since publication, RAVEN has been used in GEMs reconstruction for a wide variety of organisms, ranging from bacteria [ 11 ], archaea [ 12 ] to human gut microbiome [ 13 ], eukaryotic microalgae [ 14 ], parasites [ 15 – 17 ], and fungi [ 18 ], as well as various human tissues [ 19 , 20 ] and generic mammalian models with complex metabolism [ 21 , 22 ]. As such, the RAVEN toolbox has functioned as one of the two major MATLAB-based packages for constraint-based metabolic modelling, together with the COBRA Toolbox [ 23 – 25 ]. Here, we present RAVEN 2.0 with greatly enhanced reconstruction capabilities, together with additional new features ( Fig 1 , Table 1 ). A prominent enhancement of RAVEN 2.0 is the use of the MetaCyc database in assisting draft model reconstruction. MetaCyc is a pathway database that collects only experimentally verified pathways with curated reversibility information and mass-balanced reactions [ 26 ]. RAVEN 2.0 can leverage this high-quality database to enhance the GEM reconstruction process. While the functionality of the original RAVEN toolbox was illustrated by reconstructing a GEM of Penicillium chrysogenum [ 9 ], we here demonstrate the new and improved capabilities and wide applicability of RAVEN 2.0 through reconstruction of a GEM for Streptomyces coelicolor . 10.1371/journal.pcbi.1006541.g001 Fig 1 Schematic overview of RAVEN toolbox version 2.0. RAVEN was significantly enhanced for de novo draft model reconstruction by integrating knowledge from different sources (e.g. KEGG and MetaCyc). It has better import/export support to relevant formats, and especially improved compatibility with the COBRA Toolbox by resolving previously conflicting function names and providing a bi-directional model conversion function. As a MATLAB toolbox, RAVEN provides one unified environment for both model reconstruction and simulation analysis (e.g. FSEOF) and allows scripting for more flexible operations. 10.1371/journal.pcbi.1006541.t001 Table 1 Feature comparison of GEM reconstruction toolboxes. Features MEMO Sys FAME Microbes Flux CoReCo Pathway Tools RAVEN 1.0 COBRA 3.0 a Model SEED merlin RAVEN 2.0 Reconstruct GEM for - Prokaryote ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ - Eukaryote ✔ ✔ ✔ ✔ ✔ - Tissues/cell type ✔ ✔ Reconstruct GEM based on - KEGG ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ - MetaCyc ✔ ✔ - HMR ✔ - Template model b ✔ ✔ ✔ Import/Export - SBML ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ - YAML e ✔ - Excel ✔ ✔ ✔ ✔ ✔ ✔ Mass and charge balance c ✔ ✔ ✔ ✔ ✔ ✔ Define sub-cellular localisation d ✔ ✔ ✔ ✔ ✔ ✔ Annotate transporters during reconstruction ✔ ✔ ✔ ✔ Include spontaneous reactions during reconstruction ✔ ✔ Flux balance analysis simulation ✔ ✔ ✔ ✔ ✔ Pathways visualisation f ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ a. While COBRA does not support GEM reconstruction, it is included here as widely used MATLAB toolbox for GEM analysis. b. An existing GEM can be used as template for model reconstruction based on sequence homology. c. The mass- and charge-balanced reactions are derived from the MetaCyc database. d. The sub-cellular localization of reactions can be estimated using the predictLocalization function in RAVEN. e. COBRApy supports YAML. f. RAVEN allows the visualization of simulation results by overlaying information on pre-drawn metabolic maps. S . coelicolor is a representative species of soil-dwelling, filamentous and Gram-positive actinobacterium harbouring enriched secondary metabolite biosynthesis gene clusters [ 27 , 28 ]. As a well-known pharmaceutical and bioactive compound producer, S . coelicolor has been exploited for antibiotic and secondary metabolite production [ 29 ]. The first published GEM for S . coelicolor , iIB711 [ 30 ], was improved through an iterative process resulting in the GEMs iMA789 [ 31 ] and iMK1208 [ 32 ]. The most recent GEM, iMK1208, is a high-quality model that includes 1208 genes and 1643 reactions and was successfully used to predict metabolic engineering targets for increased production of actinorhodin [ 32 ]. Here, we demonstrate how the new functions of RAVEN can be used for de novo reconstruction of a S . coelicolor GEM, using comparison to the existing high-quality model iMK1208 as benchmark. The use of three distinct de novo reconstruction approaches enabled capturing most of the existing model, while complementary reactions found through the de novo reconstructions gave the opportunity to improve the existing model. After manual curation, we included 402 new reactions into the GEM, with 320 newly associated enzyme-coding genes, including a variety of biosynthetic pathways for known secondary metabolites (e.g. 2-methylisoborneol, albaflavenone, desferrioxamine, geosmin, hopanoid and flaviolin dimer). The updated S . coelicolor GEM is released as Sco4, which can be used as an upgraded platform for future systems biology research on S . coelicolor and related species.",
"discussion": "Discussion The RAVEN toolbox aims to assist constraint-based modeling with a focus on network reconstruction and curation. A growing number of biological databases have been incorporated for automated GEM reconstruction ( Fig 1 ). The generation of tissue/cell type-specific models through task-driven model reconstruction (tINIT) has been incorporated to RAVEN 2.0 as built-in resource for human metabolic modeling [ 19 , 39 ]. RAVEN 2.0 was further expanded in this study by integrating the MetaCyc database, including experimentally elucidated pathways, chemically-balanced reactions, as well as associated enzyme sequences (21). This key enhancement brings new features toward high-quality reconstruction, such as inclusion of transport and spontaneous reactions ( Table 1 ). The performance of RAVEN 2.0 in de novo reconstruction was demonstrated by the large overlap of reactions between the automatically obtained draft model of S . coelicolor and the manually curated iMK1208 model [ 32 ]. This indicates that de novo reconstruction with RAVEN is an excellent starting point towards developing a high-quality model, while a combined de novo reconstruction can be produced within hours on a personal computer. We used the de novo reconstructions to curate the existing iMK1208 model, and the resulting Sco4 model was expanded with numerous reactions, metabolites and genes, in part representing recent progress in studies on metabolism of S . coelicolor and related species ( Fig 5 ). We have exploited this new information from biological databases to predict novel targets for metabolic engineering toward establishing S . coelicolor as a potent host for a wide range of secondary metabolites ( Fig 7 ). Therefore, RAVEN 2.0 can be used not only for de novo reconstruction but also model curation and continuous update, which would be necessary for a published GEM to synchronize with the incremental knowledge. We thus deposited the Sco4 as open GitHub repositories for collaborative development with version control. 10.1371/journal.pcbi.1006541.g007 Fig 7 Flow-chart of the functions for MetaCyc-based de novo GEM reconstruction in RAVEN 2.0. The MetaCyc-based module (indicated by the dashed line) reads in data files of MetaCyc pathways, reactions, compounds and enzymes, and converts the information into a model structure, which can then be utilized for automatic draft GEM generation through querying the protein sequences in MetaCyc (protseq.fsa). Different components are color coded as MetaCyc data files in blue, MATLAB functions in red, and data structures in green. While RAVEN 2.0 addresses several obstacles and significantly improves GEM reconstruction and curation, a number of challenges remain to be resolved. One major obstacle encountered is matching of metabolites, whether by name or identifier (e.g. KEGG, MetaCyc, ChEBI). Incompatible metabolite nomenclature, incomplete and incorrect annotations all impede fully automatic matching and rather requires intensive manual curation, especially when comparing and combining GEMs from different sources. Efforts have been made to address these issues, e.g. by simplifying manual curation using modelBorgifier [ 40 ]. Particularly worth noting is MetaNetX [ 41 ], where the MNXref namespace aims to provide a comprehensive cross reference between metabolite and reactions from a wide range of databases, assisting model comparison and integration. Future developments in this direction ultimately leverage this information to automatically reconcile metabolites and reactions across GEMs. Another major challenge is evaluation and tracking of GEM quality. Here we evaluated Sco4 with growth and gene essentiality simulations ( Fig 5 , S8 Table , S9 Table ), however, the GEM modelling community would benefit from such and additional quality tests according to community standards. Exciting ongoing progress here is memote: an open-source software that is under development that contains a community-maintained, standardized set of metabolic model tests [ 42 ]. Given the YAML export functionality in RAVEN already supports convenient tracking of model changes in a GitHub repository, this should ideally be combined with tracking model quality with memote, rendering RAVEN suitable for future GEM reconstruction and curation needs."
} | 3,006 |
36909500 | PMC10002741 | pmc | 8,309 | {
"abstract": "The biofilm matrix is composed of exopolysaccharides, eDNA, membrane vesicles, and proteins. While proteomic analyses have identified numerous matrix proteins, their functions in the biofilm remain understudied compared to the other biofilm components. In the Pseudomonas aeruginosa biofilm, several studies have identified OprF as an abundant matrix protein and, more specifically, as a component of biofilm membrane vesicles. OprF is a major outer membrane porin of P. aeruginosa cells. However, current data describing the effects of OprF in the P. aeruginosa biofilm is limited. Here we identify a nutrient-dependent effect of OprF in static biofilms, whereby ∆ oprF cells form significantly less biofilm than wild type when grown in media containing glucose or low sodium chloride concentrations. Interestingly, this biofilm defect occurs during late static biofilm formation and is not dependent on the production of PQS, which is responsible for outer membrane vesicle production. Furthermore, while biofilms lacking OprF contain approximately 60% less total biomass than those of wild type, the number of cells in these two biofilms is equivalent. We demonstrate that P. aeruginosa ∆ oprF biofilms with reduced biofilm biomass contain less eDNA than wild-type biofilms. These results suggest that the nutrient-dependent effect of OprF is involved in the maintenance of mature P. aeruginosa biofilms by retaining eDNA in the matrix.",
"introduction": "INTRODUCTION Biofilms are aggregates of bacterial cells encased in a self-produced extracellular matrix. The matrix protects resident cells from external assaults and is composed of exopolysaccharides, extracellular DNA (eDNA), membrane vesicles, and proteins ( 1 ). Many studies have reported the effects of exopolysaccharides, eDNA, and membrane vesicles on biofilm function. However, relatively few have investigated the roles of biofilm matrix proteins, even though matrix proteins have been suggested to play many vital functions in the biofilm ( 2 , 3 ). Since the late 2000s, researchers have used proteomic approaches to identify biofilm matrix proteins and gain insight into their roles, including several studies in the model biofilm organism Pseudomonas aeruginosa . Four different studies have identified OprF as an abundant matrix protein ( 4 – 7 ). Additionally, homologs of P. aeruginosa OprF have been identified in biofilm matrices of other organisms ( 8 – 10 ). Within the P. aeruginosa biofilm, two populations of OprF protein exist: cell-associated and matrix-associated. In its more established cell-associated role, OprF is an OmpA family member and the major non-specific porin in P. aeruginosa , where it facilitates diffusion across the outer membrane ( 11 ). Multiple studies have examined biofilm formation after deletion of oprF or ompA , which eliminates both the cell- and matrix-associated protein pools ( 12 – 14 ). However, the impact of OprF and its OmpA homologs on biofilm formation is somewhat conflicting and may depend on conditions, such as oxygen or nutrient availability ( 15 ). One study shows that under aerobic conditions, a P. aeruginosa oprF interruption mutant produces twice as much biofilm as the parental strain ( 13 ). This result conflicts with a separate study in which an oprF mutant produced less biofilm when grown under anaerobic conditions ( 12 ). Furthermore, the OprF homolog OmpA, which is abundant in Escherichia coli biofilms ( 16 ), increases biofilm formation on hydrophobic surfaces ( 17 ). Mirroring this effect, in the pathogen Acinetobacter baumannii , ompA mutants are deficient in biofilm formation on abiotic surfaces and have decreased attachment to host cells ( 18 ). Together, these data suggest that OprF may play an important role in biofilm function. Within the biofilm matrix, OprF is highly abundant in membrane vesicles, which are a major matrix component involved in biofilm structure and cell-to-cell signaling ( 5 , 19 ). Two membrane vesicle synthesis pathways have been established: the bilayer couple model, which produces outer membrane vesicles (OMVs), and the explosive cell lysis model, which results in membrane vesicles ( 20 , 21 ). Interestingly, OprF has been suggested to play a role in OMV production via the bilayer couple model. An OprF mutant overproduces OMVs relative to wild-type cells due to its overproduction of the quorum-sensing signal PQS ( 22 ). Since increased production of PQS and OMVs is correlated with biofilm dispersal ( 23 ), OprF may be important for this stage of the biofilm lifecycle. However, the role of vesicle-associated OprF in the biofilm is currently unknown ( 11 ). Here we identified a nutrient-dependent biofilm defect in ∆ oprF strains of P. aeruginosa . Upon dissection of the medium components, we found that ∆ oprF biofilm formation was significantly reduced in the presence of glucose or low sodium chloride concentrations without affecting overall bacterial growth. The biofilm defect in the absence of OprF occurs during late-stage biofilm development and is not dependent on PQS production. Interestingly, we observed equivalent numbers of cells in mature wild-type biofilms and ∆ oprF biofilms (that have reduced biofilm biomass). However, there was a significant reduction in eDNA in ∆ oprF biofilms. Together, our data suggest that OprF is involved in the retention of eDNA during mature biofilm maintenance under certain growth conditions.",
"discussion": "DISCUSSION Our results highlight that growth conditions, specifically glucose and sodium concentrations, impact P. aeruginosa oprF mutant biofilm phenotypes. P. aeruginosa ∆ oprF strains formed significantly less biofilm in TSB than LB. The decrease in ∆ oprF biofilm in TSB occurred between 16–24 hours and did not result in fewer P. aeruginosa cells. Instead, ∆ oprF biofilms in TSB contained significantly less eDNA than wild-type biofilms. The mechanisms underlying how glucose and low sodium led to decreased biofilms in cells lacking OprF is an exciting topic for future studies, as is determining how matrix-associated OprF affects eDNA levels. Bouffartigues and colleagues previously found that an oprF interruption mutant forms approximately twice as much biofilm as the parental strain in LB, suggesting that a lack of OprF results in the overproduction of biofilm ( 13 ). Our results in LB using the same oprF interruption mutant strain agree with this conclusion. While these results follow the overall trend we saw in our ∆ oprF strain in TSB and LB ( Fig. 1 ), we did not observe hyperbiofilm formation in our ∆ oprF strain in LB. Since both strains are of the PAO1 lineage and whole genome sequencing of our ∆ oprF strain confirmed that no other differences exist between this strain and the parental, the difference in biofilm phenotypes suggests that there may be additional genetic factors at play. It is possible that the insertion in oprF in H636 affects biofilm formation or that the strain has accumulated secondary mutations within or outside the oprF interruption that affect biofilm formation in LB. These possibilities could be sorted out via future whole genome sequencing of H636 and comparing it to its parental strain. Matrix-associated OprF, a membrane protein containing many hydrophobic residues, is abundant in biofilm membrane vesicles ( 4 , 5 ). OMV production in biofilms is dependent on PQS production ( 32 ), but in our experiments, abolishing PQS production did not impact the ∆ oprF biofilm phenotype ( Fig. 3 ). In a wild-type biofilm, cells produce OMVs via the bilayer couple model with PQS, and MVs via explosive cell lysis ( 32 ). In ∆ pqsA biofilms, MVs are still produced ( 32 ), and we saw no defect in biofilm formation ( Fig. 3 ). Similarly, MVs are likely still produced by cell lysis in the defective biofilms of both the ∆ oprF and ∆ oprF ∆ pqs strains. Notably, these mutant strains would produce vesicles with no OprF. Given that these strains exhibit 60% less biofilm than wild type, we conclude that this decline is due to the lack of OprF, independent of OMV production. Overall, the results of the current study indicate that in a ∆ oprF background, PQS-mediated OMV synthesis is not related to the decrease in biofilm observed in TSB, which raises several questions outside the scope of this study: 1) do ∆ oprF mutants in a biofilm produce more OMVs, as has been reported for planktonic oprF mutants ( 22 )? 2) is matrix-associated OprF found only in vesicles? 3) how do glucose and low sodium affect the typical functions of OprF in biofilms? Further research probing these questions would expand our understanding of the roles of OprF and OmpA homologs in biofilm matrices. OprF significantly affects the P. aeruginosa biofilm when grown under certain conditions. It is tempting to assume that the 60% decline in ∆ oprF biofilms grown in TSB ( Fig. 4A ) is a proportional loss of all biofilm components. However, the static microtiter biofilm assay quantifies total biomass with crystal violet that stains the negatively charged components of the biofilm, namely cell surfaces, matrix membrane vesicles, and eDNA. Our biofilm cell viability assays demonstrate that ∆ oprF biofilms do not lose 60% of their cells ( Fig. 4B ). Instead, the ∆ oprF biofilms contain approximately 60% less eDNA than wild-type biofilms ( Fig. 5 ). eDNA is an essential matrix component primarily produced by biofilm cell lysis ( 21 , 33 ). It has been proposed that membrane vesicles stabilize the matrix of wild-type biofilms through their interactions with eDNA ( 34 ). Therefore, OprF, which is abundant in membrane vesicles, may be involved – directly or indirectly – in these eDNA interactions and thereby in biofilm structural maintenance. The maintenance of mature biofilms as an active, discrete stage in the biofilm lifecycle has been a recent topic of discussion ( 30 ). In this model, established biofilms respond to environmental changes to persist as a community. In a static microtiter biofilm, these changes include depletion of nutrients and waste accumulation over time. Our data indicate that OprF affects mature static biofilms in TSB, with the established ∆ oprF biofilm decreasing between 16–24 hours of incubation. This phenotype suggests that in the absence of OprF, biofilm formation progresses to maturity and subsequently degrades. When combined with our biofilm cell viability results ( Fig. 4 ), mature ∆ oprF biofilm degradation does not appear to be due to dispersion since cell numbers are maintained. Therefore, we hypothesize that OprF may be involved in matrix retention in mature static biofilm maintenance via 1) matrix-bound OprF interactions with eDNA or 2) intracellular regulatory effects of deleting oprF . Future research into these lines of questioning is necessary and will contribute to an expanded understanding of the role of OprF in mature biofilm maintenance."
} | 2,736 |
36386615 | PMC9650238 | pmc | 8,310 | {
"abstract": "Due to their extreme water depths and unique physicochemical conditions, deep-sea ecosystems develop uncommon microbial communities, which play a vital role in biogeochemical cycling. However, the differences in the compositions and functions of the microbial communities among these different geographic structures, such as seamounts (SM), marine trenches (MT), and cold seeps (CS), are still not fully understood. In the present study, sediments were collected from SM, MT, and CS in the Southwest Pacific Ocean, and the compositions and functions of the microbial communities were investigated by using amplicon sequencing combined with in-depth metagenomics. The results revealed that significantly higher richness levels and diversities of the microbial communities were found in SM sediments, followed by CS, and the lowest richness levels and diversities were found in MT sediments. Acinetobacter was dominant in the CS sediments and was replaced by Halomonas and Pseudomonas in the SM and MT sediments. We demonstrated that the microbes in deep-sea sediments were diverse and were functionally different (e.g., carbon, nitrogen, and sulfur cycling) from each other in the seamount, trench, and cold seep ecosystems. These results improved our understanding of the compositions, diversities and functions of microbial communities in the deep-sea environment.",
"introduction": "Introduction Deep-sea sediment environments contain a diverse array of abundant microorganisms, which play a vital role in biogeochemical cycling ( Zhuang et al., 2019 ). The deep-sea environment contains several different geographic structures with distinct characteristics, including seamounts, trenches, cold seeps, and hot springs ( Ingole and Koslow, 2005 ). Due to the obvious differences in water depths, temperatures, and physico-chemical conditions among the different geographic structures in deep-sea environments, they may have different microbial communities and perform different ecological functions ( Levin et al., 2001 ). Chemosynthetic microorganisms that are enriched in deep-sea hot springs can fix carbon at high rates, which results in higher primary productivity than in other deep-sea regions ( McNichol et al., 2018 ). Another study related to the Yap Trench revealed that the microorganisms in deep-sea sediments played an important role in the weathering process of volcanic materials ( Li et al., 2020 ). Moreover, typical sulfur-oxidizing bacteria have been detected in deep-sea hydrothermal vents with substantial abundances ( Ding et al., 2017 ). Therefore, it is necessary to obtain information on the microbial communities in different deep-sea ecosystems and compare the differences in their capacities for biogeochemical cycling. This information can help us to better understand the underlying mechanisms that regulate microbial ecological processes. An important issue in microbial ecology investigations is to obtain the compositions of microbial communities in different environmental ecosystems and to link them with ecological processes, such as the cycles of basic chemical elements ( Martiny et al., 2006 ). Determinations of these connections rely on accurate compositional and functional data of microbial communities ( Guo et al., 2019 ). In recent years, the development of high-throughput sequencing technologies, especially amplicon and metagenomics, has provided an effective tool for comprehensively studying the compositions and functions of complex microbial communities in natural ecosystems ( Mallick et al., 2017 ). Amplicon sequencing of barcode genes has been used to determine the composition of complex microbial communities ( Lundberg et al., 2013 ). Meanwhile, the functions of microbial communities can be comprehensively evaluated via metagenomic sequencing ( Ju and Zhang, 2015 ). Moreover, combined with in-depth binning analysis, we can obtain microbial genomes without isolation and culture and then accurately determine the relationships among microbial compositions and their ecological functions ( Hua et al., 2019 ). At present, these technical solutions have been applied in a variety of ecosystems, such as rivers ( Liu et al., 2020 ), soils ( Su et al., 2017 ), permafrosts ( Hultman et al., 2015 ), oceans ( Delmont et al., 2018 ), and animal intestines ( Xie et al., 2021 ). In this study, sediments from seamounts, trenches, and cold seeps in the deep sea of the Southwest Pacific Ocean were collected. The diversity and composition of the microbial communities in these sediments were determined by using high-throughput sequencing based on the 16S rRNA gene. Meanwhile, high-depth metagenomics combined with binning analysis was applied to recover the main microbial genomes in these sediments, which can link the microbial taxa to biogeochemical cycling. Based on the unique water depth and environmental conditions, we expected significant differences in the diversity, composition, and function of the microbial communities in the sediments from different deep-sea ecosystems. In addition, we forecasted the metabolic patterns of the microbial communities from different geographic regions in the deep sea.",
"discussion": "Discussion The microorganisms present in deep-sea sediments remain largely unknown due to the complexity of the sediment communities, and the special environmental conditions restrict the isolation of deep-sea microorganisms ( Sogin et al., 2006 ). The distinct geochemical environments in different geographic structures of the deep sea may select unique microbial communities ( Zhang et al., 2021 ). These microorganisms could play different roles in utilizing inorganic nutrients and transforming the organic compounds that are involved in deep-sea biogeochemical cycles ( Huang et al., 2019 ). As a culture-independent technology, the metagenomics of marine sediments has revealed a broad diversity of uncultured microorganisms and provided insights into their metabolic capabilities ( Wang et al., 2016 ; Zhang et al., 2016 ). Here, we presented a glimpse of the differences in the sediment microbial communities among different geographic structures in deep-sea environments. Moreover, the assembly and binning of the in-depth metagenomes recovered many uncultured genomes that enlarged the tree of life and improved the understanding of biogeochemical cycling in deep-sea sediments. The benthic community compositions in marine environments are depth stratified and reflect environmental gradients that are correlated with depth, such as temperature, oxygen content, light level, and pressure ( Clark et al., 2010 ). In this study, we observed significant differences in the diversities and compositions of the microbial communities within the CS, MT, and SM sediments ( Figure 1 ). Significantly higher richness and diversity of the microbial communities were observed in the SM and CS sediments compared to those in the MT samples, which was consistent with the lower depths of SM (~ 1,500 m) and CS (~ 1,100 m) than MT (~ 6,000 m). With increasing water depth, the seawater pressure gradually increases, and the microbial diversity decreases due to increased selection in more extreme environments ( Wakeham et al., 2007 ). A negative correlation between biodiversity and water depth has been observed in multiple previous studies ( Varliero et al., 2019 ; Wu et al., 2020 ; Zhang et al., 2021 ). Similar to the alpha diversity indices, the beta diversity, as shown in the PCoA plot in Figure 1D , also demonstrated significant variations in the bacterial communities among SM, CS, and MT. Remarkable shifts in the bacterial communities with the water depth gradient were also found in studies of multiple marine areas regardless of whether they were associated with water or sediment columns ( Yang et al., 2019 , 2021 ). The results of present and previous studies both confirmed the vital effects of water depth on the diversity and composition of bacterial communities in marine environments. A comparison of the results between the present and previous studies on the abundances of the main bacterial phyla was further performed. In agreement with our results, the presence of Proteobacteria, as the dominant bacterial phylum in deep-sea environments, has been proven by multiple studies ( Cui et al., 2019 ; Peoples et al., 2019 ; Serrano et al., 2021 ). Comparable relative abundances of Actinobacteria and Bacteroidetes were also detected in other CS systems in the South China Sea ( Cui et al., 2019 ). However, both Firmicutes and Chloroflexi represented approximately 15% in other CS systems in the South China Sea but nearly disappeared in this study ( Cui et al., 2019 ). For the MT sediments, Proteobacteria and Actinobacteria were the top two bacterial phyla, which occupied over 95% of the microbial communities ( Figure 2A ). In the sediments obtained from a seamount in the Mariana volcanic arc, Firmicutes was the dominant bacterial phylum, which was followed by Proteobacteria ( Liu et al., 2017 ). However, Firmicutes was nearly absent in the MS sediments obtained in the present study and was replaced by Actinobacteria, Bacteroidetes, Acidobacteria, and Chloroflexi ( Figure 2A ). In contrast, although Proteobacteria was also the most dominant phylum in the sediments from other trenches, the other microorganisms with high relative abundances were quite different from our results ( Peoples et al., 2019 ). In this work, Actinobacteria was the second most abundant bacterial phylum in the MT sediments ( Figure 2A ), while Thaumarchaeota, Bacteroidetes, Planctomycetes, and Chloroflexi were found to have substantial abundances in the sediments from hadal trenches ( Peoples et al., 2019 ). These results indicated that the compositions of the microbial communities in deep-sea sediments varied greatly among and within different ecosystems, which were probably affected by the habitat heterogeneity among different geographic locations and niche differentiation at the local scale. It was noteworthy that Acinetobacter was the dominant genus in the CS sediments ( Figure 2B ). Acinetobacter is a gram-negative coccobacillus that has been recognized as an organism of questionable pathogenicity to an infectious agent of importance to hospitals worldwide for decades ( Munoz-Price and Weinstein, 2008 ). In addition to being implicated in a wide spectrum of infectious diseases, Acinetobacter is more concerning because it has multiple resistance to most available antimicrobial agents ( Manchanda et al., 2010 ). In 2017, the World Health Organization recognized carbapenem-resistant Acinetobacter baumannii as the critical, number 1 priority, clinically significant pathogen ( World Health Organisation, 2017 ). The high abundance of Acinetobacter that was revealed in the CS sediments from our study suggested that CS could be one of the reservoirs and sources of Acinetobacter . Unfortunately, no MAG belonged to Acinetobacter was obtained in our study by the binning technology ( Figure 5A ), which prevented us from deeper comparisons. The relationship between Acinetobacter in the clinical and CS environments requires further detailed investigations and comparisons. Aa a central component of the ocean’s biological carbon pump, organisms in surface water can fix carbon to form particulate organic matters, and then sink to seafloor for transporting carbon and energy to abyssal depths ( Poff et al., 2021 ). In deep-sea sediments, these sinking particles are further decomposed and transformed, which in turn stores carbon in the sediment ( Koeve, 2005 ). For CAZYyme classes, glycoside hydrolases and carbohydrate esterases were responsible for degrading carbohydrates ( Drula et al., 2022 ), and they were more abundant in CS sediments compared to SM and MT ( Figure 3B ). This result suggested CS ecosystem could possess stronger ability of carbon sequestration than MT and SM sediments. Except for carbon cycle, nitrogen cycle is of great significance to maintain the balance of marine ecosystems ( Hutchins and Capone, 2022 ). Nitrification and denitrification are a pair of ecological processes that regulate the balance of ammonium and nitrate mainly mediated by aerobic microorganisms ( Barnard et al., 2005 ). Although previous studies reported the influences of temperature, pH, and nutrients on nitrification and denitrification, oxygen is the most important factor for these processes ( Ji et al., 2018 ). More powerful nitrification and denitrification have been observed in areas closer to the sea surface with higher dissolved oxygen ( Sun and Ward, 2021 ). Stronger nitrification and denitrification capacities in SM found in this study ( Figure 4B ) are probably due to the lower water depth of SM compared with MT and CS. In addition, higher abundance of genes related to nitrogen fixation in MT compared with SM and CS ( Figure 4B ) indicated that high level of N 2 could release in the MT ecosystem. Arising nitrogen from the decomposition of organic matter in deep-sea trenches, such as the Cariaco Trench ( Richards and Benson, 1961 ) and the Mariana Trench ( Liu and Peng, 2019 ), which could support the powerful nitrogen fixation in MT ecosystem. Moreover, sulfur compounds are used as both electron donor and acceptor by deep-sea microorganisms for energy conservation ( Yamamoto and Takai, 2011 ). Most of previous studies about sulfur cycle in deep-sea ecosystems focused on the sulfur oxidation at hydrothermal vents ( Sievert et al., 2007 , 2008 ; Zeng et al., 2021 ). Our findings in this study provided evidence of the sulfur reduction in other deep-sea ecosystems ( Figure 4C ). In this study, in-depth metagenomics combined with binning technology was used to explore the relationships among key microorganisms and the carbon, nitrogen, and sulfur metabolic pathways in different deep-sea geographic structures. Compared with traditional PCR-based approaches, metagenomic analyses result in greater sequence depths and also provide an overview of genetic capabilities ( Zeyaullah et al., 2009 ). The results of the present study showed an important phenomenon in which certain biochemical processes in deep-sea sediments could be conducted by various microbial groups. Genes involved in the rTCA, WL, and methanogenesis pathways in carbon cycling were detected in the results based on contigs but were not found in any bins. It should be noted that these genes were detected only in the contigs from the SM sediments ( Figure 4A ). For the genes involved in nitrogen and sulfur cycling, many genes were detected in the contigs that were not observed in the bins ( Figure 6 ). Moreover, the cases of SM.9 (Flavobacteriaceae), MT.6 (Algoriphagus), and CS.6 (Chitinophagaceae) deserve to be mentioned, which had the highest abundances in the SM, MT, and CS sediments, respectively ( Figure 5B ). All of them belong to the Bacteroidota phylum, which has been frequently detected in diverse environments, including soil ( Delgado-Baquerizo et al., 2018 ), sediments ( Probandt et al., 2017 ), sea water ( Royo-Llonch et al., 2017 ), and the guts and skins of animals ( Cuskin et al., 2015 ; Cuscó et al., 2017 ). In these diverse ecological niches, Bacteroidota are increasingly regarded as specialists for the degradation of high molecular weight organic matter, e.g., proteins and carbohydrates ( Li et al., 2017 ). The recent sequencing of Bacteroidota genomes also confirmed the presence of numerous carbohydrate-active enzymes covering a large spectrum of substrates from plant, algal and animal origins ( Han et al., 2009 ; Del Rio et al., 2010 ; Qin et al., 2010 ). These results obtained by previous studies can be mutually confirmed with the multiple carbon cycling-related genes that were detected in SM.9, MT.6, and CS.6, indicating the potential carbon metabolism by Bacteroidota in deep-sea sediments. Here, we further assigned six Bacteroidota (SM.20, MT.3, MT.9, MT.17, CS.4, and CS.5) as potentially metabolizing the nitrogen and sulfur cycles and coupling the electron flow from organic matter to the reduction of nitrate and sulfate. Bacteroidota genomes appear to be highly plastic and are frequently reorganized through genetic rearrangements, gene duplications, and horizontal gene transfers, which are features that could have driven their adaptation to distinct ecological niches ( Francois et al., 2011 ). Horizontal gene transfer, a process mediated by mobile gene elements, such as plasmids, integrons, and transposons, has been widely investigated due to its role in the acquisition and spread of antibiotic resistance genes ( Gillings, 2013 ; Mikalsen et al., 2015 ). Evidence is accumulating that certain environmental characteristics, such as the organic matter content, shape the compositions of the microbial communities in diverse environments ( Patel et al., 2014 ). Horizontal gene transfer can provide microbes with tools to degrade otherwise refractory substances ( Hehemann et al., 2012 ). Thus, the metabolic potential of Bacteroidota may be due to horizontal gene transfer and selection pressure from the extreme deep-sea environment. This partially explained why the distribution of Bacteroidota is more abundant in the organic matter-rich ecosystems (e.g., SM and CS) compared to the oligotrophic environment (e.g., MT; Figure 2A ). However, it should be noted that the metabolic pathways analyzed in this study were based on genomic DNA analysis, such that our interpretation implies the functional capabilities of these microbiomes rather than their actual activities. Further studies directly based on RNA, protein or metabolite levels are necessary to further explore the active functions in different mariculture systems. In summary, sediments from SM, MT, and CS located in the South China Sea were collected and analyzed by high-throughput sequencing technologies. Based on the amplicon sequencing of 16S rRNA gene, significant differences in the diversity and composition of microbial communities were observed among the SM, MT, and CS sediments. In addition, the metagenomics approach unveiled the power of biogeochemical cycling of microbial communities in different deep-sea sediments. The results appeared that genes involved in most of the biogeochemical cycle pathways were rarely or almost absent in the CS sediments. In contrast, SM sediments possessed the most diverse biogeochemical cycle genes, and some pathways were only detected in MT sediments. Moreover, resources of functional bacteria participated in the biogeochemical cycling were recognized by binning analysis. Among them, Proteobacteria and Bacteroidota were most dominant, meanwhile, some of them were novel bacteria without accurate taxonomic annotation at species or genus levels. Overall, such information can not only improve the understanding of the deep-sea microbial communities, but also help to explain the molecular mechanism of biogeochemical cycling in the deep sea using a combination of bioinformatic technology."
} | 4,751 |
32627824 | PMC7458392 | pmc | 8,311 | {
"abstract": "Metals are a finite resource and their demand for use within existing and new technologies means metal scarcity is increasingly a global challenge. Conversely, there are areas containing such high levels of metal pollution that they are hazardous to life, and there is loss of material at every stage of the lifecycle of metals and their products. While traditional resource extraction methods are becoming less cost effective, due to a lowering quality of ore, industrial practices have begun turning to newer technologies to tap into metal resources currently locked up in contaminated land or lost in the extraction and manufacturing processes. One such technology uses biology for the remediation of metals, simultaneously extracting resources, decontaminating land, and reducing waste. Using biology for the identification and recovery of metals is considered a much ‘greener’ alternative to that of chemical methods, and this approach is about to undergo a renaissance thanks to synthetic biology. Synthetic biology couples molecular genetics with traditional engineering principles, incorporating a modular and standardised practice into the assembly of genetic parts. This has allowed the use of non-model organisms in place of the normal laboratory strains, as well as the adaption of environmentally sourced genetic material to standardised parts and practices. While synthetic biology is revolutionising the genetic capability of standard model organisms, there has been limited incursion into current practices for the biological recovery of metals from environmental sources. This mini-review will focus on some of the areas that have potential roles to play in these processes."
} | 422 |
39791879 | PMC11837538 | pmc | 8,312 | {
"abstract": "ABSTRACT Rhizobia are soil bacteria capable of establishing symbiosis within legume root nodules, where they reduce atmospheric N 2 into ammonia and supply it to the plant for growth. Australian soils often lack rhizobia compatible with introduced agricultural legumes, so inoculation with exotic strains has become a common practice for over 50 years. While extensive research has assessed the N 2 -fixing capabilities of these inoculants, their genomics, taxonomy, and core and accessory gene phylogeny are poorly characterized. Furthermore, in some cases, inoculant strains have been developed from isolations made in Australia. It is unknown whether these strains represent naturalized exotic organisms, native rhizobia with a capacity to nodulate introduced legumes, or recombinant strains arising from horizontal transfer between introduced and native bacteria. Here, we describe the complete, closed genome sequences of 42 Australian commercial rhizobia. These strains span the genera, Bradyrhizobium , Mesorhizobium , Methylobacterium , Rhizobium , and Sinorhizobium , and only 23 strains were identified to species level. Within inoculant strain genomes, replicon structure and location of symbiosis genes were consistent with those of model strains for each genus, except for Rhizobium sp. SRDI969, where the symbiosis genes are chromosomally encoded. Genomic analysis of the strains isolated from Australia showed they were related to exotic strains, suggesting that they may have colonized Australian soils following undocumented introductions. These genome sequences provide the basis for accurate strain identification to manage inoculation and identify the prevalence and impact of horizontal gene transfer (HGT) on legume productivity. IMPORTANCE Inoculation of cultivated legumes with exotic rhizobia is integral to Australian agriculture in soils lacking compatible rhizobia. The Australian inoculant program supplies phenotypically characterized high-performing strains for farmers but in most cases, little is known about the genomes of these rhizobia. Horizontal gene transfer (HGT) of symbiosis genes from inoculant strains to native non-symbiotic rhizobia frequently occurs in Australian soils and can impact the long-term stability and efficacy of legume inoculation. Here, we present the analysis of reference-quality genomes for 42 Australian commercial rhizobial inoculants. We verify and classify the genetics, genome architecture, and taxonomy of these organisms. Importantly, these genome sequences will facilitate the accurate strain identification and monitoring of inoculants in soils and plant nodules, as well as enable detection of horizontal gene transfer to native rhizobia, thus ensuring the efficacy and integrity of Australia’s legume inoculation program.",
"conclusion": "Conclusions and future perspectives Australian commercial legume inoculants are composed of a wide diversity of organisms, which span five known rhizobial genera and at least 19 different species. Twenty-three strains could be definitively identified at the species level, while 19 strains could only be conclusively defined at the genus level. With our knowledge of rhizobia-legume symbioses built upon a narrow suite of strains and host organisms ( 1 , 119 , 120 ), sequencing of a greater diversity of rhizobia is required to bolster databases and make them more representative of rhizobial populations. Within the genera analyzed in this study, varying degrees of incongruency between core and symbiosis gene phylogenies were observed, with the level of discordance suggesting a genus-based hierarchy of HGT frequency of Mesorhizobium > Rhizobium > Sinorhizobium > Bradyrhizobium . While there is clear evidence for HGT and its impact on N 2 fixation for Mesorhizobium spp. in the field ( 9 , 10 , 77 , 121 ), the same is not true for the other rhizobia genera. For Rhizobium spp. and Sinorhizobium spp., where pSyms have been shown to be mobile in vitro ( 14 , 15 ), their environmental transfer is yet to be directly observed. Similarly, several studies have concluded symbiosis gene HGT is likely an important driver for the evolution of Bradyrhizobium spp. symbionts ( 58 , 122 – 124 ), but the mechanism of transfer for these genes has yet to be elucidated. Inoculant strains isolated from Australian soils were shown to be similar to exotic strains. This suggests that these inoculants are the result of inadvertent introductions of rhizobia, possibly arriving along with exotic soil, legume seed, or plant material, colonizing Australian soils and subsequently being isolated and developed as inoculants. However, with a paucity of sequence data of native rhizobia, the possibility remains that some of these commercial strains may be indigenous bacteria with a capacity to fix N 2 with introduced legumes. This is particularly pertinent for the Bradyrhizobium inoculant strains, where there is substantial evidence that native legumes are nodulated by this genus ( 125 – 129 ). Sequencing more rhizobia isolated from indigenous legume hosts would improve the resolution of this analysis and enable this question to be answered. The rhizobia analyzed in this study are a cohort of strains with high saprophytic competence and N 2 fixation efficiency for targeted host legumes. They provide a blueprint to allow the development of a sequence-based approach to identify nodule occupancy in the field. These genome sequences are also a highly valuable resource for interrogation into free-living persistence of rhizobia and their evolution as legume N 2 -fixing symbionts.",
"introduction": "INTRODUCTION Legumes have long been recognized for their ability to improve soil fertility, which results from the N 2 -fixing symbiotic associations they can establish with root nodule-forming bacteria known as rhizobia ( 1 ). In Australia, agricultural forage and grain legumes are exotic species introduced following European colonization during the late 18th century ( 2 ). Australian soils lacked effective rhizobia compatible with these introduced legumes, which led to the practice of inoculation ( 3 ). Since the mid-1950s, this approach has been developed into an organized system, whereby strains are collected from edaphically matched regions of the world and screened for effective N 2 fixation with target legumes in glasshouse experiments ( 4 ). Highly effective strains are further tested in field experiments, designed to assess their symbiotic capacity and saprophytic competence across hosts and a range of agroecological conditions ( 5 ). For ease of manufacture, commercial viability, and quality control, a single strain is selected for a compatible group of legumes, known as an inoculant group. Inoculant Mother Cultures are maintained within the Australian Inoculant Research Group (AIRG) and supplied to manufacturers. The commercially available rhizobial inoculant strains cover almost 100 cultivated legume species ( 6 ) and are drawn from five of the 20 ( 7 ) currently recognized genera: Bradyrhizobium , Mesorhizobium , Methylobacterium , Rhizobium , and Sinorhizobium (also known as Ensifer ) ( 8 ). Beyond this broad delineation, our understanding of the species diversity of inoculant strains is poor, which limits our capacity to identify these organisms in the field and accurately attribute N 2 fixation and yield benefit from inoculation. Horizontal gene transfer (HGT) complicates the management of legume inoculation. Transfer of symbiosis genes from inoculant strains to non-symbiotic rhizobia present in Australian soils can lead to the evolution of new legume symbionts that differ in their N 2 fixation capacity. Rhizobial symbiosis genes (e.g., nod , nif , and fix ) are generally considered to form part of the accessory genome, and as such are often found encoded chromosomally on symbiosis islands (SIs) or on plasmids ( 1 ). For instance, horizontal transfer of the symbiosis Integrative and Conjugative Element (ICE) from the Cicer arietinum inoculant strain Mesorhizobium ciceri CC1192 and Biserrula pelecinus strain M. ciceri WSM1271 to resident soil bacteria resulted in the evolution of new legume-nodulating rhizobia ( 9 – 13 ). In some cases, the evolved strains are effective symbionts ( 9 , 10 ). However, poorly effective novel strains have also been observed, with a potential to out-compete an inoculant strain for nodulation of the target legume ( 10 – 13 ). Transfer of symbiosis plasmids (pSyms) has also been shown in vitro for several well-characterized Rhizobium and Sinorhizobium strains ( 14 , 15 ), and while studies of sympatric populations suggest pSym transfer occurs frequently in the rhizosphere ( 16 – 19 ), the potential impacts of pSym transfer on symbiotic N 2 fixation in the field have not been investigated. The advent of whole genome sequencing (WGS) facilitated by NextSeq 500 and MinION Mk1B platforms has made the generation of complete genomes for rhizobia, which are generally between ~7 and 9 Mb in size, a cost-effective undertaking for a research laboratory. Complete bacterial genomes consist of a single high-quality and gap-free sequence for each bacterial DNA molecule (i.e., chromosome or plasmid(s)), providing an accurate replicon structure and an unbiased template for phylogenetic and evolutionary analyses ( 20 ). An array of genome-based taxonomy tools, such as digital DNA–DNA hybridization (dDDH), average nucleotide identity (ANI), and average amino acid identity (AAI), and the capacity to generate core genome phylogenetic analyses with many hundreds ( 21 ) or even thousands of shared single-copy genes ( 22 ) provide a high-resolution means of characterizing and classifying bacteria than was previously achievable. Furthermore, complete genome sequences allow for symbiosis gene loci to be accurately identified and their transfer in the environment following inoculation to be monitored ( 9 , 10 ). While most of the commercial inoculants used in Australia originated from other countries, 13 of the current inoculant strains were isolated from Australian soils. The origin of these strains has been debated ( 3 , 23 ) with three possibilities being that they are (i) exotic, originating from previous undocumented deliberate or accidental introductions, and have subsequently colonized soils and become “naturalized strains”; (ii) native strains, which not only nodulated indigenous legumes prior to colonization but also have a capacity to also engage in symbiosis with introduced agricultural legumes; or (iii) recombinant strains, evolving through HGT between strains from (i) and (ii), generating new rhizobia capable of nodulating introduced legumes. Significant gaps in our understanding of the taxonomy, genomic structure, and origin of many strains within the suite of commercial inoculants hamper current efforts to accurately quantify their efficacy. Here, we sought to address this dearth in knowledge by completely sequencing the genomes of all the legume inoculants commercially available in Australia. We provide an updated taxonomy of these bacteria, identify their replicon structures, and examine their relationship to each other, as well as to the broader set of publicly available genome sequences. We also analyze and speculate on the origins of inoculant strains originally isolated from Australia.",
"discussion": "RESULTS AND DISCUSSION Updated taxonomy of Australian commercial inoculants The genomes of 37 inoculant rhizobia strains were sequenced, and the sequences were submitted to the TYGS ( https://tygs.dsmz.de/ ) for taxonomic analysis, along with the genomes of five previously sequenced strains: WSM1325 ( 24 ), CB782 ( 29 ), CC1192 ( 26 ), WSM1497 ( 27 ), and SU343 ( 28 ), which are the recommended inoculants for annual Trifolium spp., Trifolium semipilosum , Cicer arietinum , Biserrula pelecinus , and Lotus corniculatus , respectively. The 42 total sequences all matched known rhizobial genera, with 19 strains grouping within Bradyrhizobium , four within Mesorhizobium , one within Methylobacterium , 14 within Rhizobium , and four within Sinorhizobium ( Table 2 ). Twenty-three of the submitted genomes met the species threshold with a known type strain (dDDH ≥70, ANI ≥96, AAI ≥96). For inoculant strains CB1717 and SU303, AAI values were greater than 96% for the closest TYGS match for Bradyrhizobium huanghuaihaiense CGMCC 1.10948 T and Rhizobium laguerreae FB206 T , respectively, but dDDH and ANI values were both below the species thresholds for these type strains. Therefore, species names could not be definitively assigned for either strain, and instead, both strains were identified to the genus level as Bradyrhizobium sp. CB1717 and Rhizobium sp. SU303. Historical Biserrula pelecinus inoculant strain WSM1558 was not previously assigned to a species when its genome was reported in Colombi et al. ( 56 ). We found WSM1558 matched with M. opportunistum WSM2075 T slightly below the d 4 dDDH threshold of 70, but both ANI and AAI values were above the species threshold for this type strain, suggesting that M. opportunistum was likely an appropriate species name for this strain. The remaining 19 inoculant strains did not match closely enough with a known type strain to be assigned a definitive species, and so were only classified to the genus level, with 10 Bradyrhizobium TABLE 2 Summary of inoculant strain classification taxonomic updates to the commercial and historical rhizobial inoculants a Species Strain Top TYGS subject d 4 dDDH [CI b ] ANI AAI Assembly \n \n B. arachidis \n \n \n CB756 \n \n B. arachidis LMG 26795 T \n \n 87.7 [85.2–89.9] \n \n 98.4 \n \n 98.7 \n \n ASM2105226v1 \n \n \n B. barranii \n \n \n CC829 \n \n B. barranii 144S4 T \n \n 72.8 [69.8–75.6] \n \n 96.4 \n \n 97.1 \n \n ASM2105236v1 \n \n B. barranii subsp. apii \n \n CC1502 \n \n B. barranii subsp. apii 38S5 T \n \n 74.0 [70.9–76.8] \n \n 96.8 \n \n 97.5 \n \n ASM2971422v1 \n \n \n B. brasilense \n \n \n 5G1B \n \n B. brasilense UFLA03-321 T \n \n 70.2 [67.2–73.0] \n \n 96.3 \n \n 97.2 \n \n ASM2971434v1 \n \n \n B. brasilense \n \n \n CB627 \n \n B. brasilense UFLA03-321 T \n \n 93.4 [91.5–94.9] \n \n 98.9 \n \n 99.0 \n \n ASM2971476v1 \n \n \n B. diazoefficiens \n \n \n CB1809 \n \n B. diazoefficiens USDA 110 \n \n T \n \n \n 89.4 [87.1–91.4] \n \n 98.7 \n \n 99.0 \n \n ASM2105228v1 \n \n \n B. huanghuaihaiense \n \n \n CB3035 \n \n B. huanghuaihaiense CGMCC 1.10948 T \n \n 77.7 [74.7–80.4] \n \n 97.4 \n \n 98.0 \n \n ASM2520088v1 \n \n \n B. pachyrhizi \n \n \n CB1923 \n \n B. pachyrhizi PAC48 T \n \n 84.4 [81.7–86.6] \n \n 98.0 \n \n 98.1 \n \n ASM2971454v1 \n \n \n B. yuanmingense \n \n \n CB1024 \n \n B. yuanmingense CCBAU 10071 T \n \n 81.9 [79.0–84.4] \n \n 97.9 \n \n 98.3 \n \n ASM2520090v1 \n Bradyrhizobium sp. CB82 B. centrolobii BR 10245 T 30.2 [27.8–32.7] 86.0 85.8 ASM2971440v1 Bradyrhizobium sp. CB1015 B. yuanmingense CCBAU 10071 T 52.4 [49.7–55.1] 93.6 94.6 ASM2520092v1 Bradyrhizobium sp. CB1650 B. neotropicale BR 10247 T 40.2 [37.7–42.7] 90.0 91.3 ASM2976191v1 Bradyrhizobium sp. CB1717 B. huanghuaihaiense CGMCC 1.10948 T 59.6 [56.8–62.4] 95.0 96.1 ASM2971432v1 Bradyrhizobium sp. CB2312 B. arachidis LMG 26795 T 53.8 [51.1–56.5] 93.9 95.3 ASM2971442v1 Bradyrhizobium sp. CB3481 B. hereditatis WSM 1738 T 33.4 [31.0–35.9] 87.6 90.3 ASM2971430v1 Bradyrhizobium sp. CIAT3101 B. rifense CTAW71 T 44.1 [41.5–46.6] 91.7 92.9 ASM2971494v1 Bradyrhizobium sp. NC92 B. glycinis CNPSo 4016 T 50.1 [47.5–52.7] 92.8 94.1 ASM2520086v1 Bradyrhizobium sp. WSM471 B. canariense BTA-1 T 57.7 [54.9–60.5] 94.4 95.6 ASM2105224v1 Bradyrhizobium sp. WU425 B. canariense BTA-1 T 57.3 [54.6–60.1] 94.4 95.9 ASM2105230v1 \n \n M. ciceri \n \n \n CC1192 \n \n M. ciceri LMG 14989 T \n \n 88.7 [86.2–90.7] \n \n 98.6 \n \n 98.8 \n \n ASM161882v1 \n \n \n M. ciceri \n \n \n WSM1497 \n \n M. ciceri LMG 14989 T \n \n 86.7 [84.1–88.9] \n \n 98.2 \n \n 98.4 \n \n ASM167245v2 \n \n \n M. jarvisii \n \n \n SU343 \n \n M. jarvisii LMG 28313 T \n \n 99.8 [99.6–99.9] \n \n 99.9 \n \n 99.9 \n \n ASM1317086v1 \n \n \n M. opportunistum \n \n \n WSM1558 \n \n M. opportunistum WSM2075 T \n \n 67.4 [64.5–70.3] \n \n 96.0 \n \n 96.8 \n \n ASM2338000v1 \n Methylobacterium sp. CB376 M. nodulans ORS 2060 T 28.8 [26.4–31.3] 85.6 83.2 ASM2971420v1 \n \n R. hidalgonense \n \n \n CB782 \n \n R. hidalgonense FH14 T \n \n 87.9 [85.4–90.0] \n \n 98.6 \n \n 98.6 \n \n ASM52087v1 \n \n \n R. laguerreae \n \n \n WSM1455 \n \n R. laguerreae FB206 T \n \n 85.0 [82.3–87.4] \n \n 98.4 \n \n 98.7 \n \n ASM2105232v1 \n \n \n R. leguminosarum \n \n \n CC283b \n \n R. leguminosarum USDA 2370 \n \n T \n \n \n 98.2 [97.3–98.7] \n \n 99.5 \n \n 99.5 \n \n ASM2971424v1 \n \n \n R. ruizarguesonis \n \n \n TA1 \n \n R. ruizarguesonis UPM1133 T \n \n 83.2 [80.4–85.7] \n \n 98.0 \n \n 98.5 \n \n ASM2105242v1 \n \n \n R. sophoriradicis \n \n \n CC511 \n \n R. sophoriradicis CCBAU 03470 T \n \n 84.6 [81.8–87.0] \n \n 98.3 \n \n 98.6 \n \n ASM2520079v1 \n \n \n R. sullae \n \n \n WSM1592 \n \n R. sullae IS123 T \n \n 89.0 [86.6–91.0] \n \n 98.4 \n \n 98.6 \n \n ASM2520071v1 \n Rhizobium sp. CB3060 R. leucaenae USDA 9039 T 38.7 [36.2–41.2] 89.6 92.6 ASM2520073v1 Rhizobium sp. CB3090 R. leucaenae USDA 9039 T 41.9 [39.4–44.4] 90.7 93.4 ASM2971428v1 Rhizobium sp. CB3171 R. leucaenae USDA 9039 T 38.5 [36.0–41.0] 89.5 92.6 ASM2971426v1 Rhizobium sp. SRDI969 R. ruizarguesonis UPM1133 T 55.1 [52.4–57.8] 94.0 95.6 ASM2515272v1 Rhizobium sp. SU303 R. laguerreae FB206 T 63.2 [60.3–66.0] 95.6 96.7 ASM2105234v1 Rhizobium sp. WSM1274 R. ruizarguesonis UPM1133 T 55.0 [52.2–57.7] 93.9 95.4 ASM2520077v1 Rhizobium sp. WSM1325 R. indicum MCC 3961 T 55.7 [53.0–58.4] 93.4 95.6 ASM2318v1 Rhizobium sp. WSM4643 R. ruizarguesonis UPM1133 T 54.7 [52.0–57.4] 94.0 95.4 ASM2515274v1 \n \n S. medicae \n \n \n SU277 \n \n S. medicae USDA 1037 T \n \n 98.3 [97.6–98.9] \n \n 99.7 \n \n 99.7 \n \n ASM2520069v1 \n \n \n S. medicae \n \n \n WSM1115 \n \n S. medicae USDA 1037 T \n \n 94.7 [93.0–95.9] \n \n 99.4 \n \n 99.2 \n \n ASM2105256v1 \n \n \n S. meliloti \n \n \n RRI128 \n \n S. meliloti NBRC 14782 T \n \n 90.1 [87.8–92.0] \n \n 98.5 \n \n 98.7 \n \n ASM2105266v1 \n \n \n S. terangae \n \n \n CB3126 \n \n S. terangae USDA 4894 T \n \n 91.6 [98.5–93.4] \n \n 98.7 \n \n 98.6 \n \n ASM2971436v1 \n \n \n a \n \n Strains that match a TYGS subject are in bold; TYGS subjects reported as Ensifer have been changed to Sinorhizobium for clarity. \n \n b \n \n CI, confidence interval. sp., one Methylobacterium sp., and eight Rhizobium sp. The high proportion (~45%) of inoculant strains that did not match a known type strain reflects the inherent diversity of these commercial strains. Replicon architecture and genomic location of symbiosis genes Given the taxonomic diversity of inoculant rhizobia sequenced, we next wanted to understand the overall architecture of the inoculant genomes and compare them to the general replicon configuration across rhizobia genera. We also wanted to identify the genomic location of the symbiosis genes, with a particular focus on common nodulation ( nodABCIJ ) and nitrogen fixation ( nifHDKEN and fixABCX ) genes ( Table S1 ), to better understand the potential of these genes to undergo HGT, generating novel rhizobia strains. Bradyrhizobium spp. represent the largest group of commercial inoculants, with the 19 strains composed of three inoculants for temperate legumes ( B. barranii CC829 for Lotus peduculatus , Bradyrhizobium sp. WU425 for Lupinus sp., and Bradyrhizobium sp. WSM471 for Ornithopus sp.) and 16 inoculants for tropical/subtropical legumes (including B. diazoefficiens CB1809 for Glycine max , Bradyrhizobium sp. NC92 for Arachis hypogaea , and Bradyrhizobium sp. CB1015 for Vigna sp.). Inoculant Bradyrhizobium strains sequenced here harbor a chromosome of ~7.8–9.8 Mb, similar to the well-studied Glycine max strain B. diazoefficiens USDA110 T ( Table 3 ). Bradyrhizobium spp. are rarely reported to harbor plasmids ( 57 , 58 ), and here, only three strains carry accessory plasmids ( B. barranii CC829, B. barranii subsp. apii CC1502, and Bradyrhizobium sp. CB82). Strains of B. barranii , including the type strain B. barranii 144S4 T , have between two and four plasmids, suggesting that multipartite genome architecture may be a characteristic of this species ( 59 ). Symbiosis genes in all inoculant Bradyrhizobium strains were also found chromosomally encoded within an SI, as is the case for well-studied model strain B . TABLE 3 Genome structure of commercial and historical inoculant strains as well as select model organisms a Strain Genome size (Mbp) Chromosome (Mbp) Accessory replicons (Mbp) \n Bradyrhizobium diazoefficiens USDA 110 T \n \n 9.10 \n \n 9.10 \n B. arachidis CB756 9.82 9.82 B. barranii CC829 10.5 9.63 pCC829_1 (0.43) pCC829_2 (0.34) pCC829_3 (0.11) B. barranii subsp. apii CC1502 10.2 9.76 pCC1502_1 (0.44) B. brasilense 5G1B 8.97 8.97 B. brasilense CB627 9.15 9.15 B. diazoefficiens CB1809 9.14 9.14 B. huanghuaihaiense CB3035 9.61 9.61 B. pachyrhizi CB1923 8.85 8.85 B. yuanmingense CB1024 7.84 7.84 Bradyrhizobium sp. CB82 10.1 9.24 pCB82_1 (0.90) Bradyrhizobium sp. CB1015 8.41 8.41 Bradyrhizobium sp. CB1650 9.43 9.43 Bradyrhizobium sp. CB1717 9.23 9.23 Bradyrhizobium sp. CB2312 9.78 9.78 Bradyrhizobium sp. CB3481 8.04 8.04 Bradyrhizobium sp. CIAT3101 9.37 9.37 Bradyrhizobium sp. NC92 8.36 8.36 Bradyrhizobium sp. WSM471 7.79 7.79 Bradyrhizobium sp. WU425 7.87 7.87 \n Mesorhizobium japonicum R7A \n \n 6.53 \n \n 6.53 \n M. ciceri CC1192 6.94 6.30 pMc1192 (0.65) M. ciceri WSM1497 7.20 6.66 pWSM1497 (0.53) M. jarvisii SU343 7.20 6.93 pMLSU343a (0.24) pMLSU343b (0.02) M. opportunistum WSM1558 6.88 6.88 \n Methylobacterium nodulans ORS 2060 T \n \n 8.84 \n \n 7.77 \n pMNOD01 (0.49 ) pMNOD02 (0.46 ) pMNOD03 (0.04 ) pMNOD04 (0.04 ) pMNOD05 (0.02 ) pMNOD06 (0.01 ) pMNOD07 (0.01 ) Methylobacterium sp. CB376 7.69 7.64 pCB376_1 (0.06) \n Rhizobium johnstonii 3841 T \n \n 7.75 \n \n 5.06 \n pRL12 (0.87 ) pRL11 (0.68 ) pRL10 (0.49 ) pRL9 (0.35 ) pRL8 (0.15 ) pRL7 (0.15 ) R. hidalgonense CB782 6.70 4.38 pCB782_1 (1.56) pCB782_2 (0.51) pCB782_3 (0.25) R. laguerreae WSM1455 6.97 4.82 pWSM1455_1 (1.06) pWSM1455_2 (0.52) pWSM1455_3 (0.29) pWSM1455_4 (0.28) R. leguminosarum CC283b 8.06 5.07 pCC283b_1 (1.16) pCC283b_2 (0.60) pCC283b_3 (0.56) pCC283b_4 (0.55) pCC283b_5 (0.13) R. ruizarguesonis TA1 7.62 5.04 pTA1_1 (0.81) pTA1_2 (0.66) pTA1_3 (0.61) pTA1_4 (0.50) R. sophoriradicis CC511 6.97 4.50 pCC511_1 (0.76) pCC511_2 (0.65) pCC511_3 (0.47) pCC511_4 (0.40) pCC511_5 (0.19) R. sullae WSM1592 7.60 4.14 pWSM1592_1 (2.76) pWSM1592_2 (0.47) pWSM1592_3 (0.23) Rhizobium sp. CB3060 7.28 4.07 pCB3060_1 (2.26) pCB3060_2 (0.54) pCB3060_3 (0.28) pCB3060_4 (0.14) Rhizobium sp. CB3090 6.49 3.91 pCB3090_1 (1.77) pCB3090_2 (0.56) pCB3090_3 (0.24) Rhizobium sp. CB3171 7.88 3.97 pCB3171_1 (2.83) pCB3171_2 (0.54) pCB3171_3 (0.21) pCB3171_4 (0.19) pCB3171_5 (0.10) pCB3171_6 (0.035) Rhizobium sp. SRDI969 6.84 4.90 pSRDI969_1 (0.64) pSRDI969_2 (0.55) pSRDI969_3 (0.43) pSRDI969_4 (0.33) Rhizobium sp. SU303 7.07 5.01 pSU303_1 (0.94) pSU303_2 (0.53) pSU303_3 (0.33) pSU303_4 (0.26) Rhizobium sp. WSM1274 7.08 4.91 pWSM1274_1 (1.18) pWSM1274_2 (0.42) pWSM1274_3 (0.31) pWSM1274_4 (0.25) Rhizobium sp. WSM1325 7.42 4.77 pR132501 (0.83) pR132502 (0.66) pR132503 (0.52) pR132504 (0.35) pR132505 (0.29) Rhizobium sp. WSM4643 6.44 4.86 pWSM4643_1 (0.63) pWSM4643_2 (0.55) pWSM4643_3 (0.41) \n Sinorhizobium meliloti 1021 \n \n 6.69 \n \n 3.65 \n pSymB (1.68 ) pSymA (1.35 ) S. medicae SU277 6.76 3.75 pSU277_1 (1.53) pSU277_2 (1.02) pSU277_3 (0.37) pSU277_4 (0.07) S. medicae WSM1115 7.06 4.10 pWSM1115_1 (1.55) pWSM1115_2 (1.13) pWSM1115_4 (0.28) S. meliloti RRI128 7.27 3.73 pRRI128_1 (1.61) pRRI128_2 (1.29) pRRI128_3 (0.31) pRRI128_4 (0.18) pRRI128_5 (0.15) S. terangae CB3126 6.63 3.97 pCB3126_1 (2.09) pCB3126_2 (0.57) \n \n a \n \n Well-studied model organisms are in bold; replicons with gray shading contain major symbiosis genes. diazoefficiens USDA110 T ( 60 ). SIs are commonly found in Bradyrhizobium genomes and have a range of characteristics suggesting they may have been acquired by HGT, including a lower GC content and a different codon usage compared to non-SI genes ( 61 ). SIs were identified in every inoculant Bradyrhizobium genome sequence and are highly variable in size, ranging from 0.49 to 1.2 Mbp with an average size of 0.79 Mbp ( Table 4 ). In most cases, the SIs were flanked by a tRNA gene at one end, and a recombinase family protein encoding gene at the other. Four strains had atypical SI borders, with B. brasilense 5G1B and B. huanghuaihaiense CB3035 containing a helicase-related protein, and Bradyrhizobium sp. CB1650 containing a DDE-type integrase/transposase/recombinase instead of the characteristic recombinase family protein, while Bradyrhizobium sp. CB3481 was missing both typical markers; instead, the SI was bordered by a transposase and a tyrosine-type recombinase/integrase ( Table 4 ). Bradyrhizobium SIs have been previously reported to be divided across three loci, labeled as regions A, B, and C ( 58 , 62 ). While regions corresponding to SI A were readily identifiable ( Table 4 ), loci corresponding to regions B and C could not be conclusively identified in many of the Bradyrhizobium inoculant genomes ( Table S2 ), indicating that a variation in SI number and configuration exists across this genus. TABLE 4 Size, GC content, and flanking loci of SIs within the chromosome of the Bradyrhizobium sp. inoculant genomes a Strain Genome size (bp) Chromosome GC content (%) SI size (bp) SI GC content (%) Left flank (locus tag) Right flank (locus tag) \n B. diazoefficiens USDA110 T \n \n 9,105,828 \n \n 64.1 \n \n 682,397 \n \n 59.4 \n Recombinase (BJA_RS07875 ) tRNA-Val (BJA_RS10650 ) B. barranii CC829 10,502,301 63.6 485,737 59.9 Recombinase (BjapCC829_RS06045) tRNA-Val (BjapCC829_RS08225) B. diazoefficiens CB1809 9,136,715 64.0 672,335 59.4 tRNA-Val (BdzoCB1809_RS36155) Recombinase (BdzoCB1809_RS39005) Bradyrhizobium sp. WU425 7,868,972 63.3 569,581 59.5 tRNA-Val (BcanWU425_RS31625) Recombinase (BcanWU425_RS34010) Bradyrhizobium sp. CB1015 8,412,734 63.8 577,690 58.9 tRNA-Val (N2604_RS05620) Recombinase (N2604_RS08035) B. yuanmingense CB1024 7,840,305 63.9 845,148 58.6 tRNA-Ile (N2605_RS25055) Recombinase (N2605_RS28330) B. arachidis CB756 9,825,352 63.6 881,791 58.1 Recombinase (BaraCB756_RS07940) tRNA-Ile (BaraCB756_RS11540) Bradyrhizobium sp. NC92 8,355,732 63.9 844,827 58.4 tRNA-Ile (N2602_RS31355) Recombinase (N2602_RS34865) Bradyrhizobium sp. WSM471 7,785,529 63.4 569,250 59.1 Recombinase (BcanWSM471_RS06555) tRNA-Val (BcanWSM471_RS09065) B. brasilense 5G1B 8,976,199 63.8 1,186,148 60.3 Helicase-related protein (QA635_RS30235) tRNA-Glu (QA635_RS35330) Bradyrhizobium sp. CB82 10,142,525 62.5 1,005,155 58.6 tRNA-Val (QA640_RS35850) Recombinase (QA640_RS40010) B. brasilense CB627 9,148,457 63.7 823,972 58.6 Recombinase (QA636_RS33020) tRNA-Glu (QA636_RS36455) Bradyrhizobium sp. CB1650 9,430,299 63.1 1,098,987 59.1 tRNA-Val (QA641_RS35410) DDE-type integrase/transposase/recombinase (QA641_RS40185) Bradyrhizobium sp. CB1717 9,232,581 63.8 647,706 58.6 tRNA-Ile (QA649_RS34405) Recombinase (QA649_RS37030) B. pachyrhizi CB1923 8,851,675 63.8 636,927 58.9 tRNA-Val (QA639_RS33045) Recombinase (QA639_RS35855) Bradyrhizobium sp. CB2312 9,780,375 63.4 899,892 58.6 tRNA-Ile (QA642_RS38195) Recombinase (QA642_RS41855) B. huanghuaihaiense CB3035 9,614,050 63.8 1,208,964 59.1 tRNA-Val (N2603_RS35765) Helicase-related protein (N2603_RS41080) Bradyrhizobium sp. CB3481 8,035,627 62.6 772,783 58.7 Transposase (QA643_RS25030) Tyrosine-type recombinase/integrase (QA643_RS28720) B. barranii subsp. apii CC1502 10,198,616 63.5 761,276 59.6 Recombinase (QA633_RS06225) tRNA-Val (QA633_RS09570) Bradyrhizobium sp. CIAT3101 9,368,452 63.4 605,622 57.8 Hypothetical protein (QA645_RS38425) tRNA-Ile (QA645_RS40690) \n Inoculant average \n 9,079,289.26 63.52 794,410.05 58.94 \n \n a \n \n Model strain B. diazoefficiens USDA110 T is included for reference in bold. The Mesorhizobium spp. inoculants for Cicer arietinum ( M. ciceri CC1192), Biserrula pelecinus ( M. ciceri WSM1497 and M. opportunistum WSM1558), and Lotus corniculatus ( M. jarvisii SU343) all contain chromosomes of similar size (6.29–6.95 Mbp), with symbiosis genes inside ICEs on the chromosomes ( Table 2 ). These elements are either monopartite (CC1192) or tripartite (WSM1497 and SU343) in structure, as has been previously reported ( 26 – 28 ). WSM1558 also contains a tripartite ICE with a total size of 619.9 kb comprising α (6,169,134 to 6,756,269 bp), β (2,982,080 to 2,997,617 bp), and γ (2,736,515 to 2,758,719 bp) regions. CC1192, WSM1497, and SU343 also possess RepABC-type accessory plasmids, which are absent in WSM1558 and model strain M. japonicum R7A ( Table 3 ). The CC1192 plasmid does encode some recognized symbiosis genes ( fixNOQP and fixGHI ), but these are not essential to symbiosis ( 9 ), likely due to additional copies of these genes encoded on the CC1192 ICE. No plasmid-encoded symbiosis genes were identified on the plasmids of either WSM1497 or SU343. The genome of Methylobacterium sp. CB376 consists of a chromosome and a plasmid, with the symbiosis genes encoded from the chromosome. While the symbiosis genes are in relatively close proximity, the overall region of ~330 kb (% GC 72.5) does not exhibit the consistent low GC content characteristic of an SI. Closer inspection of the ~6.5 kb region containing nodulation genes nodD and nodABCIJH showed it had a GC content of 52.5% compared to a chromosomal GC content of 71.6%. Model strain M. nodulans ORS 2060 T also encodes its symbiosis genes on the chromosome but carries seven additional plasmids ( Table 3 ), which amount to an additional 1.14 Mbp size difference between the two Methylobacterium spp. Genomes of members of Rhizobium are generally composed of a chromosome and several plasmids of varying sizes ( 1 ), such as the well-studied strain R. johnstonii 3841 T , which contains a 5.06 Mbp chromosome and six plasmids between 0.15 and 0.87 Mbp. Symbiosis genes in Rhizobium spp. are usually encoded on one of these plasmids (designated pSym), which for 3841 is pRL10 ( 63 ). Genomes of the Rhizobium inoculants sequenced in this study similarly contained between two and six plasmids ranging in size from 0.035 to 2.83 Mbp, with an average size of 0.66 Mbp. Symbiosis genes are encoded on a single plasmid (pSym) for 13 of the 14 Rhizobium inoculant strains. The exception was for the Vicia faba inoculant Rhizobium sp. SRDI969, where the symbiosis genes are unusually encoded on the chromosome in a ~ 75 kb region, as recently reported ( 64 ) ( Table 3 ). Comparison of the symbiosis region of Rhizobium sp. SRDI969 to plasmid-borne symbiosis gene regions from other Vicia / Pisum / Lens -nodulating strains, Rhizobium sp. SU303 (on pSU303_1) and R. johnstonii 3841 T (pRL10), showed all three regions shared a highly similar overall gene arrangement over ~57 kb and >97% nucleotide identity between coding sequences ( Fig. 1 ). All three regions exhibited a GC content of between 56.2% and 56.3%; the plasmid-borne regions are flanked by transposase encoding genes while the chromosomal region has a transposase on one side and rpoN encoded on the other. This suggests that the chromosomal SRDI969 sym genes may have resulted from a transfer event from a pSU303_1 or pRL10-like plasmid to the SRDI969 chromosome. Inter-replicon symbiosis gene transfer has been documented ( 65 ), and Mazurier and Laguerre reported nod and nif genes in lentil-nodulating R. leguminosarum strains were localized on either the chromosome or a large extrachromosomal replicon ( 66 ). It is therefore possible that other Rhizobium strains may encode their symbiosis genes chromosomally, like SRDI969. Fig 1 Comparison of SIs from Rhizobium sp. SRDI969, Rhizobium sp. SU303, and R . johnstonii 3841 T . Dot plots are shown on the left, and alignments are shown on the right. Comparative genomic alignment depicts BLASTn similarity among pRL10 SI, SRDI969 SI, and pSU303_1 SI sequences. Conserved regions are represented in shades. Gene arrangements are shown with arrows along genomic coordinates. Genomes of Sinorhizobium inoculants for Medicago spp. ( S. meliloti RRI128 and S. medicae WSM1115), Trigonella foenum-graecum ( S. medicae SU277), and Desmanthus ( S. terangae CB3126) all have two large megaplasmid replicons, with symbiosis genes present on the smaller of the two ( Table 3 ), as is the case with model strain S. meliloti 1021 (also referred to as Sm1021) ( 67 ). In Sm1021, the largest megaplasmid (pSymB) is considered a hybrid replicon with features of both chromosomes and plasmids, called a chromid ( 68 ). This classification is, in large part, due to essential protein synthesis genes engA and tRNA-Arg, which are located on pSymB in Sm1021 ( 69 ) but typically found on chromosomes in other bacteria ( 51 ). Consistent with Sm1021, replicons pRRI128_1 ( S. meliloti RRI128), pWSM1115_1 ( S. medicae WSM1115), and pSU277_1 ( S. medicae SU277) each encode engA and tRNA-Arg. These replicons also have a GC content more similar to that of their corresponding chromosomes and encode plasmid-like RepABC-type replication systems, suggesting that they, like pSymB, may be chromids. Interestingly, neither engA , tRNA-Arg, nor any other essential genes described by Parks et al. ( 51 ) were found on plasmids pCB3126_1 or pCB3126_2 of S. terangae CB3126; they are instead encoded from the chromosome, suggesting chromids may not be a conserved feature of this species’ genome. S. meliloti RRI128, S. medicae WSM1115, and S. medicae SU277 also carry one to three additional plasmids, which are absent in Sm1021 and CB3126 ( Table 3 ). These accessory or “cryptic” plasmids have been previously observed in other Sinorhizobium spp. ( 70 – 72 ), but their specific roles remain unknown ( 19 ). Inoculant core and symbiosis gene phylogenies and origin of strains isolated from Australia With an understanding of the identity and genomics of the Australian commercial inoculants, we next phylogenetically compared strains from the four significant inoculant genera to other well-studied rhizobia for which complete genome sequences were available. This comparison was performed on core and symbiosis gene sets, to explore the relationship between inoculant strains and other rhizobia and detect potential instances of symbiosis gene transfer. Mesorhizobium The four sequenced Mesorhizobium inoculant strains were phylogenetically compared to 54 strains of Mesorhizobium at both the core and the symbiosis gene levels ( Fig. 2 ). Strains isolated from Biserrula , Cicer , and Lotus were interspersed in the core genome tree, in contrast to the symbiosis tree where they clustered into three host-based groups ( Fig. 2 ). The high level of discordance between these two phylogenies is characteristic of the frequently observed horizontal transfer of symbiosis genes encoded on Mesorhizobium symbiosis ICEs ( 73 ). All four Mesorhizobium inoculant strains are known to have been introduced into Australia, and the core genome phylogeny of M. ciceri CC1192 (originating from Israel [ 74 ]) and M. ciceri WSM1497 (Greece [ 75 ]) largely reflects this, with both grouping with other strains from the Mediterranean. M. jarvisii SU343 from the USA ( 76 ) clustered most closely with Lotus -nodulating strains isolated from Japan and New Zealand. Lotus spp. are native to Japan and the USA, but not to New Zealand, where Lotus spp. are introduced forage legumes ( 77 ). Biserrula inoculant strain M. opportunistum WSM1558 isolated from Italy ( 78 ) grouped most closely with M. opportunistum WSM2075 T isolated from Australia. WSM2075, along with M. australicum WSM2073 T , has previously been shown to have acquired the symbiosis ICE from M. ciceri WSM1271, a Biserrula strain introduced from Italy to Australia in 1995 ( 79 – 81 ). Both M. australicum and M. opportunistum are present in Australian soils, as non-symbiotic strains (i.e., those lacking any discernible symbiosis genes) were recently isolated from cultivated and uncultivated Western Australian soils ( 56 ). In addition to Italy and Australia, strains of M. opportunistum have been isolated from the nodules of C. arietinum from Portugal ( 82 ) and C. canariense from the Canary Islands ( 83 ), suggesting that M. opportunistum is a widely distributed species capable of accepting symbiosis ICEs. Fig 2 Core ( A ) and symbiosis ( B ) gene phylogenies of the genus Mesorhizobium . Strains are overlayed with their host legume of isolation using the colors pink, purple, and green to represent Cicer , Biserrula , and Lotus , respectively. Trees were constructed using RAxML with Sinorhizobium americanum CFNEI 73 set as the outgroup. Nodes with 100% bootstrap support are marked with a black circle. Phylogenetic trees highlight Mesorhizobium species grouped by host plants (Cicer, Biserrula, Lotus). Subgroups are color-coded for Cicer, Biserrula, and Lotus. Broad host range Mesorhizobium subgroups are outlined with dashed lines. Ten strains appear to contain near-identical symbiosis genes to M. ciceri CC1192 despite different genetic backgrounds ( Fig. 2 ). Eight of these strains (WSM4303 through WSM4906) are known recipients of ICE Mc Sym 1192 that were recently isolated from Australian soils ( 9 , 10 ). Additionally, genomic analysis of Mesorhizobium sp. M7D.F.Ca.US.005.01.1.1 and M. ciceri USDA 3378 shows that symbiosis ICE from CC1192 aligns to M7D.F.Ca.US.005.01.1.1 with 99.997% pairwise identity across a single region, and to USDA 3378 with 89.1% pairwise identity across 17 contigs, suggesting that they too share a ICE Mc Sym 1192 with M. ciceri CC1192. Mesorhizobium sp. M7D.F.Ca.US.005.01.1.1 was isolated from Washington, USA, in 2013 ( 84 ) and has a core genome more similar to M. loti DSM 2626 T , while USDA 3378 was also isolated from the USA, but in 1940 (Rhizobium Dataset, USDA-ARS GRIN, https://www.ars-grin.gov/Rhizobium ). Sinorhizobium In contrast to the high discordance of the Mesorhizobium strains, core and symbiosis gene phylogenies of inoculant Sinorhizobium and the 62 selected genomes showed a relatively low level of discordance ( Fig. 3 ). Two of the four inoculant strains are exotic ( S. medicae WSM1115 from Greece [ 3 ] and S. terangae CB3126 from Mexico [ 85 ]), and both cluster closely to other strains from these areas on core and symbiosis gene trees. The remaining strains, S. medicae SU277 for Trigonella foenum-graecum and S. meliloti RRI128 for Medicago spp., were both isolated from Australia ( 86 , 87 ). The S. medicae and S. meliloti clades on the core gene tree are in proximity ( Fig. 3 ), which is consistent with previous reports describing the geographic origins as well as the close phylogenetic and symbiotic relationship between these species ( 19 , 88 – 91 ). S. medicae SU277 is similar to strains isolated from Greece, Italy, and Iraq, suggesting that it may have originated from the Mediterranean Basin and/or the Fertile Crescent region. Additionally, S. meliloti RRI128 is similar to many strains from Europe and the Americas but appears most similar to strains isolated in Kazakhstan, Ukraine, and Hungary, suggesting it may have originated from eastern Europe ( Fig. 3 ). This is consistent with the proposed center of Medicago diversity in the Caucasus region ( 89 ). Importantly, the only available genome sequence of a Sinorhizobium strain isolated from an Australian native legume ( Indigofera sp.) —Sinorhizobium sp. WSM1721 ( 92 )—does not cluster with the introduced inoculant strains, further supporting the likely exotic origin of the inoculant strains isolated from Australia. Fig 3 Core ( A ) and symbiosis ( B ) gene phylogenies of the genus Sinorhizobium . Strains are overlayed with their host legume of isolation using the colors pink, blue, brown, purple, and green to represent Phaseolus , Acacia , Prosopis , Glycine , and Medicago , Melilotus , and Trigonella , respectively. Select strains within the core genome phylogeny are overlayed with a gray box in which the country of isolation is indicated. Trees were constructed using RAxML with Mesorhizobium ciceri CC1192 set as the outgroup. Nodes with 100% bootstrap support are marked with a black circle. Phylogenetic trees depict relationships among Sinorhizobium species grouped by host plants: Phaseolus, Acacia, Prosopis, Glycine, and Medicago/Melilotus/Trigonella. Inoculant status and country of isolation are also indicated. Rhizobium The 14 Rhizobium inoculants are widely dispersed throughout the core genome phylogeny ( Fig. 4 ), which is consistent with their taxonomic assignments ( Table 2 ). Eight of these strains that nodulate either Trifolium or Vicia , Lens , and Lathyrus fall within the Rhizobium leguminosarum species complex (Rlc), which is a cluster of related strains representing at least 18 clades, each of which could be considered a distinct species ( 21 , 93 ). Five of these strains were introduced as inoculants: Rhizobium sp. WSM1274 and R. laguerreae WSM1455 for Vicia faba from Greece, Rhizobium sp. WSM4643 for Pisum and Vicia from Italy, R. leguminosarum CC283b for Trifolium ambiguum from Russia, and Rhizobium sp. WSM1325 for a range of annual Trifolium spp. from Greece ( 5 , 94 – 96 ). The core genomes of the Australian isolated strains Rhizobium sp. SRDI969 for Vicia faba , R. ruizarguesonis TA1 for Trifolium , and Rhizobium sp. SU303 for Lathyrus clustered with other strains from across Europe ( Fig. 4 ), suggesting that these strains may have originated from Southern or Eastern Europe. Fig 4 Core ( A ) and symbiosis ( B ) gene phylogenies of the genus Rhizobium . Strains are overlayed with their host legume of isolation using the colors pink, purple, and green to represent Phaseolus , Trifolium , and Vicia , Lens , and Lathyrus , respectively. Select strains within the core genome phylogeny are overlayed with a gray box in which the country of isolation is indicated. Trees were constructed using RAxML with Neorhizobium galegae HAMBI 540 set as the outgroup. Nodes with 100% bootstrap support are marked with a black circle. Phylogenetic trees depict Rhizobium species grouped by host plants: Phaseolus, Trifolium, and Vicia/Lens/Lathyrus. Introduced and naturalized inoculants are highlighted. R. leguminosarum species complex is outlined with dashed lines. On the symbiosis gene tree, the inoculants and other strains within the Rlc cluster into two clear clades based on their host of isolation ( Fig. 4 ). This separation is consistent with known host-strain specificities for these strains, as Rhizobium spp. that nodulate Trifolium or Vicia , Lens , and Lathyrus are known to be specific to those hosts ( 97 ). Strain TA1, originally isolated from Tasmania, is symbiotically nearly identical to R. leguminosarum 31B from Russia. Inoculant strains Rhizobium sp. SRDI969 and Rhizobium sp. SU303 both grouped with strains from the UK (248 and 3841) ( 98 ) and Tunisia (FB206) ( 99 ). Given that both the core and the symbiosis genes match closely with strains isolated overseas, as well as the geographical isolation of Tasmania and the lack of native Trifolium spp. prior to European colonization, it is highly likely that SRDI969, TA1, and SU303 are exotic strains that were introduced into Australia post-European colonization. Six inoculant strains are located outside the Rlc group on the core gene phylogeny. R. sullae WSM1592, the inoculant for the temperate legume Hedysarum coronarium , was isolated from Italy ( 25 ) and grouped separately on both the core and the symbiosis gene trees from other characterized strains, being most closely related to strains of R. gallium isolated from France, Spain, Mexico, and Canada ( 100 , 101 ). R. sophoriradicis CC511, the inoculant for Phaseolus spp. isolated from the USA ( 102 ), groups with other Phaseolus -nodulating Rhizobium on both the core and the symbiosis gene trees. Although R. hidalgonense CB782, the inoculant strain from Kenya for sub-tropical Trifolium semipilosum , did not cluster with other temperate Trifolium -nodulating strains within the Rlc on the core genome tree, it did group with these strains on the symbiosis tree. For Rhizobium sp. CB3060, the inoculant strain for the tropical legume Leucaena leucocephala isolated from north-eastern Australia, both the core and the symbiosis genes align closely with inoculant strain Rhizobium sp. CB3090 for Gliricidia spp. from Sri Lanka and Calliandra spp. inoculant Rhizobium sp. CB3171 from Nicaragua, as well as the Phaseolus -nodulating strain Rhizobium sp. NXC24 from Mexico. This suggests Rhizobium sp. CB3060 may have originated from Central America, which is the native origin of L. leuococephala , and have been introduced when the legume was brought to northern Australia in the 1890s ( 103 ). Strain R. laguerreae WSM1455 was a previously long-standing commercial inoculant strain for Vicia faba between 2002 and 2022, and was originally isolated from V. faba in Greece ( 3 , 5 ). The core genome of this strain groups with R. laguerreae FB206 T from Tunisia, while its symbiosis genes group with the historical inoculant for V. faba , WSM1274 from Greece and the current Pisum , Lens , and Lathyrus inoculant strain Rhizobium sp. WSM4643 from Italy ( Fig. 4 ). An incomplete genome sequence was submitted to NCBI in June of 2012, under the name R. leguminosarum bv. viciae WSM1455 (GCF_000271805.1). However, the chromosome sequence of GCF_000271805.1 clusters with R. acaciae , while the symbiosis genes cluster with R. laguerreae WSM1455, Rhizobium sp. WSM1274, and Rhizobium sp. WSM4643. The R. laguerreae WSM1455 and R. acaciae GCF_000271805.1 sequences share an ANI value of only 94.3%. The original source material for R. acaciae GCF_000271805.1 is no longer available, therefore it was not possible to verify why this sequence is so different from the R. laguerreae WSM1455 reported here. Nevertheless, the R. laguerreae WSM1455 strain sequenced in this work was sourced directly from AIRG, which houses the Australian mother culture collection that is supplied to inoculant manufacturers, and therefore represents the genome of this commercial inoculant strain. Discordance between core and symbiosis gene phylogenies has been noted previously ( 104 ) and is likely strongly influenced by conjugative transfer of symbiosis plasmids between Rhizobium strains ( 105 , 106 ). There is discordance between the core and the symbiosis gene trees. However, the strains that nodulate Phaseolus are not intermingled with the Trifolium and the Vicia , Lens , and Lathyrus rhizobia of the Rlc on the core gene tree ( Fig. 4 ). This is probably due to Pisum , Lens , Vicia spp. and many Trifolium spp. having centers of origin in Europe or western Asia ( 107 – 111 ), in contrast to Phaseolus bean, which was introduced more recently to this region from the Americas ( 112 ). This introduction likely brought Phaseolus bean-nodulating rhizobia in plant matter, seed, or soil, as has been suggested by García-Fraile et al. ( 113 ). That Phaseolus bean-nodulating rhizobia were geographically separated from other Rhizobium spp. for much of their evolutionary histories may contribute to the apparent separation of Phaseolus -nodulating rhizobia from the Rhizobium within the Rlc. Bradyrhizobium Comparison of the 19 Bradyrhizobium inoculant strains to 127 sequenced Bradyrhizobium strains showed they were widely dispersed throughout the core gene tree ( Fig. 5 ), indicating a high level of diversity among the commercial Bradyrhizobium inoculants. Even among seven Bradyrhizobium inoculant strains isolated from Australia, there was a significant diversity, with these strains belonging to six different species. Three strains are inoculants for temperate legumes, and the remaining four nodulate tropical legumes. Both Bradyrhizobium sp. WSM471 and WU425 are inoculant strains for Ornithopus and Lupinus . These strains are closely related to B. canariense BTA-1 T (Spain) and Bradyrhizobium sp. WSM1253 (Greece) on the core genome tree, and they cluster in a closely related clade on the symbiosis tree ( Fig. 5 ). Both Ornithopus and Lupinus spp. are introduced legume genera that have been grown in Australia for over a century. Initially, these legumes required inoculation with compatible rhizobia; however, over time, compatible strains have become established in Australian soils ( 114 ). Both WSM471 and WU425 are therefore exotic strains of European origin, which is consistent with a previous report based on multilocus sequence typing (MLST) analysis ( 114 ). Strain B. barranii CC1502, which is the inoculant strain for Chamaecytisus palmensis , an introduced tree legume species, grouped with the introduced Lotus pedunculatus strain B. barranii CB829 (from the USA) on the core genome tree but clustered within the Lupinus / Ornithopus group within the symbiosis gene tree. This suggests that like WSM471 and WU425, CC1502 likely represents an exotic strain that has become naturalized to Australian soils. Fig 5 Core ( A ) and symbiosis ( B ) gene phylogenies of the genus Bradyrhizobium . Strains are overlayed with their host legume of isolation using the colors brown, blue, purple, green, and pink to represent Acacia , Aeschynomene , Glycine max , Lotus , and Lupinus and Ornithopus , respectively. Select strains within the core genome phylogeny are overlayed with a gray box in which the country of isolation is indicated. Trees were constructed using RAxML with Bradyrhizobium sp. WD16 set as the outgroup. Nodes with 100% bootstrap support are marked with a black circle. Strains B. commune BDV5040 T and B. semiaridum WSM 1704 T are missing several symbiosis genes used for alignment construction and were therefore left off the symbiosis gene tree. Phylogenetic trees illustrate Bradyrhizobium species grouped by host plants: Acacia, Aeschynomene, Glycine max, Lotus, and Lupinus/Ornithopus. Introduced and naturalized inoculants, as well as countries of isolation, are highlighted. For the remaining four inoculant strains isolated from Australia that nodulate tropical legumes, a complex picture of core and symbiosis gene phylogenies is evident. This is partly influenced by the paucity of host range data for many of these strains. Where host range data exist, there is a tendency for some Bradyrhizobium strains to fix N 2 effectively across many legume genera. For example, Bradyrhizobium sp. CB756 is effective on species across 19 genera, Bradyrhizobium sp. CB1015 is effective across eight genera, and B. yuanmingense CB1024 is effective across 15 genera ( 85 ). Nevertheless, the geographical origins of inoculant strains can still be inferred from core and symbiosis gene phylogenies of these organisms. Inoculant strain B. huanghuaihainense CB3035 for Cyamopsis tetragonoloba (native to India, Pakistan, and Western Himalaya) was most similar to B. huanghuaihainense CGMCC1.10948 T isolated from Glycine max from China, and more distantly similar to CB1717 for Macroptilium bracteatum (native to South America), which was isolated from Brazil, on the core genome tree, but its symbiosis genes were more closely related to Bradyrhizobium sp. CB1650, which is the commercial inoculant strain for Stylosanthes hamata from Brazil. In contrast, the commercial inoculant strain for a range of Stylosanthes spp. (excluding S. hamata ), Bradyrhizobium sp. CB82, which was isolated from Australia, was only distantly related to CB1650 on the core and symbiosis genome trees, instead being more closely related to B. arachidis SM32 from China, and B. centrolobii BR 10245 T and B. neotropicale BR 10247 T isolated from Centrolobium paraense in Brazil on both trees ( Fig. 5 ). Similarly, Aeschynomene americana and A. falcata inoculant strain Bradyrhizobium sp. CB2312, which was isolated in Australia, is closest to B. arachidis strain CB756 (from Zimbabwe) on both core and symbiosis gene trees but not to other Aeschynomene sp.-nodulating strains, which form a distinct cluster in both phylogenies and were isolated from the North America, Africa, and Japan ( 115 – 118 ). Finally, B. brasilense 5G1B, which is the inoculant strain for Vigna angularis isolated from Australia, is most closely related to B. pachyrizi CB1923 from Brazil, the inoculant strain for Centrosema pascuorum and C. pubescens at the core genome level, but is symbiotically closer to G. max strains B. elkanii USDA76 T and USDA61. There is a suite of well-characterized strains of Bradyrhizobium spp. isolated from Australian native legumes including B. agreste CNPSo 4010 T , B. archetypum WSM1744 T , B. australiense WSM1791 T , B. cenepequi CNPSo 4026 T , B. commune BDV 5040 T , B. diversitatis CNPSo 4019 T , B. glycinis CNPSo 4016 T , B. hereditatis WSM1704 T , and B. murdochi WSM1790 T . The genome sequences of these strains do not cluster with the seven Bradyrhizobium sp. inoculants isolated from Australia on the core gene tree. This suggests that these seven inoculant strains are exotic organisms. However, we cannot rule out the possibility that they represent geographically widespread organisms that could also have broad host ranges. Conclusions and future perspectives Australian commercial legume inoculants are composed of a wide diversity of organisms, which span five known rhizobial genera and at least 19 different species. Twenty-three strains could be definitively identified at the species level, while 19 strains could only be conclusively defined at the genus level. With our knowledge of rhizobia-legume symbioses built upon a narrow suite of strains and host organisms ( 1 , 119 , 120 ), sequencing of a greater diversity of rhizobia is required to bolster databases and make them more representative of rhizobial populations. Within the genera analyzed in this study, varying degrees of incongruency between core and symbiosis gene phylogenies were observed, with the level of discordance suggesting a genus-based hierarchy of HGT frequency of Mesorhizobium > Rhizobium > Sinorhizobium > Bradyrhizobium . While there is clear evidence for HGT and its impact on N 2 fixation for Mesorhizobium spp. in the field ( 9 , 10 , 77 , 121 ), the same is not true for the other rhizobia genera. For Rhizobium spp. and Sinorhizobium spp., where pSyms have been shown to be mobile in vitro ( 14 , 15 ), their environmental transfer is yet to be directly observed. Similarly, several studies have concluded symbiosis gene HGT is likely an important driver for the evolution of Bradyrhizobium spp. symbionts ( 58 , 122 – 124 ), but the mechanism of transfer for these genes has yet to be elucidated. Inoculant strains isolated from Australian soils were shown to be similar to exotic strains. This suggests that these inoculants are the result of inadvertent introductions of rhizobia, possibly arriving along with exotic soil, legume seed, or plant material, colonizing Australian soils and subsequently being isolated and developed as inoculants. However, with a paucity of sequence data of native rhizobia, the possibility remains that some of these commercial strains may be indigenous bacteria with a capacity to fix N 2 with introduced legumes. This is particularly pertinent for the Bradyrhizobium inoculant strains, where there is substantial evidence that native legumes are nodulated by this genus ( 125 – 129 ). Sequencing more rhizobia isolated from indigenous legume hosts would improve the resolution of this analysis and enable this question to be answered. The rhizobia analyzed in this study are a cohort of strains with high saprophytic competence and N 2 fixation efficiency for targeted host legumes. They provide a blueprint to allow the development of a sequence-based approach to identify nodule occupancy in the field. These genome sequences are also a highly valuable resource for interrogation into free-living persistence of rhizobia and their evolution as legume N 2 -fixing symbionts."
} | 14,145 |
24191970 | null | s2 | 8,314 | {
"abstract": "Quorum sensing, a group behaviour coordinated by a diffusible pheromone signal and a cognate receptor, is typical of bacteria that form symbioses with plants and animals. LuxIR-type N-acyl L-homoserine (AHL) quorum sensing is common in Gram-negative Proteobacteria, and many members of this group have additional quorum-sensing networks. The bioluminescent symbiont Vibrio fischeri encodes two AHL signal synthases: AinS and LuxI. AinS-dependent quorum sensing converges with LuxI-dependent quorum sensing at the LuxR regulatory element. Both AinS- and LuxI-mediated signalling are required for efficient and persistent colonization of the squid host, Euprymna scolopes. The basis of the mutualism is symbiont bioluminescence, which is regulated by both LuxI- and AinS-dependent quorum sensing, and is essential for maintaining a colonization of the host. Here, we used chemical and genetic approaches to probe the dynamics of LuxI- and AinS-mediated regulation of bioluminescence during symbiosis. We demonstrate that both native AHLs and non-native AHL analogues can be used to non-invasively and specifically modulate induction of symbiotic bioluminescence via LuxI-dependent quorum sensing. Our data suggest that the first day of colonization, during which symbiont bioluminescence is induced by LuxIR, is a critical period that determines the stability of the V. fischeri population once symbiosis is established."
} | 354 |
38966714 | PMC11223108 | pmc | 8,316 | {
"abstract": "Stochastic Calculus-guided Reinforcement learning (SCRL) is a new way to make decisions in situations where things are uncertain. It uses mathematical principles to make better choices and improve decision-making in complex situations. SCRL works better than traditional Stochastic Reinforcement Learning (SRL) methods. In tests, SCRL showed that it can adapt and perform well. It was better than the SRL methods. SCRL had a lower dispersion value of 63.49 compared to SRL's 65.96. This means SCRL had less variation in its results. SCRL also had lower risks than SRL in the short- and long-term. SCRL's short-term risk value was 0.64, and its long-term risk value was 0.78. SRL's short-term risk value was much higher at 18.64, and its long-term risk value was 10.41. Lower risk values are better because they mean less chance of something going wrong. Overall, SCRL is a better way to make decisions when things are uncertain. It uses math to make smarter choices and has less risk than other methods. Also, different metrics, viz training rewards, learning progress, and rolling averages between SRL and SCRL, were assessed, and the study found that SCRL outperforms well compared to SRL. This makes SCRL very useful for real-world situations where decisions must be made carefully. • By leveraging mathematical principles derived from stochastic calculus, SCRL offers a robust framework for making informed choices and enhancing performance in complex scenarios. • In comparison to traditional SRL methods, SCRL demonstrates superior adaptability and efficacy, as evidenced by empirical tests.",
"conclusion": "Conclusion The study's conclusion establishes that Stochastic Calculus-guided Reinforcement Learning (SCRL) is an encouraging option for making decisions in ambiguous settings, promising to choose and perform better in complicated scenarios. By employing mathematical concepts rooted in stochastic calculus, SCRL provides a solid theoretical basis for decision-making and enhances performance. Field tests reveal SCRL's superiority over traditional Stochastic Reinforcement Learning (SRL) techniques, proving more adaptive and effective. Both short-term and long-term risks and dispersion are lower with SCRL than those with SRL: The former's dispersal value is 63.49 against the latter's 65.96. Additionally, SCRL demonstrates significantly lower short-term risk (0.64) and long-term risk (0.78) compared to SRL, which has short-term and long-term risk values of 18.64 and 10.41, respectively. Lower risk values indicate more excellent stability and reliability in decision-making processes. Moreover, assessments of various metrics, including training rewards, learning progress, and rolling averages, consistently show SCRL's superior performance over SRL. However, it's essential to acknowledge some study limitations, such as exploring SCRL in diverse environments further and considering additional performance metrics. Future studies could focus on refining SCRL algorithms, exploring their applicability in real-world scenarios, and addressing scalability issues to enhance their practical utility."
} | 775 |
33294406 | PMC7691556 | pmc | 8,320 | {
"abstract": "Highlights • Glucose dehydrogenase from Pseudomonas taetrolens could produce maltobionic acid. • The glucose dehydrogenase gene was homologously expressed in P. taetrolens . • Maltose-oxidizing activity and MBA production titer of P. taetrolens was improved. • MBA productivity of the recombinant P. taetrolens was increased to 9.52 g/L/h. • High-level production of MBA from HMCS was achieved by the recombinant P. taetrolens .",
"conclusion": "4 Conclusion We discovered that P. taetrolens could produce MBA from maltose and that the quinoprotein glucose dehydrogenase (GDH) of P. taetrolens could act as a maltose-oxidizing enzyme. The homologous expression of GDH in P. taetrolens improved maltose-oxidizing activity and MBA production. Optimized culture conditions improved the MBA production by recombinant P. taetrolens . The MBA production, yield, and productivity of 200 g/L, 100 %, and 9.52 g/L/h, respectively, in batch fermentation using pure maltose were higher than any other reported results of MBA production, and 200 g/L, 100 % and 6.67 g/L/h, respectively for HMCS. The increased cost due to the decreased MBA productivity from HCMS was offset by the 94.6-fold lower cost of HMCS. Thus, HMCS was a better substrate for industrial MBA production than pure maltose. Therefore, recombinant P. taetrolens and HMCS are suitable for the commercial production of MBA based on microbial fermentation.",
"introduction": "1 Introduction Sugar acids are organic acids derived from the direct oxidation of mono- or oligosaccharides such as gluconic acid, glucaric acid, xylonic acid, and lactobionic acid (4- O - β -galactopyranosyl-D-gluconic acid; LBA) [ 1 ]. Among these sugar acids, lactobionic acid, a type of aldonic acid, has received considerable attention because it can chelate metal, and is biocompatible, biodegradable, antioxidant, antimicrobial, moisturizing, and nontoxic [ 2 , 3 ]. Lactobionic acid is used in the cosmetic, pharmaceutical, food, and chemical industries [ 4 , 5 ]. At present, LBA is mainly synthesized by chemical oxidation, which requires high-energy metal catalysts that are costly, harmful, and generate unwanted byproducts [ 6 , 7 ]. Biological method of producing LBA by microbial fermentation has been intensively investigated as an alternative to chemical oxidation because of its high selectivity, efficiency, and eco-friendliness [ 2 , 4 ]. Nonpathogenic P. taetrolens does not generate byproducts from lactose and it can produce large quantities of LBA [ 4 , 8 ]. Maltobionic acid (4- O - α -D-glucopyranosyl-D-gluconic acid, MBA) is another aldonic acid obtained from maltose oxidation. It is a stereoisomer of LBA with similar physicochemical characteristics to those of LBA, such as biocompatible, biodegradable, antioxidant, metal chelating, nontoxic, and moisturizing properties [ 9 ] that are useful to the food, cosmetic, and pharmaceutical industries [ [10] , [11] , [12] ]. However, the applications of MBA are somewhat narrower than those of LBA because the period of application research of MBA is shorter than that of LBA and the production volume of MBA is lower. Therefore, the development of an efficient production process is important for expanding the applications of MBA, which has also been chemically produced like LBA [ 13 ]. Because the problems associated with the chemical production of MBA and LBA are similar, microbial fermentation method of MBA production is of interest. However, microbial MBA production has not been investigated in detail. To date, two reports have described microbial MBA production by Pseudomonas sp. and Pseudomonas fragi , but the production was much lower (94.7 g/L) than that of LBA (400 g/L) [ [14] , [15] , [16] ]. We previously characterized the substrate specificity of P. taetrolens for various saccharides and found that it could produce not only LBA but also MBA (submitted for publication). We also found that the glucose dehydrogenase (GDH, GenBank accession number WP048384179), which is one of the GDHs from P. taetrolens , could convert lactose into LBA. Some GDHs have been reported to convert lactose into LBA [ 17 , 18 ]. Up to now, all these GDHs have been originated from gram-negative bacteria and are known to quinoprotein GDH (EC 1.1.5.2) using PQQ as a cofactor. Moreover, some quinoprotein GDHs can also convert maltose into MBA [ 18 , 19 ]. Here, we aimed to enhance MBA production by optimizing the substrate concentration, growth temperature, aeration, and cell density of seed cultures using a genetically modified P. taetrolens strain, which homologously expresses the quinoprotein GDH. To our knowledge, the MBA titer, yield, and productivity from pure maltose or high-maltose corn syrup (HMCS) substrates were higher than those of a previous study of applying microorganisms to produce MBA [ 14 , 15 ] and comparable to those obtained in microbial LBA production [ 20 ].",
"discussion": "3 Results and discussion 3.1 Heterologous expression of quinoprotein glucose dehydrogenase in E. coli to confirm maltose conversion into maltobionic acid In our previous study, we found that the quinoprotein GDH (GenBank accession number WP048384179) from P. taetrolens could convert lactose into LBA, as described in detail in another manuscript (submitted for publication). Some quinoprotein GDHs can not only convert lactose into LBA, but also convert maltose into MBA [ 18 , 19 ]. To examine whether the GDH from P. taetrolens could also convert maltose into MBA, we tried to heterologously express GDH genes in bacterium unable to produce MBA and confirm formation of MBA in the recombinant bacterium. In our previous work, we found that E. coli strain, which the quinoprotein glucose dehydrogenase (GCD) gene was inactivated, could not produce MBA from maltose. Thus, we used E. coli Δgcd strain to verify the GDH from P. taetrolens could produce MBA from maltose. The GDH gene was cloned into pKK223-3 to generate pKK-GDH, which was then transformed into E. coli Δgcd to construct recombinant E. coli Δgcd (pKK-GDH). This recombinant strain was cultivated to assess the formation of MBA from maltose by HPLC. Fig. 1 shows that E. coli Δgcd (pKK-GDH) converted maltose into MBA. This result indicated that the GDH from P. taetrolens had maltose-oxidizing activity and could produce MBA from maltose. Fig. 1 Assessment of MBA production in recombinant E. coli by HPLC. Fig. 1 3.2 Effects of GDH expression on maltose-oxidizing activity and MBA production in P. taetrolens We cloned gdh into pDSK519 to generate pDSK-GDH, which was then transformed into P. taetrolens . Wild-type and recombinant P. taetrolens (pDSK-GDH) strains were incubated in non-baffled flasks for 12 h. The crude intracellular maltose-oxidizing activity of P. taetrolens (pDSK-GDH) was approximately 38 % higher than that of wild-type P. taetrolens (1.98 ± 0.11 vs. 1.44 ± 0.21 U/μg; Fig. 2 ). This result indicated that homologous GDH expression effectively increased the intracellular maltose-oxidizing activity in P. taetrolens . The production of MBA was also compared between wild-type and recombinant P. taetrolens (pDSK-GDH) after 12 h of cultivation. The MBA production (g/L) of P. taetrolens (pDSK-GDH) was approximately 15 % higher than that of wild-type P. taetrolens (15.3 ± 1.03 vs. 13.3 ± 1.25 g/L; Fig. 2 ). Thus, the homologous expression of GDH could increase maltose-oxidizing activity and MBA production in P. taetrolens . We then optimized the culture conditions to improve MBA production by the recombinant P. taetrolens strain. In our previous study, we achieved high-level production of LBA using the wild-type P. taetrolens by optimizing the culture conditions such as pH and growth temperature [ 20 ]. We could highly improve the LBA production concentration in flask and fermenter cultures by using CaCO 3 as a pH control agent. Thus, in this study, we also used CaCO 3 as a pH control agent for increasing MBA production concentration. Fig. 2 Intracellular maltose-oxidizing activity and MBA production of wild-type (W.T.) and recombinant P. taetrolens (pDSK-GDH) in flask culture. Bars, Intracellular maltose-oxidizing activity (blank) and MBA production (filled) at 12 h of cultivation time. Error bars represent standard deviation of three independent experiments. MBA, maltobionic acid; W.T., wild-type. Fig. 2 3.3 Effects of initial maltose concentration on MBA production To investigate the effects of initial maltose concentration on MBA production, recombinant P. taetrolens cells were cultivated in NB supplemented with 50, 100, 150, 200, 300, 400, and 500 g/L of maltose) in non-baffled flasks at 25 °C. The seed culture was inoculated at an initial cell density of 0.1 at OD 600nm . Cell growth was inhibited as the initial maltose concentration increased to > 200 g/L ( Fig. 3 A). The maltose consumption rate and MBA productivity also decreased at initial maltose concentrations > 200 g/L ( Fig. 3 B and 3C). The MBA productivity was maximal at 200 g/L maltose, which was approximately 211 % higher than that at 500 g/L maltose (3.33 vs. 1.07 g/L/h). These findings indicated that maltose concentrations ≤ 200 g/L did not obviously affect cell growth and increased the MBA productivity of the recombinant P. taetrolens . Thus, 200 g/L maltose was applied for subsequent investigations. Fig. 3 Effects of initial maltose concentration on MBA production by P. taetrolens (pDSK-GDH). P. taetrolens cells (initial cell density at OD 600nm , 0.1) were cultivated in 300-mL baffled flasks containing 50 mL NB with 50, 100, 150, and 200 g/L of maltose and 30 g/L CaCO 3 at 200 rpm and 25 °C. Time course of cell growth (A), maltose consumption (B) and MBA production (C). Error bars, standard deviation of three independent experiments. GDH, glucose dehydrogenase; MBA, maltobionic acid; NB, nutrient broth; OD, optical density. Fig. 3 3.4 Effects of growth temperature on MBA production To examine the effects of growth temperature on MBA production, recombinant P. taetrolens strains were cultivated at 20 °C, 25 °C, 30 °C, and 35 °C in NB containing 200 g/L maltose in non-baffled flasks. The seed culture was inoculated at an initial cell density of 0.1 at OD 600nm . At 35 °C, the growth of P. taetrolens was significantly hindered ( Fig. 4 A) and maltose conversion into MBA was negligible ( Fig. 4 C). The maltose consumption rate and MBA productivity (3.33 g/L/h, respectively) were maximal at 25 °C ( Fig. 4 B and C). Therefore, we selected a growth temperature of 25 °C for subsequent investigations. Fig. 4 Effects of temperature on MBA production by P. taetrolens (pDSK-GDH). P. taetrolens cells (initial cell density at OD 600nm , 0.1) were cultivated in 300-mL baffled flasks containing 50 mL of NB with 200 g/L of maltose and 30 g/L CaCO 3 at 20 °C, 25 °C, 30 °C, and 35 °C. Time course of cell growth (A), maltose consumption (B) and MBA production (C). Error bars, standard deviation of three independent experiments. GDH, glucose dehydrogenase; MBA, maltobionic acid; NB, nutrient broth; OD, optical density. Fig. 4 3.5 Effects of aeration on MBA production To investigate the effects of aeration on MBA production, P. taetrolens (pDSK-GDH) was cultivated in 300-mL baffled and non-baffled flasks at 25 °C and 200 rpm in 50 mL of NB supplemented with 200 g/L maltose. The seed culture was inoculated at an initial cell density of 0.1 at OD 600nm . The maximum cell density at OD 600nm was 5.5 in both types of flasks ( Fig. 5 A). Although 200 g/L of maltose was fully converted into MBA, the maltose consumption rates and MBA productivity were approximately 123 % higher in the baffled, than the non-baffled flasks (7.41 vs. 3.33 g/L/h; Fig. 5 B and C), showing that a high aeration rate obviously improves MBA productivity. Therefore, baffled flasks were used in subsequent investigations. Fig. 5 Effects of aeration on MBA production in P. taetrolens (pDSK-GDH). P. taetrolens cells (initial cell density at OD 600nm , 0.1) were cultivated in 300-mL baffled and non-baffled flasks containing 50 mL of NB with 200 g/L of maltose and 30 g/L CaCO 3 at 200 rpm and 25 °C. Time course of cell growth (A), maltose consumption (B) and MBA production (C). Error bars, standard deviation of three independent experiments. GDH, glucose dehydrogenase; MBA, maltobionic acid; NB, nutrient broth; OD, optical density. Fig. 5 3.6 Effects of seed culture cell density on MBA production To examine the effects of seed culture cell density on MBA production, cultured recombinant P. taetrolens was inoculated at densities of 0.1, 0.3, 0.5, and 1.0 at OD 600nm in 50 mL of NB containing 200 g/L maltose and cultivated at 25 °C. The maltose consumption rate and MBA productivity were improved as the cell density of seed culture increased ( Fig. 6 A, B and C). The MBA productivity was approximately 28 % higher and maximal (9.52 vs. 7.41 g/L/h), at an initial cell density of 1.0, compared with 0.1. Thus, we seeded cells at a density of 1.0 in further investigations. Fig. 6 Effects of density of seed culture on MBA production in P. taetrolens (pDSK-GDH). Cultured P. taetrolens cells were inoculated at densities of 0.1, 0.3, 0.5, and 1.0 (OD 600nm ) in 50 mL of NB with 200 g/L maltose, and 30 g/L CaCO 3 , and cultivated at 25 °C. Time course of cell growth (A), maltose consumption (B) and MBA production (C). Error bars standard deviation of three independent experiments. GDH, glucose dehydrogenase; MBA, maltobionic acid; NB, nutrient broth; OD, optical density. Fig. 6 3.7 Batch fermentation for MBA production in the bioreactor We scaled up MBA production by batch fermentation using P. taetrolens (pDSK-GDH) in a 5-L fermenter under the culture conditions optimized in the flask culture investigation. The oxygen supply should be adequate for MBA, which is synthesized by the oxidation of maltose in P. taetrolens . When the level of DO was maintained at 30 % during the fermentation process, the MBA production, yield from pure maltose, and productivity were 200 g/L, 100 %, and 9.52 g/L/h, respectively ( Fig. 7 ). These results surpassed those of previous findings of microorganisms as MBA producers [ 14 , 15 ]. Fig. 7 Batch fermentation of pure maltose by P. taetrolens (pDSK-GDH) for MBA production under optimized culture conditions in 5-L jar bioreactors. P. taetrolens cells (the initial cell density was 1.0 at OD 600nm ) were cultivated in 5-L bioreactors containing 2 L NB medium with 200 g/L of maltose and 30 g/L CaCO 3 at 25 °C. The error bars represent the standard deviation from three independent experiments. GDH, glucose dehydrogenase; MBA, maltobionic acid; NB, nutrient broth; OD, optical density. Fig. 7 Batch fermentation of P. taetrolens (pDSK-GDH) also proceeded using the substrate HMCS in a 5-L fermenter using the optimized culture conditions. The MBA production titer, yield from maltose in the HMCS, and productivity were 200 g/L, 100 %, and 6.67 g/L/h, respectively ( Fig. 8 ). The MBA production titer and yield in batch fermentation using HMCS and pure maltose were the same, but the MBA productivity (6.67 g/L/h) was approximately 70.1 % of that (9.52 g/L/h) using pure maltose. Fig. 8 Batch fermentation of P. taetrolens (pDSK-GDH) for MBA production using high-maltose corn syrup under optimal culture condition in 5-L jar bioreactor. P. taetrolens cells (initial cell density at OD 600nm , 0.1) were cultivated in 5-L bioreactor containing 2 L NB with 200 g/L of maltose and 30 g/L CaCO 3 at 25 °C. Error bars, standard deviation of three independent experiments. GDH, glucose dehydrogenase; MBA, maltobionic acid; NB, nutrient broth; OD, optical density. Fig. 8 In an industrial point of view, the decrease in productivity is a disadvantage because it lengthens the process and thus increases the cost [ [25] , [26] , [27] ]. However, the bulk price of the HMCS ($370/ton) was approximately 94.6-fold lower than that of the pure maltose ($35,000/ton) ( https://www.alibaba.com ). This fairly low price of the HMCS, compared to that of the pure maltose, can sufficiently offset the relatively low MBA productivity and render the MBA production process more economically feasible."
} | 4,059 |
30541521 | PMC6292164 | pmc | 8,321 | {
"abstract": "Background Large-scale processing of lignocellulosics for glucose production generally relies on high temperature and acidic or alkaline conditions. However, extreme conditions produce chemical contaminants that complicate downstream processing. A method that mainly rely on mechanical and enzymatic reaction completely averts such problem and generates unmodified lignin. Products from this process could find novel applications in the chemicals, feed and food industry. But a large-scale system suitable for this purpose is yet to be developed. In this study we applied simultaneous enzymatic saccharification and communition (SESC) for the pre-treatment of a representative lignocellulosic biomass, cedar softwood, under both laboratory and large-scale conditions. Results Laboratory-scale comminution achieved a maximum saccharification efficiency of 80% at the optimum pH of 6. It was possible to recycle the supernatant to concentrate the glucose without affecting the efficiency. During the direct alcohol fermentation of SESC slurry, a high yield of ethanol was attained. The mild reaction conditions prevented the generation of undesired chemical inhibitors. Large-scale SESC treatment using a commercial beads mill system achieved a saccharification efficiency of 60% at an energy consumption of 50 MJ/kg biomass. Conclusion SESC is very promising for the mild and clean processing of lignocellulose to generate glucose and unmodified lignin in a large scale. Economic feasibility is highly dependent on its potential to generate high value natural products for energy, specialty chemicals, feed and food application.",
"conclusion": "Conclusion The simultaneous enzymatic saccharification and comminution (SESC) process was recently developed to create a mild and environmentally-friendly process for the treatment of lignocellulosic biomass. Process optimization and commercial feasibility evaluation were conducted to further enhance its performance in generating natural products that may be used for a wide variety of novel applications. The pH has a significant effect on the saccharification efficiency of SESC, reaching 80% at pH 6. Solution recycling resulted in higher product concentration without affecting overall sugar recovery. The direct alcohol fermentation of SESC-generated slurry allows for the complete conversion of glucose to ethanol. SESC has huge potential to generate energy and high value products from woody biomass for industrial, agricultural and even feed and food applications, thus possibly countering major concerns regarding its enzyme and energy requirements.",
"discussion": "Discussion Mechanical comminution coupled with enzymatic reaction is the eco-friendliest approach for lignocellulosic processing. However, due to the negative views on the use of enzymes and intensive energy, measures must be undertaken to compensate for these limitations. First, it is essential to maximize the enzymatic activity through the optimization of reaction conditions. The results from the pH evaluation during SESC treatment confirmed that better saccharification is achieved at pH 6 due to reduced non-productive enzyme binding. However, quantification of free enzymes indicated the difficulty of completely eliminating this phenomenon even at the optimum pH. The operation protocol may also be modified to maximize enzyme productivity. We have confirmed the feasibility of enzyme recycling during SESC with a favorable effect on sugar recovery. The resulting concentrated product also makes downstream processing more practical. Finally, in-house enzyme production using substrates from lignocellulosic processing, such as nutrient-rich waste streams, could also help address the economic shortcomings of enzyme-based pre-treatment [ 19 ] by reducing purification, concentration, storage, and shipment expenses that are generally incorporated in the price of commercial enzymes [ 13 ]. The uncontaminated and natural characteristic of any SESC-generated waste is well adapted for this strategy. SESC optimization showed that an enzyme concentration of 0.2 mL/g wood biomass (corresponding to 30 FPU/g glucan above) for a single stage milling is necessary to achieve around 60% saccharification efficiency. On the other hand, established methods such as the SPORL process, which combines mild thermo-chemical treatment with enzymatic reaction has been reported to achieve a higher efficiency (90%) at lower enzyme loading (15 FPU). This is to be expected considering the possible synergism between acid and enzymatic hydrolysis. However, the mild operating conditions of SESC eliminates energy requirement for thermo-chemical reaction and at the same time generate less unwanted chemical by-products including furfural derivatives. In fact, the complete absence of inhibitors allowed for the direct alcohol fermentability of the sugar-rich SESC slurry at high sugar to ethanol conversion. To address the intensive energy requirement of SESC, energy usage generally declines at higher operation scale. For example, a lab-scale wet planetary ball milling requires at least 2000 MJ/kg biomass (for rice straw), but an industrial system operating at 20 tons of biomass per h with a 400-kW engine only consumes 0.64 MJ/kg for the same level of treatment [ 15 ]. This provides a positive energy forecast for large-scale processing, but actual evaluations are necessary for SESC. Another path to economic feasibility would be to utilize alternative energy sources such as watermills or windmills. Unlike thermo-chemical methods, beads milling mainly requires a continuous rotary motion, so that it may be configured to work from these natural sources. However, due to their maximum power output limitations (3–10 HP) [ 37 ], an assembly of milling units functioning at the operation scale threshold must be considered. It is also important to point out that the hardware for the commercialization of SESC are already available. Milling systems of various configurations and capacities are currently being employed for pharmaceutical, special chemicals and even food manufacturing. A grinding and dispersing machine (LME3000, Ashizawa Finetech) equipped with a 220-kW motor can be used to process 3 tons of wood slurry in a single operation. For much larger output, ultrafine grinders used in the mining industry, such as ISAMill™ with motors having power ratings up to 8000 kW and which operates under the same principle of rotating media comminution are suitable candidates [ 38 ]. These enormous milling machines may drive biorefineries of the future. The commercial viability of a biorefinery could best be elevated through product diversification. Lignocellulosics utilization should not only be limited to biofuels but instead, high-value chemicals must be prioritized [ 39 ]. This path to offset costs is highly suited for SESC. Figure 6 provides a list of products that may be directly and indirectly produced through this process. The mild nature of mechanical comminution and the specificity of the enzymatic reactions means that the major output will be a sugar-rich supernatant and a native lignin-rich precipitate. The glucose-rich SESC supernatant and even the slurry itself can be directly used as a fermentation substrate to produce fuel compounds such as ethanol and methane as well as other important chemicals like lactic acid and acetic acid. Though not covered in this work, SESC also generated significant amounts of xylose, which is another high-value product (Fig. 3 ). Fig. 6 Valorization of lignocellulosic biomass by the SESC process. The major products, consisting of sugars and unmodified lignin, may be processed to generate energy, specialty chemicals, materials and even food-related compounds Non-degraded lignin is an important component that may brighten the financial outlook of SESC. Currently, majority of lignin are derived from the kraft process of wood pulping. Kraft lignin represents 90% of all lignin produced, but the presence of aliphatic thiol groups gives the material an unpleasant odor, thus confining its use to in-house fuel [ 40 ]. The availability of high-quality sulfur-free lignin in large quantities could create novel applications. For instance, we have used the lignin-rich SESC precipitate to synthesize composite materials with high fraction strain, elasticity and fire resistance, thus making the biopolymer an ideal antiplasticizer or thermoplastic elastomer [ 12 ]. Non-deteriorated lignin was also found to impart excellent heatproof property to synthetic polymers such as poly(ethylene carbonate) [ 41 ]. Lignin may also be decomposed to generate aromatic building blocks for various materials. Generally, lignin obtained through thermo-chemical reactions has undergone extensive modification so that they have fewer β-O-4 linkages. This lignin is highly condensed due to the formation of strong and recalcitrant carbon-carbon bonds [ 42 ], making the polymer less susceptible to complete depolymerization by chemical or microbiological means. In this connection, we have previously shown that SESC-lignin is comparable to neat lignin in terms of the aromatic yield following nitrobenzene oxidation, a clear indication of its non-denatured characteristics [ 12 ]. Easier liberation and higher yields of aromatics during depolymerization is essential for the production of various platform chemicals. One of the most significant lignin-derivatives that we have developed is 2-pyrone-4,6-dicarboxylic acid (PDC), which can be employed as an epoxy adhesive [ 43 ], flocculant for radioactive Ce trapping [ 44 ] as well as a building block for biobased polymers such as polyamide, polyester and polyurethane [ 45 – 49 ]. Lignin may also be used as a raw precursor for other compounds, such as DMSO, vanilla, phenol, and ethylene [ 39 ]. Different types of biopolymers may also be produced from other lignin-derived platform chemicals, such as muconate, muconolactone and β-ketoadipate [ 50 – 53 ]. Future trends in lignocellulosic processing could also be directed for feed or even food manufacturing [ 54 ]. Feed production alone enhances food security by decreasing the fraction of arable land that must be used to grow crops for animal use [ 55 ]. In this connection, wood molasses has long been used to supplement animal feeds [ 56 ]. As an additive to swine, cattle or poultry feeds, its nutritional value is comparable to molasses from other sources such as beet and sugarcane [ 54 ]. However, thermo-chemical pre-treatment complicates purification and lowers the quality and stability of wood molasses [ 57 ]. In contrast, the uncontaminated and natural SESC extract would only require a simple evaporation step to generate a molasses-like syrup (Fig. 6 ). In addition to sugar, sulfur-free SESC lignin could also be utilized as animal feed supplement. Pure lignin and its derivatives provide health benefits to monogastric animals [ 58 ]. Food-related uses of wood are currently limited. Wood cellulose may serve as thickener or binding agent in food products [ 59 ] or may be included in dietary supplements for fat adsorption [ 60 ]. The utilization of wood for its caloric value, however, is yet to be realized. By employing food-grade buffer and enzymes, SESC sugar concentrate can be made edible for this function. This applies to other fermentation products including ethanol and acetic acid. In fact, our laboratory is currently developing a spirit liquor from SESC-treated wood. As a strong indication of the mildness of reaction, the distilled alcohol has the aroma of the particular tree species employed. Previously, a novel procedure that converts lignocellulosic cellulose into starch was reported; it employs a cocktail of enzymes that direct the transformation of cellulose into amylose [ 11 ]. Despite the initially limited production capacity and economic hurdles, the researchers were very confident of its potential in addressing future food shortages. We have a similar outlook for the SESC process. Though more detailed and strict tests are necessary to realize this goal, ‘wood for food’ research is where SESC-based lignocellulosic valorization can create a unique and vital niche."
} | 3,042 |
31018013 | PMC6851700 | pmc | 8,323 | {
"abstract": "Abstract We propose four postulates as the minimum set of logical propositions necessary for a theory of pulse dynamics and disturbance in ecosystems: (1) resource dynamics characterizes the magnitude, rate, and duration of resource change caused by pulse events, including the continuing changes in resources that are the result of abiotic and biotic processes; (2) energy flux characterizes the energy flow that controls the variation in the rates of resource assimilation across ecosystems; (3) patch dynamics characterizes the distribution of resource patches over space and time, and the resulting patterns of biotic diversity, ecosystem structure, and cross‐scale feedbacks of pulses processes; and (4) biotic trait diversity characterizes the evolutionary responses to pulse dynamics and, in turn, the way trait diversity affects ecosystem dynamics during and after pulse events. We apply the four postulates to an important class of pulse events, biomass‐altering disturbances, and derive seven generalizations that predict disturbance magnitude, resource trajectory, rate of resource change, disturbance probability, biotic trait diversification at evolutionary scales, biotic diversity at ecological scales, and functional resilience. Ultimately, theory must define the variable combinations that result in dynamic stability, comprising resistance, recovery, and adaptation.",
"introduction": "Introduction Pulse events, defined as abrupt changes in ecological parameters, are ubiquitous in ecosystems and include a wide array of phenomena, such as heat waves, marine upwelling, mass reproductive and mortality events, and biomass‐altering disturbances (Yang and Naeem 2008 ). Understanding pulse events is important because of this ubiquity and because the frequencies and magnitudes of such events as droughts, fires, floods, windstorms, and pest outbreaks, are changing because of human influences including, most importantly, climate change, land‐use change, and species invasions (Franklin et al. 2016 , Seidl et al. 2017 , Loehman et al. 2018 , McDowell et al. 2018 ). In turn, pulse events can affect the responses of ecosystems to these influences, e.g., increasing invasion rates (Hobbs and Huenneke 1992 ) and accelerating or otherwise affecting responses to climate warming (DeFrenne et al. 2013 ). Characterizing pulse dynamics is also important as the basis for determining the degree of novelty of events (Hallett et al. 2013 ). Finally, event characteristics are central to understanding ecological resilience (Ratajczak et al. 2017 ), because pulse characteristics reflect ecosystem resistance, create the initial conditions for recovery, and are, by definition, the change to which the system may or may not be resilient (Carpenter et al. 2001 ). In this paper, we use a deductive approach to derive the minimum set of propositions, here called postulates (after Marquet et al. 2014 ), that create a general explanation for pulse dynamics across ecosystems and for places with different biogeographic histories. We then develop seven generalizations, phrased as predictions, for an important class of pulse events, biomass‐altering disturbances, that emerge from these postulates. Our interest in a theory of pulse dynamics developed from the challenge of generalizing disturbance and ecosystems dynamics (White and Jentsch 2001 ). Disturbance ecology has produced a number of conceptual frameworks over the last several decades (Shugart 1984 , White and Pickett 1985 , Pulsford et al. 2016 ), including the intermediate disturbance hypothesis (Connell 1978 , Fox 2013 , Shiel and Burselm 2013 ), the dynamic equilibrium model (Huston 1979 , 2014 ), the theory of nutrient dynamics (Vitousek and Reiners 1975 , Vitousek 1984 ), the theory of forest dynamics (Shugart 1984 ), and the theory of landscape dynamics (Turner et al. 1993 ). Important recent frameworks include disturbance interactions and cross‐scale perspectives (Peters et al. 2004 , Raffa et al. 2008 , Buma and Wessman 2011 , Buma 2015 , Cannon et al. 2017 ), the concepts of biological legacy and ecological memory (Johnstone et al. 2016 ), generalizable biogeochemical responses to ecosystem disturbance (Kranabetter et al. 2016 ), the network‐based view on the role of disturbance in biodiversity and productivity (Gross 2016 , Seidl et al. 2017 ), and patterns of trait diversity shaped by evolutionary trade‐offs (Diaz et al. 2016 ). We build on this past work by proposing a general theory of pulse dynamics and disturbance that serves as an overall structure for the insights that have developed over the last several decades. This structure consists of a set of four fundamental postulates and the relationships among them that together circumscribe the common denominators and rules for all pulse events. Although disturbances can initiate successional change, we do not review theories of successional mechanisms and pathways here (see Glenn‐Lewin et al. 1992 , Meiners et al. 2015 , Peet 1992 , Walker and del Moral 2003 , Pickett et al. 2011 , Walker and Wardle 2014 , Pulsford et al. 2016 ); nor do we treat the dynamics among species that occur after pulse initiation, such as trophic interactions (see Holt 2008 , Nowlin et al. 2008 , Karakoç et al. 2018 ). Definition and types of pulse events A pulse event is any abrupt change, positive or negative, in system parameters (Yang et al. 2008 ), with abruptness defined as the magnitude of the change divided by its duration (White and Jentsch 2001 ). Pulse events can be characterized by seven continuous variables that describe the dimensions of a particular parameter change: magnitude, duration, abruptness, initial pulse rate, the rate of recovery, the degree of recovery, and the area under the pulse curve (Fig. 1 ). Because magnitude and duration are continuous variables, pulse events become gradual changes as magnitude decreases and duration increases (sometimes called “presses”). This means that “pulsedness” (Yang et al. 2008 ) is, itself, a continuous variable (Fig. 2 A, modified from Yang et al. 2008 ). However, we posit that resistance to pulse forces, discreteness of individual organisms, finite life spans, limits to niche breadth, and the interval between pulse events often result in thresholds and, therefore, discontinuities and patchiness. In essence, abruptness and discreteness can develop, at least at the scale of the individual, because event forces meet with biotic resistance that is ultimately limited. Similarly, the spatial propagation of some pulse events is characterized by thresholds leading to either low‐ or high‐magnitude events, for example, for disturbances such as insect outbreak and fire (Peters et al. 2004 ). Figure 1 Quantifying the dimensions of pulse events. Seven variables that define pulse events: magnitude, duration, abruptness (magnitude/duration), initial rate of change, rate of recovery, magnitude of recovery, and the total pulse effect (area under the curve). Figure 2 The characteristics that define pulse dynamics. (A) Pulse events vary by magnitude of change and abruptness. (B) Pulse regimes vary in periodicity, with the degree of variation in periodicity determining predictability. (C) The inverse relationship between magnitude and frequency of pulse events; the area below the curve is shaded because low‐magnitude pulses can occur at any frequency, but high‐magnitude pulses are generally constrained to low frequency (modified from White and Jentsch 2001 ). (D) The rate of change after pulse events (illustrated here by a high‐magnitude biomass‐altering disturbance) is initially limited by rate of colonization and organism response (lower box), and is finally limited by diminishing resources or space (upper box), with a maximum recovery rate at intermediate time since pulse initiation (middle box). (E) The initial rate of change (dashed lines) varies with pulse magnitude. (F) Pulse events initiate secondary pulses that can lead to synergisms such as feedback loops or cascades. Disturbance has been defined both broadly and narrowly. In its broad sense, disturbance encompasses all pulse events (White and Pickett 1985 ). In its narrow sense, disturbance applies to a special class of pulse events characterized by direct alteration of biomass and ecosystem structure (Grime 1979 ), termed here “biomass‐altering disturbance” (see below). Under both definitions, the pulse perspective focuses on the dimensions of resource change and the mechanisms of response, whether a single disturbance event causes one resource pulse or a cascade of pulses over time. We can recognize four types of pulse events (modified from Yang and Naeem 2008 ): (1) fluctuation in physical environmental conditions such as heat waves or droughts; (2) abiotic changes in resource supplies such as those caused by ocean upwelling or lake turnover; (3) changes in biotic resources through sudden demographic events, such as mass reproductive or mortality events (Holt 2008 , Yang and Naeem 2008 , Yang et al. 2008 ); and (4) changes through abrupt alteration of biotic structure, that is, biomass‐altering disturbances. Spatial subsidies, that is, the transfer of resources across space, have been recognized as a fifth category (Yang and Naeem 2008 ); however, these transfers involve forces such as wind, water flow, and gravity in the movement of resources, organic materials, soils, or geological substrates and thus can be considered forms of the second (abiotic changes in resource supply) or fourth (biomass‐altering disturbances) pulse types. Indeed, pulse events almost always result in the spatial movement of resources, though the scale of this movement varies from local to global. By including biomass‐altering disturbance within the larger framework of pulse dynamics, we explore the insight that disturbance can be generalized through the description of pulses of resource change, varying in direction, magnitude and rate, observable at a wide range of scales, and occurring at any trophic level. Three scales are immediately apparent in pulse dynamics: the scale of the individual pulse event in time and space, the scale of multiple patches and events (i.e., the landscape scale, called the multipatch scale in White and Jentsch 2001 ), and the biogeographic scale (i.e., continent to global variation in environment and species pools). At the patch scale, the initial pulse sets off a sequence of further changes (Figs. 2 and 3 ) that are determined by abiotic and biotic processes, including changing ratios among resources and the transfer of resources to and from biotic and abiotic pools and across trophic levels (Bazzaz 1983 , Bender et al. 1984 , Holt 2008 , Yang and Naeem 2008 ). At large scales, pulse dynamics are described by the spatial and temporal distributions of pulse events, such as size, dispersion, magnitude, frequency, and predictability (the “disturbance regime”; Fig. 2 ). The phrase “patch dynamics” has been used to describe ecosystem pattern and process at multiple scales in the disturbance ecology literature (Thompson 1978 , White and Pickett 1985 ). Here, we use “pulse dynamics,” after Yang et al. ( 2008 ), to emphasize change in resources and environment. We retain the concept and phrase “patch dynamics” as the third of the four postulates to treat the spatial and temporal distribution of patches and events. The third scale in our treatment, the biogeographic scale, incorporates variation in environmental conditions and species pools. There are no fixed spatial or temporal dimensions for these three scales, but analytic scales can be derived in units relative to the size, dispersal characteristics, growth rates, and life spans of organisms, or from experimental and observational designs that allow analysis of the scale dependence of ecological responses. Figure 3 Resource stoichiometry and hierarchy in limiting factors in pulse dynamics. Lengths of arrows indicate resource amounts, from which resource ratios can be calculated, for example, in relation to the limiting resource. Time 1 = resource ratios during prepulse reference dynamics, Time 2 = resource ratios after the pulse event, Time 3 = resource ratios during return to prepulse conditions. Bold arrow shows the limiting resource at each time, which changes throughout recovery, for example, from light to water to nitrogen. Solid line shows the ratios among all resources. Dashed lines show the resources that are in excess (vulnerable to loss). Some resources change in a correlated pattern, others independently (not shown). Some resources remain static in amount but become limiting when other resources change."
} | 3,177 |
25525497 | null | s2 | 8,324 | {
"abstract": "The distribution of food resources in space and time is likely to be an important factor governing the type of foraging strategy used by ants. However, no previous systematic attempt has been made to determine whether spatiotemporal resource distribution is in fact correlated with foraging strategy across the ants. In this analysis, I present data compiled from the literature on the foraging strategy and food resource use of 402 species of ants from across the phylogenetic tree. By categorizing the distribution of resources reported in these studies in terms of size relative to colony size, spatial distribution relative to colony foraging range, frequency of occurrence in time relative to worker life span, and depletability (i.e., whether the colony can cause a change in resource frequency), I demonstrate that different foraging strategies are indeed associated with specific spatiotemporal resource attributes. The general patterns I describe here can therefore be used as a framework to inform predictions in future studies of ant foraging behavior. No differences were found between resources collected via short-term recruitment strategies (group recruitment, short-term trails, and volatile recruitment), whereas different resource distributions were associated with solitary foraging, trunk trails, long-term trail networks, group raiding, and raiding. In many cases, ant species use a combination of different foraging strategies to collect diverse resources. It is useful to consider these foraging strategies not as separate options but as modular parts of the total foraging effort of a colony."
} | 404 |
34278013 | PMC8260767 | pmc | 8,325 | {
"abstract": "Ubiquitously distributed microorganisms are natural decomposers of environmental pollutants. However, because of continuous generation of novel recalcitrant pollutants due to human activities, it is difficult, if not impossible, for microbes to acquire novel degradation mechanisms through natural evolution. Synthetic biology provides tools to engineer, transform or even re-synthesize an organism purposefully, accelerating transition from unable to able, inefficient to efficient degradation of given pollutants, and therefore, providing new solutions for environmental bioremediation. In this review, we described the pipeline to build chassis cells for the treatment of aromatic pollutants, and presented a proposal to design microbes with emphasis on the strategies applied to modify the target organism at different level. Finally, we discussed challenges and opportunities for future research in this field.",
"introduction": "1 Introduction With rapid growth of the global population and material consumption, discharge of various pollutants continues to increase, and environmental pollution has become one of the most severe issues affecting human health [ 1 ]. Pollutants refer to substances that can cause environmental pollution and have an adverse effect on the environment if discharged into the atmosphere, water or soil during human daily life [ 2 ]. Environmental pollution have direct damage to the ecological systems, such as water deterioration, forest destruction, and desertification, or indirect damage to human [ 3 ]. Among the major environmental pollutants, aromatic compounds are of great concern because they will be persistent in the environment due to the high thermodynamic stability of the benzene group. How to degrade these pollutants is currently a key challenge in environmental pollution control. Thanks to diverse types of microbial metabolism, most pollutants can be degraded or transformed by certain microorganisms [ 4 ]. The microbial degradation of aromatic pollutants has been developed for 40 years and has always been a hot topic in environmental protection (see Fig. 1 ). Naturally genes involved in aromatic pollutants degradation generally exist in clusters and are often located on the plasmids with low copy numbers and large sizes [ 5 ]. The gene clusters comprise catabolic genes encoding enzymes, transport genes encoding proteins for uptake of the aromatic compounds, and regulatory genes responsible for regulating the expression of both catabolic and transport genes. Gene clusters from Comamonas sp. strain E6 can degrade o -phthalate, terephthalate, and isophthalate via the protocatechuate 4,5-cleavage pathway [ 6 , 7 ]. Microbes generally only contain the catabolic genes for a single compound, and the degradation of multiple compound pollutants by a single strain is currently not well resolved [ 8 ]. Microbial remediation has been drawing increasing attention in the recent years, especially in the rising era of synthetic biology. Synthetic biology provides a strategy to construct engineered microorganisms that can monitor, aggregate and degrade environmental pollutants, with the aim to eliminate water pollution, remove garbage, and reduce air pollution [ 9 ]. So far, there are only limited reports on the construction of microorganism to degrade aromatic compounds using synthetic biology [ 10 ]. However, many efforts have been made to degrade 1,2,3-trichloropropane via synthetic biology [ 11 ], and synthetic biosensors for rapid detection of water contaminants [ 12 ], these successful cases bring us the hope to the degradation of aromatic compounds. Here, we review current progresses in aromatic compounds degradation using microbes and present a proposal for the rational design and construction of microbial strains to degrade aromatic compounds (see Fig. 2 ). Such a proposal usually follows the Chassis selection-Pathway design-Metabolism optimization-Tolerance engineering cycle, where iteration of each cycle leads to the improvement of the microbes. Fig. 1 Development of microbial degradation of aromatic pollutants. In the 1980s, it was an exciting era of microbe discovery; In the 1990s, naturally occurring microbes already have considerable ability to remove many environmental pollutants; In the 2000s, sanger sequencing leads to the discovery of microbial degradation gene clusters; In the future, the emerging of synthetic biology technologies brings a new artificial microorganism for pollutants degradation. Fig. 1 Fig. 2 Schematic overview of synthetic biology strategies applying to microbial degradation of aromatic pollutants (Naphthalene, Toluene, and Phenanthrene). The workflow includes chassis selection, pathway design, metabolism optimization, and tolerance engineering. (A) Not just model microbes but also some nonconventional microbes can serve as a chassis cell for the degradation of pollutants, such as Naphthalene, Toluene, and Phenanthrene. When selecting a host, consideration should be given to the characteristics of the pollutant, the chassis's genetic manipulation tools, genetic databases, and growth characteristics. (B) Biodegradation pathways containing gene clusters can be integrated into the chromosome or plasmid, and pathway design rely on genome data (gene clusters), mining tools (KEGG and MRE), and engineering tools (DNA assembly, CRISPR/Cas editing and Enzyme engineering). (C) Recently developed synthetic biology tools will accelerate the optimization of catabolism pathways for pollutants (AI-based design parts). (D) Most of the efforts in tolerance engineering have relied on improving the native gene function ( nah , tmo , xyl and phn ) and capabilities of a chassis cell. Fig. 2"
} | 1,426 |
40050621 | PMC11885466 | pmc | 8,326 | {
"abstract": "Current computer vision is data-intensive and faces bottlenecks in shrinking computational costs. Incorporating physics into a bioinspired visual system is promising to offer unprecedented energy efficiency, while the mismatch between physical dynamics and bioinspired algorithms makes the processing of real-world samples rather challenging. Here, we report a bioinspired in-materia analogue photoelectronic reservoir computing for dynamic vision processing. Such system is built based on InGaZnO photoelectronic synaptic transistors as the reservoir and a TaO X -based memristor array as the output layer. A receptive field inspired encoding scheme is implemented, simplifying the feature extraction process. High recognition accuracies (>90%) on four motion recognition datasets are achieved based on such system. Furthermore, falling behaviors recognition is also verified by our system with low energy consumption for processing per action (~45.78 μJ) which outperforms most previous reports on human action processing. Our results are of profound potential for advancing computer vision based on neuromorphic electronics.",
"introduction": "Introduction Computer vision is important for a wide spectrum of applications, including video retrieval, human-robot interaction, and entertainment, while it is always involved with accessing of a great amount of data in the event streams. Computer vision algorithms run on a von Neumann computing system is always energy-costly due to the efficiency loss with the increase of data scale 1 – 3 . On the contrary, our human brain has an unerring instinct for processing huge amounts of information in an energy-saving way. The energy efficiency of the biological visual system may rise from the filtering capability and the spike encoding scheme of the biological visual system. The former is achieved through the hierarchical structure and the receptive fields in each layer, which allows only a fraction of incoming visual events to be perceived, remembered or acted on. The latter enables event-driven computing, which consumes energy only when the population of cells are elicited by a sensory event. In this way, we can classify, locate, detect, and segment targets in video inputs that are transmitted by retina at roughly 10 million bits per second (10 Mb·s –1 ) with high accuracy and efficiency 4 – 6 . The neuromorphic visual system (NVS), which emerged recently, is aimed at extending the efficiency and capability of computer vision by replicating biological superiorities from the bottom up 7 – 10 . For example, optoelectronic graded neurons based on MoS 2 optoelectronic transistors were proposed for in-sensor motion perception, in which motion speeds can be effectively perceived with gate voltage modulation 11 . A van der Waals (vdW) heterostructure array was fabricated with the recognition capability to classify the motion modes (e.g., direction and velocity) of a target, via its gate-tunable electronic and optoelectronic properties 12 . An oxide-based retinomorphic photomemristor-reservoir computing system was proposed for motion modes recognition and prediction 13 . More recently, a fully memristive elementary motion detector has been proposed, achieving a high accuracy and low computational cost in lane-changing maneuvers prediction 14 . However, the identification of the motion and verification on real-world datasets have seldom been reported by previous reports, which greatly limit the applications in real scenarios. This might be due to the complexity of data preprocessing and the difficulty in tailoring the properties of an NVS device to match the requirements of appropriate algorithms. In this work, we demonstrate a bioinspired in-materia analog photoelectronic reservoir computing (Alpho-RC) system based on indium-gallium-zinc-oxide (IGZO) photoelectronic synaptic transistors as the reservoir and a TaO X -based memristor array as the output layer for human action processing. The human actions are captured and represented by 3D human skeleton sequences. Such sequences are then encoded into spike trains by several bioinspired Gaussian receptive field (GRF) neurons, and no feature extraction process on skeleton sequence is required. The spike train from each GRF neuron is applied to an IGZO-based photoelectronic synaptic transistor. Such transistor exhibits gate voltage tunable shading memory and nonlinear properties based on its photoelectronic coupling dynamics. In this case, it can effectively map the population encoded spike trains into high-dimension space and can provide abundant states as the virtual nodes for reservoir computing. Human action recognition tasks are stimulated with high accuracies (>90%) based on four standard datasets including UTD-MHAD, MSR Action3D, MSR Action Pairs, and Florence 3D. Furthermore, a one-transistor-one-memristor (1T1R) array used as the readout layer is integrated with the photoelectronic reservoir to construct the Alpho-RC system. Identification of falling behaviors is achieved based on such a system, and an energy consumption of only ~45.78 μJ/action is achieved, at least two orders of magnitude lower than digital processors. Our work can be regarded as a platform for next-generation neuromorphic computing which would prosper the development of high energy-efficient virtual reality, medical care, and visual surveillance.",
"discussion": "Discussion In conclusion, we developed a biologically plausible Alpho-RC architecture for implementing high energy efficient human action processing. Such architecture provides an advanced bioinspired in-materia reservoir computing framework for processing dynamic coordinates of skeleton joints without prior feature extraction process, which would facilitate the development of full-hardware implementation of reservoir computing. In such framework, the feature extraction could be incorporated in the population encoding by the GRF neurons and the nonlinear mapping by the photoelectronic synaptic transistors. In this case, the computation complexity is greatly reduced, in comparison with the previous methods that require complex feature extraction algorithms and multiple iterations in most feedforward networks. In many previous reports, not only prior feature extraction process but also feature enhancement process are required. Feature enhancement process is always implemented in complex computational modules like CAG (Coordinate Aware Grouping) and VAG (Virtual-part Aware Grouping) modules 53 . In this work, the frames of dynamic coordinates were encoded into only three spike trains, greatly simplifying the complexity of the RC system. The richness of reservoir states can be originated from the electron/proton electrostatic coupling at the IGZO/electrolyte interface. The introduction of light stimulation can further enrich the reservoir states based on the persistent photoconductivity effects in the IGZO channel. More importantly, the light can convey the spike encoding information in a fast, noise-robust, and high bandwidth manner which might promote the throughput and reliability of system 54 , 55 . For example, the usage of light can alleviate the congestion of electrical operations and broaden the data capacity range in computing. This could improve the efficiency for processing multiple floating-point operations, such as human action recognition 56 , 57 . Such biologically plausible architecture is very efficient for both human action recognition and prediction, indicating the remarkable feasibility and versatility. More importantly, a 1T1R array as readout layer is integrated with the photoelectronic reservoir to construct an analog photoelectronic reservoir computing system. High accuracy of over 96% on real-world falling action recognition demonstrates the Alpho-RC system’s application potential in the field of smart healthcare. Meanwhile, the estimated energy consumption for processing per action is two orders of magnitude lower than the reported digital methods. Comparison of PRC devices for human action processing also illustrates the advantages of our Alpho-RC system (Supplementary Table 3 ). Our devices showed the capability of recognizing human actions with the largest number and a simple training process. Additionally, our devices have been verified by homemade datasets, and the potential for fully analog reservoir computing has been verified by using memristor array as the readout layer. Future improvements may be devoted to the optimizing of photoelectronic synaptic transistor with respect to operating speed, and the development of large-scale system for high throughput processing. Previously, a SrTiO 3 -based memristor with multiple synaptic functions emulations has been proposed. A modified short-term plasticity neuron (m-STPN) was built based on such memristor. Such m-STPN has been applied on a bio-inspired deep neural networks (DNN), and an estimated gain in energy efficiency between 96× and 966× were achieved 58 . This concept can be borrowed for further optimizing our Alpho-RC system. Our Alpho-RC system provides a bionic computing paradigm with low energy consumption for the applications in virtual reality system, human-computer interaction system, smart medical care and many more."
} | 2,308 |
25540776 | PMC4261833 | pmc | 8,327 | {
"abstract": "Metabolic modeling provides the means to define metabolic processes at a systems level; however, genome-scale metabolic models often remain incomplete in their description of metabolic networks and may include reactions that are experimentally unverified. This shortcoming is exacerbated in reconstructed models of newly isolated algal species, as there may be little to no biochemical evidence available for the metabolism of such isolates. The phenotype microarray (PM) technology (Biolog, Hayward, CA, USA) provides an efficient, high-throughput method to functionally define cellular metabolic activities in response to a large array of entry metabolites. The platform can experimentally verify many of the unverified reactions in a network model as well as identify missing or new reactions in the reconstructed metabolic model. The PM technology has been used for metabolic phenotyping of non-photosynthetic bacteria and fungi, but it has not been reported for the phenotyping of microalgae. Here, we introduce the use of PM assays in a systematic way to the study of microalgae, applying it specifically to the green microalgal model species Chlamydomonas reinhardtii . The results obtained in this study validate a number of existing annotated metabolic reactions and identify a number of novel and unexpected metabolites. The obtained information was used to expand and refine the existing COBRA-based C. reinhardtii metabolic network model iRC1080. Over 254 reactions were added to the network, and the effects of these additions on flux distribution within the network are described. The novel reactions include the support of metabolism by a number of d -amino acids, l -dipeptides, and l -tripeptides as nitrogen sources, as well as support of cellular respiration by cysteamine- S -phosphate as a phosphorus source. The protocol developed here can be used as a foundation to functionally profile other microalgae such as known microalgae mutants and novel isolates.",
"conclusion": "Conclusion Phenotypic profiling has tremendous utility in modeling and understanding algal metabolism and is essential in elucidating genotypic differences in algae and the effects of environmental conditions on metabolism. The method presented here demonstrates the first reproducible study utilizing PM assays in profiling microalgae using C. reinhardtii as a model. We observed positive growth on 148 nutrients (one positive assay for C-source utilization, four positive assays for the S-source and P-source utilization, and 139 positive assays for N-source utilization). The wealth of phenotypic data can be used along with other references to compare organisms with known mutants or unknown isolates. This wealth of information will also shed light on new and novel metabolic pathways. The substrate utilization information and the newly identified metabolites were used for metabolic network expansion and refinement of the iRC1080 metabolic model. The study also provides a framework to bridge the missing links between genomics and metabolomics in microalgae. The described work provides an excellent method for the initial characterization of newly isolated or uncharacterized strains of algae. This combination of high-throughput phenotypic screening with metabolic modeling can allow for rapid refinement of existing metabolic network models as demonstrated and also provides biochemical evidence to support de novo reconstruction of new algal models.",
"introduction": "Introduction Optimization of algal metabolism toward improved bioproduct production while maintaining strain robustness remains a challenge that requires experimental strategies informed through systems-level analyses of metabolism. The use of metabolic network models can guide the development of optimization strategies that would be otherwise difficult through rational designs (Oberhardt et al., 2009 ; Schmidt et al., 2010 ; Koskimaki et al., 2013 ; Koussa et al., 2014 ). While an increasing number of algal species are being isolated and sequenced for biofuel or other applications, to date, there are only a handful of reconstructed algal networks available (Koussa et al., 2014 ). A major obstacle in the reconstruction of high-quality network models for algae remains hinged on the inability to obtain rapid and high-throughput metabolic phenotypic data to guide and validate reconstruction efforts. One potential high-throughput phenotypic analysis technology is the Biolog OmniLog ® phenotype microarray (PM) (Biolog, Hayward, CA, USA) (Bochner et al., 2001 ; Bochner, 2003 , 2009 ). By assaying cellular metabolism in response to thousands of metabolites, signaling molecules, and effector molecules (as well as osmolites), the Biolog PM assays have greatly boosted functional metabolic profiling by providing insight into function, metabolism, and environmental sensitivity (Bochner et al., 2001 ; Bochner, 2003 , 2009 ). Biolog PM assays rely on the measurement of metabolite utilization of cells in 96-well microplates. Each well contains different nutrients, metabolites, and pH and osmolarity solutes. Other bioactive molecules such as antibiotics and hormones may also be assayed. This utilization is assessed and measured in the form of cell respiration determined by the amount of color development produced by the NADH reduction of a tetrazolium-based redox dye (Bochner et al., 2001 ; Bochner, 2003 , 2009 ). Plates can be monitored automatically over time with the OmniLog platform. A common set of 20 96-well microplates are designed to measure carbon, nitrogen, sulfur, phosphorus utilization phenotypes, along with osmotic/ion, and pH effects. This high-throughput and standardized approach has the ability to provide a quick method for the phenotypic comparison of different strains and organisms in a convenient manner leading to insights into the metabolic state of the cell. While the PM technology has been used for metabolic phenotyping of various microbial species including bacteria and fungi, it has not been reported for the phenotyping of microalgae. Likewise, the technology has been successfully used for verification and expansion of a number of existing microbial metabolic network models (Bochner et al., 2001 ; Bochner, 2003 , 2009 ; Bartell et al., 2014 ), yet its use for improvement of microalgal models remains unreported. The goal of the present study is to establish a reliable method for characterizing metabolic phenotypes of microalgae that can be used to expand existing network models or guide the reconstruction of new algal metabolic models. We present the implementation of the PM platform for metabolic phenotyping of microalgae using Chlamydomonas reinhardtii as a model organism then expand a well-curated existing metabolic network model of C. reinhardtii accordingly.",
"discussion": "Discussion Algae are a group of diverse photosynthetic eukaryotes, which are polyphyletic in origin (Pröschold and Leliaert, 2007 ). Algal lineages include the viridiplantae, which the green algae (or Chlorophyta) belong to; stramenopiles that include brown, golden, and yellow algae and diatoms; rhodophyta or the red algae; and photosynthetic alveolates that include dinoflagelates (Barton et al., 2007 ). Given the evolutionary distances between these lineages, significant differences in genome size and coding potential, environmental niche, and metabolic properties can be expected. Members of green algae may be aquatic or soil organisms with mixotrophic or autotrophic modes of metabolism (Pröschold and Leliaert, 2007 ). In addition, microalgae may or may not require co-factors for their growth. Studies on microalgal growth requirements have indicated that more than half require cobalamin (vitamine B12), while 22% require thiamine and 5% need biotin (Croft et al., 2006 ). Interestingly, these requirements are not reflected in algal phylogeny (Helliwell et al., 2011 ). Genomic approaches powered by next-generation sequencing technologies help to improve the understanding of the encoded algal metabolic potential; however, the full characterization of algal metabolism requires phenotypic data. For instance, the metabolome of C. reinhardtii has been studied under a number of conditions, including sulfur deprivation (Matthew et al., 2009 ; Shu and Hu, 2012 ; Aksoy et al., 2013 ), nitrogen deprivation (Blaby et al., 2013 ; Courant et al., 2013 ), and response to irradiance (Mettler et al., 2014 ) to provide insight into regulatory and metabolic responses of the species to environmental perturbations. In addition, transcriptomics, proteomics, and metabolomics studies have guided non-targeted profiling approaches for the detection and quantification of metabolites. Those non-targeted profiling approaches have included nuclear magnetic resonance (NMR), liquid chromatography mass spectrometry (LC-MS/MS), and gas chromatography mass spectrometry (GC/MS) (Veyel et al., 2014 ; Wase et al., 2014 ). The ability to study functional responses and phenotypes has been classically limited to targeted serial studies that usually employ mutagenesis, genetic knockouts, genetic over-expression, and physiological studies (Bochner, 2003 ; Dent et al., 2005 ; Morgan et al., 2009 ; Tshikhudo et al., 2013 ; Greetham, 2014 ). The wealth of phenotypic information gained from the PM technology, as demonstrated in this study, can help provide more complete systems-level knowledge when combined with other omics data, and help develop and refine metabolic models. Genome-scale metabolic networks provide predicted genotype-phenotype relationships through metabolic flux optimization-based approaches. We previously reconstructed a genome-scale model for C. reinhardtii (iRC1080) (Chang et al., 2011 ) based on literature evidence (entailing ~250 sources), structurally verified genomic evidence, and predicted gene function and cellular localization information. This model has 1,706 metabolites with 2,191 reactions. Through the pipeline that we have described in this work, we were able to expand the network significantly to include 1,959 metabolites, 2,445 reactions, and 1,106 associated genes. A clear advantage that the PM provides is functional assays for entry metabolites to inform model refinement. Whereas mass spectrometry approaches give information on intermediate- and final-level metabolites, PM assays have the unique capability, due to the accounting for entry-level metabolites, to inform more complete models from the start of metabolic pathways. PM assays and mass spectrometry can therefore be considered as complementary approaches when characterizing organisms’ metabolic profiles, with each technology refining and filling in specific gaps in metabolic models. Yet, PM’s contribution to a metabolic model’s refinement is made through a rapid, high-throughput, and convenient manner with an entire set of metabolites assayed in 5–7 days. New metabolites We have identified a number of di and tripeptides, and d -amino acids that significantly expand the list of nitrogen utilization compounds in C. reinhardtii . While we found d -amino acids can support metabolism of C. reinhardtii , they may be involved in additional functions. A serine-type d -alanyl- d -alanine carboxypeptidase was found in C. reinhardtii ’s genome that could potentially be involved in d -alanine metabolism. Serine-type d -alalyl- d -alanine carboxypeptidases have been shown to play a variety of protective roles including protection against ionic and hyperosmotic stress (Príncipe et al., 2009 ). A d -alanine ligase was found in C. reinhardtii ’s genome that is potentially involved in d -alanine multimerization. Recent research using 15 N NMR spectroscopy found that d -alanine accumulated in plants during UV exposure and this finding is supported by previous research under various stress signals (Monselise et al., 2014 ). Therefore, the possibility that d -amino acids might have additional cellular functions in C. reinhardtii , aside from providing a source of nitrogen, can be a subject of future investigations. Chlamydomonas reinhardtii is known to be able to use a variety of amino acids as a sole nitrogen source as long as acetate is present (Munoz-Blanco et al., 1990 ). In C. reinhardtii , arginine is the only amino acid known to be imported with high affinity; the rest are believed to be deaminated extracellularly (Kirk and Kirk, 1978 ) or transported passively (Zuo et al., 2012 ). We note that a search in the literature for d -amino acid transports has not provided any information on the mode of transport for this class of amino acids in C. reinhardtii , nor is it known if the C. reinhardtii deaminase can deaminate d -amino acids. However, C. reinhardtii has been shown to exhibit amino acid racemerase activity (Takahashi et al., 2012 ), which could explain the ability to assimilate d -amino acids intracellularly. This also provides indirect evidence that these amino acids may be absorbed or transported into the cell for conversion to their L counterparts. A biological function for d -amino acids has not been clearly defined; however, d -alanine and d -aspartate were detected in algae using a reversed-phase HPLC; d -alanine was present in some marine diatoms while d -aspartate was found in all the selected freshwater green microalgae and marine diatoms (Yokoyama et al., 2003 ). In many microbes, dipeptides are imported into the intracellular compartment before they are eventually hydrolyzed. For instance, Francisella tularensis relies on an amino acid transporter of the major facilitator superfamily of secondary transporters for transporting amino acids intracellularly. Furthermore, dipeptides containing asparagine were effective at restoring cellular multiplication in the infection cycle of a F. tularensis mutant that lacked that essential amino acid transporter (Gesbert et al., 2014 ). In this study, a variety of dipeptides were found to promote heterotrophic respiration in C. reinhardtii . The latest version of C. reinhardtii ’s genome contains a gene annotated as a peptide hydrolase Cre02.g078650.t1.3. We note that the detected utilization of the dipeptides is not without sequence specificity as 159 out of 267 of the dipeptides and 9 out of 14 of the tripeptides did not result in positive assay results. From these newly identified metabolites, three phosphorus compounds were found: (1) cysteamine- S -phosphate (C 2 H 7 NO 3 PS), which is an organic phosphorothioate anion that is derived from deprotonation of thiophosphate OH groups and protonation of the amino group, (2) thiophosphate (or phosphorothioate), and (3) dithiophosphate, which is the product of the reaction of a base with phosphorus pentasulfide. The only new sulfur source that was identified, tetrathionate, is a sulfur oxoanion and is derived from the compound tetrathionic acid and is commonly found in soils. It is a key intermediate in the oxidation of various reduced inorganic sulfur compounds. Several species of bacteria including Salmonella enterica (Winter et al., 2010 ) and Acidithiobacillus ferrooxidans (Rohwerder et al., 2003 ; Holmes and Bonnefoy, 2007 ; Chen et al., 2012 ) are known to be able to assimilate tetrathionate. Strains of A. ferrooxidans overexpressing tetrathionate hydrolase (tetH) were found to grow better on both sulfur and tetrathionate. In the archeon Acidianus hospitalis , tetrathionate is secreted to form filaments from tetrathionate homomultimers (Krupovic et al., 2012 ). These remarkable filaments are believed to play a role in sulfur metabolism and adaptation to A. hospitalis ’s extreme environment. Prokaryotes have also been shown to use tetrathionate as an electron acceptor in cobalamin (coenzyme B12) synthesis (Roth et al., 1996 ). Sulfur is commonly assimilated as reduced sulfur for most living organisms, but bacteria are known to reduce tetrathionate, thiosulfate, sulfite, sulfur, and dimethyl sulfoxide in dissimilatory reactions as well (Barrett and Clark, 1987 ). Tetrathionate is often used as an electron sink for oxidative phosphorylation (Chen et al., 2012 ). Bacteria that are known to respire using tetrathionate are often found to have the capability of reducing thiosulfate as well, but thiosulfate is not found to be reduced among organisms that do not respire thiosulfate (Rohwerder et al., 2003 ). Considering that C. reinhardtii is a soil organism, the ability to assimilate this compound is likely to provide an important survival advantage in Chlamydomonas ’ natural life cycle. iBD1106 model vs. iRC1080 Different behaviors can be observed for iBD1106 than those for iRC1080 under different conditions. When the biomass production was set as the objective function, a differential change can be noticed as a result of growth conditions. The addition of the new nitrogen sources ( d -amino acids, dipeptides, and tripeptides) has a significant and differential effect on the shadow prices of metabolites under light and dark conditions for biomass production (Figures 5 A,B, respectively). Under light growth, the d -aspartate in iBD1106 showed significant effect on the behavior of the chloroplastic metabolites of the riboflavin pathway. In iBD1106, d -aspartate is converted into l -aspartate through racemase, and l -aspartate can be produced through hydrolysis of its dipeptides (Asp–Leu, Asp–phe, Pro–Asp, Asp–Ala, Asp–Gln, and Asp–Gly). Also the oxidation of d -asparagine produces d -aspartate as oxo-carboxylate (Eq. 1 ). The addition of l -aspartate increases its consumption in purine metabolism, which yields to more production of 2,5-Diamino-6-hydroxy-4-(5′-phosphoribos ylamino)-pyrimidine (25dhpp). The latter can be converted into 5-Amino-6-(5′-phosphoribosylamino)uracil (5apru) in the riboflavin metabolism resulting in an excess of 4-(1- d -Ribitylamino)-5-aminouracil (4r5au) and 5aprbu, with shadow prices of 0.168 and 0.158, respectively. Those results were not observed in iRC1080. Another example of model discrepancy under light growth is the effect of adding d -serine reactions in iBD1106. Addition of d -serine limited the availability of the metabolite 1-(9Z)-octadecenoyl,2-(11Z)-octadecenoyl-sn-glycerol-3-phosphate (pa1819Z18111Z) (shadow price −0.009 in iRC1080 and −0.65 in iBD1106). This metabolite is produced and consumed by the reactions of glycerolipid metabolism for the production of Palmitoyl-CoA (n-C16:0CoA) (pmtcoa). The addition of l -serine in iBD1106 results in more consumption of pmtcoa in the sphingolipid metabolism through the reaction serine C-palmitoyltransferase (SERPT) that produces 3-dehydrosphinganine. Under dark growth conditions, 4-aminobutanoate was in excess in iRC1080 and became limiting in iBD1106 with shadow price values of 0.18 and −0.05, respectively. The reason for this limiting availability is the addition of d -histidine and d -glutamate dipeptides hydrolysis reactions, e.g., Ala–His, and inversion into l -histidine and l -glutamate through a racemase. This addition increases the consumption of l -glutamate and l -histidine along with 4-aminobutanoate in glutamate and arginine and proline metabolisms, respectively. Moreover, the dark growth condition did not affect the behavior of 4-aminobutanoate significantly in iBD1106; however, in iRC1080 it was shifted from a limiting metabolite (−0.07) into an excess metabolite (0.18) (Table 4 ). The excessiveness of 4-aminobutanoate in iRC1080 under dark conditions might be related to the high consumption of NADPH under dark growth conditions. In proline metabolism, NADPH and 4-aminobutanoate are consumed more rapidly in dark than that in light conditions. As such, the addition of d -histidine and d -glutamate compensates the effect of growth under dark in the proline metabolism."
} | 4,961 |
28729530 | PMC5519710 | pmc | 8,329 | {
"abstract": "With its light weight, low cost and high efficiency even at low operation frequency, the triboelectric nanogenerator is considered a potential solution for self-powered sensor networks and large-scale renewable blue energy. As an energy harvester, its output power density and efficiency are dictated by the triboelectric charge density. Here we report a method for increasing the triboelectric charge density by coupling surface polarization from triboelectrification and hysteretic dielectric polarization from ferroelectric material in vacuum ( P ~ 10 −6 torr). Without the constraint of air breakdown, a triboelectric charge density of 1003 µC m −2 , which is close to the limit of dielectric breakdown, is attained. Our findings establish an optimization methodology for triboelectric nanogenerators and enable their more promising usage in applications ranging from powering electronic devices to harvesting large-scale blue energy.",
"introduction": "Introduction Intensive research efforts have been devoted to sustaining the huge energy consumption of modern society while minimizing the environmental cost. Harvesting energy from renewable natural resources such as the sun, the wind and biomass, has been demonstrated to be a sustainable alternative for energy crisis, and plays an increasingly important role with the fast depletion of fossil fuels 1 . With their light weight, low cost and high efficiency even at low operation frequency, the newly invented triboelectric nanogenerators (TENGs) have been proven to be a promising complementary solution for harvesting ambient mechanical energy that is ubiquitous but wasted in our everyday life 2 – 8 . The operation of TENGs is based on triboelectrification (or contact electrification) and electrostatic induction 9 , and the fundamental theory lies in Maxwell’s displacement current and change in surface polarization 10 . Previous work on TENG has demonstrated its potential of wide application ranging from powering small electronic devices for self-powered systems, to functioning as active sensors for medical, infrastructural, human–machine, environmental monitoring, and security 11 – 17 . Moreover, it can be effectively used for scavenging energy from low frequent ocean waves for the prospect of large-scale blue energy 18 – 21 . As an energy harvester, the commercialization and application of TENGs highly depend on their power density, which is quadratically related to the triboelectric charge density 22 . Therefore, large efforts have been devoted to increasing the amount of triboelectric charges by means of material improvement, structural optimization, surface modification, and so on 23 – 26 . Artificial injection of ions, for example, by using corona discharging, was considered a straight forward way to increase the charge density, resulting in a high output charge density of 240 µC m −2 , but long-term stability remains an issue 27 . Very recently, a high-output charge density of 250 µC m −2 was realized on a TENG through elastomeric materials and a fragmental contact structure 28 . Nevertheless, the achievable output charge density has still limited been by the phenomenon of air breakdown in all previous studies. For triboelectrification in the air atmosphere, it is also known that the effective contact area is significantly smaller than the overall surface area, and thus optimizing the contact area and structure of the TENG can effectively increase the overall triboelectric charge density. Herein, we first demonstrate that with the improved soft-contact and fragmental structure, the triboelectric charge density can be increased from 50 to 120 µC m −2 in air when compared to a conventional TENG with only hard contact. By applying high vacuum (~10 −6 torr), the charge density is further boosted to 660 µC m −2 without the limitation of air breakdown. With the coupling of surface polarization from triboelectrification and hysteretic dielectric polarization from a ferroelectric material, it can jump to 1003 µC m −2 , which elevates the maximum output power density of a conventional TENG from 0.75 to 50 W m −2 even at a low-motion frequency of about 2 Hz, the normal frequency of human walking and ocean waves. These findings open more possibilities for TENGs both as highly efficient mechanical energy harvesters for large-scale energy sources such ocean waves, and as self-powered modules integratable with devices beyond wearable electronics and sensors. Our findings may also give new insights into long-lasting debates over the mechanism of triboelectrification and its kinetics.",
"discussion": "Discussion In this work, the triboelectric charge density of TENG is first improved to 660 µC m −2 in vacuum where the limitation of air breakdown is eliminated, and further to 1003 µC m −2 via coupling of surface and dielectric polarization. High vacuum environments cannot only guarantee better performance of TENGs, but also spare TENGs from performance degradation caused by the natural accumulation of dust and air moisture. The progress here sets a performance record for TENGs and establishes an optimization methodology for them. This work also provides an insight into the restricting factors on performance of TENGs, making it necessary to re-estimate the upper limit of TENGs to be much higher than previously expected. The surface charge density of TENG is simultaneously limited by the triboelectrification charge density, air breakdown and dielectric breakdown, described formally as follows: 3 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\sigma _{{\\rm{TENG}}}}{\\rm{ = min}}\\left( {{\\sigma _{{{\\rm triboelectrification}}}},\\,{\\sigma _{{{\\rm air\\_breakdown}}}},\\,{\\sigma _{{{\\rm dielectric\\_breakdown}}}}} \\right)$$\\end{document} σ TENG = min σ t r i b o e l e c t r i f i c a t i o n , σ a i r _ b r e a k d o w n , σ d i e l e c t r i c _ b r e a k d o w n \n Without the concern of air breakdown (Supplementary Fig. 1 and Supplementary Note 1 ), dielectric breakdown may become the next bottleneck of TENG. Concerning the PTFE film, its σ \n dielectric_breakdown is calculated to be around 1115 µC m −2 (Supplementary Fig. 7 and Supplementary Note 6 ), which is not far from our result of 1003 µC m −2 . Future work will involve the optimization of dielectric material to further explore the limitations of triboelectrification. Without loss of generality, the paradigm to enhance the triboelectric charge density of TENG in this work can be applied to other technologies involving contact electrification. Furthermore, it will benefit the long-lasting debate over the underlying mechanism of triboelectrification and its kinetics, which calls for more experimental evidence to test existing hypotheses, such as electron transfer by tunneling effect, mass transport, and even a hybrid of both 35 . In high vacuum, interfering factors with triboelectrification such as dust and moisture can be reduced or even eliminated, and a much higher amount of transferred charge can be detected, both of which will favor the evidence seeking process. In practice, our study points to an effective approach for enhancing the output power of TENGs, which greatly improves the prospect of large-scale blue energy using nanogenerators networks 36 . Considering that existing TENGs proposed for harvesting water wave energy are already enclosed and sealed with waterproof containers 37 , 38 , it is rather straightforward and cost-effective to make them airtight and maintain an internal vacuum. The resulted boost in electrical output can reduce the required spread of blue-energy nets, and thus minimize the environmental impacts while meeting energy needs especially in extreme weather conditions."
} | 1,980 |
35333568 | PMC8956263 | pmc | 8,331 | {
"abstract": "Accurate transmission of biosignals without interference of surrounding noises is a key factor for the realization of human-machine interfaces (HMIs). We propose frequency-selective acoustic and haptic sensors for dual-mode HMIs based on triboelectric sensors with hierarchical macrodome/micropore/nanoparticle structure of ferroelectric composites. Our sensor shows a high sensitivity and linearity under a wide range of dynamic pressures and resonance frequency, which enables high acoustic frequency selectivity in a wide frequency range (145 to 9000 Hz), thus rendering noise-independent voice recognition possible. Our frequency-selective multichannel acoustic sensor array combined with an artificial neural network demonstrates over 95% accurate voice recognition for different frequency noises ranging from 100 to 8000 Hz. We demonstrate that our dual-mode sensor with linear response and frequency selectivity over a wide range of dynamic pressures facilitates the differentiation of surface texture and control of an avatar robot using both acoustic and mechanical inputs without interference from surrounding noise.",
"introduction": "INTRODUCTION Human-machine interfaces (HMIs) play a key role in the interaction between humans and machines by allowing the facile and intuitive control of machines for nondisabled and disabled people, who may or may not have knowledge regarding complicated software and hardware. Most of the existing wearable HMI devices use low-frequency (1 to 10 Hz) touch or hand motions such as tapping, bending, and flutter to deliver simple commands to machines ( 1 , 2 ). Recent developments in HMIs requiring high-frequency signal detection for robots, virtual reality (VR), augmented reality (AR), and Internet of Things (IoTs) demand precise and intuitive control of HMIs to deliver various senses and biosignals from humans. In addition to the low-frequency tactile mapping of objects ( 3 ), the perception of roughness and surface texture based on high-frequency vibration (80 to 300 Hz) detection will be required for a robotic skin to precisely perceive and manipulate objects ( 4 ). The mechano-acoustic sensing of electrophysiological signals such as cardiac operation and vocal vibration from skin requires a frequency bandwidth of 10 to 2000 Hz ( 5 ). A dynamic sensor with a high signal-to-noise ratio (SNR), a broad linear response range, and a wide frequency bandwidth is essential for the accurate perception of high-frequency dynamic signals using HMIs. Among the various candidates for dynamic sensors, triboelectric sensors (TESs) instantly generate high power in response to dynamic stimuli without additional power supply. Thus, TESs are considered self-powered voice recognition devices for biometric identification ( 6 ), hearing aids ( 7 ), and skin-attachable microphones ( 8 ). In addition, TESs recognize multiple physical touch ( 9 ), motion ( 10 , 11 ), fine texture ( 12 ), and displacement of objects ( 13 ), demonstrating great potential for dynamic HMI applications ( 14 ). Moreover, the electrical power generated due to mechanical deformation can be used to power luminescence devices ( 15 ) and actuators ( 16 ) that can offer feedback on dynamic information to the user through user-interactive devices. Dynamic HMIs require selective recognition of desired frequency information without interference from surrounding noises. Hence, developing dynamic sensors with controllable resonance frequency that perceive the dynamic response with a specific frequency while canceling undesirable noise frequency is necessary. In the human ear, thousands of basilar membranes in the cochlea provide gradient stiffness depending on width, thickness, and location, allowing the selective detection of an individual acoustic vibration from a broad range of frequencies, thus enabling us to recognize only the desired sound from the complex noise ( 17 ). Inspired by the frequency tunability of the inner structure of cochlea, various frequency-selective acoustic sensors have been created by modulating the form factors of channel materials including sensor size, membrane length, and location or by introducing diaphragms with different hole shapes, sizes, and thicknesses on the top layer of acoustic sensors ( 18 – 21 ). However, narrow tunable range of resonance frequencies and difficulties in miniaturization are the main limitations of these sensors. Considering the frequency selectivity using vibrational behavior depending on different modulus, control of the modulus or stiffness of active materials would result in the variation of resonance frequency. Recently, porous, crack, and biophotonic nanostructures of active materials have been studied for acoustic identification and improvement in acoustic sensitivity ( 22 – 25 ). Improved designing of materials and structural parameters would broaden the tunable range of frequency and increase the sensitivity, providing a new platform for the demonstration of advanced dynamic sensor that can cover the human voice frequency (100 to 4000 Hz) and dynamic tactile sensation (<100 Hz). In contrast to conventional HMIs based on single dynamic signals, dual-mode dynamic HMIs based on multiple dynamic signals from acoustic and tactile/physical stimuli with high sensitivity and frequency selectivity enable facile control of machines without surrounding interferences. However, dual-signal perceptive dynamic HMIs based on a single device have not been reported thus far. Therefore, we propose a frequency-selective acoustic and haptic smart glove for dual-mode HMIs. For the dual-mode HMIs, we suggest a hierarchical ferroelectric composite comprising surface macrodome (MD) and inner micropore (MP) structures decorated with nanoparticles (NPs) to develop a self-powered frequency-selective TES with high sensitivity and linear response ( Fig. 1A ). Upon pressure application, the gradual deformation of hierarchical ferroelectric composite results in the linear electrical output over a wide dynamic pressure range when external pressure is applied ( Fig. 1B ). In the hierarchical structure, the surface MD structure provides a larger variation in contact area, leading to increased sensitivity within a low-pressure range. The high deformability of inner MP structure prevents the rapid saturation of sensitivity. This extends the detectable pressure over a wide range. In addition, the NP in MP structure increases the dielectric property of the film and the localized stress under mechanical deformation, resulting in enhanced pressure sensitivity over a wide pressure range. The porosity and pore size of hierarchical structures, which affect the density and mechanical modulus of ferroelectric composites, are easily controllable, resulting in the active frequency selectivity ( Fig. 1C ). Furthermore, modulating form factors such as thickness and area of ferroelectric composites significantly increase the dynamic range of frequency selectivity (145 to 9000 Hz), covering the audible frequency range. As a proof of demonstration, we developed multichannel acoustic sensor arrays with different resonance frequencies capable of canceling unexpected noise and enabling the fabrication of high-accuracy voice recognition devices ( Fig. 1D ). The qualitative comparison of voice recognition can be performed using a machine learning technique based on the artificial neural network (ANN)–based training method that facilitates higher accuracy and selectivity in the proposed sensor than a commercial microphone. Last, we fabricated a texture-perceptive smart glove integrated with TESs on the tip of each finger that can detect and differentiate between various surface textures. Combined with a wireless platform, our smart glove can be used as a dual-signal perceptive HMI device with dual-mode operation using both mechanical and acoustic signals for the control of avatar robotic arms, thus providing a next-generation platform of HMI devices. Fig. 1. Hierarchically designed ferroelectric composites for dynamic interfacing applications. ( A ) Schematic of the TES using the hierarchical architecture of macrodome (MD), micropore (MP), and nanoparticles (NPs). ( B ) Schematic of the dependency of each structural component of the hierarchical ferroelectric composite on the pressure sensitivities of TESs. ( C ) Graph showing the frequency selectivity of TESs depending on the structural designs of hierarchical ferroelectric composites. ( D ) Applications of TESs in various dynamic interfacing devices including noise-independent voice recognition, texture perception, and dynamic motion detection and interfacing using robotic hands.",
"discussion": "DISCUSSION We developed highly linear and sensitive TESs based on the ferroelectric composites with a hierarchical architecture comprising MD, MP, and ceramic NP. The hierarchical geometry induced stress concentration at the interface of heterogeneous materials with different moduli and high deformability, providing linear gradient of stress-induced polarization. This facilitated the high sensitivity (36 nA/kPa) and linearity (1 V/kPa) of the proposed TESs over a wide dynamic pressure range (0 to 70 kPa). We demonstrated the capabilities of the proposed TESs by using them in dynamic interfacing devices used to recognize the acoustic wave, surface texture, and dynamic movements. The facile tunability of the resonance frequency for TESs using the structural designs of hierarchical TESs allows the realization of a high acoustic selectivity over a wide frequency range (145 to 9000 Hz). This leads to a high accuracy of over 95% in the case of noise-independent voice recognition devices. Furthermore, the high flexibility and linear responsivity of TESs assist in detecting and distinguishing between the fine textures of surfaces and versatile motion of robotic hands. Thus, hierarchical TESs exhibit great potential as the next-generation sensor for dynamic interfacing applications. The capabilities of hierarchical TESs based on ferroelectric composites provide a solid platform for improving the existing conventional sensors and application of TESs in humanoid robots, wearable devices, and biometrics."
} | 2,541 |
37315145 | PMC10266731 | pmc | 8,335 | {
"abstract": "Nature has evolved eight different pathways for the capture and conversion of CO 2 , including the Calvin-Benson-Bassham cycle of photosynthesis. Yet, these pathways underlie constrains and only represent a fraction of the thousands of theoretically possible solutions. To overcome the limitations of natural evolution, we introduce the HydrOxyPropionyl-CoA/Acrylyl-CoA (HOPAC) cycle, a new-to-nature CO 2 -fixation pathway that was designed through metabolic retrosynthesis around the reductive carboxylation of acrylyl-CoA, a highly efficient principle of CO 2 fixation. We realized the HOPAC cycle in a step-wise fashion and used rational engineering approaches and machine learning–guided workflows to further optimize its output by more than one order of magnitude. Version 4.0 of the HOPAC cycle encompasses 11 enzymes from six different organisms, converting ~3.0 mM CO 2 into glycolate within 2 hours. Our work moves the hypothetical HOPAC cycle from a theoretical design into an established in vitro system that forms the basis for different potential applications.",
"introduction": "INTRODUCTION Globally, more than 350 Gt of carbon dioxide (CO 2 ) is fixed by autotrophic organisms each year ( 1 ). Of this, more than 95% is funneled through a single enzyme, ribulose-1,5-bisphosphate carboxylase-oxygenase (Rubisco), the carboxylating enzyme in the Calvin-Benson-Basham (CBB) cycle ( 2 ). In addition to Rubisco and the CBB cycle, seven other autotrophic cycles have been found, which revolve around eight alternative carbon-fixing enzymes ( 3 ). Yet, these eight pathways represent only a small fraction of the possible thousands of autotrophic CO 2 -fixation pathways that have been available for nature to explore during evolution ( 4 ). Notably, there are several CO 2 -fixing enzymes, which are not known to operate in autotrophic CO 2 fixation, but instead in other metabolic processes, such as the assimilation of organic compounds, anaplerosis, or biosynthesis ( 5 ). One prime example are enoyl–coenzyme A (CoA) carboxylase/reductases (Ecr, EC 1.3.1.85). These enzymes are the most efficient CO 2 -fixing enzymes described to date, outcompeting Rubisco by more than one order of magnitude ( 1 , 6 ) and, unlike Rubisco, do not show any side reaction with oxygen. However, despite their highly favorable kinetic and biochemical properties, Ecrs have thus far only been described in the context of heterotrophic acetate assimilation via the ethylmalonyl-CoA pathway and polyketide biosynthesis but were apparently not recruited for autotrophic CO 2 fixation. The ostensible inability of evolution to systematically assemble the most promising enzymes into new pathways and cycles has inspired several approaches to overcome this roadblock by creating alternative, “synthetic” CO 2 fixation pathways from the bottom-up. This effort has mainly been a theoretical exercise so far ( 1 , 6 – 10 ) but has demonstrated that multiple pathways can indeed be designed that outcompete natural CO 2 fixation in respect to kinetic and thermodynamic considerations. In efforts to leverage the potential of Ecrs for synthetic CO 2 fixation, we (and others) have drafted several new-to-nature CO 2 -fixation pathways that are based on this newly found CO 2 -fixation principle ( 1 , 11 ). Notably, these designs are all superior to the natural CBB cycle. We recently realized one of these designs, the CETCH cycle (for Crotonyl-CoA/EThylmalonyl-CoA/Hydroxybutyryl-CoA cycle), in vitro and optimized it in several steps ( 1 , 12 ). The CETCH version 5.4 cycle included 17 enzymes from nine different species including three re-engineered enzymes, and two reductive carboxylation reactions working in tandem to convert CO 2 into glyoxylate/glycolate in vitro at fixation rates that are comparable to natural CO 2 fixation pathways. Further optimization of CETCH using a machine learning–guided approach increased yield by almost an order of magnitude, achieving a system that can fix 5.9 mM CO 2 within ~3 hours (from 100 μM starting substrate) in vitro. Yet, the CETCH cycle is only one of the potential designs that we had originally envisioned ( 1 ). In ongoing efforts in our laboratory to realize other solutions, we present here the successful reconstruction of the HydrOxyPropionyl-CoA/Acrylyl-CoA (HOPAC) cycle ( Fig. 1A ). The overall topology of the HOPAC cycle is similar to the naturally existing 3-hydroxypropionyl-CoA (3HP) cycle ( 13 , 14 ). However, the HOPAC cycle was designed to be energetically more efficient than the 3HP cycle. To realize the HOPAC cycle, we dissected the cycle into two partial pathways, a reductive and an oxidative part, which allowed for the convenient screening and prototyping of alternative variants of different enzyme chemistries ( Fig. 1, B and C , and Table 1 for different combinations). Successful versions of the two parts were combined into a complete HOPAC cycle, which was further optimized through rational and machine learning–guided efforts to reach activities that compare to other in vitro CO 2 -fixation pathways ( 1 , 11 ). Notably, the final, optimized version of the HOPAC cycle differs in several key reactions from the 3HP cycle, which makes it a truly new-to-nature CO 2 fixation pathway with potential for different in vitro and in vivo applications. Fig. 1. The HOPAC cycle and its alternative designs explored in this study. ( A ) Overall scheme of the HOPAC cycle. The cycle converts two molecules of inorganic carbon (HCO 3 − and/or CO 2 ) into one molecule of glyoxylate. The HOPAC cycle can be divided into a reductive and an oxidative part, for each of which two different variants were developed in this study. Color coding indicates key metabolites that were analyzed in this study. Enzyme abbreviations are expanded in table S6. Option 1: Scs, Smt, or Sch; option 2: Smt or Mcs; option 3: GabT or βapt; option 4: βcl or βct. ( B ) Variants of the reductive part of the HOPAC cycle tested in this study with the respective cofactors ( C ) Variants of the oxidative part of the HOPAC cycle variants tested in this study with the respective cofactors. ATP and NADPH balances for all possible combinations of the reductive and oxidative parts of the HOPAC cycle are given in Table 1 . Table 1. Cofactor requirements for all possible combinations of the different reductive and oxidative part variants. HOPAC cycle combination Cofactor Reductive part variant + Oxidative part variant ATP NADPH β-Alanine route + CoA route 2 2* β-Alanine route + Free acid route (Smt or Scs/Mcs variant) 2 2* 3HP route (Ccr variant) + CoA route 2 3 3HP route (Ccr variant) + Free acid route (Smt or Scs/Mcs variant) 2 3 β-Alanine route + Free acid route (Sch variant) 3 2* 3HP route (Ccr variant) + Free acid route (Sch variant) 3 3 3HP route (Pcs variant) + CoA route 3† 3 3HP route (Pcs variant) + Free acid route (Smt or Scs/Mcs variant)‡ 3† 3 3HP route (Pcs variant) + Free acid route (Sch variant) 4† 3 *These values relate to the number of NADPH consumed by the core chemistry of the cycle, an additional NADPH is required if the amino acid is regenerated. †These values relate to the number of ATP consumed by the core chemistry of the cycle, an additional ATP is required if the AMP is regenerated to ATP. ‡This version of the HOPAC cycle, using the Pcs variant of the 3HP route and the Smt variant of the free acid route, is the option used in the first cycle of the natural 3-hydroxypropionate bicycle.",
"discussion": "DISCUSSION In this work, we prototyped and optimized a new-to-nature CO 2 -fixation pathway, the HOPAC cycle. The core cycle consists of 11 enzymes from six different organisms and is similar to the naturally existing 3HP bicycle. However, for the conversion of acrylyl-CoA into (2 S )-methylmalonyl-CoA, the HOPAC cycle uses a reductive carboxylation instead of a combination of acrylyl-CoA reduction and ATP-dependent carboxylation. This makes the HOPAC Ccr 33% more ATP-efficient than the 3HP, which underlines the potential of synthetic biology to realize more efficient metabolic pathways than natural evolution. Beyond this difference in acrylyl-CoA carboxylation, the HOPAC Ccr cycle also differs in respect to other enzyme chemistries from the 3HP. The central oxidation of succinyl-CoA to malyl-CoA proceeds through a new-to-nature metabolite, fumaryl-CoA, and involves two new-to-nature reactions: one reaction producing fumaryl-CoA (through oxidation of succinyl-CoA) and another reaction that consumes fumaryl-CoA (through hydration into malyl-CoA). We succeeded in establishing these reactions through screening multiple enzymes candidates for promiscuous activities with these metabolites that we could potentially exploit for construction of the cycle. Thus, the final implementation of our design differs not only from 3HP but also from other natural reaction sequences and pathways, qualifying the HOPAC Ccr as a bona fide “new-to-nature” pathway. During our design, we explored other variants to the HOPAC, such as the free-acid variant in the oxidative part or the β-alanine variant in the reductive part. Although we did not pursue these options further in the framework of this study, these pathway variants are still viable and could be used in the context of other in vitro setups or efforts to establish the HOPAC in vivo. One of the most promising alternatives is the Pcs/Pcc variant of the HOPAC cycle, in which we used Pcs in combination with Pcc for the conversion of 3HP into (2 S )-methylmalonyl-CoA—at the cost of one extra ATP. The acrylyl-CoA reductase domain of Pcs has been recently engineered toward a reductive carboxylase, reaching 97% carboxylation activity at saturating CO 2 concentrations, albeit at lowered overall catalytic activity of the enzyme. Using a “carboxylating” Pcs (with improved overall rates) could omit the need for Pcc in the future. This would lower the extra ATP need for carboxylation and make the Pcs/Pcc-based variant energetically equivalent to HOPAC Ccr . Having established a first version of the HOPAC cycle, we could further optimize the system through iterative rounds of rational and machine learning–guided efforts. Addition of auxiliary enzymes, such as MeaB, and adjusting the concentration of individual components was crucial to improve the system by 10-fold compared to version 1.0, which strongly emphasizes the potential and the need of design-build-test-cycles in complex in vitro biology. In its current form, HOPAC Ccr fixes CO 2 at a rate of 2.4 nmol min −1 mg −1 , which matches well the efficiencies of other in vitro CO 2 -fixation pathways and provides the basis for the further development of the HOPAC cycle toward different in vitro and in vivo applications."
} | 2,681 |
19543724 | PMC2729981 | pmc | 8,336 | {
"abstract": "The performance of a full-scale (500 m 3 ) sulfidogenic synthesis gas fed gas-lift reactor treating metal- and sulfate-rich wastewater was investigated over a period of 128 weeks. After startup, the reactor had a high methanogenic activity of 46 Nm 3 ·h −1 . Lowering the carbon dioxide feed rate during the first 6 weeks gradually lowered the methane production rate. Between weeks 8 and 93, less than 1% of the hydrogen supplied was used for methanogenesis. Denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified 16S rRNA gene fragments showed that the archaeal community decreased in diversity but did not disappear completely. After the carbon dioxide feed rate increased in week 88, the methane production rate also increased, confirming that methane production was carbon dioxide limited. Even though lowering the carbon dioxide feed appeared to affect part of the sulfate-reducing community, it did not prevent achieving the desired rates of sulfate reduction. The average sulfate conversion rate was 181 kg∙h −1 for the first 92 weeks. After 92 weeks, the sulfate input rate was increased and from week 94 to 128, the average weekly sulfate conversion rate was 295 kg·h −1 (SD ± 87). Even higher sulfate conversion rates of up to 400 kg·h −1 could be sustained for weeks 120–128. The long-term performance and stability together with the ability to control methanogenesis demonstrates that synthesis gas fed reactor can be used successfully at full scale to treat metal and sulfate-rich wastewater.",
"introduction": "Introduction Sulfate- and metal-rich wastewaters that are low in organic carbon are produced as a result of several industrial activities, such as metal smelting, flue gas scrubbing, and mining. Sulfate-reducing bioreactors have been shown to be suitable systems to remediate these types of wastewater (Boonstra et al. 1999 ). These bioreactors utilize the sulfidogenic activity of sulfate-reducing bacteria (SRB) to simultaneously remove sulfate and metals in the form of metal sulfides (Muyzer and Stams 2008 ). Paques B.V. (Balk, The Netherlands) has developed a sulfidogenic SULFATEQ® system, which is based on a gas-lift reactor fed with hydrogen gas as the electron donor for sulfate reduction. For large-scale applications, synthesis gas is an attractive source of hydrogen (van Houten and Lettinga 1994 ). Synthesis gas composition varies depending on the organic carbon source used for its production, but it contains mainly H 2 , CO 2 , and CO with minor levels of other components, such as methane and nitrogen. After tests conducted at pilot-scale, the first full-scale synthesis gas fed reactor was constructed to treat wastewater from the Nyrstar Budel Zink zinc smelting plant (Budel-Dorplein, The Netherlands). The full-scale sulfidogenic bioreactor was chosen over the original lime stone neutralization process because the zinc sulfide produced in the bioreactor is suitable to be re-used as a zinc ore in the zinc smelting process. As a consequence, no solid waste is produced. Furthermore, since the solubilities of most metal sulfides are much lower than those of their respective hydroxides, considerably lower effluent metal concentrations can be achieved. During startup the full-scale reactor was fed with purified synthesis gas containing mainly hydrogen (van Houten et al. 2006 ). While startup of the reactor was successful and the desired rate of sulfate reduction was achieved, hydrogenotrophic methanogenesis was not suppressed despite the predominance of SRB (van Houten et al. 2006 ). Methanogenesis is an unwanted process, as it increases operational costs. Persistence of methanogenesis was unexpected, as prior research had shown suppression of methanogenesis by sulfate reduction in lab-scale systems (van Houten et al. 1994 ; Weijma et al. 2002 ). The suppression of methanogenesis by sulfate reduction in anaerobic bioreactors is often attributed to the fact that SRB can reach much lower hydrogen threshold concentrations compared to methanogens. Based on the hydrogen thresholds of the methanogens and SRB detected within the sludge (van Houten et al. 2006 ), methanogenesis should have been suppressed rapidly by sulfate reduction. However, because of the relatively short sludge retention time (SRT) of 4–7 days employed for the full-scale reactor, hydrogen threshold concentrations cannot be reached. During operation at full scale, process conditions fluctuate and mass transfer of hydrogen to the liquid phase exceeds the requirement for sulfate reduction for the majority of time. Consequently, a continuous state of hydrogen limitation is not reached, and as a result, methanogens are not readily outcompeted by SRB. A potential strategy to suppress methanogenesis would be to limit the carbon dioxide feed to the reactor. As hydrogenotrophic methanogens require carbon dioxide as their terminal electron acceptor, limiting the carbon dioxide feed should lower their activity. SRB also require carbon dioxide for growth but only as source of carbon. Heterotrophic SRB were found to be the predominant microorganisms in the reactor (van Houten et al. 2006 ). These SRB utilize carbon dioxide as a carbon source but only to a minor extent, as they obtain most of their carbon (Sorokin 1966a , b ; Badziong and Thauer 1979 ) from the organic carbon source provided. Information on the long-term performance and the microbial community composition and dynamics is still very limited for this type of bioreactor, especially at full scale. Additional information, especially on community dynamics in relation to operational changes, may provide information that can be used to understand how to consequently control the performance of synthesis gas fed bioreactors. Therefore, we have studied the performance based on sulfate reduction rates and monitored methane production over a prolonged period of 128 weeks. During this period, the carbon dioxide feed to the reactor was lowered. For the same period the composition and dynamics of the dominant bacterial and archaeal species was studied. Temporal community changes were assessed using denaturing gradient gel electrophoresis (DGGE), while the relative abundance and identity of dominant bacterial species in the sludge were estimated using restriction fragment length polymorphism (RFLP) and sequence analysis of 16S rRNA gene clone libraries.",
"discussion": "Discussion The community analysis results clearly show that even a mainly hydrogen fed sulfate-reducing bioreactor can sustain a diverse bacterial community not restricted to just hydrogenotrophic SRB. The notable presence and potential role of Bacteriodetes , Chloroflexi , and even Thermotogales in a mesophilic reactor community, which was fed with limited amounts of acetate as the only organic carbon and hydrogen as the main electron donor is intriguing. Even though these phyla are underrepresented when it comes to cultured and described species, these organisms appear to be generally involved in degradation of carbohydrates and proteins. For instance, the closest physiologically characterized relative to Bacteriodetes clones, Proteiniphilum acetatigenes strain TB107 UNI-1T, has been shown to grow on complex substrates such as sugars and amino acids but not on hydrogen. Okabe et al. ( 2005 ) presented data that strongly suggest that members of the Chloroflexi preferentially utilize microbial products derived from biomass decay. As the ingoing wastewater did not contain complex organic compounds, a plausible explanation for their presence would be that these organisms are involved in the degradation of microbial cell material in sludge. Under normal conditions, due to the relatively high growth rate in the reactors (estimated doubling time of 0.14–0.25 day −1 based on SRT), there will be a continuous production of dead cells allowing these putative scavengers to persist in the system. Koizumi and co-workers have observed coexistence of SRB and Chloroflexi in the top sediment of a saline meromictic lake (Koizumi et al. 2004 ). They proposed a syntrophic relationship between SRB and Chloroflexi , where members of subphylum I might be initial degraders of macromolecules, providing fatty acids and hydrogen as electron donors for SRB. If this interaction also takes place in the Budel reactor, these intermediates might enable a more diverse SRB community to be sustained that is not restricted to just hydrogenotrophic sulfate reduction. The fluctuations in operating conditions that continuously occur during operation at full scale are probably the main reason why multiple SRB are sustained in the reactor. The lack of a continuous prolonged state of one limiting substrate will allow multiple hydrogenotrophic SRB to coexist. This appears to have been the case during the first 101 weeks of operation. Two bands, marked B and C in Fig. 4 , were visible throughout this period and became relatively abundant during the stable period with low methane production (weeks 18 – 101). The DGGE migration patterns of the abundant clones BUD03 and BUD11 coincided with bands B and C. These two taxa, belonging to the sulfate-reducing genera of Desulfomicrobium and Desulfovibrio (Fig. 5 ), appear to have contributed to the sulfidogenic performance of the reactor. Band B was detectable over the longest period of time (weeks 1 – 121). By week 128, a dramatic shift has taken place to one predominant band A. Band A coincides with the DGGE migration pattern of clone BUD15 and possibly also represents BUD16. Clones BUD16 and BUD15 were also found to be predominant as indicated by their RFLP abundance, 27.7% and 21.3% respectively. Even though the hydrogen concentration of the recycle gas was low during this period, it is highly unlikely that this shift was caused by hydrogen limitation considering the high methanogenic activity in the same period. If a continuous state of hydrogen limitation had occurred, methanogenesis should have suppressed rapidly by sulfate reduction (Schönheit et al. 1982 ). A much more plausible explanation would be that growth of BUD15 and BUD16 had been carbon limited, more specifically carbon dioxide limited. Band A was visible throughout the whole monitored period. It was also more predominant in weeks 10–26 but gradually became less so following the gradually lowered carbon dioxide feed rate. The increased carbon dioxide feed rate from week 96 would have alleviated this limitation allowing it to become more abundant. This would imply that the microorganisms represented by the D. giganteus like clones BUD16 and BUD15 possess a high growth rate ( µ \n max ) on and possibly a higher affinity ( K \n m ) for hydrogen, giving them a selective advantage over other SRB when carbon dioxide is not growth limiting. This observation of potential carbon dioxide limitation affecting part of the SRB community may also explain in part why the two clones belonging to the genus Desulfomicrobium found to be dominant during startup (van Houten et al. 2006 ) became less dominant during the first 10 weeks of gradually lowering the carbon dioxide feed to the reactor. In addition, the calamity in week 4 , where too much metal-rich waste water was fed to the reactor-causing metal toxicity would also have impacted their relative abundance. Desulfovibrio paquesii strain SB1 (van Houten et al. 2009 ) was isolated from the Budel full-scale reactor from a sample taken in week 7. It was found to be one of the dominant SRB at that time based on a MPN dilution counts. The DGGE migration pattern of strain SB1 consisting of the two bands indicating D (Fig. 4 ) is visible up to week 18. It was found that strain SB1 has a relatively high maximum growth rate of 1.9–2.8 day −1 , which may explain its abundance at the time of isolation (van Houten 2006 ) shortly after the calamity. Its gradual decline in abundance thereafter based on DGGE analysis may also have been due to carbon dioxide limitation with the SRB community gradually shifting toward populations better adapted to the prevailing low carbon dioxide concentrations. The observed redundancy within the SRB present in the reactor could be an important factor in maintaining a stable prolonged reactor performance under fluctuating process conditions. Based on prevailing conditions, only a limited number of SRB will become abundant and contribute to the reactors performance. In addition, Kaksonen et al. ( 2004 ) demonstrated that two lab-scale sulfate-reducing fluidized bed reactor communities, enriched and maintained on a single electron donor treating acidic metal-containing wastewater, were composed of a diverse mixture of bacteria including several SRB. This diversity and flexibility of microbial communities were also suggested to enhance the robustness of the reactor under varying operational conditions. Lowering the carbon dioxide feed resulted in a clear decrease in methanogenic activity to the extent that less than 1% of the hydrogen supplied was used for methanogenesis between week 8 and 93. Lowering the carbon dioxide feed also decreased the archaeal diversity observed by DGGE. The lowered methanogenic activity suggests that carbon dioxide became the growth-limiting substrate for methanogenic archaea. In addition, the competition for a single growth-limiting substrate could be expected to decrease diversity. The increase in the rate of methanogenesis after the carbon dioxide feed had increased from week 88 onwards confirms that the archaeal community was carbon dioxide limited between week 13 and 95. Hence, limiting the carbon dioxide feed proved to be an effective tool to control the methanogenic activity, but it did not result in a complete washout of methanogens from the reactor. Consequently the reactor remained vulnerable to an increased methanogenic activity after the carbon dioxide feed rate increased. These results demonstrate clearly that it is crucial to tightly control the carbon dioxide feed during operation of synthesis gas fed reactors to minimize methanogenesis. In summary, this study demonstrates that full-scale synthesis gas fed reactors fed with hydrogen as the main electron donor can sustain a relatively diverse microbial community that is not restricted to sulfate-reducing bacteria. During operation at full scale, fluctuations in process conditions enabled multiple populations of hydrogenotropic SRB to co-exist. Lowering the carbon dioxide feed proved to be an effective tool to control the methanogenic activity. It also resulted in lowering the archaeal diversity, but it did not result in a complete washout of methanogens from the reactor. In addition, it appeared to affect part of the sulfate-reducing community, but it did not impact the diversity in the same way or prevented achieving the desired rates of sulfate reduction. This, together with the ability to minimize methanogenesis, demonstrates that synthesis gas fed sulfidogenic bioreactors can be used successfully at full scale to treat metal and sulfate-rich wastewater."
} | 3,762 |
26198539 | PMC4510527 | pmc | 8,337 | {
"abstract": "Although protected areas can lead to recovery of overharvested species, it is much less clear whether the return of certain predator species or a diversity of predator species can lead to re-establishment of important top-down forces that regulate whole ecosystems. Here we report that the algal recovery in a Mediterranean Marine Protected Area did not derive from the increase in the traditional strong predators, but rather from the establishment of a previously unknown interaction between the thermophilic fish Thalassoma pavo and the seastar Marthasterias glacialis . The interaction resulted in elevated predation rates on sea urchins responsible for algal overgrazing. Manipulative experiments and field observations revealed that the proximity of the seastars triggered an escape response in sea urchins, extending their tube feet. Fishes exploited this behavior by feeding on the exposed tube feet, thus impairing urchin movement, and making them vulnerable to predation by the seastars. These findings suggest that predator diversity generated by MPA establishment can activate positive interactions among predators, with subsequent restoration of the ecosystem structure and function through cascading consumer impacts.",
"discussion": "Discussion The hypothesis that a synergistic facilitative interaction between the seastar M. glacialis and the fish T. pavo makes sea urchins vulnerable to seastar predation was corroborated by the results of this study. Sea urchin resulted the preferred food item of M. glacialis while sea urchin tube feet were frequently found in the T. pavo gut contents. The experimental removal of sea urchin tube feet was quantitatively similar to the action of wrasses during the escape behaviour triggered by the seastar. Our results suggested that T. pavo facilitated M. glacialis in the barren habitat by removing tube feet and reducing escape speeds of sea urchins allowing higher rates of capture by M. glacialis . At the same time, T. pavo seems to have benefited from the escape response of sea urchins triggered by the seastar, since extended tube feet became accessible to the fish. However, in our study we lacked the data to demonstrate whether the absence of seastars affects the ability of T. pavo to capture sea urchin tube feet. The establishment of a facilitative interaction between T. pavo and M. glacialis is particularly interesting since these species usually occupy non-overlapping habitats, and neither of them generally feed on adult of sea urchins 42 43 . The simultaneous occurrence of sea urchins, M. glacialis and T. pavo in the shallow rocky shore of Ustica, where high densities of the seastar occurred 14 , likely created the condition for the described interaction to happen. Species of the Thalassoma genus display different feeding strategies. This fish can use stones as anvils to crash bivalve that are too large to swallow; it acts as facultative cleaner-fish; it follows “nuclear fish ” eating particles stirred up from the bottom 44 ; and it follows divers to obtain food 53 . Importantly, wrasses are able to remember what, where and when food can be found 54 . Data from gut contents of T. pavo at Ustica clearly showed that sea urchin tube feet are an important component of its diet. Hence, T. pavo may have “learned” to follow the attacks of M. glacialis on sea urchins in order to easily feed on the sea urchin tube feet. This is of particular interest since T. pavo is a thermophilic species and its Mediterranean range has moved progressively northward, following increases in water temperature 55 . It is possible that interactions between T. pavo and seastars or other weak sea urchin predators will arise in other regions, with important consequences on sea urchin populations and benthic community structure. The correlation between the decrease in sea urchin abundance, the increase in seastar population and the recovery in algae forest ( Fig. 1 ) indicates that the high predation on sea urchins by M. glacialis , facilitated by T. pavo , is a key process for the recovery of the algae forest at Ustica island 14 40 . The establishment of MPAs is a widespread tool for marine managers to fight the detrimental effect of multiple stressors, such as loss of species and climate change, on coastal systems worldwide 11 . Several studies indicate a positive effect of protection on targeted species within MPAs and, in some cases, the recovery of former ecosystems (e.g., kelp forest) through trophic cascades 33 35 36 . In other cases, however, the response of communities to protection seems to be idiosyncratic 10 . The variability in the effects of protection could depend on the time scale considered, as both direct and indirect effects of protection are slow, unstable and asynchronous 16 . Moreover, even if the goal of MPAs is the protection of the whole assemblages of organisms and their interactions, monitoring programs often focus only on targeted and conspicuous species, or limited assemblages (e.g., fish assemblages). The potential role of other species and the unique assemblages that may result due, in part, to their establishment, on the dynamics of protected ecosystems is thus largely overlooked. Predator diversity may enhance the strength of trophic cascades by providing alternative consumers when environmental conditions change (i.e., Insurance Hypothesis) 56 57 . However, many systems are characterized by key predators and weak alternative consumers, whose effect on prey populations is mild 58 59 . In other cases, multi-predator assemblages may alter the strength of trophic control on preys by the so-called emergent multi-predator effect. Different predators can establish synergistic interactions, either positive (i.e., facilitation) or negative (i.e., interference), affecting predation rate on common prey 19 59 . Here we presented a case where the paucity of traditional key predators is compensated by an unexpected interaction between two weak predators (a seastar and a fish), leading to the ecosystem restoration ( Fig. 4 ). More broadly, our study represents an example where predator assemblage diversity strengthened the trophic control of keystone grazers, allowing the recovery of the algal forest habitat (i.e. the three-dimensional component). Our results confirm the importance of long term studies for proper assessment of the direct and indirect effects of protection on ecosystems. This study is also a reminder of the importance of field observation for revealing arising interactions among novel species assortments in a changing community."
} | 1,652 |
35044829 | PMC8769544 | pmc | 8,340 | {
"abstract": "Chemocatalytic lignin valorization strategies are critical for a sustainable bioeconomy, as lignin, especially technical lignin, is one of the most available and underutilized aromatic feedstocks. Here, we provide the first report of an intensified reactive distillation–reductive catalytic deconstruction (RD-RCD) process to concurrently deconstruct technical lignins from diverse sources and purify the aromatic products at ambient pressure. We demonstrate the utility of RD-RCD bio-oils in high-performance additive manufacturing via stereolithography 3D printing and highlight its economic advantages over a conventional reductive catalytic fractionation/RCD process. As an example, our RD-RCD reduces the cost of producing a biobased pressure-sensitive adhesive from softwood Kraft lignin by up to 60% in comparison to the high-pressure RCD approach. Last, a facile screening method was developed to predict deconstruction yields using easy-to-obtain thermal decomposition data. This work presents an integrated lignin valorization approach for upgrading existing lignin streams toward the realization of economically viable biorefineries.",
"introduction": "INTRODUCTION Lignocellulosic biomass (LCB) is the most abundant form of biomass on Earth, and its valorization has been studied extensively for producing sustainable fuels, chemicals, and materials ( 1 – 8 ). The composition of LCB is approximately 40 to 60% cellulose, 10 to 40% hemicellulose, and 15 to 30% lignin ( 9 ), and the specific percentages depend on the type of biomass ( 1 ). Cellulose and hemicellulose are polysaccharides that are readily converted into fuels and chemicals, such as hydroxymethylfurfural or sugar alcohols such as xylitol, through straightforward acid catalysis ( 5 , 10 ), while lignin is a complex aromatic polymer network whose valorization is challenging because of its inherent recalcitrance ( 1 , 5 ). Presently, lignin is separated from biomass at a rate of approximately 70 to 100 million metric tons/year through pulping or biorefining processes ( 11 , 12 ), but most isolated lignins have a dark color, strong odor, broad molecular weight distribution, and limited reactivity ( 5 , 13 – 15 ), which restricts them to low-value applications (e.g., fillers for tires, asphalt, or concrete) ( 16 , 17 ). Furthermore, there is substantial variability in the composition, chemical structure, cost, and environmental impact of technical lignins due to differences among feedstocks and pulping/refining techniques ( 13 , 18 – 20 ). Lignin deconstruction or depolymerization is a promising valorization approach for technical lignin, as it can generate products much more valuable than bulk lignin ( 1 , 2 , 4 , 5 , 21 , 22 ). Reductive catalytic deconstruction (RCD) and reductive catalytic fractionation (RCF) are particularly promising strategies that deconstruct lignin into its phenolic constituents at high yields, with the RCF process also incorporating a simultaneous LCB fractionation step ( 1 , 23 – 27 ). RCF or RCD typically uses a supported metal catalyst and an alcohol solvent (e.g., methanol or ethanol) to concurrently dissolve the lignin and deconstruct it at ~200° to 250°C ( Fig. 1A ) ( 1 ). Hydrogen gas often is added under pressure to promote hydrogenation over lignin recombination by “capping” reactive species; however, the alcohol solvent also can act as a hydrogen donor to avoid these recondensation reactions ( 24 , 28 ). A major drawback of current RCF/RCD-based techniques is the use of volatile, flammable solvents at high temperatures and pressures, resulting in significant safety hazards, high capital costs, and considerable energy requirements ( 29 ). Furthermore, high-pressure reactions typically are restricted to batch processing and have a limited throughput for a given system size, resulting in higher unit operation costs ( 29 , 30 ). Comparatively, low-pressure RCF/RCD has been reported in less volatile solvents (e.g., ethylene glycol) ( 24 , 31 ), but there are still solvent toxicity and sustainability challenges with these low-volatility alcohols ( 32 , 33 ). We note that the conventional RCF and conventional RCD process conditions are identical in this work; however, we use RCD going forward because the technical lignin specimens do not require the standard LCB fractionation step. Fig. 1. Overview of RCD processes. ( A ) Conventional RCD using methanol as a solvent, 40-bar external H 2 , and 5 weight % (wt%) Ru/C as a catalyst and ( B ) RD-RCD developed in this work using glycerin as a solvent and no external H 2 . At an operating temperature of 250°C, conventional RCD is pressurized to between 80 and 120 bar, whereas RD-RCD operates at ambient pressure. Here, we report an intensified reactive distillation (RD)–RCD process (see Fig. 1B ) that uses glycerin, an inexpensive, sustainable, and low-volatility by-product of biodiesel production, as the solvent. The high boiling point of glycerin (~290°C) enables ambient-pressure operation, along with simultaneous RD for more efficient separation of phenolic deconstruction products. Although RCF-type processes typically are used in lignin-first valorization strategies, this work focuses on the valorization of technical lignins obtained from common pulping methods and uses conventional RCD in methanol as a benchmark for deconstruction efficiency. The phenolic deconstruction products were functionalized with photopolymerizable acrylate groups, incorporated into a photocurable three-dimensional (3D) printing resin for additive manufacturing, and printed using a commercially available stereolithography (SLA) 3D printer. Furthermore, technoeconomic analysis (TEA) for both conventional and RD-RCD illustrated the effect of the process intensification and lignin source on economics for the production of a pressure-sensitive adhesive (PSA). Overall, the RD-RCD process improved the deconstruction efficiency versus RCD and resulted in more cost-competitive bioproducts. Last, a fast and inexpensive lignin screening method was established to predict deconstruction yields from thermal degradation behavior, providing a streamlined alternative to bench-scale RCD-type experimentation or 31 P nuclear magnetic resonance (NMR) spectroscopy to assess the economic feasibility of technical lignin valorization.",
"discussion": "RESULTS AND DISCUSSION Reactive distillation–reductive catalytic deconstruction RD-RCD and conventional RCD were used to deconstruct technical lignins obtained via Kraft, organosolv, soda, and thermomechanical pulping processes. A summary of the lignin feedstocks is shown in Table 1 . Phenolic product yields [weight % (wt%) on a lignin basis] for both processes with different technical lignins are provided in Fig. 2 . Conventional RCD resulted in yields ranging from 4.7 to 45.6%, whereas RD-RCD yields were between 8.3 and 31.7%. RD-RCD also generated solvent reforming by-products, such as substituted dioxolanes, solketal, and cyclopentenones, that were collected in the extracted distillate. These secondary products accounted for between 45 and 76% of the product mixture, depending on the lignin feedstock. Gas chromatography–mass spectrometry (GC-MS) plots, method validation, a representative mass balance, and yields for a catalyst-free control experiment are shown in figs. S1 to S9 and tables S1 to S4. Table 1. Lignin samples and key characteristics. \n Sample \n \n Process type \n \n Biomass type \n \n Precipitation method \n \n Region \n \n Species \n KHA1 Kraft Hardwood CO 2 acidification South America Eucalyptus KSA Kraft Softwood CO 2 acidification Europe Pine SNWA Soda Nonwood Mineral acidification Asia Wheat straw KHA2 Kraft Hardwood CO 2 acidification Europe Birch OHP Organosolv Hardwood Solvent precipitation Europe Beech BHF Biorefinery Hardwood Filtration North America Birch Fig. 2. Comparison of RD-RCD and RCD deconstruction yields. RD-RCD phenolic product yields (solid) and conventional RCD phenolic product yields (striped) for six lignin samples. All reactions were run for 15 hours at 250°C. The intensified RD-RCD provides several benefits over conventional RCD. Specifically, operating at ambient pressure overcomes a key scale-up hurdle, reduces capital requirements, and mitigates safety hazards ( 29 ). The use of glycerin is also advantageous because it is environmentally benign and is obtained as a by-product of biodiesel production ( 34 ). Additionally, RD-RCD simplifies product purification and reduces energy intensity because a cooling step is not required before extracting phenolic products in contrast to conventional RCD ( 6 ). The yields for both processes were comparable for most samples; however, conventional RCD resulted in higher phenolic product yields than RD-RCD for more “deconstructable” lignins (e.g., BHF, biorefinery hardwood filtration), whereas the reverse was true for more recalcitrant lignins (e.g., KSA, Kraft softwood CO 2 acidification). The concurrent removal of phenolic products through distillation as they form may have driven these differences in yields between the two processes; however, further studies are needed to elucidate the underlying reaction mechanisms and confirm this hypothesis. The two processes generated bio-oils with varying product distributions. As expected, softwood lignin (sample KSA) yielded almost exclusively guaiacyl (G) units; hardwood lignins (samples KHA1, Kraft hardwood CO 2 acidification1, KHA2, Kraft hardwood CO 2 acidification2, OHP, organosolv hardwood solvent precipitation, and BHF) produced a mixture of G and syringyl (S) compounds; and herbaceous lignin (sample SNWA, soda nonwood mineral acidification) formed p -hydroxyphenyl (H), G, and S units. The RCD products generally contained larger amounts of S compounds than RD-RCD because S units are less volatile than H and G compounds and do not distill completely in RD-RCD. For instance, the selectivity of S compounds (i.e., the sum of the selectivities of S-type compounds) generated from sample BHF was 74% for RCD, whereas this value was 22% for RD-RCD. This behavior potentially accounts for the differences in yields between RD-RCD and RCD for samples that generated comparatively large amounts of S compounds via RCD, such as BHF and OHP. The identified phenolic compounds and their respective selectivities (i.e., the proportion of all detected phenolic compounds) for each technical lignin are shown in Fig. 3 . Fig. 3. Product selectivities for phenolic RCD and RD-RCD products. The sample name indicates conventional RCD (e.g., KSA), and RD indicates reactive distillation (e.g., RD KSA). Dashed lines separate feedstock type (horizontal lines) and lignin subunit type (vertical lines). Selectivities above 40% are shown in white text for clarity. Beyond the proportions of H, G, and S products, the substituent groups at the 4-position varied depending on the process and the type of lignin. The substitutions common to lignin oils from both RD-RCD and RCD were hydrogen, methyl, ethyl, and propyl chains, and in the case of RD-RCD, some propenyl and vinyl moieties were generated. Conventional RCD also produced hydroxypropyl-substituted phenolics. Between the two deconstruction approaches, RCD generated higher amounts of propyl-substituted phenolics, whereas RD-RCD resulted in compounds with shorter alkyl chains or no 4-substitution. This trend is most apparent with sample KSA; the selectivity for propylguaiacol was 48% for RCD, and the selectivity for guaiacol was 57% for RD-RCD. Among the different pulping methods, Kraft lignins (KHA1, KSA, and KHA2) produced greater amounts of unsubstituted compounds (e.g., guaiacol and syringol), and soda, biorefinery, and organosolv lignins (i.e., lignins from milder pulping/fractionation techniques) resulted in bio-oils rich in propyl substituents. For instance, the selectivity for propyl-substituted phenolics was 64% for BHF RD-RCD bio-oil and only 12% for KHA1 RD-RCD bio-oil. Pulping and deconstruction methods significantly affect selectivities within the group of shared RCD/RD-RCD products, and the relationship between the feedstock (biomass type and separation method), deconstruction technique, and product selectivities could be leveraged to tailor the slate of biobased chemicals generated in a biorefinery. Additionally, RD-RCD generates compounds with unsaturated substituents and solvent reforming by-products that are absent in conventional RCD conducted with external H 2 . For instance, isoeugenol, propenylsyringol, and vinylguaiacol are present in RD-RCD distillates but not in the conventional RCD bio-oils. Products with unsaturated functional groups likely are generated because RD-RCD removes them via distillation, whereas in conventional RCD, the products undergo further hydrogenation. However, note that conventional RCD can be altered to generate more unsaturated products by eliminating external H 2 or removing the catalyst altogether at the expense of reduced yields ( 24 , 35 , 36 ). Among the lignin samples, those from harsher pulping processes, such as sample KSA, produced the lowest amount of unsaturated compounds in RD-RCD (1.4% of the phenolic products), and lignins from gentle processes, such as sample BHF, had the highest proportion of unsaturated products (11.4%), which can be rationalized on the basis of acid content in the feedstocks ( 37 , 38 ). Varying levels of glycerin decomposition/reforming compounds, including cyclopentanones, cyclopentenones, solketal, and dioxolanes, were also detected in all RD-RCD bio-oils (see figs. S1 to S7 and table S1). These reforming products are generated even in the absence of lignin (fig. S10). The boiling points of these chemicals generally are lower than those of the phenolic products (~150°C versus >200°C) such that they could be separated via distillation. Alternatively, condensed-phase (e.g., chromatographic) methods could be used to further improve economics and reduce the environmental impact of the process in comparison to distillation ( 30 ). The purified compounds then could be treated as co-products with applications as platform chemicals or green solvents. For example, cyclopentenones are valuable precursors for pharmaceuticals and fragrances ( 39 ). Biobased 3D printing RCF/RCD deconstruction products or “bio-oils” have been used as precursors for numerous bio-derived products, such as polymers and pharmaceuticals ( 6 , 35 , 40 ), and the generation of biobased materials from RCF/RCD oils is particularly attractive, as it can minimize the need for significant product purification ( 5 , 6 ). Additive manufacturing is gaining traction within the biobased materials space as a versatile fabrication approach ( 3 , 4 , 41 ), and SLA 3D printing is a technique that uses visible or ultraviolet light to photocure resins ( 42 ). Lignin and lignin-derivable compounds have been incorporated into photocurable resins; however, to the authors’ knowledge, the use of RCF/RCD oil mixtures for SLA printing has not been reported ( 3 , 41 , 43 ). Furthermore, most SLA resins are composed of alkyl methacrylates, and translating biobased, lignin-derived monomers to 3D printing resins offers a new class of materials for additive manufacturing. To demonstrate the utility of RD-RCD lignin oils, acrylated SNWA bio-oil was incorporated into an SLA resin and 3D printed using a commercially available printer. SNWA bio-oil was chosen because it had a high syringol content and was produced from sulfur-free lignin. The complete 3D printing formulation contained 40 wt% acrylated bio-oil obtained directly from the RD-RCD products generated in this work, 40 wt% lignin-derivable vanillyl alcohol diacrylate, 15 wt% Peopoly resin, and 5 wt% photoinitiator. The diacrylate and Peopoly resin were added to improve overall print quality (fig. S11). The acrylation reaction and a photo of a printed object from our formulation are shown in Fig. 4 . A second resin was prepared from the acrylated bio-oil and vanillyl alcohol diacrylate without Peopoly resin, and the material was cured following the same procedure. However, the print without Peopoly exhibited breaking/flaking after several minutes (fig. S11). Fig. 4. SLA 3D printing. ( Top ) Functionalization reactions for RD-RCD bio-oil and vanillyl alcohol. ( Bottom ) 3D printed UD using the SLA resin in a commercially available SLA printer. The photo was taken after rinsing with isopropanol and post-curing under an ultraviolet lamp for 10 min. R.T., room temperature. The printed SLA resin is a salient example of the multiple potential applications for RD-RCD bio-oils and other co-products. The material properties (e.g., glass transition temperature and solvent resistance) can be tuned by leveraging structure-property relationships ( 5 , 6 , 44 – 47 ) to choose mixtures of bio-oils from different feedstocks for a target monomer composition. Furthermore, formulation for SLA printing required the addition of a cross-linker because inert (i.e., non-acrylated) components in the bio-oil mixture solvated the polymer and prevented the formation of a solid print, although polymers from the phenolic acrylates are expected to be in a glassy state at room temperature ( 6 ). This behavior demonstrates the potential utility of the glycerin reforming by-products as green solvents and presents an exciting opportunity for further process intensification by avoiding the separation of functionalized monomers and inactive by-products using energy-intensive methods such as distillation. TEA of lignin-based PSA production RCD and RD-RCD were compared using a biobased PSA as an example product. The PSA was chosen over SLA printing resins because the market is more mature and higher in volume than the 3D printing resin market. The adhesive market was estimated at $52.6 billion in 2017 and is projected to grow at 5.6% annually through 2025, with PSAs accounting for ~27% of the revenue ( 48 ). In a previous study, a triblock polymer PSA synthesized by Wang et al. ( 6 ) using a poplar wood RCF oil exhibited superior adhesive performance in comparison to incumbent commercial products when evaluated via American Society for Testing and Materials (ASTM) standard tests, even without added tackifiers. Given the high performance and versatility of this biobased PSA, a similar triblock polymer was used for TEA in this work. The lignin deconstruction yields and product distributions obtained from the conventional and RD-RCD experiments were used to design two lignin-based PSA production processes in Aspen Plus V11 (Aspen Technology) ( 49 , 50 ), starting with the RCD or RD-RCD deconstruction reactions and followed by purification, functionalization, and polymerization. Additional details are included in the Supplementary Materials (table S5 and fig. S12), and process flow diagrams are provided in figs. S13 to S15. The lignin capacity of each system was assumed to be 18,144 metric tons/year as a base case. First, lignin was deconstructed by RCD or RD-RCD to produce a mixture of phenolic compounds at 250°C. In the case of RD-RCD, solvent reforming by-products also were generated, which were assumed to be sold as fuel. In both RCD and RD-RCD, the phenolic products required further purification by liquid-liquid extraction with hexane. For RCD, the extraction separated the phenolic compounds from the residual lignin and lignin oligomers. In RD-RCD, the extraction served to isolate the phenolic products in the distillate from hydrophilic by-products and glycerin. After extraction, the phenolic mixtures were functionalized using methacrylic anhydride (MAAH) with 4-dimethylaminopyridine (DMAP) as a catalyst at 45°C. The residence time for the functionalization reaction was 3 hours to ensure complete conversion. Next, quicklime (CaO) was used to neutralize the unreacted MAAH and methacrylic acid (MAA) by-product. Then, the functionalized acrylic monomers were extracted from the aqueous phase with anisole. The extracted monomers were copolymerized with butyl acrylate (BA) by reversible addition-fragmentation chain transfer (RAFT) polymerization in anisole at 70°C using butan-2-one-3-ethyltrithiocarbonate as the chain transfer agent (CTA) and 2,2′-azobis(isobutyronitrile) as the initiator ( 6 ). After polymerization, the anisole was removed under reduced pressure to obtain the PSA polymer. Three representative lignin samples (KSA, OHP, and BHF) were considered in this analysis, and because the biobased PSA polymer’s market price is unknown, the minimum selling price (MSP) was chosen as the metric to compare the production costs. The most significant driver was the deconstruction yield, as illustrated in Fig. 5 , figs. S16 and S17, and tables S6 to S12. For instance, with KSA lignin, RD-RCD generated PSA polymer at 1202 kg/hour and solvent reforming by-product at 271 kg/hour, and RCD produced PSA at just 592 kg/hour. The low throughput of RCD resulted in a MSP of $20,732/t, whereas the comparatively higher throughput of RD-RCD enabled a 62% lower MSP of $7901/t, as shown in Fig. 5, A and B . The differences between the two processes were less pronounced for other lignin samples. RD-RCD outperformed RCD in all cases, even when RD-RCD yields were lower. For example, the RCD PSA production rate for BHF was significantly higher than that for RD-RCD, 3545 kg/hour versus 2478 kg/hour, but the MSP was $6440/t for RCD in comparison to $6163/t for RD-RCD ( Fig. 5, C and D ) because the glycerin reforming by-products from RD-RCD can be sold as biofuel. Similarly, with OHP lignin, RCD and RD-RCD generated PSA at 2030 and 1845 kg/hour, respectively, leading to MSPs of $8773/t and $7067/t (figs. S16 and S17). Fig. 5. MSP breakdown for different feedstocks. ( A ) RCD with KSA lignin feedstock. ( B ) RD-RCD with KSA lignin feedstock. ( C ) RCD with BHF lignin feedstock. ( D ) RD-RCD with BHF lignin feedstock. WWT, wastewater treatment. The RD-RCD process exhibited improved unit economics in comparison to conventional RCD, particularly in the deconstruction stage. The solvent reforming by-product biofuel revenue and significantly lower capital requirement for the reactor were the primary drivers of the favorable economic performance ( Fig. 6A ). Both processes had the same plant configuration for the functionalization and polymerization stages, and because the expenses at this stage are dominated by the raw material costs, including MAAH and BA ( Fig. 6A ), the contributions of these operations to the MSP are comparable for both systems and nearly independent of the deconstruction yields. Fig. 6. Contribution of process stages to MSP and impact of plant capacity. ( A ) Economic contribution of process stages to the MSP for RD-RCD (solid) and RCD (striped) using BHF lignin at a scale of 18,144 t/year. Raw materials for each stage did not include the products from the previous stage; for instance, the phenolic compounds from RCD were not included in the raw material cost for the functionalization step. ( B ) MSP of PSA relative to plant capacity for RCD (purple squares) and RD-RCD (green circles) using BHF lignin. func., functionalization; decon., deconstruction (A). The TEA also provided insights into future cost-saving and revenue-generating opportunities. The most efficient approach to reduce the cost of the deconstruction stage for both RCD and RD-RCD is to reduce the capital requirements, especially the cost of the deconstruction reactors. This goal could be achieved by optimizing the catalyst and reaction conditions to reduce the reaction time and solvent consumption. Lowering raw material costs also would have a substantial impact on process economics. For instance, the cost of MAAH is a leading contributor to the MSP, and converting the MAA back to MAAH would significantly reduce the final PSA price ( 51 ). Additionally, expanded implementation of commercial lignin recovery technologies will increase the supply of lignin and likely contribute to reduced feedstock costs and MSPs. Plant economics also could be drastically improved by generating saleable products from the residual, partially deconstructed lignin. In this work, the residual lignin is considered waste to simplify the comparison between RCD and RD-RCD; however, as technical lignins tend to be somewhat recalcitrant, this stream accounts for most of the input lignin feedstock. Oligomeric lignin has been used in numerous materials applications ( 27 , 52 – 56 ), and it is possible to recover this partially deconstructed lignin through precipitation for further upgrading (see fig. S9 and table S3). The economic impact of the chemical plant capacity was evaluated using BHF lignin feedstock capacities ranging from 5000 to 30,000 t/year in a sensitivity analysis ( Fig. 6B ). With an increase in plant capacity to 30,000 t/year, the MSPs of both RCD and RD-RCD PSAs were reduced to $6362/t and $5924/t, respectively. At this scale, the MSP of the RD-RCD PSA was lower than that of RCD, even without selling the solvent reforming by-products as fuel. However, at a significantly smaller scale, such as 5000 t/year, RCD and RD-RCD had comparable MSPs of $7167/t and $7264/t, respectively. Furthermore, the economic impact of plant capacity was more pronounced for RD-RCD than conventional RCD, and thus, the scale-up advantages of RD-RCD are more substantial. Additional sensitivity analyses were performed using other process parameters, as shown in figs. S18 and S19, with more details provided in the Supplementary Materials. Overall, RD-RCD was economically favorable with all three of the feedstocks included in this analysis. For lignins with lower yields, such as KSA lignin, the benefits of RD-RCD were more pronounced than for lignins with higher deconstruction yields (e.g., BHF). MSPs generally were in the range of $6000/t to $8000/t and are thus competitive with incumbent commercial PSAs. Key parameters for process economics were plant size, solvent consumption, energy usage, and operation mode (batch versus continuous). Technical lignin screening method Lignin deconstruction yields had a significant impact on process economics, and thus, simple screening methods to assess the “deconstructability” of technical lignins are needed to determine the economic viability of valorization approaches. To this end, RCD phenolic yields from both conventional RCD and RD-RCD were used to develop an empirical relationship to predict phenolic yields with thermal degradation characteristics measured by thermogravimetric analysis (TGA). Each sample exhibited a characteristic TGA decomposition peak (%/°C), and the temperature at which 40% mass loss occurred ( T 40% ) indicated a point just after the peak for all samples ( 57 ). Lignin phenolic product yields (on a lignin basis) were plotted against T 40% and fit to exponential functions using least squares. The experimental data and exponential fits for RD-RCD and RCD are shown in Fig. 7 (A and B, respectively), and TGA curves are shown in figs. S20 to S25. Similarly, RD-RCD and RCD yields were correlated with the amount of condensed phenolic units as determined by 31 P NMR spectroscopy (figs. S26 to S31 and table S13), as shown in Fig. 7 (C and D). Fig. 7. Screening method for technical lignin feedstocks. ( A ) RD-RCD phenolic yields versus T 40% (from TGA). ( B ) RCD phenolic yields versus T 40% . ( C ) RD-RCD phenolic yields versus condensed phenol content (from 31 P NMR spectroscopy). ( D ) RCD phenolic yields versus condensed phenol content. Experimental data (RD-RCD: orange triangles; RCD: green circles) and an exponential fit (dashed lines) are shown on logarithmic scales. The exponential fits to deconstruction yields and T 40% had R 2 values of 0.87 and 0.75 for RCD and RD-RCD, respectively. The fits to deconstruction yields using condensed phenol content were of similar quality, with R 2 values of 0.85 and 0.76, respectively. The inverse correlation of deconstruction yield and T 40% arises from the distribution of interunit linkages; lignins with higher thermal stabilities contain more recalcitrant bonds, whereas lignins with relatively low thermal stabilities contain more labile linkages that are more easily broken during RCD or RD-RCD. The effect of thermal stability on yield was stronger for conventional RCD in comparison to RD-RCD ( Fig. 7, A and B ), likely because RD-RCD removed phenolic products as they formed and prevented recondensation, whereas RCD relies on hydrogenation to mitigate condensation reactions. Condensed phenol content also was correlated with RCD yield for both processes to serve as a benchmark ( Fig. 7, C and D ), and the two correlations were of similar quality, as suggested by comparable R 2 values ( Fig. 7 ). Although the correlations are comparable, the condensed phenol metric necessitated using previously reported 31 P NMR spectroscopy data from experiments that were comparatively more complex than was TGA ( 58 ). For instance, lignin samples first were phosphorylated for NMR spectroscopy analysis, whereas no pretreatment was needed for TGA ( 58 ). Thus, the development of the TGA-based correlation enables facile screening of potential technical lignin feedstocks for RCD valorization strategies with a level of accuracy comparable to the more complex NMR spectroscopy method. This correlation is currently limited to technical lignins because cellulose, hemicellulose, extractives, and other biomass components may lead to more complex thermal decomposition profiles. Despite this constraint, this method to predict deconstruction yields via TGA can screen technical lignin samples for compatibility with other lignin valorization processes, such as biological, pyrolytic, and other catalytic valorization approaches. In summary, RD-RCD is introduced to simultaneously fractionate biomass, deconstruct lignin, and distill the phenolic products at ambient pressure. It enables facile scale-up and product separations compared to conventional lignin-first fractionation/deconstruction strategies, mitigates safety hazards associated with high-pressure RCF/RCD, and eliminates the energy-intensive H 2 . It also replaces alcohol solvents with glycerin, a more environmentally friendly, inexpensive, and abundant by-product of biodiesel production. The RD-RCD bio-oil was functionalized and incorporated into an ~80% biobased resin and printed successfully with a commercially available SLA 3D printer, illustrating the utility of these lignin-derived mixtures in high-performance materials applications. TEA for the production of a PSA demonstrated that RD-RCD provides favorable MSPs compared to traditional RCD for three different technical lignin feedstocks, with as much as a 60% reduction for softwood Kraft lignin. Last, we introduced a straightforward, fast, and inexpensive thermal decomposition method for screening potential feedstock candidates. Together, this work demonstrates the potential of technical lignin as an inexpensive and abundant resource for producing value-added chemicals and materials and a scalable valorization pathway for feedstock selection, intensified deconstruction, product fabrication, and economic evaluation."
} | 7,807 |
32843539 | PMC7449608 | pmc | 8,343 | {
"abstract": "Ecosystem management must be viewed in the context of increasing frequencies and magnitudes of various disturbances that occur at different scales. This work provides a glimpse of the changes in assembly mechanisms found in microbial communities exposed to sustained changes in their environment. These mechanisms, deterministic or stochastic, can cause communities to reach a similar or variable composition and function. For a comprehensive view, we use a joint evaluation of temporal dynamics in assembly mechanisms and community structure for both bacterial taxa and their functional genes at different abundance levels, in both disturbed and undisturbed states. We further reverted the disturbance state to contrast recovery of function with community structure. Our findings are relevant, as very few studies have employed such an approach, while there is a need to assess the relative importance of assembly mechanisms for microbial communities across different spatial and temporal scales, environmental gradients, and types of disturbance.",
"conclusion": "Concluding remarks. The joint evaluation of assembly mechanisms, community structure, and function of bacterial taxa and functional genes provided in-depth understanding of the response of complex microbial systems to a doubling of the organic loading rate. This press disturbance altered community function, structure, and assembly mechanisms. Disturbance had an effect not only on community function but also on its functional potential, emphasizing the relevance of assessing communities of organisms together with communities of genes. Through null-model analyses, community assembly was found to have a stronger deterministic component in the common fraction, whereas the role of stochastic mechanisms was higher for the less abundant portion of the community. Also, reactors that recovered functions after returning to low-organic-loading conditions maintained different α- and β-diversities compared to reactors that had not been disturbed, in terms of both taxonomic and functional genes, showing that resilience based on community function does not necessarily translate into resilience based on community structure. Further, we urge caution when assessing microbial community assembly mechanisms, as results can vary depending on the approach (16S rRNA gene metabarcoding and shotgun metagenomics taxonomic or functional gene community profiling) and whether the focus is on taxa or genes. Not only can genes suggest different dominance patterns of stochastic versus deterministic mechanisms of community assembly compared to taxa, but the fraction of the community driving such assembly mechanisms—common versus rare—can also differ. Finally, this study employed alteration in the substrate feeding scheme as a type of press disturbance. More research is needed on different types of disturbances (e.g., pollutant additions, pH shifts, and temperature changes) within different complex microbial systems at different scales to broadly validate our observations. Additional studies covering different spatial and temporal scales, environmental gradients and types of disturbance could lead to a general framework of how press disturbances alter the structure, function, and assembly mechanisms of microbial communities.",
"introduction": "INTRODUCTION Microbes drive all biogeochemical cycles on Earth, with microbial communities providing important ecosystem functions that impact all other forms of life ( 1 ). Community structure, often described in terms of α- and β-diversity, is thought to have an effect on ecosystem function ( 2 ). However, our capacity to predict and manage the functions of microbial communities and how they are linked to community structure is still limited ( 3 ). In this regard, engineered systems like sludge bioreactors for wastewater treatment constitute model systems for microbial ecology studies ( 4 ), with measurable ecosystem functions, such as carbon and ammonia removal, that not only are important in practice ( 5 ) but also involve complex microbial communities in a controlled environment ( 6 ). In ecology, disturbances are believed to have direct effects on ecosystems by altering community structure and function ( 7 ). Press disturbances that impose a long-term continuous change of species abundances by altering the environment ( 8 ) are of interest in microbial ecology, as they can drive systems to alternative stable states with different community function and structure ( 9 ). These disturbances could occur in the form of environment modifications that are not directly harmful to organisms, while still providing less abundant community members opportunities to grow ( 10 ). In sludge bioreactors, a continuous alteration in the substrate feeding scheme can trigger changes in community function and structure, yet whether these changes are reproducible ( 11 ) and whether they can be reversed when the disturbance ceases remain unknown. Community assembly mechanisms are inherently linked with ecosystem function, as they play an important role in shaping community structure ( 12 ). These mechanisms can be either deterministic ( 13 ) or stochastic ( 14 ), and they may act in combination to shape patterns of community assembly ( 15 – 18 ). Disturbance is thought to be a main factor driving these underlying mechanisms of community assembly ( 19 ), yet a predictive understanding of its effects is missing ( 20 ). Disturbance can promote stochastic assembly mechanisms that lead communities to divergent states of structure and function ( 21 , 22 ); therefore, studies assessing its effects require replicated designs ( 23 ). Also, microbial communities within wastewater treatment systems have been shown to harbor a core group of 100 to 800 abundant (i.e., common) operational taxonomic units (OTUs) across plants and countries ( 24 , 25 ). Activated sludge community dynamics have been suggested to differ for common and less abundant (i.e., rare) taxa on the grounds of distance-based analyses ( 26 – 29 ), proposing that common OTUs are driven by deterministic mechanisms of assembly. The contribution of the remaining rare taxa to biochemical transformations is as yet unknown, but their potential as a seed bank for gene and taxonomic diversity ( 30 ) merits consideration. Indeed, rare taxa and genes may become very abundant under conditions ( 31 ) that can be elicited by disturbance. The contribution of assembly mechanisms is often quantified via null-model analyses ( 32 ). In studies of microbial systems, such analyses are usually focused on community data obtained from a fragment of the 16S rRNA gene ( 17 , 18 , 33 – 38 ), which is known to be highly conserved between different species of prokaryotes ( 39 ). On the other hand, the use of metagenomics would allow inference of assembly mechanisms from the whole-community DNA ( 40 ), enabling the assessment of variability in genes that are less conserved and that could unveil important aspects of community assembly. However, studies usually employ these sequencing methods separately, as their results are thought to be hard to reconcile ( 41 ). Microbial communities can also be assessed in terms of their functional gene structure ( 42 ), and while functional gene assembly can be assessed via microarray ( 22 , 43 ) or clone library ( 44 ) approaches, a more complete view of the functional gene structure of the microbial community can be obtained through shotgun metagenomics. Hence, an assessment of community assembly mechanisms combining both 16S rRNA gene amplicon and shotgun metagenomics sequencing would provide important insights toward a better understanding of the effect of disturbance in community assembly, structure, and function, while still allowing a comparison with prior findings. Further, such evaluation is seldom carried out in complex and engineered microbial systems like bioreactors for wastewater treatment. The objective of this work was to test the effect of a press disturbance by doubling the organic load in a replicated set of activated sludge bioreactors at mesocosm scale. Based on our findings in a prior study at microcosm scale ( 21 ), we expected the disturbance to affect community function, structure, and assembly, with the hypothesis of a stronger deterministic effect at the disturbed level. Samples were analyzed using shotgun metagenomics, 16S rRNA gene metabarcoding, and effluent chemical characterization. Patterns of α- and β-diversity were employed to assess temporal dynamics of community structure. Assembly mechanisms were evaluated through a mathematical null model on the effective bacterial turnover expressed as a proportion of total bacterial diversity. Finally, the disturbance was removed during the last 14 days of the study to evaluate if community function, structure, and assembly would display signs of recovery.",
"discussion": "DISCUSSION Disturbance leads to different community function and structure. Community function and structure, in terms of α- and β-diversity, were clearly different in reactors with low and high organic loading. Diversity metrics are among the fundamental descriptive variables of community ecology ( 46 ), on both a local (α-diversity) and a spatiotemporal (β-diversity) scale ( 47 ). In our study, high-organic-loading disturbance led to a reduction of the nitrite oxidation function in these reactors, coinciding with a marked community differentiation from the low-organic-loading reactors in terms of β-diversity ( Fig. 2B ), although only a few taxa are known to perform this function. However, we also found higher taxonomic α-diversity in the high-organic-loading reactors ( 2 D ASV and 2 D Genus ) ( Fig. 1 ), which means that the relative abundances were more evenly distributed among taxa for these communities. Since the low-organic-loading reactors displayed better chemical oxygen demand (COD) removal and complete nitrification with almost no residual NH 4 + -N or NO 2 − -N, it was expected that they would harbor more diverse communities, as community evenness was suggested to be a key factor in preserving the functional stability of an ecosystem ( 48 ). We did indeed observe higher functional gene α-diversity ( 2 D IP2G ) ( Fig. 1C ) for the low-organic-loading reactors toward the end of the study, which is similar to the opposing trends of taxonomic and functional gene α-diversity previously reported after autoclave sterilization of soil microbial communities ( 49 ). Further, the dominant genera of the acclimation phase gave way to two separate clusters with different dominant genera at the end of the study ( Fig. 3 ; Fig. S2 ). Bacterial successional dynamics have been described before for activated sludge systems, but studies often present analysis at the phylum or class level of taxonomy ( 26 , 27 ). Analysis of broad taxonomic categories may hide important assembly mechanisms operating at lower taxonomic levels ( 50 ), which is why we present successional changes at the genus level in this study. For example, the two main genera across low- and high-organic-loading reactors ( Fig. 3 ) were Thauera and Paracoccus , which are denitrifying organisms in activated sludge systems ( 51 ). These genera likely benefited from their versatility in carbon substrate uptake and ability to reduce available nitrite and nitrate during the anoxic phase of the bioreactor cycle ( 52 ). The accumulation of nitrite during the aerobic phase in high-organic-loading reactors seemed to have benefited Paracoccus more than Thauera . The main nitrifying genera were Nitrosomonas and Nitrospira , whose relative abundance was diminished in high-organic-loading reactors, consistent with a reduction in the nitrite oxidation function (details in reference 45 ). We further evaluated the succession of trait complexes, which are a product of the expression of multiple true traits ( 53 ), to identify patterns suggesting differential functional gene investment at different stages of the study. Traits like cell motility and cell wall were enriched in low-organic-loading reactors, while traits of replication and repair prevailed in high-organic-loading ones ( Fig. 4 ). The enrichment of maintenance functions across disturbed reactors exemplifies how organisms have to invest resources to adapt to changes in the environment ( 54 ). High-organic-loading reactors also showed an increased prevalence of ATP-binding cassette transporter ( Fig. S3A and B ) and stress response ( Fig. S3C ) genes, which encode traits related to cell survival. The prevalence of certain functional genes suggests community-level tradeoffs under disturbance similar to the ones described by life history strategy theory ( 55 ), comparable to those reported for sludge bioreactors under pollutant disturbance ( 56 ). Taken together, these observations highlight how organisms face tradeoffs when allocating resources to certain traits to maximize their fitness, which depends on abiotic and biotic interactions within their habitat ( 57 , 58 ). Additionally, two of the disturbed reactors recovered the nitrite oxidation function after returning to low-organic-loading conditions for 14 days ( Fig. 5 ). However, the α- and β-diversity with respect to taxonomy and functional genes did not revert to previous levels ( Fig. 1 and 2 ). The fact that an altered community was still able to provide the functions of the original one (carbon removal and nitrite oxidation) supports the notion of functional redundancy ( 59 ) in complex microbial systems like bioreactors, where different types of organisms are capable of performing a wide range of functions ( 60 ). Community assembly mechanisms differ for taxa and genes, as well as for common and rare fractions. In terms of relative deterministic strength, disturbed reactors showed a stronger role of deterministic mechanisms for all three types of data sets evaluated ( Fig. 6 ), likely due to the selective pressure via environmental filtering ( 61 ) imposed by the disturbance. Deterministic mechanisms dominated bacterial community assembly at the ASV taxonomic level across all reactors and were stronger under high organic loading ( Fig. 6A ). This agrees with previous studies of press disturbance in mesocosm sludge bioreactors using a micropollutant (3-chloroaniline), which reported higher similarity among disturbed reactors than among control reactors via 16S rRNA gene amplicon analysis ( 62 , 63 ), although without quantifying assembly mechanisms. Partitioning the contribution of common (90% of the reads accumulated) and rare (<10% of the reads accumulated) portions of the data revealed that the common fraction was significantly driven by deterministic assembly mechanisms ( Fig. 6B and 8B ), while the rare portion displayed high stochasticity ( Fig. 6C and 8C ). Stochastic mechanisms of ecological drift can operate in the portion of taxa at low relative abundances ( 20 ), while the overall community assembly is mainly deterministic. These findings are in agreement with prior studies that reported higher variability for rare taxa than common taxa in full-scale and mesocosm bioreactor sludge systems ( 26 – 29 ), albeit without the use of null-model analysis. Although dissimilarity observations can highlight deterministic effects, they have low power for inferring stochasticity ( 23 ), which is why we partitioned community assembly mechanisms via null-model analysis in the present work. Similarly, a recent study on granular biofilm reactors using one simple carbon source also reported stronger homogeneous selection for abundant taxa and higher stochastic assembly via drift for low-abundance taxa ( 38 ). This was done using the type of null model that was applied on the metabarcoding data set in this study ( Fig. 8 ), and it has the advantage of incorporating phylogeny into the analysis ( 64 ) but does not take advantage of replicated designs the way the null model of Kraft et al. ( 65 ) does. As the results from null-model analyses are very sensitive to the models, algorithms, and diversity metrics employed ( 66 ), concordant outcomes in studies using different approaches are desired, and this requires more research ( 23 ). Further, the aforementioned studies employed 16S rRNA gene metabarcoding with OTU-type clustering for their community analyses. Here, we inferred community assembly mechanisms from ASV data, which has several benefits over traditional OTU clustering ( 67 ), including generating 10 to 100 times fewer spurious units ( 68 ), which reduces bias for α-diversity estimations. Although classification criteria for rare and common fractions are arbitrary ( 30 , 69 ), there is consistency between our community assembly observations from 16S rRNA gene metabarcoding data and the current literature. Nonetheless, further research is needed to evaluate whether the rare fraction is biologically active ( 70 ). Metagenomics data also revealed higher deterministic strength at high organic loading for taxa ( Fig. 6D ) and slightly higher DS for genes ( Fig. 6G ). This concurs with the higher deterministic assembly reported previously for sludge microcosm reactors that were press disturbed with 3-chloroaniline and compared to undisturbed ones, via null-model analysis on metagenomics genus-level data ( 21 ). The study included a different type of press disturbance in a bioreactor scale 2 orders of magnitude smaller (20 ml) than the ones in this study (5 liters), and also in a shorter time span (35 days). Since diversity is multidimensional and scale dependent ( 71 ), assembly mechanisms and the results of null-model analyses can differ at different spatial and temporal scales ( 20 , 66 ); therefore, these concurring effects of press disturbance in community assembly that were observed at microcosm and mesocosm bioreactor scales are relevant. Indeed, it is necessary to quantify the relative importance of stochastic and deterministic assembly mechanisms for microbial communities across different spatial and temporal scales, environmental gradients, systems, and types of disturbance ( 20 , 23 ). Most of the DS points for the overall IP2G gene data set fell below 50% ( Fig. 6G ), implying that stochastic mechanisms of assembly were stronger for genes than for taxa. Both data sets used the same number of individuals per sample for the null-model analysis (100,000) and thus represent a balanced sampling effort for such a comparison. Separate evaluation of the common and rare fractions in terms of DS showed that deterministic assembly mechanisms were stronger in the common fraction ( Fig. 6E and H ), while the rare portion was more influenced by stochasticity ( Fig. 6F and I ), which is similar to what we observed for ASV data using two different null-modeling approaches. Further, β-diversity was always greater than expected for bacterial taxa under the null model regardless of the sequencing method ( Fig. 7 ; Table S3 ), suggesting that taxa tended to be more aggregated within replicate bioreactors than expected by chance. Aggregation can be explained by processes of habitat filtering ( 72 ) due to the recurrent conditions at each low and high organic loading level, as well as by dispersal limitation ( 73 ), which is a condition of the experimental design using reactors as closed systems without immigration. However, the opposite was shown for the functional gene data set, where the mean expected β-diversity of individual IP2G genes under the null model was higher than the observed β-diversity; i.e., replicate bioreactors were more similar than expected by chance ( Fig. 7 ). These results highlight that functional gene data can indicate patterns of dominant stochastic versus deterministic mechanisms of community assembly that differ from patterns obtained after analysis of taxon data. Sequencing methods, diversity, and assembly mechanisms: what is consistent and what is not. Using two different community profiling approaches on the same samples, we found that press disturbance favored deterministic assembly mechanisms, where bioreactor bacterial communities at low and high organic loading levels clearly separated into two different clusters in terms of β-diversity for taxa and genes. Both methods identified a greater influence of deterministic assembly mechanisms in the common fraction of the community, whereas stochasticity was more important in the low-abundance fraction. These complementary analyses suggest that one should be able to assess the effect of a press disturbance on both β-diversity and the overall assembly mechanisms of microbial communities, regardless of the sequencing approach employed and whether the focus is on taxa or functional genes. However, methodologies were inconsistent when the relative contributions of deterministic and stochastic mechanisms of community assembly were probed, with metagenomics data displaying higher stochasticity for genera than when ASVs from 16S rRNA gene metabarcoding were used. This was not an effect of assessing different levels of taxonomic resolution, as higher resolution levels were shown to be more conserved than lower ones ( 50 ) and therefore displayed assembly mechanisms that were more deterministic. However, this could have been the effect of a different sampling coverage under the null-model analysis, because metabarcoding renders a lower total number of reads per sample, about 1 order of magnitude, than metagenomics analysis. When the focus was only on metagenomics data, functional gene assembly was found to be more stochastically driven than that of taxa at the genus level, even under a balanced sampling coverage. Further, there were differences in terms of which fraction of the community, common or rare, affected overall community assembly. Overall functional gene assembly appeared to be driven by a balance between common and rare fractions. On the other hand, overall assembly of bacterial taxa had DS values similar to those of the rare fraction and βNTI values similar to those of the common fraction. Similarly, the effect of disturbance on α-diversity was different for genes and taxonomic data. The challenge of reconciling results from different sequencing methods has been recognized and requires further research ( 41 ). Differences in assembly mechanisms can arise due to the distinct biological signature being evaluated when a fraction of a specific DNA marker like the 16S rRNA gene is targeted, compared to using DNA of the whole meta-community as the basis of classification of taxa and functional genes. Currently, the 16S rRNA amplicon approach is expected to result in less ambiguity, because fewer singletons are generated than with shotgun metagenomics. Not surprisingly, the majority of existing studies on microbial community assembly focused on taxonomic OTU data sets from 16S rRNA gene metabarcoding ( 17 , 18 , 33 – 38 ), with very few studies also using shotgun metagenomics data to quantify community assembly mechanisms in bioreactor systems for wastewater treatment ( 21 ). Future studies on the impact of disturbance in assembly mechanisms would benefit from also incorporating whole-community DNA information, enabling the assessment of less conserved genes that could unveil important aspects of community assembly ( 40 ). The latter is promising given the increasing availability of methods that are faster, are PCR independent, and allow long-read sequencing ( 74 ). Concluding remarks. The joint evaluation of assembly mechanisms, community structure, and function of bacterial taxa and functional genes provided in-depth understanding of the response of complex microbial systems to a doubling of the organic loading rate. This press disturbance altered community function, structure, and assembly mechanisms. Disturbance had an effect not only on community function but also on its functional potential, emphasizing the relevance of assessing communities of organisms together with communities of genes. Through null-model analyses, community assembly was found to have a stronger deterministic component in the common fraction, whereas the role of stochastic mechanisms was higher for the less abundant portion of the community. Also, reactors that recovered functions after returning to low-organic-loading conditions maintained different α- and β-diversities compared to reactors that had not been disturbed, in terms of both taxonomic and functional genes, showing that resilience based on community function does not necessarily translate into resilience based on community structure. Further, we urge caution when assessing microbial community assembly mechanisms, as results can vary depending on the approach (16S rRNA gene metabarcoding and shotgun metagenomics taxonomic or functional gene community profiling) and whether the focus is on taxa or genes. Not only can genes suggest different dominance patterns of stochastic versus deterministic mechanisms of community assembly compared to taxa, but the fraction of the community driving such assembly mechanisms—common versus rare—can also differ. Finally, this study employed alteration in the substrate feeding scheme as a type of press disturbance. More research is needed on different types of disturbances (e.g., pollutant additions, pH shifts, and temperature changes) within different complex microbial systems at different scales to broadly validate our observations. Additional studies covering different spatial and temporal scales, environmental gradients and types of disturbance could lead to a general framework of how press disturbances alter the structure, function, and assembly mechanisms of microbial communities."
} | 6,420 |
24812591 | PMC4013104 | pmc | 8,344 | {
"abstract": "Horizontal gene transfer plays an essential role in evolution and ecological adaptation, yet this phenomenon has remained controversial, particularly where it occurs between prokaryotes and eukaryotes. There are a handful of reported examples of horizontal gene transfer occurring between prokaryotes and eukaryotes in the literature, with most of these documented cases pertaining to invertebrates and endosymbionts. However, the vast majority of these horizontally transferred genes were either eventually excluded or rapidly became nonfunctional in the recipient genome. In this study, we report the discovery of a horizontal gene transfer from the endosymbiont Wolbachia in the C6/36 cell line derived from the mosquito Aedes albopictus . Moreover, we report that this horizontally transferred gene displayed high transcription level. This finding and the results of further experimentation strongly suggest this gene is functional and has been expressed and translated into a protein in the mosquito host cells.",
"introduction": "Introduction Horizontal gene transfer (HGT) is the exchange of genetic elements between phylogenetically distant and reproductively isolated species or organisms. 1 HGT was traditionally thought to be unlikely between prokaryotes and eukaryotes. 2 However, recent studies have revealed that microbial endosymbiotic DNA fragments can integrate into unrelated eukaryotic organisms. 3 - 5 Although, this process is still considered extremely rare, endosymbiotic gene transfer is believed to be possible because of the intimate symbiotic relationship. 1 Wolbachia is a very important and common bacterial endosymbiont, 6 infecting around 66% of arthropods worldwide including mosquitoes, Drosophila , and beetles. 7 , 8 Wolbachia is able to invade rapidly and spread widely among the host arthropod population, and can manipulate its host’s reproduction and also prevent its host from infecting humans with RNA viruses and other pathogens. 9 - 11 This behavior of Wolbachia , along with the increasing resistance of many arthropods to insecticides, has led to considerable research aimed at using Wolbachia as a biological control for insect-borne diseases. 12 , 13 Research efforts have particularly focused on mosquito-borne diseases including malaria, dengue fever, West Nile fever, lymphatic filariasis, which become a significant concern to global human health and cause the mortality and morbidity of hundreds of thousands of people annually. 14 - 16 A large range of insect hosts can be naturally infected with Wolbachia , including four mosquito species, Culex pipiens , 17 Culex quinquefasciatus , Aedes fluviatilis , 18 and Aedes albopictus , 19 such an important vector of dengue, arguable the most important arboviral diseases of humans, leading to tens of thousands of deaths globally each year. Recent studies have revealed that genetic fragments, ranging in size from single genes to even entire genome, have horizontally transferred from Wolbachia to their insect hosts as illustrated below. Genes originating from Wolbachia were identified and located on the X chromosome of the adzuki bean beetle, Callosobruchus chinensis . 20 Hotopp et al., 3 meanwhile, found nearly the entire Wolbachia genome in Drosophila ananassae Hawaii, while much smaller fragments were found in the other three insects and four nematode species. However, these horizontally transferred genes were all found during tetracycline treatment and their transcription levels were far lower than those of the control gene (act5C) in their recipient, suggesting they could not be expressed or were potentially nonfunctional. 3 Therefore, the mechanism and process of evolution of Wolbachia mediated HGT remains poorly understood. A natural gene-exchange model is required to more fully understand the processes of HGT evolution. C6/36 cell line was derived from Ae. albopictus which can be naturally infected with both Wolbachia wAlbA and wAlbB strains. 19 However, it is accepted that Ae. albopictus C6/36 cell lines lack Wolbachia endosymbionts (based on professor Scott O’Neill’s personal communication). In this study, we report the discovery of a gene WP0273(C6/36) within Wolbachia -uninfected C6/36 cells of Ae. albopictus and it is highly similar to the gene originating from Wolbachia of Cx. quinquefasciatus Pel wPip strain (the putative transcriptional regulator, WP0273). We further revealed the high transcription level of the horizontally transferred gene WP0273(C6/36) and demonstrated that WP0273(C6/36) encodes a protein in the host cell. Taken together, these findings strongly suggest that this represents a natural HGT event from Wolbachia to Ae. albopictus C6/36 cell line, which is involved in a particular functional role.",
"discussion": "Discussion Our results strongely support the conclusion that WP0273(C6/36), which is highly similar to the gene WP0273 originating from Wolbachia , has horizontally transferred from the symbiotic bacteria Wolbachia to the Ae. albopictus C6/36 cell line. C6/36 cell line was isolated from Ae. albopictus (Singh), we speculate this case of HGT event has occurred in the original Ae. albopictus (Singh) before the C6/36 cell line constructed, so we couldn’t detect the WP0273(C6/36) in Wolbachia -free Ae. albopictus strains (China). Most of these genes horizontally transferred between Wolbachia and their hosts (including mosquitoes, Drosophila , filarial nematodes, and beetles) appear to be nonfunctional in the recipient genome. These transferred genes are commonly reported to be transcriptionally inactive, or to contain stop codons or other structural disruptions such as frameshifts and retroelements insertions. 28 In contrast, this study demonstrated that the horizontally transferred gene, WP0273(C6/36), is transcriptionally active in the C6/36 cell line (as measured by comparison of transcriptional levels with a control gene, act5C). Our results indicate that WP0273(C6/36) is translated into a functional protein which participates in protein-protein interaction. In addition, we confirmed (by western blotting analysis) that the endosymbiontic gene WP0273(C6/36) can be expressed into a protein in eukaryotic cells. In general, horizontally transferred genes must meet three requirements to attain a novel functionality. First, the donor DNA has to be delivered into the recipient cell. Second, the target sequences must integrate into the host’s genome. Third, these genes could be expressed in host organisms or cells. 29 Obviously, the horizontally transferred gene WP0273(C6/36) has met all the needs described above, suggesting WP0273(C6/36) has obtained a new function. However, what the new capacity of WP0273(C6/36) is and which role it plays in the host cell remains unknown. Further studies are now required to determine whether WP0273(C6/36) acts to regulate transcription or engenders other effects. Based on our results examining protein structure and motif analysis, we hypothesize that WP0273(C6/36) probably regulates gene expression in C6/36 cells via DNA-binding sites. We will aim to test this hypothesis in future work using Chromatin Immunoprecipitation (Chip) and EMSA-electrophoretic mobility shift assays. Ankyrin repeats consist of 33 amino-acid sequences and the first ANK-containing proteins to be characterized were the yeast cell cycle regulator Swi6/Cdc10 and the Drosophila cell signaling Notch protein. 30 - 32 ANK abounds in arthropod-infecting Wolbachia (for example, 54 in wPip, 23 in wMel, and 34 in wAna 33 ). Ank may play an interesting role in the relationship between host and Wolbachia because ANK is one of the most common protein-protein interaction motifs in nature, involved in many physiological processes including cell signaling, apoptosis, and cell cycle control. 27 Previous results have showed that the protein WP0273(C6/36) could interact with the protein RPL39, indicating it is worthy of further study as there are any other features of WP0273(C6/36). HGT is a crucial driving force in bacterial evolution, with a remarkable impact on pathogenicity and antibiotic resistance of human associated microbes. 29 , 34 Transfer of the vanA gene from an E. faecium isolate of animal origin to an E. faecium isolate of human origin can occur in the intestines of humans. 35 “Pathogenicity islands” horizontally acquired were major contributors to the virulence of many pathogenic bacteria. 36 , 37 Successful adaptive HGT is traditionally considered beneficial either to hosts or to the transferred genes. 29 However, the mechanisms by which Wolbachia is able to manipulate its insect host (to affect reproductive characters, 11 disrupt pathogen infection 38 , 39 or shorten hosts lifespan 18 ) is still poorly understood. One hypothesis posits that some Wolbachia strains interfere with a range of human pathogens in a manner correlated with the innate immune response in the insect. 18 , 40 Here, we propose that interaction of Wolbachia -host and the regulation of insect behavior by Wolbachia are mediated by the effect of horizontally transferred genes that can obtain a new functionin the recipient. It is unknown whether there are other cases of HGT between Wolbachia and Ae. albopictus C6/36 cell line. Elucidating this will require a large-scale exhaustive search using PCR and specific quantitative PCR detection in the C6/36 cell line. Should additional transcriptionally active genes be detected, then the Ae. albopictus C6/36 cell line will offer an ideal model to study HGT mechanisms and evolutionary processes. It will nevertheless be necessary to identify the biological function of any further transferred genes identified, including WP0273(C6/36). The results of this study provide strong evidences that WP0273(C6/36) discovered in Ae. albopictus C6/36 cells is an event of HGT and it appears to be functional. Importantly, our results highlight a novel mechanism of Wolbachia -host interaction and establish a basis for the further study of HGT between endosymbionts and eukaryotes."
} | 2,527 |
39274645 | PMC11396725 | pmc | 8,347 | {
"abstract": "Today, smart materials are commonly used in various fields of science and technology, such as medicine, electronics, soft robotics, the chemical industry, the automotive field, and many others. Smart polymeric materials hold good promise for the future due to their endless possibilities. This group of advanced materials can be sensitive to changes or the presence of various chemical, physical, and biological stimuli, e.g., light, temperature, pH, magnetic/electric field, pressure, microorganisms, bacteria, viruses, toxic substances, and many others. This review concerns the newest achievements in the area of smart polymeric materials. The recent advances in the designing of stimuli-responsive polymers are described in this paper.",
"conclusion": "8. Conclusions The changing needs of society require the creation of new products that are useful in everyday life. The answer in polymer chemistry is smart materials. The development of this field of polymers allows for better adaptation to the needs. The demonstrated use of these materials confirms that their development can revolutionize many industries. The interest of groups of scientists in the development of this field has already been visible for many years.",
"introduction": "1. Introduction The most common definition of polymers is that they are molecules consisting of repeating units characterized by various properties. This is a very general definition, however, that does not reflect the essence of polymers. The multitude of their types, methods of preparation, and application—topics on which paper can be written. Numerous studies in this area show that polymers offer more possibilities than limitations. They are an integral part of our lives, and it can undoubtedly be said that getting rid of them may cause a major regression in our civilization. Being an element of our development, they are also subject to changes dictated by the environment, society’s requirements, and evolution. Some examples of polymer applications are shown in Figure 1 [ 1 , 2 , 3 ]. The presence of polymers in our lives is undeniable. Being part of our environment, they are also subject to some modifications. One result of this is the creation of new types of polymers and the emergence of a new group—smart (intelligent) polymeric materials [ 4 ]. The beginning of the era of intelligent materials dates back to 1950 when Katchalsky’s group started working on hydrogels [ 5 ]. Since then, interest in stimulus-responsive materials has been constantly growing. This fact is supported by numerous papers published every year. As shown in Figure 2 , in the years 2000–2011, a relatively small number of articles on the subject of smart materials were published (about 2000 articles). In the following years, this value gradually increased. After about 5 years, the number of papers had doubled. The year 2019 can be considered as a breakthrough in the research on novel stimuli-responsive materials. Significantly, from 2019 to 2023, about 6000 publications on this topic have been reported. The increased number of published works show how important searching for advanced materials for technology purposes is. If we want to specify what these intelligent materials are, we can compare them to human intelligence or the recently increasingly mentioned artificial intelligence. In relation to this, intelligence means the ability to recognize, name, and respond appropriately to what is happening around us, solve a given problem, and learn. It is similar to smart polymers [ 4 ]. It defines a specific group of polymers that respond to external environmental factors by changing their physical or chemical parameters, which can be detected as changes in solubility, swelling, hydrophilicity/hydrophobicity, or micellization. This specific answer became the basis for designing materials useful in various industries. These factors may be physical, chemical, or biological in nature [ 5 , 6 ]. Sometimes, and more and more often, the term “multi-stimuli” can be heard, which means the ability of polymers to react to several factors [ 7 ]. Changes caused by a given factor are most often reversible, i.e., when they stop, the polymer begins to return to its original state [ 7 ]. A more detailed breakdown is shown in Figure 3 [ 5 , 7 ]. Smart polymers find a special area in medicine. Their specific features justifying this use are summarized below ( Table 1 ) [ 8 ]. Taking into account the methods of synthesis of smart polymers, several basic ones can be distinguished: ▪ Traditional radical polymerization—conventional, which is characterized by mild reaction conditions and can be used in the presence of most monomers; ▪ Controlled radical polymerization—to which belong: (a) reversible addition-fragmentation chain transfer (RAFT) and (b) atom transfer radical polymerization (ATRP) [ 9 ]. On the other hand, due to their physicochemical form, stimuli-responsive polymers can also be classified into various groups, such as gels, solutions, self-organized clusters, coatings, solid materials, and others [ 5 ]. Another classification of intelligent polymers refers to the working mechanism of these materials. Taking into consideration this aspect, smart polymeric materials can be divided as follows: shape-memory polymers (SMP), self-healing materials, polymeric hydrogels, and other responsive polymers [ 4 ]. The future use of smart materials is primarily based on activities aimed at compliance with the principle of sustainable development. They are probably searching for new sources of natural polymers, modifying them, and looking for new intelligent properties. Although the nature of each material is different, their future seems bright due to the many advantages they present: reversible nature of changes and real-time response, they often respond to various environmental stimuli—expand their applicability to various fields by learning the response mechanisms—their reaction is clearly visible and predictable [ 10 ]. The aim of this review is to introduce the reader to the topic of smart polymers in a very short but clear way. The main idea of this paper is to characterize polymeric materials that show sensitivity to various stimuli. This article primarily describes the classification of smart polymers depending on the type of factor (physical, chemical, biological) to which they are responsive. Moreover, the application of these materials in various areas, as well as some examples of the latest achievements of various types of smart materials, such as hydrogels, shape-memory polymers, self-healing materials, and others, are presented. This review does not focus on a specific type of smart polymer but describes the topic comprehensively. In order to highlight the huge potential of this group of smart materials, their selected advanced applications in medicine, chemical industry, agriculture, and modern technologies are presented."
} | 1,725 |
25232892 | PMC4207535 | pmc | 8,350 | {
"abstract": "Lignin\nbiosynthesis occurs via radical coupling of guaiacyl and\nsyringyl hydroxycinnamyl alcohol monomers (i.e., “monolignols”)\nthrough chemical condensation with the growing lignin polymer. With\neach chain-extension step, monolignols invariably couple at their\nβ-positions, generating chiral centers. Here, we report on activities\nof bacterial glutathione- S -transferase (GST) enzymes\nthat cleave β-aryl ether bonds in lignin dimers that are composed\nof different monomeric units. Our data reveal that these sequence-related\nenzymes from Novosphingobium sp. strain PP1Y, Novosphingobium aromaticivorans strain DSM12444, and Sphingobium sp. strain SYK-6 have conserved functions as\nβ-etherases, catalyzing cleavage of each of the four dimeric\nα-keto-β-aryl ether-linked substrates (i.e., guaiacyl-β-guaiacyl,\nguaiacyl-β-syringyl, syringyl-β-guaiacyl, and syringyl-β-syringyl).\nAlthough each β-etherase cleaves β-guaiacyl and β-syringyl\nsubstrates, we have found that each is stereospecific for a given\nβ-enantiomer in a racemic substrate; LigE and LigP β-etherase\nhomologues exhibited stereospecificity toward β( R )-enantiomers whereas LigF and its homologues exhibited β( S )-stereospecificity. Given the diversity of lignin’s\nmonomeric units and the racemic nature of lignin polymers, we propose\nthat bacterial catabolic pathways have overcome the existence of diverse\nlignin-derived substrates in nature by evolving multiple enzymes with\nbroad substrate specificities. Thus, each bacterial β-etherase\nis able to cleave β-guaiacyl and β-syringyl ether-linked\ncompounds while retaining either β( R )- or β( S )-stereospecificity.",
"introduction": "Introduction Lignin,\na major component of plant cell walls, is a recalcitrant\npolymer composed of monomeric units (i.e., components derived from\nguaiacyl and syringyl monomers), 1 − 3 providing plants with both pathogenic\nresistance and structural integrity. 4 , 5 The β-O-4′-ether\n(hereafter termed β-ether) is the most prevalent type of intermolecular\nbond through which the guaiacyl (monomethoxylated) and syringyl (dimethoxylated)\naromatic units are linked. 6 Thus, the development\nof methodologies for β-ether cleavage and depolymerization of\nthe lignin backbone may reveal novel aspects of catalysis and lead\nto lignin-derived products of high economic value. 7 − 10 The formation of lignin\npolymers by radical coupling of monomeric\nunits generates a racemic product containing both β( R )- and β( S )-ether bonds. Here, we\nreport on enzyme activity with a set of newly analyzed substrates\nfor a group of sequence-related bacterial β-etherases that are\nglutathione- S -transferase (GST) superfamily member\nenzymes, each of which catalyzes cleavage of β-ether bonds that\nare characteristically found in lignin polymers. Specifically, we\nreveal that each of these enzymes has activity with guaiacyl- and\nsyringyl-containing substrates and that each enzyme exhibits stereospecifity\nfor cleavage of either β( R )- or β( S )-ether-linked enantiomers. The bacterium Sphingobium sp. strain SYK-6 possesses\nseveral metabolic enzymes that mediate metabolism of lignin-derived\ncompounds. 11 “Lig enzymes”\nthat act in the proposed β-etherase pathway enable this organism\nto derive monoaromatic growth substrates from β-ether-linked\nα-keto diguaiacyl compounds such as α-(4-O-Me)-guaiacylglycerone-β-(1′-formyl)-guaiacyl\nether (GβG). The β( R )- and β( S )-enantiomers of GβG (Gβ( R )G and Gβ( S )G, Figure 1 A) arise as β-etherase pathway intermediates from the activities\nof nicotinamide adenine dinucleotide (NAD)-dependent Lig dehydrogenases,\nwhich oxidize the corresponding benzylic alcohols to α-ketones. 12 , 13 It has been shown that, using glutathione (GSH) and GβG as\ncosubstrates, the β-etherases (LigE, LigP, and LigF1) cleave\nthis aromatic dimer, 14 − 17 producing vanillin and a GSH-conjugated guaiacyl monomer (Gβ-SG)\nas reaction products. 18 Gβ-SG is\nfurther degraded by LigG (and other enzymes that have not yet been\nidentified), yielding glutathione disulfide (GSSG) and the monoaromatic\ncompound β-deoxy-α-(4- O -Me)-guaiacylglycerone\n(Figure 1 A). 15 Figure 1 β-Etherase\npathway-mediated conversion of β-enantiomers\nof substrates GβG, GβS, SβG, and SβS, in which\nvanillin and syringaldehyde are formed from cleavage of β-guaiacyl\n(in panels A and C) and β-syringyl (in panels B and D) ether-linked\ncompounds. Compound names are displayed below each structure, and\n3-methoxylated (i.e., guaiacyl) and 3,5-dimethoxylated (i.e., syringyl)\nunits are shown in blue and red. Catabolism of (A) Gβ( R )G and Gβ( S )G, as well as (B) Gβ( R )S and Gβ( S )S, yields aromatic monomers\nGβ( S )-SG, Gβ( R )-SG,\nand β-deoxy-α-(4- O -Me)-guaiacylglycerone\nas metabolic intermediates. Catabolism of (C) Sβ( R )G and Sβ( S )G, as well as (D) Sβ( R )S and Sβ( S )S, yields aromatic monomers\nSβ( S )-SG, Sβ( R )-SG,\nand β-deoxy-α-(4- O -Me)-syringylglycerone\nas metabolic intermediates. The racemic nature of the lignin backbone 19 − 22 and the existence of both β( R )- and β( S )-configurations in lignin\nnecessitate the ability to degrade both Gβ( R )G and Gβ( S )G enantiomers (Figure 1 A). In Sphingobium sp. strain SYK-6,\nthis is accomplished via the activities of multiple β-etherases\nwith complementary stereochemical properties. 12 , 15 Sphingobium sp. strain SYK-6 LigE and LigP catalyze\nstereospecific cleavage of Gβ( R )G, and LigF1\nexhibits stereospecificity for the Gβ( S )G enantiomer.\nIn this organism, β-ether cleavage is coupled to GSH-conjugation,\ninversion of β-chirality, and stereoselective formation of Gβ( S )-SG (LigE and LigP) and Gβ( R )-SG\n(LigF1). 18 The existence of guaiacyl\nand syringyl units in the lignin polymers\nof all land plants other than softwoods also necessitates the existence\nof enzymes that will cleave β-ether-linked units of different\nsubunit composition (i.e., guaiacyl-β-guaiacyl (GβG),\nguaiacyl-β-syringyl (GβS), syringyl-β-guaiacyl (SβG),\nand syringyl-β-syringyl (SβS); Figure 1 ). Although the activities of Sphingobium sp. strain SYK-6 β-etherases have been shown to contribute\nto the stereospecific and stereoselective degradation of model compounds\ncontaining guaiacyl units, such as Gβ( R )G and\nGβ( S )G, 15 , 18 , 23 , 24 the role served by Lig enzymes\nin the catabolism of native lignin-derived compounds is largely unknown\nbecause (a) investigation of enzymes that might be involved in this\npathway has been limited to those encoded in the genome of Sphingobium sp. strain SYK-6 and (b) the activities of β-etherase\npathway enzymes have not been tested with the range of β-ether-containing\noligomers composed of guaiacyl and syringyl subunits that are typically\nfound in lignin (Figure 1 B–D). In this work, we reveal the ability of β-etherases from Sphingobium sp. strain SYK-6 (SsLigE, SsLigP, and SsLigF1)\nto cleave model dimeric lignin compounds containing GβS, SβG,\nand SβS β-ether linkages, in addition to the previously\nreported GβG substrate. 15 , 18 Further, we identify\nseveral additional sequence-related proteins with β-etherase\nactivity from Novosphingobium aromaticivorans ( N. aromaticivorans ) DSM12444 (NaLigE, NaLigF1, and NaLigF2)\nand Novosphingobium sp. strain PP1Y (NsLigE). We\ndemonstrate that each enzyme catalyzes cleavage of all four combinations\nof β-ether-linked substrates, GβG (Figure 1 A), GβS (Figure 1 B), SβG\n(Figure 1 C), and SβS (Figure 1 D), where each LigE/LigP β-etherase homologue\nhas the conserved function of degrading β( R )-enantiomers whereas each LigF1/LigF2 β-etherase homologue\nexhibits stereospecificity for the β( S )-enantiomers.\nThus, we show that several bacteria possess β-etherases that\nhave a previously unreported ability to cleave lignin dimers containing\nGβG, GβS, SβG, and SβS β-ether linkages.\nOur results also reveal that each of these enzymes exhibits similar\nstereospecifity to that previously described for the enzymes from Sphingobium sp. strain SYK-6. 15 , 18 , 25 These observations reveal important features of a\nconserved class of bacterial enzymes that have utility in the conversion\nof lignin during either plant biomass processing or the potential\nproduction of valuable compounds from this abundant polymer.",
"discussion": "Discussion Recently, it has been shown that GST superfamily enzymes from Sphingobium sp. strain SYK-6 have the ability to act as\nstereospecific β-etherases using lignin model compounds as substrates. 15 , 18 These so-called Lig β-etherases have been shown to cleave\nlignin dimers composed of guaiacyl monomers. In this study, we investigated\nwhether Lig β-etherases from Sphingobium sp.\nstrain SYK-6 also exhibit enzyme activity with substrates that contain\nsyringyl units, the other major monomeric constituent of lignin. Further,\nwe investigated whether other bacteria possess sequence-related proteins\nwith similar or different substrate or stereospecificities as those\nreported for the Sphingobium sp. strain SYK-6 enzymes. This study reveals for the first time that (a) several species\nof sphingomonads encode glutathione-dependent enzymes that catalyze\ncleavage of β-ether linkages that are found in lignin, (b) each\nLig homologue cleaves guaiacyl-β-guaiacyl, guaiacyl-β-syringyl,\nsyringyl-β-guaiacyl, and syringyl-β-syringyl β-ether-linked\nsubstrates, and (c) with each substrate, LigE/LigP and their homologues\nexhibit β( R )-stereospecificity whereas LigF\nand its homologues have β( S )-ether stereospecificity.\nThese results show that methoxy group ring substitutions on the aromatic\nmonomeric units are not inhibitory to the function of these β-etherase\nenzymes. Rather, sphingomonads use enzymes with active sites that\nare receptive to variably methoxylated rings. Also, these findings\ngive insight into how a set of β-etherase pathway enzymes from\ndifferent species accommodate substrates containing the multiple chiral\ncenters (i.e., at carbons α and β) that exist in the β-ether-linked\nstructures found in lignin. 21 , 22 The NAD-dependent dehydrogenases\noxidize and eliminate the chiral center at carbon α, forming\nα-keto-β( R )- and α-keto-β( S )-enantiomers. Further, the existence of both β( R )- and β( S )-ether enantiomers in\nnature is overcome by the evolution of separate glutathione-dependent\nenzymes with either β( R )- or β( S )-ether-cleaving reaction mechanisms. Our results\npredict that a single organism may contain multiple\nβ( R )-etherases (e.g., SsLigE and SsLigP) or\nnumerous β( S )-etherases (e.g., NsLigF1 and\nNsLigF2), each of which is capable of catalyzing cleavage of GβG,\nGβS, SβG, and SβS enantiomers. In sphingomonads Sphingobium sp. SYK-6, Novosphingobium sp.\nstrain PP1Y, N. aromaticivorans strain DSM12444,\nand another Novosphingobium strain with sequence-related\nhomologues to Lig enzymes, strain B-7 (Figure 6 ), it appears that metabolism of α-keto-β-ether-linked\ncompounds is achieved via catalysis by multiple Lig β-etherases\nwith overlapping function. However, it is possible that variations\nof the pathway may exist in closely related bacteria. For example,\na phylogenetic tree constructed from an alignment of LigE/LigP and\nLigF homologues (Figure 6 ) reveals that five\nLigE/LigP homologues belonging to four sphingomonad strains (Figure 6 , LigE cluster) were more closely related to each\nother than the next 11 sequences identified in the SsLigE BLASTP search\n(Figure 6 , HypGST cluster A). The genome of\neach sphingomonad encodes multiple LigF homologues. Five such sphingomonad\nstrains encode closely related putative LigF enzymes (Figure 6 , LigF cluster) that exhibited sequence dissimilarity\nwith the three nonsphingomonad LigF homologues (Figure 6 , HypGST cluster B), perhaps because the HypGST sequences\nencode different functions. Overall, BLASTP analysis predicts\nthat six sphingomonads (α-Proteobacteria\nof the order Sphingomonadales ) encoded Lig homologues\nthat were aligned in the phylogenetic tree (Figure 6 ). Additional BLASTP searches within the genomes of Sphingobium sp. strain SYK-6 and Novosphingobium strains B-7, PP1Y, and DSM12444, each of which had multiple sequences\nin the phylogenetic tree, revealed that each organism encoded both\nthe LigE homologue needed for β( R )-enantiomer\ndegradation, and the LigF homologue required for catabolism of β( S )-enantiomers. Of these, only Sphingobium sp. strain SYK-6 encoded multiple β( R )-specific\n(SsLigE and SsLigP) and multiple β( S )-specific\nenzymes (SsLigF1 and SsLigF2). However, Novosphingobium sp. strain PP1Y, Novosphingobium sp. strain B-7,\nand N. aromaticivorans strain DSM12444 were each\nfound to encode a single LigE homologue and multiple sequences with\nLigF homology. Also, all four sphingomonad strains additionally encode\nmultiple NAD-dependent dehydrogenases that catalyze the formation\nof the α-ketones required for β-ether cleavage activity. The fifth sphingomonad that encodes putative β-etherases, Sphingomonas wittichii ( Sm. wittichii )\nstrain RW1, had a single LigE homologue (Figure 6 , HypGST cluster A) that, based on sequence analysis, is more similar\nto RpHypGST (which had a shorter sequence and did not exhibit β-etherase\nactivity) than to the confirmed β-etherases in the LigE cluster.\nFurther, the Sm. wittichii genome did not encode\na protein related to those that have β( S )-etherase\nactivity or putative NAD-dependent Lig dehydrogenase activity, suggesting\nthat the LigE homologue in Sm. wittichii does not\nencode a function related to β-ether catabolism. Another sphingomonad, Sphingobium xenophagum ( Sb. xenophagum ),\nencoded three homologues with potential β( S )-specific activity (SxLigF1, SxLigF2, and SxLigF3), but did not\nencode a LigE homologue. Given the high sequence similarity to enzymes\nwith demonstrated β( S )-etherase activity, it\nis possible that Sb. xenophagum carries out β( S )-enantiomer catabolism with its various LigF homologues\nbut uses alternative metabolic pathways for the degradation of β( R )-enantiomers. Thirteen of the thirty-one sequences\nin the phylogenetic tree (Figure 6 ) are derived\nfrom nonsphingomonads, one from each\nof α- (of the order Rhodospirillales ), β-,\nγ-, and δ-Proteobacteria, and nine from α-Proteobacteria\n(of the order Rhizobiales ). Amorphus coralii ( A. coralii ) was the only nonsphingomonad that\nencoded both a LigE- and a LigF-like protein. However, unlike in Novosphingobium sp. strains B-7 and PP1Y, N. aromaticivorans strain DSM12444, and Sphingobium sp. strain SYK-6,\nthe A. coralii genome encoded no sequences with homology\nto the NAD-dependent Lig dehydrogenases, suggesting that homologues\nfrom A. coralii are HypGSTs with alternative functions\nto those of the Lig β-etherases. Further, the A. coralii LigE homologue clustered with the other homologues with shorter\nsequences that we predict not to have β-etherase activity (Figure 6 , HypGST cluster A). The genomes of Glaciecola\npolaris and Variovorax paradoxus EPS, each\nencode a single LigF-like sequence (Figure 6 , HypGST cluster B), but did not encode homologues of any of the\nother essential β-etherase pathway enzymes. We therefore propose\nthat the HypGST proteins in clusters A and B do not have activity\nas β-etherases with the lignin compounds used in this study. Given that each of the LigE/LigP enzymes that we tested catalyzed\nβ( R )-ether cleavage, whereas each LigF enzyme\nexhibited β( S )-stereospecificity, we propose\nthat the β-etherase pathway functions similarly in Novosphingobium sp. strains B-7 and PP1Y, N. aromaticivorans strain\nDSM12444, and Sphingobium sp. strain SYK-6. These\norganisms appear to have adapted to the racemic nature of lignin by\nevolving multiple glutathione-dependent enzymes with complementary\nβ-etherase stereospecificities. It will be intriguing to learn\nif the functions of the β-etherase pathway are unique to the\nsphingomonads as the availability of additional genome sequences pave\nthe way for future studies of lignin catabolism in other bacteria."
} | 3,991 |
34908127 | PMC8715522 | pmc | 8,351 | {
"abstract": "Abstract \n Tepidimonas taiwanensis is a moderately thermophilic, Gram-negative, rod-shaped, chemoorganoheterotrophic, motile bacterium. The alkaline protease producing type strain T. taiwanensis LMG 22826 T was recently reported to also be a promising producer of polyhydroxyalkanoates (PHAs)—renewable and biodegradable polymers representing an alternative to conventional plastics. Here, we present its first complete genome sequence which is also the first complete genome sequence of the whole species. The genome consists of a single 2,915,587-bp-long circular chromosome with GC content of 68.75%. Genome annotation identified 2,764 genes in total while 2,634 open reading frames belonged to protein-coding genes. Although functional annotation of the genome and division of genes into Clusters of Orthologous Groups (COGs) revealed a relatively high number of 694 genes with unknown function or unknown COG, the majority of genes were assigned a function. Most of the genes, 406 in total, were involved in energy production and conversion, and amino acid transport and metabolism. Moreover, particular key genes involved in the metabolism of PHA were identified. Knowledge of the genome in connection with the recently reported ability to produce bioplastics from the waste stream of wine production makes T. taiwanensis LMG 22826 T , an ideal candidate for further genome engineering as a bacterium with high biotechnological potential.",
"introduction": "Introduction The majority of current plastics, for example, polyethylene, polyvinyl chloride, polystyrene, and nylon, are made from petroleum. Although their production is cheap, the environmental burden and the resources for their production will be depleted in the future. Therefore, a search for alternatives is needed. A solution has re-emerged in bio-based plastics ( Kawashima et al. 2019 ). A promising group of bioplastics is now presented by polyesters of hydroxyalkanoic acids, that is, polyhydroxyalkanoates (PHA). These environmentally friendly alternatives to petroleum-based polymers are accumulated naturally by numerous prokaryotic microorganisms ( Muhammadi et al. 2015 ; Sabapathy et al. 2020 ). Unfortunately, less than 1% of the total plastic production comes from the bioplastics industry ( Shogren et al. 2019 ). The main obstacle preventing wider utilization of PHA in viable industrial processes is the cost of the carbon resources and the cost of the fermentation and downstream processing. A promising strategy which might help to reduce the cost of PHA is the use of inexpensive or waste carbon substrates that do not compete with the human food chain ( Koller 2018 ) as well as the employment of extremophilic microorganisms ( Obruca et al. 2018 ). Although some pivotal work has been completed and the fact that microorganisms use PHA to store unused energy and carbon in the cytoplasm in the form of intracellular granules is known ( Obruca et al. 2018 ), additional knowledge that can be mined from various genomes of PHA producers is of high importance. The type strain Tepidimonas taiwanensis LGM 22826 T (=BCRC 17406 T , I1-1 T ) is a thermophilic, Gram-negative bacterium that was isolated from a hot spring in the Pingtung area in southern Taiwan ( Chen et al. 2006 ). The rod-shaped cells are motile via a single polar flagellum. The bacterium was originally investigated for its strong alkaline protease activity, which is usable in different industries ( Gupta et al. 2002 ). Nevertheless, other important features of the strain remained hidden, which may be due to the missing high-quality complete genome assembly and functional annotation of the genome. Only recently was its ability to utilize glucose and fructose to produce PHA reported ( Kourilova et al. 2021 ). As it is a thermophile, PHA production takes place within the temperature range 45–55 °C, which reduces the risk of microbial contamination. Therefore, the strain presents an ideal organism for utilization under unsterile conditions, known as the “next-generation industrial biotechnology” concept ( Chen and Jiang 2018 ). In this article, we present its first high-quality complete genome sequence. We annotated the genome, identified key genes in PHA metabolism and in coding extracellular proteases, and searched for prophage DNA and CRISPR arrays.",
"discussion": "Results and Discussion Genome Assembly and Properties The complete genome sequence of T. taiwanensis LMG 22826 T was reconstructed using more than 415,000 Oxford Nanopore Technologies (ONT) reads with average length of 17 kb and polished by an additional 2.3 million high-quality (average Phred score Q ≈ 35) Illumina PE reads. The overall coverage of the final assembly consisting of a single circular chromosome was 2785x. The genome has been deposited at the DDBJ/EMBL/GenBank under accession No CP083911.1 . More than 2.9 Mb long, the genome of T. taiwanensis consists of 2,764 genes, some of them organized into 569 predicted operons that comprise two or more structural genes. From 2,700 coding genes, 66 were marked as pseudogenes, which in most cases are made of incomplete gene sequences according to NCBI Prokaryotic Genome Annotation Pipeline (PGAP). All rRNA genes are present in three copies and 16S rRNA copies have similarity >99%. Further analysis of tRNAs encoded in the genome with tRNA-scanSE ( Chan and Lowe 2019 ) revealed the differences between numbers of different isoaceptor tRNAs which can correlate with codon usage bias as has been reported ( Rocha 2004 ). For example, three of four possible types of alanine amino acid isoaceptors are encoded in the genome in the ratio 1:1:3, so the abundance of the codon corresponding to the third isoaceptor is expected to be higher. In addition, the tRNA analysis revealed a high number, precisely 42, of tRNA isoaceptors that can be affected by relatively high GC content ( Kanaya et al. 1999 ), in this case, almost 69%. All genome features of the T. taiwanensis genome are summarized in Table 1 . Using the complete genome sequence, Tepidimonas thermarum was found to be the closest species to T. taiwanensis . Whole-genome sequence-based phylogeny of the ten most closely related species is available under supplementary figure S1, Supplementary Material online. Genome similarity of T. taiwanensis to these ten species expressed as digital DNA to DNA hybridization reached values from 19.8% to 25.6%. Table 1 Genomic Features of Tepidimonas taiwanensis LMG 22826 T Feature Chromosome Length [bp] 2,915,587 GC content [%] 68.75 Genes 2,764 Operons 569 CDSs 2,700 Pseudogenes 66 ncRNAs 3 rRNAs (5S, 16S, 23S) 3, 3, 3 tRNAs 52 Functional Annotation Protein-coding sequences (CDSs) were divided into 21 categories according to Cluster of Orthologous Groups (COGs). The most abundant known gene function is the “Energy production and conversion group of genes (C)” with 7.59% from all CDSs. Additionally, many genes belong to “Amino acid transport and metabolism group (E),” which makes up 7.44%. The high number of genes in these two groups corresponded to housekeeping functions of cells but was also related to industrially utilizable features, for example, the ability to produce PHA. Its production by the strain T. taiwanensis LMG 22826 T was proved recently, and the presence of phaC gene was confirmed by PCR ( Kourilova et al. 2021 ). Here, we identified loci of phaC (LCC91_05560) and neighboring phaR (LCC91_04215) genes that are necessary for PHA production as well as a locus of phaZ (LCC91_05500) coding PHA depolymerase. Unfortunately, almost 17% of coding genes were assigned to “Unknown function (S)” category and 8.85% genes were not recognized at all. All categories with gene counts are available under supplementary table S1, Supplementary Material online. The arrangement of all genes in the T. taiwanensis LMG 22826 T genome is shown in figure 1 where every COG and every type of RNA is distinguished by a different color. Fig. 1. A genome map of the T. taiwanensis LMG 22826 T . The first two outermost circles represent CDS on the forward and backward strands, respectively. The next two circles contain pseudogenes on the forward and backward strands. Colors represent the functional classification of a COG. The fifth and the sixth circles consist of various types of RNA genes. Two inner circles show GC skew and GC content, respectively. Although the strain was reported to produce extracellular alkaline proteases, due to the lack of genome sequence the enzymes were never identified. KEGG searches revealed 21 orthologues for alkaline proteases, see supplementary table S2, Supplementary Material online. Three of them, sppA (LCC91_01535), degP/htrA (LCC91_09805), and prpL (LCC91_10460), coded for extracellular proteases. Moreover, their predicted molecular weights 36.4, 52.8, and 69.9 kDa matched experimental evidence provided by zymography ( Chen et al. 2006 ). These enzymes might be of great industrial importance. For example, the optimum enzymatic activity of protease IV coded by sppA was reported to be at pH 10 and temperature of 45 °C ( Engel et al. 1998 ), which makes this enzyme utilizable in the detergent industry for production of washing powder. The genome was searched for CRISPR arrays and, as a consensus from three prediction methods, four large arrays were reported in the T. taiwanensis LMG 22826 T genome. The largest array consisted of 42 spacers and was 2,590 bp long. Moreover, cas genes, such as cas1 or cas2 , whose proteins are responsible for spacers acquisition into CRISPR arrays ( Yosef et al. 2012 ), were found in the genome. Unfortunately, neither of these genes was the gene that encodes the Cas9 protein well known for its high utilization in genetic engineering. The summary of CRISPR arrays is included under supplementary in table S3, Supplementary Material online. Finally, the presence of prophage DNA was checked. PHASTER found three prophages: two of them were labeled as incomplete and one as intact prophage. The sequence that was labeled as intact prophage consists of 70 proteins, and 46 of them match to phage proteins such as phage tail protein or phage virion protein. Eight of these proteins correspond to Escherichia phage vB_EcoM_ECO1230-10 , which has not been reported as a phage able to survive life conditions of thermophilic bacteria such as T. taiwanensis . Although Prophage Hunter did not label any of the phages as active, the previously mentioned phage sequence achieved the highest score and corresponded to Acidithiobacillus phage Aca ML1, which has been found in thermophilic, acidophilic bacterium Acidithiobacillus caldus ( Covarrubias et al. 2018 ), so it is possible to presume this phage has the ability to survive the life conditions of T. taiwanensis . Overall, the reliable statement of whether the prophage is active would need further analysis. The summary table of present prophage DNA from PHASTER tool is available under supplementary table S4, Supplementary Material online."
} | 2,761 |
35497439 | PMC9049229 | pmc | 8,352 | {
"abstract": "In this study, the anti-biofouling effect of a thin film nanocomposite (TFN) membrane with a functionalized-carbon-nanotube-blended polymeric support layer was analyzed to determine the applicability of this membrane for the pressure-retarded osmosis (PRO) process. The anti-biofouling property of TFN membranes for the PRO process was characterized by SEM, FTIR, and AFM, as well as contact angle measurements and zeta potential analysis of the bottom side of the support layer. The anti-biofouling effect of the fabricated membrane for the PRO process was analyzed by bacterial attachment tests on the bottom surface of the support layer and biofouling tests in a cross-flow operation system in the PRO mode (AL-DS). The TFN membrane with 0.5 wt% fCNTs exhibited enhanced anti-biofouling properties of the bottom surface of the support layer compared to the bare TFC membrane due to the low roughness, high negative surface charge, and hydrophilicity. Compared to the bare TFC membrane, the support layer of the fCNT0.5-TFN membrane exhibited a 35% decrease in bacterial attachment. In a laboratory-scale biofouling test, the water flux of the fCNT0.5-TFN membrane was ∼10% less than that of the bare TFC membrane in the PRO mode.",
"conclusion": "4. Conclusions The TFN membrane with 0.5 wt% of fCNTs blended in the support layer exhibited enhanced membrane performance and anti-biofouling property in the PRO mode via the change in its bottom surface characteristic of the support layer. Compared to bare TFC membrane (CNT 0 wt%), the blended membrane exhibited an increase in the water flux, decrease in the roughness, increase in the negative surface charge, lower hydrophobic property, and decrease in the bacterial attachment. In the laboratory-scale biofouling test, compared to the bare TFC membrane, the fCNT0.5-TFN membrane exhibited a low flux decline in the PRO mode. All things considered, CNT-composite TFN membranes demonstrate immense potential of improving the performance of membrane in the PRO process by overcoming the current biofouling problem via the decrease in the bacterial attachment and inhibition of bacterial growth as well as the increase in water flux. Furthermore, fCNT-blended membranes exhibited better performance in the anti-biofouling property via the additional functionalization of fCNTs with bio-toxic material.",
"introduction": "1. Introduction As a water crisis and energy shortage have become serious issues around the world, membrane technology has become increasingly popular compared to other water treatment technologies due to less chemical and thermal consumption as well as low energy consumption. Pressure-retarded osmosis (PRO) is one of the membrane technologies used for producing water and energy by the osmotic pressure difference between two solutions. 1 As saline water ( i.e. , draw solution) draws feed solution ( i.e. , wastewater) through a membrane, the volume of the draw solution increases, which leads to the movement of the turbine to produce electricity and water. 2,3 In the PRO process, a thin film composite (TFC) membrane, comprising a polyamide active layer on a support layer, has been widely used for applications. 4 This TFC membrane is limited by serious fouling due to the membrane orientation. The orientation of the TFC membrane in the PRO process is known as the PRO mode (AL-DS), where the active layer contacts the draw solution, while the support layer contacts the feed solution. Due to the membrane orientation, foulants in the feed solution are easily deposited on the bottom surface of the support layer and within the support layer structure. 3,5 Due to the high concentrations of organic matter and bacteria in the feed solution, biofouling on membranes leads to a significant reduction in the system performance via the decrease in the membrane water flux and degradation of the membrane in the PRO process. 6 The development stages of biofouling on membranes include the initial attachment of bacteria and formation of biofilms with the secretion of extracellular polymers. Therefore, it is imperative to modify membranes for the purpose of minimizing bacterial attachment and bacterial growth. The modification of membranes to reduce biofouling is categorized into anti-adhesion and antimicrobial approaches. The anti-adhesion approaches involve the control of the initial attachment of the microorganisms to inhibit the biofilm development via the change in the membrane surface properties such as hydrophilicity, 7 surface charge, 8,9 and surface roughness. 10 Compared to hydrophobic membranes, hydrophilic membranes would exhibit less bacterial attachment due to the low interaction between the membrane surface and bacteria. In addition, the low roughness reduces the contact between the membrane surface and bacterium. The negatively charged surface of the membrane repels negatively charged bacteria around neural pH via electrostatic repulsive forces. 11 By contrast, the antimicrobial approaches involve the prevention of the biofilm growth on the membrane via the use of biocidal agents that can disrupt and kill the attached microorganisms. 12 Recently, carbon nanotubes (CNTs) have been applied for the synthesis of advanced membranes for use in water purification and desalination. 13 CNTs can change membrane properties such as hydrophilicity, roughness, and surface charge. These property changes can improve the membrane performance, such as the water flux, thermal stability, mechanical stability, and antifouling. 14–16 In addition, studies have reported the potential antimicrobial effect of CNTs. 17–20 Antimicrobial mechanisms are hypothesized to occur via oxidative stress, physical cell damage, and intracellular metabolic pathway disruption. 19,21 Previous studies have reported the incorporation of CNTs in the active layer of TFC membranes to enhance separation performance and anti-biofouling effect on the active layer for the Forward Osmosis (FO) or Reverse Osmosis (RO). 17–20 In this purpose, the functionalized CNT was embedded in the polyamide thin film during interfacial polymerization or additional coating. 22–24 Nevertheless, few studies have described carbon nanotubes employed in the TFC membrane for PRO process. In this study, the CNTs incorporated in the support layer of TFN membranes was analyzed to determine their anti-biofouling ability in the PRO mode for potential applications of the PRO process. In addition, the relationship between the anti-biofouling effect and characterization of the bottom surface of the support layer on the TFN membrane was investigated by scanning electron microscopy (SEM) and atomic force microscopy (AFM), as well as contact angle measurements and zeta potential analysis. In particular, this study played a significant role in the analysis of the anti-biofouling effect of the fCNT-TFN membrane in a cross-flow system by biofouling tests in the PRO mode (AL-DS). To the best of our knowledge, it is the first paper of reporting characterization of the bottom surface of support layer on the fCNT-TFN membrane and biofouling test of the fCNT-TFN membrane using the feed solution containing microorganisms for application in PRO process ( Fig. 1 ). Fig. 1 Anti-biofouling properties of the TFN membrane with fCNT blended in the support layer for the PRO process.",
"discussion": "3. Results and discussion 3.1 Characterization of fabricated membranes 3.1.1 FTIR analysis of functionalized CNT and fabricated membranes Functional groups of functionalized MWCNTs and the 0.5 wt% functionalized MWCNT-blended support layer were examined by FTIR (Fig. S1 † ). Functionalized CNTs exhibited peaks of –OH (∼3440 cm −1 ) and (–COOH) (∼1380 cm −1 ), corresponding to bare CNTs. 26–28 The carboxylic and hydroxyl groups of the functionalized CNTs led to the increased dispersion of CNTs in the polymeric support layer via hydrogen bonding with the PES sulfonic groups. 14,29 Compared to the bare TFC membrane, the fCNT composite TFN membrane, comprising the functionalized CNTs in the support layer, exhibited increased peak intensities. 3.1.2 Morphology of fabricated membranes with SEM Fig. S2 † shows the cross-sectional image of the fabricated TFN membranes and the surface image of the bottom side of the support layer. In the cross-sectional image of the bare TFC membrane (Fig. S2a † ) and fCNTs composite TFN membranes (Fig. S2b–d † ), the polymer support exhibited an asymmetric porous structure with a polyamide active layer on the top of the surface. Compared to the bare TFC membrane, TFN membranes exhibited a slightly straight finger-like structure. In the surface SEM image, the bottom side of fCNT composite support layers exhibited larger pores (Fig. S2f–h † ) than those observed on the bare PES support layer (Fig. S2e † ). 30 Pore size was affected by the concentration of fCNTs, which were deposited on the bottom side of the support layer (Fig. S3 † ). The addition of 0.5 wt% of fCNTs led to the formation of larger pores on the bottom of the support layer than bare PES support layer due to the hydrophilic functional groups of fCNTs and hydrogen bonding with the PES sulfonic groups. At an fCNT concentration greater than 0.5 wt%, the high viscosity of the polymer solution led to the delayed rate of phase separation, leading to small pores and aggregation. 3.1.3 AFM analysis of the roughness of the fabricated TFC membranes The addition of fCNT in the support layer strongly affected the roughness of the bottom surface of the support layer of the TFN membranes. The roughness parameters ( R a , R q ) of the membrane bottom surface initially decreased with the addition of up to 0.5 wt% fCNTs and then increased with the further addition of the fCNT content of the support layer ( Table 2 ). Fig. S4 † shows the three-dimensional surface images of the bottom side of the support layer with different concentrations of fCNTs at a scan size of 8 μm × 8 μm. Although the bottom surface of support layer has more porous structure than top surface of support layer, the bottom surface microstructure of the support layer became smoother with the increase in the concentration of fCNTs in the casting solution of up to 0.5 wt% because fCNT located on the bottom side of the support layer led to a slower counter-diffusion velocity of the solvent and non-solvent, resulting in a smoother membrane surface. 31 The further increase in the fCNT concentration (1 wt%, 2 wt%) of the casting solution led to a rough membrane surface by the aggregation of fCNTs in the support layer, as the high amount of fCNT was excessive for electrostatic interactions and good compatibility with the PES support layer. 32,33 Surface roughness parameters of the support layer of the fabricated TFN membranes Membrane Roughness \n R \n a (nm) \n R \n q (nm) \n R \n z (nm) CNT0 49.976 62.17 340.394 CNT0.5 23.874 31.3 277.413 CNT1 28.43 37.76 334.207 CNT2 31.139 40.708 439.176 3.1.4 Hydrophilicity of the fabricated TFN membranes The hydrophilicity of the support layer on the as-synthesized TFN membranes was determined by static water contact angle measurements ( Fig. 2 ). With the increase in the concentration of fCNTs deposited on the bottom surface of support layer, the hydrophilicity of the bottom surface of the membrane increased due to presence of hydroxyl and carboxylic groups. 14,16,28 In addition, wettability is directly associated to the change in the microstructure, and the good dispersion of fCNTs could increase hydrophilicity 34 fCNTs were well dispersed on the bottom surface of the support layer at a concentration of 0.5 wt% and led to high hydrophilicity. The further increase in the fCNT concentration (1 wt% and 2 wt%) led to the aggregation of fCNTs on the membrane and decrease in their dispersion in the support layer. This tendency reduced the hydrophilicity of the bottom side of the support layer at an fCNT concentration greater than 0.5 wt%. Fig. 2 Contact angle measurement of the support layer of the fabricated membrane. 3.1.5 Surface charge of the support layer The surface charge of the bottom side of the support layer was examined to determine the anti-biofouling property using different fCNT contents ( Fig. 3 ). The zeta potential of the bottom surface of the bare support layer (CNT0) was −38.1 mV, which changed to −63 mV (CNT0.5) by blending with fCNTs due to the negative charges of the hydroxyl and carboxylic groups after both groups ionized in 10 mM NaCl solution. 28 At an fCNT content of greater than 0.5 wt%, a high amount of fCNTs led to the increase in the surface charge of the zeta potential as the high fCNT content led to their increased aggregation and decreased dispersion in the support layer. For the most part, the zeta potential values of both Gram positive and Gram negative bacteria were negative (∼−20 mV). 35,36 The negatively charged membrane surfaces repel the bacterial surface, which is also negatively charged, via electrostatic repulsions. 37 Therefore, the high negative surface charge of the support layer (CNT0.5) might increase the resistance of negatively charged foulants and bacteria. Fig. 3 Surface zeta potential of the support layer. 3.2 Water flux performance \n Fig. 4 shows the water flux and RSF of the bare TFC membrane/fCNT composite TFN membranes. In the FO mode, the water flux of the CNT0.5 membrane slightly increased from 7.09 L m −2 h −1 (LMH) to 7.32 LMH compared with that observed for CNT0 membranes. In the PRO mode, the water flux of CNT0.5 increased by ∼30% from 11.48 LMH (CNT0 membrane) to 16.06 LMH due to the increase of the larger pores in the support layer than bare PES support layer. 14,28 With the increase in the fCNT content of the membrane to greater than 0.5 wt%, the high viscosity of the casting solution led to the delayed phase separation and afforded smaller pores and lower water flux. Generally, the water flux in the PRO mode is more rapid than that in the FO mode due to the higher effective driving force, which is caused by membrane orientation. 38 In the PRO mode, as water passes through the support layer to the draw solution, the highly porous support layer with the blended fCNTs can increase the effective osmotic difference (Δπ eff ) and lead to the notable increase in the water flux. On the other hand, in the FO mode, as water passes through the active layer to the draw solution, there is a lower increase in the effective osmotic difference. Hence, in the PRO process, it is effective to blend 0.5 wt% fCNTs in the support layer of the TFN membrane due to the highest water flux. Fig. 4 Water flux/RSF of bare TFC membranes and fCNT composite TFN membranes (0.5 wt%, 1 wt%, 2 wt%) in the FO mode (AL-FS)/PRO mode (AL-DS). RSF is a membrane selectivity parameter in osmotically driven membrane processes. In the FO mode, with the increase in the water flux, RSF increased from 5.45 gMH (CNT0) to 6.31 gMH (CNT0.5). In the PRO mode, with the increase in the water flux, RSF increased from 6.02 gMH (CNT0) to 10.17 gMH (CNT0.5). RSF can increase due to the high effective area of the support layer with larger pores than bare PES support layer. With the increase in the pore size, salt can diffuse to the feed side instead of staying within the internal space between the active and support layers. 39,40 3.3 Analysis of bacterial attachment on the support layer \n Fig. 5 shows the attachment of bacteria on the bottom surface of the support layer of TFN membranes. There is an increased number of bacteria attached on the bottom side of the support layer due to the facile attachment of bacteria on the porous support on the surface and inside the membrane structure. 41 The number of E. coli attached to the support layer blended with 0.5 wt% fCNT (CNT0.5) was 35% less than that attached on the bare support layer of the TFC membrane (CNT0). At an fCNT concentration of greater than 0.5 wt%, the increased FCNT content of the support layer led to the slight increase in the number of attached bacteria. The reduction of bacterial attachment was considerably affected by the bottom surface characteristics of negative charge, low roughness, and hydrophilicity. The highest negative surface charge ( Fig. 3 ) of the CNT0.5 membrane might reduce bacteria attachment via electrostatic repulsions with the negatively charged bacterial surface. 37 In addition, the highest hydrophilicity ( Fig. 2 ) and lowest roughness ( Table 2 ) of the fCNT composite support layer might have reduced anti-adhesion on the membrane via the limitation of the contact between the membrane surface and bacterium. 31 Furthermore, the antibacterial property of CNT might inhibit bacterial growth. A previous study has already explained the high potential and mechanisms of the antibacterial property of CNTs such as the intracellular metabolic pathway disruption, oxidative stress, and physical stress. 19 Fig. 5 Bacterial attachment on the support layer of the fabricated TFN membrane at different fCNT concentrations. 3.4 Biofouling test of the fabricated TFN membrane in the PRO mode Biofouling test in the cross-flow operation mode was performed to understand the anti-biofouling effect of the 0.5 wt% fCNT composite TFN membrane for the PRO process. The biofouling formation between the CNT0.5 membrane and CNT0 membrane ( i.e. , bare TFC membrane) was compared in the PRO mode. Dynamic cross-flow possibly led to the increased bacteria detection by shear stress and to the low formation of biofouling on the bottom surface of the support layer in the fCNT0.5-TFN and bare TFC membranes. Even so, compared to the bare TFC membrane, the fCNT0.5-TFN membrane exhibited a low flux decrease ( Fig. 6 ). First, bacteria in the feed solution entered the porous support layer and inevitably reduced the water flux until 500 min. After 500 min, the water flux of the membrane fabricated with 0.5 wt% of fCNTs exhibited a low decrease. After 2000 min, compared to the bare TFC membrane (CNT0), the 0.5 wt% fCNT membrane exhibited a low decline of ∼10%. This relatively less reduction of water flux might have several causes: high initial water flux of 0.5 wt% of fCNTs membrane, the reduced accumulation and growth of bacteria on the bottom surface of the support layer of the fCNT0.5-TFN membrane caused by the modified characteristic of the bottom surface – the high electrostatic repulsive force with surface charge, low roughness, and hydrophilicity of fCNT on the bottom side of the support layer. Fig. 6 Comparison of the biofouling flux decline of the bare TFC membrane and CNT0.5 TFN membrane in the PRO mode. The reduced accumulation and growth of biomass on the bottom of the TFN support layer was observed by CLSM imaging. A sparser biofilm was developed on the fCNT0.5-TFN membrane compared to the bare TFC membrane ( Fig. 7 ). Fluorescence images revealed green and red spots, corresponding to live bacteria and dead bacteria, respectively. As the biofouling test was processed in 2000 min, bacteria grew and died continuously, and it was difficult to examine the dead bacteria due to the membrane antibacterial property. Fig. 7(a and b) shows the biofilm formation on the bottom surface of the bare TFC membrane support layer: Green and red spots were observed almost everywhere, indicating that bacteria are easily attached on the bare TFC membrane support layer and grow and form a biofilm. In contrast, live or dead bacteria ( E. coli ) were randomly located on the bottom of the fCNT0.5 TFN support layer ( Fig. 7c and d ). Fig. 7 Orthogonal view of the confocal laser scanning microscopy (CLSM) image of the biofouled layers on CNT0/CNT0.5 membrane in the PRO mode: (a and b) CNT0 membrane/(c and d) CNT0.5 membrane and (a and c) live bacteria on membrane/(b and d) dead bacteria on the membrane."
} | 4,949 |
38667951 | PMC11051020 | pmc | 8,355 | {
"abstract": "The crucial functional arbuscular mycorrhizal fungi (AMF) and diazotrophs play pivotal roles in nutrient cycling during vegetation restoration. However, the impact of managed vegetation restoration strategies on AMF and diazotroph communities remains unclear. In this study, we investigated the community structure and diversity of AMF and diazotrophs in a karst region undergoing managed vegetation restoration from cropland. Soil samples were collected from soils under three vegetation restoration strategies, plantation forest (PF), forage grass (FG), and a mixture of plantation forest and forage grass (FF), along with a control for cropland rotation (CR). The diversity of both AMF and diazotrophs was impacted by managed vegetation restoration. Specifically, the AMF Shannon index was higher in CR and PF compared to FF. Conversely, diazotroph richness was lower in CR, PF, and FG than in FF. Furthermore, both AMF and diazotroph community compositions differed between CR and FF. The relative abundance of AMF taxa, such as Glomus , was lower in FF compared to the other three land-use types, while Racocetra showed the opposite trend. Among diazotroph taxa, the relative abundance of Anabaena , Nostoc , and Rhizobium was higher in FF than in CR. Soil properties such as total potassium, available potassium, pH, and total nitrogen were identified as the main factors influencing AMF and diazotroph diversity. These findings suggest that AMF and diazotroph communities were more sensitive to FF rather than PF and FG after managed vegetation restoration from cropland, despite similar levels of soil nutrients among PF, FG, and FF. Consequently, the integration of diverse economic tree species and forage grasses in mixed plantations notably altered the diversity and species composition of AMF and diazotrophs, primarily through the promotion of biocrust formation and root establishment.",
"conclusion": "5. Conclusions This study expands the understanding of the effects of managed vegetation restoration in vulnerable karst areas on AMF and diazotroph diversity, as well as community composition. Our study revealed that the diversity and community composition of AMF and diazotrophs were more sensitive to the FF strategy compared to PF and FG. The combination restoration strategy (i.e., FF) may promote N accumulation by increasing diazotroph diversity and strengthening the correlation among certain AMF taxa (e.g., Racocetra ) and diazotroph taxa (e.g., Anabaena and Nostoc ), although AMF diversity could be suppressed in the FF. The findings of our study provide valuable new insights for evaluating ecological restoration efforts in fragile karst regions. Given the uncertainty surrounding the long-term effects of various replantation strategies on AMF and diazotroph communities, it is crucial to continue to assess these microbial populations as a key aspect of karst vegetation restoration efforts.",
"introduction": "1. Introduction Soil key functional microorganisms serve as vital indicators for assessing soil quality and maintaining soil fertility. AMF are known to effectively improve nutrient utilization efficiency and regulate the stability of plant community ecosystems [ 1 , 2 , 3 ]. The mutualistic relationship between AMF and plants is particularly significant during vegetation restoration, considering that approximately 80% of plant species engage in mycorrhizal associations [ 3 , 4 ]. Additionally, free-living nitrogen (N) fixation driven by diazotrophs is an essential source of N input, particularly in the absence of symbiotic N fixation by legumes [ 5 , 6 ]. Therefore, comprehending how variations in the community composition, diversity, and abundance of AMF and diazotrophs respond to different vegetation restoration strategies can offer valuable insights into promoting positive vegetation succession and ensuring ecosystem stability [ 7 ]. The communities of AMF and diazotrophs are influenced by a range of biotic and abiotic factors [ 3 , 8 , 9 ]. Vegetation restoration indirectly impacts AMF by altering soil nutrient availability and plant communities, a relationship that has been widely studied [ 7 , 10 , 11 ]. These studies consistently reveal a significant correlation between the structure and function of AMF and diazotroph communities and plant diversity and soil quality. Plant diversity can have a positive, negative, or neutral effect on AMF and diazotroph diversity. These varying impacts may be attributed to a myriad of factors, including the specific plant species involved, the type of soil, and the duration of vegetation recovery [ 3 , 12 , 13 , 14 , 15 ]. Compared to agricultural farming systems, vegetation cover resulting from restoration efforts affects AMF and diazotroph diversity and populations by influencing litter input and root exudates [ 7 , 16 ]. Increased availability of soil nutrients, such as carbon (C) sources, promotes microbial growth, leading to higher abundances of AMF and diazotrophs [ 4 , 6 , 16 ]. Conversely, excessive N fertilizer input has been shown to negatively impact diazotroph abundance by suppressing N 2 fixation activity [ 17 , 18 ]. Furthermore, tillage practices can disrupt hyphae, resulting in lower AMF colonization in croplands [ 19 , 20 ]. Therefore, the effects of vegetation restoration on AMF and diazotroph communities are complex due to differences in soil properties, plant characteristics, and management practices. Understanding the response of AMF and diazotroph diversity and community composition to vegetation restoration, especially in managed systems, is crucial for effective land management. The southwestern region of China boasts the largest karst landscape in the world, an environmentally fragile terrain characterized by shallow soil layers, limited soil nutrients, and a unique dual structure of aboveground and underground features [ 21 , 22 ]. Over the past century, the area has experienced exacerbated rock desertification due to intense tilling activities. In response, various ecological restoration initiatives, such as the “Grain to Green” project, have been implemented [ 21 , 23 ]. A key aspect of restoration efforts lies in managed vegetation restoration, encompassing the strategic planting of economic tree species and forage grasses to restore degraded land in karst areas [ 24 , 25 , 26 ]. Due to the accelerated depletion of soil nutrients caused by unreasonable land-use in the early stages, vegetation restoration efforts in this area were frequently constrained by soil nutrient deficiencies, particularly N [ 27 ]. As a result, enhancing the functional capacity of key microorganisms, such as AMF and diazotrophs, in the vegetation restoration process significantly contributes to the improvement in soil nutrient levels. Many studies have reported on the impact of vegetation restoration on soil nutrient accumulation and microbial communities [ 7 , 16 , 25 , 26 , 28 ]. However, the majority of these investigations have focused on natural vegetation restoration, with limited research conducted on managed vegetation restoration. Previous studies conducted in the karst region have shown that planting economic tree species and elephant grass, compared to cropland, leads to increased C and N stocks and enhances the abundance of total phospholipid fatty acids [ 24 , 26 ]. Moreover, managed vegetation restoration has been found to alleviate microbial C limitation but exacerbate limitations in microbial N and phosphorus (P) [ 28 ]. These findings indicate that managed restoration strategies exert a significant influence on soil nutrient levels and microbial abundance. Nevertheless, despite the recognized significance of AMF and diazotrophs in managed vegetation restoration efforts, the impact of plantations with forage grass and economic tree species on these microbial communities remains unclear to date. To gain a deeper understanding of the changes in AMF and diazotroph characteristics across various managed vegetation restoration strategies, this study aims to assess the community composition and diversity of AMF and diazotrophs in soils under three strategies. We hypothesized that managed vegetation restoration, compared to cropland, would increase the diversity of both AMF and diazotrophs and alter the community composition as well.",
"discussion": "4. Discussion This study investigated the community structure and diversity of AMF and diazotrophs under three managed vegetation restoration strategies. Our findings suggest that those managed restoration strategies, especially the FF approach, have a significant impact on the diversity and community composition of both AMF and diazotrophs. Importantly, this research highlights the distinct responses of key microbial groups to different restoration strategies. This suggests that the specific functions of those microbes within each restoration context should be carefully considered when designing and implementing future vegetation restoration efforts. 4.1. The Effect of Managed Vegetation Restoration on AMF and Diazotroph Diversity Many studies have underscored the importance of changes in soil properties and plant communities in shaping the diversity of AMF and diazotrophs during vegetation restoration [ 7 , 10 , 32 , 33 ]. The findings of our study demonstrate that diazotroph and AMF diversity exhibits distinct responses to managed vegetation restoration. Notably, both AMF and diazotroph diversities are found to be more sensitive to the FF strategy than the PF and FG strategies. A previous study reported higher biomass of biocrusts in the FF restoration strategy compared to other land-use types [ 26 ]. The increased formation and development of biocrusts in the FF may contribute to promoting biological N fixation [ 26 , 34 , 35 ]. Another study in the same area found higher root biomass associated with planting forage grass compared to economic tree species, thereby increasing C resource availability in the soil under forage grass planting conditions [ 24 ]. Therefore, the increased N fixation activity and C exudation observed in the FF strategy could potentially stimulate a greater diversity of diazotroph species, leading to a higher richness of diazotrophs in the FF compared to CR, PF, and FG strategies. Contrary to our initial hypothesis, the AMF Shannon index was lower in the FF treatment compared to CR and PF. Several studies have shown that AMF diversity is correlated with plant diversity [ 4 , 36 ]. In the present study, economic tree species were planted in the PF treatment on downhill positions for 25 years. The presence of various grass and shrub species growing under these trees in the PF led to higher plant diversity compared to the forage grass treatments in FG and FF. Therefore, the increased plant species diversity with grasses and shrubs under PF may contribute to higher AMF diversity in PF than in FF. Additionally, a previous study found an increased saturated water adsorption ratio of biocrust in the combined restoration strategy of FF [ 26 ]. This suggests that more AMF species may be induced in the drier conditions of CR and PF compared to FF. Moreover, the combined restoration strategy involving plantation trees and forage grassland may reduce light availability due to shading, potentially suppressing AMF colonization and subsequently decreasing AMF diversity in FF [ 37 , 38 ]. Those findings indicate that the different changes in AMF and diazotroph diversity to managed vegetation restoration are influenced by soil condition and plant properties. Consequently, it is crucial to consider these factors when designing and implementing vegetation restoration strategies in order to maximize microbial diversity and enhance ecosystem functionality. 4.2. Effect of Managed Vegetation Restoration on AMF and Diazotroph Community Compositions Consistent with our previous research in karst regions for agricultural or natural ecosystems, the genera Glomus and Bradyrhizobium emerged as the dominant genera for AMF and diazotrophs, respectively [ 15 , 16 ]. That suggests that Glomus and Bradyrhizobium play important roles in improving nutrient accumulation during managed vegetation restoration. Both AMF and diazotroph community compositions differed significantly between CR and FF. Glomus typically exhibits a wide range of adaptability to various environmental conditions, including harsh soil conditions [ 39 , 40 , 41 ]. The relative abundance of Glomus was lower in the FF compared to CR, PF, and FG, indicating that FF with low light availability may suppress the growth of Glomus . Conversely, the relative abundance of Racocetra was higher in FF compared to other land-use types, indicating that mixed plantation restoration strategies may influence the distribution of AMF taxa. Regarding diazotroph taxa, FF exhibited a higher relative abundance of Anabaena , Nostoc , and Rhizobium compared to CR. As mentioned previously, FF favors the establishment and development of biological crusts, leading to enhanced C exudation [ 26 ]. That explains the abundant presence of N-fixing Nostoc and Rhizobium in FF, supported by other studies that suggest biological crusts may result in high proportions of N-fixing Nostocales under grassland degradation in the Tibetan Plateau [ 42 ]. In addition, low soil nutrient (e.g., NO 3 − , AP, and AK) conditions may serve as the primary stimulus for the proliferation of cyanobacteria genera like Anabaena and Nostoc . These cyanobacteria have been observed to release nutrients, particularly N, thereby providing additional nutrients for plants. Additionally, Nostoc has been noted to thrive in barren soils, indicating its resilience in nutrient-poor environments [ 43 , 44 ]. We further investigate the relationship between AMF and diazotroph taxa in order to predict and assess nutrient cycling within the soil. The highest relative abundance of Glomus was negatively correlated with diazotroph groups containing Anabaena and other low-relative-abundance taxa, suggesting a competitive relationship between Glomus and Anabaena . In contrast, AMF taxa such as Racocetra and unclassified taxa were positively correlated with diazotroph groups containing Anabaena and Nostoc . A previous study found that strengthening mutualistic associations among certain diazotrophs (e.g., Bradyrhizobium and Azotobacter ) and AMF (e.g., Racocetra ) groups mainly contributes to free-living N fixation [ 6 ]. Consequently, N-fixing Anabaena and Nostoc may rely on Racocetra to enhance N fixation during managed vegetation restoration efforts. 4.3. Implications for Future Managed Vegetation Restoration Consistent with previous studies, managed vegetation restoration increased soil C and N accumulation [ 24 , 26 , 28 ]. The NH 4 + was higher in the PF, FG, and FF than in the CR because managed vegetation restoration biocrusts promoted N fixation but decreased nitrification. One study also found the positive correlation between N availability and diazotrophs [ 31 ], which agrees with the findings of this study that soil N content (e.g., DON and TN) showed significant contributions to various diazotroph diversities. Conversely, higher TP and AP contents were observed in the CR than in other restoration strategies. This suggests that excessive P availability may have an inhibitory effect on nitrogen-fixing microorganisms. Previous studies have shown that soil P levels are negatively correlated with N-fixing microorganisms and plant growth, as high P levels can limit plant growth and N assimilation [ 45 ]. Additionally, soil properties (e.g., NH 4 + , TP, and AP) were similar in the FF compared to PF and FG, likely due to the relatively short duration (20 years) of the different restoration strategies. However, only FF, but not PF and FG, had an effect on the diversity and certain taxa of AMF and diazotrophs, possibly because the combination of forest and grass in the FF could benefit the formation of biocrusts [ 26 ]. As aforementioned, biocrusts in FF may enhance biological N fixation by stimulating diazotrophs. Furthermore, variations in AMF diversity were primarily explained by TK, soil pH, SOC, and TN. TK is an essential nutrient for promoting plant growth, which indirectly affects AMF diversity during managed vegetation restoration [ 46 , 47 ]. In summary, our study demonstrates that the FF strategy offers advantages in terms of improving diazotroph diversity while also resulting in a decline in AMF diversity. This is likely due to the rich biocrust biomass, root biomass, and C exudates present in the FF strategy. Future efforts to restore ecological balance in vulnerable karst areas may benefit from considering the inoculation of AMF and diazotrophs in replanting strategies."
} | 4,194 |
36013451 | PMC9410007 | pmc | 8,357 | {
"abstract": "Nitrogen (N) is a gas and the fifth most abundant element naturally found in the atmosphere. N’s role in agriculture and plant metabolism has been widely investigated for decades, and extensive information regarding this subject is available. However, the advent of sequencing technology and the advances in plant biotechnology, coupled with the growing interest in functional genomics-related studies and the various environmental challenges, have paved novel paths to rediscovering the fundamentals of N and its dynamics in physiological and biological processes, as well as biochemical reactions under both normal and stress conditions. This work provides a comprehensive review on multiple facets of N and N-containing compounds in plants disseminated in the literature to better appreciate N in its multiple dimensions. Here, some of the ancient but fundamental aspects of N are revived and the advances in our understanding of N in the metabolism of plants is portrayed. It is established that N is indispensable for achieving high plant productivity and fitness. However, the use of N-rich fertilizers in relatively higher amounts negatively affects the environment. Therefore, a paradigm shift is important to shape to the future use of N-rich fertilizers in crop production and their contribution to the current global greenhouse gases (GHGs) budget would help tackle current global environmental challenges toward a sustainable agriculture.",
"conclusion": "10. Conclusions and Future Perspectives Nitrogen (N) may be to the plant, to some extent, what the seed is to agriculture. Efficient plant growth, development, and productivity are conditioned by the quality of nutrients available, coupled with the ability of the plant to take up and use them. On the one hand, N is one of the most important macronutrients, and by far the most abundantly used in agricultural production. Today, conceiving agriculture without N is nearly utopic, considering that N is vital for the plant and is involved in almost all physiological and biological processes. It is believed that, in order to reduce the impact of current agricultural practices on the environment, optimizing applications of N-rich fertilizers, enhancing the N use efficiency, and implementing better crop management practices are key factors to alleviate the impact of climate change, while maintaining a balanced productivity and quality.",
"introduction": "1. Introduction Agriculture is one of the major components of many development programs aiming at transforming economies, and it is perceived as a core economic sector with great potential to induce sustainable economic growth in several countries [ 1 ]. Agriculture is also central to achieving essential development goals, including ensuring food security and improving human and animal nutrition, as well as achieving employment stability and social growth [ 2 ]. To sustain their growth and productivity, plants require nutrients, which are essential for successful growth and achieving optimum yields. Among them, nitrogen (N), phosphorus (P), and potassium (K), known as primary macronutrients, and sulfur (S), calcium (Ca), and magnesium, identified as secondary macronutrients, have been reported to play a preponderant role. Of all these nutrients, N has proven indispensable in several ways, and N deficiency in the soil has been shown to result in impaired growth and development that culminates in compromised productivity and quality of crops [ 3 , 4 ]. To date, dosages of primary macronutrient applications for various crop species have been optimized, and evidence of increased crop productivity has been recorded. One should note that the dosage of a particular fertilizer is a function of many factors, including plant nutrient requirement and growth stage, the variety of a crop species under cultivation, the type and the concentration of the fertilizer, and soil type and density. In the same way, the availability of nutrients to the plant is affected by the soil type, the pH, and the cationic exchange capacity (CEC) of the soil [ 5 ]. Likewise, abiotic factors, such as salinity, drought stress, or heavy-metal toxicity can impair the availability of nutrients to the plant, thus resulting in poor growth and productivity. In recent years, many reports highlighted a growing interest in the use of an optimized plant nutrition, as well as the promotion of fertilizer use efficiency, which have emerged over the concerns that excessive application of mineral fertilizers, especially those rich in N, has proven to be one of the major causes of greenhouse gases (GHGs) emissions in agriculture [ 6 , 7 , 8 ]. Some voices emphasized the necessity to transition to sustainable agricultural production systems, while others sustained that the use of N-rich fertilizers is necessary for more food production to feed the growing global population. The paradigm of plant nutrition lies in the necessity to tackle the issue of soil fertility for efficient and successful crop cultivation and increased productivity, associated with an enhanced nitrogen use efficiency (NUE) by plants, while considering the sustainability and the resilience of food production systems. Recent alarming global warming records have set an imperative to reduce the emission of GHGs in all sectors. In agriculture (crop production), reducing the excessive application of N-rich fertilizers is a major target. This review aims to provide an expanded view of the roles of N and N-containing compounds in plants, and it presents multiple facets of N under various conditions in a single location. We equally discuss the dynamics of N availability, use, remobilization, deficiency, and toxicity in soil and how these events affect the plant. This work also assesses and highlights the progress made in our understanding of N metabolism in plants, diving from the empirical background to the future use of N, from the perspective of global environmental challenges."
} | 1,485 |
36134169 | PMC9417798 | pmc | 8,358 | {
"abstract": "Electronic textiles (e-textiles) typically comprise fabric substrates with electronic components capable of heating, sensing, lighting and data storage. In this work, we rationally designed and fabricated anisotropic light/thermal emitting e-textiles with great mechanical stability based on a sandwich-structured tri-electrode device. By coating silver nanowire network/thermal insulation bilayer on fabrics, an anisotropic thermal emitter can be realized for smart heat management. By further covering the emissive film and the top electrode on the bilayer, light emitters with desirable patterns and colors are extracted from the top surface via an alternative current derived electroluminescence. Both the light and thermal emitting functions can be operated simultaneously or separately. Particularly, our textiles exhibit reliable heating and lighting performance in water, revealing excellent waterproof feature and washing stability.",
"conclusion": "Conclusions In summary, we present a waterproof and flexible sandwich-structured tri-electrode electronic textile device with lighting and heating dual functions. By coating silver nanowire network/thermal insulation bilayer on fabrics, an anisotropic thermal emitter could be realized for smart heat management with the saturation temperature ranging from 37 °C to 81 °C as the applied voltage was increased from 1.0 to 3.0 V. Particularly, the inside temperature of the device was evidently higher than that of the outside, benefiting from the Cs x WO 3 thermal insulation film. On the other hand, light emitters with defined patterns and tunable colors ranging from blue, yellow to bright white were extracted from the top surface via alternative current-derived electroluminescence. Both the light and thermal emitting functions could be operated simultaneously or separately. Importantly, the washable properties of the e-textiles were studied, and they exhibited stable heating and lighting performances in water, thereby revealing an excellent waterproof feature and washing stability.",
"introduction": "Introduction Developing multifunctional textile is the primary goal of smart cloth for personalized healthcare. Generally, electronic textiles (e-textiles) with personal thermal management (PTM) are recognized as a new effective and energy-saving way to maintain the temperature focusing on the human body. 1–6 A PTM device should assure the basic needs of being wearable, washable, 7,8 and capable of raising the body temperature or cooling if desired. 9 For personal heating, coating conductive nanomaterials on fabrics has been demonstrated to be an effective approach via reflecting infrared radiation (IR) back to the human body. 10–12 Typically, silver nanowires (AgNWs) with high IR reflection efficiency and excellent electrical conductivity are the most promising choice among various metallic nanofibers. 13,14 Although e-textiles for PTM have been developed rapidly, functional diversity realized on a single device is highly desirable, which is of great significance for the facile fabrication and long-term operation. With an increasing demand for intelligent fabrics, the studies on the multifunctionalization of e-textiles have been explored recently. By integrating current electronic or optoelectronic devices, such as nanogenerators, 15 sensors, 16 energy storage 17 or light-emitting devices, 18 with textile substrates, various multifunctionalized e-textiles have been demonstrated. 19,20 For instance, Dong et al. reported a series of multifunctional conductive hydrogel/thermochromic elastomer hybrid fibers with core–shell segmental configuration and their application as flexible wearable strain and temperature sensors to monitor human motion and body/surrounding temperatures. 21 Chen et al. simultaneously integrated a triboelectric nanogenerator with a light emitting diode (LED) on clothes to achieve lighting driven by the triboelectric nanogenerator. 22 Although multifunctional e-textiles have made great progress, further simplification of the preparation process as well as the circuit structure is in urgent need. Especially, intrinsically producing all-in-one wearable electronics with satisfying multiparameter management is still challenging. Herein, we report washable multifunctional e-textiles based on a sandwich-structured tri-electrode device. Tunable heating effect could be realized by depositing a conductive AgNW film on the fabric for efficient heat management. By covering the thermal insulation film, the emissive layer and the top electrode on the surface of the AgNW film, anisotropic thermal regulation and light emission could be achieved simultaneously or separately on a single device. Particularly, the fabricated e-textiles exhibited excellent waterproof features and washing stability, benefiting from all the materials involved in the textile device being waterproof, endowing them with excellent application potential in smart wearable electronics.",
"discussion": "Results and discussion \n Fig. 1 demonstrates the application concept and the working principle of the wearable electronic textile device with a vertically stacked sandwich structure consisting of AgNW–polymer electrodes and ZnS–polymer emissive layer. The cross-sectional scanning electron microscopic (SEM) image of the device is shown in Fig. S4 (ESI † ). The SEM image showed that the thickness of the thermal insulation layer, the luminous layer and the NOA63 protective layer were about 4 μm, 12 μm and 6 μm, respectively. The tunable heating effect was realized by coating conductive AgNW film on the fabric due to the generation of joule heat when a direct current (DC) voltage was applied on the bottom electrode ( Fig. 1b ). 24,25 By subsequently covering the thermal insulation film, the emissive layer and the top electrode on the surface of the AgNW film, the tunable light emission could be achieved by applying an alternating current (AC) between the top and the bottom electrodes ( Fig. 1b ). These two functions exhibited by the fabric device suggested a smart application in functional fabrics. To guarantee an efficient and directed thermal management, a layer of cesium tungsten bronze (Cs x WO 3 ) as the thermal insulation film was incorporated between the AgNW–NOA63 heating film and the light-emitting layer to reduce any unnecessary heat loss due to external environment and maintain the temperature focusing on the human body ( Fig. 1c ). 26,27 Fig. 1 (a) Schematic illustration of the wearable textile device with lighting and heating dual-functions. (b) Schematic diagram of the device structure and working principle of the dual-mode operation. (c) Working principle of the dual-mode device with directed thermal management and light emission. To explore the electrical heating performance of the dual-mode device, the time-dependent temperature profiles of the AgNW–NOA63 film ( R S ≈ 12 Ω sq −1 ) under applied voltages in a range of 1.0–3.0 V were investigated. Fig. 2a shows that the device demonstrates almost the same fast response time under three different applied voltages and reaches a saturation temperature within 70 seconds. As the applied voltage increased from 1.0 to 3.0 V, the saturation temperature increased from 37 °C to 81 °C, suggesting an obvious voltage-dependent heating property. 28 Fig. 2b shows a fast thermal response behavior of the device, and the saturation temperature and response time of the device exhibit almost no evident decline during the 40 repeated heating/cooling cycles with durations of ∼3 min under an input voltage of 3.0 V, indicating a high heating stability of the device. Fig. 2c, d show that the inside temperature of the device is clearly higher than that of the outside, benefiting from the Cs x WO 3 thermal insulation film. Moreover, bending tests with 300 repeated cycles of two-side bending and release process were carried out to further study the heating reliability of the device. Fig. 2e shows that the heating performance of the device remains stable over 300 bending cycles, confirming the excellent long-term heating reliability of the device due to the embedding of AgNWs. Fig. 2 (a) Time-dependent surface temperature profiles of a AgNW–NOA63 film heater ( R S ≈ 12 Ω sq −1 ) at different input voltages. (b) Cyclic on–off test results for the AgNW–NOA63 film heater. (c) Function of the temperature on both sides of the insulation film. (d) Infrared photographs of the heater on both sides of the thermal insulation film. (e) Temperature of the heater as a function of the bending cycle. Besides the outstanding electrical heating performance, the as-fabricated electronic textile device also demonstrated an excellent light emission property. The as-fabricated fabric device demonstrated a bright and uniform light emission powered by an alternating current bias and enabled excellent mechanical flexibility ( Fig. 3a ). Fig. 3b shows that the fabric device exhibits a very stable and constant real-time luminance upon repeated bending with a radius of curvature of 20 mm, suggesting excellent mechanical stability and luminance stability. Moreover, no obvious change in luminance was observed when the radius of bending curvature was gradually changed, further verifying the satisfactory emission property of the device. Fig. 3c shows that the device begins to emit light at a bias voltage of about 100 V, and the emission intensity increases rapidly thereafter to reach as high as 98 Cd m −2 at 300 V and 400 Hz. The relationship between the relative bias voltage and the electroluminescence (EL) intensity could be written as the following formula: 29–31 1 L = L 0 exp(− b / V 1/2 ) where L is the brightness, V is the applied voltage, and L 0 and b are constants associated with the device (the particle size of the phosphor, the concentration of the EL powder in the dielectric, the dielectric constant of the embedding medium and the thickness of the emitting layer). Considering the simple device structure, outstanding flexibility and luminance stability, fabric alternating current electroluminescent (ACEL) devices with patterned shapes could be rationally designed and fabricated via various approaches. For example, by dipping a pen with ZnS–polymer light-emitting layer and directly writing on AgNW–NOA63–fabric substrate, devices with defined shapes as well as different colors were fabricated ( Fig. 3d–f ). In addition, flexible devices with more complicated patterns could be obtained by the embossing approach ( Fig. 3g–i ). Fig. 3 (a) The flexible lighting device under a mechanically distorted state. (b) The luminance variation of the lighting device under different curvatures and the luminance variation during 500 times bending. (c) Luminance versus alternating voltage properties of the device at different frequencies. (d) Schematic diagram of the device patterned by the dipping pen writing. (e) Optical picture of the device patterned by the dipping pen writing. (f) Luminescence picture of the device patterned by the dipping pen writing. (g) Schematic diagram of the device patterned by embossing. (h) Optical picture of the device patterned by embossing. (i) Luminescence picture of the device patterned by embossing. Benefiting from good stability of both heating and lighting performances, these two functions of the device were further investigated. When AC power was applied on the device with a bottom AgNW electrode with a sheet resistance of 13 Ω sq −1 , the device was in lighting mode and emitted a bright blue light with a surface temperature of 37 °C ( Fig. 4a ). In the case of heating mode, when DC power was applied, the device emitted no light while the temperature increased to 74 °C ( Fig. 4b ). If both AC and DC power were applied, the device exhibited a high surface temperature of 74 °C and emitted light ( Fig. 4c ). These results revealed that the heating and lighting functions of the e-textiles could work separately, and the device could simultaneously achieve a good heating/lighting performance. Fig. 4d shows that the brightness of the device does not change significantly as the surface temperature increases due to an increase in the applied voltage, which confirms that the dual functions perform independently without any effect on each other when working simultaneously. Fig. 4 (a) Photograph of the device in the lighting mode. (b) Photograph of the device in the heating mode. (c) Photograph of the device both in lighting and heating mode simultaneously. (d) The relationship between the heating temperature and the luminance of the device. To explore the washable property of the flexible e-textiles, the heating and lighting performances of the device after washing for 50 cycles were investigated. Fig. 5 shows that the dual-mode device still maintains integrated structure in water and works normally. As the washing time prolonged, the heating temperature and the electroluminescence of the device almost remained unchanged, which suggested an excellent waterproof performance and washing stability of the flexible fabric devices, endowing them with great application potential in functional wearable electronics. Fig. 5 (a) Schematic diagram of the water-proof property and washability of the device. (b) Infrared photograph of the device in heating mode in water. (c) Photo of the device in lighting mode in water. (d) Heating performance of the device after different washing cycles. (e) Lighting performance of the device after different washing cycles."
} | 3,380 |
30278125 | null | s2 | 8,359 | {
"abstract": "Initial attachment to a surface is a key and highly regulated step in biofilm formation. In this study, we present a platform for reversibly functionalizing bacterial cell surfaces with an emphasis on designing biofilms. We engineered the Lap system of Pseudomonas fluorescens Pf0-1, which is normally used to regulate initial cell surface attachment, to display various protein cargo at the bacterial cell surface and control extracellular release of the cargo in response to changing levels of the second messenger c-di-GMP. To accomplish this goal, we fused the protein cargo between the N-terminal retention module and C-terminal secretion signal of LapA and controlled surface localization of the cargo with natural signals known to stimulate or deplete c-di-GMP levels in P. fluorescens Pf0-1. We show this system can tolerate large cargo in excess of 500 amino acids, direct P. fluorescens Pf0-1 to surfaces it does not typically colonize, and program this microbe to sequester the toxic medal cadmium."
} | 252 |
39747896 | PMC11696221 | pmc | 8,360 | {
"abstract": "Active biological molecules present a powerful, yet largely untapped, opportunity to impart autonomous regulation of materials. Because these systems can function robustly to regulate when and where chemical reactions occur, they have the ability to bring complex, life-like behavior to synthetic materials. Here, we achieve this design feat by using functionalized circadian clock proteins, KaiB and KaiC, to engineer time-dependent crosslinking of colloids. The resulting material self-assembles with programmable kinetics, producing macroscopic changes in material properties, via molecular assembly of KaiB-KaiC complexes. We show that colloid crosslinking depends strictly on the phosphorylation state of KaiC, with kinetics that are synced with KaiB-KaiC complexing. Our microscopic image analyses and computational models indicate that the stability of colloidal super-structures depends sensitively on the number of Kai complexes per colloid connection. Consistent with our model predictions, a high concentration stabilizes the material against dissolution after a robust self-assembly phase, while a low concentration allows for oscillatory material structure. This work introduces the concept of harnessing biological timers to control synthetic materials; and, more generally, opens the door to using protein-based reaction networks to endow synthetic systems with life-like functional properties.",
"introduction": "Introduction The current state-of-the-art in next-generation materials design is to create structures that can achieve desired functions in response to external perturbations, such as self-repair in response to damage 1 – 6 . Looking beyond this stimulus-response framework, we envision autonomously functional materials that can not only respond directly to their environment, but also have the capability to store a memory of past events and dynamically change their own properties 7 – 11 . Such materials could be used to create dynamic sequestration devices that filter toxins on a programmable schedule, or medical implants that self-assemble and restructure to protect and suture wounds and dissolve once fully healed. An attractive strategy to equip materials with robust autonomous function is the use of distributed information processing throughout the material, rather than a central controller. This concept is similar to the function of interacting networks of biomolecules in living cells, which provide finely-tuned spatiotemporal regulation of physiology. In many cases, a small number of interacting network components can be isolated from the cell and retain modular function to achieve tasks such as defining structures with a specific size 12 , generating spatial patterns 13 , or keeping time 14 . On a larger scale, the collective action of these biomolecules provides a way for energy flux to impart non-equilibrium properties into structures to create active matter. The last two decades have seen tremendous progress in identifying and understanding the emergent properties of active matter, from active colloids and molecular motor-driven active biomaterials, to soft robotics and living concrete 15 – 23 . However, engineering autonomous materials with robust, kinetically-controlled activity, inherent to living systems, remains a grand challenge in materials science 24 – 26 . Here, we develop the proof-of-concept of an autonomous material with properties that are temporally programmed by biological signaling molecules using protein components derived from the cyanobacterial circadian clock (Fig. 1 ). In their natural context, KaiA, KaiB, and KaiC proteins generate a self-sustaining ~24-h rhythm that is used to synchronize physiology with the external light-dark cycle. Moreover, these proteins can be removed from their cellular context and can generate stable oscillations in a reconstituted in vitro system 14 , 27 – 29 . Fig. 1 Harnessing circadian clocks to engineer non-equilibrium materials across scales. A We functionalize cyanobacteria clock proteins—hexameric KaiC rings (blue), KaiA dimers (cyan), and KaiB monomers (green)—to couple to materials by incorporating biotinylated KaiB (b-KaiB). B KaiB biotinylation: (left) example sites of possible amine-reactive biotinylation (magenta) overlaid on the KaiB crystal structure (green) 60 , 61 , (right) SDS-PAGE gel of unlabeled KaiB and biotinylated KaiB (b-KaiB), showing successful biotinylation indicated by a mobility shift of the biotinylated product (molecular weight standards [M] are 10 and 15 kDa). C KaiABC reactions in the presence of biotinylated KaiB. Oscillations are measured by fluorescence polarization of FITC-labeled KaiB (0.2 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${{{\\rm{\\mu }}}}{{{\\rm{M}}}}$$\\end{document} μ M ), a read-out of KaiB–KaiC complex formation. All conditions contain 3.5 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${{{\\rm{\\mu }}}}{{{\\rm{M}}}}$$\\end{document} μ M KaiB, with the specified fraction being b-KaiB. Oscillatory association of KaiB with KaiC is sustained with 55% b-KaiB (magenta), the percentage used in subsequent experiments, but not with 80% (pink). Each curve is the mean of two replicates. D KaiB monomers bind cooperatively to KaiC rings in a phosphorylation-dependent manner (indicated by the orange “P” circles), mediated by KaiA, and are subsequently released as KaiC dephosphorylates over a 24-h cycle. We exploit the transition from free KaiB to KaiB fully assembled on a KaiC hexamer to create a time-dependent and phosphorylation-dependent change in crosslinking valency. E We incorporate the “circadian crosslinkers” depicted in ( D ) into suspensions of 1- \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${{{\\rm{\\mu }}}}{{{\\rm{m}}}}$$\\end{document} μ m streptavidin-coated colloids to drive time-dependent crosslinking of colloids. F Microscope images of fluorescent streptavidin-coated colloids, mixed with KaiB, b-KaiB, and KaiC phosphorylation site mutants that cannot bind KaiB (left) or constitutively bind KaiB (right), show that KaiB–KaiC assembly selectively causes mesoscale clustering and connectivity of colloids. G Sedimentation of colloidal clock suspensions shown in ( F ) demonstrates pronounced settling of colloids after a day of incubation with the mutant that forms constitutively KaiB–KaiC complexes (right) compared to colloids mixed with the non-binding KaiC mutant (left). Cartoons depict the expected state of the suspension for constitutively binding (left) and non-binding (right) mutants (not drawn to scale). Source data for C are provided as a Source Data file. These oscillations can be detected as an ordered pattern of multisite phosphorylation on KaiC, which acts as a signaling hub that binds and releases protein partners throughout the cycle 30 – 32 . In brief, KaiC consists of two tandem ATPase domains, CI and CII, arranged into a hexameric ring (one blue circle in Fig. 1A represents one subunit consisting of a CI-CII pair). KaiA binds to the CII domain of KaiC, which stimulates autophosphorylation 33 , 34 . Phosphorylation accumulates slowly throughout the day, first on Thr432, then on Ser431. When Ser431 is heavily phosphorylated (shown in Fig. 1D ), corresponding to dusk, ring-ring stacking interactions allow the CII domain to regulate the slow ATPase cycle in CI 35 . The post-hydrolysis state of CI allows KaiB to bind to KaiC 36 and six KaiB molecules to assemble cooperatively on the CI ring 37 , 38 . KaiB is a metamorphic protein that is not competent to bind KaiC in its ground-state fold and must refold to form the KaiB–KaiC complex. Though ground-state KaiB can tetramerize, previous data suggests it is largely monomeric at standard concentrations 37 , 39 . This nighttime KaiB–KaiC complex sequesters and inactivates KaiA, closing a negative feedback loop to inhibit further phosphorylation and allowing KaiC to dephosphorylate. Unphosphorylated KaiC then releases KaiB and is ready to begin the cycle again. The kinetics of the phosphorylation rhythms are robust to temperature and protein concentration, yet can be tuned over a wide range by single amino acid substitutions in the Kai proteins 40 , 41 . The system is also unusually thrifty in its energy consumption—with each KaiC molecule consuming only 15 ATP per day 42 . Thus, the Kai protein system is a uniquely attractive choice to develop into a synthetic tool to endow materials with programmable, autonomously time-dependent properties.",
"discussion": "Results and discussion Biotinylation of KaiB allows KaiC to selectively mediate material crosslinking To harness the Kai protein system for materials activation, we aimed to exploit the changing oligomeric state of KaiB, induced by interaction with KaiC, throughout the circadian cycle mediated by KaiA. Namely, KaiB transitions from being free in solution to forming a hexameric KaiB–KaiC complex (KaiBC). We reasoned that functionalized KaiB molecules would not be effective crosslinkers of material components when KaiB is free in solution, but that assembly into the KaiBC complex might create a potent multivalent crosslinker where time-dependent KaiB–KaiC interactions can bridge multiple biotin-streptavidin bonds (Fig. 1D ). To develop this tool and characterize its effect, we chose a commercial colloidal suspension as a model material platform (Fig. 1E ). We hypothesized that as KaiBC complexes form over time, the number of colloids able to participate in KaiB-mediated crosslinks would increase, and we would observe a transition from a fluid-like suspension of single colloids, to a gel-like state with larger connected clusters of colloids (Fig. 1E, F ). Consistent with this prediction, we expect macroscopic changes in the ability of the colloidal material to sediment (Fig. 1G ). To endow the KaiBC complex with time-dependent material crosslinking properties, we first needed to functionalize KaiB to bind to the colloids strongly and statically, which we achieved through biotinylation of KaiB (Fig. 1A, B ) and the use of streptavidin-coated colloids (Fig. 1E–G ). We next needed to ensure that biotinylation of KaiB did not interfere with oscillatory assembly of the KaiBC complex (in the presence of KaiA), and that biotinylated KaiB (b-KaiB) could still bind to KaiC in a phosphorylation-dependent manner. To achieve the former, we used a fluorescence polarization (FP) assay to monitor rhythmic complex formation in the KaiABC reaction, finding that the reaction could tolerate the majority of the KaiB proteins being replaced by b-KaiB while still producing high amplitude rhythms (Fig. 1C ). The altered amplitude of rhythms in the presence of b-KaiB may reflect a change in complex stability. In pull-down experiments, we found that b-KaiB retains its ability to interact with KaiC as well as its preference for the pS431 state (Fig. S1 ). We next aimed to test the ability of KaiC to selectively crosslink b-KaiB-coated colloids via KaiBC complex assembly. To this end, we mixed b-KaiB into a suspension of 1- \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${{{\\rm{\\mu }}}}{{{\\rm{m}}}}$$\\end{document} μ m diameter streptavidin-coated colloids, then added either the pS KaiC mutant or the pT KaiC mutant to the suspension. pS and pT are mutated at the phosphorylation sites of KaiC to mimic either a state that permanently allows KaiB binding (pS—S431E;T432A, corresponding to the night phase) or prevents binding (pT—S431A;T432E, corresponding to the morning phase). By imaging the fluorescent-labeled colloids, we found that the colloids remained largely as isolated microspheres in the presence of the non-binding pT mutant, exhibiting no preferential self-association even after a day of incubation (Figs. 2A and S2 ). Fig. 2 KaiB–KaiC complexes crosslink colloids with high specificity in a phosphorylation-dependent manner. Fluorescence microscopy images of suspensions of 1-µm diameter colloids taken at 1 h (top), 7 h (middle), and 28 h (bottom) after mixing with KaiC mutants that are frozen in A non-binding (pT) or B binding (pS) states show substantial clustering and assembly of pS-colloids over time that is absent for pT-colloids. Fluorescence microscopy images of suspensions of colloids of 2 µm ( C ) and 6 µm ( D ) diameter in the presence of pT and pS KaiC proteins show that timed aggregation, dependent on the phosphorylation state of KaiC, is preserved for different sizes of colloids. The concentrations of colloids, proteins, and reagents, as well as imaging parameters, are identical to those in ( A , B ). E Images of a suspension of 1-µm diameter colloids undergoing sedimentation in a 12 mm long capillary over 28 h in the presence of pT (left) and pS (right) show that pS-colloids sediment more quickly, as indicated by dark regions extending further down the images. The time that each image is captured is listed at the top. F The same suspension parameters as in ( A , B ) but without KaiC (including only KaiB and b-KaiB) show minimal clustering over the course of 1 h (top) to 28 h (bottom), demonstrating that the KaiB–KaiC complex formation is essential to the colloidal self-assembly shown in ( A – E ). G Suspensions of streptavidin-coated colloids, with identical conditions to those in ( A , B ), but with Kai proteins replaced with alternative biotinylated constructs that could, in principle, crosslink streptavidin-coated colloids: (left) 1 kDa biotin-PEG-biotin with 1 biotin on each end, (middle) 20 kDa biotin-PEG-biotin with 1 biotin on each end, and (right) biotin-BSA with 8-16 biotins. The molarity of PEG and BSA matched the KaiC molarity used in ( A – E ), and minimal clustering is observed from 1 h (top) to 28 h (bottom), demonstrating that the effect shown in ( A – F ) is unique to the KaiB–KaiC binding interaction. This minimal self-association is similar to that observed without b-KaiB (Fig. S2 ), indicating that non-specific crosslinking is low. In contrast, Fig. 2B shows that b-KaiB-coated colloids incubated with the binding-competent pS KaiC mutant grow into large colloidal aggregates. This crosslinking mechanism is robust, producing qualitatively similar effects with different-sized colloids (Fig. 2C, D ), and forming structures that are system-spanning and relatively immobile compared to pT KaiC samples in all cases (Fig. S3 ). Thus, b-KaiB can act as a potent material crosslinker that functions only in the presence of appropriately phosphorylated KaiC. To demonstrate that these mesoscopic structural changes can translate to macroscopic material changes visible to the naked eye, we imaged colloidal suspensions undergoing sedimentation in capillaries on the centimeter scale over the course of a day. We observed macroscopic sedimentation dependent on the phosphorylation of KaiC: the larger microscopic clustering seen for pS is mirrored by pronounced sedimentation of the suspension, while the pT colloids remain suspended (Fig. 2E ). To further establish the robustness and specificity of this self-assembly process, we performed experiments in which we: (1) removed KaiC (Fig. 2F ), (2) removed b-KaiB (replacing with wild-type KaiB) (Fig. S2 ), and (3) replaced streptavidin-coated colloids with passivated non-functionalized colloids (Fig. S4 ). We observed no clustering in any of these control experiments (Fig. S5 ). We also performed experiments in which we replaced Kai proteins with bovine serum albumin (BSA) or PEG polymers that have multiple biotinylation sites, but are present on the same linker molecule rather than brought together by macromolecular complex assembly, as in the Kai protein system. While, in principle, these polymer linkers have the potential to act as crosslinkers to bind streptavidin-coated colloids, we found minimal clustering for all linker sizes and number of biotin sites (Figs. 2G and S6 ). Thus, the self-assembled structures we observe specifically require the formation of KaiBC complexes. We understand this unique feature as arising from the fact that KaiC acts as a “middle-man” to crosslink colloids via binding to b-KaiB. In this way, all biotin moieties on a single b-KaiB can be bound to the same colloid without preventing KaiC association. In the alternative designs, all biotin moieties that could crosslink two colloids are on the same construct, which strongly favors their binding to the same colloid over a neighboring one. This unique selectivity is an important functional feature of the KaiABC crosslinking system, independent of its potential for oscillatory behavior. Crosslinking proceeds with kinetics programmed by clock proteins To characterize the clock-driven clustering kinetics that program the colloidal suspensions to transition from freely floating particles to large, connected superstructures, we collected images of the emerging clusters at nine different time points over the course of a day. To promote mixing and limit colloid settling and surface adsorption, we kept all samples under continuous rotation between imaging intervals. Overlaying temporally color-coded images from these time-course experiments confirms that structure emerges over time in the pS KaiC sample, while minimal clustering is seen in the presence of pT (Fig. 3A, B ). Fig. 3 KaiBC crosslinking mediates robustly timed self-assembly of colloidal clusters that are synced with KaiBC complex formation. Colorized temporal projections of time-lapses of pT KaiC ( A , yellow) and pS KaiC ( B , cyan) over the course of 28 h, with colors indicating increasing time from dark to light according to the color scales. (Insets) Zoomed-in regions of the projections highlighting pS-specific cluster growth over time that is absent for pT. Pixel intensity probability distributions for pT ( C , yellow), and pS ( D , cyan) at different times over 28 h, with lighter shades denoting later times according to the color scales in ( A , B ). Distributions show broadening and emergence of high-intensity peaks at later times for pS. Dashed gray line denotes the full width at 1% of the maximum probability (FW1%), which serves as a clustering metric used in ( F ). Data are generated from the same images used to generate colormaps in ( A , B ). E Spatial image autocorrelation functions \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) versus radial distance \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$r$$\\end{document} r (in units of colloid diameter) for 5 different times between 1 and 28 h for pT (yellow) and pS (cyan) with color shade indicating time according to the legends in ( A , B ). The characteristic correlation length \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ , determined by fitting each \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) curve to an exponential function, is denoted by the intersection of the dashed horizontal line at \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g={e}^{-1}$$\\end{document} g = e − 1 . Data shown are the mean and SEM across 36 images from two replicates. F Correlation lengths \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ (open squares), FW1% (half-filled triangles), and median cluster size (filled circles, see Figs. S7 and S8 ), each normalized by their initial pT value, show that the time course of cluster assembly over 28 h for pT (gold) and pS (cyan) correlate with the fluorescence polarization (FP) of fluorescently labeled KaiB (right axis, mP), which serves as a proxy for KaiBC complex formation. Both the degree of clustering and FP remain at a minimum for pT, while for pS, both steadily increase for the first ~15 h. \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ , FW1%, and cluster size data are the mean and SEM across 36 images from two replicates. FP data are the mean and SEM (too small to see) of three replicates. Source data are provided as a Source Data file. To quantify the time-dependent colloidal self-assembly that is apparent in microscopy images, we use spatial image autocorrelation (SIA) analysis to measure the average size of colloidal clusters at each time point. SIA quantifies the correlation \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) between pixels separated by a radial distance \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$r$$\\end{document} r (Figs. 3C and S7 ), which decays exponentially from \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g\\left(0\\right)=1$$\\end{document} g 0 = 1 with the decay rate indicating the characteristic size of features in an image. Slower decay of \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) with \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$r$$\\end{document} r indicates larger features (i.e., clusters), as seen for pS compared to pT and long compared to short times (Figs. 3C and S7 ). By fitting each \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) curve to an exponential decay, we quantify a characteristic correlation lengthscale \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ of the colloidal system, indicated by the distance \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$r$$\\end{document} r at which the dashed horizontal line intersects each curve in Fig. 2C . We also implemented alternative image analysis algorithms to assess clustering, including quantifying the distribution of pixel intensities (Figs. 3D, E and S7 ) and directly detecting clusters as connected regions in a binarized image (Figs. S7 and S8 ), both yielding similar results to SIA (Figs. 3D, E , S7 and S8 ). Specifically, the full pixel intensity distribution width at 1% (FW1%) and median cluster size both display similar time-dependence as \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ (Fig. 3F ). To directly compare these different clustering metrics, we normalize each quantity by the corresponding initial value for pT such that the values indicate the degree of clustering, which is one in the absence of clusters (Fig. 3F ). Given that pS KaiC is locked into a binding-competent state, the gradual self-assembly of colloids over many hours, suggests that the rate-limiting step in self-assembly is KaiB–KaiC complex formation rather than the time needed for colloids to come into close contact. Indeed, KaiBC complexes are known to form on the timescale of many hours, likely due to both the slow ATPase cycle in the KaiC CI domain 43 and the time required for KaiB to refold into an alternative fold-switched structure 39 , 44 . To test this hypothesis, we measured the kinetics of the KaiBC interaction using FP of labeled KaiB, which increases with increasing formation of KaiBC complexes, and compared the results to the kinetics of material self-assembly. Figure 3F shows in overlay the time evolution of the relative FP, demonstrating that KaiBC complex formation grows approximately linearly for the first 15 h after which it approaches saturation, likely reflecting a regime where the majority of both KaiB and KaiC molecules are in complex and have been depleted from solution. The agreement between the kinetics of KaiBC interactions and material self-assembly shown in Fig. 3F , as well as the robust specificity of the colloidal assembly (Fig. 2A, B, F ), is strong evidence that the biochemical properties of the Kai proteins, such as the KaiC catalytic cycle, are regulating the rate of cluster growth. Brownian Dynamics simulations recapitulate timing of cluster formation The correlation of KaiB fluorescence polarization with the clustering of colloids suggests that Kai protein interactions control the kinetics of clustering. In order to assess this mechanism, we developed a numerical simulation that captures the key components of our experimental system (see section “Methods” and SI ). In the simulations, 1- \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${{{\\rm{\\mu }}}}{{{\\rm{m}}}}$$\\end{document} μ m diameter colloids move via Brownian motion in a 50 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${{{\\rm{\\mu }}}}{{{\\rm{m}}}}$$\\end{document} μ m × 50 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${{{\\rm{\\mu }}}}{{{\\rm{m}}}}$$\\end{document} μ m 2D plane, and, when the surfaces of two colloids are within 10 nm of each other (comparable to the size of the KaiBC complex 45 ), they can form a bond mediated by b-KaiBC complexes. KaiB and KaiC are assumed to be present at constant concentrations and their interaction to form crosslinks is treated phenomenologically. The probability of complex formation during an encounter is a constant value chosen to match the solution-binding kinetics (see SI ). We allow simulations to run for 30 h and capture the state of the colloids at the same time intervals as in experiments (Fig. 4 ). Fig. 4 Kinetic simulations of Kai-mediated crosslinking recapitulate slow formation of colloidal clusters. A Simulation snapshots showing clustering of colloids (red circles) crosslinked by permanent bonds (blue lines), analogous to the experimental pS-colloid system, at 1 (left), 7 (inset), and 28 (right) hours. Colorized temporal projections of ( B ) simulation snapshots for colloids with permanent crosslinker bonds (permanent bonds, P) and ( C ) experimental snapshots for pS-colloids show similar features emerging over the course of a day. Colorized temporal projections of ( D ) simulation snapshots for colloids with no crosslinker bonds (no bonds, N) and ( E ) experimental snapshots for pT colloids both show minimal clustering or restructuring over the course of a day. Times and color-coding used in projections are the same as in Fig. 3 , as indicated by the color scales. F \n \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) computed for simulation snapshots, taken at times specified in the legend, for colloids with no bonds (N, yellow squares) and permanent bonds (P, cyan triangles). Time course of the ( G ) correlation lengths \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ and ( H ) colloid connectivity number CCN determined from simulations with permanent bonds (P, cyan) and no bonds (N, gold). I Multiple metrics of clustering and self-assembly resulting from permanent crosslinker bonding in experiments (pS) and simulations (P), each normalized by its maximum value to indicate the fractional clustering index (left axis) measured using each metric. Metrics include: experimental correlation lengths ( \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ , open squares), simulated correlation lengths (sim \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ , filled squares), full width at 1% (FW1%, half-filled triangles), and median cluster size (cluster size, filled circles). Trends in both simulation and experimental data track with the time course of KaiB fluorescence polarization (right axis (mP), translucent triangles) in a reaction with pS KaiC. All simulation data shown is the mean and SEM across five replicates. Experimental data shown in ( C , D , I ) are reproduced from Fig. 3 . Source data are provided as a Source Data file. To model our experimental pS KaiC and pT KaiC colloidal systems, we consider cases in which, respectively, bonds between colloids are incapable of releasing once they are formed (Fig. 4B ) and bond formation probability is zero (Fig. 4D ). The color-coded temporal overlays of simulation images with “permanent bonds” ( P ) and “no bonds” ( N ) show qualitative similarities with the experimental overlays of pS and pT (Fig. 4C, E ). To quantitatively compare simulated and experimental cluster assembly kinetics, we perform the same SIA analysis that we use for experimental images (Fig. 3C ) to compute time-dependent autocorrelation curves (Fig. 4F ) and corresponding correlation lengths (Fig. 4G ). Similar to the \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) trends we observe for experimental pT and pS images (Fig. 3C ), Fig. 4F shows that \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) for the “no bonds” system exhibits minimal time-dependence and fast decay with distance \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$r$$\\end{document} r , indicative of small features that do not change size over time. Conversely, \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$g(r)$$\\end{document} g ( r ) for the “permanent bonds” case ( P ) decays more slowly than N at all time points and broadens substantially over time, indicative of larger clusters that grow over time. The time course of the corresponding correlation lengths of the 30-h simulation are similar to the experimental trends shown in Fig. 3F , with \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi$$\\end{document} ξ values for P growing over time and transitioning to slower increase in the latter half of the simulation. The continued cluster growth for pS (experiments) and P (simulations) is somewhat unexpected given the saturation of the fluorescence polarization at ~15 h. Specifically, FP saturation suggests that all possible KaiBC complexes have formed, while the colloid data suggest that clusters continue to form and grow after this saturation. To shed light on this seeming paradox, we compute the colloid connectivity number (CCN) from simulations, which measures how many neighboring colloids are connected to a single colloid. Because of the 2D geometry and the size of the colloids, the maximum possible CCN is six. Figure 4I shows that CCN increases to saturating levels over the course of ~10–15 h, similar to the KaiBC FP data, while the simulated correlation lengths continue to increase after this time, albeit less dramatically than the first half of the time course (Fig. 4G ). These data indicate that cluster growth can proceed even when the majority of colloids are saturated with permanent crosslinks. Such assembly kinetics may arise if the majority of saturated colloids are in the interiors of clusters, while those on the boundaries have available b-KaiB binding sites to crosslink to other colloids on the edges of neighboring clusters. Self-assembly thus transitions from that of single colloids coming together to form clusters, to one in which most colloids are participating in clusters that then merge to form larger superstructures. Figure 4I corroborates this physical picture by comparing the kinetics of cluster formation in the experimental and simulation data with the KaiBC assembly kinetics. The similarity in the shapes of the experimental and simulation curves indicates that the model is indeed capturing the underlying process of generating clusters. The clear shift in kinetics at ~15 h in all data further corroborates the robustness of the simulations, and demonstrates that self-assembly is rate-limited by the timescale of KaiBC complex formation. Oscillations in colloidal clustering depend on the crosslinker density Having demonstrated that material self-assembly can be temporally programmed by the phosphorylation state of the circadian clock proteins, we now investigate the effect of oscillatory interactions between KaiB and KaiC in the wild-type system when KaiA is present. To achieve oscillatory crosslinking, we replace the phosphorylation-locked mutants with wild-type (WT) KaiC and add KaiA, creating a circadian rhythm in both KaiC phosphorylation (mediated by KaiA) and the KaiB–KaiC interaction (Fig. 1C, D ). To guide our experiments, we first aimed to understand how oscillating crosslinkers may translate to the dynamics of material self-assembly. To do so, we extended our model shown in Fig. 4 to allow sinusoidally varying colloid binding and unbinding rates (see section “Methods”, SI ). In brief, we consider the same binding rate amplitude \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${p}_{o}$$\\end{document} p o as in the permanent bond case but we incorporate an oscillation of this rate, \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${p}_{{on}}={p}_{o}{\\cos }^{2}(\\pi t/T)$$\\end{document} p o n = p o cos 2 ( π t / T ) , where \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$T$$\\end{document} T is the oscillation period. This construction models the coherent bulk oscillations in the biochemical properties of the KaiABC reaction. We also add a dissociation rate with the same amplitude and functional form as the binding rate, but that is \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\pi /2$$\\end{document} π / 2 radians out of phase, i.e., \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${p}_{d}={p}_{o}{\\sin }^{2}(\\pi t/T)$$\\end{document} p d = p o sin 2 ( π t / T ) . This framework assumes that each connection between two colloids is bonded by a single KaiBC complex ( \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 1). Figure 5A shows that this minimal bonding allows for oscillatory connectivity, with peaks in CCN observed at times that are in phase with the peaks in the input oscillatory binding and that roughly correlate with the measured peaks in KaiC phosphorylation (Figs. 5D and S9 ). However, the peak CCN values are low compared to the saturating value of 6, and non-zero CCN values are only maintained for a small fraction of the oscillation period, suggesting very weak oscillatory clustering. Fig. 5 Oscillatory material properties depend on crosslinker density. A – C Simulations model oscillatory colloidal crosslinkers with different numbers of KaiB–KaiC complexes ( \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n$$\\end{document} n , bonds) participating in each connection between colloids. A Colloid connectivity (CCN) versus time for systems with different numbers of bonds per colloid connection, from light to dark gray: \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 1, 4/3, 5/3, 2, 3. Arrow indicates direction of increasing \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n$$\\end{document} n . Intermediate bond numbers 1 < \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n \\, < \\, 2$$\\end{document} n < 2 result in oscillating connectivity, while \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=1$$\\end{document} n = 1 is not sufficient for pronounced clustering and \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n\\ge 2$$\\end{document} n ≥ 2 promotes sustained cluster growth with minimal dissolution. B The fractional clustering index (see text) versus time for bond densities shown in ( A ) reveal oscillatory clustering for \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n < 2$$\\end{document} n < 2 that is most pronounced for \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=5/3$$\\end{document} n = 5 / 3 . The colored boxes enclose the data points corresponding to the simulation images with color-matched borders shown in ( C ). C Simulation snapshots that correspond to troughs (red, orange) and peaks (green, blue) shown in ( B ) demonstrate that peaks and troughs correspond to substantial and minimal clustering, respectively. D Fluorescence polarization (FP) of KaiB (left axis, triangles) and percentage of phosphorylated KaiCs (%P, right axis, circles) during a KaiB–KaiC reaction. %P measurements were performed in the presence (filled circles) and absence (open circles) of colloids, showing that oscillatory KaiC phosphorylation dynamics are unaffected by the presence of colloids (see Fig. S9 ). E , F The fractional clustering index versus time for colloid experiments performed with KaiC concentrations of 6.67 µM (1 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × , dark gray circles), 3.33 µM (0.5 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × , gray squares), and 1.67 µM (0.25 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × , light great diamonds) reveal oscillatory clustering for the lowest concentration, similar to the simulated \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=5/3$$\\end{document} n = 5 / 3 case, while the two higher concentrations steadily become increasingly clustered over time, similar to the simulated \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n\\ge 2$$\\end{document} n ≥ 2 cases. Colored boxes enclose the data points corresponding to the microscope images with color-matched borders shown in ( F ). F Microscope images that correspond to troughs (red, orange) and peaks (green, blue) shown in ( E ) show strong similarities to simulated images and demonstrate minimal and substantial clustering, respectively. Simulated colloid connectivity ( G ) and fractional clustering index ( H ) for extended times (30–48 h) for the same bond numbers examined in ( A – C ). Simulation snapshots are shown for the \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=5/3$$\\end{document} n = 5 / 3 case at 33 (blue), 39 (red), 45 (brown), and 48 (green) hours, as well as the \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=1$$\\end{document} n = 1 and \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=3$$\\end{document} n = 3 cases at 48 h (right column). Translucent boxes that match the snapshot borders indicate the corresponding connectivity and fractional clustering index. All simulated and experimental images shown are 50 µm \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × 50 µm. All experimental data shown is the mean and SEM across 36 images from two replicates. All simulation data shown is the mean and SEM across five replicates. Source data are provided as a Source Data file. However, given the saturating level of Kai proteins in our experiments (~10 5 b-KaiB proteins per colloid) and the two orders of magnitude smaller size of the crosslinkers compared to the colloid surface area, we anticipate that more than one KaiBC bond participates in a typical connection between two colloids in experiments. To incorporate multiple bonds per connection into our simulations we modify the dissociation rate to include the number \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n$$\\end{document} n of bonds that participate in each colloid connection, as \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${p}_{d}={p}_{0}^{n}{\\sin }^{2}(\\pi t/T)$$\\end{document} p d = p 0 n sin 2 ( π t / T ) . Additional curves in Fig. 5A show that \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 2 and \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 3 produce saturating connections that are unable to appreciably dissociate during a bond oscillation cycle. However, for intermediate cases \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 4/3 and 5/3, we observe pronounced oscillations in connectivity, suggesting similar oscillatory clustering of colloids. Similar to Fig. 4 , we translate connectivity to clustering kinetics by computing the correlation length for each time point that is captured in experiments. To compare the time dependence of complex formation for different bond numbers we evaluate the fractional clustering index, which we define as the baseline-subtracted correlation length, \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\xi \\left(t\\right)-{\\xi }_{\\min }$$\\end{document} ξ t − ξ min , normalized by the corresponding maximum value, \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\xi }_{\\max }-{\\xi }_{\\min }$$\\end{document} ξ max − ξ min : \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${FCI}=\\left(\\xi \\left(t\\right)-{\\xi }_{\\min }\\right)/\\left({\\xi }_{\\max }-{\\xi }_{\\min }\\right)$$\\end{document} F C I = ξ t − ξ min / ξ max − ξ min . All values of this function lie between 0 and 1 to allow us to isolate the time dependence of the clustering. Figure 5B shows that robust oscillatory clustering is achieved for \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 4/3 and 5/3, with the initial peak being more pronounced for \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 5/3. Figure 5C shows the simulation snapshots that correlate with the peaks and troughs of the clustering index, visually demonstrating the presence of large superstructures at the peaks and minimal clustering at the troughs. We understand this complex dependence on bond density as follows: a low density of crosslinkers (i.e., \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=$$\\end{document} n = 1) does not allow superstructures to form, simply because many particles will not be able to find an attachment point, even if the Kai proteins in the system are in a binding-competent state. However, at high crosslinker density (i.e., \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n\\ge$$\\end{document} n ≥ 2), multivalent effects prevent superstructures from easily disassembling once formed, even when the KaiBC binding probability falls. Thus, the model predicts a “sweet spot” in crosslinker density (i.e., 1 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$ < \\, n \\, < $$\\end{document} < n < 2) where the underlying molecular rhythm in KaiBC interaction will be transduced into material properties with high amplitude (Fig. 5B, C ). Armed with these predictions, we performed experiments at different Kai concentrations, to mimic varying \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n$$\\end{document} n values in simulations. We first aimed to demonstrate that the ~24 h oscillation in KaiBC complex formation is not disrupted by the presence of colloids. Figure 5D confirms that the expected oscillation in KaiBC complex formation, measured by FP, is unperturbed by the buffer conditions used for assembling colloidal materials; and that the oscillatory phosphorylation of KaiC is similarly preserved (Fig. S9 ) and largely unaffected by the presence of streptavidin-coated colloids. We then performed the same full time course of microscopy measurements as for the pS and pT mutants (Fig. 2 ) at 1 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × , 0.5 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × , and 0.25 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × of the Kai concentration used in the mutant experiments (6.66 µM, Figs. 2 and 3 ). Evaluating the same fractional clustering index as in simulations (Fig. 5B ), we find that for the two higher concentrations, colloid superstructures assemble slowly over the course of the day, but show no detectable disassembly, similar to the \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n\\ge$$\\end{document} n ≥ 2 simulations (Fig. 5B, E ). However, at the lower Kai concentration (0.25 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\times$$\\end{document} × , 1.67 µM), oscillatory clustering appears (Fig. 5E, F ), with peak and trough times corresponding approximately to the peak and trough KaiBC interaction detected by FP (Fig. 5D ). The clustering, dissolution and re-clustering quantified by the fractional clustering index (Fig. 5E ) can be observed in the corresponding microscope images (Fig. 5F ) which have very similar features to the simulated images (Fig. 5C ). These results demonstrate the achievement of timed assembly and disassembly of a material and the power of predictive modeling to identify the appropriate region of phase space to achieve this engineering feat. To further demonstrate the oscillatory nature of the assembly, we extend the timescale of simulations out to 48 h over which we expect to capture an additional assembly oscillation. As shown in Fig. 5G, H , we observe a similar oscillation in connectivity and clustering index as in the first 24 h, with similar dependence on \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n$$\\end{document} n . Simulation snapshots at the peaks and trough of the oscillations are also highly similar to those shown in Fig. 5C , demonstrating repeatable oscillatory assembly and disassembly. While oscillatory material assembly indeed represents a transformative advance in materials design, we also point out that the robust time constant associated with the material assembly, dictated by KaiBC complex formation, provides an additional important technological advance. Indeed, the kinetics of cluster assembly are robustly regulated, nearly independent of protein concentration, when we use non-oscillating mutants to program the assembly phase of the material (Fig. S10 ). That the timescale of protein complex assembly is nearly invariant to concentration over a broad range 43 is a unique feature of the KaiB–KaiC system, likely because the binding timescale is rate-limited by the KaiC ATPase cycle and the foldswitching of KaiB. This robustness of timing against fluctuating concentration would be difficult to achieve via other assembly control mechanisms (Fig. S11 ). Outlook Biomolecular signaling systems typically must maintain their function with high fidelity while interacting with numerous other components crowded within living cells and while subject to unpredictable fluctuations in their environment. These constraints equip networks of interacting biomolecules with unique robustness properties that may allow them to be harnessed to endow synthetic materials and systems with functionality, programmability, and autonomous reconfigurability. However, coupling biomolecular systems to synthetic materials to impart desired properties remains a grand challenge in active matter and biomaterials research 24 , 46 – 49 . Here, we break new ground by using the KaiABC circadian clock as a prototypical example of a robust biomolecular signaling system. We demonstrate that this system can maintain its natural activity even when functionalized to act as a material crosslinker, and that it can then be used to autonomously regulate timed material self-assembly and oscillation. Specifically, we show that Kai proteins can assemble colloidal suspensions into networks of mesoscopic clusters at rates and efficiencies that are controlled by the phosphorylation state of KaiC. These molecular interactions translate to bulk changes in the sedimentation properties of the materials, visible by the naked eye. Moreover, our mathematical models show that the valency of the clock protein crosslinkers can be used as a switch to allow either sustained self-assembly or rapid dissolution of the material. In intermediate regimes of valency, this system drives oscillatory material properties. This proof-of-concept opens the door to Kai-mediated scheduled crosslinking of a diversity of synthetic and natural materials, such as hydrogels, polymeric fluids, cellulose, and granular materials, to drive user-defined autonomous changes in material properties on a programmable schedule. Importantly, the timing of material self-assembly can not only be robustly programmed by Kai clock proteins, but that timing can be precisely tuned with the use of KaiC mutants that operate on cycles of different durations from ~15 to 158 h 40 . The intrinsic temperature compensation property of KaiC would further ensure that the kinetics of assembly are robust against environmental fluctuations 50 . Other accessory proteins, including SasA and CikA can be incorporated and functionalized to allow for material interactions peaking at other phases of the cycle 27 . These designs can potentially be used to create technologies such as dynamic filtration and sequestration devices, self-healing infrastructure, and programmable wound suturing. Beyond material crosslinking, the Kai system could be used as a synthetic scaffold to gate enzymatic activity to control the release of drugs or achieve metabolic channeling by enforcing spatial proximity between other entities. Beyond time-keeping, biological systems are capable of many information-processing tasks including thresholding, fold-change detection, and sign-sensitive filtering of input signals. Because these systems all function based on high molecular specificity, they represent a natural library of computational devices that can be coupled to non-biological systems to achieve autonomous control."
} | 16,212 |
37576678 | PMC10413832 | pmc | 8,361 | {
"abstract": "Promoted model architectures or algorithms are crucial for intelligent\nmanufacturing since developing them takes a lot of trial and error\nto embed the domain knowledge into the models correctly. Especially\nin semiconductor manufacturing, the whole processes depend on complicated\nphysical equations and sophisticated fine-tuning. Therefore, we use\na neuroevolution-based model to search the optimized architecture\nautomatically. The collector current value at a particular bias of\nthe silicon–germanium (SiGe) heterojunction bipolar transistor,\ngenerated by technology computer-aided design (TCAD), is used as the\ntarget dataset with six process parameters as the inputs. The processes\ninclude oxidation, dry and wet etching, implantation, annealing, diffusion,\nand chemical–mechanical polishing. Our work can build a suitable\nmodel network with a fast turnaround time, and practical physical\nconstraints are fused in it without domain knowledge extraction. Take\nthe case with 3840 data and one output as an instance. The mean square\nerrors of the train set and validation set, as well as the mean absolute\npercentage error of the test set, are 1.317 × 10 –6 , 7.215 × 10 –7 , and 0.216 while using multilayer\nperceptron (MLP) and they are 3.285 × 10 –7 ,\n1.661 × 10 –7 , and 0.097 while using NE. The\nconsequences show that the work in this vein is promising. According\nto the trend plot and results, the ability to extract physic is much\nbetter than the traditional (MLP) model.",
"conclusion": "Conclusions This work can robustly and automatically tune the model architecture\nand find the model capable of extracting domain knowledge in semiconductor\nmanufacturing problems. The autoevolved network architecture is shown\nto be self-adaptive to the domain knowledge, which can be important\nin many practical fields where data are expensive, or the problems\nare highly complex. With the combination of NE and Adam, the optimal\nmodel architectures can achieve convergence with the train set loss\nand validation set loss settled at smaller values in reference to\nMLP baselines. Compared with MLP, overfitting can be avoided, and\nless data are required for training. After training, the prediction\nof the test data and the trend graph drawn reflect that the optimal\nNE model has a better comprehension of the domain knowledge, denoting\nthat the method can reduce the need for complex formulas and tremendous\nhuman effort to build appropriate model architectures.",
"introduction": "Introduction Many machine learning (ML) and deep learning technologies have\nbeen developed for intelligent manufacturing to deal with various\ntasks and apply them to process monitoring, process control, optimization,\nfault detection, and prediction. 1 Even\nin semiconductor manufacturing, sophisticated and intricate industries,\nML methods can still be used for them. This includes using optimized\nmodel architectures based on traditional ML models 2 or innovative ones, 3 and improved\noptimization algorithms 4 in yield prediction\nand analysis, manufacturing process excursion detection, and manufacturing\nflow simplification. 5 In model architectures, most current models are designed mainly\nby human experts, which is very time-consuming due to the need for\nsufficient domain knowledge and trial and error for suitable architectures.\nTherefore, the neural architecture search (NAS) or neuroevolution\n(NE) is a field attracting more attention as it can automatically\nfind the optimized model for different cases by using search methods\nand fitting scores. 6 These methods are\ngenerally divided into several categories. First, the evolution-based\nmethods, or neuroevolution (NE), where the genetic algorithm is the\nmost regularly used technique in which a specific population of individuals\nis generated by algorithms, and a series of the process with selection,\ncrossover, and mutation will be conducted after being evaluated and\nsorted according to their fitness score. Then, the new generation\nof individuals will develop in the direction of increasing the overall\nfitness from generation to generation until the termination condition\nis satisfied. 7 , 8 Second, the reinforcement learning\n(RL) method uses the accuracy of the child network defined by action\nspace as a reward to set the training direction. 9 Third, the recently flourishing gradient optimization methods\nin which the model architecture composed of mixed operations are processed\nin a differentiable form and parameterized, and then, using the gradient\ndescent method to evaluate the quality of the architecture. 10 Fourth, the surrogate model-based optimization\ncan be regarded as a progressive model-constructing process. Its core\nconcept is to search from a simple model, build a surrogate model\nof the objective function iteratively by recording past evaluations,\ngradually evolve to more complex architectures, and use it to predict\nthe most promising architecture. 11 − 13 Bayesian Optimization\n(BO) is one of the most popular hyperparameter optimization methods,\nand many recent studies have attempted to apply them to NAS, 11 , 12 in which the validation results, in some cases, 11 are modeled as a Gaussian process (GP) guiding the search\nfor an optimal architecture. Nevertheless, in GP-based BO methods,\nthe inference time can cubically increase with the number of observations\nand cannot validly deal with variable-length problems. 14 Finally, other methods, including random search, 15 grid search, 16 simulated\nannealing, 17 etc., have been applied to\nexecute NE/NAS. Although more and more applications of neural architecture\nsearch have been utilized in various problems, 18 in semiconductor fields, NAS studies have not been prevailing.\nOnly a few studies using RL methods are located in the field of semiconductor\nmanufacturing, 19 and almost none of them\nare based on NE or gradient optimization methods to optimize model\narchitectures. Achieving high accuracy needs a great deal of data which is sometimes\ntricky, costly, or impractical to attain. 20 Furthermore, data from science and engineering are apt to be dispersed\nand noisy because real-world experiments are expensive and subject\nto the environment and equipment. Hence, predictions may not be robust,\nowe interpretability, and even violate physical constraints, leading\nto generalization errors. 21 Integrating\nhuman knowledge into ML can help overcome these difficulties to some\nextent by significantly reducing data requirements, ameliorating reliability\nand robustness, and building interpretable ML systems. 20 , 22 Domain-knowledge-based ML is a practical approach employed in primarily\nscientific fields and can be generally classified into domain-knowledge-informed\nmachine learning architecture, hybrid domain-knowledge-informed machine\nlearning model, multifidelity framework, and Bayesian framework. 23 By embedding prior information into the neurons\nor layers, combining domain-knowledge-based models with ML models,\nor incorporating domain knowledge constraints into learning algorithms,\nit can enhance the information content of the available data and promote\nthe learning algorithm to attain the correct solution. In this work,\ntechnology computer-aided design (TCAD) is used for simulating semiconductor\nprocesses under various parameters and producing datasets, 24 which offer sufficient, densely distributed\ndata when experimental data measured from real devices are limited. The MLP neural network is usually used as the basic structure in\ndomain-knowledge-based ML. Still, its architecture must be adjusted\nfirst whenever it is applied to different tasks. Besides, accurately\nintegrating prior domain knowledge into the architecture is also a\nconundrum. Thus, we take advantage of the automatic search ability\nof NAS to find out a particular network architecture as a type of\ndomain knowledge constraint to help convergence. Previously, we have\nreceived some achievements in optimizing ML compact device models\nvia NE algorithms. 25 In our work, the NE\nalgorithm capable of dynamically tuning model architectures is utilized\nas a model optimization method to extract the physical characteristics\nof semiconductor devices manufactured under different process parameters.\nThere have been very few studies in the field of semiconductors using\nNE with evolved topologies. 26 Therefore,\nwe want to apply the method to clarify the effect of NE in semiconductor\nprocesses and device problems. Even in the entire field of chemistry\nand physics, NE-related works do not prevail, and after a careful\nliterature review, we only found these studies 27",
"discussion": "Result and Discussion Table 1 shows the\nMLP and NE-based models trained under six inputs with one I c value as an output. In each dataset size,\nthe train set is divided into different proportions, and the rest\nis the test set. During training, 20% of the train set is assigned\nto the validation set, and training is conducted for 20,000 epochs\nwith the patience 500 of early stop setting. The best case of the\nMLP models selected among multiple architectures is used to compete\nwith the optimized model based on NE. It should be emphasized that\nboth MLP and NE models are saved at the lowest validation loss during\ntraining, i.e., restoring the best weights. From the performance on\neither the training or the test set, NE is better than MLP. It is\nobserved that the optimized model architecture of the NE-based model\ncan attain better results with fewer iterations when using gradient\ndescent regardless of the dataset sizes or train-test splits. This\nphenomenon also indicates the effect of data efficiency. The dataset\nsizes and train-test splits in Table 2 are the same as the conditions in Table 1 , but the only difference is\nthat the three I c values at different\nbiases are used as the outputs in MLP and NE-based models. The purpose\nof training under six inputs and three outputs is to increase the\ncomplexity of the problems to show the superior fitting and domain-knowledge\nextraction capability of NE. The results externalize that NE can still\noutperform MLP in this case. Table 1 MLP and NE Results on SiGe HBT Semiconductor\nProcessing Datasets with 6 Inputs and 1 Outputs Networks Used in This\nCase, and Training Epochs Are 20,000 with the Patience 500 of Early\nStop Setting and Restoration of the Best Weights model architecture dataset size train-test\nsplit epochs train set\nMSE train set\nMAPE (%) validation\nset MSE validation\nset MAPE (%) test set\nMSE test set\nMAPE (%) numbers of\nparameters multilayer\nperceptron 1920 0.2 2373 1.095 × 10 –4 1.515 1.109 × 10 –4 1.511 1.274 × 10 –4 1.692 353 0.5 7066 1.160 × 10 –6 0.166 8.905 × 10 –7 0.173 9.880 × 10 –7 0.183 281 0.8 6136 6.305 × 10 –7 0.129 4.746 × 10 –7 0.125 5.985 × 10 –7 0.139 217 2880 0.2 6220 1.001 × 10 –5 0.474 1.088 × 10 –5 0.542 2.374 × 10 –5 0.709 217 0.5 8837 5.114 × 10 –7 0.103 5.774 × 10 –7 0.117 4.473 × 10 –7 0.112 281 0.8 4208 4.695 × 10 –6 0.292 2.942 × 10 –6 0.300 2.862 × 10 –6 0.298 217 3840 0.2 8144 2.383 × 10 –6 0.350 2.851 × 10 –6 0.407 4.139 × 10 –6 0.459 353 0.5 5353 5.047 × 10 –7 0.146 3.537 × 10 –7 0.157 3.490 × 10 –7 0.152 353 0.8 8132 1.317 × 10 –6 0.210 7.215 × 10 –7 0.222 6.861 × 10 –7 0.216 217 neuroevolution 1920 0.2 7554 1.480 × 10 –7 0.070 2.007 × 10 –7 0.077 2.589 × 10 –7 0.086 346 0.5 9200 1.283 × 10 –7 0.045 6.031 × 10 –8 0.041 8.690 × 10 –8 0.049 258 0.8 9885 1.103 × 10 –7 0.045 6.578 × 10 –8 0.046 7.711 × 10 –8 0.051 228 2880 0.2 11,054 4.996 × 10 –7 0.088 4.172 × 10 –7 0.095 7.624 × 10 –7 0.121 203 0.5 9155 2.722 × 10 –7 0.062 2.231 × 10 –7 0.067 2.026 × 10 –7 0.069 258 0.8 6935 3.016 × 10 –7 0.069 1.561 × 10 –7 0.071 1.695 × 10 –7 0.069 228 3840 0.2 7043 3.303 × 10 –7 0.125 2.937 × 10 –7 0.142 3.182 × 10 –7 0.142 343 0.5 4940 2.803 × 10 –7 0.104 1.739 × 10 –7 0.107 1.872 × 10 –7 0.107 346 0.8 6904 3.285 × 10 –7 0.099 1.661 × 10 –7 0.097 1.582 × 10 –7 0.097 228 Table 2 MLP and NE Results on SiGe HBT Semiconductor\nProcessing Datasets with 6 Inputs and 3 Outputs Networks Used in This\nCase, and Training Epochs Are 20,000 with the Patience 500 of Early\nStop Setting and Restoration of the Best Weights model architecture dataset size train-test\nsplit epochs train set\nMSE train set\nMAPE(%) validation\nset MSE validation\nset MAPE(%) test set\nMSE test set\nMAPE(%) numbers of\nparameters MLP 1920 0.2 8023 5.607 × 10 –6 0.403 6.355 × 10 –6 0.444 7.465 × 10 –6 0.476 297 0.5 3716 6.984 × 10 –6 0.456 3.770 × 10 –6 0.370 7.007 × 10 –6 0.455 297 0.8 5383 3.058 × 10 –7 0.092 2.517 × 10 –7 0.096 2.569 × 10 –7 0.096 453 2880 0.2 10,843 6.271 × 10 –6 0.366 7.191 × 10 –6 0.439 7.515 × 10 –6 0.453 231 0.5 5621 9.154 × 10 –7 0.152 8.010 × 10 –7 0.158 7.904 × 10 –7 0.164 371 0.8 7086 9.709 × 10 –7 0.166 8.011 × 10 –7 0.172 7.960 × 10 –7 0.171 297 3840 0.2 7283 2.133 × 10 –6 0.327 4.059 × 10 –6 0.475 4.616 × 10 –6 0.433 543 0.5 3745 2.486 × 10 –6 0.378 2.443 × 10 –6 0.397 2.338 × 10 –6 0.395 453 0.8 6480 8.150 × 10 –7 0.208 6.943 × 10 –7 0.211 6.003 × 10 –7 0.201 453 NE 1920 0.2 20,000 2.564 × 10 –7 0.079 2.487 × 10 –7 0.089 2.798 × 10 –7 0.097 272 0.5 16,594 1.283 × 10 –7 0.059 1.129 × 10 –7 0.066 1.128 × 10 –7 0.065 272 0.8 5078 2.505 × 10 –7 0.074 1.564 × 10 –7 0.075 1.593 × 10 –7 0.076 488 2880 0.2 19,633 4.930 × 10 –7 0.123 4.917 × 10 –7 0.130 7.505 × 10 –7 0.151 242 0.5 12,477 5.497 × 10 –7 0.104 3.417 × 10 –7 0.107 3.383 × 10 –7 0.107 365 0.8 5113 3.613 × 10 –7 0.089 2.737 × 10 –7 0.094 2.254 × 10 –7 0.087 272 3840 0.2 13,761 5.657 × 10 –7 0.161 5.273 × 10 –7 0.189 5.743 × 10 –7 0.190 499 0.5 15,175 5.445 × 10 –7 0.172 3.803 × 10 –7 0.170 3.801 × 10 –7 0.169 468 0.8 11,437 2.455 × 10 –7 0.112 1.969 × 10 –7 0.116 1.716 × 10 –7 0.111 421 As shown in Figure 2 , the domain-knowledge trends are plotted to see the effect of NE\nfor the different cases in Tables 1 and 2 . Figure 2 a–d shows predictions under 3840 data,\ntrain-test split = 0.8, and one output. Figure 2 a,b shows the linear graphs and logarithm\ngraphs of I c to En dose2 , while Figure 2 c,d shows graphs of I c to En dose1 and C dose2 , respectively. As shown in Figure 2 a,b, the predictions of MLP are roughly the same as\nthose of NE. Nevertheless, in Figure 2 c,d, due to the slight change in collector current,\nit is evident that NE is better than MLP at predicting the domain-knowledge\ntrends, showing the fitting capability and domain-knowledge extraction\nof NE. Figure 2 e, f\nshows images similar to those in Figure 2 a–d, which are only changed to three\noutputs. Since the simultaneous prediction of three outputs enhances\nthe complexity of the model, the poor predictions of the curves in\nMLP are observed in Figure 2 e–f. In this case, it can also be observed that NE’s\nfitting capability and domain-knowledge extraction is better, as shown\nin Figure 2 g,h. Figure 2 Trend charts by the prediction of MLP and NE with dataset size\n= 3840 and train-test split = 0.8. (a) I c – En dose2 in the linear scale\nwith one output, (b) I c – En dose2 in log scale with one output, (c) I c – En dose1 with one output, (d) I c – C dose2 with one output, (e) I c – En dose2 in the\nlinear scale with three outputs, (f) I c – En dose2 in the log scale with\nthree outputs, (g) I c – En dose1 with three outputs, and (h) I c – C dose2 with three\noutputs. Figure 3 a–d\nare the one-output cases, and Figure 3 e–h are for three outputs. In (a), (b), (e),\nand (f), train loss and validation loss of the best individual in\na generation of NE are selected. They indicate that after the evolution\nof NE, the convergence speed of the architecture training process\nincreases in both the train set and validation, stably surpassing\nthe performance of MLP. In (c), (d), (g), and (h) the training loss\nand the validation loss of the model with one output and three outputs\nduring training are shown. (c) and (g) show both for MLPs, while (d)\nand (h) are all models after NE optimization. The graphs reveal that\nthe optimized NE model has a fast convergence speed and less fluctuation. Figure 3 MLP and NE models’ fitness score with dataset size = 3840,\ntrain-test split = 0.8. (a) MLP and NE model’s loss in the\ntrain set with one output. (b) MLP and NE model’s loss in the\nvalidation set with one output. (c) MLP model’s training loss\nand validation loss comparison with one output. (d) NE model’s\ntraining loss and validation loss comparison with one output. (e)\nMLP and NE model’s loss in the train set with three outputs.\n(f) MLP and NE model’s loss in the validation set with three\noutputs. (g) MLP model’s training loss and validation loss\ncomparison with three outputs. (h) NE model’s training loss\nand validation loss comparison with three outputs. Finally, Figure 4 a,b shows the performance of NE and MLP models in 1 output and three\noutput cases, respectively. The graphs record the accuracy of each\ngeneration of NE and the best hand-tuned MLP for the predictions from\nthe test set, in which the mean absolute percentage error (mape) is\nused as an evaluation. At the beginning of the process, the mape scores\nof each individual of NE are relatively scattered. However, the individual\nof each generation in the search space will be stably evolved and\nconverged to a specific architecture that is good at predicting the\ndomain knowledge trend of TCAD simulations, and then, the optimal\nmodel is obtained, as shown in Figure 5 a,b. Figure 4 (a) MLP and NE models’ performance in the test set with\nsix inputs and one output. Dataset size = 3840 and train-test split\n= 0.8. (b) MLP and NE models’ performance in the test set with\nsix inputs and three outputs. Figure 5 (a) Optimized NE model by GA with six inputs and one output. Dataset\nsize = 3840 and train-test split = 0.8. (b) Optimized NE model by\nGA with six inputs and three output. In smart manufacturing, NAS is suitable for solving the issue that\nthe development time of domain-knowledge-based model construction\nis time-consuming. NE methods are used as a strategy to search the\nentire search space, such as neural networks through augmenting topologies\n(NEAT). 7 By the genetic algorithm, NEAT\nevolves architecture and weights incrementally through adding neurons\nand removing connections, which can continuously explore search space.\nHowever, for the complex nonlinearity in the dataset, NEAT only uses\nthe genetic algorithm to solve it, and it may take a long time to\napproach excellent performance compared to gradient descent. In our\nwork, Adam is used to train the network, and the genetic algorithm\nis used to evolve the architecture and weights. Selection operation\nrandomly chooses parents to reproduce, which maintains the genetic\ndiversity of networks. Parts of the parents’ characteristics\nare kept under various operations, such as crossover and mutation,\nand the architectures are robustly evolved. In the way of the gradient\ndescent, the nonlinearity issue can be solved, and the genetic algorithm\ncan find the optimized model for domain-knowledge extraction. In the field of using domain knowledge in ML, Li et al. eliminate\nthe nonphysical behavior produced from MLP model predictions by embedding\ndomain knowledge into layers of the model. The domain knowledge is\nto use the tan h function to fit the linear region\nof I D – V DS at small V DS and the saturation\nregion at large V DS in thin-TFET devices.\nBesides, sigmoid functions are used to describe the I D – V TG curve in the\nsubthreshold region that turns on exponentially and then becomes polynomial\nin the on region. Using these two activation functions as the domain-knowledge\nconstraints can solve the counter-theory prediction caused by the\nlack of devices’ domain knowledge in a specific part of the\nMLP model. 38 Kao et al. also proposed a\nhybrid physics-based BSIM model and ML model architecture, where the\noutput of the BSIM is the current value and the output of the ML model\nis a bias-dependent correction function ε( V GS , V GD ) for the nonidealities\nnot included in the BSIM model. The output of the BSIM and ML models\nare multiplied to obtain the final current value. This method can\nachieve satisfactory generalization and Gummel symmetry of devices\nin most device operation regions. 39 Although\nthe above two methods can successfully reach their aim, they still\nneed human tuning in ML model architectures to find a better one.\nIn our approach, our model uses the GA-based NAS method to generate\nthe model architecture and activation function selection, which is\na fully automated process. As long as the search rules of the algorithm\nare given, the architecture search can be automatically tuned without\ntrial and error by human hands. Compared with the MLP model, the NE\nmodel shows better generalization and is closer to the actual value\nin the result prediction. The nonlinearity and high complexity of the dataset is a significant\nproblem in many fields. The MLP model makes use of an activation function\nto solve this problem. The domain knowledge itself can be complicated\nand thus results in high data nonlinearity. Therefore, MLP that only\nuses a single activation function often cannot effectively solve it.\nAlthough using an MLP model with more parameters can help fit a dataset\nwith complex changes, it is also more likely to cause the model to\noverfit the training set. Eventually, the model’s prediction\naccuracy on the test set will decline. Consequently, the GA-based\nNE method employed in this work is to adjust the model architecture,\nwhich is tantamount to integrating domain knowledge constraints into\nthe framework after modifications. Combining the connections with\ntwo activation functions between blocks here provides another assistance\nto the architecture. It helps nonlinearity fitting since the connections\nare selected by GA with a uniform random selection scheme to ensure\nthe search space’s diversity. Parents’ strengths are\npreserved for the next generation through crossover and mutation to\nevolve the network architectures robustly. The whole process here\nis to limit the counter-theory behavior that the MLP model is prone\nto predict because of insufficient data information. Thus, our work\ncan retain generalizability with customized network architecture. Among the current optimizers, Adam combines the advantages of momentum\non the RMSProp method’s base, stabilizing it via additional\nhyperparameters, such as β 1 , β 2 ,\nand ϵ. 40 Parameter updates can be\nmade more stable when initializing or encountering small gradients\ncontinuously. Nevertheless, Adam tends to have poor generalization\nthan traditional SGD. Even though a fast convergence is accomplished\nduring training, this causes the fact that errors during testing are\nalmost much worse than those during training. 41 Some studies also mention that Adam has convergence problems in\nsome cases. The adaptive learning rate algorithm in Adam may lead\nto suboptimal solutions due to exponential moving averages. 42 MLP models are more restricted models with redundant\narchitectures than the NE model, so the neural networks trained in\nsuch a space with extremely high dimensions easily get stuck in the\nsaddle and stagnate. In our approach, since the model architectures\nare constantly optimized in addition to the parameter update during\neach training, the abovementioned dilemma can be overcome during training.\nGiven a problem, the NE model has more chance to smooth the error\nsurface owing to the calibration through GA, that is, an easier way\nto approach the minimum in comparison with the MLP model."
} | 5,880 |
32872654 | PMC7559569 | pmc | 8,363 | {
"abstract": "This paper discusses the surface-engineered nanomaterials (adaptive nano-structured physical vapor deposition (PVD) thin-film coatings) that can effectively perform under severely non-equilibrium tribological conditions. The typical features of these nanomaterials are: (a) Dynamically interacting elements present in sufficient amounts to account for its compositional/structural complexity; (b) an initial non-equilibrium state; (c) optimized micro-mechanical characteristics, and (d) intensive adaptation to the external stimuli. These could be considered as functionally graded nanomaterials that consist of two major layers: an underlying (2–3 microns) thin-film PVD coating, the surface on which an outer nanoscale layer of dynamically re-generating tribo-films is produced as a result of self-organization during friction. This tribo-film nanolayer (dissipative structures) was discovered to represent complex matter, which exhibits characteristic properties and functions common to naturally occurring systems. These include adaptive interaction with a severely non-equilibrium environment; formation of compounds such as sapphire, mullite, and garnet, similar to those that arise during metamorphism; ability to evolve with time; as well as complexity and multifunctional, synergistic behavior. Due to several nanoscale effects, this nanolayer is capable of protecting the surface with unprecedented efficiency, enabling extensive control over the performance of the entire surface-engineered system. These surface-engineered nanomaterials can achieve a range (speed and level) of adaptability to the changing environment that is not found in naturally occurring materials. Therefore, these materials could be classified as metamaterials. The second major characteristic of these materials is the structure and properties of the coating layer, which mostly functions as a catalytic medium for tribo-film generation and replenishment. A functioning example of this type of material is represented by an adaptive hard thin-film TiAlCrSiYN/TiAlCrN nano-multilayer PVD coating, which can efficiently work in an extreme environment, typical for the dry machining of hard-to-cut materials.",
"conclusion": "3. Conclusions This paper outlines the potential benefits to nanoscience that can be provided by the study of phenomena occurring under harsh environmental conditions. A high compositional complexity in combination with nano-structural characteristics may result in the development of a surface-engineered nanomaterial with unprecedented ability to sustain these conditions. This approach could be effectively applied in a variety of challenging practical fields. A novel generation of adaptive nano-structured PVD thin-film coatings and, in particular, the TiAlCrSiYN/TiAlCrN bi-multilayer PVD coating is introduced in this work, which could successfully function under strongly non-equilibrium tribological conditions combined with high temperatures (about 1000–1200 °C) and stresses (up to 5 GPa). The surface-engineered nanomaterial considered in this paper has some characteristics typical of complex adaptive systems: A large number of dynamically interacting elements that result in compositional/structural and architectural complexity, an initial non-equilibrium state of the coating layer, openness of the tribo-system via adaptation to a severely nonequilibrium environment, and, lastly, optimized mechanical characteristics. These surface-engineered layers could be considered as functionally graded nanomaterials that consist of two major, differently scaled layers: (1) The outer, 2–5 nm-thick layer of dynamically regenerating tribo-films that form as a result of interaction with the environment. This layer is produced as a consequence of self-organization during friction with the formation of dissipative structures on top of (2) the underlying 2–3 micron-thick PVD coatings layer. The dynamically formed top nano-layer of the tribo-films represents complex matter. As such, it exhibits the following characteristics and functions common to natural systems: openness by means of adaptive interaction with a non-equilibrium severe environment, formation of the compounds such as sapphire, mullite, garnet, and others, similar to those that develop as a result of metamorphism, temporal behavior and ability to evolve with time, complexity and multifunctionality, emergent (synergetic) performance. The tribo-film nanolayer can efficiently protect the frictional surface, thereby enabling control over the performance of the entire coating layer. Two recently discovered nanoscale phenomena are described: The “trigger’’ effect and the ‘’heat flow reflection’’ effect. Due to these effects, the surface-engineered nanomaterials can reach a range of adaptability to the increasingly severe environment that is not found in natural materials. Therefore, the nanomaterials outlined in this study could be considered as metamaterials. The second functionally graded nanomaterial layer discussed here, i.e., the micron-thick coating layer, mostly serves as a medium for the catalysts of tribo-film regeneration. To promote this function, its architecture, structure, and properties have to be carefully optimized. As the presented nanomaterials are highly ordered adaptive systems, dynamically interacting with an increasingly severe environment, they are capable of effectively sustaining extreme operating conditions with unattainable levels of adaptability and wear performance. A true understanding of such metamaterials can only be obtained by a holistic approach.",
"introduction": "1. Introduction A top-down approach that mostly focuses on the miniaturization of structural components still predominates in nanoscience and nanotechnology [ 1 ]. This miniaturization approach is quite efficient in many fields of science and technology [ 2 ]. However, the predominant reductionist paradigm in material science and engineering barely applies to natural systems [ 3 ] that are endowed with differing levels of complexity. In light of this problem, a new trend has emerged in surface science, namely nature-inspired surface engineering. Several leading scientific institutions are now moving in the nature-mimicking research direction [ 4 ], and principal international conferences have recently taken place on this topic. It has already been established that the investigation of dissipative, non-equilibrium processes, similar to those that can even be observed in the living world, constitute highly promising grounds for future scientific advances [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 ]. Therefore, the next step in nanomaterial and nanotechnology development should be the introduction of complex adaptive systems that function in a way designed to imitate natural processes [ 14 ]. This nature-mimicking approach should not just be limited to the simple copying of the exact design of natural systems. The time has come to incorporate a temporal approach to the design of such materials [ 15 ]. To further proceed in this direction, it is necessary to introduce a somewhat new philosophy based on the paradigm of complexity and the associated theory of self-organization [ 6 , 16 , 17 ]. Complexity demands a fundamentally different attitude from the mainstream reductionist approach. Although the universal laws are undoubtedly valid, each complex system is different. These differences are of such importance that specific case studies need to be analyzed and understood in each complex system for any conclusion to be carried over to another one [ 6 ]. This is why we strongly contend that the approach to the nanomaterial design laid out in this paper would prove useful in a wide variety of applications. One of the major topics covered in this work is the self-organization phenomenon [ 18 , 19 , 20 , 21 ] in a specific area of tribology. A key condition for self-organization is the presence of a strong gradient of characteristics during the non-equilibrium process (for instance, in Benard cells [ 21 ]). A common example of such a process is friction [ 22 ], which typically exhibits a strong gradient of various characteristics (for example, temperatures and stresses) among the surfaces in tribo-contact [ 22 ]. This presents an environment under which potential self-organization phenomena may readily emerge to their full extent [ 22 ]. Currently, the intense demand to improve the efficiency of various engineering processes results in increasingly severe and even extreme operating environments [ 23 ]. In the present paper, we use extreme tribological conditions as such an example. In this case, tribo-systems can almost always be found to operate under strongly non-equilibrium conditions [ 24 ]. One significant example of such conditions is manifested under the ultra-performance dry (lubricant-free) machining (in particular, ball nose end milling) of hard-to-cut materials when the tribo-system operates under heavy (up to 3–5 GPa) loads and high (1000–1200 °C and above) temperatures [ 25 ]. Such severe cutting conditions can be withstood by cutting tools with adaptive nanostructured thin-film wear-resistant physical vapor deposited (PVD) coatings [ 23 , 26 ]. A major feature of such coatings is their ability to form highly protective/lubricating tribo-films on the friction surface as a result of the interaction with the external environment [ 24 , 27 ]. The application of these adaptive coatings can lead to an unprecedented increase in machining productivity [ 28 ]. A new generation of wear-resistant thin-film coatings is introduced in this work, which represents a specific example of a complex adaptive system. Such systems are capable of fully developing their unique properties during temporal interaction with the ever-changing, severe environment they are embedded in [ 29 , 30 , 31 , 32 , 33 , 34 ]. The probability of self-organization’s incidence is higher in complex systems [ 35 ]. Therefore, complex systems, contrary to expectation, can spontaneously exhibit stunning degrees of order, which, in turn, is essential for the understanding of their emergent (synergistic) behavior [ 36 , 37 , 38 ]. A significant recent achievement in surface engineering was the introduction of coatings with a complex composition, which contain at least five alloying elements. This was accomplished in two different ways: High-entropy alloy coatings [ 39 , 40 , 41 , 42 , 43 , 44 , 45 ] and coatings with emergent performance [ 27 ]. The coating with emergent properties [ 27 ] showed much promise under severe external conditions associated with the machining of hard-to-cut materials. The next step to improve upon was the coating’s architecture: multilayered and bi-multilayered [ 28 ]. Combining these two approaches (compositional and structural optimization) resulted in the development of efficiently operating complex adaptive systems. The goal of this paper is to demonstrate a complex adaptive system using up-to-date results from the gradual compositional and architectural optimization of a surface-engineered layer, which is represented by a bi-nano-multilayer hard thin-film TiAlCrSiYN/TiAlCrN coating deposited by PVD."
} | 2,790 |
39819972 | PMC11739516 | pmc | 8,364 | {
"abstract": "We introduce a novel neuromorphic network architecture based on a lattice of exciton-polariton condensates, intricately interconnected and energized through nonresonant optical pumping. The network employs a binary framework, where each neuron, facilitated by the spatial coherence of pairwise coupled condensates, performs binary operations. This coherence, emerging from the ballistic propagation of polaritons, ensures efficient, network-wide communication. The binary neuron switching mechanism, driven by the nonlinear repulsion through the excitonic component of polaritons, offers computational efficiency and scalability advantages over continuous weight neural networks. Our network enables parallel processing, enhancing computational speed compared to sequential or pulse-coded binary systems. The system’s performance was evaluated using diverse datasets, including the MNIST dataset for image recognition and the Speech Commands dataset for voice recognition tasks. In both scenarios, the proposed system demonstrates the potential to outperform existing polaritonic neuromorphic systems. For image recognition, this is evidenced by an impressive predicted classification accuracy of up to 97.5%. In voice recognition, the system achieved a classification accuracy of about 68% for the ten-class subset, surpassing the performance of conventional benchmark, the Hidden Markov Model with Gaussian Mixture Model.",
"introduction": "Introduction The rapid development of artificial neural networks and applied artificial intelligence is predominantly aimed at the efficient processing of large data sets and pattern recognition 1 . Traditional approaches, however, are increasingly facing constraints in terms of computational speed and energy efficiency, particularly in hardware implementations of these networks 2 . These constraints have spurred interest in neuromorphic systems, where hardware mimics the structure and function of the human brain. The exploration of novel materials and mechanisms is crucial for the development of efficient neuromorphic systems 3 . A particularly promising direction in this research area is the exploitation of exciton-polariton interactions within specially designed semiconductor microcavities in the strong light-matter coupling regime 4 , 5 . Exciton-polaritons are quasiparticles emerging from the coupling of photons and excitons 6 . They possess dual light and matter properties, enabling strong optical nonlinearity and picosecond scale reaction times. These characteristics enable the development of polariton based high-speed neuromorphic systems with high efficiency 7 . The term “polariton neuron” was first introduced in ref. 8 , devoted to planar waveguide structures that translate polariton coherence over extended distances. This research laid the groundwork for using polariton neurons to construct binary logic gates in semiconductor microcavities, serving as a sort of precursor for neuromorphic computing. Much later, the reservoir computing scheme has emerged as a key approach for the development of polariton-based neural networks 9 , 10 . This technique employs a network with fixed, random connections, simplifying the architecture as compared to traditional neural networks. Historically, quantum computing has been considered the “holy grail” of polaritonics, particularly in terms of the applicability of its outcomes. This is why the concept of reservoir computing has been highly regarded, as it extends its utility from classical to quantum computing domains 11 . Complementing this quantum focus, recent advances of polaritonics might bring important applications in classical information processing, particularly in pattern recognition. Reflecting the latter statement, the authors of ref. 4 leveraged the polariton properties to achieve a 93% level in recognizing handwritten digits in the Mixed National Institute of Standards and Technology dataset (MNIST dataset) 12 , 13 , that is a benchmark in pattern recognition tasks. Complementing this, ref. 5 reported not only a 96.2% classification accuracy on the same dataset in the experiment but also demonstrated the efficiency of backpropagation training in their exciton-polariton-based neuromorphic hardware. This efficiency in processing complex patterns, however, often comes with increased computational and memory demands in traditional neuromorphic networks 14 , which necessitates either a reduction in input resolution or more complex processing architectures for feasible operation times. In contrast, the introduction of binarized neural networks, which streamline the network by utilizing two-level activations or weights and performing simple binary operations, marks a significant advancement in the field. These networks are distinguished by their enhanced speed and energy efficiency, with only a minimal trade-off in inference accuracy 15 . By efficiently using memory to store binary rather than continuous data and simplifying computational demands, binarized networks offer a compelling alternative to conventional networks with continuous activation functions. In the context of high-speed neuromorphic systems, binary networks have shown a significant progress. Recent advancements in the area of neuromorphic binarized polariton networks have showcased their remarkable capabilities. As shown in ref. 15 , this approach, involving input encoding using nonresonant picosecond laser pulses to excite localized condensation sites, each representing binary inputs with distinct pulse energies, has led to significant achievements in pattern recognition. Notably, the system achieves approximately 96% classification accuracy on the MNIST dataset, even in a noisy experimental environment, using a single-hidden-layer network. This level of accuracy, attained through binary operations, is particularly impressive given the challenges of the experimental setup. In scenarios where information is encoded through individual or paired pulses, there’s a marked tendency to opt for a sequential processing approach, employing a single transformation gate for each pulse set. However, this approach presents challenges in terms of operational speed within these binary systems. A tool for overcoming the issue of parallelizing input uploading in neuromorphic binarized polariton networks has been proposed in a recent study 16 . Their method involves spatial encoding of input information, with all input pulses designed to arrive at the network simultaneously. By enabling parallel input encoding, this method effectively addresses the operational speed limitations inherent in the sequential approach. The next natural step towards parallelizing neuronal triggering, in conjunction with ensuring the interaction among individual neurons, is the use of spatial lattices of neurons. Lattices of mesoscopic coherent condensates of exciton polariton have evolved as a sophisticated extension of the principles observed in chains and lattices of ultracold atoms 17 – 21 harnessing the unique properties of light-matter interactions to delve into new frontiers of quantum simulations and condensed matter physics 22 – 25 . Advancing beyond their predecessors, they offer greater control and versatility, operating at higher temperatures and allowing for more dynamic configurations, thereby enhancing their practicality in exploring complex quantum phenomena 23 . Various techniques have been employed to manipulate the spatial potential for trapping and arranging polaritons in a microcavity plane 22 . Among these techniques are etching lattices of coupled micropillars from planar microcavities 26 , variation of the thickness of the cavity layer 27 , deposition of metallic films onto the surface of the microcavity 28 . An alternative approach to creating polariton lattices in a microcavity plane is by using regular spatial patterns of the pumping light. In this geometry, each condensate in a lattice is created by a separate nonresonant optical pump beam, while spatial coherence between the condensates across the lattice is provided by the exchange of ballistically propagating polaritons 29 . Besides replenishing the polariton condensate state, the pump also facilitates their trapping, contributing to the formation of an effective complex potential for the trap. Depending on the combination of kinetic properties of the polaritons and the gain-loss balance, either dissipative trapping can occur, where the condensates are predominantly localized within the pump spot 24 , 29 , 30 (phase locking regime), or they are localized in the minima of the real part of the effective potential created by the pumping light 31 , 32 . The complex effective potential is formed by the repulsive reservoir of hot incoherent excitons, which are excited by the pump light. Spatial light modulators (SLMs) offer an advantage of high spatial resolution and the ability to control the intensity and distribution of the reservoir with great accuracy, enabling precise manipulation of the locations of polariton condensates and their coupling strengths within the lattice. Additionally, these optically induced potentials can be effectively combined with stationary potentials, offering even more versatility of control over polariton condensates 33 – 36 . In this manuscript, we present a neuromorphic network architecture employing lattices of exciton-polariton condensates, interconnected and energized through nonresonant optical pumping. This design capitalizes on benefits of a binary framework, wherein each neuron, aided by the spatial coherence of coupled condensates, executes binary operations. The lattice structure facilitates parallel uploading and processing of signals, enhancing neuron-neuron interactions. The effectiveness of our system was evaluated using the MNIST dataset. It demonstrated promising results compared to current neuromorphic systems. Additionally, we have developed a technique for input signal densing that notably improves the system performance, that achieves an accuracy rate of 97.5%, surpassing that of existing polariton-based neuromorphic systems. To further validate our model’s robustness and versatility, we extended our testing to the Speech Commands dataset 37 , 38 , which includes audio clips for voice command recognition. Our system’s performance, considerably exceeding that of both the linear classifier and the advanced Hidden Markov Model with Gaussian Mixture Model (HMM-GMM) 39 , 40 , confirms its potential for effective application in complex speech recognition tasks.",
"discussion": "Discussion Polaritonic neural networks, both previously proposed and presented in this work, are constructed based on a single hidden neuronal layer. In this context, the nature of neuronal connections plays a crucial role in determining the accuracy of such a network. In ref. 15 , optical XOR gates function as neurons in the hidden layer. Input to each gate consists of a pair of optical pulses encoding two random pixels from a binarized initial image, and the output is the result of nonlinear interaction of these pulses. Thus, in this architecture of the neural network, nonlinearity is achieved at the level of interaction of the input layer neurons, while the triggering of neurons in the hidden layer merely reflects the outcome of these interactions. In the architecture without input signal densing, proposed by us, input neurons do not interact with each other. Nonlinearity is achieved at the level of interaction between the input neurons and the neurons in the hidden layer. Thus, the neurons in the hidden layer not only transmit the result of the interaction but are themselves the subjects of this interaction. Our proposed neural network architecture is highly advantageous as it allows for the integration of both approaches to facilitate nonlinear interactions among neurons of different layers. The integration was realized through the introduction of an input signal densing procedure, leading to record-breaking level of predicted accuracy in polaritonic neural networks, substantially surpassing those of previously proposed architectures. The path towards further enhancing the accuracy of our system remains open. Due to the flexible geometry of the lattice and the genuinely nonlinear nature of the interactions among the lattice-forming polariton condensates, there is potential to increase accuracy by involving a larger number of neurons from the hidden layer in contact with input signals. Additionally, a modification of the nature of neuronal interactions, such as by replacing the OR operation with a XOR operation in the hidden layer, could also contribute to the improvement of accuracy. To address the challenge of scaling polariton neural networks beyond current limitations, we propose several strategies. These strategies, detailed in Sec. S7 of the Supplementary Information, include tiling multiple SLMs, utilizing optical waveguides, implementing advanced micro-optics, and employing dynamic reconfiguration techniques. By using multiple SLMs placed adjacent to each other, one can create a larger composite grid that significantly increases the addressable area for the pump beams. This modular approach allows for the expansion of the network by simply adding more SLMs, with careful alignment and synchronization ensuring seamless operation across the entire grid. Optical waveguides can be integrated into the sample to direct light precisely to designated spots, overcoming the spatial limitations of free-space optics. These waveguides can be fabricated using advanced lithography techniques to create efficient light paths with minimal loss, enabling the creation of larger networks without increasing the physical footprint. Incorporating advanced micro-optics, such as microlens arrays, allows for precise focusing and directing of pump beams. This enhances the density of pump spots within the available area, enabling more neurons to be addressed simultaneously. Techniques like diffractive optical elements can further optimize the beam distribution and intensity profile. Dynamic reconfiguration techniques enable a single SLM to sequentially address different regions of the sample at high speeds or to reproduce different configurations of the neural network on the same region. This time-multiplexing approach effectively increases the number of controllable neurons without increasing the physical size of the SLM or the sample, by dynamically reconfiguring the pump spots or the neural network fragments. Each of these approaches is designed to overcome spatial constraints and enhance the efficiency of pump beam delivery, thereby enabling the creation of larger and more complex polariton neural networks. For more detailed descriptions and technical aspects of these strategies, refer to the Supplementary Information. The main advantage of the polaritonic part of the neural network is its unprecedented speed, with processing times of only a few tens of picoseconds required for the generation and evolution of polariton condensates in response to laser pump pulses. Delays caused by electronic components are inevitable for neural networks of any nature during data loading and result reading stages. However, all other stages in a polaritonic network can be realized without electronic components. In each specific experiment, the input signal transformation masks, once set initially, remain unchanged throughout the experiment. This means that while the input signals change, the optical paths of individual signals remain constant. Among the operations mentioned, randomization and expansion are linear operations that can be described as matrix-vector multiplications. These operations can be optically implemented, as mentioned in 4 , 5 and detailed in 47 , 48 . An input light field representing the vector of input signals passes through an optical element that applies a spatially varying phase or amplitude modulation, effectively performing a matrix-vector multiplication at the speed of light. The signal densing operation is more complex, but solutions exist for its optical implementation as well. Our studies on the dependence of neuron performance on signal pulse intensity show that, with other parameters fixed, the intensity of signals can vary within several folds while ensuring correct neuron switching and not compromising the isolation of dyads. In this mode, signal densing can be achieved by simply combining multiple pulses at a single point on the sample, resulting in a proportional increase in the intensity of the resulting pulse. Such a combination can be realized using optical components like beam splitters and combiners. Another approach involves temporal multiplexing, where the steps of the signal densing process are spread out over time. This technique is similar to the dynamic reconfiguration strategy discussed for network scaling, where different steps of the process occur sequentially in time but within the same spatial region. By carefully timing these steps, we can ensure that the combined effect of the pulses is achieved without significant delay. Even with an increase in the resulting operation time to hundreds of picoseconds, this approach still maintains a substantial speed advantage over electronic neural networks. Transitioning to the aspects of practical realization of the proposed architecture, it becomes essential to consider the possible experimental realization of polariton condensate lattices designed above. Spatially-distributed systems comprising chains and lattices of interconnected quantum entities have gained recognition as platforms for information storage, transmission and processing, as well as simulators of complex phenomena 20 , 22 , 49 – 51 . Polariton lattices, a recent breakthrough in spatially-distributed quantum systems, are notable for their exceptional spatial coherence 24 , 25 , 29 , 30 . This coherence significantly exceeds that of individual condensates and it facilitates phase locking of nodes across the entire lattice. This widespread phase synchronization could potentially enhance the nonlinearity of interactions essential for neuromorphic network operation, see, e.g., 4 . However, this scenario leads to the loss of a key advantage emphasized in our work: the feasibility of neuron response binarization would be significantly compromised, impacting computational speed and resource efficiency. To address this issue while retaining the advantage of polariton lattices, which is the optical controllability of connections between condensates, we propose ensuring pairwise interactions of condensates within the lattice by selectively severing non-contributing links. For this purpose, a variety of approaches exists. The first possible approach involves optical induction of potential barriers for ballistic polaritons within a planar microcavity, similar to the excitation of the polariton lattice itself as well as the input signals. For this purpose, the nonresonant excitation of the exciton reservoir can be used 52 . We propose to employ it for both the polariton lattice and the input signals generation. However, it should be noted that in this scenario, the pump will contribute not only to the separation of the condensates but also to the changes in their occupation numbers. This factor can be mitigated by using a reservoir of dark excitons as a barrier 53 , 54 . Dark excitons do not participate in optical interactions and do not directly influence the population of polariton condensates. Meanwhile, the strong repulsive nature of polariton-exciton interactions is equally characteristic of bright and dark excitons. A recent paper 54 demonstrates the feasibility of excitation of a dark exciton reservoir through the two-photon absorption. Given that this approach results in record-long exciton lifetimes, over 20 ns, it suggests that such a reservoir would not contribute to replenishing the polariton condensate. An alternative approach, described in ref. 55 , consists in the separation of condensates within the lattice through the creation of spatially varying dissipation profiles by controlling the decay rates of polaritons at different lattice sites. Among the experimental methods mentioned in 55 , one is proton implanting into quantum wells, which enables independent control of exciton and cavity photon energies, influencing polariton decay rates 56 . Additionally, controlled stress applied to the substrate can create spatial traps, affecting the coupling of exciton and photon states, thus varying polariton lifetimes 57 . For dynamic dissipation control, electrical carrier injection can be used, causing localized losses through the absorption by excited states 58 . One can also exploit the biexciton formation regime to alter polariton interactions 59 . As numerical simulations illustrated in Fig. 1h, l show, both potential landscape profiling and dissipation control are comparably effective tools of the condensate-condensate coupling control. The choice of the appropriate method then would depend primarily on the experimental capabilities available and the specific goals set for the experiment. The previously described approaches for the pairwise coupling of condensates in a lattice primarily involved operations within a planar microcavity. However, traditional methods of structuring cavities, such as deep etching techniques, offer alternative avenues for exploration, allowing, e.g., for the creation of clusters 60 and chains 61 of micropillars. Structures, crafted through precise etching processes, may form a distinct spatial arrangement within the microcavity plane. Another approach for clustering polariton condensates within a lattice, detailed in ref. 33 , involves creating controllable Josephson junctions of the condensates. This is achieved through nanostructuring of cavity mirror surfaces via direct laser writing 62 , creating local potential minima, and dynamically tuning the potential landscape using a thermo-responsive polymer affected by a heating laser to vary the optical medium’s refractive index. This method allows for controlled polariton tunnelling between condensates, facilitated by finite height potential barriers, leading to the formation of condensates in a thermo-optically adjustable potential landscape. Our investigation into binary polariton neural networks has demonstrated their fundamental operational principles and potential for energy-efficient computing, suitable for certain applications where high-speed and low power consumption are critical. Energy consumption and efficiency analysis of the proposed system are discussed in Sec. S8 of the Supplementary Information. While binary networks inherently trade off precision for efficiency, our results with a basic logistic regression model on the MNIST dataset have shown promising accuracy levels. Moreover, the proposed network architecture, while simple, suggests several avenues for enhancement that could address tasks requiring higher precision. We propose that extending the network’s complexity through additional hidden layers could amplify its computational power. To address the challenge of integrating multiple hidden layers in a polaritonic neural network, the output signals from one hidden layer can be utilized as a template for generating input signals for the next layer. This approach may involve electronic components for signal regeneration or amplification, which, while enhancing accuracy, compromises the system’s computations speed advantage. Alternatively, integrated optical waveguides for polaritons can link neurons across hidden layers. These waveguides utilize the bistability effect, where a polariton condensate in a low-intensity state is switched to a high-intensity state by an input trigger 8 . The hidden layer photoluminescence signals can serve as these triggers, effectively generating signals in neurons at the opposite ends of the waveguides in subsequent layers. This approach maintains the speed advantage by avoiding electronic mediation and capitalizes on the intrinsic properties of polaritons for rapid signal transmission. This approach, although requiring adjustments to the hidden layer geometry, offers a feasible and efficient solution for multi-layer integration in polaritonic neural networks. Adjusting the network’s geometric layout from a square to a hexagonal lattice could further improve its performance by increasing the number of neurons each input signal can potentially activate, thus enhancing the structural nonlinearity of the network. Moreover, replacing the OR gate response with other types of logical operations, such as XOR gates, could add another layer of nonlinearity. An XOR gate could toggle the state of each neuron in response to paired inputs without changing the overall parity, introducing a dynamic component to the neuron’s response. We also propose the idea of accommodating continuous input signals alongside binary outputs. This approach would utilize the varying intensities of signals, such as those from grayscale images, to modulate input signals, enabling a more nuanced response based on pixel brightness. By doing so, we could significantly enrich the input feature space without compromising the inherent advantages of binary systems, such as low memory usage and high processing speeds. Finally, exploring controlled interactions between neurons—what is often considered a parasitic effect of crosstalk—could be harnessed to advantageous effect. By fine-tuning the isolation and interaction among neuron pairs, we could potentially enhance the network’s robustness and precision. In summary, we have developed a neuromorphic network architecture leveraging lattices of exciton polariton condensates. The design takes advantage of a binary framework, where each neuron, facilitated by the spatial coherence of pairwise coupled condensates, performs binary operations. This coherence ensures efficient network-wide communication, with the binary neuron switching driven by nonlinear repulsion through the excitonic component of polaritons. The binary nature of a network offers computational efficiency and scalability advantages, setting this system apart from conventional continuous weight models and sequential binary neuromorphic systems. The network’s effectiveness was demonstrated using the MNIST dataset for handwritten digit recognition. Our network has not only shown competitive performance against existing systems, but also surpassed them when taking advantage of the original signal densing technique. The developed approach allowed the network to achieve a remarkable 97.5% classification accuracy, theoretically. Further validation was conducted on the Speech Commands dataset, which contains diverse and complex one-second audio clips of spoken words. This additional testing phase highlighted the adaptability and robustness of our architecture in processing intricate audio data and handling a variety of speech recognition tasks. By employing a binary operational framework and exploring various lattice structuring techniques, this study opens new pathways for developing efficient, scalable, and high-speed neuromorphic systems. We are confident that polaritonic systems have high potentiality as creating powerful tools for complex pattern recognition and data processing tasks."
} | 6,847 |
25028422 | PMC4161255 | pmc | 8,365 | {
"abstract": "ABSTRACT Soil microbial diversity represents the largest global reservoir of novel microorganisms and enzymes. In this study, we coupled functional metagenomics and DNA stable-isotope probing (DNA-SIP) using multiple plant-derived carbon substrates and diverse soils to characterize active soil bacterial communities and their glycoside hydrolase genes, which have value for industrial applications. We incubated samples from three disparate Canadian soils (tundra, temperate rainforest, and agricultural) with five native carbon ( 12 C) or stable-isotope-labeled ( 13 C) carbohydrates (glucose, cellobiose, xylose, arabinose, and cellulose). Indicator species analysis revealed high specificity and fidelity for many uncultured and unclassified bacterial taxa in the heavy DNA for all soils and substrates. Among characterized taxa, Actinomycetales (Salinibacterium), Rhizobiales (Devosia), Rhodospirillales (Telmatospirillum), and Caulobacterales ( Phenylobacterium and Asticcacaulis ) were bacterial indicator species for the heavy substrates and soils tested. Both Actinomycetales and Caulobacterales (Phenylobacterium) were associated with metabolism of cellulose, and Alphaproteobacteria were associated with the metabolism of arabinose; members of the order Rhizobiales were strongly associated with the metabolism of xylose. Annotated metagenomic data suggested diverse glycoside hydrolase gene representation within the pooled heavy DNA. By screening 2,876 cloned fragments derived from the 13 C-labeled DNA isolated from soils incubated with cellulose, we demonstrate the power of combining DNA-SIP, multiple-displacement amplification (MDA), and functional metagenomics by efficiently isolating multiple clones with activity on carboxymethyl cellulose and fluorogenic proxy substrates for carbohydrate-active enzymes.",
"introduction": "INTRODUCTION Soil microorganisms catalyze Earth’s biogeochemical reactions, including the degradation of organic matter and recycling of nutrients. Soils host diverse microhabitats with varied physicochemical gradients and environmental conditions. In this context, soil microorganisms live in consortia, interacting physically and biochemically with other members of the soil biota ( 1 ). Attesting to the heterogeneity, interactivity, and connectivity of the soil niche, traditional culture-based techniques grossly underestimate microbial diversity. Readily cultured microorganisms typically represent a very small proportion of soil microbial communities ( 2 ); the “uncultured majority” harbor an enormous reservoir of uncharacterized organisms, genes, and enzymatic processes ( 3 ). An outstanding methodological question remains: how best to access the biotechnological potential contained within the DNA of soil’s uncultured microorganisms? Degradation of plant organic matter by the combined action of glycoside hydrolase (GH) enzymes is an important soil function. The GH group of enzymes is distributed across a wide variety of organisms. They catalyze the hydrolysis of glycosidic bonds in complex carbohydrates (e.g., cellulose and hemicellulose) to release simple sugars (e.g., pentoses and hexoses), and as a result, GHs include important enzymes for biotechnological applications. Because glycosidic bonds are considered among the most stable linkages that occur naturally, GHs are credited as some of the most proficient catalysts ( 4 ). Recent research suggests a broad diversity of bacteria contribute to plant polymer degradation ( 5 – 8 ), supporting the use of cultivation-independent methods, such as metagenomics, as most strategic for the recovery of genes and enzymes from these microorganisms. Metagenomics captures the genomes of environmental community microbes, circumventing the need for cultivation and enabling the exploration of microbial genetic diversity and biotechnological potential ( 9 ). Metagenomic analyses have exposed new microbial pathways and reactions, yielding novel enzymes and products of economic importance. Given that metagenomic studies demonstrate that the majority of total genetic diversity space remains unexplored, “it will be far more efficient and productive to seek new enzymes from metagenome libraries than to tweak the activities of existing ones” ( 10 ). Indeed, there are several recent examples of GHs (e.g., cellulases) recovered by functional screening of metagenomic libraries from terrestrial environments (e.g., see references 11 , 12 , 13 , and 14 ). These studies reflect a laborious limitation of bulk DNA metagenomic library construction: in the absence of suitable selections for phenotype, many clones (e.g., tens of thousands) must be screened prior to recovering targets of interest. In addition, recovered clones are theoretically the most abundant target genes in the microbial community of interest. Targeted metagenomic approaches, such as those involving an enrichment culture step ( 15 ), thus offer the potential to filter for sequences specific to an activity of environmental or industrial relevance. Stable-isotope probing (SIP) is a culture-independent method for targeting microorganisms that assimilate a particular growth substrate ( 16 – 18 ). For the analysis of genomic DNA of active organisms, a SIP substrate (e.g., 13 C labeled or 15 N labeled) is incorporated into the DNA (DNA-SIP) or RNA (RNA-SIP) of active organisms, and isopycnic ultracentrifugation can differentiate labeled nucleic acids from an abundant background of unlabeled community genomes. Combining SIP with metagenomics provides access to the genomes of less-abundant community members and offers insight into complex environmental processes, such as biodegradation (as reviewed in references 19 , 20 , and 21 ). Several studies have combined DNA-SIP and metagenomic sequencing to identify high proportions of genes from active microorganisms, such as those using glycerol ( 22 ), C 1 compounds ( 23 – 26 ), and biphenyl ( 27 , 28 ). Previous SIP studies reported that in an agricultural soil (clay loam soil, pH 6.6), cellulose was metabolized by Bacteroidetes , Chloroflexi , and Planctomycetes ; cellobiose and glucose were degraded predominantly by Actinobacteria ( 8 ). The results also suggested that cellulolytic bacteria are different from saccharolytic bacteria and that oxygen availability defined the different taxonomic groups involved. Under anoxic conditions, cellulose was metabolized by Actinobacteria , Bacteroidetes , and Firmicutes ; carbon from cellobiose and glucose were assimilated by Firmicutes . Others found that members of the Burkholderiales , Caulobacteriales , Rhizobiales , Sphingobacteriales , Xanthomonadales , and Group 1 Acidobacteria were associated with three different soils amended with cellulose ( 29 ). A recent survey of active bacteria in an Arctic tundra sample found Clostridium and Sporolactobacillus involved in 13 C-glucose assimilation and Betaproteobacteria , Bacteroidetes , and Gammaproteobacteria involved in the assimilation of carbon derived from 13 C-cellulose ( 30 ). Others have used SIP and labeled cellulose to identify Dyella , Mesorhizobium sp., Sphingomonas sp., and an uncultured deltaproteobacterium (affiliated with Myxobacteria ) linked to cellulose degradation ( 6 ). The ability to identify genes based on function, instead of sequence homology, is arguably the most powerful application of metagenomics for the recovery of novel genes ( 31 ) and a natural partner of the SIP approach for targeting active-yet-uncultured microorganisms ( 21 ). Previous studies were focused on the analysis of single substrates or individual samples. In addition, only one previous study combined SIP and functional metagenomic screens, expressing labeled DNA within a surrogate Escherichia coli host for identification of enzyme activity ( 22 ). In this study, we expand on previous efforts to combine SIP and metagenomics (as reviewed in reference 21 ), enriching soil microorganisms active in degrading plant-derived carbohydrates and screening GHs through activity-based functional metagenomics. We combined SIP, high-throughput sequencing of labeled 16S rRNA genes and metagenomic DNA, multiple-displacement amplification (MDA), and functional metagenomics to identify active microorganisms and associated GH enzymes. We also isolated GH-positive clones from a cosmid library in a much higher frequency than would be expected with traditional efforts using conventional metagenomics.",
"discussion": "RESULTS AND DISCUSSION Characterization of active soil bacteria. We used DNA-SIP as a targeted approach for enriching active soil microorganisms involved in the metabolism of five plant-derived carbohydrates (glucose, cellobiose, xylose, arabinose, and cellulose). Three disparate soil samples were obtained from the CM 2 BL soil collection based on maximal physicochemical diversity ( Table 1 ) ( http://www.cm2bl.org/ ). In particular, soil pH was low for the Arctic tundra and temperate rainforest soil samples, suggesting that the microbial composition and diversity of these two samples would be fundamentally different from those in agricultural soil ( 32 , 33 ). The water-filled pore space (WFPS) was maintained between 50% and 60% to avoid decreased aerobic microbial activity at WFPS values of >60% ( 34 , 35 ). TABLE 1 Location and physicochemical characteristics of the soil samples selected for DNA stable-isotope probing incubations a Sample Location Latitude and longitude Bulk density (g/cm3) Amt of carbon (% dry wt) pH Moisture (% dry wt) Amt of nitrogen (% dry wt) Soil type Total Inorganic Organic Arctic tundra (1AT) Daring Lake, North-West Territories, Canada 64°52′N, 111°35′W 0.2 46.9 BDL b 46.9 3.9 417.7 1.42 Organic Temperate rainforest (7TR) Pacific coastal rainforest, Vancouver Island, Canada 48°36′N, 124°13′W 0.6 10.8 BDL 10.8 4.9 69.8 0.35 Coarse sandy loam Agricultural soil-wheat (11AW) Elora Research Station, Ontario, Canada 43°38′N, 80°24′W 1.1 1.85 0.12 1.7 7.4 17.9 0.19 Silt loam a For more details, see http://www.cm2bl.org/ . b BDL, below detection limit. Because 13 C-labeled cellulose was commercially unavailable at the time of this research, both native cellulose and 13 C-labeled cellulose were produced as the substrates for SIP incubations by Gluconacetobacter xylinus , generating predominantly amorphous cellulose ( 36 ), which is more readily degraded than crystalline cellulose ( 37 ). To ensure detectable labeling, similar to a previous experimental approach ( 8 ), glucose, cellobiose, arabinose, and xylose were added weekly (1.5 mmol of C) for 3 weeks, reaching levels approximately 5 to 500 times higher than those normally detected in soils ( 38 , 39 ). Although substrate concentrations were higher than typical bulk soil concentrations, higher polysaccharide substrate concentrations would be expected in the root rhizosphere and in areas of active plant matter decomposition (as reviewed in reference 39 ), suggesting that our incubation conditions would not be unrealistic for some naturally occurring soils. These concentrations were chosen to ensure that labeled isotope was more abundant than endogenous soil carbon sources for the success of DNA-SIP, enabling the separation and purification of labeled DNA for subsequent molecular analyses ( 16 , 40 ). Similar substrate concentrations and incubation times with glucose and cellulose were used previously ( 30 ), demonstrating minimal-yet-detectable labeling of DNA in an Arctic tundra soil sample. Metabolism of labeled substrates in DNA-SIP incubations was confirmed by higher headspace CO 2 production in all substrate-amended serum vials compared to uninoculated controls for each of the three soils ( Fig. 1 ). In all cases, cellulose-amended vials demonstrated reduced CO 2 production compared to the other substrates, further justifying an extended incubation time for this comparably recalcitrant substrate. The average amount of CO 2 released after 6 days was 13% of the headspace, which, after subtraction of the average CO 2 produced in uninoculated vials, was approximately equivalent to 1.4 mmol of carbon. This represents 93% of the total weekly carbon added (~1.5 mmol of carbon). FIG 1 Carbon dioxide production for Arctic tundra (1AT) (A), temperate rainforest (7TR) (B), and agricultural (11AW) (C) soils. Soil samples were amended with labeled ( 13 C) or unlabeled ( 12 C) substrates, and serum bottles were aerated weekly to replenish oxygen and deplete carbon dioxide. The “control” represents a soil sample incubated without substrate. In addition to monitoring CO 2 production in all vials, separate soil incubations were prepared with a defined helium-oxygen headspace and glucose amendment in order to monitor O 2 consumption. As expected, the addition of glucose stimulated O 2 consumption, but the headspace remained oxic for each of the weekly incubation periods over the first 3 weeks (see Fig. S1 in the supplemental material), indicating that weekly aeration of experimental vials was sufficient to deplete CO 2 and replenish O 2 . Maintaining oxic conditions was important to ensure that the DNA-SIP incubation recovered DNA from microorganisms involved in aerobic degradation of complex carbohydrates in addition to capturing DNA from microorganisms involved in anaerobic metabolism ( 41 ). Indeed, recent oxic incubations demonstrated activity of anaerobic clostridia ( 8 , 30 , 42 ), presumably because anoxic microenvironments exist even within oxic experimental microcosms. Confirmation of isotope labeling. At the two time points of all incubations (1 and 3 weeks for all substrates, except for cellulose, which was sampled at 3 and 6 weeks), DNA was retrieved for the analysis of bacterial community composition by agarose gel electrophoresis and denaturing gradient gel electrophoresis (DGGE) ( 43 ). All DNA extracts from microcosm soils were subjected to density gradient ultracentrifugation and recovered in 12 fractions, which were analyzed in agarose gels. The results demonstrated that all soils possessed more DNA in 13 C-incubated heavy fractions (i.e., 1 to 7) than in 12 C-control fractions (i.e., 8 to 12) from glucose, cellobiose, arabinose, and xylose SIP incubations (see Fig. S2 to S6 in the supplemental material). For cellulose, only temperate rainforest and agricultural soil incubations resulted in heavier DNA visible in agarose gels corresponding to 13 C-labeled sample heavy DNA fractions (see Fig. S6 ) for the 6-week time point. Similar results were observed for all earlier time points but with less DNA associated with heavy fractions for 13 C-incubated samples compared to the later time points (data not shown). Although extended incubation times were important, one caveat of extended incubation times for SIP incubations (e.g., for cellulose) is that labeled carbon might have been distributed more broadly within the microbial community, which may result in less-specific enrichment of substrate-degrading microbial genomes in the resulting data and libraries. The presence of distinct fingerprint profiles in heavy fractions for 13 C-incubated samples, but not for the corresponding 12 C-control fractions, demonstrates isotopic enrichment of nucleic acids ( 16 ). Bacterial DGGE fingerprints corresponding to all late-time-point fractions demonstrated unique patterns associated with the heavy fractions (e.g., fractions 1 to 7) for all 13 C-incubated SIP microcosms (see Fig. S2 to S6 in the supplemental material). Although some cross-gradient fingerprint variations were associated with 12 C-control DNA, these differences were likely GC content shifts because they were pronounced only in the lightest fractions (e.g., fractions 10 to 12) and were distinct from shifts associated with fractionated 13 C-DNA. Substrate- and soil-specific heavy fraction patterns were consistent for early- and late-time-point samples (data not shown), which indicated that detected active bacteria were stable over time rather than changing due to food web dynamics ( 40 ). Heavy DNA fingerprints were used to identify fractions containing 13 C-labeled DNA for subsequent 16S rRNA gene sequencing, bulk DNA sequencing, and functional metagenomics. Based on DGGE patterns, we identified fraction 5 and/or 6 as being representative of heavy DNA and fraction 10 as representing light DNA for all soils, substrates, and incubation times (see Fig. S2 to S6 in the supplemental material). Although fractions 1 to 5 also may have captured DNA from labeled microorganisms, these fractions were not analyzed further because the vanishingly small proportions of DNA recovered from these gradient fractions would have made PCR and subsequent metagenomic library preparation problematic. Taxonomic characterization of heavy DNA. We selected representative gradient fractions from all soils, substrates, and incubation times for profiling of the bacterial V3 region of 16S rRNA genes. Based on DGGE data, we selected fractions 6 (heavy) and 10 (light) for Arctic tundra and fractions 5 (heavy) and 10 (light) for temperate rainforest and the agricultural soil. In addition, we sequenced V3 regions of 16S rRNA genes from DNA extracted from the initial soil samples used to establish SIP incubations to determine whether light fractions resembled the original soil community as would be expected. Following paired-end-read assembly, we analyzed 630,000 assembled sequences (10,000 sequences per sample) using an AXIOME management of the QIIME pipeline and additional custom analyses (e.g., multiresponse permutation procedure [MRPP] and indicator species analysis). Good’s coverage ( 44 ) for the heavy fraction samples ranged from 84 to 92%, and light fraction samples ranged from 68 to 85%, which indicates that this level of sequencing captured the majority of bacterial taxa in these samples. β diversity was assessed by weighted UniFrac distances visualized within principal coordinate analysis (PCoA) plots. The results indicated that all samples from within each of the three soil treatments were grouped distinctly according to soil type ( Fig. 2A ), which was highly significant based on MRPP analysis ( A = 0.18 [chance-corrected within-group agreement], T = −20.4 [test statistic], P < 0.001). Both the Arctic tundra and temperate rainforest soil profiles grouped more closely with one another, which is likely a result of both soils sharing low pH ( Table 1 ), a major determinant of soil bacterial diversity and taxonomic composition ( 45 , 46 ). In addition, all heavy and light fraction profiles for the three soils were clustered distinctly ( Fig. 2A ), which was also highly significant ( A = 0.40, T = −28.3, P < 0.001). Native soil phylogenetic profiles clustered with their respective light fractions, indicating that the “background” bacterial community remained relatively constant throughout the SIP incubation. Although the two time points for some 13 C-labeled substrates grouped together ( Fig. 2B ), the differences between heavy and light fractions were much greater than those observed between the five substrates used in this study. FIG 2 Principal coordinate analysis (PCoA) biplots of weighted UniFrac distances for 16S rRNA gene sequences generated by assembled paired-end Illumina reads. Samples separated by soil type and fraction (A) as well as by carbon source (B). Native soils were associated with their respective light fractions. Gray spheres represent taxonomic affiliations of OTUs that correlated most strongly within the ordination space. Many operational taxonomic units (OTUs) were affiliated with SIP-derived heavy DNA, but multiple permutations of the analysis were required to summarize indicator OTUs for different sample subsets. We used indicator species analysis ( 47 ), with an indicator value (IV) threshold of 70% and a >250 minimum sequence sum threshold to identify the strongest significant OTUs ( P < 0.01) associated with (i) all heavy DNA samples (versus all light DNA samples) ( Fig. 3 ; see Table S1 in the supplemental material), (ii) all heavy DNA samples within each soil type (versus all light DNA samples for the same soil type) (see Table S2 in the supplemental material), (iii) each substrate across all heavy DNA samples from all soil types (versus the heavy DNA for the other substrates from all soil types) (see Table S3 in the supplemental material), and (iii) each substrate from heavy DNA within each soil type (versus the other substrates for the same soil type heavy DNA) (see Tables S4 to S6 in the supplemental material). FIG 3 Cleveland plot of operational taxonomic unit (OTU) abundance for OTUs possessing the highest indicator values (i.e., >70%) for an association with DNA-SIP heavy DNA (black squares [average abundance]) for all substrates and soils combined, in comparison to light DNA (gray squares [average abundance]). Taxonomic affiliations are included for phyla, with additional classifications for order (o_), family (f_), and genus (g_). For additional details, see Table S1 in the supplemental material. When we compared OTUs associated with all heavy DNA samples versus all light DNA samples from all soils, indicator species analysis revealed multiple poorly classified indicators, in addition to genus-classified OTUs associated with the Salinibacterium ( Actinobacteria ), Devosia ( Alphaproteobacteria ), Telmatospirillum ( Alphaproteobacteria ), Phenylobacterium ( Alphaproteobacteria ), and Asticcacaulis ( Alphaproteobacteria ) genera ( Fig. 3 ; see Table S1 in the supplemental material). The indicator species analysis from all heavy DNA samples versus all light DNA samples within each soil type showed that the predominant genus-classified OTUs identified in heavy fractions from tundra soil (1AT) were Salinibacterium ( Actinobacteria ), Rhodanobacter ( Gammaproteobacteria ), Conexibacter ( Actinobacteria ), Telmatospirillum ( Alphaproteobacteria ), Asticcacaulis ( Alphaproteobacteria ), and Burkholderia ( Betaproteobacteria ), in addition to OTUs within orders such as Sphingomonadales and Acidobacteriales (see Table S2 in the supplemental material). The temperate rainforest soil (7TR) heavy DNA was dominated by OTUs classified to the genera Paucibacter ( Betaproteobacteria ), Burkholderia ( Betaproteobacteria ), Spirochaeta ( Spirochaetes ), Salinibacterium ( Actinobacteria ), Telmatospirillum ( Alphaproteobacteria ), Labrys ( Alphaproteobacteria ), Mesorhizobium ( Alphaproteobacteria ), and Phenylobacterium ( Alphaproteobacteria ), in addition to uncharacterized genera from other phyla, such as Verrucomicrobia (see Table S2 ). The agricultural soil wheat (11AW) heavy DNA OTUs were represented by the genera Pseudomonas ( Gammaproteobacteria ), Devosia ( Alphaproteobacteria ), Pseudoxanthomonas ( Gammaproteobacteria ), Salinibacterium ( Actinobacteria ), Ramlibacter ( Betaproteobacteria ), Ochrobactrum ( Alphaproteobacteria ), Paenibacillus ( Firmicutes ), and Aeromicrobium ( Actinobacteria ) and further unclassified members of the orders Pseudomonadales , Rhizobiales , Xanthomonadales , Actinomycetales , Burkholderiales , and Bacillales (see Table S2 ), among others. Orders associated with the metabolism of cellulose were dominated by Actinomycetales and Caulobacterales (genus Phenylobacterium ) (see Table S3 in the supplemental material). Members of the Alphaproteobacteria were associated with the metabolism of arabinose, and members of the order Rhizobiales were strongly associated with the metabolism of xylose. There were no specific indicator species associated with glucose or cellobiose across all soils (see Table S3 ), which might also suggest that abundant soil OTUs were also active in assimilating these substrates. The predominant indicator species for the agricultural soil fed with [ 13 C]glucose were associated with Paenibacillus ( Bacillales ) (see Table S4 in the supplemental material). The use of cellulose was associated with Mesorhizobium ( Alphaproteobacteria ), Devosia ( Alphaproteobacteria ), and Cellvibrio ( Gammaproteobacteria ), in addition to other poorly classified OTUs from the Sphingomonadales and Actinomycetales . The use of cellulose in temperate rainforest soil was associated with the Myxococcales ( Deltaproteobacteria ) (see Table S5 in the supplemental material). An OTU affiliated with Caulobacterales was associated with the metabolism of glucose in Arctic tundra. Nevskia ( Gammaproteobacteria ), and two OTUs affiliated with the Acidobacteria were associated with tundra cellulose assimilation (see Table S6 in the supplemental material). No other OTUs were significant indicators for the remaining substrates (i.e., cellobiose, arabinose, and xylose) for the three soils, which might indicate that active taxa were also abundant soil bacteria. Although our DNA-SIP incubation revealed many poorly classified indicator taxa (see Tables S1 to S6 in the supplemental material), many of the indicator species associated with heavy DNA were expected based on previous studies. For example, Salinibacterium was associated with frozen soils from glaciers ( 48 ) and Antarctic permafrost ( 49 ). This genus has been associated with the metabolism of a variety of carbon sources, including sucrose, glucose, cellobiose, mannose, melibiose, maltose, galactose, arabinose, and fructose ( 48 , 50 ). In addition, Devosia species were isolated from greenhouse soil and beach sediments, testing positive for the hydrolysis of esculin, β-galactosidase, β-glucosidase, and N -acetyl-β-glucosaminidase, although unable to degrade carboxymethyl cellulose (CMC) ( 51 , 52 ). Phenylobacterium and Burkholderia are abundant in forest soils ( 53 ) and the genus Asticcaulis was identified among aerobic chemoorganoheterotrophs in tundra wetlands, able to use glucose, sucrose, xylose, maltose, galactose, arabinose, lactose, fructose, rhamnose, and trehalose, among other carbon sources ( 54 ). The genus Spirochaeta has some species that are free-living saccharolytic and obligate or facultative anaerobes and were isolated from diverse environments, mainly from extreme aquatic environments ( 55 , 56 ). Spirochaeta americana was reported to be a consumer of d -glucose, fructose, maltose, sucrose, starch, and d -mannitol ( 56 ), and Spirochaeta thermophila was reported to be a cellulolytic organism; the study of its genome revealed a high proportion of genes encoding more than 30 GHs ( 55 ). TABLE 2 Substrate-specific activities of positive metagenomic clones from the [ 13 C]cellulose DNA-SIP library Clone Insert size (kb) Activity (μM MU released) a CMC activity b α- l -Arabinofuranoside pyranoside β- d - Cellobiopyranoside β- d - Glucopyranoside β- d - Xylopyranoside N -Acetyl-β- d - galactosaminide C122 21.6 0.4 0.2 0.6 0.7 124.2 − C424 8.2 0.9 57.6 109.4 1.6 0.7 − C762 13.5 2.4 5.4 21.2 0.7 0.4 − C1024 16.8 123.8 6.5 35.8 1.7 0.5 − C1088 11.9 0.5 25.6 79.2 1.2 0.6 − C2194 12.9 0.5 0.3 0.6 0.4 39.6 − C2380 14.9 0.38 0.46 0.53 0.41 0.40 +++ C2044 14.7 0.40 0.40 0.52 0.39 0.36 ++ a Cellulase activity was scored by Congo red staining of clones on the LB-CMC plate. Other activities were measured in cell-free extracts using methylumbelliferone-based substrates. MU, methylumbelliferone units based on equal volumes of sample for each assay. b CMC, carboxymethyl cellulose. Plate-based clearing (high, +++; medium, ++; negative, −) was detected by Congo red stain and activity based on comparison to those of positive and negative controls. TABLE 3 Analysis of cosmid insert end sequences Clone BLASTx result for a : Forward read Reverse read Description E value (% identity [no. positive/total]) Description E value (% identity [no. positive/total]) C122 Porphyromonas gingivalis (4-amino-4-deoxy- l -arabinose transferase) 4e–5 (29 [40/139]) Cellvibrio japonicus Ueda107 (β-xylosidase) 8e–136 (82 [131/162]) C424 Cellvibrio sp. strain BR (DNA-directed DNA polymerase) 1e–28 (69 [66/80]) Cellvibrio sp. strain BR (Glucuronate isomerase) 2e–103 (91 [157/163]) C762 Chthoniobacter flavus (putative PAS/PAC sensor protein) 1e–86 (78 [151/171]) Sorangium cellulosum (hypothetical protein) 2e–28 (54 [83/125]) C1024 Cellvibrio sp. strain BR (glucuronate isomerase) 2e–17 (95 [34/40]) Cellvibrio sp. strain BR (gluconolaconase) 2e–46 (80 [85/96]) C1088 Saccharophagus degradans (SSS sodium solute transporter superfamily) 6e–61 (68 [123/150]) Cellvibrio sp. strain BR (auxin efflux carrier) 5e–44 (75 [101/114]) C2194 Dyadobacter fermentans (ROK family protein) 1e–91 (95% [140/142]) Failed sequencing reaction C2380 Alicyclobacillus acidocaldarius (Glyoxalase/bleomycin resistance protein/dioxygenase) 2e–15 (52 [51/69]) Cellvibrio sp. strain BR (glucosamine fructose-6-phosphate aminotransferase, isomerizing) 3e–105 (96 [162/163]) C2044 Cellvibrio sp. strain BR (DNA polymerase III subunit delta) 1e–71 (96 [116/118]) Dyadobacter fermentans (hypothetical protein) 9e–129 (97 [181/184]) a Cosmids were end sequenced with M13 forward and reverse primers flanking the site of metagenomic DNA insertion. For each clone, two end sequences were obtained and are referred to as “reverse” and “forward” reads. Top matches for BLASTx analyses are shown. Positive results are the number of amino acids from the query that match the amino acids from the subject sequence. The total number of amino acids from the subject is shown. MG-RAST analysis and functional annotation. We used next-generation sequence analysis of bulk 13 C-labeled DNA to survey the prevalence of annotated GHs within three pooled samples that were targeted for subsequent functional metagenomic screens. Guided by the UniFrac-based PCoA plot ( Fig. 2 ), we pooled heavy DNA samples representing all substrates (except cellulose) associated with low pH (i.e., temperate rainforest, Arctic tundra), heavy DNA for all substrates (except cellulose) from the agricultural soil, and the cellulose-enriched DNA from the three soils. Analysis of paired-end reads was performed by MG-RAST using annotations derived from the Swiss-Prot/Uniprot database. Only 19.4% (low-pH library), 19.6% (cellulose library), and 22.0% (agricultural library) of sequences were annotated by Swiss-Prot in MG-RAST using an E value threshold of 0.01, which is an important consideration for any subsequent analysis of annotation data based on this minority of sequences. Nonetheless, using a custom Perl script to convert Swiss-Prot annotations to CAZy GH identifiers, we detected 81 distinct GH families for the pooled-cellulose library and 80 GH families for each of the low-pH and agricultural soil composite libraries. The distribution of annotated GHs varied between samples, and the most abundant families in the three pooled samples were GH1, -2, -3, -5, -9, -13, -23, -28, and -35 (see Table S7 in the supplemental material). In addition, the three next-generation sequence data sets were very similar in their distributions (i.e., r > 0.99) for the three libraries ( Fig. 4 ), and all had representation among GH families commonly associated with known cellulases (GH1, -3, -5 to -9, -12, -45, -48, and -61), hemicellulases (GH8, -10 to -12, -26, -28, -53, and -74), and debranching enzymes (GH51, -54, -62, -67, -78, and -74) as reviewed elsewhere ( 57 , 58 ). The GH families involved in the hydrolysis of cellulose that were most abundant in our data were GH families 3, 5, and 9 ( Fig. 4 ; see Table S7 ). However, given that most GH family annotations were not represented by known CAZy identifiers and that only ~20% of our paired-end reads were annotated by Swiss-Prot, the abundance and distribution of functional GH families in our pooled DNA is underrepresented. As a result, we used functional screens of large-insert metagenomic libraries for the recovery of GHs to help circumvent the limitations of sequence-based analysis of our heavy DNA samples. FIG 4 Glycoside hydrolase (GH) families associated with pooled heavy DNA. Functional annotation of the metagenomic data revealed diverse GH gene representation within the pooled heavy DNA. Enriched metagenomic library. Pooled high-molecular-weight DNA from the 13 C-cellulose-enriched SIP incubations for the three soils were captured in cosmid libraries and screened for GHs involved in the degradation of cellulose and other plant-derived polymers based on activity in E. coli . Multiple-displacement amplification (MDA) increased the amount of nucleic acids obtained from pooled cellulose DNA-SIP incubations prior to the isolation of 25- to 75-kb DNA fragments via pulsed-field gel electrophoresis (PFGE). The cellulose-SIP metagenomic library contained ~83,000 clones with an average insert size of 31 kb based on restriction digestion of a subset of 40 random clones (data not shown). These results compare favorably to a library of ~10,500 clones generated from MDA-amplified SIP-enriched seawater DNA, which had an average insert size of 27 kb, ranging from 17 to 40 kb ( 26 ). We used a combined parallel approach for functional screening of 2,876 randomly selected clones from the cellulose-enriched metagenomic library. Growth of colonies on LB supplemented with carboxymethyl cellulose (CMC), followed by poststaining with Congo red ( 59 ), facilitated identification of clones expressing either endoglucanase (EC 3.2.1.X) or glucosidase (EC 3.2.1.X) activities ( 60 ). From the 2,876 clones screened, we identified eight positive clones, two of which (C2380 and C2044) were capable of hydrolyzing CMC ( Table 2 ). Restriction mapping showed that these two clones were distinct ( Fig. 5 ). Clones C122 and C2194 carried dissimilar DNAs encoding β- N -acetyl-galactosaminidases (EC 3.2.1.53). β-Glucosidase activities (EC 3.2.1.21) were detected in clones C424, C762, and C1088. Clones C424 and C1088 contained overlapping DNA—probably from the same organism—consistent with the substrate activity profiles. Restriction pattern of clone C1024 was similar to C1088 and C424 ( Fig. 5 ), but C1024 had both α- l -arabinofuranosidase (EC 3.2.1.55) and β-glucosidase (EC 3.2.1.21). The open reading frame (ORF) encoding the β-glucosidase was likely located in the overlapping region. FIG 5 Restriction of cosmid DNA with EcoRI-HindIII-BamHI. DNA sizes in kb are marked on the left and right. M, molecular size markers. The sizes of digested DNA fragments except for the cosmid backbone (the very top band) were added up to obtain the insert sizes of the cloned metagenomic DNA. End sequencing of the positive isolates demonstrated that most clones had at least one end sequence matching the known cellulolytic member of the Gammaproteobacteria, Cellvibrio sp. ( 61 ), with 69 to 95% identity ( Table 3 ). Other top BLAST matches included Saccharophagus degradans , Dyadobacter fermentans , Alicyclobacillus acidocaldarius , and Chthoniobacter flavus ( Table 3 ), with 29 to 97% identity. Although these bacteria are not well characterized to date, other researchers have reported that they use cellulose and other carbohydrates as a carbon source and/or they contain GHs encoded in their genome ( 62 – 65 ). As predicted, the end sequence identities for C424 and C1088 were very similar taxonomically (i.e., Cellvibrio sp.). On the other hand, end sequence data for C122 and C2194 did not suggest a similar genomic origin ( Table 3 ), consistent with the restriction pattern of these cosmids ( Fig. 5 ). Posterior analysis of reverse and forward end sequences of the positive clones was done by comparing end sequences to Illumina forward and reverse reads from whole-genome sequencing of the three SIP libraries (see Table S8 in the supplemental material). The results showed that the majority of end sequences were represented in the cellulose library, as expected, and only a few sequence matches were found in other libraries using the selected threshold. The high frequency of positive clones after screening of DNA-SIP-derived clones compares favorably to those from previous soil functional metagenomic studies reporting the recovery of single positive cellulose hits from screening of tens of thousands of clones. For example, a single cellulose-encoding clone and two xylanase-encoding clones were recovered from functional screening of 13,800 clones from three fosmid metagenomic libraries derived from grassland in Germany, with an insert size range of between 19 and 30 kb ( 11 ). Also, one cellulase-encoding clone was retrieved from the functional screening of 3,024 clones from a bacterial artificial chromosome metagenomic library derived from red soil in China, with insert sizes ranging from 25 to 165 kb ( 12 ). In another study, one cellulase-encoding clone was recovered from functional screening of 14,000 clones with an average insert size of 5 kb from a metagenomic phagemid library from a forest soil in China ( 13 ). Finally, a CMC-positive clone was retrieved from a metagenomic fosmid library derived from wetland soil in South Korea, after screening of 70,000 clones with an average insert of 40 kb ( 14 ). Although not conducted here, a well-replicated direct comparison of GH gene recovery from metagenomic libraries prepared from SIP-derived heavy DNA, light DNA, and the original soil DNA would be necessary to confirm the effectiveness of DNA-SIP. In addition, the ability to recover GH genes in high proportions using cultivation-based enrichment approaches is a well-established alternative to direct metagenomics ( 15 ). DNA-SIP incubations are designed to be less dependent on rapid growth of a readily cultivated subset of the microbial community ( 40 ). Indeed, our labeled DNA contained many OTUs that were classified poorly within described bacterial taxonomies (see Tables S1 to S6 in the supplemental material). Direct DNA-SIP and enrichment culture comparisons would be valuable but have not yet been conducted to our knowledge. In summary, the combination of DNA-SIP and metagenomics helped recover soil GHs in higher proportions than all previously reported efforts via direct metagenomics, which demonstrates the power of using DNA-SIP as an activity-based prefilter for targeted metagenomic approaches. Our study demonstrated the capability of scaling DNA-SIP analysis for the interrogation of multiple environmental samples with multiple substrates, with sampling at multiple time points. A high-quality cosmid library with >31-kb inserts was constructed from heavy DNA originating from a 13 C-cellulose-incubated sample, and highly efficient screening of GHs from a small set of clones (0.3% positive hits) showed strong potential of the techniques combined in this study for functional metagenomics. Identification of the genes encoding GHs and characterization of these enzymes are ongoing and further functional screening of the 13 C-cellulose DNA-SIP library clones in other surrogate hosts will be assessed to identify additional GH representation."
} | 9,788 |
38506512 | PMC11064492 | pmc | 8,366 | {
"abstract": "ABSTRACT Understanding the interactions between microorganisms and their impact on bacterial behavior at the community level is a key research topic in microbiology. Different methods, relying on experimental or mathematical approaches based on the diverse properties of bacteria, are currently employed to study these interactions. Recently, the use of metabolic networks to understand the interactions between bacterial pairs has increased, highlighting the relevance of this approach in characterizing bacteria. In this study, we leverage the representation of bacteria through their metabolic networks to build a predictive model aimed at reducing the number of experimental assays required for designing bacterial consortia with specific behaviors. Our novel method for predicting cross-feeding or competition interactions between pairs of microorganisms utilizes metabolic network features. Machine learning classifiers are employed to determine the type of interaction from automatically reconstructed metabolic networks. Several algorithms were assessed and selected based on comprehensive testing and careful separation of manually compiled data sets obtained from literature sources. We used different classification algorithms, including K Nearest Neighbors, XGBoost, Support Vector Machine, and Random Forest, tested different parameter values, and implemented several data curation approaches to reduce the biological bias associated with our data set, ultimately achieving an accuracy of over 0.9. Our method holds substantial potential to advance the understanding of community behavior and contribute to the development of more effective approaches for consortia design. IMPORTANCE Understanding bacterial interactions at the community level is critical for microbiology, and leveraging metabolic networks presents an efficient and effective approach. The introduction of this novel method for predicting interactions through machine learning classifiers has the potential to advance the field by reducing the number of experimental assays required and contributing to the development of more effective bacterial consortia.",
"introduction": "INTRODUCTION A microbial consortium is a group of different species or strains of microorganisms that interact to execute specific behaviors. Within microbial consortia, different types of interactions occur among microorganisms and their neighbors ( 1 ). In this way, cross-feeding and competition interactions allow for the optimization of parallel metabolic processes of different microorganisms to increase productivity, efficiently consume certain nutrients, or maintain consortium stability against environmental perturbations over time( 2 – 4 ). Cross-feeding interactions may be unidirectional, where one microorganism benefits and grows by utilizing secreted metabolites from another, or bidirectional, involving a reciprocal exchange of metabolites among different microorganisms in the community( 5 – 9 ). Notably, understanding these interactions is key for understanding individual microorganism behavior within a consortium and, even more importantly, for unraveling the collective behavior of an entire community of microorganisms. Microbial consortia engineering (MCE) has recently evolved into an established scientific discipline on its own ( 10 ). The principal objective in this field is to create communities that exhibit stability over time, enhanced productivity of specific metabolites, and improved metabolic functionalities ( 1 , 2 , 11 – 13 ). In essence, MCE seeks to design consortia of microorganisms with specific properties of interest. The most commonly used MCE methodology involves assembling the metabolism of different microorganisms to achieve a particular behavior that promotes interactions among the different cells and their environment ( 12 , 13 ). Other approaches, such as ecology-based models ( 14 ) and ODE-based ( 15 – 17 ) mechanistic models, do not rely on the whole genome. It is widely acknowledged that knowledge about each isolated microorganism is insufficient to explain the behavior and properties of consortia ( 3 , 4 , 18 – 21 ). Therefore, understanding the behavior of a microbial consortium is contingent on determining the relationships between the microorganisms within it and how these relationships influence the community’s behavior ( 13 , 22 – 24 ). Moreover, when aiming to design a consortium with a specific behavior, it is necessary to previously identify the key microbial species that contribute to active beneficial processes and the positive interactions supporting the growth and stability of these populations in the community ( 25 ). Regardless of the method used for MCE, genomic information is deemed essential for achieving optimal results. Advances in genome sequencing technologies have enabled the comprehensive description of the metabolism of different organisms at a whole-genome scale ( 26 , 27 ). Over the last 20 years, metabolic network reconstruction has significantly expanded its range of applications ( 20 , 28 ). For instance, current uses of metabolic networks include understanding various metabolic processes in a microorganism to optimize the production of a particular metabolite ( 29 – 32 ) or to comprehend the metabolic properties of different microorganisms ( 22 , 28 , 31 ). The reconstruction of a metabolic network involves using an organism’s genomic information to understand its metabolism. This process typically comprises several consecutive steps ( 24 ). In the first step, a draft reconstruction of the network is generated based on the genome annotation to identify a collection of metabolic functions encoded in the genome. Several tools can be employed to create the initial reconstruction of a metabolic network ( 33 ). The second step involves the manual curation of the draft network, requiring a meticulous review of each enzyme and reaction in the metabolic network based on previous knowledge of the organism and known metabolic reactions. In the third step of reconstruction, the metabolic network is converted into mathematical computable functions or models, usually written in a standardized format. The last step involves the verification, evaluation, and validation of the metabolic model to identify missing metabolic functions, often necessitating a repetition of the second and third steps ( 23 ). Notably, there is also an automated reconstruction tool for genome-scale metabolic models, CarveMe, which employs an innovative top-down reconstruction approach ( 34 ). Although genome-scale metabolic reconstruction has different applications, the majority of published uses are related to improving the understanding of a bacteria’s metabolism at a molecular level and increasing the production of certain metabolites in a single organism ( 25 , 29 ). Different strategies to define, understand, and characterize cross-feeding interactions between bacteria are based on experimental or mathematical approaches ( 34 ). For instance, to predict cross-feeding interaction in the gut microbiome, GutCP ( 14 ) combines machine learning techniques with an ecological-guided model of the microbiome. At the same time, MICOM ( 35 ) integrates taxonomic abundance based on metagenomic samples with dietary constraints to generate personalized metabolic models. Another tool for analyzing metabolic interactions and microbial consortia, the Microbiome Modeling Toolbox ( 36 ), employs metagenomic data and microbial metabolic reconstructions as input. Baldini et al. ( 37 ) used nutrient dynamics with a coarse-grained description of cell metabolism, integrating an ecology model for the population to develop cross-feeding models. Freilich et al. ( 7 ) developed a novel methodology to predict overall potential interspecies interactions in different environments using metabolic reconstruction and ecological co-occurrence patterns. In their study, the authors simulated the possible competitive and cooperative interactions between pairs of bacteria under different growth conditions. It is also relevant to mention tools that predict non-gut microbiome consortia, such as SMETANA ( 38 ), which can predict higher-order community interactions based on in vitro media and conditions. Importantly, these approaches rely on highly curated metabolic networks, often requiring manual intervention, and also demand access to extensive computational infrastructures ( 35 ) . In this work, we focus on identifying novel cross-feeding and competition interactions between bacteria by developing a new computational approach. This innovative tool, based on a relatively simple machine learning algorithm, employs metabolic networks automatically reconstructed from genome annotations to predict bacterial interactions. Importantly, our method demands lighter computational resources, and once trained, it is usable on standard computers. Initially, we compiled a data set of pairs of bacteria for which their interaction had been previously reported. Subsequently, we explored various methods to automatically reconstruct metabolic networks and encode this information into a linear vector describing pairs of bacteria. Through comprehensive testing, we demonstrate how our relatively simple approach effectively distinguishes between cross-feeding and competition interactions.",
"discussion": "DISCUSSION In this work, we developed a novel bacterial interaction predictor based on metabolic network features using different classifier algorithms. The utilization of metabolic information for analyzing and assembling microbial consortia has been considered in different studies ( 7 , 28 , 43 – 47 ). Motivated by this, we aimed to create a model capable of predicting the type of interaction between pairs of microorganisms with minimal information and resources while still effectively describing a bacterium. We curated a data set of cross-feeding and competition interactions sourced from the literature ( 48 , 49 ). This data set includes a pool of 3,141 reactions obtained from automatically generated metabolic networks for 244 bacteria and 6 archaea. On average, each genome exhibits 793 reactions, showcasing a distinct pattern for each microorganism, reinforcing the differences among the analyzed microorganisms. The data set contains 1,053 cross-feeding and 273 competing pairs of microorganisms. To mitigate biases stemming from the order of microorganisms in the numeric vectors employed for training classification algorithms, we implemented data augmentation. This process involved creating new vectors for each unique combination of bacteria A and B, considering both AB and BA arrangements. The data set forms the foundation of our method and was used to train several machine learning classification algorithms. Given the relatively small number of examples in our data set, we opted to perform a fourfold cross-validation instead of dividing it into test and training sets. To minimize similarities between folds, we applied k-means clustering. This approach ensures that the reported performance is based on training and testing data sets that are as dissimilar as possible, decreasing the chances of overfitting. We initiated our exploration with a KNN algorithm, using the optimal k as determined by the accuracy of 10% of randomly selected data, using 67% for training and 33% for testing. Notably, we achieved promising results with a precision of 0.76 and recall of 0.73 when competition was the target class, considering that a random prediction should give a precision of 0.21. For cross-feeding as the target class, we obtained a precision of 0.95 and a recall of 0.94, surpassing the expected precision of 0.79 for a random prediction. This result obtained with the simplest classification algorithm is very promising, indicating the ability to correctly define the interactions between two microorganisms based on automatically annotated metabolic reactions from their genome. To discard potential artifacts from the machine learning algorithm and explore more complex classification method, we compared the baseline KNN with the algorithms: RF, SVMs (using linear, polynomial, RBF, and sigmoid kernels), and XGBoost. For each algorithm, we consistently observed worse results using competition as the target class compared to cross-feeding (the most numerous class). This difference suggests that competition interactions in our data set might not be sufficiently representative to distinguish between cross-feeding and competition interactions. Another possibility is that competition interactions are more difficult to identify, but we would require larger data sets to validate this hypothesis. Nonetheless, the standout performer across all assessments was XGBoost, which confirms its efficacy in handling unbalanced data sets. There are other methods available to predict bacterial interactions. One method that uses machine learning to predict cross-feeding interactions in the gut microbiome is GutCP ( 14 ). This method combines an ecological-guided model of the microbiome with machine learning techniques. Another approach, MICOM ( 36 ), integrates taxonomic abundance based on metagenomic samples with dietary constraints to generate personalized metabolic models. Furthermore, the Microbiome Modeling Toolbox ( 37 ) uses metagenomic data and microbial metabolic reconstructions to analyze metabolic interactions and microbial consortia. In contrast to these methods, our approach relies solely on metabolic network information automatically generated from the annotated genome of two microorganisms, making it the most straightforward and cost-effective approach available. The study of Freilich et al. ( 7 ) also used metabolic reconstruction along with ecological co-occurrence patterns to predict and simulate interactions between pairs of bacteria under different conditions. For this reason, we used our predictor to forecast the interactions and compared the PCPS (potential cooperation scores) matrix (with interactions predicted) with the species combinations created by them. We obtained 1,895 pairs of consensus interactions with their study, constituting approximately one-third of the complete data set. Although the consensus corresponds to only one-third of the complete data set, it should be noted that our method enhances the accuracy and quality of the information used, as we have used real interactions between pairs of bacteria rather than simulations, and this information has increased in recent years. Additionally, if we only considered the interaction data from the Freilich et al. data set (discarding NI), we obtained an accuracy of 0.72, an F1 score of 0.85, and a recall of 0.83. Since 2017, the NJS16 database has housed a collection of experimentally validated interactions between pairs of bacteria. The availability of experimentally validated interaction data presents immense potential for understanding the interactions between pairs of bacteria in a community. However, unraveling the behavior of microbial consortia remains a challenge that computational approaches hold promise in addressing in the future. Additionally, another important limitation of our method is that the predictions do not offer insights into the specific metabolites exchanged between pairs or resources over which different genotypes compete in case of competition or shared in the context of the cross-feeding case. We must stress that our approach relies solely on literature data that has undergone experimental validation. Therefore, there is a potential bias in the representativeness of the strains used in the examples. Nevertheless, our trained models are compatible with standard computers and can predict interactions within minutes. Our method, grounded in experimentally validated interactions between pairs of bacteria rather than simulations, has yielded promising results, suggesting the encouragement of our machine learning approach. However, there is room for improvement and refinement, particularly as more validated information, especially concerning competition examples, becomes available. In conclusion, unraveling the interactions among microorganisms within a consortium and understanding their impact on community behavior are paramount. The integration of metabolic networks and machine learning has facilitated rapid data analysis, enabling the evaluation of new methodologies to expand our knowledge of microbial community relationships. The inclusion of information regarding the culture medium in our method for predicting bacterial interactions could, in the future, enhance the predictive ability by capturing the influence of the environment on metabolic responses. By considering different environmental conditions, our model could simulate more realistic scenarios and adapt to changes in the environment, thereby improving predictive accuracy. This addition allows us to explore indirect interactions and provide a more complete representation of complex microbial dynamics, which is especially valuable for practical applications in fields such as microbiological research, biotechnology, and health. The concept of representing metabolic networks as inputs for machine learning has been previously explored by DiMucci et al. ( 50 ). They employ a representation of bacterial interactions, incorporating features derived from relevant traits, such as metabolic functions, presence/absence of specific genes, or any other pertinent trait by each bacterium, to train machine learning algorithms. In this study, we introduced a novel machine learning method that utilizes automatically reconstructed metabolic networks to predict cross-feeding and competition interactions between pairs of microorganisms. While there is still room for improvement, our method exhibits excellent performance in distinguishing between cross-feeding and competitive interactions. This tool holds the potential to aid the selection of microbial consortia components and advance our understanding of microbiota relationships."
} | 4,518 |
38318491 | PMC10840354 | pmc | 8,368 | {
"abstract": "The biosynthesis of bioactive secondary metabolites, specifically antibiotics, is of great scientific and economic importance. The control of antibiotic production typically involves different processes and molecular mechanism. Despite numerous efforts to improve antibiotic yields, joint engineering strategies for combining genetic manipulation with fermentation optimization remain finite. Lincomycin A (Lin-A), a lincosamide antibiotic, is industrially fermented by Streptomyces lincolnensis . Herein, the leucine-responsive regulatory protein (Lrp)-type regulator SLCG_4846 was confirmed to directly inhibit the lincomycin biosynthesis, whereas indirectly controlled the transcription of SLCG_2919 , the first reported repressor in S. lincolnensis . Inactivation of SLCG_4846 in the high-yield S. lincolnensis LA219X (LA219XΔ 4846 ) increases the Lin-A production and deletion of SLCG_2919 in LA219XΔ 4846 exhibits superimposed yield increment. Given the effect of the double deletion on cellular primary metabolism of S. lincolnensis , Plackett-Burman design, steepest ascent and response surface methodologies were utilized and employed to optimize the seed medium of this double mutant in shake flask, and Lin-A yield using optimal seed medium was significantly increased over the control. Above strategies were performed in a 15-L fermenter. The maximal yield of Lin-A in LA219XΔ 4846-2919 reached 6.56 g/L at 216 h, 55.1 % higher than that in LA219X at the parental cultivation (4.23 g/L). This study not only showcases the potential of this strategy to boost lincomycin production, but also could empower the development of high-performance actinomycetes for other antibiotics.",
"conclusion": "5 Conclusions In this study, a strategy that combined genetic and fermentation engineering was used to increase lincomycin production ( Fig. 8 ). Genetic engineering focused on knocking out two repressors, SLCG_4846, which is involved in lincomycin biosynthesis by controlling the lin cluster, and SLCG_2919, the first reported negative regulator in S. lincolnensis , to increase lincomycin yield. Furthermore, fermentation engineering was investigated to optimize the seed medium using a rational response surface method, which resulted in a further increase in lincomycin yield, whether in shake flasks or fermenters. Therefore, this strategy will be beneficial in boosting the development of high-performance actinomycetes for other antibiotics. Fig. 8 Schematic diagram to improve lincomycin yield. Fig. 8",
"introduction": "1 Introduction Antibiotics, which are bioactive secondary metabolites produced by the fermentation of actinomycetes, are extensively used in pharmacology, agriculture, and other fields [ 1 ]. Meanwhile, the biosynthesis of antibiotics is complex and depends on various factors, including intracellular gene expression, regulation of extracellular medium components, and process control [ 2 , 3 ]. Over the past few decades, most actinomycetes used for the industrial-scale production of antibiotics have been obtained via random mutagenesis programs [ 4 , 5 ]. In recent years, the modification of factors affecting antibiotic production and the optimization of fermentation process have been widely used to improve the yield of antibiotics [ 3 , 5 , 6 ]. Response surface methodology can avoid the drawbacks of classical methods and is an empirical technique for modeling and optimizing fermentation processes [ 7 ]. Recombinant DNA technologies are also efficient tools for increasing antibiotic yields in actinomycetes [ 8 ]. In Saccharopolyspora erythraea ZL1004, the P450 hydroxylase gene eryK and O-methyltransferase gene eryG were co-overexpressed for titer improvement of erythromycin [ 9 ]. Subsequently, fermentation process of the genetically engineered Sac. erythraea was optimized to improve the titer [ 10 , 11 ]. Therefore, combining genetic manipulation with fermentation engineering can enhance the biosynthetic yield of mutant strains, which is necessary for the industrial overproduction of antibiotics. The expression of gene clusters for antibiotic biosynthesis in actinomycetes typically occurs in the early stationary phase, followed by a transition phase involving complex metabolic alterations [ 12 ]. During this process, numerous transcriptional regulators control antibiotic biosynthesis by responding to extracellular and intracellular signals [ [13] , [14] , [15] ]. Changes in the quantity and activity of these regulators lead to variations in nutrient uptake and utilization [ 16 , 17 ]. Thus, the deliberate engineering of regulators in actinomycetes is an effective strategy to improve antibiotic production. Engineering multiple regulatory elements to comprehensively adjust gene expression could boost antibiotic titers [ [18] , [19] , [20] , [21] ]. However, the original fermentation conditions could not fully tap the potential of the engineering strains, resulting in the inability to further increase the antibiotic yield. Fermentation optimization may be a practical approach to solve this problem. To date, a joint engineering strategy for coupling regulator manipulation with fermentation optimization has not been reported. Lincomycin is clinically used to treat bacterial infections in patients who cannot use penicillin, cephalosporins, or macrolide antibiotics. Lincomycin A (Lin-A), comprising an α-methylthiolincosamide and N-methylated 4-propyl- l -proline, is a major fermentation product of the actinomycete Streptomyces lincolnensis [ 22 ]. Owing to the global market of hundreds of tons per year, improving the yield of lincomycin is of significance [ 23 ]. Random mutagenesis and fermentation optimization have been frequently used to increase lincomycin production [ 24 , 25 ], and considerable efforts have been recently devoted to understand lincomycin biosynthesis and its regulation [ 22 , [26] , [27] , [28] , [29] ]. Notably, several transcription factors (TFs) in S. lincolnensis have been discovered and used to improve lincomycin [ 19 , [30] , [31] , [32] ]. However, understanding the regulatory landscape of lincomycin biosynthesis is limited. In particular, no studies have been published by coupling TF-based genetic manipulation with fermentation optimization for lincomycin titer improvement. In this study, a leucine-responsive regulatory protein (Lrp)-type TF, SLCG_4846, was identified to directly repress lincomycin biosynthesis, whereas indirectly control the transcription of SLCG_2919, which was the first reported negative TF in S. lincolnensis [ 30 ]. The double deletion of SLCG_4846 and SLCG_2919 in a high-yield S. lincolnensis resulted in the significant increase in lincomycin yield. The Plackett-Burman, response surface designs and steepest ascent method were used to optimize the seed medium for this double mutant in shake flasks. The experiments were performed in a fermenter, which could significantly boost industrial lincomycin production.",
"discussion": "4 Discussion Biosynthesis is a sophisticated process that utilizes microorganisms to obtain an abundance of commercial products. The application of strains to enhance the titer during industrial production is critical for such efforts [ 35 ]. Strain modification and fermentation optimization are common strategies for improving the biosynthesis of secondary metabolites [ 36 ]. In antibiotic-producing actinomyces, it is difficult to further scale up the production of antibiotics owing to the monolithic strategy. In this report, we developed a strategy to jointly integrate genetic manipulation and fermentation optimization in S. lincolnensis , leading to a marked improvement in lincomycin production. Lrp, widely present in prokaryotes, participates in the biosynthesis of antibiotics [ 15 , 19 , 34 ]. We previously confirmed that SLCG_Lrp directly stimulates the biosynthesis of lincomycin [ 19 ]. Meanwhile, SLCG_Lrp was found to be directly repressed by the TetR-type regulator SLCG_2919 [ 19 ]. However, little is known about the remaining Lrps in S. lincolnensis . Herein, a novel Lrp protein from S. lincolnensis , SLCG_4846, was identified and proven to directly repress the expression of the lin cluster except lmbE , further regulate the lincomycin biosynthesis. In addition, SLCG_4846 and SLCG_2919 regulated each other at the transcriptional level, in which SLCG_4846 indirectly promoted the expression of SLCG_2919 and SLCG_2919 directly controlled the expression of SLCG_4846 . Consequently, these findings exhibit a sophisticated regulatory network for lincomycin biosynthesis. In addition, the relationship between Lrp and other regulators need to be further exploited. The multiplex and rational engineering of transcriptional regulators in actinomycetes is an ideal strategy to upgrade the titer of antibiotics [ 18 , 37 , 38 ]. In our previous study, deletion of TetR-type regulatory genes SACE_3986 and SACE_3446 in Sac. erythraea increased the erythromycin yield [ 21 ]. Deletion of negative regulatory genes cebR and txtR in S . albidoflavus J1074 results in an improvement in thaxtomins yield [ 39 ]. Herein, double deletion of SLCG_4846 and SLCG_2919 resulted in a significant increase in the Lin-A yield. Therefore, this work reveals the potential for broader industrial applications in improving other secondary metabolites. Microbial fermentation is the basis to boost the yields of a range of secondary metabolites with economic importance [ 7 , 40 , 41 ]. During the fermentation process, seed conditions, including the optimum age and physiological state, are regarded as the key to the biosynthetic overproduction of antibiotics [ 10 ]. Nevertheless, the seed at the optimum age and physiological state is often ignored, which could induce failure at the fermentation stage. In this work, we utilized a Plackett-Burman design, the steepest ascent method, and response surface design to optimize the seed medium of LA219XΔ 4846-2919 . The maximum yield of 3.875 g/L of Lin-A was achieved with LA219XΔ 4846-2919 in flask, which was approximately 25.6% higher than that when using the engineering strain with un-optimized seed medium. When cultivated in a 15-L bioreactor, Lin-A yield of LA219XΔ 4846-2919 reached 6.56 g/L. In summary, our results have provided a viable approach for improving the titers of most antibiotics in actinomycetes."
} | 2,595 |
22347874 | PMC3276360 | pmc | 8,370 | {
"abstract": "Under anoxic conditions in sediments, acetogens are often thought to be outcompeted by microorganisms performing energetically more favorable metabolic pathways, such as sulfate reduction or methanogenesis. Recent evidence from deep subseafloor sediments suggesting acetogenesis in the presence of sulfate reduction and methanogenesis has called this notion into question, however. Here I argue that acetogens can successfully coexist with sulfate reducers and methanogens for multiple reasons. These include (1) substantial energy yields from most acetogenesis reactions across the wide range of conditions encountered in the subseafloor, (2) wide substrate spectra that enable niche differentiation by use of different substrates and/or pooling of energy from a broad range of energy substrates, (3) reduced energetic cost of biosynthesis among acetogens due to use of the reductive acetyl CoA pathway for both energy production and biosynthesis coupled with the ability to use many organic precursors to produce the key intermediate acetyl CoA. This leads to the general conclusion that, beside Gibbs free energy yields, variables such as metabolic strategy and energetic cost of biosynthesis need to be taken into account to understand microbial survival in the energy-depleted deep biosphere.",
"conclusion": "Conclusion If energy yields per substrate are the only important variable controlling microbial metabolism in energy-starved subsurface sediments, then acetogenic microbes should be outcompeted by other anaerobic microbes that perform energetically more favorable pathways, such as sulfate reduction and methanogenesis. While this may be the case in some places, recent δ 13 C-isotopic analyses that indicate a significant acetogenic contribution to total acetate turnover have suggested otherwise (Heuer et al., 2009 ; Lever et al., 2010 ). In this study, I discuss several potentially advantageous traits of acetogenic microbes that may enable them to coexist with sulfate reducers and methanogens in spite of lower energy yields per substrate. Using conservative calculations, I show that most acetogenic substrates are likely to occur at concentrations that vastly exceed the thermodynamic threshold concentration for acetogenesis and are thus potential energy substrates to acetogens in the deep biosphere. Due to their ability to metabolize certain substrates via multiple different reactions, e.g., methanol alone, methanol + H 2 , or methanol + formate, acetogens have a remarkable metabolic flexibility compared to sulfate reducers and methanogens, which in some cases may enable them to gain higher energy yields per substrate than these two groups. Acetogens also have a greater metabolic versatility with respect to the number and breadth of substrates utilized than sulfate reducers and methanogens. As a result, they may avoid competition via niche differentiation, i.e., by feeding on substrates not utilized by most sulfate reducers or methanogens. The greater substrate breadth furthermore means that acetogens are able to access energy from a greater overall number of substrates. Rather than evolving to become highly efficient and specialized consumers of abundant single substrates, acetogens are therefore likely to be substrate generalists with the capacity to draw on a large pool of less abundant (rare) substrates. A further advantage of the acetogenic lifestyle may lie in the ability of acetogens to curb energy spent on biosynthesis. Acetogens use the reductive acetyl CoA pathway, the energetically least costly of all C fixation pathways. By using this pathway for both energy production and biosynthesis, they may cut back on energy that other groups spend on the maintenance of additional genes and enzymes. Use of organic compounds rather than H 2 /CO 2 as starting blocks of biomass synthesis may moreover enable acetogens to circumvent energetically costly lithoautotrophic C fixation. Given the high energetic cost of amino acid synthesis in deep subsurface sediments, and the fact that synthesis and maintenance of enzymes for DNA and protein repair are likely to be the main energy expenditures of microbes in starvation mode, acetogens may be able to save crucial energy for survival by virtue of the simplicity and versatility of their biochemical pathway. Given that the vast majority of cells in deep subsurface sediments are probably in starvation mode with generation times of hundreds to thousands of years (Biddle et al., 2006 ; Jørgensen et al., 2006 ), basic questions regarding the ecology of these organisms remain unanswered. Are the cells found highly recalcitrant survivors from surface environments, or have they adapted to the conditions of extreme energy limitation? Have microbes actively colonized sediments long after their accumulation, or have they been present since their initial deposition? Irregardless of the answers to these questions, it is likely that the ability of acetogens to use wide substrate ranges and perform biosynthesis at low energetic cost represent valuable survival traits in the deep biosphere – even if they did not originally evolve as adaptations to this environment.",
"introduction": "Introduction Past studies on anoxic sediments have demonstrated a redox zonation among terminal organic matter remineralizing microbes in relation to electron acceptor availability (e.g., Froelich et al., 1979 ; Canfield et al., 1993 ). Organisms using the electron acceptor with the highest Gibbs free energy yields dominate over groups using energetically less favorable electron acceptors (e.g., Cappenberg, 1974 ; Lovley and Goodwin, 1988 ; Hoehler et al., 1998 ). Higher energy yields support faster growth rates and result in competitive exclusion of groups using less favorable electron acceptors (Cord-Ruwisch et al., 1988 ). Energy substrates with high turnover rates, e.g., hydrogen (H 2 ) and acetate, can even be drawn down to thermodynamic threshold concentrations, at which only the most energetically favorable electron acceptor present provides sufficient energy for proton translocation across the cell membrane, ATP formation, and growth (Hoehler et al., 2001 ; Hoehler, 2004 ). Evidence supporting the notion of biological redox zonation comes from freshwater and coastal marine sediments, as well as laboratory-based chemostat and coculture experiments. Consistent with the notion of redox zonation, the processes of denitrification, manganese and iron reduction, and sulfate reduction should exclude energetically less favorable reactions involving CO 2 reduction or disproportionation, wherever the available nitrate, manganese (IV), iron (III), and sulfate pools are not rate-limiting (e.g., Froelich et al., 1979 ; Canfield et al., 1993 ). In freshwater and coastal marine sediments depletion of the most favorable oxidants often occurs shallowly owing to an excess of electron donors produced by fermentation and hydrolysis reactions (Capone and Kiene, 1988 ). This creates a niche for methane-producing Archaea (methanogens) and acetate-synthesizing microbes (acetogens), groups that are able to harvest energy from CO 2 reduction in underlying layers (e.g., Phelps and Zeikus, 1984 ; Avery et al., 2002 ; Ferry and Lessner, 2008 ; Liu and Conrad, 2011 ). Contrastingly, in more oligotrophic offshore marine sediments, which cover most of the Earth’s surface, organic matter and hence electron donor availability are typically limiting. Depletion of nitrate, oxidized metals, sulfate, and/or even dioxygen (O 2 ) does not occur until tens of meters below the seafloor – if at all (e.g., D’Hondt et al., 2004 ; D’Hondt et al., 2009 ). Accordingly methanogens and acetogens should be absent or at best lead fringe existences – dormant, or surviving in small numbers on non-competitive energy substrates not used by the other groups, such as methylated C1 compounds or methoxylated aromatic compounds (Franklin et al., 1988 ; Lever et al., 2010 ). Hence, recent evidence from deeply buried marine sediments indicating significant accumulation of biogenic methane in the presence of sulfate and metal reducing populations seems surprising (Wang et al., 2008 ). Moreover, even though sulfate reducers and methanogens gain more energy than acetogens from shared energy substrates, there is increasing evidence that acetogens play a quantitatively important role in organic carbon cycling in the marine and terrestrial deep biosphere (Heuer et al., 2006 , 2009 ; Griebler and Lueders, 2008 ; Pedersen et al., 2008 ; Lever et al., 2010 ). In the following sections I will examine possible reasons for the coexistence of acetogenesis with other pathways that are considered to be energetically more favorable in the deep subsurface. In my analyses, I will (1) conservatively calculate the energy yields of widespread acetogenesis reactions in the subsurface, (2) examine the potential for substrate generalism as a successful strategy under extreme energy limitation, and (3) examine the cost of biosynthesis and potential ways by which acetogens may reduce energy expended on biosynthesis.",
"discussion": "Results and Discussion The thermodynamic argument To assess the energetic feasibility of microbial metabolic reactions (Table 1 ) in deep subseafloor sediments, it is helpful to conservatively calculate their energy yields under conditions that resemble those found in situ . In this section, I examine the energetic potential of various acetogenesis reactions to occur in deep subseafloor sediments by examining (1) Gibbs free energy yields per reaction ( Δ G r ′ ) , (2) Gibbs free energy yields per substrate ( Δ G s ′ ) , (3) thermodynamic threshold concentrations of substrates for acetogenesis reactions to be thermodynamically favorable, (4) in situ energy yields of the reactions for which educt and product concentrations have been quantified in subseafloor sediments, and (5) energy yields per hydrogen molecule (H 2 ) of the various litho- and organotrophic acetogenesis reactions involving H 2 compared to competing hydrogenotrophic sulfate reduction and methanogenesis reactions. Which acetogenesis reactions are thermodynamically favorable? Calculated Gibbs free energies indicate that most acetogenesis reactions are thermodynamically favorable in deep subseafloor sediments – with energy yields exceeding the BEQ ( Δ G r ′ = - 10 k J m o l - 1 ) under a wide range of temperatures, pressures, and hydrogen concentrations (Table 3 ). The highest energy values with Δ G r ′ < - 100 k J m o l - 1 are in carbohydrates, pyruvate, methyl chloride, methoxylated aromatic compounds, and lactate. Other substrates, such as glycolate, oxalate, methanol, and ethanol also produce energy yields exceeding the BEQ. By contrast, the classic autotrophic (“homoacetogenic”) reaction from H 2 –CO 2 and reactions from formate are endergonic at 0.1 nM H 2 concentrations (Table 3 ), and only yield energy at 1 μM H 2 concentrations and low to intermediate temperatures (Table 3 ). For energy-yielding substrates that can be used with or without H 2 , i.e., carbon monoxide, lactate, methanol, and syringate, reactions not involving H 2 yield more energy at [H 2 ] = 0.1 nM than reactions involving H 2 ; in the case of carbon monoxide and lactate, this difference is crucial, since reactions without hydrogen produce high-energy yields, whereas reactions with H 2 are endergonic (Table 3 ). At [H 2 ] = 1 μM this changes, i.e., energy yields of some of the reactions with hydrogen are exergonic, yielding more energy than the BEQ (Table 3 ); in one case (lactate, −1.9°C) free energy yields even exceed those of acetogenesis from lactate alone at low to intermediate temperatures (also see Results and Discussion on Table 4 in next section). Table 3 Gibbs free energy yields of the various acetogenesis reactions at a wide range of temperatures, pressures, and H 2 concentrations . Temperature −1.9°C +25°C +122°C Pressure 1 atm 1000 atm 1 atm 1000 atm 1 atm 1000 atm A H 2 –CO 2 45.2 41.4 72.4 68.6 170 167 Carbon monoxide ND ND − 54.7 ND ND ND Carbon monoxide + H 2 ND ND −2.57 ND ND ND Formate 24.0 22.5 43.0 41.5 111 110 Formate + H 2 24.2 21.6 46.3 43.6 126 123 Lactate − 99.5 − 98.6 − 101 − 100 − 106 − 105 Lactate + H 2 18.0 12.8 58.1 52.9 203 198 Glycolate − 38.0 − 36.9 − 16.8 − 15.6 59.9 61.0 Oxalate − 93.9 − 94.4 − 84.1 − 84.6 − 48.8 − 49.4 Methanol − 104 − 105 − 91.8 − 92.6 − 45.9 − 46.8 Methanol + H 2 − 14.8 − 13.7 −4.84 −3.68 31.1 32.3 Methanol + formate − 25.3 − 25.9 − 17.9 − 18.5 8.82 8.23 Ethanol − 37.7 − 37.9 − 34.3 − 34.5 − 22.0 − 22.2 Pyruvate − 172 − 170 − 157 − 155 − 102 − 100 Glucose − 332 − 331 − 345 − 344 − 394 − 393 Cellobiose − 715 ND − 747 ND − 861 ND Methyl chloride − 210 ND − 198 ND − 152 ND Syringate ND ND − 275 ND ND ND Syringate + H 2 ND ND − 101 ND ND ND Vanillate ND ND − 263 ND ND ND B H 2 –CO 2 − 37.9 − 41.7 − 18.9 − 22.7 49.4 45.7 Carbon monoxide + H 2 ND ND − 48.2 ND ND ND Formate + H 2 − 17.3 − 20.0 0.610 −2.04 65.3 62.7 Lactate + H 2 − 107 − 112 − 78.9 − 84.1 21.2 16.0 Methanol + H 2 − 35.6 − 34.4 − 27.7 − 26.5 0.891 2.05 syringate + H 2 ND ND − 130 ND ND ND For reactions that yield more energy than a BEQ of Δ G r ′ = - 10 k J m o l - 1 these values are indicated in bold. (A) [H 2 ] = 0.1 nM in reactions with H 2 ; (B) [H 2 ] = 100 nM. ND = not determined, due to absence of published Δ H f ∘ and Δ V f ∘ values . Table 4 Gibbs free energy yields per substrate for the various acetogenesis reactions at a wide range of temperatures (°C ) , pressures (atm), and H 2 concentrations . Substrates reaction −1 −1.9°C +25°C +122°C 1 atm 1000 atm 1 atm 1000 atm 1 atm 1000 atm A H 2 –CO 2 4 11.3 10.3 18.1 17.1 42.6 41.7 CO 4 ND ND − 13.7 ND ND ND CO + H 2 2 ND ND −1.28 ND ND ND Formate 4 6.00 5.62 10.7 10.4 27.9 27.5 Formate + H 2 2 12.1 10.8 23.1 21.8 62.9 61.6 Lactate 2 − 49.8 − 49.3 − 50.5 − 50.0 − 53.0 − 52.5 Lactate + H 2 1 18.0 12.8 58.1 52.9 203 198 Glycolate 4 − 9.51 − 9.23 − 4.20 − 3.91 15.0 15.3 Oxalate 4 − 23.5 − 23.6 − 21.0 − 21.2 − 12.2 − 12.3 Methanol 4 − 26.1 − 26.3 − 22.9 − 23.2 − 11.5 − 11.7 Methanol + H 2 1 − 14.8 − 13.7 −4.84 −3.68 31.1 32.3 Methanol + formate 1 − 25.3 − 25.9 − 17.9 − 18.5 8.82 8.23 Ethanol 2 − 18.9 − 19.0 − 17.2 − 17.3 − 11.0 − 11.1 Pyruvate 4 − 43.0 − 42.6 − 39.1 − 38.8 − 25.4 − 25.1 Glucose 1 − 332 − 331 − 345 − 344 − 394 − 393 Cellobiose 1 − 715 ND − 747 ND − 861 ND Methyl chloride 4 − 52.6 ND − 49.5 ND − 38.1 ND Syringate 2 ND ND − 137 ND ND ND Syringate + H 2 1 ND ND − 101 ND ND ND Vanillate 4 ND ND − 65.9 ND ND ND B H 2 − CO 2 4 − 9.48 − 10.4 − 4.73 − 5.68 12.4 11.4 CO + H 2 2 ND ND − 24.1 ND ND ND Formate + H 2 2 − 8.67 − 9.99 0.305 −1.02 32.7 31.3 Lactate + H 2 1 − 107 − 112 − 78.9 − 84.1 21.2 16.0 Methanol + H 2 1 − 35.6 − 34.4 − 27.7 − 26.5 0.89 2.05 Syringate + H 2 1 ND ND − 130 ND ND ND Reactions yielding energy in excess of a BEQ of Δ G r ′ = - 10 k J m o l - 1 are indicated in bold. (A) [H 2 ] = 0.1 nM in reactions with H 2 ; (B) [H 2 ] = 100 nM . Within the calculated ranges, temperature has a much greater impact on free energy yields of acetogenesis reactions than pressure. For pressure changes from 1 to 1000 atm, the largest effect is in reactions that include H 2 , with lactate + H 2 having the greatest change (5.2 kJ mol −1 ); for all reactions without H 2 , the difference in Δ G r ′ between 1 and 1000 atm pressure is ≤2 kJ mol −1 . By comparison, the difference in Δ G r ′ due to a temperature change from −1.9 to +122°C is always greater than the difference in Δ G r ′ caused by a pressure change from 1 to 1000 atm. For reactions from H 2 –CO 2 , formate–H 2 , lactate–H 2 , and cellobiose, the change in Δ G r ′ values going from −1.9 to +122°C even exceeds 100 kJ mol −1 of the reaction. The magnitude of temperature effects seems to follow trends. First of all, of the six reactions involving H 2 or formate for which temperature effects could be calculated, none are thermodynamically favorable at 122°C. This, and the vast overall decrease in free energy yields for acetogenesis reactions involving H 2 or formate in response to temperature, suggests a strong selection against these reactions at temperatures approaching the known upper limit of life – unless H 2 or formate concentrations at high temperatures are much higher than assumed here. Secondly, there appears to be a systematic difference in how temperature affects energy yields. For all C 1 –C 3 substrates except reactions from lactate alone, free energy yields decrease with temperature; for reactions from lactate alone, there is a slight increase with temperature. By contrast, the free energy yields of acetogenesis reactions from the carbohydrates glucose (C 6 ) and cellobiose (C 12 ) show strong increases in response to temperature. Based on the small number of substrates included in these calculations and that the only two large substrates included are carbohydrates, it is premature to argue that larger carbon substrates should be consumed preferably at high temperatures. Yet, the fact that certain substrates or acetogenesis reactions increase, while others decrease in energy yield in response to temperature, suggests that temperature exerts an important control over which substrates are consumed and energy-yielding reactions performed by acetogens in the deep biosphere. Which acetogenesis reactions are most likely under the conditions examined? The Gibbs free energy yield of a metabolic reaction indicates whether this reaction can be used as a source of energy in a given environment. Under substrate-limiting conditions, as are likely in the deep biosphere, one might, however, expect microbial consumer choices – assuming they follow optimum foraging behavior – to be driven by energy yields per mole of substrate – as long as the overall reaction produces more energy than the BEQ. For substrates that can be metabolized via multiple reactions that each yield more energy than the BEQ, e.g., methanol at −1.9°C (Table 3 ), one might expect consumers to show a preference toward the reactions with the highest energy yield per substrate. Additionally, it is possible that organisms, despite being energy-starved, show a preference toward certain substrates over others based on energy content per substrate molecule. To examine possible consequences of an optimum foraging behavior that is driven by energy yields per substrate, the latter were calculated (Table 4 ). Comparing different substrates on a per-substrate-level, Gibbs free energy yields remain high ( Δ G s ′ < - 100 k J m o l - 1 ) for cellobiose, glucose (all T and P), lactate + H 2 (only at [H 2 ] = 1 μM, −1.9 and 25°C; all unchanged compared to Table 3 ), as well as syringate (Table 4 ). Provided their availability in the deep biosphere, and that acetogens make choices based on energy per substrate molecule, these substrates should be consumed preferentially over the others examined. Other good substrates may include – in order of descenting energy yields – vanillate, methyl chloride, lactate, pyruvate, methanol, and oxalate. The classic lithoautotrophic reaction from H 2 –CO 2 , and reactions from formate and glycolate, are the least energy-yielding on a per-substrate-level, and therefore the least likely to be consumed, should energy content on a substrate-level determine acetogenic substrate choice. When comparing energy yields of acetogenesis substrates for which multiple reaction pathways are known, i.e., CO, formate, lactate, methanol, and syringate, the same overall trends seen on a per-reaction-level still hold for formate and syringate – independent of H 2 concentrations (Tables 3 and 4 ). For carbon monoxide, lactate, and methanol, the same trends occur at [H 2 ] = 1 nM, but not at [H 2 ] = 1 μM. In spite of the overall reaction from carbon monoxide yielding more energy than the reaction from carbon monoxide + H 2 at high [H 2 ] (Table 4 ), acetogenesis from carbon monoxide + H 2 yields more energy on a per-substrate-level (Table 4 ). For lactate, reactions from lactate + H 2 at high H 2 yield more energy per lactate than reactions from lactate alone – not only, as previously, at −1.9°C (Table 3 ), but also at +25°C (Table 4 ). And for methanol – unlike before (Table 3 ) – reactions with H 2 yield more energy per methanol at high H 2 and low temperature (−1.9 to 25°C) than reactions with methanol alone (Table 4 ). These results confirm the importance of calculating energy yields on a per-substrate-level. Moreover, they underscore the likely importance of temperature in regulating which acetogenesis reactions are occurring in situ – even when these reactions involve the same carbon substrate. The extent to which microbes can detect and respond to (minor) differences in energy yields of different reactions involving the same substrates, thereby optimizing their foraging behavior with respect to energy yields per substrate, is poorly understood. The potential advantages for survival in energy-starved environments are apparent. Yet, it is not known whether microbes express any form of substrate selectivity in the energy-starved deep biosphere, or rather indiscriminately consume any metabolizable substrate that enters their reach. The strategy employed by an individual cell may not solely depend on the energy yield per substrate. Other factors, such as substrate turnover rate, energetic cost of substrate/metabolite transport across the cell membrane, and energy return on investment for each enzyme that needs to be synthesized to catabolize an additional energy substrate will most likely also affect which substrates are consumed. Which acetogenesis reactions are likely to occur in situ ? The calculated free energy yields presented so far are based on limited published information on concentrations of acetogenic substrates. Only H 2 , formate, and acetate concentration data have been published for the deep subseafloor biosphere (Shipboard Scientific Party, 2003 ; Lorenson et al., 2006 ; Heuer et al., 2009 ; Lever et al., 2010 ; Expedition 329 Scientists, 2011 ); concentrations of the other substrates had to be approximated using data from surficial marine sediments (Meyer-Reil, 1978 ; Sørensen et al., 1981 ; King et al., 1983 ; Smith and Oremland, 1983 ; Parkes et al., 1989 ; Martens, 1990 ; Hoehler et al., 2001 ; Dhillon et al., 2005 ; Finke et al., 2006 ; King, 2007 ), freshwater sediments (King et al., 1982 ; Lovley and Goodwin, 1988 ; Chidhaisong et al., 1999 ; Keppler et al., 2000 ), marine water columns (Edenborn and Litchfield, 1987 ; Ballschmiter, 2003 ), and the terrestrial deep biosphere (Chapelle and Bradley, 2007 ). Since concentrations of H 2 , formate, and acetate in the deep biosphere overlap with the concentrations of these species in other sedimentary environments, it seems realistic to conservatively approximate subseafloor concentrations of other substrates, such as glucose or oxalate, with the lowest values measured in other sedimentary environments. This cannot hide the fact that actual concentrations have not been measured, however. An alternative to calculating energy yields at assumed substrate concentrations is therefore to calculate the threshold concentrations required for acetogens to meet the BEQ from a substrate. This can be done conservatively, since the concentrations of most other reaction educts and products, i.e., H + , H 2 O, HCO 3 − , acetate, H 2 , Cl − , are well-constrained for the deep biosphere and/or set to conservative values (see Materials and Methods ). Thermodynamic threshold concentrations were, as previously, calculated for [H 2 ] = 0.1 nM and 1 μM (Table 5 ). Table 5 Thermodynamic threshold concentrations of widespread acetogenesis reactions at a wide range of temperatures and pressures, assuming a biological energy quantum of \n Δ G f ′ = - 10 kJ mol - 1 . Temperature −1.9°C +25°C +122°C Pressure 1 1000 1 1000 1 1000 A H 2 –CO 2 4.5E − 08 3.0E − 08 4.1E − 07 2.8E − 07 9.2E − 05 6.9E − 05 CO ND ND 1.1E − 12 ND ND ND CO + H 2 ND ND 4.5E − 10 ND ND ND Formate 4.3E − 06 3.7E − 06 2.1E − 05 1.8E − 05 1.0E − 03 9.2E − 04 Formate–H 2 2.0E − 04 1.1E − 04 8.5E − 03 5.0E − 03 9.5E + 01 6.4E + 01 Lactate 2.4E − 16 2.9E − 16 1.1E − 15 1.3E − 15 4.6E − 14 5.2E − 14 Lactate + H 2 2.5E − 02 2.5E − 03 8.6E + 04 1.0E + 04 1.3E + 21 2.7E + 20 Glycolate 4.5E − 09 5.1E − 09 5.0E − 08 5.7E − 08 2.0E − 05 2.2E − 05 Oxalate 9.2E − 12 8.6E − 12 5.7E − 11 5.4E − 11 5.2E − 09 5.0E − 09 Methanol 2.8E − 12 2.6E − 12 2.6E − 11 2.4E − 11 6.5E − 09 6.1E − 09 Methanol + H 2 1.2E − 08 2.0E − 08 8.0E − 07 1.3E − 06 2.8E − 02 3.9E − 02 Methanol + formate 1.1E − 10 8.6E − 11 4.1E − 09 3.2E − 09 3.1E − 05 2.6E − 05 Ethanol 2.1E − 10 2.0E − 10 7.4E − 10 7.1E − 10 1.6E − 08 1.6E − 08 Pyruvate 1.6E − 19 1.9E − 19 3.8E − 18 4.4E − 18 9.4E − 15 1.0E − 14 Glucose 2.7E − 96 4.1E − 96 4.8E − 93 7.1E − 93 5.4E − 85 7.2E − 85 Cellobiose 6.5E − 204 ND 3.4E − 197 ND 1.4E − 180 ND Methyl chloride 5.1E − 24 ND 5.9E − 23 ND 2.5E − 20 ND Syringate ND ND 6.6E − 34 ND ND ND Syringate + H 2 ND ND 1.0E − 23 ND ND ND Vanillate ND ND 7.9E − 19 ND ND ND B H 2 –CO 2 4.5E − 08 3.0E − 08 4.1E − 07 2.8E − 07 9.2E − 05 6.9E − 05 CO + H 2 ND ND 4.5E − 14 ND ND ND Formate + H 2 2.0E − 08 1.1E − 08 8.5E − 07 5.0E − 07 9.5E − 03 6.4E − 03 Lactate + H 2 2.5E − 26 2.5E − 27 8.6E − 20 1.0E − 20 1.3E − 03 2.7E − 04 Methanol + H 2 1.2E − 12 2.0E − 12 8.0E − 11 1.3E − 10 2.8E − 06 3.9E − 06 Syringate + H 2 ND ND 1.0E − 31 ND ND ND Concentrations are for the first substrate listed, e.g., H 2 for H 2 –CO 2 . (A) [H 2 ] = 0.1 nM in reactions with H 2 ; (B) [H 2 ] = 100 nM . At first glance it is clear that the concept of threshold concentrations is only relevant for a subset of acetogenic substrates. For glucose, cellobiose, syringate, as well as syringate + H 2 at 1 μM [H 2 ], threshold concentrations are lower than a single molecule of the substrate per liter. In fact, thermodynamic threshold concentrations for glucose and cellobiose are even orders of magnitude lower than one molecule per Earth’s entire ocean volume (Table 5 )! For carbon monoxide, lactate, pyruvate, methyl chloride, syringate + H 2 at low [H 2 ], or lactate + H 2 at high [H 2 ] it also seems unlikely that meeting the BEQ is a realistic obstacle. For these reactions, threshold concentrations are at most in the low picomolar range – and with that ~2–3 orders of magnitude lower than microbes are known to draw limiting metabolite concentrations down to (e.g., Fuhrman and Ferguson, 1986 ; Hoehler et al., 2001 ; Stolper et al., 2010 ). If previously measured concentrations of organic substrates in deep subsurface sediments, which have for the most part been (0.1 μM (e.g., Shipboard Scientific Party, 2003 ; Mitterer, 2006 ; Heuer et al., 2009 ; Lever et al., 2010 ), are a good reference, then we are left with the same conclusions as before (Table 4 ), i.e., that most acetogenesis reactions produce energy yields in excess of the BEQ, even at substrate concentrations that are low for the deep biosphere. More interestingly, perhaps, examining those substrates that were previously considered less likely to be used by acetogens (based on Tables 3 and 4 ), suggests that even formate and H 2 –CO 2 are potential acetogen substrates in some subsurface environments. Formate concentrations from low micromolar to tens of micromolar (Table 5 ) have been documented for sites ranging from organic-rich (ODP Site 1230) to highly oligotrophic (ODP Site 1231; Shipboard Scientific Party, 2003 ). Thermodynamic calculations based on measured formate concentrations suggest that formate could be a substrate of acetogenesis at certain depths in subsurface sediments on the Juan de Fuca Ridge Flank (Lever et al., 2010 ). Accurate quantifications of hydrogen concentrations in the deep biosphere are fraught with uncertainty, with two different methods yielding results differing by up to two orders of magnitude (Lin et al., 2011 ). Yet, independent of the method used, measured concentrations exceeding 10 nM are not uncommon (Shipboard Scientific Party, 2003 ; Lorenson et al., 2006 ; Expedition 329 Scientists, 2011 ; Lin et al., 2011 ), and suggest that even acetogenesis from H 2 –CO 2 is possible in some places, if not widespread. In situ energy yields of acetogenesis reactions based on measured concentrations To the best of my knowledge, the only subseafloor sediment samples for which all educt and product concentrations of acetogenesis reactions have been quantified are from ODP Leg 201 (Shipboard Scientific Party, 2003 ). The seven sites sampled during this expedition vary from organic-rich to oligotrophic and cover a range of energy conditions that is likely to include most anoxic subseafloor sediments on Earth. For these samples, the concentrations of formate and hydrogen (and no other acetogenic substrates) were measured in parallel with concentrations of acetate, DIC (proxy for bicarbonate), and pH, allowing for the calculation of in situ energy yields of acetogenesis reactions from H 2 –CO 2 , formate, and formate + H 2 (Figure 1 ). Figure 1 Depth profiles of energy yields associated with acetogenesis reactions from (A) H 2 –CO 2 , (B) formate, and (C) formate–H 2 for ODP sites 1225-31 . All calculations are based on measurements from sediment cores collected during ODP Leg 201 (Shipboard Scientific Party, 2003 ). Calculated free energy yields for the three reactions show clear trends: reactions from H 2 –CO 2 are mostly endergonic, and only yield energy in excess of the BEQ value at a few shallow depths at ODP Site 1231 (Figure 1 A). Reactions from formate are exergonic with energy yields around or exceeding the BEQ across all sites and depths sampled (Figure 1 B). Reactions from formate + H 2 , are for the most part slightly exergonic, but only exceed the BEQ at a few depths at ODP Sites 1225, 1230, and 1231 (Figure 1 C). Based on these results, one might suppose that acetogenesis from formate is possible across a wide range of subseafloor habitats, whereas acetogenesis reactions involving H 2 –CO 2 or formate + H 2 are unlikely. Yet, the high uncertainty associated with the quantification of porewater H 2 concentrations needs to be taken into account. H 2 concentrations measured during ODP Leg 201 were obtained via an incubation method, which assumes headspace hydrogen to be in equilibrium with dissolved hydrogen in pore fluids after an incubation period (Lovley and Goodwin, 1988 ; Hoehler et al., 1998 ). When compared to a new, extraction-based method on the same samples, the incubation method yields concentration measurements that are consistently lower by one to two orders of magnitude (Lin et al., 2011 ). If in situ concentrations of H 2 in sediments sampled during Leg 201 are one order of magnitude higher than measured previously, this would lower the Δ G r ′ for acetogenesis from H 2 –CO 2 by ~22 kJ mol −1 . In this case, close to half of the samples would have energy values exceeding the BEQ (Figure 1 A). If in situ concentrations are two orders of magnitude higher, this will lower the Δ G r ' for acetogenesis from H 2 –CO 2 by an additional ~22 kJ mol −1 – by a total of ~44 kJ mol −1 of the reaction compared to the measured H 2 data. In this case, energy yields of acetogenesis from H 2 –CO 2 would exceed the BEQ in the overwhelming majority of samples collected during Leg 201. Energy yields of acetogenesis reactions involving H 2 compared to competing sulfate reduction and methanogenesis reactions The main empirical support for the concept of redox zonation comes from isotopic tracer studies and measurements of hydrogen concentrations in sulfate-reducing and methanogenic freshwater and coastal surface sediments (e.g., Cappenberg, 1974 ; Capone and Kiene, 1988 ; Lovley and Goodwin, 1988 ; Hoehler et al., 1998 ; Heimann et al., 2010 ). Acetogenesis has received less attention in sediments due to the difficulty of detecting the process; after all the end product acetate is also a key substrate to sulfate reducers and methanogens, and rapid turnover results in acetate typically not accumulating to high concentrations – unlike the end products of sulfate reduction and methanogenesis, sulfide, and methane. Moreover, acetogenesis is often equated with the “homoacetogenic” reaction from H 2 –CO 2 , which is thermodynamically unfavorable under thermodynamic control of H 2 concentrations in sulfate-reducing or methanogenic sediments. Only rarely have energy yields of organotrophic acetogenesis reactions that include H 2 been taken into account (Liu and Suflita, 1993 ; Lever et al., 2010 ). Here I compare the energy yields of various acetogenesis reactions involving H 2 to those of the widespread hydrogenotrophic sulfate reduction and methanogenesis reactions at a wide range of H 2 concentrations (Figure 2 ). Figure 2 Relationship between H 2 concentrations and energy yields for sulfate reduction, methanogenesis, and acetogenesis from H 2 –CO 2 , as well as acetogenesis from formate–H 2 , CO–H 2 , methanol–H 2 , lactate–H 2 , and syringate–H 2 . (A) energy yields per reaction ( Δ G r ′ ) , the black line indicates the BEQ, (B) energy yields per substrate ( Δ G s ′ ) . All calculations were done assuming standard temperature and pressure, and using educt and product concentrations as outlined in the Materials and Methods, except for [H 2 ]. At first glance it is apparent that acetogenesis from H 2 –CO 2 is thermodynamically less favorable than sulfate or methanogenesis reactions from H 2 –CO 2 , independent of H 2 concentrations (Figure 2 A). Under the conditions used in calculations, sulfate reducers can meet the BEQ down to H 2 concentrations of ~0.6 nM, methanogens down to 11 nM, whereas acetogens require 410 nM H 2 concentrations. Acetogenesis from formate + H 2 is also unlikely, as its energy yields are below the BEQ unless H 2 concentrations are in the micromolar range. More energetically favorable than sulfate reduction or methanogenesis is, however, the acetogenic reaction from syringate + H 2 , which even at H 2 concentrations as low as 0.01 nM produces high-energy yields (~−90 kJ mol −1 ) – concentrations at which both sulfate reduction and methanogenesis are endergonic. Moreover, while sulfate reduction from H 2 –CO 2 is the overall second most energy-yielding reaction, acetogenesis reactions from CO, methanol, and lactate produce more energy than hydrogenotrophic methanogenesis at H 2 concentrations within the typical range measured in deep subseafloor sediments. When energy yields are considered on a per hydrogen molecule level, the results are even more striking. Acetogenesis reactions from syringate + H 2 , methanol + H 2 , and CO + H 2 all provide more energy per H 2 molecule than sulfate reduction from H 2 –CO 2 (Figure 2 B). The reaction from lactate + H 2 yields less energy than hydrogenotrophic sulfate reduction, but slightly more than hydrogenotrophic methanogenesis, while acetogenesis reactions from formate + H 2 and H 2 –CO 2 produce the least amount of energy per H 2 molecule and are endergonic except at micromolar H 2 concentrations (Figure 2 B). One might argue that carboxydo- and organotrophic acetogenesis reactions involving H 2 are unlikely in marine sediments. After all, reactions involving the same carbon substrate without H 2 yield more energy except when high H 2 concentrations coincide with low temperature (Tables 3 and 4 ) – a condition that has traditionally only been observed during season-induced temporary disequilibria in shallow sediments (Hoehler et al., 1999 ) and is perhaps unlikely in seasonally stable subsurface sediments. Yet, the method-dependent discrepancies in measured H 2 concentrations in deep subsurface sediments (Lin et al., 2011 ) leave room for high uncertainty; if the higher [H 2 ] measurements obtained with a new extraction-based method (Lin et al., 2011 ) turn out to be accurate, then carboxydo- and organotrophic acetogenesis reactions with H 2 may be competitive, if not energetically favorable, over carboxydo- and organotrophic reactions without H 2 in the predominantly cold, deep biosphere. Evidence supporting the importance of organotrophic reactions with H 2 comes from subsurface sediments of the Atlantic Coastal Plain (Liu and Suflita, 1993 ). An acetogenic isolate from these sediments only showed growth through O -demethoxylation of syringate under an H 2 –CO 2 atmosphere, while no growth on syringate was observed under an N 2 –CO 2 or N 2 atmosphere. And by metabolizing syringate with H 2 , this organism was able to outcompete hydrogenotrophic methanogens for H 2 in the initial sediment enrichment. The ability to gain energy from the demethoxylation of syringate or other lignin monomers is widespread among acetogens, but not among sulfate reducers or methanogens, suggesting that methoxy-groups on aromatic rings might represent non-competitive substrates (Lever et al., 2010 ). The same is not true for the other substrates, lactate, CO, and methanol (Figure 2 ). Lactate serves as a growth substrate not only to many acetogens (Lever et al., 2010 ), but also to many sulfate reducers (Rabus et al., 2006 ), which can be expected to have higher energy yields from competing sulfate reduction reactions. Similarly, despite being less widely used as growth substrates than among acetogens (Lever et al., 2010 ), CO, and methanol can also serve as energy substrates to certain sulfate reducers (reviewed in Mörsdorf et al., 1992 ), as well as several methanogens (reviewed in Whitman et al., 2006 , and in Ferry, 2010 ). Both sulfate reducers and methanogens can be expected to gain more energy from reactions involving CO or methanol than acetogens. Hence, evidence suggesting an important role for acetogenesis in the cycling of CO and methanol in marine and freshwater sediments might be surprising (King, 2007 ; Jiang et al., 2010 ). In the following section, I will argue that the ability of acetogens to use a wide range of substrates is a viable survival strategy under conditions of energy limitation – despite lower energy yields per substrate. Metabolic strategies of acetogens A striking feature of acetogens as a metabolic guild is the widespread ability to use a large number and wide diversity of carbon compounds as energy substrates. Over half of all cultured strains test positively for growth on H 2 /CO 2 , carbon monoxide, formate, methanol, ethanol, other aliphatic compounds such as lactate, and methoxylated aromatic compounds (Lever et al., 2010 ). Further widely used growth substrates include carbohydrates, other short-chain fatty acids and alcohols, methoxylated aliphatic compounds, betaines, amino acids, and aldehydes (Drake et al., 2006 ). Even complex organic polymers, such as cellulose or carboxymethylcellulose, are used by some strains (Wolin and Miller, 1994 ; Karita et al., 2003 ; Wolin et al., 2003 ). Due to the limited number of energy substrates on which growth is typically tested, substrate ranges of acetogens may significantly exceed the currently known spectrum. Any methyl or methoxyl groups of compounds found in the environment represent potential energy substrates, that, given thermodynamically favorable conditions, might be combined with CO 2 to form acetate. Considering the striking metabolic versatility of acetogens, it seems plausible that the resulting plasticity with respect to substrate use is part of the strategy that enables acetogens to coexist with sulfate reducers and methanogens. In the following sections, I will examine two hypotheses that seek to explain the benefits of a wide metabolic spectrum. The first hypothesis is that acetogens can coexist with sulfate reducers and methanogens due to niche differentiation with respect to substrate use. In other words, acetogens may avoid competition by consuming substrates not used by sulfate reducers or methanogens. The second hypothesis is that the ability to pool energy from a wide range of metabolic reactions enables coexistence despite lower energy from shared substrates. These two hypotheses are not incompatible, but should rather be regarded as two complementary advantages of a generalist metabolic strategy. Niche differentiation based on substrate use When viewed collectively, sulfate reducers, though not to the same extent as acetogens, can also exploit a large variety of substrates. When examined more closely, however, it appears that only the ability to use H 2 , short-chain fatty acids, and ethanol is truly widespread across the various genera (Rabus et al., 2006 ). Common acetogenic substrates such as methanol, glucose, fructose, carbon monoxide, and methoxylated lignin monomers are not substrates to the vast majority of sulfate reducers (Mörsdorf et al., 1992 ; Rabus et al., 2006 ). The thermodynamic advantage of higher energy yields of sulfate reduction compared to acetogenesis thus only plays out for a subset of acetogenic substrates that are also utilized by sulfate reducers. Tracer experiments indicating H 2 and short-chain fatty acids as the main electron donors used by sulfate reducers in estuarine and marine sediments support this conclusion (Sørensen et al., 1981 ; Parkes et al., 1989 ). Similarly, experimental evidence indicating acetogens as key consumers of CO and methoxyl groups in sulfate-reducing marine sediments (Küsel et al., 1999 ; King, 2007 ) suggests that, despite overlaps in substrates, acetogens, and sulfate reducers practice a form of niche differentiation in which each group favors different energy substrates where they coexist. Vastly less metabolically versatile than acetogens or sulfate reducers, the substrate range of methanogens is limited to (1) CO 2 reduction (H 2 /CO 2 , formate, a few use carbon monoxide or alcohols), (2) acetate disproportionation, and (3) demethylation of C1 compounds (methanol, methyl sulfides, and methylamines). With the exception of one genus ( Methanosarcina ), most methanogens are substrate specialists and only capable of growth on one of these three substrate groups (Whitman et al., 2006 ). Hence, potential competition between acetogens and methanogens for substrates is limited to a small subset of acetogenic substrates. Niche differentiation, resulting in use of different energy substrates where the groups coexist, may thus explain why sulfate reducers and methanogens do not competitively exclude acetogens in anoxic sediments. Why have sulfate reducers not adapted to use the full spectrum of substrates used by acetogens? And, given that both acetogens and methanogens utilize the reductive acetyl CoA pathway for energy production and/or C fixation and overlap in substrate use (Drake et al., 2006 ; Whitman et al., 2006 ), why might these two groups differ so drastically with respect to their metabolic versatility? The ultimate evolutionary explanations remain subject to speculation. On a more proximal level, differences in energy yields and turnover rates of energy substrates may have played a role in driving differences in metabolic strategies. The most common substrates used by sulfate reducers and methanogens, i.e., H 2 and acetate, so called central intermediates of organic carbon degradation (e.g., Valentine, 2001 ; Dolfing et al., 2008 ), are presumably the electron donors with the highest turnover rates in anoxic sediments. Other short-chain fatty acids, which represent important energy substrates to sulfate reducers are also known to have high turnover rates (Sørensen et al., 1981 ; Parkes et al., 1989 ); even methylated compounds used by methanogens, i.e., methanol, methylamines, and methyl sulfides, are known to have high turnover rates in certain environments (e.g., Zhilina and Zavarzin, 1990 ; Mitterer et al., 2001 ; Jiang et al., 2010 ; Lin et al., 2010 ). Substrates with high turnover rates are typically small, as they derive from a variety of individually less abundant, larger source molecules; they also often harbor less energy than larger organic molecules with lower turnover rates (Tables 3 and 4 ). Why might certain groups utilize substrates with high turnover rates but low-energy yields, while others use substrates with low turnover rates but high-energy yields? Part of the answer may lie in the universal requirement of cells to meet maintenance energy requirements. Meeting maintenance energy requirements is especially challenging in deep subsurface sediments, as these have typically been cut off from fresh organic matter supplies for thousands to millions of years. Here the vast majority of cells is likely to be in a permanent state of starvation (D’Hondt et al., 2004 ; Jørgensen et al., 2006 ) and starvation may even represent the primary source of mortality. Based on chemostat experiments, the following relationship between maintenance energy and temperature has been established (Tijhuis et al., 1993 ; Harder, 1997 ): (1) ME = A * e − E a / R T where ME is the maintenance energy (kJ (g dry mass) −1 d −1 ), A a constant [4.99 × 10 12 kJ (g dry mass) −1 d −1 ], E a the activation energy (69.4 kJ mol −1 K −1 ), R the universal gas constant (0.008314 kJ mol −1 K −1 ), and T the temperature (K). The value of the constant A was calculated from the energy supply rate at which microbial cell growth and replication stops in chemostat experiments (Tijhuis et al., 1993 ). It has since been estimated that the actual threshold energy required for cell maintenance is three orders of magnitude lower than the threshold for growth/replication (Price and Sowers, 2004 ; Biddle et al., 2006 ). Therefore, I will use a value of 4.99 × 10 9 kJ (g dry mass) −1 d −1 for A in all calculations of ME. Based on this value of A , a maintenance energy of 1.26 kJ (g dry mass) −1 year −1 can be calculated at standard temperature. Tijhuis et al. ( 1993 ) propose that 26 g of cell dry mass on average contain 12 g cell carbon. Combined with the published estimate of 10 fg C per cell for sediment-inhabiting microbes (Whitman et al., 1998 ), one can then calculate a cell-specific maintenance energy, ME cell , of 2.74 × 10 −14 kJ cell −1 year −1 at standard temperature. The relationship between ME cell , the Gibbs free energy yield per substrate, Δ G s ′ (kJ mol −1 ), and the cell-specific substrate turnover rate, k cell (mol cell −1 year −1 ) that is required for a cell to meet maintenance energy requirements, can be expressed as follows: (2) M E cell = ∑ Δ G ′ s,A × k cell , A + Δ G ′ s,B × k cell , B + … where A and B indicate substrates A and B, respectively. If cells are only consuming one substrate, this expression simplifies, so the equation can be solved for k cell , if ME cell and Δ G s ′ are known: (3) k cell = ME cell / Δ G s ' The relationship between k cell and Δ G s ′ is hyperbolic (Figure 3 A). This has implications for metabolic strategies among microbes: for instance, microbes can meet maintenance energy requirements by consuming substrates with low-energy yields as long as turnover rates are high and the BEQ is met (ME A ); alternatively, microbes can meet ME requirements at low turnover rates, as long as energy yields per substrate are high (ME B ). Figure 3 Relationships between energy yields per substrate ( Δ G s ′ ) and turnover rate ( k cell ) . (A) Hyperbolic relationship between Δ G s ′ and k cell assuming ME cell = 2.74 × 10 − 14 kJ cell − 1 year − 1 . ME A and ME B indicate two different strategies to meet ME cell , ME A for a substrate with low-energy yields ( Δ G s ′ = - 10 k J m o l - 1 ) and a high k cell , and ME B for a substrate with high-energy yields ( Δ G s ′ = - 100 k J m o l - 1 ) and a low k cell . (B) Illustration of the effect of 10 kJ mol − 1 increments in energy yields per substrate on the turnover rate required to meet ME cell . (C) Turnover rates for ME A and ME B (“plus 0”) compared to competing reactions “plus 10,” “plus 20,” and “plus 30” with 10, 20, and 30 kJ mol − 1 higher energy yields per substrate, respectively. A further implication is that small changes in Δ G s ′ greatly influence the turnover rate required to meet ME cell if energy yields per substrate are small (here <50 kJ mol −1 ), but not if they are high (here ≥100 kJ mol −1 ; Figure 3 B). If Δ G s ′ changes from −10 to −20 kJ mol −1 , the turnover rate required to meet maintenance energy requirements drops by 50%. By comparison, if Δ G s,B ′ changes from −100 to −110 kJ mol −1 , the decrease in required turnover rate is only ~9%. What does this mean regarding the substrates used by sulfate reducers and methanogens compared to the substrates used by acetogens? In Figure 3 C, maintenance turnover rates are illustrated for the same metabolic reactions, A and B, as in Figure 3 A (plus 0), as well as for three hypothetical pathways that produce higher energy yields per substrate (plus 10, plus 20, plus 30); “plus 0” exemplifies acetogenesis reactions from a high turnover, low-energy substrate (ME A ), and a low turnover, high-energy substrate (ME B ), respectively; the pathways behind “plus 10,” “plus 20,” and “plus 30” are energetically more favorable methanogenesis and sulfate-reducing reactions involving the same two substrates. The differences in Δ G s ′ of −10, −20, and −30 kJ mol −1 relative to acetogenic reactions are based on typical differences in Δ G s ′ calculated for methanogenesis/sulfate reduction vs. acetogenesis reactions involving the substrates formate, methanol, and lactate across a wide range of temperatures (275–337 K) and sulfate concentrations (0–28 mM) in subsurface sediments (Lever et al., 2010 ). The comparison illustrates that acetogenesis reactions operating near the thermodynamic threshold ( ME A , Δ G s 1 = − 10 k J m o l − 1 ) ; Figure 3 C) have a tremendous disadvantage in meeting ME cell compared to competing methanogenic or sulfate-reducing reactions, which can operate at one-half, one-third, or one-fourth the substrate turnover (Table 6 ). The advantage of vastly lower turnover rates required to meet ME cell diminishes with increasing energy yields per substrate (ME B , Figure 3 C). It follows that the minimum turnover rates of the three competing methanogenesis and sulfate reduction reactions are only lower by 9, 17, and 23%, when the energy yield per substrate is −100 kJ mol −1 for the acetogenesis reaction (Table 6 ). Table 6 Overview of cell-specific substrate turnover rates ( k cell ; fmol cell −1 year −1 ) required to meet the theoretical maintenance energy requirement of 2.74 × 10 −1 4 kJ cell −1 year −1 at different free energy yields per substrate ( Δ G s ′ ; kJ mol −1 ) . Plus 0 Plus 10 Plus 20 Plus 30 Ratios of turnover rates Δ G s ′ k cell Δ G s ′ k cell Δ G s ′ k cell Δ G s ′ k cell k cell, plus 0 : k cell, plus 10 k cell, plus 0 : k cell, plus 20 k cell, plus 0 : k cell, plus 30 A 10 2.74 20 1.37 30 0.91 40 0.68 0.50 0.33 0.25 20 1.37 30 0.91 40 0.68 50 0.55 0.67 0.50 0.40 50 0.55 60 0.46 70 0.39 80 0.34 0.83 0.71 0.63 75 0.36 85 0.32 95 0.29 105 0.26 0.88 0.79 0.71 100 0.27 110 0.25 120 0.23 130 0.21 0.91 0.83 0.77 200 0.14 210 0.13 220 0.12 230 0.12 0.95 0.91 0.87 500 0.05 510 0.05 520 0.05 530 0.05 0.98 0.96 0.94 Plus 0 (A + B) Plus 10 Plus 20 Plus 30 Ratios of turnover rates Σ Δ G s ′ k cell Δ G s ′ k cell Δ G s ′ k cell Δ G s ′ k cell k cell, plus 10 : k cell, plus 0 k cell, plus 20 : k cell, plus 0 k cell, plus 30 : k cell, plus 0 B 20 1.4 20 1.4 30 0.91 40 0.68 1.00 0.67 0.50 40 0.69 30 0.91 40 0.68 50 0.55 1.33 1.00 0.80 100 0.27 60 0.46 70 0.39 80 0.34 1.66 1.43 1.25 150 0.18 85 0.32 95 0.29 105 0.26 1.76 1.58 1.43 200 0.14 110 0.25 120 0.23 130 0.21 1.82 1.66 1.54 400 0.07 210 0.13 220 0.12 230 0.12 1.90 1.82 1.74 1000 0.03 510 0.05 520 0.05 530 0.05 1.96 1.92 1.88 Calculations are shown for four different “pathways”; “plus 0” is the pathway with the lowest Gibbs free energies, and an analog for acetogenesis; “plus 10,” “plus 20,” and “plus 30” are energetically more favorable pathways that produce 10, 20, and 30 kJ more energy per mole of substrate. These energetically more favorable pathways are analogs for methanogenesis and sulfate reduction reactions. The ratios in required substrate turnover rates for the plus 0 pathway vs. the other three pathways to match the above maintenance energy requirement are shown on the far right. (A) all four pathways only use one substrate, A; (B) plus 0 pathway uses two substrates, A and B, that are equal in Δ G s ′ and k cell , while plus 10, plus 20, and plus 30 pathways still only use substrate A . Returning to the question raised earlier in this section, i.e., whether methanogens/sulfate reducers and acetogens may practice a form of niche differentiation, in which each group uses different substrates where they co-occur, the model presented here provides clear answers. Feeding on high-energy, low turnover substrates is a viable survival strategy for microbes, as is feeding on low-energy, high turnover substrates. The two strategies may, at least in part, explain the coexistence of acetogenic with sulfate-reducing and/or methanogenic microbial populations. A reason why sulfate reducers/methanogens may mainly use low-energy, high turnover substrates is their vast energetic advantage over acetogens in metabolizing these substrates. At high Δ G s ′ , the energetic advantage of sulfate reducers/methanogens over acetogens decreases, and other guild-specific traits may increase in importance. One of these will be discussed in the following section. Specialist vs. generalist arguments The potential advantages of a specialist vs. a generalist life style have been the subject of discussions among ecologists for over five decades. A traditional view is that selectivity (specialization) pays off under non-limiting energy conditions, while less discrimination toward food sources (generalism) is the more effective survival strategy under energy limitation (e.g., Emlen, 1966 ; Dykhuizen and Davies, 1980 ; for reviews see Pianka, 1994 ; Egli, 1995 ). With respect to microbial ecology, this argumentation has been called into question by certain microbial growth experiments, in which long starvation periods appeared to favor specialists (e.g., Kuenen, 1983 ). A fundamental difference between microbes and macrobiota under energy limitation is that microbes not only struggle to meet maintenance energy requirements, but also to acquire a minimum free energy (BEQ) from a metabolic reaction to even be able to produce ATP. Hence, substrate specialists are often equipped with high substrate affinities down to very low concentrations that enable them to outcompete not only other metabolic guilds, but also other members of their own metabolic guild. A classic example among methanogens is the obligately aceticlastic Methanosaeta genus which can grow on acetate concentrations below 10 μM (Jetten et al., 1992 ). By comparison, the (by methanogen standards) “generalistic” Methanosarcina genus, members of which can grow via CO 2 reduction, acetate disproportionation, and demethylation of C1 compounds, requires acetate concentration of at least 200 μM for growth (Jetten et al., 1992 ). High substrate affinity appears to be a strategy among specialists by which they can more efficiently take up substrates and drive substrate concentrations below the threshold concentrations required by generalists. This form of energetic (i.e., thermodynamic) exclusion, which only occurs in microbes, provides a gateway for substrate specialists, provided they can meet maintenance energy requirements. The costs of substrate specialization are a smaller accessible energy pool due to utilization of fewer energy substrates and lower energy yields per substrate. In energy-starved environments, such as deep subsurface sediments, a specialist metabolic strategy may work effectively for substrates with high turnover rates. High turnover rates and relatively high cell densities are necessary to maintain the low substrate concentrations that enable specialists to thermodynamically exclude less efficient consumers of the same substrate. As shown earlier (Tables 1 – 3 ), thermodynamic exclusion of acetogens is only likely for the low-energy substrates H 2 and formate, however. Other, less common, but more energy-rich substrates occur at concentrations exceeding the thermodynamic threshold, and provide little incentive for specialization due to the impossibility of thermodynamically excluding other groups, such as acetogens, from consuming them. Hence, a more generalist metabolic strategy may be more effective among consumers of these more rare, energy-rich substrates. As discussed in a previous section, substrate generalism is a widespread trait among acetogens. With respect to meeting maintenance energy requirements, there are clear advantages to using more than one substrate (Figure 4 ); for instance, at high turnover rates and low Δ G s (−10 kJ mol −1 substrate) combining the energy yields of two substrates, A and B, with the same Δ G s ′ and k cell may enable acetogens to lower their required substrate turnover by 50% and successfully compete with methanogens/sulfate reducers that gain 10 kJ more energy per mole of substrate A (ME A ; Figure 4 ; Table 6 ). While this is an improvement of competitiveness, pooling the energy from two substrates is still insufficient to compete with methanogens/sulfate reducers gaining 20 or 30 kJ more per mole of substrate, which still can meet ME cell on turnover rates that are 33 and 50% lower, respectively (Table 6 ). To match energy yields of the latter, energy from three or four substrates with the properties of substrate A would need to be pooled – a considerable disadvantage in terms of energy efficiency as one might argue. This changes for substrates with high-energy yields (e.g., −100 kJ mol −1 substrate) and resultingly low required turnover rates (ME B ; Figure 4 ). For these, pooling the energy from two substrates would enable an acetogen to grow at a significantly lower turnover rate than any of the competing methanogens/sulfate reducers utilizing only one substrate – a significant advantage (Figure 4 ; Table 6 ). It follows from this that pooling the energy from multiple substrates would increase acetogenic competitiveness with sulfate reducers and methanogens overall – and in particular for energy-rich substrates. Figure 4 The same relationship as in Figure 3 C, except that turnover rates required for ME A and ME B (“plus 0”) are now calculated for the sum of two substrates, A and B [“plus 0 (A + B)”] . All other values are the same as before. The same principle as in comparing the benefits of using two vs. one substrates applies in comparing the benefits of using more, e.g., 3 vs. 2, 5 vs. 4, 10 vs. 5, etc., substrates. The main point is that acetogens typically have wider substrate spectra than sulfate reducers or methanogens, and that pooling energy from a larger number of substrates may enable acetogens to in some cases survive on lower substrate turnover rates than the other groups, despite lower energy yields per substrate. Experimental evidence that confirms pooling of energy sources as the explanation for the coexistence of acetogenic with sulfate-reducing and/or methanogenic populations in the deep biosphere is still missing. However, the same principle has been demonstrated in continuous-flow cultures involving other groups of microorganisms that were grown under carbon-limiting conditions: here several studies have shown substrate generalists to grow at lower substrate concentrations than substrate specialists when incubations included multiple substrates (Gottschal et al., 1979 ; Dykhuizen and Davies, 1980 ; reviewed in Egli et al., 1993 ; Egli, 1995 ). Based on the calculations presented (Figure 4 ; Table 6 ), one might conclude that pooling energy is only an effective strategy for subseafloor acetogens to meet ME cell if it involves high-energy substrates. For low-energy substrates, more specialized organisms with higher energy yields per substrate, i.e., sulfate reducers and methanogens, should have a vast advantage (Figure 4 ), provided that energy yields of acetogenesis reactions even exceed the BEQ. Even the most efficient specialist will reach a limit when substrate turnover rates drop below the threshold required to meet ME cell , however; at this point the specialist is either forced to consume additional substrates, or to allow substrate concentrations above the thermodynamic threshold. Evidence potentially supporting the latter comes from oligotrophic sediments of the South Pacific Gyre and Equatorial Pacific, where H 2 concentration peaks in the tens of nanomolar range have been reported for subsurface horizons with exceedingly low microbial activities (Shipboard Scientific Party, 2003 ; Expedition 329 Scientists, 2011 ). If substrate specialists are forced to allow substrate concentrations above the thermodynamic threshold, they become vulnerable to less efficient, more generalistic organisms competing for their preferred substrate. Ultimately, because of the larger accessible substrate and hence energy pool, one might therefore expect substrate generalists to dominate under the most energy-depleted conditions. The results presented thus far suggest that it is very difficult to predict the outcome of the complex competition between acetogens and other groups for substrates in the deep biosphere. Beside physical variables, such as temperature and pressure, it may be necessary to measure concentrations of all educts and products of relevance – a very challenging task with acetogens, due to their wide substrate spectra – as well as measure substrate-specific turnover rates – a seemingly impossible undertaking given the very low turnover rates in the deep biosphere. Even with complete knowledge on concentrations and turnover rates, predicting competitive outcomes on a substrate-level would be compromised by our still limited knowledge on the metabolic capabilities of microbes inhabiting anoxic (subseafloor) sediments, as well as other important life history traits. One of the latter is the energetic cost of biosynthesis – a variable that is likely to vary widely across microbes and microbial metabolic guilds. The energetic cost of biosynthesis among acetogens Of the currently known six pathways of autotrophic carbon fixation, the reductive acetyl CoA pathway is the simplest and energetically most favorable due to the absence of complex biochemical intermediates (Russell and Martin, 2004 ; Berg et al., 2010 ). This strictly anaerobic pathway only consists of a carbonyl branch, in which CO 2 is reduced to an enzyme-bound carbonyl group, and a methyl branch, in which CO 2 is reduced to a cofactor-bound methyl group. The bifunctional enzyme CO dehydrogenase/acetyl CoA synthase (CODH/ACS) carries out both the reduction of CO 2 to CO, as well as the synthesis of the end product, acetyl CoA, by joining the carbonyl and methyl groups (e.g., Hügler and Sievert, 2011 ). The reductive acetyl CoA pathway is unique among C fixation pathways in that it is linear; given geochemically favorable conditions, e.g., in alkaline hydrothermal vent environments, each step is exergonic, meaning that CO 2 fixation can occur spontaneously (Martin and Russell, 2007 ). It has therefore been suggested that this pathway started as a geochemical pathway (Russell and Martin, 2004 ). Moreover, due to it being the only known C fixation pathway that occurs in both Bacteria and Archaea, it has been conjectured that the reductive acetyl CoA pathway is the most ancient C fixation pathway (Fuchs and Stupperich, 1985 ), or even the very first biochemical pathway to have evolved on Earth (Peretó et al., 2004 ). The great simplicity and low energetic cost suggest that anaerobic organisms using this pathway have an energetic advantage over organisms using other C fixation pathways, such as the reverse tricarboxylic acid cycle. The reductive acetyl CoA pathway is found in all known acetogens and methanogens, as well as several autotrophic sulfate reducers and anammox bacteria (Schauder et al., 1989 ; Drake et al., 2006 ; Strous et al., 2006 ; Whitman et al., 2006 ). Certain methanogens and autotrophic sulfate reducers use the pathway exclusively for biosynthesis, while others, including acetate-oxidizing sulfate reducers, aceticlastic methanogens, and syntrophic acetate oxidizers, can produce energy by reversing the pathway so it becomes oxidative (e.g., Spormann and Thauer, 1988 ; Hattori et al., 2005 ; Liu and Whitman, 2008 ). Acetogens, and among these I include facultative acetogens, such as certain sulfate reducers, methanogens, and anaerobic acetate oxidizers (e.g., Jansen et al., 1984 ; Rother and Metcalf, 2004 ; Hattori et al., 2005 ; Lessner et al., 2006 ; Henstra et al., 2007 ), are the only group known to perform this pathway both for biosynthesis and energy production. A possible advantage of using the same pathway for energy production and biomass assimilation is that smaller genomes and fewer enzymes need to be produced and maintained. Since starvation mode is likely to be the rule rather than the exception among microbes in energy-deprived subseafloor sediments (Jørgensen et al., 2006 ), reducing the energetic cost of genome maintenance and enzyme synthesis may confer a significant advantage to microbes that are able to carry out energy production and biosynthesis via the same pathway. It has, in fact, been postulated that the synthesis and maintenance of enzymes to repair DNA from depurination reactions and proteins from racemization reactions are the main energy expenditures among microorganisms in survival mode (Price and Sowers, 2004 ). While too little is known about cell-specific enzyme concentrations and turnover rates in the deep subseafloor to calculate the energetic cost of synthesizing and maintaining these enzymes, concentrations of protein building blocks, i.e., certain amino acids (aspartate, glutamate, serine, glycine), have been measured in subsurface sediments of the Peru Margin and Equatorial Pacific (Mitterer, 2006 ). I use these here to calculate the energetic cost of their lithoautotrophic synthesis. Irrespective of the site, a high energetic cost can be expected for the synthesis of all four amino acids (Figure 5 ). Assuming that this is a general trend across amino acids, the lithoautotrophic synthesis of proteins, and hence enzymes, can be expected to be an energetically costly process in subseafloor sediments. A key intermediate during amino acid synthesis under anaerobic conditions is the energy-rich acetyl CoA molecule, which is also a crucial intermediate during acetogenesis. Assuming that acetogenesis from H 2 –CO 2 is associated with an energetic cost (Figure 1 A), then obligate lithoautotrophs, including many methanogens and sulfate reducers, which synthesize amino acids from H 2 and CO 2 , will spend significant amounts of energy on the reductive synthesis of acetyl CoA alone. By contrast, the majority of organotrophic acetogenesis reactions are exergonic (Table 3 ). Most acetogens may therefore be able to cut back energy expenditures during enzyme synthesis, compared to obligately autotrophic organisms, by using organic substrates to synthesize the amino acid precursor acetyl CoA. Figure 5 Depth profiles of energetic cost of the lithoautotrophic synthesis of the amino acids (A) aspartic acid [asp 2− ; 4 HCO 3 − + NH 4 + + 6 H 2 + H + → − OOCCH(NH 2 )CH 2 COO − + 8 H 2 O], (B) glutamic acid [glu 2− ; 5 HCO 3 − + NH 4 + + 9 H 2 + 2 H + → − OOC(CH 2 ) 2 CH(NH 2 )COO − + 11 H 2 O], (C) serine [3 HCO 3 − + NH 4 + + 5 H 2 + 2 H + → CH 2 OHCH(NH 3 + )COO − + 6 H 2 O], and (D) glycine (2 HCO 3 − + NH 4 + + 3 H 2 + H + → NH 3 + CH 2 COO − + 4 H 2 O) at ODP sites 1225-31 . All calculations are based on measurements obtained from sediment cores collected during ODP Leg 201 (Shipboard Scientific Party, 2003 ; Mitterer, 2006 )."
} | 17,096 |
29057877 | PMC5714965 | pmc | 8,371 | {
"abstract": "Large-scale and high-efficient water collection of microfibers with long-term durability still remains challenging. Here we present well-controlled, bioinspired spindle-knot microfibers with cavity knots (named cavity-microfiber), precisely fabricated via a simple gas-in-water microfluidic method, to address this challenge. The cavity-microfiber is endowed with unique surface roughness, mechanical strength, and long-term durability due to the design of cavity as well as polymer composition, thus enabling an outstanding performance of water collection. The maximum water volume collected on a single knot is almost 495 times than that of the knot on the cavity-microfiber. Moreover, the spider-web-like networks assembled controllably by cavity-microfibers demonstrate excellent large-scale and high-efficient water collection. To maximize the water-collecting capacity, nodes/intersections should be designed on the topology of the network as many as possible. Our light-weighted yet tough, low-cost microfibers with high efficiency in directional water transportation offers promising opportunities for large-scale water collection in water-deficient areas.",
"introduction": "Introduction The spider silk 1 , 2 is well-known for its intriguing ability to collect water from humid air, and has thus inspired the design for materials of unique wettability. The water-wetted spider silk composes of periodic spindle-knots and joints with different surface roughness 3 . The unique structure of the natural microfiber enables a surface energy gradient, as well as a difference in Laplace pressure between the knots and joints 4 . Both result in the directional transport of water droplets towards the knots continuously. Guided by this insight, microfibers with polymer spindle knots have been fabricated to mimic the spider silks for water collection 5 – 11 . The functionalities of these microfibers depend crucially on their geometrical properties, such as the knot size and surface nanostructures. The microfibers with spindle-knots can be fabricated by methods including electrospinning 12 , 13 , dip-coating 14 , 15 , and microfluidic approaches 16 . In the electrospinning approach, a viscous inner liquid and a less-viscous shell liquid are electrified coaxially, forming a hydrophobic fiber with hydrophilic knots 12 , 13 . With the dip-coating, smooth microfibers are dipped into polymer solutions, and then droplets form along the fiber due to Rayleigh-plateau instability 14 , 15 . Subsequently, these droplets are solidified to generate the knots. These two approaches have, however, limited control over the microstructure of the fabricated fiber, such as the separation between the knots and the size of the knots. Microfluidics enables good controllability of microscale jets and droplets 17 , and is thus capable of producing microfibers with precisely-tuned spindle-knot structures. With the microfluidic approach, a liquid jet encapsulating discrete core oil droplets is typically templated for a gel fiber 18 . After dehydration, the oil cores wrapped in a thin layer of gel fiber serve as knots. Thus, the spindle-knot fiber by microfluidics is usually uniform in the material composition, almost with no difference in surface roughness. Moreover, the oil drops evaporate over time and thus the knots deform gradually, compromising the long-term functionality. Furthermore, although the functions of a single fiber have been studied, the integrated collective performance of assembled spindle-fibers as topological networks has not been sufficiently demonstrated. Therefore, it is urgently demanding to fabricate durable, functional spider-silk-mimicking microfibers, and assemble these fibers into topological networks for large-scale water collection. Here we fabricate microfibers with spindle cavity-knots that mimic the structure and surface roughness of the spider silk from composite hydrogels (named cavity-microfiber) by simple microfluidics, for assembled networks and large-scale water collection. The cavity-microfibers are templated from jet phase with gas bubbles, via cross-linking and drying. The knot size and distance between the knots are controlled by the flow rates of the jet phase and the pressure of the gas phase. The surface roughness is enabled by incorporating phase-separated polymers in the jet phase. Owing to the cavity design, the surfaces of the knot part are much rougher than the rest of the cavity-microfiber, enhancing the driving force of the directional water transport. Due to the robustness of the cavity-microfiber, the cavity knots maintain their shape and functions for cycles of water-collecting. Furthermore, we demonstrate the water-collecting efficiency of different topological fiber-networks made from the cavity-microfibers in bio-mimicking spider-web. We show that the structure of cavity-microfiber and its network topology dictate the water-collecting performance. Our facial and economic approach offers light-weighted, yet tough spindle cavity-knot microfibers with high efficiency of water collection. These cavity-microfibers are promising building blocks for spider-web-like networks for water treatment, drug delivery, tissue engineering and cell culture.",
"discussion": "Discussion In conclusion, we employ a simple microfluidic approach to fabricate spindle-knot microfibers with cavity knots for assembled topological networks and large-scale water collection. The spindle-knot structures such as the knot size and distance between knots are precisely controlled by the gas-phase pressure and flow-rate of jet phase. The cavity design as well as the polymer compositions enables the desirable surface roughness, mechanical strength and long-term durability. The maximum water volume collected on a single knot is almost 495 times than that of the knot on the cavity-microfiber we fabricated. On the basis of these results, the cavity-microfibers are assembled controllably into spider-web-like networks for large-scale water collection. To maximize the water-collecting capacity, the topology of the network should be designed to have as many nodes/intersections as possible. Our light-weighted yet tough, low-cost microfibers with high efficiency in directional water transportation creates opportunities for large-scale water collection. Moreover, the spider-silk-like microfibers with cavity-knots, as well as the assembled networks are potential candidates for tissue engineering, encapsulation and controlled release, and controlled liquid transport."
} | 1,627 |
23879629 | null | s2 | 8,372 | {
"abstract": "Contact-dependent growth inhibition (CDI) is a phenomenon in which Gram-negative bacteria use the toxic C-terminus of a large surface-exposed exoprotein to inhibit the growth of susceptible bacteria upon cell-cell contact. Little is known about when and where bacteria express the genes encoding CDI system proteins and how these systems contribute to the survival of bacteria in their natural niche. Here we establish that, in addition to mediating interbacterial competition, the Burkholderia thailandensis CDI system exoprotein BcpA is required for biofilm development. We also provide evidence that the catalytic activity of BcpA and extracellular DNA are required for the characteristic biofilm pillars to form. We show using a bcpA-gfp fusion that within the biofilm, expression of the CDI system-encoding genes is below the limit of detection for the majority of bacteria and only a subset of cells express the genes strongly at any given time. Analysis of a strain constitutively expressing the genes indicates that native expression is critical for biofilm architecture. Although CDI systems have so far only been demonstrated to be involved in interbacterial competition, constitutive production of the system's immunity protein in the entire bacterial population did not alter biofilm formation, indicating a CDI-independent role for BcpA in this process. We propose, therefore, that bacteria may use CDI proteins in cooperative behaviours, like building biofilm communities, and in competitive behaviours that prevent non-self bacteria from entering the community."
} | 394 |
36442122 | PMC9894190 | pmc | 8,373 | {
"abstract": "Significance Soft microrobots that can climb in open, unstructured environment have crucial applications in exploration and monitoring. Small-scale soft actuators with complex, large, customized 3D-to-3D shape morphing capability could equip soft microrobots with enriched locomotion modes and functionalities. Here, we present design strategies and fabrication techniques of voltage-driven 3D soft actuators that can reversibly morph between different configurations at small scales. These developments provide routes to soft microrobots that can change their 3D shapes on demand. A soft microrobot capable of both climbing on flat and curved surfaces and transitioning between different surfaces is demonstrated. This microrobot can carry an integrated microcamera for navigation, while surviving extreme compression and climbing bamboo/leaf, indicating promising potential in practical applications.",
"discussion": "Discussion and Conclusions In summary, the design concepts/methods and assembly technologies reported herein provide access to soft LIG–LCE actuators capable of voltage-driven, customized 3D-to-3D shape morphing (bending angle > 200°) at millimeter scales (from 1 to 10 mm). The soft LIG–LCE actuators enable developments of morphable electroadhesive footpads, smart joints, and highly deformable body, which can be assembled into multigait soft climbing microrobots at different length scales (body length from 6 to 90 mm). The soft microrobot features capabilities of i) climbing on flat and curved surfaces of different materials (e.g., transparent glass, paper, and leaf), ii) transitioning between different surfaces, and iii) flipping over barriers, enabled by three locomotion gaits of the robot. SI Appendix , Table S2 presents a comparison of our soft microrobot with other existing soft climbing robots, in terms of adhesion strategy, actuation method, body length of the robot, robot mass, inclined angle, adaptability of wall shapes, wall transition capability, and working space. Previous works of soft climbing robots mainly focused on motion agility/maneuverability (e.g., locomotion speed and turning ability) on a single type of surface (e.g., on flat wall, rod, pipe) ( 6 – 8 ), new actuation mechanisms ( 3 , 6 , 63 ) (e.g., pneumatic actuation, dielectric elastomer-based actuation, and SMA-based actuation), motion control synergy ( 5 , 6 , 10 ), innovative adhesion mechanism ( 2 , 9 , 63 – 68 ) (e.g., dry adhesion, wet adhesion, chemical adhesion, electroadhesion, pneumatically suction, and mechanical friction), and so on. None of the previously reported soft climbing robots can climb, simultaneously, on flat and curved surfaces, in an open space. To our knowledge, the microrobot presented in the current work is the smallest soft robot that can climb in an open space. These advances of our soft microrobot (i.e., climbing on flat/curved surfaces, wall transitioning) are enabled by the developed morphable footpads capable of customized shape morphing and smart joints that allow on-demand switch between the stepping and flipping gaits of the microrobot. The improved adaptability and miniaturization of soft climbing robots can enlarge the range of their workspace. Two deformable bodies can be incorporated into the microrobot to enable the steering ability, as demonstrated in SI Appendix , Fig. S34 and Movie S12 . In this design, the straight motion and rightward/leftward motion can be achieved through synchronous and asynchronous deformations of two bodies. Further extension of the presented concept allows us to design microrobots with more complicated 3D shapes as shown in SI Appendix , Fig. S35 and S36 and Movie S13 . SI Appendix , Fig. S35 shows an inchworm-like microrobot with three articulations (left, middle, and right parts). The body of this microrobot contains three individually addressed actuating components, resulting in more deformation morphologies than the microrobot with a single actuating component. SI Appendix , Fig. S35 and Movie S13 demonstrated three deformation morphologies observed during three typical motions (I, II, and III) of inchworms. In the Motion I ( SI Appendix , Fig. S35 C ), the actuation of the middle part allows the microrobot to first lift the right foot, and a subsequent actuation of the left and right parts enables the elongation of the body, until the right foot drops to the ground. Removal of the applied voltage from the three body parts and the left electroadhesive footpad allows the left foot to slide rightward by contracting the body, which completes the rightward locomotion. In the Motion II ( SI Appendix , Fig. S35 D ), the microrobot moves to the right direction using the stepping gait, with both feet contacting with the ground during the motion. In the Motion III ( SI Appendix , Fig. S35 E ), the microrobot just lifts the right foot until the foot contacts a pencil, followed by removing the voltage to allow the microrobot to recover the initial configuration. SI Appendix , Fig. S36 demonstrates the design, fabrication, and upward climbing of a snake-like microrobot on a glass of rod. The deformable body is assembled from a zigzag-shaped 2D precursor, with well-designed local molecular orientation of LCE and 2D pattern of PI frame. The body can be divided into three parts (top, middle, and bottom parts). The top and bottom parts can wrap around the rod to generate frictional forces that balance the weight, and the shape morphing of the middle part can increase/decrease the distance between the top and bottom parts. Therefore, a rational control of these three individually actuated components enables the upward climbing on a rod. To achieve an untethered climbing microrobot ( 36 , 69 – 74 ), an on-board battery (e.g., lithium/polymer battery) or solar cell component can be integrated into our microrobot to serve as power sources. For the microrobot with a body length of 30 mm, the costs of transport (COTs) are estimated as 2.0 × 10 4 , 9.0 × 10 4 , and 1.2 × 10 4 , for climbing upward vertically with stepping gait, moving horizontally with forward-type flipping gait, and transitioning upward between two walls of 90° angle with transition-type flipping gait (refer to SI Appendix , Cost of Transport Calculation ). The magnitudes of the COTs are comparable to or less than other soft robots based on thermally controlled actuators ( 36 , 75 , 76 ). Further improvements of the locomotion speed could focus on accelerating heating and cooling processes of LCE, e.g., by embedding liquid metal droplet ( 77 ) and carbon black nanoparticle ( 78 ), reducing the actuation temperature of LCE ( 35 ), and integrating microfluidic channels to pump cold fluids ( 79 )."
} | 1,680 |
27886299 | null | s2 | 8,374 | {
"abstract": "The Nobel prize in chemistry in 2016 was awarded for 'the design and synthesis of molecular machines'. Here we designed and assembled a molecular machine for the detection of specific RNA molecules. An association of several DNA strands, named multifunctional DNA machine for RNA analysis (MDMR1), was designed to (i) unwind RNA with the help of RNA-binding arms, (ii) selectively recognize a targeted RNA fragment, (iii) attract a signal-producing substrate and (iv) amplify the fluorescent signal by catalysis. MDMR1 enabled detection of 16S rRNA at concentrations ∼24 times lower than that by a traditional deoxyribozyme probe."
} | 157 |
23795153 | null | s2 | 8,375 | {
"abstract": "No abstract available"
} | 5 |
22089132 | PMC3272368 | pmc | 8,377 | {
"abstract": "Legumes ( Fabaceae or Leguminosae ) are unique among cultivated plants for their ability to carry out endosymbiotic nitrogen fixation with rhizobial bacteria, a process that takes place in a specialized structure known as the nodule. Legumes belong to one of the two main groups of eurosids, the Fabidae, which includes most species capable of endosymbiotic nitrogen fixation 1 . Legumes comprise several evolutionary lineages derived from a common ancestor 60 million years ago (Mya). Papilionoids are the largest clade, dating nearly to the origin of legumes and containing most cultivated species 2 . Medicago truncatula ( Mt ) is a long-established model for the study of legume biology. Here we describe the draft sequence of the Mt euchromatin based on a recently completed BAC-assembly supplemented with Illumina-shotgun sequence, together capturing ~94% of all Mt genes. A whole-genome duplication (WGD) approximately 58 Mya played a major role in shaping the Mt genome and thereby contributed to the evolution of endosymbiotic nitrogen fixation. Subsequent to the WGD, the Mt genome experienced higher levels of rearrangement than two other sequenced legumes, Glycine max ( Gm ) and Lotus japonicus ( Lj ). Mt is a close relative of alfalfa ( M. sativa ), a widely cultivated crop with limited genomics tools and complex autotetraploid genetics. As such, the Mt genome sequence provides significant opportunities to expand alfalfa’s genomic toolbox."
} | 370 |
21258353 | PMC3075872 | pmc | 8,378 | {
"abstract": "Robust high-throughput synthesis methods are needed to expand the repertoire of repetitive protein-polymers for different applications. To address this need, we developed a new method, overlap-extension rolling circle amplification (OERCA), for the highly parallel synthesis of genes encoding repetitive protein-polymers. OERCA involves a single PCR-type reaction for the rolling circle amplification of a circular DNA template and simultaneous overlap extension by thermal cycling. We characterized the variables that control OERCA and demonstrated its superiority over existing methods, its robustness, throughput and versatility by synthesizing variants of elastin-like polypeptides (ELPs) and protease-responsive polymers of a glucagon-like peptide-1 analog. Despite the GC-rich, highly repetitive sequences of ELPs, we synthesized remarkably large genes without recursive ligation. OERCA also enabled us to discover “smart” biopolymers that exhibit fully reversible thermally responsive behavior. This powerful strategy generates libraries of repetitive genes over a wide and tunable range of molecular weights in a “one-pot” parallel format."
} | 286 |
20589842 | null | s2 | 8,379 | {
"abstract": "Genome-scale metabolic network reconstructions are built from all of the known metabolic reactions and genes in a target organism. However, since our knowledge of any organism is incomplete, these network reconstructions contain gaps. Reactions may be missing, resulting in dead-ends in pathways, while unknown gene products may catalyze known reactions. New computational methods that analyze data, such as growth phenotypes or gene essentiality, in the context of genome-scale metabolic networks, have been developed to predict these missing reactions or genes likely to fill these knowledge gaps. A growing number of experimental studies are appearing that address these computational predictions, leading to discovery of new metabolic capabilities in the target organism. Gap-filling methods can thus be used to improve metabolic network models while simultaneously leading to discovery of new metabolic gene functions."
} | 230 |
35200470 | PMC8871090 | pmc | 8,380 | {
"abstract": "Based on the good self-healing ability to repair mechanical damage, self-healing hydrogels have aroused great interest and been extensively applied as functional materials. However, when partial failure of hydrogels caused by breaking or dryness occurs, leading to recycling problems, self-healing hydrogels cannot solve the mentioned defects and have to be abandoned. In this work, a novel recyclable and self-healing natural polymer hydrogel (Chitosan/polymethylacrylic acid-: CMA) was prepared. The CMA hydrogel not only exhibited controlled mechanical properties from 26 kPa to 125 kPa with tensile strain from 1357% to 3012%, but also had good water retaining property, stability and fast self-healing properties in 1 min. More importantly, the CMA hydrogel displayed attractive powder self-healing performance. After drying–powdering treatment, the mentioned abandoned hydrogels could easily rebuild their frame structure to recover their original state and performance in 1 min only by adding a small amount of water, which could significantly prolong their service life. These advantages guarantee the hydrogel can effectively defend against reversible mechanical damage, water loss and partial hydrogel failure, suggesting great potential applications as a recyclable functional hydrogel for biomaterials and electronic materials.",
"conclusion": "3. Conclusions In this work, we designed and prepared a novel highly flexible, self-healing and recyclable natural polymer hydrogel. The CMA hydrogel exhibited good flexibility and controlled mechanical properties with the content of MA (tensile strain from 1357% to 3012% under 26 kPa to 125 kPa, and 50% compressive strain under 32 kPa to 176 kPa). Besides, the CMA-3 hydrogel demonstrated remarkable self-healing ability, which could effectively repair the mechanically damaged hydrogel in 1 min and recover 92.9% mechanical strength of its original state in 5 min. More importantly, due to its reversible and dynamic interactions, the CMA-3 hydrogel showed excellent powder self-healing performance. The failure of hydrogel caused by dryness or missing parts could be easily repaired and a complete hydrogel rebuilt with the required shape in 1 min after drying–powdering and addition of water, and its mechanical strength recovered, indicating excellent recyclability. Based on these advantages, the CMA hydrogel should be potentially useful in diverse areas such as tissue engineering, stretchable electronics and wearable or implantable devices.",
"introduction": "1. Introduction Hydrogels are water-swollen polymeric materials displaying good hydrophilic three-dimensional (3D) networks by chemical or physical crossing links [ 1 , 2 , 3 ]. Owing to their peculiar structure, hydrogels exhibits characteristic properties [ 4 , 5 , 6 ], (e.g., high softness, hydrophilic nature, insolubility, swelling behavior and sensitivity to physiological environment). Over recent decades, hydrogel-based functional materials have aroused great interest and been extensively applied in adsorption materials, as drug carriers, in tissue engineering and wearable flexible electronic devices [ 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 ]. However, although hydrogels exhibit obvious advantages in performance, they are abandoned after suffering irreversible damage (e.g., mechanical breaks, dryness or when part of the hydrogel is missing). Due to the lack of effective recycling strategies, the abandoned materials cause serious waste and environment pollution [ 15 , 16 ]. Thus, a novel recyclable hydrogel with good performance should be developed. In recent years, various functional hydrogels have been designed to try to solve these problems. Self-healing hydrogel is considered to be an ideal candidate to address mechanical damage [ 17 , 18 , 19 , 20 ]. When impacted by reversible and dynamic interactions, self-healing hydrogel can effectively rebuild the hydrogel’s frame structure and recover its performance. Jiang et al. [ 21 ] prepared a copolymer hydrogel (poly(MAA-co-BA-co-OEGMA): methacrylic acid/oligo(ethylene glycol) methacrylate/4-Hydroxybenzaldehyde/ethylenediamine) with good self-healing abilities, based on dynamic hydrogen bonds and imine bonds. The cut hydrogel could rapidly merge into one piece in 2 min and recover 87% of the mechanical strength of its original state at 25 °C after 40 min. However, when hydrogels lose water, they cannot maintain good self-healing ability and have to be abandoned. To solve the dryness problem, Pan et al. [ 22 ] used vinylimidazole and hydroxypropyl acrylate to prepare a good self-healing hydrogel with excellent swelling–shrinking properties. The hydrogel had good self-healing ability, while exhibiting good wetness. Based on an elastic frame structure, the hydrogel could recover from a dry state to its original state with addition of water. After 5 swelling–shrinking cycles, it still could maintain stable mechanical performance, but once parts of the hydrogel were missing, the remaining hydrogel could not recover to its original state and maintain stable performance. In a recent study by our groups, a recyclable hydrogel (sodium alginate/polymethyl-acrylic acid) was developed with good self-healing ability and powder self-healing performance [ 23 ]. Under reversible and dynamic interactions, the hydrogel was capable of preventing mechanical damage and the dryness problem, as well as having good recyclability through drying–powdering treatment. Although the powder self-healing SAMA hydrogel exhibited good recyclability, its mechanical properties and self-healing response time are not enough to meet the requirements of applications. Accordingly, a new hydrogel material exhibiting good mechanical properties and fast powder self-healing properties should ideally be developed to recycle hydrogel materials, although this remains highly challenging. In this work, a novel recyclable natural polymer hydrogel (CMA: chitosan/polymethyl-acrylic acid) was designed and prepared. The CMA hydrogel exhibited good mechanical properties under 26 kPa to 125 kPa with tensile strain from 1357% to 3012%, and it could be controlled by regulating the content of hydrogel. Besides, it displayed good self-healing ability, and could effectively heal in 1 min and recover 92.9% mechanical strength of its original state after 5 min at room temperature. More importantly, the CMA hydrogel possessed excellent powder self-healing performance. Failed hydrogel could be easily recycled based on drying–powdering treatment, which could effectively solve the problems of dryness, partially missing hydrogel and recycling problems to prolong the service life in 1 min. Benefiting from the mentioned advantages, this powder self-healing hydrogel will have great potential application as a recyclable functional hydrogel in biomaterials and electronic materials.",
"discussion": "2. Results and Discussion 2.1. Characterization of CMA Hydrogels Figure 1 A illustrates the preparation of the natural polymer hydrogel. Based on the natural frame structure of Chitosan (CS) and functional monomer-methacrylic acid (MA), CMA hydrogels were synthesized through free radical polymerization by reversible hydrogen bonds with 3D network structures. The chemical structure displayed by the CMA hydrogels was analyzed by Fourier-transform infrared spectroscopy (FTIR) and X-ray diffraction (XRD). According to Figure 1 B, the broad peak at 3247 cm −1 corresponded to the –OH and –NH 2 stretching vibration. The peaks at 1642 cm −1 and 1588 cm −1 belonged to the –C=O and -NH bending vibration [ 24 , 25 ]. Moreover, the characteristic peaks of methacrylic acid (MA) at 3018 cm −1 , 1686 cm −1 and 1622 cm −1 , respectively, belonged to the stretching vibrations of –C=C, –COOH and –C=C. In the spectra of the CMA hydrogels, all the characteristic peaks of CS and MA could be observed. Besides, the significant shifting peaks from 3247 cm −1 to 3360 cm −1 and from 1588 cm −1 to 1637 cm −1 indicated the strong dynamic hydrogen bonds. The mentioned results demonstrate the successful preparation of the CMA hydrogels. Figure 1 C shows the XRD pattern of the CMA hydrogels. The two broad characteristic peaks at 2θ = 15° and 2θ = 30° belong to the amorphous structure of the hydrogels. As shown in Figure 1 D, the micro-morphology of the CMA hydrogels exhibited a continuous and regular porous structure, which would be conducive to maintaining good flexibility and water retaining ability of the hydrogels. 2.2. Mechanical and Water Retaining Properties of CMA Hydrogels Mechanical properties are a vital performance parameter against which to assess hydrogel materials. The mechanical strength of CMA hydrogels was analyzed through performing stress strain experiments. As shown in Figure 2 A, all the CMA hydrogels exhibited good stretching ability. With the increase of MA content from 1.0 g to 3.0 g ( Table S1 ), the stretching stress increased from 26 kPa to 125 kPa, with tensile strain increasing from 1357% to 3012%, and Young’s modulus from 20.3 kPa to 42.9 kPa. Besides, the compressing ability of the CMA hydrogels showed excellent performance. With the increase of MA content from 1.0 g to 3.0 g, the compressing stress increased from 32 kPa to 176 kPa under 50% compressing strain ( Figure 2 B). The CMA hydrogels demonstrated controlled mechanical properties to meet different application needs ( Table S2 , Figure S1 ). With different contents of MA, the CMA-3 hydrogel showed optimal mechanical properties with 3012% stretching strain under 125 kPa and 50% compressing strain under 176 kPa. Compared with previously reported self-healing hydrogels, especially in hydrogel biomaterials and hydrogel electrolytes, the CMA-3 hydrogel exhibited much better mechanical properties ( Table S3 ) [ 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ]. Besides, it showed good flexibility. During 4 stretching–recovering cycles, although the stress–strain curves exhibited a small hysteresis loop, the hydrogel could quickly recover to its original shape when the stress released, suggesting excellent self-recovery ability ( Figure 2 C). This could be attributed to the continuous porous structure and strong reversible hydrogen interactions, which could effectively counteract the effect external force to enhance its mechanical properties [ 34 , 35 , 36 ]. To clearly show the gel’s flexibility, the prepared CMA-3 hydrogel stick and disk were used as demonstrations. As shown in Figure 2 D,E, in the continuous stretching and bending processes, the CMA-3 hydrogel exhibited good flexibility without breaking. Besides, it showed excellent compressing ability. Even under a human weight stress ( Figure S2 ), the hydrogel could maintain stability without breaking and immediately recover to its original state when the load was removed. Due to its excellent mechanical properties, the CMA-3 hydrogel was used for the following tests. Its water-retaining property is another important factor for hydrogel materials. To assess its stability, the water loss process of CMA-3 hydrogel was recorded for 48 h at room temperature of 20 °C under 70% RH. As shown in Figure S3 , the CMA-3 hydrogel could maintain more than 85% water after 48 h, which could meet most applications’ needs. This could be attributed to the many hydrophilic groups in the hydrogel, which could effectively block the water movement and protect the water from evaporation. 2.3. Self-Healing Property of CMA- 3 Hydrogel Under reversible hydrogen bonds, the CMA-3 hydrogels exhibited good self-healing ability. To directly demonstrate this self-healing performance, two CMA-3 hydrogel disks were prepared and colored with rhodamine B and rhodamine 6G, respectively. First, the colored hydrogel disks were cut into two pieces. Subsequently, the red half and the orange half were placed in contact along the cutting line and maintained stability at room temperature without any external stimulus or healing agent. After 1 min, it was found that the two half hydrogels successfully self-healed into a single one ( Figure 3 A). Additionally, the healed hydrogel exhibited good mechanical strength to withstand continuous twisting and stretching without cracking, demonstrating good ability to recover its mechanical properties ( Figure 3 B,C). In order to quantitatively assess the self-healing property of the hydrogel, the mechanical strength changes over time of the CMA-3 hydrogel were tested during the self-healing process. As shown in Figure S4A,B , with the increase of self-healing time, the mechanical property gradually recovered. After 5 min, the CMA-3 hydrogel could recover nearly 92.9% strain and 82.7% stress of its original state, respectively. Compared with previously reported self-healing hydrogels, the CMA-3 hydrogels showed much better self-healing properties ( Table S3 ), which could be attributed to the reversible hydrogen bonds. When the hydrogels were cut, considerable reversible interactions were broken, and numerous functional groups separated and exposed on the cutting surface. While the cut hydrogel pieces were put together, the exposed functional groups would immediately interact with each other via hydrogen bonds, and efficiently rebuild the frame structure to realize the hydrogels’ self-healing. To further demonstrate the self-healing property of the CMA-3 hydrogel, the colored hydrogels were cut into small particles as shown in Figure S5 . Subsequently, these particles were put into a heart-shaped mold to record the self-healing process. Within 10 min, a complete heart-shaped hydrogel was obtained with good flexibility. These results indicated that the CMA-3 hydrogels possessed excellent self-healing property, which could effectively defend irreversible mechanical damage to prolong their service life. 2.4. Powder Self-Healing Property of CMA- 3 Hydrogel Dryness is a serious problem of hydrogels. Once water is lost, the hydrogel will quickly dry and shrink, resulting in performance failure, which significantly narrows the application of hydrogels. Fortunately, under strong reversible interactions, the CMA-3 hydrogel exhibited an excellent powder self-healing property, which could effectively solve the dryness problem of hydrogels. As shown in Figure 4 A, the failed CMA-3 hydrogel was dried and powdered. Then, the powder was added into the round shape mold with a little water at room temperature. After 1 min, the powder quickly formed a piece of hydrogel with the mold shape (SI-V1). This is attributed to the benefit of the strong dynamic hydrogen bonds in the hydrogels. The added water served as a necessary medium to help the dry hydrogel powder rebuild the frame structure through strong dynamic interactions. To further verify its good powder self-healing property, the recycled CMA-3 hydrogel was dried and powdered again. The powder was put into different shaped molds to recycle the hydrogel. As shown in Figure 4 A, the recycled powder could regenerate the hydrogels with various mold shapes, which indicated good recyclability. It was noteworthy that the recycled hydrogel also recovered its mechanical strength ( Figure 4 B). Under the continuous stretching–releasing process, the recycled hydrogel could maintain good flexibility. Based on this excellent powder self-healing performance, the CMA-3 hydrogel could easily address the dryness defect of hydrogels, suggesting good recyclable ability. Besides, due to its powder self-healing ability, the hydrogel not only could address the dryness problem, but serve as an agent to repair hydrogels with parts missing. By adding a little prepared hydrogel powder and water, it could quickly repair the defects and reform a complete hydrogel in 5 min. Meanwhile, the repaired hydrogel kept good flexibility. As shown in Figure 4 C, although a fracture could be observed at the contact position, the self-healed hydrogel exhibited good mechanical properties to withstand the tensile force without splitting. Based on the excellent advantages of good flexibility, self-healing ability and powder self-healing performance, the CMA hydrogel could effectively overcome the defects of mechanical damage, dryness and missing parts, which lead to the performance failure of hydrogel materials and their abandonment. More importantly, the attractive recyclability could guarantee the sustainability of hydrogel materials to prolong their service life."
} | 4,096 |
31504731 | PMC6878951 | pmc | 8,382 | {
"abstract": "Abstract Reverse gyrase (RG) is the only protein found ubiquitously in hyperthermophilic organisms, but absent from mesophiles. As such, its simple presence or absence allows us to deduce information about the optimal growth temperature of long-extinct organisms, even as far as the last universal common ancestor of extant life (LUCA). The growth environment and gene content of the LUCA has long been a source of debate in which RG often features. In an attempt to settle this debate, we carried out an exhaustive search for RG proteins, generating the largest RG data set to date. Comprising 376 sequences, our data set allows for phylogenetic reconstructions of RG with unprecedented size and detail. These RG phylogenies are strikingly different from those of universal proteins inferred to be present in the LUCA, even when using the same set of species. Unlike such proteins, RG does not form monophyletic archaeal and bacterial clades, suggesting RG emergence after the formation of these domains, and/or significant horizontal gene transfer. Additionally, the branch lengths separating archaeal and bacterial groups are very short, inconsistent with the tempo of evolution from the time of the LUCA. Despite this, phylogenies limited to archaeal RG resolve most archaeal phyla, suggesting predominantly vertical evolution since the time of the last archaeal ancestor. In contrast, bacterial RG indicates emergence after the last bacterial ancestor followed by significant horizontal transfer. Taken together, these results suggest a nonhyperthermophilic LUCA and bacterial ancestor, with hyperthermophily emerging early in the evolution of the archaeal and bacterial domains.",
"introduction": "Introduction Understanding the nature of the last universal common ancestor of extant life (LUCA) is one of the most difficult, yet important problems in evolutionary biology. If we were able to determine the genes encoded by the LUCA, we could make important conclusions regarding the evolutionary histories of all living organisms, as well as make predictions about the environment in which the LUCA lived. However, deciphering phylogenetic relationships dating back billions of years is a process fraught with difficulty. Not least because continual mutation over such time periods saturates sequences, erasing earlier phylogenetic signals that may exist, but also because mechanisms such as horizontal gene transfer act to introduce phylogenetic conflict between protein histories, further decreasing our ability to resolve such ancient relationships. Hence, the field of early evolutionary biology is one which is prone to disagreements, even when considering similar data sets. A poignant example of such disagreement comes from the phylogenies of reverse gyrase (RG), the only known hyperthermophile-specific protein, ubiquitously encoded by the genomes of hyperthermophilic organisms and absent from mesophiles ( Forterre 2002a ; Kaufmann 2006 ; Brochier-Armanet and Forterre 2007 ; Heine and Chandra 2009 ). Understanding the evolutionary history of RG is important as the presence or absence of this gene in ancestral genomes (such as the LUCA) would allow us to infer a crude optimal growth temperature for these long-extinct species. The presence of RG appears to be incompatible with mesophily, and conversely, the absence of RG appears incompatible with hyperthermophily. Thus, the presence of a gene encoding RG would infer a hyperthermophilic or thermophilic lifestyle excluding the option of a mesophilic lifestyle, and the absence, a mesophilic or moderately thermophilic growth condition, to the exclusion of hyperthermophily. This predictive ability is a powerful tool in evolutionary biology, where the optimal growth temperature of long-extinct organisms plays an important role in understanding genome evolution (e.g., “thermoreduction” [ Forterre 1995 ]; protein and RNA evolution [ Boussau et al. 2008 ; Groussin and Gouy 2011 ], etc.). In order to make inferences about the presence or absence of RG in ancestral organisms, it is therefore vital to have a robust phylogeny for RG. However, the limited genetic data for hyperthermophilic organisms have restricted our ability to make such generalizations, and the small body of literature regarding RG evolution seems to flip-flop between the presence and absence of RG in LUCA. Early analyses for the presence of RG were carried out experimentally by looking for positive supercoiling activity in cell lysates ( Collin et al. 1988 ; Bouthier de la Tour et al. 1990 ). Though this allowed the identification of new RG-encoding species, these early experiments failed to detect RG in bacteria (and some archaeal species) due to the presence of the antagonistic DNA gyrase ( Guipaud et al. 1997 ). Later, the discovery of RG activity in bacteria (namely in the Thermotogales ) suggested that RG may be a more ancestral protein than previously thought, potentially evolving before the divergence of the bacterial and archaeal lineages ( Bouthier de la Tour et al. 1991 ). Even at this early stage, the presence/absence of RG in the LUCA became an important factor in unraveling the nature of the ancestral life on earth ( Forterre et al. 1995 ). The first published phylogeny for RG included 13 sequences, and did not resolve the monophyly of the bacterial and archaeal domains ( Forterre et al. 2000 ), suggesting that RG could not have evolved solely by vertical descent in the two domains, casting doubt over its presence in the LUCA. A later analysis was able to recover the bacterial and archaeal monophyly using a data set of 32 RG sequences; however, the domain separation was only weakly supported, and the bacterial tree did not reflect a canonical 16S phylogeny leading the authors to hypothesize an Archaea-to-Bacteria transfer for RG ( Brochier-Armanet and Forterre 2007 ). Subsequently, another group was again able to recover the bacterial–archaeal domain separation in a phylogeny of only 15 sequences ( Heine and Chandra 2009 ). Although these monophyly-recovering analyses suggest the direct descent of RG from the LUCA, the data sets used were small and the intradomain tree topologies were not as would be expected from an ancient protein evolving independently in the two domains. The most complete RG phylogeny to date used a data set of 97 sequences, and identified RG as a candidate LUCA protein due to the recovered monophyly of the bacterial and archaeal domains, prompting the authors to conclude that the LUCA was likely a hyperthermophile ( Weiss et al. 2016 ). Unfortunately, the topology of the recovered tree was not analyzed in-depth and upon closer examination it is clear that this phylogeny suffers the same problems observed previously, namely the branch between Archaea and Bacteria was rather short and the clades produced in the analysis are atypical and do not conform to the canonical 16S or universal protein phylogenies (tree reproduced in supplementary fig. 1 , Supplementary Material online). Therefore, the conclusion that RG was encoded by the LUCA is supported weakly, at best. With the quantity of available genetic data growing exponentially, and increasing effort being made to sequence the genomes of archaeal species (many of which are hyperthermophiles), we thought it important to update the phylogeny of RG, and the evolutionary conclusions this can achieve. Using bioinformatics techniques, we reveal 376 RG sequences from 247 organisms across the bacterial and archaeal domains. Phylogenetic reconstruction of these sequences does not resolve the monophyly of the two domains, but rather reveals multiple potential horizontal transfer events. These results suggest RG was not present in the LUCA, but rather evolved after the divergence of the lineages leading to the LBCA and LACA. We therefore conclude that LUCA was a mesophile or moderate thermophile, with hyperthermophily evolving later, possibly before the emergence of the LACA.",
"discussion": "Discussion The work presented here represents the largest data set of RG sequences to date. Phylogenetic analyses of this data set strongly suggests the absence of RG in the LUCA since the archaeal and bacterial RG do not form two monophyletic clades ( fig. 2 ). Indeed, all analyzed tree topologies with monophyletic archaeal and bacterial clades were soundly rejected for the RG data set by a suite of tree topology selection tests. Furthermore, the short branch length between any interdomain clades of our phylogenies ( fig. 2 ) indicates a period of divergence inconsistent with the tempo of evolution between LUCA and the common ancestors of Archaea and Bacteria; the branches between the different archaeal and bacterial clades are all very short, suggesting the existence of a single version of RG. In contrast, using the same set of species, we have shown here that not only do Archaea and Bacteria form two monophyletic clades in phylogenies of markers inferred to be present in LUCA such as EF-G, RNA polymerase, and 16S rRNA, but that the branch between these two clades is very long ( fig. 3 ). It could be argued that the shorter interdomain branches of RG relative to other LUCA proteins simply reflects a higher rate of sequence conservation in RG; however, if true, this requirement for sequence conservation must have been transient (crown branch lengths are similar between RG and the LUCA proteins; fig. 3 ), and only present during the times where other proteins show the greatest divergence. Such a scenario is at odds with all other proteins inferred to be present in the LUCA; using different species, long branches separating Archaea and Bacteria were also systematically observed in the phylogenies of 36 universal proteins (most likely present in the LUCA) ( Da Cunha et al. 2017 ). Our RG results are in contradiction with recent attempts to reconstruct the proteins of the LUCA ( Weiss et al. 2016 ) ( supplementary fig. 1 vs. supplementary fig. 3 , Supplementary Material online). This could be explained by the difference in the number of sequences used in the two analyses (97 vs. 376) and the fact that Weiss and colleagues do not include the branch length between the Archaea and Bacteria as a criterion to indicate the presence of a protein in the LUCA. This branch was very short in the RG tree of Weiss and colleagues ( supplementary fig. 1 , Supplementary Material online) and we could not recover the monophyly of Archaea and Bacteria using their data set, suggesting that the monophyly versus paraphyly of Archaea and Bacteria is sensitive to some parameters of tree reconstruction, further suggesting a nondistinct separation of these usually highly divergent domains. The absence of RG in LUCA, combined with the apparent requirement of RG for growth at high temperature, suggests the existence of a nonhyperthermophilic LUCA. Although RG knockout strains do appear viable at high temperatures in the laboratory (at least in some species), such strains invariably suffer from temperature-sensitive growth defects ( Atomi et al. 2004 ; Lipscomb et al. 2017 ; Zhang et al. 2018 ). Thus the possibility exists that hyperthermophily evolved in the absence of RG, with the subsequent emergence of RG conferring a huge selective advantage such that it rapidly became fixed in all hyperthermophilic lineages. Additionally, we cannot exclude the possibility that another protein served a similar function in the LUCA, conferring a hyperthermophilic growth condition; however, there is no evidence for the existence of such a protein in modern organisms. Thus, this hypothetical protein would have to have been lost in an intermediate mesophilic state of both of the post-LUCA lineages (leading to the LBCA and LACA), or lost in one lineage and replaced with the emergent RG in another. Such a scenario seems unlikely, especially considering the consistency of our RG results with those observed through independent methods. For example, work on ancestral protein and rRNA reconstructions ( Galtier et al. 1999 ; Boussau et al. 2008 ; Groussin and Gouy 2011 ) suggest that the LUCA was either a mesophile or a moderate thermophile. Additionally, thermoadaptations observed in membrane lipids ( Langworthy and Pond 1986 ; Wiegel and Michael 2014 ) and modifications of tRNA ( Edmonds et al. 1991 ; Lorenz et al. 2017 ) are nonhomologous between bacteria and archaea, suggesting hyperthermophily evolved independently in each lineage rather than being a shared trait from the LUCA. A nonhyperthermophilic LUCA is also in agreement with the idea that LUCA was an organism simpler than modern ones, with smaller ribosomes ( Fox 2010 ) and possibly an RNA genome ( Poole and Logan 2005 ). Indeed, the origin of most DNA replication proteins cannot be traced back to LUCA ( Forterre 2002b , 2013 ), and it seems that RG is not an exception. The transition from a LUCA with an RNA genome to archaea and bacteria with DNA genomes could also explain why the tempo of evolution drastically slowed between LUCA and the two prokaryotic ancestors, considering that DNA can be replicated and repaired much more faithfully than RNA ( Forterre 2006 ). With respect to our RG phylogenies, and RG evolution in general, the short branch lengths between bacterial and archaeal clades would place the emergence of RG in the age of DNA cells, that is, more recently than the time of a rapidly evolving RNA-based LUCA (and post-LUCA lineage). This, perhaps, would seem logical considering the strict DNA substrate-dependence of RG, and RG conferring adaptation to hyperthermophilic growth temperatures—a state likely incompatible with RNA genomes ( Ginoza and Zimm 1961 ). Finally, our work highlights the fact that a widespread distribution across bacterial and archaeal taxa is not sufficient evidence for inferring the presence of a protein in the LUCA. Rather, a clear, well-separated monophyly of Archaea and Bacteria, and deep congruence with canonical phylogenetic relationships should be demonstrated (e.g., those exemplified by RNA polymerase, EF-G, 16S rRNA etc.)."
} | 3,514 |
30984090 | PMC6449629 | pmc | 8,383 | {
"abstract": "The adaptive benefits of individual specialization and how learning abilities correlate with task performance are still far from being well-understood. Red wood ants are characterized by their huge colonies and deep professional specialization. We hypothesized that red wood ants Formica aquilonia form aversive learning after having negative encounters with hoverfly larvae differently, depending on their task specialization. We tested this hypothesis, first, by examining whether hunters and aphid milkers learn differently to avoid the nuisance of contacts with syrphid larvae, and, second, by analyzing the difference between learning in “field” and laboratory-reared (naïve) foragers. During the first interaction with the syrphid larva in their lives the naïve foragers showed a significantly higher level of aggressiveness than the members of a natural colony. Naïve foragers applied the “mortal grip,” “prolonged bites,” and “nibbling” toward the enemy with a significantly higher frequency, whereas members of both “field” groups behaved more carefully and tried to avoid encounters with the larva. The aphid milkers, who had a negative experience of interaction with the larva, being “glued” with its viscous secretion, behaved much less aggressively in the follow-up experiments after 10 min and even 3 days, thus exhibiting the shaping of both short- and long-term memories. However, both “field” hunters and naïve foragers demonstrated no signs of aversive learning. These data provide some new insights into the relationship between task specialization and learning performance in ants. Given our previous results, we speculate that scouts and aphid milkers are the most cognitively gifted specialists in red wood ants, whereas hunters and guards are rather brave than smart.",
"introduction": "Introduction Eusocial insect species form persistent colonies, where groups of individuals (castes) perform different tasks (division of labor). Not only members of the reproductive castes, such as queens, have clearly defined responsibilities with workers, but the workers themselves also have particular tasks such as caring for the young, defense, foraging, nest-building activities, and so on (for a review see: Robinson, 1992 ). The main characteristic of insect societies is the way tasks are distributed among group members. The adaptive benefits of individual specialization and how learning abilities correlate with task performance are still far from being understood. Ants are good candidates for studying these problems as a highly diverse and successful group of social hymenopterans, which, unlike bees and wasps, consists only of eusocial species (for a review see: Reznikova, 2018 ). There are more than 12,000 ant species in the world, with different colony sizes (from tens to millions of individuals), social life and styles of cooperation, from single foraging to mass recruiting. Ant species display different modes of the division of labor and the specialization of workers on various tasks. Caste polyphenism is rather expressive in some ant species, which harbor special sub-castes of workers profoundly different by morphology, physiology, and behavior (for a review see: Hölldobler and Wilson, 1990 ). In leaf-cutting ants, for instance, tiny “mini workers” cultivate fungi in the subterranean nest to feed the larvae. Other workers have an up to 200-fold increased body weight, leaving the nest for long foraging trips, and bringing back leaves which are used as substrate for the fungi ( Wilson, 1980 ). However, a morphological caste system may also have costs, as it may prevent a colony from rapidly adjusting caste ratios, increase the energetic cost of rearing or limit the task repertoire ( Oster and Wilson, 1978 ). In the majority of ant species, workers only specialize in different tasks behaviorally. In some ant species, behavioral specialization is strongly determined by chronological age and physiological development (temporal polyethism): young workers typically perform safe tasks inside the nest, such as nursing the brood, and only later in life move on to more risky tasks outside the nest, such as foraging or territorial defense ( Hölldobler and Wilson, 1990 ; Beshers et al., 2001 ; Giraldo and Traniello, 2014 ). In other species, task specialization among workers may be shaped by their size polymorphism, genetic background, experience, and social interactions, which is also partly influenced by age ( Tripet and Nonacs, 2004 ; Schwander et al., 2005 ; Ravary et al., 2007 ; Helanterä et al., 2013 ; Giehr et al., 2017 ; Doering et al., 2018 ). Behavioral and cognitive mechanisms of the specialization of workers on different tasks remain the least studied in this area. Red wood ants (the Formica rufa group) are possibly the most promising group for studying the role of learning and experience in task specialization among workers ( Reznikova, 2018 ). In comparison with many sympatric species, the mound-building red wood ants have hundreds of times more individuals in their colonies and spacious feeding territories ( Dlussky, 1967 ; Rosengren and Sundström, 1987 ). Every day red wood ants face complex vital problems: for example, in order to obtain honeydew, the basic food for adults, thousands of colony members have to find and memorize locations of a large number of aphid colonies within such a huge three-dimensional space as a tree is for an ant ( Reznikova, 2008 ). Dobrzanska (1958) demonstrated that in red wood ants, groups of individuals return repeatedly to approximately the same parts of the colony’s feeding territory on the ground and work together there. Studying site allegiance in red wood ants, Rosengren and Fortelius (1987) characterized red wood ants as “replete ants” storing not lipids in their fat bodies but habitat information in their brains, and members of this group of species used to be a relevant model for studying spatial learning and intelligence ( Reznikova and Ryabko, 1994 ; Nicholson et al., 1999 ). It was demonstrated in early studies that in red wood ants, out-nest workers include hunters and collectors of nest material acting on the ground, aphid milkers collecting honeydew on the trees and hunters collecting prey there, as well as guards defending the nest entrances ( Otto, 1958 ; Horstmann, 1973 ). Studies at the individual level revealed deep professional specialization, that is, considerable behavioral differences between members within different task groups. For instance, the task group of aphid tenders turned out to include professional subgroups such as scouts, aphid milkers (“shepherds”), aphid guards, and carriers ( Reznikova and Novgorodova, 1998a ; Reznikova, 2007 ). Experimental studies of interactions of the red wood ant Formica aquilonia with ground beetles, their eternal enemies, large and dangerous, revealed that nest guards, and hunters are much more aggressive than aphid milkers ( Reznikova and Dorosheva, 2004 ; Dorosheva et al., 2011 ). Experiments with other intruders, such as spiders and the small parasitic rove beetle, revealed a context-dependent specialization in colony defense in the red wood ant F. rufa : small workers were better at preventing brood predation than larger workers, and nurses and workers at nest entrances were more aggressive toward parasitic beetles than extranidal foragers ( Parmentier et al., 2015 ). Reznikova and Ryabko (1994 , 2011 ) revealed in red wood ants a group-retrieving mode of foraging basing on the difference of the searching activity and cognitive abilities between scouts and recruited foragers. Scouts appeared to be able to grasp regularities in the sequences of turns (right and left) in the “binary tree” maze on the way to the goal, and use them to optimize their messages to recruited foragers, whereas the recruited foragers can only memorize and not transfer the information. The sophisticated communication system between the scouts and the recruited foragers is even more complicated than the honeybee dance language (see details in: Reznikova, 2017 ). Experiments with a battery of behavioral tests demonstrated that scouts form spatial memory faster and keep information longer than recruited foragers. They were, in general, more exploratory than other out-nest workers, more readily switched between different activities in unfamiliar situations, and, although displaying an intermediate level of aggressiveness between aphid milkers and nest guards, they never attacked the enemy directly ( Atsarkina et al., 2014 ; Reznikova, 2018 ). A question then arises about the distribution of cognitive responsibilities within the ant colony (sensu: Reznikova, 2008 ), that is, about the differentiation between groups in their abilities to perform cognitively demanding tasks. In this study, we concentrate on how representatives of different task groups shape natural aversive learning in the context of repeated interactions with an enemy. Similar with Hollis et al. (2017) study, we use the term “natural aversive learning” meaning ants’ behavior toward their natural enemies or/and predators, although agreeing with Hénaut et al. (2014) that even if a species demonstrates aversive learning in the laboratory, most often it is difficult to determine whether such a capacity is likely to occur in natural conditions (see also: Bernays, 1993 ). Aversive learning has been studied in ants mainly on the basis of one-trial learning. Ants’ memory for an aversive event was tested shortly after their unpleasant experience. For example, Ectatomma tuberculatum appeared to learn how to break quickly through the spiders’ web and kept this memory during 15 min after the first experience ( Hénaut et al., 2014 ), Formica pratensis retained the memory of a single unpleasant collision with a hoverfly larva for 10–30 min after the event ( Novgorodova, 2015 ), and pavement Tetramorium ants learned to avoid antlion traps following a single successful escape from a pit for 1 min after the encounter ( Hollis et al., 2017 ). Long-term memory for an aversive event was demonstrated in the experiments of Dejean (1988) with Odontomachus troglodytes . These ants kept the memory about encounters with the chemically defensive larva of the African chrysomeline during 28 days. As far as we know, natural aversive learning in ants has never been considered in the context of task specialization. The interaction between red wood ants and hoverfly larvae can serve as a natural and somehow unique experimental model in studying differences in learning abilities between members of different task groups. Hoverfly (Diptera, Syrphidae) females use semiochemicals to locate aphid colonies and to oviposit eggs from which aphidophagous larvae hatch (see reviews in: Detrain et al., 2017 ). It is known that aphids provide a vital energy source that is essential for the survival and growth of the ant colony ( Addicott, 1978 ; Skinner and Whittaker, 1981 ), and ants actively protect their trophobionts from all aphidophages ( Novgorodova and Gavrilyuk, 2012 ). Hoverfly larvae use adhesive saliva to incapacitate an attacker ( Eisner, 1972 ; Rotheray, 1986 ). Ecological aspects of intricate interplay between ants and hoverfly larvae have been studied recently on F. pratensis by Novgorodova (2015) and on Lasius niger by Detrain et al. (2017) . Both studies showed that to prevent predation of aphids by hoverfly larvae, ants demonstrated aggressive behavior; once bitten by ants, third instar hoverfly larvae released a droplet of viscous and sticky secretion from the mouth, which hardens like glue. These actions make ants stop their attacks to clean themselves and decrease aggression toward the larvae for a few minutes. Since the impact of the larvae on ants is not dangerous, but only unpleasant for them, ants can interact with the enemy many times in a row. This experimental model allows one to check if ants are learning to avoid unpleasant encounters. We, therefore, hypothesized that the members of different task groups form aversive learning after having negative encounters with hoverfly larvae differently depending on their tasks. We tested this hypothesis, first, by examining whether hunters and aphid milkers from the natural colony of F. aquilonia learn differently to avoid the nuisance of contacts with hoverfly larvae, and, second, by analyzing the difference between learning in the “field” and naïve foragers. We successfully showed that the aphid milkers shaped both short- and long-term memories about a negative experience of interaction with the enemy, whereas hunters and naïve foragers did not change their behavior after the unpleasant event. As far as we know, this is the first demonstration of natural aversive learning in ants in the context of task specialization.",
"discussion": "Discussion In our experiments with F. aquilonia , during the first interaction with the syrphid larva in their lives, the naïve foragers showed a significantly higher level of aggressiveness than the members of a natural colony. Naïve foragers applied the mortal grip, prolonged bites, and nibbling toward the enemy with significantly more frequency, whereas members of both “field” groups behaved more carefully and tried to avoid encounters with the larva. The aphid milkers, who had a negative experience of interaction with the larva, being “glued” with its viscous secretion, behaved much less aggressively in the follow-up experiments after 10 min and even 3 days, thus shaping both short- and long-term memories. However, both “field” hunters and naïve foragers demonstrated no signs of aversive learning. As far as we know, this study provides the first link between the natural aversive learning performance and task specialization in ants. A question then arises, what behavioral and cognitive traits of the members of different task groups are responsible for the differences in their abilities to form natural aversive learning. First, the difference in the previous experience could influence the abilities to form natural aversive learning in the hunters and the aphid milkers in our experiments with hoverfly larvae. It has been recently demonstrated that in the red wood ant F. rufa workers inside the nest interact more frequently with the myrmecophile parasites than foragers, and prior encounter and greater experience in attacking these parasites could cause nurses and mound workers to recognize them more rapidly as a threat than foragers would do and to have a lower threshold to initiate aggression ( Parmentier et al., 2015 ). In our experiments, aphid milkers taken from aphid colonies on a tree were more likely to have experience of previous encounters with the hoverfly larvae and other aphidophages such as ladybird larvae, all chemically defensive, whereas land hunters were unlikely to encounter these insects, and naïve foragers completely lacked the previous experience of interaction with chemically defensive insects. This assumption, although indirect, is also confirmed by differences in the behavior of hunters and naïve foragers toward the enemy. In our experiments, naïve foragers of 13–14 months of age, applied much more aggressive reactions toward the hoverfly larvae than the members of both “field” groups. During the follow-up experiments, naïve foragers behaved less carefully than “field” hunters. Not only did they not learn to avoid the syrphids, but in some cases, they passed on to predatory behavior toward these insects. Since syrphid larvae may occasionally be part of the ant diet ( Punttila et al., 2004 ), naïve foragers can recognize them as “a general image of a victim” such as many larvae of other insects permanently found in the ants’ prey ( Iakovlev et al., 2017 ), rather than “an enemy image.” The hoverfy larva, with its relatively safe gluing secretion, is much less dangerous than, say, predatory ground beetles who can kill the ants, and also have chemical protection. It has been demonstrated earlier that red wood ants possess an innate template for perception and identification of an “enemy image” including such features of the predatory ground beetles as dark coloration, the size, the presence of “outgrowths” (legs, antennae), body symmetry, the rate of movement, and scent ( Dorosheva et al., 2011 ; Reznikova and Dorosheva, 2013 ). However, the ability to single out the key features and complete the integral image seems to require accumulation of experience ( Reznikova and Iakovlev, 2008 ), and hunters are much more cautious toward the ground beetles than both nest guards and naïve workers ( Iakovlev, 2010 ). In the present study, “field” hunters, although did not learn to avoid the negative encounters with hoverfly larvae, behaved more careful than naïve foragers, perhaps based on their experience of collisions with some chemically defensive insects on the soil surface. The second reason explaining differences in the behavior of representatives of different ants’ task groups may be the manifestation of a set of task-specific behavioral features. It was demonstrated earlier that in red wood ants members of each task group possess a stable set of distinct behavioral characteristics ( Reznikova, 2011 ). Experiments of Iakovlev (2010) with the battery of behavioral tests, in which ants interacted with artificial models of natural objects as well as with live predatory beetles, revealed particular suits of behavioral features of the members of different task groups. For instance, aphid milkers display a high level of exploratory activity with the preference for artificial grass stems, low aggressiveness, and evasion of contacts with the predators. At the sight of a stuffed blue tit, guards react aggressively and try to bite, whereas aphid milkers continue milking aphids or jump from the branch down ( Iakovlev, 2010 ; Reznikova, 2011 , 2017 ). Hunters display much more agility and aggressiveness than aphid milkers, and the low level of exploration activity ( Iakovlev, 2010 ). There are similarities in behavioral traits of hunters and nest guards; however, the hunters are more careful, and in contrast to the nest guards, they never use the most dangerous methods of dealing with predatory beetles, such as mortal grip and long bites ( Reznikova and Iakovlev, 2008 ). Similar correlations between behavior and specialization were revealed in other ant species. In Myrmica rubra colonies patrollers are bolder, more aggressive and more active than foragers and brood carers ( Chapman et al., 2011 ). In Camponotus \n aethiops , the exploratory activity of workers in the open field significantly predict learning performance: “active-explorers” were slower in appetitive olfactory learning than “inactive-explorers,” and they are also more aggressive ( Udino et al., 2017 ). In the present experiments, the ability of aphid milkers to learn how to avoid hoverfly larva can be associated with their low aggressiveness and conflict avoidance. It is likely that battles with aphidophages are left to the aphid guards who belong to the same aphid tending group and possess similar behavioral traits with the nest guards ( Reznikova and Novgorodova, 1998a ). In sum, previous experience, innate level of aggressiveness and exploration, and templates of vital attention objects, together shape the difference in learning between representatives of different task groups. It is difficult, if not impossible, to separate the task specialization from the experience gained. Naive foragers in our experiments can serve as an example. Lacking experience with natural enemies, competitors and symbionts, having only vague templates of food and enemy images, they applied highly aggressive reactions toward dangerous animals, and did not learn how to avoid encounters. Keeping in mind the high cognitive abilities of scouts as the most intelligent task group in red wood ants’ colony described earlier ( Reznikova and Ryabko, 2011 ), one can suggest that individual variation in aggressiveness, peculiarities of exploratory activity, orientation, learning, and memory underlies the specialization of workers in performing various tasks. Given our previous results, we speculate that scouts, and aphid milkers are the most cognitively gifted task groups in red wood ants, whereas hunters and guards are rather brave than smart. There is much work to be done to evaluate the fitness consequences of deep professional specialization in red wood ants. The role distribution is somewhat rigid in these species. In our experiments with aphid milkers and aphid guards belonging to the same aphid tending group, when ants were experimentally forced to change their roles, much food was lost ( Reznikova and Novgorodova, 1998a ; Reznikova, 2007 ). In species with small colony sizes of about 100–300 workers such as Temnothorax species, specialists are no better at their jobs than generalists, and sometimes even perform worse. In addition, most of the work in the colony is not performed by the most efficient workers ( Dornhaus, 2008 ). Moreover, strict specialization is disadvantageous for a colony’s annual reproduction and growth during slave raids ( Jongepier and Foitzik, 2016 ). It is possible that in red wood ants, with their colony sizes up to million individuals, the effectiveness of deep professional specialization is connected with their sophisticated communication system based on transferring from scouts to foragers the exact messages about the coordinates of remote goals ( Reznikova, 2017 ). We suggest that further study of the distribution of cognitive responsibilities within colonies of different sizes and levels of specialization in different ant species may help to revise our understanding of the benefits of colony organization."
} | 5,450 |
38439870 | PMC10909651 | pmc | 8,384 | {
"abstract": "To switch the over-reliance on fossil-based resources, curb environmental quality deterioration, and promote the use of renewable fuels, much attention has recently been directed toward the implementation of sustainable and environmentally benign ‘waste-to-energy’ technology exploiting a clean, inexhaustible, carbon-neutral, and renewable energy source, namely agricultural biomass. From this perspective, anaerobic co-digestion (AcoD) technology emerges as a potent and plausible approach to attain sustainable energy development, foster environmental sustainability, and, most importantly, circumvent the key challenges associated with mono-digestion. This review article provides a comprehensive overview of AcoD as a biochemical valorization pathway of crop residues and livestock manure for biogas production. Furthermore, this manuscript aims to assess the different biotic and abiotic parameters affecting co-digestion efficiency and present recent advancements in pretreatment technologies designed to enhance feedstock biodegradability and conversion rate. It can be concluded that the substantial quantities of crop residues and animal waste generated annually from agricultural practices represent valuable bioenergy resources that can contribute to meeting global targets for affordable renewable energy. Nevertheless, extensive and multidisciplinary research is needed to evolve the industrial-scale implementation of AcoD technology of livestock waste and crop residues, particularly when a pretreatment phase is included, and bridge the gap between small-scale studies and real-world applications.",
"conclusion": "9 Conclusion The valorization of lignocellulosic field-based residues and livestock waste as significant bioenergy resources through anaerobic co-digestion represents a potent waste management technology and a greener bioenergy production route. This approach contributes to long-term energy security, reduces GHG emissions, and mitigates environmental and health threats exacerbated by conventional agricultural waste disposal practices such as open-air burning, landfills, and random piling. It also enhances biological and physico-chemical soil quality through the land application of digestate as a bio-fertilizer and soil conditioner. Additionally, the synergistic effect of co-digesting these substrates offers better process stability, adjusts nutrient imbalances, improves buffering ability, stabilizes the C/N ratio, dilutes toxicity from inhibitory compounds, and ultimately increases CH 4 yield. Looking forward, ongoing research, technological innovations in reactor design, process monitoring, and feedstock pretreatment methods, as well as comprehensive assessments encompassing technical, economic, and environmental aspects, hold the potential to further optimize process parameters, increase biogas yields, and improve digestion stability and efficiency.",
"introduction": "1 Introduction The exponential growth of the global population, industrialization, hastened urbanization, excessive dependency, and foreseeable exhaustion of fossil fuels (coal, natural gas, and crude oil) are the most disquieting and acute challenges of the modern era [ 1 ]. All these factors give rise to widespread issues rooted in waste accumulation, oil price hikes, air pollution, climate change, and global warming due to the emissions of GHGs and the release of sequestered carbon into the atmosphere [ 2 , 3 ]. Furthermore, when considering emissions by fuel type in 2020, fossil fuel consumption accounted for 93.22% of global CO 2 emissions (the percentages are 31.81%, 21.26%, and 40.15% for oil, gas, and coal, respectively) [ 4 ]. To alleviate the over-dependence on fossil-based resources and curb the global carbon footprint, it is necessary to hunt for an alternate, renewable, and clean energy source that can significantly contribute to environmental conservation, long-term energy security, and sustainable economic growth [ 5 ]. In this perspective, developing and developed countries have realigned the “take, make, and dispose\" linear and conventional economic model of production and consumption with a greener, closed-loop, and circular bio-economy model [ 6 , 7 ]. In this model, biomass resources are efficiently and sustainably valorized in multi-output and integrated production chains such as biorefineries, while also fully recycling the waste and encouraging the long-term optimization of biomass value via cascading [ 8 ]. The circular bio-economy broadly means the sustainable valorization of renewable biological resources and waste streams into a multitude of useful and value-added products (biopolymers, food, bio-based chemicals, feed) and bioenergy (power, biofuels, heat) [ 9 , 10 ], while contributing to the Sustainable Development Goals adopted by the United Nations Framework Convention on Climate Change [ 4 ]. Among auspicious bioenergy resources, agricultural biomass emerges as a substantial candidate to satisfy future energy demands, disregarding the harmful environmental problems because it is a carbon-neutral, abundant, and inexhaustible feedstock [ 11 , 12 ]. Lignocellulosic agri-waste pertains to second-generation biomass substrates and is typically categorized into crop/plant-based residues (such as wheat straw, corn stover, prunings, sorghum stalks, etc.) and livestock wastes (manure, droppings). Due to their significant annual generation in large quantities, the conversion and proper management of these wastes through the AcoD process represent one of the most promising bioenergy production strategies and potential energy sources for the future [ 13 ]. AcoD, or biomethanation process, denotes the concurrent digestion of feedstocks with complementary properties and their decomposition through the syntrophic interaction of various microbial populations under anoxic conditions across four consecutive phases (hydrolysis, acidogenesis, acetogenesis, and methanogenesis) [ 14 ]. The expression of the overall bioconversion reaction of biowaste organic fractions is given in Eq. (1) [ 15 ]: (1) C c H h O o S s + wH 2 O → mCH 4 + nNH 3 + SH 2 S + (c-m) CO 2 Where, w = 1/4 (4c-h-2o+3n+3s) and m = 1/8 (4c + h-2o-3n-2s). The biodegradable portions of biowastes (C c H h O o S s ) are mostly composed of lipids (C 12 H 24 O 6 ), carbohydrates (C 6 H 12 O 6 ), and proteins (C 13 H 25 O 7 N 3 S) [ 15 , 16 ]. The deconstruction of feedstock and the segregation of cell-wall constituents (holocellulose and lignin) represent critical and intricate steps. This complexity arises from the inherent composition and structural features of the lignocellulosic material, encompassing cellulose crystallinity, polymerization degree, pore volume, degree of lignification, accessible surface area, and particle size. These factors often obstruct the biodegradability of cellulosic and hemicellulosic fractions and impede their accessibility to microbial attack [ 3 , 17 ]. Hence, a pretreatment step is compulsory before the AD process. The primary objective of pretreatment techniques, whether biological (e.g., fungi, ensiling, enzymes, microbial consortium), physical (extrusion, comminution, microwave, cavitation), chemical (acid, alkali, ozonolysis, ionic liquids, organosolv), or combinatorial (LHW, SE, ammonia fiber explosion), is to alleviate the inherent recalcitrance of LCB by increasing the ASA, disrupting the persistent carbohydrate-lignin shields, disintegrating the lignin sheath, and shrinking the crystalline structure of cellulose [ 18 , 19 ]. Thus, improving biogas rate and productivity. This review article is prompted by the urgent need to address contemporary environmental challenges arising from reliance on fossil-based resources, with a specific focus on the efficacy of AcoD as a potent strategy for sustainable bioenergy production. Emphasizing the critical need to transition toward bioenergy derived from renewable sources, particularly lignocellulosic residues and livestock waste, we underscore their potential not only to reduce dependence on exhaustible fossil fuels but also to contribute significantly to the reduction of carbon footprints and GHG emissions. A pivotal aspect of this review involves the identification and analysis of the 13 most prevalent models that quantify the annual potential of crop residues and livestock waste for bioenergy production. These models are meticulously drawn from various literature-based case studies conducted in different countries. Notably, to the best of our knowledge, this review stands as the first to present these models cohesively and collectively in a single review, providing a distinctive contribution to the field. This approach not only provides invaluable insights for farm owners, stakeholders, and researchers but also establishes a robust foundation for assessing total waste quantities when actual data may be lacking. Furthermore, this manuscript comprehensively explores various facets of AcoD, including the intrinsic structural and compositional features of lignocellulosic residues, the different biotic and abiotic parameters impacting the overall co-digestion performance, and the physico-chemical factors influencing pretreatment efficiency. Additionally, we delve into the latest trends, limitations, and noteworthy research advancements in diverse pretreatment methods applied prior to the AcoD process, contributing to a comprehensive understanding of the complexities involved in advancing sustainable bioenergy production from lignocellulosic residues and livestock waste.",
"discussion": "8 Discussion and future recommendations The integration of livestock waste and crop residues as feedstocks for anaerobic co-digestion holds significant promise in the realms of renewable energy, waste management, and agricultural sustainability, particularly in rural areas. This approach stands out as a technology that is aligned with the principles of the circular economy, converting livestock waste and crop residues from liabilities to valuable assets, with positive implications for both environmental stewardship and agricultural economy. As can be seen from the literature, a considerable amount of research has delved into estimating the annual potential of crop residues and livestock manure generated for bioenergy production, using various theoretical models. This estimation is indeed a critical aspect when assessing the feasibility of implementing AcoD plants. However, theoretical estimations often rely on generalized data that may overlook regional or seasonal variations, leading to inaccuracies in the results (i.e., underestimation or overestimation of the actual potential). Hence, integrating more localized and context-specific data, considering factors such as weather conditions, soil quality, waste management practices, and accounting for local variations, can contribute to more accurate assessments that reflect the actual potential of livestock waste or crop residues in a particular region or farm. The implementation and development of small- and large-scale agricultural plants, while offering significant benefits in terms of waste management and renewable energy production, is subject to various technical, economic, political, and social constraints. On the technical front, the primary challenge lies in the variability of feedstock quality and composition. Animal manure and crop residues can exhibit fluctuations in their organic content, nutrient composition, and moisture levels due to seasonal changes, crop type variations, soil conditions, livestock diet and management, and farming practices, making it challenging to maintain optimal AcoD conditions and achieve a consistent balance for microbial activity. Thus, sophisticated monitoring and control systems are required to optimize operating conditions and prevent process upsets. Furthermore, the chemical properties (pH, moisture level, carbohydrates, toxicity, crude protein, etc.) and physical composition (particle size, density, porosity, etc.) of the input substrates should be thoroughly studied before the AcoD process since they have a significant impact on the overall performance of the conversion process, the quality of the end products (biogas and digestate), and, more importantly, the microbial ecosystems in the bio-digester. Specifically, the analytical characterization of feedstocks prior to the process enables the selection of appropriate co-digestion conjugates and the determination of the optimal mixing ratio of raw materials. The microbiological aspect is also one of the most important factors influencing both digestion stability and efficiency. Although there are several studies in the literature describing the microbial pathways of the co-digestion process, there is still a deficiency in the thorough understanding of the biochemistry and microbial ecology in anaerobic digesters when it comes to the co-digestion of various livestock wastes and lignocellulosic residues. Therefore, more in-depth research should be conducted to develop instructive approaches that may reveal a complete and accurate identification of all microbial species involved in the process, their complex structure, their relationship with feedstock compositions, their metabolic capacity, and their quantitative and qualitative relationships to the functional performance of the digester. Moreover, it is instrumental to deploy innovative technologies (artificiel intelligence, kinetic modeling, machine learning) and robotic tools (device mobiles, smart cameras) that can predict the dynamical behavior of substrates during conversion and monitor fluctuations in the microbial environment and their impact on other operational parameters. The inclusion of a pretreatment stage is beneficial for improving co-digestion efficiency and optimizing biogas yield by breaking down the recalcitrant structure of the feedstock and rendering polysaccharides more accessible to enzymes and microorganisms. Physical pretreatments are widely applied on an industrial scale and pose no inherent risk of forming inhibitory compounds. However, they are high-severity methods and involve high power consumption and maintenance costs, making the entire process expensive. Chemical pretreatment processes reduce cellulose crystallinity and moderately boost biogas production, but they require expensive reagents, generate caustic intermediary products, and necessitate large quantities of water for washing the substrates before introducing them into the reactor. On the other hand, biological pretreatment is one of the most promising technologies for enhancing biogas production efficiency, as it is environmentally friendly, requires milder reaction conditions, and produces fewer side-stream products. However, its application to animal manure is underexplored in the literature. Compared to single treatments, multi-stage (combined) pretreatment techniques significantly increase biomethane production, decrease pretreatment severity, and reduce the formation of inhibitory compounds. Table 3 recapitulates the advantages and drawbacks of the aforementioned pretreatments. Nevertheless, comparing pretreatment technologies with each other proves challenging due to the non-standardized conditions such as the substrate's composition and structure, pretreatment operational conditions, and AcoD process types. The effect of pretreatment on both the innate composition of the feedstock and methane yield improvement has been extensively investigated in the literature at bench-scale utilizing BMP assays ( Table 4 ) . While this methodology proves effective in determining optimal pretreatment conditions and assessing substrate degradation rates, as well as the ultimate methane yield, it remains challenging to extrapolate laboratory test findings and improvements to industrial-scale, continuously loaded AcoD systems. Hence, the implementation of specific pretreatment techniques for agricultural residues at the industrial level often constitutes a significant impediment [ 227 ]. Moreover, for both energetic viability and economic profitability, the input for pretreatments (in terms of extra energy and cost requirements) must be significantly lower than the output energy (biogas, heat) and economic gain (represented by methane yield increase) [ 40 , 228 ]. The effectiveness and selection of the appropriate pretreatment method for lignocellulosic biomass or livestock residues are influenced by several crucial parameters, including the physico-chemical properties of the substrate, pretreatment complexity, pre-treatment operating conditions, the formation of inhibitory products, economic and energy costs, environmental impacts, cost considerations (cost of chemicals, thermal/electrical energy input, cost of biological agents), and the methane improvement achieved [ 229 ]. Therefore, an extended and interdisciplinary investigation into techno-economic and life cycle assessments, energy and exergy analysis, and exergo-environmental evaluation is necessary to scrutinize the environmental sustainability, economic and energy feasibility of pretreatment methods on an industrial scale. This will help bridge the gap between laboratory findings and the real-world application of pretreatment technologies. Table 3 Pros and cons of pretreatment techniques [ 36 , 70 , 220 , 230 , 231 ]. Table 3 Pretreatment method Pros Cons Biological • Eco-friendly and sustainable process •No release of inhibitory compounds due to mild operation conditions (atmospheric pressure, low temperatures) •Low capital and operating cost requirements •Less energy and chemical requirement •Degradation of both hemicellulose and lignin •Long incubation time •Slow rate of delignification •It requires careful control operation conditions •Lower reaction rates •large space requirement Physical •Low environmental impacts •Increase of the ASA •Easy handling especially for mechanical pretreatment •No need for chemical catalysts •High power, energy and water expenditure •High maintenance cost Chemical •Effective solubilization due to reduced cellulose crystallinity and DP •It increases accessibility to cellulose •faster rates and better efficiencies •High amount and cost of reagents •It mandates expensive corrosion-resistant materials due to caustic properties of chemicals especially acids •Requirement of neutralization and detoxification steps •Release of fermentation inhibitors during the process Table 4 Effects of biological, chemical and physical pretreatments on biogas yield and composition of crop residues and animal waste. Table 4 Pretreatment methods Feedstock Process conditions Outcomes References Pretreatment AD Fungi Wheat straw Ligninolytic fungi, batch, 28 °C for 7 days Thermophilic condition (50 °C), 6 weeks •48.2% decrease of lignin content •enhancement of biogas yield 5 times compared with untreated WS •407.1% increase in methane yield [ 232 ] Bean straw WRF Pleurotus ostreatus, 30 °C (1, 10 and 30 mg fungus/g straw) for 14, 21 and 28 days Batch, 30 °C •Maximum lignin (18%) and hemicellulose (44%) degradation at 30 mg fungus/g straw •Highest total methane yield (38 CH4/g VS loaded) [ 168 ] Rice straw WRF Pleurotus ostreatus (PO), Phanerochaete chrysosporium (PC), Ganoderma lucidum (GL), 30 °C, 5 weeks – •2.22-fold increase in methane yield with (PC) pretreated RS. •(GL) and (PO) resulted in 1.88-fold, 1.64-fold increase in methane yield, respectively. [ 159 ] Corn stover Phanerochaete chrysosporium, SSF at 28 °C, 30 days Batch, 37 °C, 30 days •Highest methane yield (265 mL/g VS) compared to untreated (215.5 mL/g VS); 49.5 mL/g VS increased biomethane [ 233 ] Dairy cattle manure Pleurotus ostreatus, 28 °C, 14 days 37 °C 7% increase in methane production [ 234 ] Pleurotus ostreatus, 2–17 °C min and 10–31 °C max, 2 months 111% increase in methane production Ensiling Mixture of wheat straw + Sugar beet leaves (after chopping) Lab-scale: storage in vacuum bags in a barn (5–15 °C) for two months and subsequently at room temperature (20 ± 0.5 °C) for 7 months. Batch, 37 °C, 61 days •BMP increase ranged from 19 to 34% [ 171 ] Pilot-scale: storage in silos with volume of 2.6 m 3 for approx. 6 months (177–189 days) Batch, 38–39 °C, 58 days •BMP increase ranged from 18 to 32% Sugarcane stalks (SCS) Storage in vacuumed and double sealed silos in the dark under ambient temperature of 20–25 °C for 70 days 38 °C, 30 days •Increase in methane potential by 24.0% (without additives) [ 179 ] Sugarcane trash (SCT) •Increase in methane potential by 23.4% (without additives) •Addition of sugarcane molasses + commercial silage inoculant resulted in 51.4% higher methane potential than ensiled SCT (without additives) Mixture of fresh cattle manure + wheat straw 3.5 L airtight round plastic storage drums, 25 ± 2 °C for 120 days Batch, 35 °C •Co-ensiling led to 67% methane potential losses (without additives) •Limitation of energy losses to 25% after formic acid addition •Full preservation of methane potential after glucose addition [ 235 ] Enzymatic Chicken manure Mixture of commercial enzymes (Onozuka R-10 and Macerozyme R-10), 40 °C, 24h Batch, 37 °C for 21 days •35% increase in biogas production compared to the control without enzymatic pre-treatment [ 236 ] Corn stover Laccase (LA) and peroxidases (manganese peroxidase + versatile peroxidase), 30 °C for 0, 6, 12 and 24 h – •25% increase in biomethane production after 24 h (with laccase) •17% increase in biomethane production after 6h (with peroxidases) •Treatment with both enzyme groups increased biomethane production with 16% and 14% after respectively 6 and 24 h of treatment [ 237 ] Microbial consortia Rice straw (RS) + Pig manure (PM) Cellulolytic microflora (Clostridium, Petrobacter, Defluviitalea, and Paenibacillus), 55 °C, 30 h 35 °C, 15–20 days •62.20, 59.58, and 33.77% decrease in the content of cellulose, hemicellulose, and lignin, respectively •45% icrease in the cumulative methane production of RS and PM (342.35 ml (g-VS) −1 ) compared to untreated (236.03 ml·(g-VS) −1 ) [ 187 ] Corn straw (CS) Mixed microbes: Phanerochaete chrysosporium, Coriolus versicolor, Trichoderma viride, Aspergillus niger, Gloeophyllum trabeum, Bacillus circulans, Pseudomonas aeruginosa and Streptomyces badius; 30 °C for 14 days 35 °C, 30 days •131.6% increase in methane yield [ 238 ] Mechanical pretrearment giant reed stems (Arundo Donax) Two stages (hammer mill + pin mill) dry milling device with working capacity up to 1,2 t h −1 38 °C; 28 days 137.7% gain in the cumulative methane production [ 239 ] Wheat straw 49.1% gain in the cumulative methane production Wheat straw Knife mill; particle size reduction to 2 mm 40 °C, 60 days 83.5% improvement in methane yield [ 240 ] Barley straw Knife mill; particle size reduction to 5 mm 54.2% improvement in methane yield Horse manure Prototype ball mill, rotational speeds (6,10,14 rpm) 37 °C, 35 days Increase in specific methane yield by 37.3 % [ 241 ] Cattle manure Mobile and fixed hammer mills; wet sieving 35 °C Increase in methane production rate by 15% and 27% for mobile hammer mill and fixed hammer mill, respectively [ 242 ] Steam explosion Rice straw 200 °C; 120 s 38 °C; 21 days • 51 % increase in biogas production • 13.72 % and 16.79% increases in degradation rates of cellulose and hemicellulose, respectively, as compared to untreated straw [ 243 ] Corn stover 160 °C; 2 min mesophilic conditions (37.5 °C); 49 days 22 % improvement in methane yield [ 17 ] Pig manure 170 °C; 30 min 35.1 °C 206.9 % improvement in methane yield [ 244 ] Liquid hot water Wheat straw 175 °C for 30 min; pressure: 0.4 ∼ 2.5 MPa Mesophilic conditions 36 °C 62.9 % increase in methane yield [ 245 ] Extrusion Rice straw twin-screw extruder (55 kW, 120 rpm) mesophilic temperature 37 °C; 60–90 days 72.2 % increase in methane yield [ 246 ] Acid pretreatment Wheat straw Dilute H 2 SO 4 (1%), 121 °C, 10–120 min mesophilic, 37 °C, 30 days 16% increase in methane yield [ 247 ] Corn straw Dilute H 2 SO 4 , HCl, CH 3 COOH and H 2 O 2 (1,2,3,4%), 25 °C, 7 days mesophilic, 37 °C, 35 days 115.4% increase in methane yield with 3% H 2 O 2 [ 248 ] Cow manure Peracetic acid (0.01–0.10 g/g VS), 6 and 12h 38 °C, 45 days 39.1% increase in biogas production [ 249 ] Dairy cow manure HCl (2%), 37 °C, 12h 37 °C Increase in methane potential by 20.6% [ 250 ] Alkali pretreatment Wheat straw Ammonia (2,4,6%), 35 °C, 7 days 35 °C, 60 days 52% increase in methane yield [ 251 ] Rice straw Ammonia (2,4,6%), 30 °C, 7 days 35 °C, 55 days 28.55% increase in methane yield at 4% ammonia concentration [ 252 ] Cow manure Calcium oxide (0.05–0.15 g gTS-1); 6–12h 38 °C 26 % increase in biogas production [ 249 ] Dairy cow manure NaOH (10%), 100 °C, 5min 37 °C Increase in methane potential by 23.6% [ 250 ] Chicken litter NaOH (5%), 120 °C, 90 min 37 °C Up to 50% improvement in biogas production [ 253 ] From an economic standpoint, it is necessary to consider the collection sites for residual biomass, the transportation to the processing plant, and the subsequent storage. Furthermore, upfront capital expenditure for digester infrastructure and ongoing operational costs, including maintenance, raw material procurement, and monitoring, contribute to the economic challenges. The economic feasibility of AcoD facilities is also closely tied to governmental policies, incentives, and subsidies. Government policies play a crucial role in shaping the regulatory framework, incentives, and support mechanisms that can either promote or hinder the implementation of AcoD technology involving crop residues and livestock waste by agricultural enterprises and waste management facilities. Governments can establish grant programs, subsidies, feed-in tariffs, and tax credits for entities investing in the waste management and conversion sectors. These financial incentives stimulate the uptake of sustainable practices, bolster the economic feasibility of the biogas plant, offset the upfront capital costs and operational expenses associated with setting up the co-digestion facility, and encourage farm owners to invest, particularly in remote areas where animal waste and crop residues are generated in substantial quantities. China, Denmark, and Italy promote the production of biogas and biomethane from agricultural waste. In the United Kingdom, AD facilities primarily employ municipal biowaste, sewage sludge, and wastewater as main feedstocks. This is attributed to regulations that restrict the utilization of energy crops as raw materials to a maximum of 50% [ 254 ]. To achieve lasting change, awareness campaigns and educational programs are also essential. Local farmers must be apprised about the environmental and health implications of inadequate waste management and empowered with knowledge about sustainable alternatives. This can instill a sense of responsibility and ownership in adopting more eco-friendly practices."
} | 6,726 |
25267793 | PMC4178370 | pmc | 8,385 | {
"abstract": "Metabolic networks have become one of the centers of attention in life sciences research with the advancements in the metabolomics field. A vast array of studies analyzes metabolites and their interrelations to seek explanations for various biological questions, and numerous genome-scale metabolic networks have been assembled to serve for this purpose. The increasing focus on this topic comes with the need for software systems that store, query, browse, analyze and visualize metabolic networks. PathCase Metabolomics Analysis Workbench (PathCaseMAW) is built, released and runs on a manually created generic mammalian metabolic network. The PathCaseMAW system provides a database-enabled framework and Web-based computational tools for browsing, querying, analyzing and visualizing stored metabolic networks. PathCaseMAW editor, with its user-friendly interface, can be used to create a new metabolic network and/or update an existing metabolic network. The network can also be created from an existing genome-scale reconstructed network using the PathCaseMAW SBML parser. The metabolic network can be accessed through a Web interface or an iPad application. For metabolomics analysis, steady-state metabolic network dynamics analysis (SMDA) algorithm is implemented and integrated with the system. SMDA tool is accessible through both the Web-based interface and the iPad application for metabolomics analysis based on a metabolic profile. PathCaseMAW is a comprehensive system with various data input and data access subsystems. It is easy to work with by design, and is a promising tool for metabolomics research and for educational purposes. Database URL : http://nashua.case.edu/PathwaysMAW/Web",
"conclusion": "Conclusions and future work PathCase MAW is an online multi-tool system to effectively create, store, browse, query, visualize and analyze metabolic networks. The system provides an SBML parser to create networks from a SBML document, and a user-friendly interface that lets users to update the database. Stored information can be accessed online either through a Web interface or an iPad application. PathCase MAW is a useful system to access and user-created reconstructed networks online for researchers. An analysis made online on the SMDA tool has been published ( 11 ) as a proof of concept. As future work, our goal is to link the data stored in the database to outside sources such as KEGG for reaction references and HMDB for metabolite references, whenever the information is provided in the input model. Next, we plan to have a collective site that includes all of the released models for the use of research community.",
"introduction": "Introduction Metabolomics is a relatively new ‘omics’ platform in life sciences research. The advancements in analytical methodology and high-throughput rates have led to the collection of large metabolic data sets. Metabolic profiles and genome-scale metabolic networks ( 1 ) are used in various contexts, such as (i) predicting flux distribution for the metabolic activity over the network [metabolic control analysis (MCA) ( 2 ), flux balance analysis (FBA) ( 3 ) and constraint-based methods] and (ii) drug discovery and disease research ( 4–7 ). The increase in the number and importance of metabolic networks has come with the need for carefully designed databases to store/organize metabolic networks, and efficient online tools to browse/ analyze/visualize metabolic data. The goal of PathCase MAW (Metabolic Analysis Workbench) is to provide a metabolic network database and a Web- or tablet-based system that enables users to interact with the underlying metabolic network. PathCase MAW provides the following functionalities:\n A metabolic network database that captures the metabolic network with a compartment hierarchy and metabolic regulation relationships. A Web site that (i) enables users to browse pathways, reactions, metabolites/metabolite pools and compartments stored in the database, (ii) provides several built-in queries and interactive visualization and (iii) has the integrated steady-state metabolic dynamics analysis (SMDA) tool. SMDA tool takes a set of metabolite measurements and a metabolic subnetwork stored in the PathCase MAW database as input. Then, it produces all possible steady-state flow scenarios (called flow graphs) for the selected subnetwork as output (that are consistent with the observed metabolite measurements and the underlying biochemistry) (available at http://nashua.case.edu/PathwaysMAW/web/ ). An iPad application that has all capabilities of the Web-based PathCase MAW system with the exception of browsing/querying (available at Apple AppStore). An offline metabolic network editor with visualization capabilities that enables users to create their own network in a user-friendly way. An SBML Parser to parse and store genome-scale reconstructed metabolic networks [e.g. Recon 1 of humans ( 8 )] into the PathCase MAW database. Currently, the PathCase MAW system works on a manually created (and generic) mammalian metabolic network, which is obtained from the metabolic atlas by Selway et al. ( 9 ). We also have three genome-scale reconstructed networks hosted and available on the sister PathCase RCMN (PathCase Reconstructed Metabolic Networks) Web site ( 10 ). Source codes of the Web interface, PathCase MAW editor, SBML Parser, as well as the database schema are available on request for academic users to create their own networks and to host/access them. User-created networks can also be hosted on the PathCase RCMN Web site on request.",
"discussion": "Results and discussion All presented subsystems of PathCase MAW have been implemented and released to the research and community. Currently, the system runs on a manually created generic mammalian database that consists of 26 pathways, 282 reactions and 243 metabolites. We have also parsed and released three genome-scale reconstructed metabolic networks (two for Mus musculus and one for T rypanosoma \n c ruzi ) in the sister PathCase RCMN site. Next, we provide a comparison of the existing systems with PathCase MAW and then the challenges/shortcomings of the current system. Comparison of PathCase MAW with existing systems KEGG ( 33–36 ) has been a major source of metabolic pathways, which provides an application programming interface and data download options. Unlike PathCase MAW , KEGG (i) provides limited visualization and browsing capabilities, (ii) does not capture compartment information for metabolites, (iii) provides limited functionality over the data and (iv) ignores metabolic regulation, such as covalent activation/inhibition or metabolite ratios (e.g. high NAD/NADH ratio activates alpha-ketoglutarate dehydrogenase). PathCase SB and PathCase KEGG ( 37 ) have been released to provide additional functionalities such as querying and interactively visualizing pathway data. PathCase KEGG hosts KEGG data in its own schema and provides similar browsing and visualizing capabilities as PathCase MAW system. PathCase SB stores kinetic models of pathways provided by Biomodels Database ( 31 ). It integrates KEGG data with the stored models, and enables users to simulate and compose models. However, neither of these systems have metabolomics analysis goals or tools. They only work on data provided by well-known third parties (KEGG and Biomodels Database). Conversely, our goal is to provide a system that enables users to work on their own data to use the functionality provided. Finally, among the three above-mentioned systems, only PathCase KEGG has a mobile interface ( 38 ) (iPad application). BioCyc ( 39 ) is another major pathway/genome database, which hosts numerous different organism databases at various curation levels. They provide tools like PathoLogic ( 40 ) to predict and reconstruct metabolic networks directly from the corresponding genomes, and SRI’ss pathway tools to adapt and curate networks. For an organism, the database contains the following information: genome, gene products, metabolic network, regulatory network and the transporter complement of the organism. Although the data are comprehensive (e.g. PathCase MAW is only dealing with metabolic networks), browsing and visualization capabilities are limited on their Web site. BioCyc provides a SVG-based, static visualization of the complete metabolic network, and links to related sources per item, but there is no other interaction, unlike the PathCase MAW visualization interface. ‘Pathway Tools’ ( 41 ) provide such features; however, it is an offline system and can be compared with the PathCase MAW editor. Computational tools provided by pathway tools focus on prediction of metabolic features, such as pathways, choke points or operons for metabolic networks. On the other hand, SMDA tool is concerned with analyzing metabolic profiles and predicting metabolic activity on metabolic networks. They also provide a FBA tool, but only in the desktop version. BioCyc does have an iPhone application for only EcoCyc ( 42 ), and there is no application for iPad. In short, BioCyc metabolic network creation from the genome is a big (and, as they also state, difficult) and different task than PathCase MAW ’s task, which deals with already created networks such as those from already reconstructed genome-scale networks. PathCase MAW clearly provides a much smaller subset of the information provided by BioCyc; however, for metabolic networks, it provides a user-friendly, manageable and online system with nice analysis, browsing and visualizing capabilities. Reactome ( 43–46 ) is a pathway database curated and updated by life scientists. Although the visualization is static (e.g. no moving around of items), it provides beautiful images of selected pathways with zoom in/out capabilities and links to related resources (based on Systems Biology Graphical Notation). Reactome provides tools such as pathway analysis (e.g. overrepresentation analysis, comparing pathways of species). They also provide a built-in query interface similar to PathCase MAW . There is no mobile interface provided. Reactome’s goal is to have an all-inclusive one-stop-shop type of curated metabolic pathway database, whereas PathCase MAW ’s goal is to provide tools and schema for researchers to move their own data to PathCase MAW schema to have the access, query, browsing and analysis interfaces available. MEMOSys ( 47 ) is a system that provides a platform for researchers to collaborate and create reconstructed metabolic networks. The system has a version control mechanism to keep track of the history of the models. The networks that are already created are stored in a repository. They can be browsed, compared and exported via the user-friendly Web interface. They provide PDF-based maps for some models. However, the system does not offer metabolomics analysis tools, query interface, mobile interface or interactive visualizations. In summary, PathCase MAW is a unique system with its goal and capabilities. It has similarities and differences, compared with the existing systems. However, it is not a stand-alone advancement over an existing system. Challenges and shortcomings One of the biggest challenges of the PathCase MAW system is with parsing reconstructed metabolic networks. And, the reason is the misuse of the SBML format for creating networks by researchers. For instance, Edinburgh metabolic network for humans ( 48 ) does not specify ‘compartment’ attribute for species, as they do not distinguish between compartments, and assume all information is in the same single compartment, ‘cell’. Similarly, many models include information (e.g. associated pathway/subsystem for a reaction) in the tag, which is simply unstructured plain text from the parser’s point of view, and creates parsing problems for all SBML parsers. These challenges are explained in detail in Alshalwi, 2011 ( 49 ). The second challenge is with the exponential complexity of the SMDA Tool, which is also explained in detail in Cakmak et al . , 2012 ( 11 ). For networks with high numbers of reactions and, in those cases, with a small number of metabolite observations, the SMDA system may have an unacceptably long response time, and may even run out of memory. That said, note that this is a problem shared by all constraint-based methods ( 16 ) with similar goals and assumptions."
} | 3,087 |
33674757 | PMC7935979 | pmc | 8,386 | {
"abstract": "As an active interface between the host and their diet, the gut microbiota influences host metabolic adaptation; however, the contributions of fungi have been overlooked. Here, we investigate whether variations in gut mycobiome abundance and composition correlate with key features of host metabolism. We obtained animals from four commercial sources in parallel to test if differing starting mycobiomes can shape host adaptation in response to processed diets. We show that the gut mycobiome of healthy mice is shaped by the environment, including diet, and significantly correlates with metabolic outcomes. We demonstrate that exposure to processed diet leads to persistent differences in fungal communities that significantly associate with differential deposition of body mass in male mice compared to mice fed standardized diet. Fat deposition in the liver, transcriptional adaptation of metabolically active tissues and serum metabolic biomarker levels are linked with alterations in fungal community diversity and composition. Specifically, variation in fungi from the genera Thermomyces and Saccharomyces most strongly associate with metabolic disturbance and weight gain. These data suggest that host–microbe metabolic interactions may be influenced by variability in the mycobiome. This work highlights the potential significance of the gut mycobiome in health and has implications for human and experimental metabolic studies.",
"introduction": "Introduction The modern diet is dominated by processed sugar and carbohydrates that are linked to metabolic and immune-mediated diseases 1 . Disruption of the gut microbiome also influences the development of metabolic disease 2 , 3 , and dietary composition is a key driver of gut microbial community structure and function 1 . Gut microbes form an interface between the diet and the host, participate in digestion and energy extraction from otherwise indigestible fiber and oligosaccharides 4 , produce short-chain fatty acids and novel metabolites, and ultimately shape host endocrine and immune signaling. Given these findings, robust microbial communities appear crucial component to metabolic homeostasis as supported by data across diverse species, life stages, and disease states 2 , 5 . While the gut microbiome is often equated to bacteria, microbial communities contain diverse populations of archaea, viruses, protists, and fungi 6 . Collectively termed the mycobiome, gut fungal communities of molds and yeasts are crucial to maintaining gut homeostasis and systemic immunity 7 . Mounting evidence suggests other domains of life, such as viruses, can also influence host metabolic tone 8 . However, data describing the role of the mycobiome in host metabolism remain scarce 9 , 10 . Recent studies in humans and mice indicate commensal fungi have the potential to influence host metabolism directly 7 , 11 – 14 and via alterations to bacterial community composition 15 , 16 . The latter interactions between bacteria and fungi are likely to yield greater insight into the complex microbial gut ecology and health. The role of the gut mycobiome in host metabolism remains unresolved. This question is further obscured by the difficulty in discriminating environmental and transiently ingested fungi from true colonizers, especially in free-living and genetically heterogeneous human populations. Thus, we designed a study using specific pathogen-free mice with the same genetic background so that we could control for age, sex, and previous founder exposures. We also sourced these mice from four different vendors to vary their initial gut mycobiomes. Utilizing this design, we tested our hypothesis that the gut mycobiome would associate with host metabolic response to a processed diet, which is representative of typical westernized diets. First, we found the gut mycobiome of laboratory mice differs dramatically between animal vendors. Using a combinatory approach, we then identified fungal populations that shifted in response to processed diet and identified key fungal taxa that may be linked to host metabolic alterations. We show baseline correlations in the gut mycobiome strongly associate with changes in host adiposity and serum metabolic biomarkers in response to a highly processed low-fat diet. Our results support the role of the gut mycobiome in host metabolic adaptation and have important implications regarding the design of microbiome studies and the reproducibility of experimental studies of host metabolism.",
"discussion": "Discussion Here, we investigated whether variations in gut mycobiome relative abundance and composition correlate with key features of host metabolism. This question drives towards understanding the complex interkingdom interactions between bacteria and fungi and how they are both collectively shaped and potentially contribute to host homeostasis. To address this question, we asked if differing starting mycobiomes could affect host adaptation to a standardized diet and an ultra-processed diet rich in purified carbohydrates in a manner that associated with deleterious metabolic outcomes. Our key findings are that the gut mycobiome of healthy mice is shaped by the environment, including diet, and significantly correlates with metabolic changes in the host. For instance, increased triglyceride concentrations and metabolic biomarkers, including the deposition of hepatic lipids, correlate with increased abundance of the fungal genera Thermomyces and decreased Saccharomyces . Our results highlight the potential importance of the gut mycobiome in health and have implications for human and experimental metabolic studies. The bacterial microbiome is strongly influenced by dietary exposure 26 and influences host metabolism 3 . Emerging work demonstrates dietary exposure also shapes fungal communities 9 . However, despite evidence for fungal pattern recognition receptors in the human gut 7 and fungal influences on the disease in the human gut 27 , 28 , continuous gut colonization by fungi remains controversial in humans 29 . Convincing evidence indicates fungi colonize the mouse gut and influence host physiology 30 and disease 31 . Diet may have a dominant effect over host genotype on the composition of the gut bacteriome with subsequent alterations in host-microbe interactions 32 . We observed a similar strong effect of diet on the composition of both fungal and interkingdom community composition. However, because we used mice with the same genetic background, we were able to examine the impact of the founding microbial communities largely independent of host genomic influences. In our study, both prolonged exposure to a processed diet and length of isolation in a specific pathogen-free environment reduced fungal diversity. In turn, reduced fungal diversity was associated with increased adiposity and physiologic alterations seen in the metabolic syndrome. In our controlled study, we found key differences in baseline mycobiome composition reflect differences in metabolic tone in response to diet. For metabolic studies in mice, the choice of vendor, shipment, diet, and housing may have instrumental roles in shaping outcomes, which should encourage further caution in drawing causative relationships. Validating outcomes in mice from several vendors or across multiple shipments may be necessary to address this potential confounder. The implication for human microbiome studies, which often examine only bacteria and sample only fecal communities, is that the mycobiome may have unappreciated effects on microbiome-associated outcomes. Exposure to a high-fat diet may alter fungal and interkingdom community composition 9 . Our work suggests complex alterations in co-abundance networks are associated with diet. Fungal cell wall components are a major point of interaction between fungi and bacteria in the environment 7 . For example, the abundance of a major gut bacteria, Bacteroides thetaiotaomicron , can be influenced by the presence of mannan in fungal cell walls 16 . Similarly, fungal chitin influences the composition of anaerobic bacteria 33 . Our co-occurrence correlation analysis suggests the loss of key fungi during dietary exposure was closely related to differences in bacterial community composition, which may represent niche replacement in the face of a shifting dietary environment or disruption of interkingdom metabolomic networks. This study extends the previous work 9 in two important avenues. First, we examined the impact of a highly-processed diet on gut communities. This approach allowed us to identify more subtle physiologic changes in host metabolic tone. Second, by using mice from different vendors, we extended these observations by observing the effect of different founding mycobiomes on host metabolism. Among the myriad metabolites and effectors that likely connect gut microbial communities to metabolism, the gut bacteriome may influence host metabolism through several major mechanisms 34 . While classically attributed to bacteria, fungi have an often underappreciated role in the production of metabolites and may interact with host physiology via analogous mechanisms. One of the best-known fungal pattern recognition receptors in humans is Dectin-1 ( CLEC7A ), which recognizes fungal β-glucan. In addition to immune cells 7 , adipose tissue expresses Dectin-1 35 . In humans, obesity is associated with increased Dectin-1 expression in adipose tissue 35 . In mice, Dectin-1 ( Clec7a ) has a MyD88-independent role in diet-induced obesity 35 . In MyD88-deficient adipose tissue, Dectin-1 was upregulated in both adipocytes and adipose-associated macrophages. Furthermore, blockade of Dectin-1 led to improved glucose sensitivity and decreased numbers of CD11c + macrophages. The reverse was also true in Dectin-1 activation 35 . We also observed differences in Clec7a expression in the liver that correlated with differences in adiposity. This may suggest that mycobiome-driven antigen-presenting cell education in the gut could directly influence systemic metabolic tone. Another important function of gut microbes, often attributed solely to bacteria, is the production of secondary bile acids (BAs). Our understanding role of BAs has evolved from simple detergents to hormones intimately related to multiple metabolic processes, revealing important roles of BAs in dyslipidemia and type 2 diabetes 36 . BAs are a significant source of host–microbiome interaction via cellular receptors such as TGR5 ( Gpbar1 ) 37 and FXR ( Nrlh4 ) 36 . Therefore, BA composition is interrelated with metabolic hormones such as leptin 38 , 39 , resistin 40 , ghrelin, GLP-1 (glucose-like peptide-1), and peptide YY 41 . However, fungi also produce BAs and likely participate in the gut BA pool to an underappreciated extent 42 . Fusarium , which we found was key to the structure of processed diet-induced weight gain-resistant TAC mice, is a prolific metabolizer of deoxycholic acid. Other key fungi, such as Aspergillus and Penicillium , can also produce secondary BAs 42 . BAs also influence the stability of other fungal metabolites. For example, luminal BAs 43 influence the stability of a prominent lipase in Thermomyces , which we observed as the taxa most significantly associated with weight gain on a processed diet. Intriguingly, in our model, these differences correlated with differences in metabolic hormones, including leptin, resistin, and ghrelin. These findings may suggest a potential role for fungi in host metabolic processes via BA signaling. In summary, these data indicate the gut mycobiome in healthy mice is highly variable and responds to disturbances such as changes in the environment and diet. Despite these ecological pressures, resilient differences in gut mycobiome composition in healthy mice strongly associated with differences in host metabolic tone, including differential fat deposition, metabolic biomarkers, and gene expression in metabolic tissues. We also highlighted two fungal-derived products with plausible effects on host metabolism. While there are potentially thousands of metabolically active fungal products, our work argues for future work in defined gnotobiotic animal models to further elucidate the mechanisms of mycobiome-host interaction. Finally, our findings suggest that differences in the gut mycobiome may be an underappreciated source of variability in health and metabolic outcomes in response to dietary interventions."
} | 3,115 |
31458242 | PMC6643867 | pmc | 8,388 | {
"abstract": "The development of\nartificial nanosystems that mimic directional\nwater-collecting ability of evolved biological surfaces is eagerly\nawaited. Here we report a new type of addressable water collection\nthat is induced by coupling both vapor gradients, like a road drawn,\nand the temperature-tuned condensation in nanopores as step signals.\nWhat distinguishes the motion described here from the motions reported\nearlier is the fact that neither bulk liquid infiltration nor displacement\nof droplet is required. Instead, the motion results from a scanned\nwater capture because of the temperature-dependent condensation command\nacting on the vapor pressure gradient track originated by a droplet\nwithout a bulk fluidic connection with a mesoporous film. This novel\nworking principle demands only a small-range surface temperature control,\nwhich was entirely generated by a thermoelectric cell integrated to\nthe mesoporous substrates. The strategy opens the route to achieving\nprecise control over wetting location (from a few to hundreds of micrometers)\nand hence over the direction of water collected by these widely employed\nnanomaterials. Furthermore, as water is collected from condensation\ninto the pores, the system naturally involves purification and subsequent\ndelivery of clean water, which provides an added value to the proposed\nstrategy.",
"conclusion": "Conclusions The development of systems that operate on “open\npit”\nscenarios to provide addressable water collection, which is intrinsic\nto evolved natural systems, is becoming a key requirement for versatile\ndevices, presenting a significant scientific and engineering challenge.\nWhereas studies of water collection into pore media in the past used\npressure chambers at a given temperature, we have here explored the\nsurface temperature–actuated condensation in a natural vapor\ngradient guide generated by water droplets under room conditions as\nthe driving forces. Water-collecting directionality in the mesoporous\nfilms is the result of the condensation features tuned by surface\ncooling on drop vapor environment at ambient conditions. We have demonstrated\nthat mesoporous films not only yield water condensation but also enable\na directional water collection from the humidity of small droplets\nthrough relatively small MTF temperature variations. The temperature\ncontrol has been established by voltage inputs, which offer additional\nopportunities for the development of microdevices with integrated\nsmart water collection capabilities. Unlike the nanofluidic platforms\nbased on capillary infiltration reported at present, in this noncontact\nmode the motion of the wetting front is fully driven by water uptake\nfrom the vapor phase. This strategy takes advantage of coupling both\nthe vapor gradient around sessile drops, like road drawn, and the\ntemperature-dependent condensation pressure in nanopores as step signals.\nWe anticipate that this route can be extended readily to other nanoporous\nstructures (nanostructured porous silicon, 23 metal–organic frameworks, 24 and\nso forth) to guide their water uptake. This new working principle\nopens the way for nanopore networks to be used as an accurately controllable\nwetting location versatile system and hence to direct the water collection.\nWe therefore envision that this study will aid the development of\nnovel addressable systems for water collection and purification. Owing\nto its high modularity and scalability, this self-purifying water\nsupply may find applications in advanced microdevices.",
"introduction": "Introduction Notable biological\nsurfaces on the micro- and nanometer scale,\nin both plant and animal kingdoms, display special features that control\ntheir interaction with water from humid air. 1 A fascinating example is provided by certain beetles living in the\ndesert that can capture water from the fog on their backs. 2 Cactus species can also harvest water from humid\nair using exquisite multilevel surface structures. 3 In addition to the adaptive characteristics that facilitate\nefficient fog capture, these adapted species present directional water\ncollection as a notable ability of its biological structure. Mesoporous\nmaterials, which are composed of humidity-sensitive nanopores, enjoy\na reputation as surfaces that collect water from humid air. 4 − 6 Another interesting but less studied feature is the nanofluidic\nability of mesoporous thin films (MTFs) to transport water by a bulk\nfluidic connection from drops deposited over its surface. 7 , 8 Liquid from these sessile droplets infiltrates the porous matrix\nvia capillary imbibition and builds a defined wetted region with a\nfluid front that advances in the mesoporous film. 7 For instance, nanofluidic platforms based on MTF have been\nused to implement localized chemical reactions, 9 for nanoflow manipulation, 10 as a fluid sensor, 11 and, notably, as\nan electrical current nanogenerator. 12 Consider now a vapor gradient that influences a nanopore media.\nThe resulting variations in relative humidity create a pathway for\nthe directional collection of water. It is reasonable to expect that\ndecreasing the temperature of the MTF can cause water capture along\nthe pressure gradient. This scanned condensation of the water would\nbe an innovative method for the creation of novel systems able to\nhandle water uptake on the microscale. As proof of this concept, this\nwork demonstrates that decreasing the temperature of the condensing\nsurface drives a mobile wetting front onto the MTF guided from the\nvapor gradient associated with an adjacent water drop, which is not\nin direct contact with the patterned MTF. Given that water uptake\nproduces a directional expansion of the wetted area, MTF exhibits\nan addressable water-collecting capability that is comparable to that\nfound in evolved biological systems. A remarkable plus of the strategy\nis that it works under room conditions, which offers advantages in\nthe development of practical devices. We further demonstrate the possibility\nto obtain and deliver purified water to micro-size nano-confined domains\nin a well-controlled fashion.",
"discussion": "Results and Discussion Thick (≈130\nnm) mesoporous silica thin films (≈10\nnm pore size) prepared as reported previously 7 , 9 were\npatterned in defined millimeter-size shapes, using conventional photolithography\ntechniques. 13 Details are given in the Experimental Section and the process is schematically\nshown in Figure S1 ( Supporting Information ). The scanning electronic microscopy (SEM) images of the patterned\nmesoporous films elucidate its structure ( Figure 1 ). The system was operated in the lab atmosphere,\nunder controlled conditions of temperature and relative humidity:\n25 °C and 46%, respectively. The water adsorption isotherm shown\nin Figure S2 ( Supporting Information ) reveals\nthat in these ambient conditions the pores remain principally empty.\nThe silicon substrate was then thermostated with a Peltier cell to\nhomogeneously tune the temperature on the mesoporous film when a voltage\nis applied. When a water drop (2 μL) was placed around 50 μm\nbeside the patterned MTF onto the silica substrate, a localized condensation\nregion develops on the MTF surface. The wet area can be optically\nobserved because water uptake produces a refractive index contrast\nin relation to the dry zone. Figure 1 Microcopy images of the patterned MTF. (A) Low-magnification\nSEM\nimage taken from as-deposited films on silicon after photolithography\npatterning. (B) Magnified image that shows mesopores. (C) Cross-sectional\nview of the MTF. Figure 2 presents\na schematic representation of the mechanisms controlling water collection\nthrough the mesoporous matrix. Water evaporation from the droplet\nsurface creates a microclimate of increased vapor pressure ( p ) in relation to the ambient vapor pressure ( p a ), as schematically shown in Figure 2 a. An estimation of the pressure profile\ncan be made by using the classical description of droplet evaporation\nfor a given temperature and without convective flows, where vapor\ndistribution into the air is governed by molecular diffusion. 14 , 15 From the characteristic diffusion time t D ≈ d 2 / D , where\nthe diffusion coefficient of water vapor in air is D = 28 10 –6 m 2 /s at 25 °C, for example,\nthe time for water molecules to diffuse a distance d ≈ 100 μm is less than 1 ms. According to this estimation,\none may consider a quasi-steady state of vapor pressure because the\ndiffusivity is practically instantaneous for small distances. In spherical\ncoordinates, around a symmetric droplet of radius, r drop , one has: ∂( r 2 ∂ p /∂ r )/∂ r = 0, where r is the radial coordinate. 16 The boundaries of this equation are the saturated\nvapor pressure on the drop surface, p sat = p ( r = r drop ), and the ambient pressure at long distances, p a = p ( r →\n∞). A straightforward integration of the governing equation\nyields 1 which describes pressure variation in the\nradial outward direction of the surrounding air. Equivalently, a gradient\nof relative humidity (RH = p / p sat ) develops around the droplet, which decreases from 100%\non the drop surface to 46% clear of the drop (room value). 17 , 18 This microenvironment of enriched humidity spans over the patterned\nMTF and induces water condensation into the mesoporous matrix ( Figure 2 b). Figure 2 Scheme of the water-collecting\nsystem. (A) Schematic representation\nof vapor pressure contours around a sessile drop on a silicon substrate\n(side view); the drop border is micrometers apart from a region of\npatterned mesoporous film. The cooling plate of the thermoelectric\ncell in direct contact with the substrate is also shown. (B) Zoom\nof the mesoporous film in close vicinity of the drop border (side\nview). The blue arrows indicate the condensation flux, which takes\nplace in the zone where the vapor pressure ( p ) is\nhigher than the condensation pressure ( p a ) at the nanopores. The black arrow is the radial coordinate ( r ) in the plane of the substrate, where r pb is the location of the patterning border and r wf is the position of the wetting front at a\ngiven temperature. Nanopore structures effectively\ncapture water from unsaturated\nvapor, as the capillary pressure ( P cap ) arising from small curvatures in the nanoscale confinement strongly\ndecreases the vapor pressure required for condensation ( p c ), in comparison to the condensation pressure on a flat\nsurface ( p sat ). 19 In fact, according to Kelvin–Laplace equation, the condensation\npressure can be written as 20 2 where V m is molar\nvolume of the liquid, R is the gas constant, and T abs is the absolute temperature. It is worth\nreminding that p sat ( T ) is a well-known function for water, 21 which increases from 2.34 to 5.63 kPa in the range 20–35\n°C. On the other hand, the air in direct contact with the patterned\nfilm has a decreasing vapor pressure, as shown in Figure 2 b. Thus, water uptake into\nthe mesoporous film occurs in the close vicinity of the droplet border,\nwhere p ≥ p c .\nBeyond a certain radial distance, where the vapor pressure is lower\nthan p c , the film remains dry. At a given\ntemperature, the wetting front reaches a steady position ( r wf ), as shown in Figure 2 b. The series of pictures in Figure 3 a display how the\nwetted region grows as the substrate\ntemperature decreases (cooling from 30.6 to 21.4 °C in this example).\nThe dynamics of this directional water collection can be vividly observed\non the Movie S1 ( Supporting Information ) accelerated by a factor of about 4. As mentioned above, when MTF\nwas exposed to vapor gradient surrounding a droplet, water started\nto condense selectively on the mesoporous film and the position of\nthe wetting front was displaced with relatively small temperature\nvariations. In this process, the nanopores serve as temperature-activated\ncondensing sites to be able to move the wetting front directionally.\nImbibition driven by capillary condensation in nanopores has been\nrecently demonstrated by theory and experiments, 19 where vapor pressure was varied in a chamber at a given\ntemperature. However, as a novel approach, our strategy consists in\nexploiting the natural vapor pressure gradient generated by water\ndroplets at room conditions, whereas condensation is induced by controlling\nthe surface temperature, thus additionally allowing localized and\ndirectional water uptake. In what follows, we describe this effect\nby using a qualitative analysis of the underlying phenomena of vapor\ntransport. Figure 3 Working principle of the water-collecting system. (A) Time-lapse\noptical microscopy sequence showing the advancement of the wetting\nfront along the mesoporous film during a gradual temperature diminution.\nScale bar = 50 μm. (B) Qualitative graph of the condensation\npressure as a function of film temperature according to eq 2 . (C) Qualitative graph of the vapor\npressure as a function of radial distance, according to eq 1 . The red arrow indicates the displacement\nof the wetting front when the film temperature decreases. (D) Position\nof the wetting front as a function of substrate temperature; the distance\nwas measured from the MTF patterning border, and corresponds to r wf ( T ) – d 0 , where the distance d 0 is\nabout r pb ( Figure 2 b). The dashed line is to guide the eye (see Figure 4 for model prediction).\nSubstrate temperature was controlled with a thermoelectric cell, where\nvoltage varied from −1 to 2 V in steps of 0.02 V s –1 . The inset shows the linear relation between the temperature measured\nat the substrate surface and the input voltages. The Kelvin–Laplace equation allows one to rationalize\nthe\nthermal effect, taking into account that the condensation pressure\nincreases with temperature (see eq 2 ) because of the dependence of both p sat ( T ) and the exponential factor (energy\nassociated with the curvature); the overall effect is depicted in Figure 3 b. Therefore, directly\ndecreasing the film temperature enables condensation at lower vapor\npressures, that is, in the outer regions of the gradient, which produces\na displacement of the wetting front in the outward radial direction,\nas illustrated in Figure 3 c (see also eq 1 ). Furthermore, assuming that the wetting front is defined by the\ncondition p ( r wf ) ≈ p c ( T ), one may infer the existence\nof an implicit function, r wf ( T ), connecting the diagrams of Figure 3 b,c. Effectively, the function r wf ( T ) was experimentally measured and is reported\nin Figure 3 d, where\nthe dashed line is just to guide the eye. In what follows we explore\nthe theory for such correlation. In fact, equalizing eqs 1 and 2 yields 3 where the pressure values taken at\n25 °C\ncome from eq 1 and correspond\nto the vapor pressure in the air, which is supposed to be kept at\nambient conditions during the experiments, as a first approximation.\nAlso in eq 3 , the capillary\npressure has been included as P cap = 2σ/ r p , where r p is the\naverage pore radius and σ is the water surface tension (this\nparameter slightly varies with temperature and may also be considered\nconstant in the range 20–35 °C). In order to compare\nthe model prediction against experimental data,\nit should be taken into consideration that the distance plotted in Figure 3 d is d ( T ) = r wf ( T ) – d 0 , where d 0 is an arbitrary constant that measures the initial position\nof the wetting front during image analysis. Here we left d 0 as a free parameter, together with the drop radius, r drop , that enters eq 3 , which is also an effective value considering\nthe oversimplifications made to derive eq 1 . Finally, the comparison is made in Figure 4 , where it is readily seen that the theoretical prediction\nclosely follows the experimental trend. It is worth to remark that\nthe model includes all the characteristic constants of water (reported\nin the caption of Figure 4 ) and the average pore radius of the mesoporous film, r p = 5 nm. The only fitting parameters were r drop = 220 μm and d 0 = 175 μm, whose values are quite reasonable for the\nexperimental set. Figure 4 Theoretical background for the working principle. Wetting-front\nposition as a function of temperature, for different pore radius.\nThe distance plotted is d = r wf ( T ) – d 0 , where r wf ( T ) is included\nfrom eq 3 . Parameter\nvalues used in calculations are: σ = 73 mN m –1 , V m = 18 mL, R = 8.314\nJ mol –1 K –1 , p sat (25) = 3.17 kPa, and p a (25) = 0.46 p sat (25). The function p sat ( T ) was included as a correlation\nfor the range 20–35 °C. 21 Fitting\nparameters are r drop = 220 μm and d 0 = 175 μm (see text for details). In Figure 4 , the\ndashed lines correspond to smaller and larger pore radius (3 and 10\nnm, respectively), for the same r drop and d 0 values. These curves illustrate the response\nof the system for mesoporous films with different pore size. For example,\nfor a given temperature, smaller pore radius yields larger wetting\nfronts, which is because smaller radius increases capillary condensation,\nand thus, the system may capture water at relatively lower vapor pressures.\nTherefore, one may observe that eq 3 establishes a strong correlation between the wetting-front\nposition and the surface temperature, which is fully based on physical\nparameters of the system. The practical interest of this correlation\nis that the position\nof the wetting front can be handled by controlling the mesoporous\nfilm temperature, which is carried out by using a Peltier cooling\nsystem in our experiments. Liquid flow because of capillary action\nmay also play a role, driving water from wet to dry zones into the\nfilm; however, the infiltration dynamics through the nanopore structure\nis very slow in comparison to condensation, 19 whereas here the wetting front instantaneously followed the temperature\nchanges. At this point, it is worth emphasizing that the film becomes\nwetted without a bulk fluidic connection with the droplet, in contrast\nto previous setups where MTF were infiltrated from liquid reservoirs. 10 Considering a top view of the water-uptake\nphenomenon, such as\nin Figure 3 a and Movie S1 , it is readily seen how the 2D shape\nof the wetting front closely follows the isobar lines of the vapor\npressure gradient. For a given film temperature, condensation occurs\nat radial positions r ≤ r wf , where p ( r ) ≤ p c ( T ). Decreasing the film temperature\nproduces an expansion of the wetted area, and the advance of the wetting\nfront resembles a scanning of the peripheral isobars in the plane\nof the substrate. The effect is clearly observed because optical reflectance\nlocally changes upon water saturation, providing a net contrast between\nwet and dry zones of the film. 7 Furthermore,\nit has been recently shown that the relationship between the amount\nof adsorbed water and light reflectance is rather linear. 19 Control experiments performed in droplet absence\nshowed no directional water collection (see Figure S3; Supporting Information ), reinforcing the hypothesis\nthat both components of the system (mesoporous matrix and water reservoir\nin noncontact mode) are necessary to produce an addressable water-uptake\nlocation. In order to test the potential utility of the addressable\nnanocondensator\nto collect and deliver purified water, we conducted chemical reaction\nexperiments on patterned MTF, in which reagents were supplied from\nsessile droplets. Two differing configurations were tested: droplets\ndirectly in contact with the MTF and droplets adjacent to the patterned\nregion without fluidic contact with the MTF. Figure 5 a shows the localized formation of AgCl (s) when droplets of 0.1 M silver nitrate (AgNO 3 )\nand 1 M sodium chloride (NaCl) solutions are deposited over the opposite\nborders of a rectangular (1.2 mm × 2.4 mm) MTF patterning. 9 Wetted regions grew from both drops via capillary\nimbibition, and the transported reagents produced an abrupt precipitation\nof silver chloride in the contact interface. 9 In contrast, no precipitation was observed when droplets were not\nin touch with the MTF patterning ( Figure 5 b). This expected result makes evident the\ndistillated nature of the collected water and simultaneously reveals\nan attractive strategy for malleable water purification on the microscale,\nfurther suggesting the potential to integrate this controlled supply\nof clean water into advanced microdevices. Although it can be argued\nthat the instability of MTFs in aqueous media represents a liability\nof this system for long-term applications, 22 our experiments show that they are usable at least within a reasonable\ntime frame. Figure 5 Purified water collector. Optical microscopy images of the region\nwhere the wetting fronts from different drops met. Drops of 0.1 M\nAgNO 3 and 1 M NaCl solutions were deposited in opposite\nsides of a patterned MTF with (A) or without (B) direct connection\nto the film. In both cases, the lower panels show the final stage\n(after decreasing temperature): as expected, silver chloride precipitation\ntakes place only if there is a bulk fluidic connection between the\ndrops and the MTF. Scale bar = 50 μm."
} | 5,288 |
22661921 | PMC3351603 | pmc | 8,389 | {
"abstract": "The introduction of rhizobacteria that tolerate heavy metals is a promising approach to support plants involved in phytoextraction and phytostabilisation. In this study, soil of a metal-mine wasteland was analyzed for the presence of metal-tolerant bacterial isolates, and the tolerance patterns of the isolated strains for a number of heavy metals and antibiotics were compared. Several of the multimetal-tolerant strains were tagged with a broad host range reporter plasmid (i.e. pPROBE-NT) bearing a green fluorescent protein marker gene ( gfp ). Overall, the metal-tolerant isolates were predominately Gram-negative bacteria. Most of the strains showed a tolerance to five metals (Zn, Cu, Ni, Pb and Cd), but with differing tolerance patterns. From among the successfully tagged isolates, we used the transconjugant Pseudomonas putida G25 (pPROBE-NT) to inoculate white mustard seedlings. Despite a significant decrease in transconjugant abundance in the rhizosphere, the gfp -tagged cells survived on the root surfaces at a level previously reported for root colonisers.",
"introduction": "Introduction The introduction of rhizosphere-competent bacteria that are able to tolerate increased concentrations of heavy metals is a promising approach for the improvement of phytoextraction and phytostabilisation carried out by some metal-tolerant plants (Kuiper et al. 2004 ; Lebeau et al. 2008 ). Plant inoculation with the released bacteria has to be accompanied with the monitoring of their survival in the rhizosphere. Therefore, inoculants should be tagged with a marker that allows the introduced cells to be identified and monitored among the populations of indigenous soil microorganisms. One of the more useful methods to do this relies on the gfp gene, which encodes the green fluorescent protein (GFP) from the jellyfish Aequora victoria (Errampalli et al. 1999 ). The unique feature of this marker is its exclusive suitability for monitoring released bacteria in soils because the gfp is absent in soil microorganisms, and the expression of green fluorescence does not require any substrate or cofactor (Unge et al. 1998 ; Cassidy et al. 2000 ; Kozdrój et al. 2004 ). In addition, the gfp marker enables the researcher to study the introduced strains in situ with a minimum of sample preparation, thus avoiding possible disturbance of the natural cell colonisation pattern. GFP-tagged bacteria have been used to inoculate the soil and the seeds/seedlings of plants exposed to high levels of heavy metals (Liao et al. 2006 ; Braud et al. 2009 ; Ma et al. 2009 ). The major goal of such studies is to identify the conditions affecting both the survival of the released mutants and their activity in terms of the stabilisation of metal concentrations in a habitat. Beneficial bacterial inoculants are selected to support the growth of plants; these generally show tolerance to heavy metals and have phytoextraction or phytostabilisation activity (Wu et al. 2006a ; Braud et al. 2009 ). This results in the establishment of relevant plant–microorganism associations that are highly compatible in removing metal from soil (Lebeau et al. 2008 ). Metal-mine wastelands, which standardly contain high concentrations of heavy metals, such as zinc (Zn), copper (Cu), lead (Pb), nickel (Ni), and cadmium (Cd), can be rich sources of metal-resistant bacteria and metal-resistant plants (Barrutia et al. 2011 ) for bioaugmentation-assisted phytoextraction and phytostabilisation. For example, Ma et al. (2009 ) reported that the Cu-resistant strain of Achromobacter xylosoxidans significantly improved Cu uptake by metal-accumulating Indian mustard ( Brassica juncea ) and promoted plant growth. However, before any bacterial strain can be used for soil bioaugmentation-assisted phytoremediation, its survival, persistence and habitat colonisation, as well as its ability to interact with a plant host must be assessed. White mustard ( Sinapis alba ) belongs to a diverse group of mustards that show an increased tolerance to heavy metals. These plants have been suggested for application in the phytoextraction or phytostabilisation of the metals in contaminated soils (Wu et al. 2006b ; Lebeau et al. 2008 ). Reporter broad host range plasmids of the pPROBE series containing GFP gene have been employed to construct bacterial metal-biosensors (Liao et al. 2006 ) or inoculants used for bioaugmentation of a metal-contaminated soil (Braud et al. 2009 ). We report here the first trials to introduce these plasmids into multimetal-tolerant bacterial isolates from the soil of a metal-mine wasteland. The aim of the research was to determine which of the gfp -marked multimetal-tolerant bacterial strains were able to survive in the rhizosphere and on the roots of white mustard. These trials represent the first step to assess an application potential of the plant–bacteria association in a metal-contaminated environment.",
"discussion": "Results and discussion Isolation of metal-tolerant strains from metalliferous soil Soil of a metal-mine wasteland is a habitat rich in heavy metals that exert a strong selecting pressure that enables the growth of different metal-tolerant bacteria. Low concentrations of nutrients and the restricted availability of water and oxygen are additional constraints on the growth of microorganisms in this habitat. It has been reported that Gram-positive bacterial species predominate among the bacteria surviving in soils polluted with high concentrations of heavy metals (Roane and Kellog 1996 ; Ellis et al. 2003 ; Åkerblom et al. 2007 ; Sułowicz et al. 2011 ). By contrast, other studies have indicated that it is Gram-negative bacteria which predominate in sites rich in heavy metals (Kunito et al. 1997 ; Brim et al. 1999 ; Piotrowska-Seget et al. 2005 ). In total, we isolated 25 Cu-tolerant bacterial strains from the soil collected at the metal-mine wasteland; of these 16 isolates were Gram-negative and nine strains were Gram-positive organisms. With the exception of two isolates, IGB 2 and IGB 8, these metal-tolerant strains were identified to the species level based on their MIDI-FAME profiles, all with SI > 0.600 (Table 1 ). The highest number (i.e. 10) of strains were identified as Pseudomonas putida . This species and other fluorescent pseudomonads have often been reported as organisms of great adaptability to harsh conditions in soil contaminated with heavy metals (Roane 1999 ; Duponnois et al. 2006 ; Wu et al. 2006a ).\n Table 1 Bacterial strains isolated from a metalliferous soil and their patterns of tolerance to selected heavy metals and antibiotics Strain MIC a (mM) Antibiotics b \n Zn Cu Ni Pb Cd Ap Tc Km \n Bacillus cereus GB1 7 5 6 2 1 r r r Isolate GB2 7 6 5 2 1 r s r \n Bacillus sphaericus GB3 4 6 3 2 0 s s s \n Arthrobacter oxydans \n 6 6 4 2 1 r r r \n Citrobacter diversus \n 10 7 7 1 2 r r r \n Klebsiella pneumoniae \n 9 8 7 2 3 r r s \n Bacillus cereus GB7 8 6 4 3 0 s s s Isolate GB8 8 10 10 3 0 s s s \n Brevibacterium acetylicum \n 5 5 6 3 2 r r r \n Bacillus sphaericus GB10 5 7 6 2 1 s s r \n Bacillus amyloliquefaciens \n 6 8 7 2 1 s s s \n Pantoea agglomerans GP12 9 10 6 2 4 r s r \n Pantoea agglomerans GP13 9 10 6 2 3 r s r \n Pantoea agglomerans GP14 9 10 6 2 4 r s s \n Pseudomonas putida G15 8 4 2 2 2 r s s \n P. putida G16 7 4 4 2 2 r s s \n P. putida G17 8 4 4 2 2 r r r \n P. putida G18 0 2 2 1 2 r s s \n P. putida G19 8 4 4 2 2 r s s \n P. putida G20 8 4 4 2 2 r s r \n P. putida G21 4 3 4 1 2 r s r \n P. putida G22 7 3 4 3 2 r s r \n Comamonas acidovorans \n 7 3 4 3 1 r s r \n P. putida G24 4 2 2 0 1 r s s \n P. putida G25 7 3 2 2 1 r s s \n a Minimum inhibitory concentrations were determined on 0.1× tryptic soy broth (TSA) amended with the metals \n b The strains are tolerant (r) or sensitive (s) to: ampicillin (Ap; 100 μg ml -1 ), tetracycline (Tc; 20 μg ml -1 ), kanamycin (Km; 20 μg ml -1 ) \n Most bacterial strains revealed tolerance to five metals; however, the tolerance patterns differed among the isolates. Three strains of Pantoea agglomerans expressed high tolerance to all the metals. In turn, Citrobacter diversus and Klebsiella pneumoniae were tolerant to 10 or 9 mM Zn, and the latter strain was also resistant to 3 mM Cd. In addition, the isolate IGB 8 tolerated up to 10 mM of Cu and Ni (Table 1 ). Overall, most strains tolerated higher concentrations of Zn, Cu and Ni than of Pb and Cd. Similar results were obtained by Piotrowska-Seget et al. ( 2005 ) for metal-tolerant bacteria isolated from polluted arable soils and barren spoil of a former silver mine. Multimetal tolerance is a characteristic feature of different heterotrophic bacteria isolated from highly polluted soils (Trojanovska et al. 1997 ; Malik et al. 2002 ; Sułowicz et al. 2011 ). Ryan et al. ( 2005 ) found that 82% of isolates from metal-polluted soil showed resistance to five out of eight tested metals. Piotrowska-Seget et al. ( 2005 ) also found that plasmid-containing bacteria were tolerant of several metals. Multimetal-tolerant bacterial strains have been found to be able to survive in metalliferous soils planted with some plants showing an increased tolerance to various metals (Epelde et al. 2010 ; Sułowicz et al. 2011 ). In terms of antibiotic tolerance, only five of the identified species (i.e. Bacillus cereus GB1, Arthrobacter oxydans , Citrobacter diversus , Brevibacterium acetylicum and Pseudomonas putida G17) showed a tolerance to Tc, Kn and/or Ap. A similar number of strains, however, all belonging to one species, namely P. putida (i.e. G15, G16, G18, G24 and G25), tolerated only Ap. By contrast, four strains (i.e. Bacillus sphaericus GB3, B. cereus GB7, B. amyloliquefaciens and the isolate IGB 8) were sensitive to all three antibiotics (Table 1 ). Antibiotic tolerance often accompanies an increased resistance to heavy metals among different bacteria isolated from sites exposed to high pollution and/or containing material rich in the metals (Berg et al. 2005 ; Stepanauskas et al. 2005 ; Baker-Austin et al. 2006 ). This situation results from the co-transfer of antibiotic resistance genes and those of metal resistance on the same plasmid under selective conditions (Foster 1983 ; Baker-Austin et al. 2006 ). However, the ecological role of this association for bacterial strains occupying a severe habitat is not fully understood. Presumably, the extra feature of antibiotic resistance increases their survival success during the competition for available niches when in the presence of compounds acting as antimicrobials and signal molecules. As a result, respective strains of bacteria may differ in their tolerance profiles, which may in turn affect their survival (Alonso et al. 2001 ; Hibbing et al. 2010 ). Our cluster analysis of the metal and antibiotic tolerance patterns of all the bacterial strains showed that they grouped into two major clusters (Fig. 1 ). The first composite cluster included only strains of Pseudomonas putida , with the one exception being Comamonas acidovorans . However, the second major cluster was composed of two subclusters, with one comprising all of the Gram-positive strains, and the second consisting of Gram-negative strains of Pantoea agglomerans (clustering together) and Klebsiella pneumoniae grouping with Citrobacter diversus separately (Fig. 1 ). This clustering pattern shows that the separation of the P. putida group may have been associated with the possession of a common mechanism(s) of metal tolerance that is chromosomally encoded (Cánovas et al. 2003 ). In contrast, the Gram-positive strains with their thicker cell envelopes react to biocides differently. Their clustering might be, at least in part, due to the presence of a common plasmid carrying genes for multimetal and drug resistance (Kamala-Kanan and Kui Jae 2008 ). Bacterial species such as P. agglomerans , K. pneumoniae and C. diversus belong to the same family of Enterobacteriaceae . Therefore, they formed the separate cluster towards the other Gram-negative group of P. putida in this study. In addition, their likeness supports the fact the multidrug resistance ( mar ) operon is widespread among enteric bacteria (Cohen et al. 1993 ). However, closer clustering of these bacteria with the Gram-positive group may be explained by their similar ability for protection against the biocides, due to the role of the cell envelopes (i.e. glycocalyx and thick cell wall, respectively). Regarding the cluster of K. pneumoniae and C. diversus , their likeness may have resulted from the presence of the same phosphatase-mediated metal accumulation process involved in the detoxification of the bacteria (Macaskie et al. 1994 ).\n Fig. 1 Dendrogram representing similarities of metal and antibiotic tolerance patterns of different bacterial strains isolated from soil of a metal-mine wasteland. Ao Arthrobacter oxydans , Ba Brevibacterium acetylicum , Bam Bacillus amyloliquefaciens , Bc Bacillus cereus , Bs Bacillus sphaericus , Ca Comamonas acidovorans , Cd Citrobacter diversus , IGB2 , IGB8 isolates GB2 and GB8, Kp Klebsiella pneumoniae , Pa Pantoea agglomerans , Pp Pseudomonas putida \n \n gfp -tagged strains and its survival in the mustard rhizosphere The underlying rationale for isolating metal-tolerant bacteria from metalliferous soils is their potential application for bioaugmentation-assisted phytoremediation of these habitats. Various markers, often with different detection frequencies, have been used to track the fates of these metal-tolerant bacterial isolates following their reintroduction into the soil (Zaidi et al. 2006 ; Ma et al. 2009 ). In our study, we successfully introduced the plasmid pPROBE-NT by triparental conjugation into four strains: Pseudomonas putida G16, P. putida G20, P. putida G25 and Comamonas acidovorans . The growth of these strains on selective agar medium containing Ap (100 μg ml -1 ), Km (20 μg ml -1 ) and 3 mM Zn was used to select for the pPROBE-NT transconjugants as neither donor strains nor recipients were able to grow on the medium amended with these markers. We also checked the putative transconjugants for GFP production under the confocal microscope. All transconjugants gave positive results, with P. putida G25 (pPROBE-NT) showing the strongest green colour, indicating intensive synthesis of the protein (Fig. 2 ). Therefore, we used transconjugant strain G25 for further survival experiments. An approach that involves the tagging of bacterial strains with gfp genes by plasmid transfer or recombination into the chromosome has been recommended by various authors because of the efficient detectability of the tag and the low energetic burden placed on the cells. (Kendall and Badminton 1998 ; Errampalli et al. 1999 ; Kozdrój et al. 2004 ).\n Fig. 2 Photograph of microscopic image of fluorescent green fluorescent protein ( gfp )-tagged transconjugant Pseudomonas putida G25. The photograph was obtained using a confocal laser scanning microscope \n The introduction of metal-tolerant bacterial strains into soil seeded with plants that tolerate increased concentrations of heavy metals has been reported as a promising approach that facilitates the survival and development of these plants in contaminated habitats (Abou-Shanab et al. 2003 ; Sheng and Xia 2006 ; Ma et al. 2009 ). However, the success of this approach is dependent on the potential of the inoculant to colonise plant roots efficiently, which in turn is related to its own survival in the rhizosphere. To estimate the level of adaptation between the released inoculant cells and white mustard, we inoculated plant seedlings growing in a sandy soil. The numbers of gfp -tagged Pseudomonas putida G25 colonising the roots of white mustard decreased from the initial log 7.48 ± 0.28 to log 4.95 ± 0.25 and log 3.62 ± 0.18 CFU g -1 dry soil on days 14 and 54 post-inoculation, respectively. The average counts of the total indigenous heterotrophic bacteria were about log 7.95 ± 0.24 CFU g -1 dry soil in the rhizosphere. However, the natural resistance to Ap, Tc and Km among the indigenous bacteria was below the detection limit of log 1.47 ± 0.15 CFU g -1 dry soil. The microscopic observation of root and rhizosphere preparations confirmed the successful survival of the transconjugant in the rhizosphere and the colonisation of the roots of white mustard seedlings on day 7 (Fig. 3 ). Although the presence of the gfp -tagged transconjugants on the roots was still visible on day 54, only a few bacterial cells were noticeable in the rhizosphere specimen (Fig. 4 ). A decrease in counts of introduced bacteria over a few days is often observed due to competition for nutrients and space with other rhizosphere microorganisms. They are grazed on by protozoa and exposed to abiotic stress; some cells die or lose culturability following release (van Veen et al. 1997 ). As a result, the inoculant population ultimately reaches a level reflecting its ability to adapt to conditions prevailing in the rhizosphere of the appropriate plant species (de Weger et al. 1995 ; Kozdrój et al. 2004 ). Errampalli et al. ( 1998 ) indicated that gfp -marked Pseudomonas sp., introduced into a creosote-contaminated soil, declined over a 26-day period, although the low numbers recovered up to 13 months after inoculation. It can also not be excluded that the decreased numbers of P. putida G25 (pPROBE-NT) may have resulted from the loss of the plasmid over time. However, this vector has been reported to be a stable one in a broad range of bacterial hosts (Miller et al. 2000 ). Belimov et al. ( 2004 ) reported a slight decrease in the numbers of inoculant rhizobacteria between days 10 and 25 during their colonisation of barley roots. By contrast, an introduced population of Bacillus sp. that was resistant to Cd was still detectable at the same density in the rhizosphere of rape 2 weeks after inoculation (Sheng and Xia 2006 ). In addition, the survival of inoculants associated with a plant host depends on changes in the physiological state of the plant (Lebeau et al. 2008 ). Wu et al. ( 2006b ) reported that young mustard seedlings are more favourable to an introduced metal-tolerant strain than flowering plants, possibly due to differences in the composition of the root exudates. We obtained similar results for the survival of gfp -tagged P. putida G25. Immobilisation of bacterial inoculants into carriers, such as alginate, clay, peat or methyl cellulose, which protects them against biotic and abiotic environmental stress, can increase both their survival in soil as well as their colonisation of soil (van Veen et al. 1997 ; Kozdrój et al. 2004 ; Braud et al. 2009 ). However, to facilitate colonisation of the entire rhizosphere and roots of growing seedlings by the released bacterial strains, the application of free-cell suspensions, instead of immobilised cells, appears to be useful (Ciccillo et al. 2002 ; Mazolla et al. 1995 ).\n Fig. 3 Photographs of microscopic images of fluorescent gfp -tagged transconjugant Pseudomonas putida G25 colonising the root surface of a 7-day-old seedling of white mustard ( a ) and those surviving in the rhizosphere ( b ). The photographs were obtained with a confocal laser scanning microscope \n Fig. 4 Photographs of microscopic images of fluorescent gfp -tagged transconjugant P. putida G25 colonising the root surface of a 54-day-old seedling of white mustard ( a ) and those surviving in the rhizosphere ( b ). The photographs were obtained with a confocal laser scanning microscope \n In conclusion, the soil of the metal-mine wasteland is a habitat favouring the selection of multimetal-tolerant heterotrophic bacteria, mostly represented by Gram-negative species, which can be differentiated according to their metal and antibiotic tolerance patterns. Although these bacteria are characterised by their high metal tolerance, only a few strains can be recipients of the gfp -bearing reporter plasmid and subsequently express the green fluorescent protein. Indeed, one recipient strain, Pseudomonas putida G25, yielded a transconjugant that distinctly expressed GFP and was able to colonise the rhizosphere and roots of white mustard seedlings. Despite a significant decrease in the counts of the transconjugant in the rhizosphere, the gfp -tagged cells persist at the level reported for root colonisers (Scher et al. 1994 ). This result is a promising indicator of plant–transconjugant interdependence that can favour both partners. Further studies are needed to determine whether P. putida G25 (pPROBE-NT) can promote the growth of white mustard in soil containing high concentrations of heavy metals, bearing in mind that the ultimate goal is the potential application of both organisms in bioaugmentation-assisted phytoremediation of polluted habitats."
} | 5,233 |
35426531 | PMC9012074 | pmc | 8,390 | {
"abstract": "Abstract The present work focuses on: (1) the evaluation of the potential of Chlorella fusca to grow and synthesize metabolites of biotechnological interest, after being exposed for fourteen days to urban wastewater (UW) from Malaga city (UW concentrations: 25%, 50%, 75%, and 100%); (2) the study of the capacity of C. fusca to bioremediate UW in photobioreactors at laboratory scale; and (3) the evaluation of the effect of UW on the physiological status of C. fusca , as photosynthetic capacity by using in vivo Chl a fluorescence related to photosystem II and the production of photosynthetic pigments. C. fusca cell density increased in treatments with 50% UW concentration, followed by the treatment with 100% UW, 75% UW, the control, and finally 25% UW. Protein content increased to 50.5% in 75% UW concentration. Stress induced to microalgal cultures favored the increase of lipid production, reaching a maximum of 16.7% in 100% UW concentration. The biological oxygen demand (BOD 5 ) analysis indicated a 75% decrease in 100% UW concentration. Dissolved organic carbon (DOC) levels decreased by 41% and 40% in 50% UW and 100% UW concentration, and total nitrogen (TN) decreased by 55% in 50% UW concentration. The physiological status showed the stressful effect caused by the presence of UW on photosynthetic activity, with increasing impact as UW concentration grew. In the framework of circular economy, we seek to deepen this study to use the biomass of C. fusca to obtain metabolites of interest for biofuel production and other biotechnological areas. Graphical Abstract",
"introduction": "Introduction At the international level, different countries have advanced the Sustainable Development Goals (SDG) adopted by the General Assembly of the United Nations in 2015, where two of the objectives agreed by the group were: “Ensure availability and sustainable management of water and sanitation for all” (SDG 6) and “Ensure access to affordable, reliable, sustainable and modern energy for all” (SDG 7). These environmental policies reflect the need for advancing jointly in the optimization of effluent treatment processes in urban agglomerates and in the use of different sources of renewable energy to limit the use of fossil fuels, which generate an increase in greenhouse gases (GHGs) (Mikhaylov et al. 2020 ). The urban effluents are commonly referred to as \"urban wastewater\" (UW) and include domestic wastewater or its mixture with industrial wastewater and/or stormwater runoff. Domestic wastewater contains effluents from housing and service areas generated mainly by human metabolism and domestic activities (European Commission 1991 ). They contain a very diverse physicochemical composition, from simple organic and inorganic compounds to substances that are known as \"emerging contaminants\", including drugs (Salibián 2014 ), microplastics (Woodward et al. 2021 ), and nanoparticles (Kühr et al. 2018 ). Currently, the problem is becoming more complex with the COVID-19 pandemic. Some researchers detected SARS-CoV-2 in UWs from different countries in America, Europe, Asia, and Oceania (Haramoto et al. 2020 ; Randazzo et al. 2020 ; Giraud Billoud et al. 2021 , among others). The UW treatment aims to avoid damage to public, private, and industrial water supplies, to water intended for recreation, fishing activities, agriculture, and also depreciation of the value of the land and impacts on the ecosystems and human health (Cossio et al. 2021 ). In recent years, the need to contribute to the development of efficient and economically convenient methods for UW treatment has motivated the study of new technologies. In this sense, microalgal applications have increased in the last decade, due to their importance in phycoremediation and to their extensive application potential in biopharmaceutical, nutraceutical, and renewable energy industries (Khan et al. 2018 ; Rao et al. 2019 ). Microalgae could generate a wide variety of bioproducts, including pigments, proteins, polysaccharides, lipids, vitamins, antioxidants, and bioactive compounds (Tang et al. 2020 ). Several species of microalgae have been investigated for their ability to synthesize and store bioproducts with pharmacological and biological properties, as well as metabolites used for biofuel production (Khan et al. 2018 ). Different species of the genus Chlorella have been described as possible sources for biodiesel production according to their oil properties (Sathish, 2020 ; Katiyar et al. 2021 ). There are records of the use of different fractions of industrial, domestic, or urban effluents as culture media for chlorophyte microalgae (Singh et al. 2017 ). This is based on the presence of compounds that can be sources of organic carbon, nitrogen, phosphorus, and trace elements. Many of these compounds can be efficiently metabolized (Pacheco et al. 2015 ). Compared to other biofuel feedstocks, microalgae have the following benefits: (1) They do not compete with cropland and freshwater crops because they can be grown in saline water and nonarable land (Cai et al. 2013 ). (2) They can grow at an exceptionally fast rate (Tredici 2010 ). (3) They show a high oil content of 20%-50% dry weight (DW) (Cai et al. 2013 ), in which the lipid can be converted into biodiesel (Chisti 2007 ; Lam and Lee 2011 ; Singh et al. 2017 ). According to Tsukahara and Sawayama ( 2005 ), a lipid content of 30% in microalgae cells is equivalent to an annual lipid production of 4.5–7.5 ton/ha. (4) They are a perfect candidate for the sequestration of CO 2 and the reduction of GHGs. (5) They can utilize nutrients from most wastewaters, providing an alternative method for their treatment (Mureed et al. 2018 ; Li et al. 2019 ). (6) Residues of microalgal biomass, after lipid extraction, can be utilized as a source of nitrogen, as a protein-rich food, or crop fertilizer (Li et al. 2019 ). Chlorophyta species are a good candidate for the remediation of effluent coming from urban and rural areas (Katiyar et al. 2021 ; Kumari et al. 2021 ). Among these, Chlorella fusca is a species with high productivity in cultures at different scales (Jerez et al. 2014 , 2016 ). However, it is important to continue studying the possible use of C. fusca biomass, to depurate UW, mitigate eutrophication in aquatic ecosystems, and decrease economic costs of both the treatment of UW and biomass production (Peralta et al. 2019 ). Based on this background, in the present study we analyzed the biomass production of C. fusca exposed to UW, as well as contaminant removal efficiency, the lipid and protein content, and the electron transport rate and maximum quantum yield, both considered indicators of microalgal physiological status. In vivo chlorophyll a fluorescence has been successful used as estimator of photosynthetic capacity in C. fusca in different type of photobioreactors and it has been related to biomass productivity and accumulation of bioactive compounds (Jetrez et al., 2016; Peralta et al., 2019 ).The use of effluents as culture media to produce metabolites of interest has the multiple purpose of contributing to their purification, through the degradation of organic and inorganic contaminants, and valuing them to obtain a product of interest.",
"discussion": "Discussion Microalgae have lately attracted great interest worldwide due to their application potential in wastewater remediation, in nutraceutical, pharmaceutical, and renewable energy industries (Pahazri et al. 2016 ; Khan et al. 2018 ). Different researchers demonstrated the purification capacity of microalgae and verified that the biomass obtained had a high content of bioproducts (Tang et al. 2020 ). On the other hand, the use of wastewater as a substitute for algae nutrients would significantly reduce the operational cost of the cultures. Wastewater contains phosphorus, nitrogen, carbon, and other constituents needed for microalgal growth, although some other undesirable compounds such as emerging pollutants and heavy metals can also be found (Morales-Amaral et al. 2015 ). In this sense, Khan et al. ( 2018 ) described the advantages of microalgae to produce biofuels and various bioactive compounds and discussed culturing parameters. According to these authors, the most important and challenging issues are increasing microalgal growth rate and enhancing bioproduct synthesis, among others. In the present study, we investigated the ability of C. fusca to grow in a medium composed of UW, as well as its efficiency for the removal of contaminants and its potential to produce biomass with a high lipid and protein content. The results show that the final cell growth values in the 50% UW, 75% UW, and 100% UW treatments were higher than in the control. In cultures with 50% UW, the highest cell density and the shortest doubling time were obtained (Fig. 1 ). According to Katiyar et al. ( 2021 ) Chlorella minutissima and Chlorella sorokiniana showed higher growth rate, lipid content, and biomass productivity, when cultured in wastewater than in control. Similar results were reported by Singh et al. ( 2017 ), although these authors recorded a higher growth rate and biomass production, and shorter doubling time, using Parachlorella kessleri -I in 100% municipal wastewater concentration. This greater growth in the treatments with UW is possibly because the concentration of ammonium as a nitrogen source prevents this nutrient from being limiting (Gómez Serrano 2012 ). Lin et al. ( 2007 ) observed that Chlorella pyrenoidosa (LK) was ammoniacal-N tolerant, with cell density increases in leachate with 405 mg l −1 of NH 4 + . Similar results were observed in this study, where the cell density of C. fusca increased in UW with about 620 mg l −1 of ammonium. In turn, the UWs provided a greater amount of nutrients to the treatments, including PO 4 3− , SO 4 2− , K + , Mg 2+ (among others) favoring growth under these conditions. However, the nutrients are not the sole requirement for microalgal development, since temperature, light, aeration, and pH, as well as mixing, can contribute to their growth (Pahazri et al. 2016 ). In algal cultures, the pH usually increases due to photosynthetic CO 2 assimilation. This pH increase can be compensated by respiration (Jyoti and Awasthi 2013 ). According to this, in our study, pH values would be expected to increase along the experiment due to photosynthetic activity; however, the pH increased only in the control, while in cultures with UW, it went down, despite their photosynthetic activity (Fig. 2 ). Nitrogen is usually present in wastewater as NH 4 + , contributing significantly to the changes in pH value. Assimilation of nitrate ions by algae tends to raise the pH, but if ammonia is used as nitrogen source, the pH of the medium may decrease (Kong et al. 2011 ; Pahazri et al. 2016 ). Scherholz and Curtis ( 2013 ) analyzed the influence of ammonium and pH on the growth of Chlorella vulgaris in photobioreactors and observed that cultures provided with 4.5% nitrogen from ammonium showed significant growth. At the same time, they observed a decrease in pH during the growth phase, followed by a pH rise indicating sequential ammonium and nitrate metabolism. These results are consistent with the almost exclusive consumption of ammonium instead of nitrate and corroborate the statements of previous works where it is mentioned that the use of nitrate is inhibited in the presence of ammonium (Florencio and Vega 1983 ). In the 1970s, the worldwide energy crisis encouraged the use of microalgae as renewable and sustainable sources to produce biofuels. Depending on the type of wastewater used in the cultivation of microalgae, the lipid percentage obtained will be different. In agricultural wastewaters Chlorella sp. showed 9% DW (Jacobson and Alexander 1981 ) and 13.6% DW (Wang et al. 2010a ) of lipid content. In industrial effluents, Chlorella saccharophila was obtained with 18.10% DW of lipid content (Chinnasamy et al. 2010 ). In this study, the lipid percentage was 14.7% DW (25% UW), 15.5 DW (50% UW), 16% DW (75% UW), and 16.7% DW (100% UW). These values are comparable to those obtained by Jebali et al. ( 2015 ) and Hernández et al. ( 2016 ). Hempel et al. ( 2012 ) showed that the strains with the highest lipid content were Chlorella sp. 589 (30.2% DW), C. saccharophila 477 (27.6% DW), and Chlorella sp. 800 (24.4% DW). It is known that the production of lipids increases up to 65% by the nutritional deprivation of microalgae (Markou and Nerantzis 2013 ). In our case, we observed a smaller increase. This is because, despite being the original purpose, the high concentration of ammonium in the UW and the ability of C. fusca to metabolize it showed it was not in a status of nutritional nitrogen deprivation. The increase in lipid production is possibly related to the stress conditions induced by physicochemical conditions of the UW. Probably the solution may be to perform assays in two phases: the first to decrease the concentration of nutrients, essentially nitrogen, and a second crop in which, with nitrogen deficiency, lipid production increases. This solution was also given by Cai et al. ( 2013 ). Moreover, it has been recorded that the maximum accumulation of lipids in Chlorella sp. is related to the pH of the medium, the optimum being pH values between 7.0 and 8.5 (Wang et al. 2010b ; Sakarika and Kornaros 2016 ). In our study, the pH range in the treatments with UW was 6.4 – 8.1, therefore this parameter should be considered in later studies, in order to increase the productivity of lipids in C. fusca . The cultivation of algae has expanded to new fields, such as feed and food, cosmetics, and biopharmaceutical products (Khan et al. 2018 ). In this regard, different authors investigated biomass productivity and the synthesis of amino acids on microalgae strains. Hempel et al. ( 2012 ) identified strains with an amino acid content of more than 40% DW: Spirulina platensis reached a protein content of 46.8% DW, Chlorella sp. 589 of 44.3% DW, and C. saccharophila 477 of 42.4% DW. In our study, a higher concentration of UW in the cultures leads to an increase in the production of proteins, implying a greater assimilation of nitrogen, which in the UW is mainly ammonium (Peralta López 2013 ). Given the high proportion of protein accumulated in 75% UW treatment (51%), this system can be considered as a form of protein production for commercial use. In addition to demonstrating that the biomass of microalgae obtained in cultures with wastewater has a high content of bioproducts, these authors verified their capacity and efficiency for bioremediation. Although C. fusca is not a species commonly used in the remediation of UW (Pahazri et al. 2016 ), we demonstrate that it has the capacity to grow mixotrophically and accumulate nutrients from it. In this study, the maximum efficiency of DOC removal was obtained on the tenth day of exposure, with removal efficiency ranges that vary between 23.35% and 45.48% (Fig. 4 ). Katiyar et al. ( 2021 ) reported a higher total organic Carbon removal efficiency by C. minutissima and C. sorokiniana (95% and 98%, respectively) in wastewater collected from India (500 ml, for 12 days), compared to that recorded in this study. This trend is justified by different authors: Eny ( 1951 ) observed that the metabolic route of Chlorella sp. could be altered with the supply of organic substrates such as glucose or organic acids, which means that they can perform not only autotrophic but also heterotrophic growth. The organic compounds can be used as an essential nutrient (Sachdev and Clesceri 1978 ) or as an accessory growth factor (Saunders 1957 ). The heterotrophic growth of microalgae (including Chlorella sp.) can be rapid, from the incorporation of organic substrates in the oxidative assimilation process for storage material production (Burrell et al. 1984 ). In this study, TN decreased at the end of the incubation period (14th day) in the 50% UW and 100% UW treatments, obtaining a removal efficiency between 55% and 24.6%. Similar percentages were reported by Katiyar et al. ( 2021 ), with TN removal rates (12 days) for C. minutissima and C. sorokiniana as 28.46% and 40% respectively. An opposite tendency was observed by Lin et al. ( 2007 ), who reported that the relative NH 4 + removal rate in lower leachate concentrations (10% and 30%) was higher than that in higher concentrations (50%, 80%. and 100%). In this sense, C. fusca is considered suitable microalgae for the degradation and elimination of nitrogenous waste present in UW. The fact that microalgal cultures subjected to a high concentration of UW maintain their capacity for carbon removal is crucial for their possible use as a bioremediation system. The UW contains a microorganism cocktail (bacteria, protozoa, rotifers, among others) that feed on organic matter giving rise to a very high BOD 5 (1000 mg l −1 ) (Madoni 2011 ). During the experiment, in the treatment with 100% UW, BOD 5 decreased by 75% (from 1000 mg l −1 to 250 mg l −1 ). The dissolved organic matter in the UW is probably assimilated by microalgae and is thus eliminated. He et al. ( 2013 ) reported that the microalgae–bacteria consortium present in wastewater is more efficient in removing BOD (97%). The dynamics of photosynthetic pigment content, changes in their ratio, as well as rETR and Fv/Fm associated with PSII, indicate C. fusca adaptation in response to the impact of UW addition to the culture medium. Different authors reported that the accumulation of Chl a is related to nitrogen metabolism (Rüdiger and López-Figueroa 1992 ). Chu et al. ( 2015 ) found that Chl a content in C. vulgaris cultures increased on the first days of incubation, when there was enough N in the medium, to decrease later when N was running low. In our case, the trend was similar, decreasing from day 10 until the end of the experiment. At the same time, in 50% UW and 75% UW concentrations, the carotenoid concentration increased at the end of the incubation period; similar results were reported by Kiran et al. ( 2014 ). In this work, the increase in UW concentration was accompanied by a reduction of the Chl a /Chl b ratio and an increase in the pigment index (carotenoids/Chl a ) in C. fusca . The Chl a /Chl b ratio can characterize the photochemical potential and biosynthetic activity of algae, controlling the absorbed light intensity (Tanaka and Melis 1997 ). Thus, under stress action, a decrease in Chl a content takes place, and accordingly the ratio between these two types of pigment decreases. When this happens, the pigment index increases due to the formation of carotenoids that perform a supporting and protective role in the photosynthesis. Chl a is the part of reaction centers and peripheral complexes of photosystems I and II, and Chl b is the component of the light-collecting complex of photosystem II. Therefore a change in the Chl a /Chl b ratio may indicate a shift of stoichiometric balance between the reaction center complexes of both photosystems and the light-collecting complex of photosystem II (Bodnar et al. 2016 ). Also, the quality of the light inside each reactor was affected by the presence of UW, which possibly has a negative impact on cultures (Peralta et al. 2019 ). In this sense, the characteristics of the medium with different UW concentrations conditioned the biomass optical properties and consequently, the ETRmax and Fv/Fm. The UW treatments showed initial Fv/Fm values lower than the control, demonstrating the stress induced by the culture medium with UW at the beginning of the experiment, improving their yield at the end when manage to acclimate to the conditions of the culture medium. Similar results were obtained by Peralta et al. ( 2019 ), who reported that Fv/Fm was correlated with nitrogen content and maximal rETR with photosynthetic performance and nitrogen metabolism. The UW treatments reached the lowest rETR values at the end of the experiment in relation to the control. This indicates that microalgae could not achieve optimal photosynthetic activity, possibly due to the toxicity and turbidity of UW. The cultivation of microalgae in UW is a promising alternative for its treatment, at the same time as it provides a culture medium with nutritional properties. Even though UW affected the photosynthetic activity of microalgae, they were able to grow and synthesize lipids and proteins. The TN, DOC, and BOD 5 reduction efficiency increased with the exposure time, showing the potential of C. fusca to remove these compounds from effluents on a laboratory scale. Therefore, this is recommended as an eco-friendly method that should be tested for wastewater treatment on a larger scale, improving the factors limiting the performance of these microalgae-based wastewater treatment systems (Acién et al. 2016 )."
} | 5,257 |
37487096 | PMC10401026 | pmc | 8,392 | {
"abstract": "Significance Plants with reduced lignin content are desirable because their biomass is more amenable to biofuel production; however, the engineering of lignin content and composition often results in impaired growth for unknown reasons. This phenotype could be caused by altered water transport through the xylem of low-lignin plants, but other models suggest that nonstructural phenylpropanoid metabolites could be involved in this phenotype. Lignans are products of phenylpropanoid metabolism that are widely distributed in the plant kingdom, but their biological functions have been understudied. Here, we present evidence suggesting that reductions in the content of the lignan pinoresinol and its precursor, coniferyl alcohol, are involved in the impaired growth phenotypes of low-lignin plants and also lead to perturbations in root development.",
"discussion": "Discussion Specialized metabolites are often considered nonessential for plant survival at least in the short term, but some, including several phenylpropanoids, are now known to have roles in plant growth and development ( 33 , 43 , 59 – 65 ). Additionally, the CA derivatives, dihydroconiferyl alcohol and dehydrodiconiferyl alcohol glucosides, have been reported to promote cell division in a manner similar to cytokinin ( 54 , 55 , 66 , 67 ), although this observation was recently refuted ( 57 ). LMID is a defect in the shoot system’s growth and development caused by alterations in phenylpropanoid metabolism. This defect is often attributed to a deficiency in lignification that leads to vascular collapse and poor water transport, but additional models suggest that LMID could be the result of the activation of a cell-wall-integrity-sensing mechanism or the hyperaccumulation/hypoaccumulation of phenylpropanoid intermediates ( 7 ). Several studies have revealed that vascular collapse and LMID can be disentangled ( 31 , 68 ), and others have suggested that hyperaccumulation of phenylpropanoid metabolites can be the cause of impaired plant growth ( 69 , 70 ). Recently, the compromised rosette and stem size of a lignin-deficient mutant with collapsed xylem was rescued by vessel-specific complementation ( 6 ), but this experiment does not rule out the possibility that a growth-promoting phenylpropanoid metabolite is synthesized in vascular tissues or other cells in which phenylpropanoid promoters are active. Investigations into the role of soluble phenylpropanoids as growth regulators have been impeded by the fact that their synthesis shares a common biosynthetic pathway with lignin; genetic manipulations that alter their content also affect lignin levels ( 71 ). Although detrimental effects on plant growth arising from genetic manipulations of the phenylpropanoid pathway may often be the result of alterations in lignin content and vascular collapse, the common origin of lignin and soluble phenylpropanoids may have obscured the possible role of these metabolites as plant growth regulators. Coniferyl Alcohol and Its Dehydrodimer Pinoresinol Rescue the Growth Phenotype of cadc cadd \n C4H:F5H . Plants overexpressing F5H deposit primarily S lignin and exhibit only a mild negative-growth phenotype ( 36 – 38 , 41 ). The compromised growth of ref3-2, ccr1 , and cadc cadd plants overexpressing F5H does not appear to be due to the synergistic effect of a decrease in lignin content and a change in composition because the lignin content of these mutant backgrounds ( 13 , 40 , 72 ) does not correlate with the severity of their growth phenotypes observed upon introduction of C4H:F5H. Furthermore, the addition of C4H:F5H to a ref3-3 background results in dwarfism ( Fig. 1 ), although the ref3-3 mutant does not have altered lignin content ( 13 ). Together, these data are inconsistent with a lignin-based phenotype, although we cannot exclude the possibility that the phenotypes of F5H overexpressors are due to distinct causes, such as the low lignin content of ref3-2 and the aldehyde-rich lignin of cadc cadd. Taken together, these data suggest that the growth impairment in cadc cadd, ccr1, ref3-2 , and ref3-3 plants overexpressing F5H is due to alterations in phenylpropanoids other than lignin. Supplementation with CA or PR restored the growth of cadc cadd C4H:F5H ( Fig. 2 B ), ref8-2 , and ref3-2 C4H:F5H ( SI Appendix , Fig. S5 ) plants. Exogenous CA was recently reported to lead to the upregulation of MYB4 , ( 42 ) which would be expected to repress expression of our C4H promoter-driven F5H overexpression construct ( 73 ). Instead, under our experimental conditions, transcript levels of the endogenous C4H were unaffected by CA treatment in cadd C4H:F5H seedlings, and the expression of the F5H transgene was only modestly affected and remained 12-fold higher than in wild type ( SI Appendix , Table S2 ). Similar levels of F5H overexpression are sufficient to effectively elevate S-lignin composition ( 36 ). This observation and the fact that PR does not affect C4H:F5H transcript accumulation show that phenotypic complementation comes about through a mechanism distinct from a decrease in F5H overexpression. The decrease in CA and coniferin in ref3-2 , ref3-3 , and ccr1 C4H:F5H is consistent with a model in which upstream mutations reduce CA content, and high levels of F5H activity caused by the C4H:F5H construct further deplete the CA pool ( Table 1 ). Despite growing relatively normally, the transcriptional profiles of cadc cadd and fah1-2 C4H:F5H plants are substantially altered and share many of the same misregulated groups of genes, suggesting that these plants’ altered transcript levels could be caused by a common factor, possibly the depletion of CA ( 74 ). Surprisingly, the levels of CA in cadd C4H:F5H and cadc cadd C4H:F5H are instead elevated, an observation that is at odds with this model. The fact that CA complements the growth phenotype of cadc cadd C4H:F5H suggests that the increase of endogenous CA levels in these plants does not occur in cells in which it is needed for growth. The synthesis of CA in the absence of CADC and CADD may be catalyzed by one or more of the additional seven CADs that the Arabidopsis genome encodes. Substantial CAD enzymatic activity can be detected in cadc cadd stem extracts ( 75 ). This activity clearly does not compensate for the lack of CADC and CADD given the lignin phenotype of the cadc cadd mutant, pointing to the importance of tissue-specific localization of phenylpropanoid enzymes and metabolites. Finally, it is notable that CA rescues the dwarf phenotype of cadc cadd C4H:F5H plants, although plants are overexpressing F5H , which uses CA as a substrate. This observation suggests that the concentrations of CA in vivo that rescue these growth phenotypes might be substantially lower than what was applied exogenously and may also explain why PR rescues growth at lower concentrations than CA does. Lignans and neolignans are widely distributed among the plant kingdom ( 51 ), but their functions in planta have been understudied. These compounds have been associated with plant defense for their antifeedant, allelopathic, and antifungal properties ( 76 ), and recently, the role of a sinapoylcholine-conjugated neolignan in seed coat protection has been demonstrated ( 77 ). Reports on coniferyl alcohol-derived molecules affecting plant growth date back to the 1970s ( 78 ); however, there are only isolated reports on this topic, and they provide limited data to support the role of these molecules in plant growth. In our studies, DCG, its aglycone, and DCA did not have a positive effect on the growth of cadc cadd C4H:F5H seedlings ( Fig. 4 A and SI Appendix , Fig. S4 ). Although PR and other CA dimers can be incorporated into lignin, only the 8–8 dimer is able to rescue the growth of these plants, whereas 8– O –4 or 8–5 dimers cannot, although the same linkages are found in native lignin and should be able to supplement lignification as readily as PR. Taken together, our results suggest a role for PR or its derivatives in plant growth regulation. Lignans Restore Lateral Root Development in cadd \n C4H:F5H . The importance of phenylpropanoids in root development has been explored in recent studies. The C4H null mutant, ref3-4 , has a deficiency in LTR development and an increase in adventitious roots which was attributed to the inhibition of auxin transport by both cis- cinnamic acid accumulation and impaired phloem-mediated transport caused by lignin deficiency ( 79 ). Addition of coniferaldehyde complements the poor water transport of ref3-4 seedlings and alleviates its adventitious root phenotype presumably by complementing the mutant’s low-lignin phenotype ( 79 ). Kaempferol negatively regulates lateral root development by scavenging ROS in lateral root primordia where it is required for emergence. As a result, the tt7 mutant, defective in FLAVONOID 3′-HYDROXYLASE , has increased levels of kaempferol and a decrease in emerged LTRs, but an increase in unemerged LTRs ( 80 ). Based on these observations, our previous definitions of hormones and secondary metabolites may have to be revisited given the importance of secondary metabolites in fine tuning growth and developmental processes. In addition to their growth defect, cadc , cadd , and cadc cadd seedlings carrying the C4H:F5H construct have fewer LTRs, and this phenotype can also be rescued by chemical complementation with CA and its derivative, PR. In addition to the vasculature, roots have an additional lignin-made structure, the Casparian strip (CS), which plays a role in controlling the diffusion of water and solutes through the endodermis ( 81 ). Mutants in C4H and MYB36 are defective in CS deposition ( 82 , 83 ), and incorporate increased levels of suberin in the endodermis, the degradation of which is necessary for LTR emergence ( 84 ). CS defects in these mutants can be complemented by exogenous CA ( 83 , 85 ), which would appear to suggest that defects in CS deposition and an increase in suberization could explain the decrease in LTRs in cadd C4H:F5H and cadc cadd C4H:F5H and their rescue by CA and PR. Nevertheless, the observation that the LTR defect in these plants can be rescued with MMPR and DMPR demonstrates that this is not the case. We selected these compounds because the methyl etherification of the phenol prevents incorporation into lignin but might not disrupt their putative activity as signaling molecules. MMPR and DMPR were more potent than PR at rescuing the LTR phenotype of cadd C4H:F5H , and DMPR was more potent than MMPR, suggesting that rescue of LTR development does not come about through catabolism of these compounds to PR. Altogether, these data suggest that the LTR defect of cadd C4H:F5H plants is due to a deficiency in lignan biosynthesis. The Aberrant Root Hair Morphology of cadd \n C4H:F5H Can Be Rescued by Coniferyl Alcohol. Root hair development is a process that requires rapid cell expansion, and it is controlled by multiple factors, including reactive oxygen species (ROS), auxin, ethylene, calcium gradients, actin, and microtubule cytoskeletons. Mutations that affect these processes lead to changes in the quantity and morphology of root hairs and in some cases lead to branched hairs ( 58 ). The importance of phenylpropanoids in root hair development has been demonstrated in flavonol-deficient mutants of tomato and Arabidopsis which have increased ROS levels and elevated numbers of root hairs ( 86 , 87 ). Trichoblasts are nonlignifying epidermal cells that develop into root hairs, and the fact that their morphology can be altered by the presence or absence of CA further suggests it, or a derivative thereof, may have a signaling role in plant development."
} | 2,955 |
39820425 | PMC11739686 | pmc | 8,394 | {
"abstract": "Large yellow croaker ( Larimichthys crocea ) is a highly economically important marine fish species in China. However, substantial individual variations in growth performance have emerged as a limiting factor for the sustainable development of the large yellow croaker industry. Gut microbiota plays a crucial role in fish growth and development by regulating metabolic processes. To explore these dynamics, we employed metagenomics, transcriptomics, and untargeted metabolomics to comprehensively analyze the structure of the intestinal microbiome and its relationship with intestinal metabolism and host gene expression. We constructed association models for “gut microbiota–differentially expressed genes”, “differentially expressed genes–metabolites,” and “gut microbiota–metabolites.” Sequencing data and LC–MS/MS raw data have been deposited in NCBI and MetaboLights databases for public access. Our findings offer critical insights into the molecular mechanisms underlying growth variations in L. crocea and provide valuable data for the selective breeding of improved strains."
} | 271 |
34870984 | PMC8678986 | pmc | 8,395 | {
"abstract": "Soft actuators allowing\nmultifunctional, multishape deformations\nbased on single polymer films or bilayers remain challenging to produce.\nIn this contribution, direct ink writing is used for generating patterned\nactuators, which are in between single- and bilayer films, with multifunctionality\nand a plurality of possible shape changes in a single object. The\nkey is to use the controlled deposition of a light-responsive liquid\ncrystal ink with direct ink writing to partially cover a foil at strategic\nlocations. We found patterned films with 40% coverage of the passive\nsubstrate by an active material outperformed “standard”\nfully covered bilayers. By patterning the film as two stripes, a range\nof motions, including left- and right-handed twisting and bending\nin orthogonal directions, could be controllably induced in the same\nactuator. The partial coverage also left space for applying liquid\ncrystal inks with other functionalities, exemplified by fabricating\na light-responsive green reflective actuator whose reflection can\nbe switched “on” and “off”. The results\npresented here serve as a toolbox for the design and fabrication of\npatterned actuators with dramatically expanded shape deformation and\nfunctionality capabilities.",
"conclusion": "Conclusions We have demonstrated that using DIW to fabricate\nLC-based patterned\nfilms leads to actuators providing alternative, controlled shape deformations\nwith additional functionalities possible. The systematic study performed\nin this work revealed that passive substrates, such as thermoplastics,\ncan be made responsive without fully covering the surface with active\nmaterial or affecting their performance, which reduces the fabrication\ncost as less of the expensive LC material is needed. Additionally,\npatterning offers the possibility of having discrete active regions\nof the foils that can be individually triggered and create localized\nstresses, similar to how tendons control motion in human hands. As\na result, we could achieve reversible, light-driven twisting, both\nright- and left-handed, and bending, parallel and perpendicular to\none of the axes, in the same film. Additionally, partially covering\nthe passive layer also allows a facile combination of different materials\nthat have distinct functionalities. This was demonstrated by fabricating\na photonic light-responsive actuator that could selectively reflect\na specific wavelength on demand. The results discussed in this contribution\nembody the potential of using DIW in fabricating a new class of patterned\nactuators with expanded functionality and shape deformation capabilities.\nNow the mechanism behind the response of the present system is mostly\nunderstood, future research should make use of simulation to predict\noptimal patterns of an active material for fabricating soft actuators\nwith novel functionalities, responsivities, and complex shape-morphing\ncapabilities.",
"introduction": "Introduction Significant progress\ntoward untethered, centimeter-scale soft actuators\ntriggered by external stimuli has been made, 1 − 4 and their appeal in robotics has\nbeen demonstrated. 5 − 9 Among responsive materials, liquid crystal (LC) polymers are considered\nexcellent options for fabricating this class of actuators, 10 − 14 delivering large, rapid, and reversible preprogramed deformations\ntriggered by a variety of stimuli, both in wet and dry environments. 15 , 16 To date, there have been essentially three options for creating\nthese LC-based actuators. In the first option, actuators are made\nfrom single film layers and generally only display single deformation\nmodes depending on the molecular alignment fixed in the network, for\nexample, contraction, 17 bending, 18 or twisting, 19 typically\ntriggered by a single stimulus. The second option is bilayer actuators\nof the active LC material on a passive foil or layer, which can improve\nthe robustness of the device while delivering the same type of motions\nas single-layer films, 20 , 21 although facilitating functional\ncombinations in a single actuator such as light and magnetic responsivity\nor temperature response while having photonic properties. 22 − 25 The third option consists of stacking multiple layers, resulting\nin a 3D object, 6 , 11 , 26 − 28 which can be accomplished via different fabrication\nmethods. 3 This latter approach has the\nadvantage that complex shape deformation can be accomplished that\ncan result in self-propulsion, 10 , 29 for example. However,\nachieving multiple deformation modes within the same actuator remains\nchallenging. This calls for a new design concept that outperforms\nboth single- and bilayer actuators to bring untethered soft actuators\none step closer to widespread use. 30 − 32 Diverse motions\nfrom a single object are achieved in nature mostly\nthrough individual, sequential responses by discrete subunits that\nform the larger object: 33 take, for example,\nthe rotation of human hands, which is accomplished by the contraction/expansion\nof individual tendons. 34 By translating\nthis concept to untethered actuators, novel motions and control could\nbe accomplished. Patterned actuators are fabricated by selectively\ndepositing an\nactive material in specific regions atop a passive substrate. The\npresence of discrete active regions on a passive layer allows the\nuse of less material than having to coat entire substrates, and for\nindividual regions of the device to be independently activated and,\nas in tendons, trigger localized stresses and specific localized motions.\nPatterning has been employed in bilayer films to deliver folding, 35 rolling, 36 and curling, 37 origami-like folding, 38 out-of-plane shape deformations, 39 or\nto induce 2D-to-3D shape changes in films. 40 Additionally, temperature-responsive LC elastomers (LCE) acting\nas a sort of “skeletal muscles” within a passive material\nthat triggers a shape change upon heating have also been reported. 6 , 41 These examples show the potential of patterning, but its use to\nobtain multiple shape deformations within the same object remains\nunexplored. For this goal, employing light as the trigger is appealing\nbecause it can be rapidly, tether-lessly, and locally applied. 42 Here, we report on actuators intermediate\nto single- and bilayer\nfilms, with multifunctionality and a plurality of possible shape changes.\nThese light-responsive-patterned actuators fabricated via direct ink\nwriting (DIW) display several distinct, controlled deformation modes\non demand. To establish the potential of locally depositing actuator\nmaterial atop a passive substrate and to compare its performance to\nconventional fully covered substrates, a single LCE stripe is written\non a polymer substrate and the device responsivities to both temperature\nand light are characterized. Interestingly, large deformations were\ntriggered by both light and temperature but in opposing directions.\nThese findings are used to fabricate actuators with multiple LCE stripes\nwith different reversible and controlled shape deformations upon specific\nillumination of the different light-responsive regions. Additionally,\npartial coverage with an active material provides partial transparency\nto the actuator and space for applying additional layers, for example,\nphotonic reflectors. In the latter case, we present an actuator capable\nof performing two tasks simultaneously: motion and selective reflection\nof different wavelengths of light.",
"discussion": "Results and Discussion Generating\nPatterned Actuators The patterned, light-responsive\nLC actuators were prepared by DIW. The alignment of the LC mesogens,\nkey to the actuator’s eventual performance, is dictated by\nthe DIW procedure ( Figure 1 ). 3 This fabrication approach permits\nthe easy deposition of discrete, light-responsive regions that can\nbe independently triggered. 6 We selected\na thin (10 μm) uniaxially stretched polyetherimide (PEI) foil\nwith nanogrooves as the passive substrate ( Figure S1 ). There is no indication that stretched PEI acts as an alignment\nlayer for oligomeric mesogens: alignment is solely dictated by the\nrelative motion of the printing head to the substrate. PEI has a storage\nmodulus of ca . 3000 MPa at 20 °C ( Figure 2 a) on the same order of magnitude\npoly(ethylene terephthalate) (PET); 21 , 23 it is transparent\nto blue light but not ultraviolet (UV) light. Figure 1 Scheme depicting the\nsynthesis of the LC ink (top) and the fabrication\nprocess of an LC-based patterned actuator via DIW (bottom). The LC\noligomer used to prepare the ink is the result of chain extension\nvia a thiol-acrylate Michael addition of molecules 1–2 using 3 as a spacer. The insets show an idealized molecular\nalignment at each stage. Figure 2 Characterization of the\npatterned film prepared via DIW. (a) Storage\nmodulus as a function of the temperature of the two layers separately.\nThe modulus was investigated parallel (∥) and perpendicular\n(⊥) to the alignment direction of the LC and to the stretching\ndirection of the PEI foil. (b) Schematic representation of a patterned\nfilm in which the different tunable design parameters are defined\n(top). Yellow and light gray represents the LCE and the PEI, respectively\n(bottom). A photograph of a fabricated 5 × 25 mm 2 film with 40% coverage. (c) 3D profile of the edge of the printed\nLCE. (d) Crossed polarized light micrographs of the printed LCE on\nthe foil. The black arrows represent the directions of the polarizer\n(P) and analyzer (A). The white dashed lines indicate the borders\nof the printed lines. The white arrows indicate the direction of printing.\nThe scale bar represents 500 μm. The LC oligomer was synthesized via a base-catalyzed thiol-acrylate\nMichael addition reaction of 1–3 as previously\nreported 44 (Figure 1, the results of the\ndifferent characterizations of the prepared oligomer may be found\nin Figures S2–S4 and Tables S1 and S2 in the Supporting Informtation ). The excess acrylate compared to\nthiol groups resulted in acrylate-terminated oligomers with molecule 2 in the main chain; the azobenzene group is responsible for\ngranting the network light responsivity with its photo-induced trans -to- cis isomerization. 30 The synthesized oligomer had a number-average\nmolecular weight ( M n ) of 7756 g mol –1 based on the 1 H NMR spectrum, with a dispersity\n( Đ ) of 2.4 as determined using gel permeation\nchromatography (GPC), and showed an isotropic-to-nematic phase transition\n(T I/N ) at 80 °C and a nematic-to-smectic C (N-SmC)\nphase transition (T N/SmC ) of 42 °C, as indicated by\nthe differential scanning calorimetry (DSC) traces and crossed polarized\nlight micrographs. The LC ink for the DIW process was prepared by\nmixing the LC oligomer (98 wt %) with a photoinitiator (2 wt % 4 ). The patterning process on PEI foils was first optimized.\nThe influence\nof the printing direction on the mesogenic alignment, either parallel\nor perpendicular to the stretching direction of the PEI, was investigated\n( Figure S5 ) with the reservoir set to 70\n°C and the printing bed to 20 °C. At these temperatures,\nprinting at a speed of 7 mm s –1 with a 335 μm\ndiameter nozzle invariably resulted in uniaxially aligned lines, independent\nof the underlying PEI stretching direction. These results verify that\nthe PEI foil is not acting as an alignment layer, and the mesogenic\nalignment is determined solely by the DIW process, as expected. After printing, the formation of the LCE network was initiated\nusing high-intensity UV light at room temperature under N 2 . Sol/gel fraction experiments revealed an average 85% gel fraction\nfor the LCE. Copolymerization of acrylate groups generally results\nin higher fractional network formations, but the high viscosity and\nthe long oligomeric chains ( ca. 9 units) of this\nsystem reduce the mobility of the radicals and lower the cross-link\ndensity, respectively, resulting in a reduced gel fraction. With the printing parameters optimized, we deposited a single 2\n× 35 mm 2 ( w × l ) LCE stripe on a PEI foil (10 × 10 × 0.001 cm 3 ). After photopolymerization, a laser cutter was employed to extract\na 5 × 25 mm 2 film with the LCE rectangle situated\nalong the center of the PEI ( Figure 2 b). As a result, the PEI film had 40% of its area covered\nby the LCE. We observed that in some cases the films showed a prebend\nwhen suspended. We hypothesize that the prebend is a consequence of\nthe high temperature experienced by the films during laser cutting\nbecause bare laser-cut PEI films having the same dimensions also showed\nprebends: a linear relationship between the prebend and laser intensity\nwas observed ( Figures S6 and S7 ). Furthermore,\ncharacterization of the patterned film revealed that the thickness\nof the deposited LCE averaged to 90 μm, as determined by an\noptical profiling system ( Figure 2 c). Across the 3D profile, the active area shows an\nundulating topography, with each “wave” representing\none of the deposited filaments that constitute the LCE stripe. The\nbirefringence of the active layer was observed between crossed polarizers\n( Figure 2 d): the LCE\nstripe appeared darker when oriented parallel to the polarizer or\nanalyzer than when at 45°. This dark–bright state indicates\na uniaxially oriented LCE: 3 in this case,\nthe mesogenic alignment was parallel to the longitudinal axis. Temperature\nResponse The response of the 5 × 25\nmm 2 , 40% covered, and 90 μm LCE film ( Figure 2 b) to temperature was evaluated\n( Figure 3 a). Upon increasing\nthe temperature from 25 to 110 °C, the film initially bent, and\nthen started to tightly roll up above 80 °C, with the LCE inside\nthe curvature. Such a response is not surprising as uniaxially aligned\nLCEs are known to display large, anisotropic shape deformations ( ca. 50%), 3 , 16 contracting along and expanding\nperpendicular to the alignment direction ( Figure S8 ). The temperature response arises from the increasing disorder\nof the mesogenic groups with increasing temperature as the network\nundergoes phase transitions. Upon heating, the network transitions\nfrom the SmC-to-N phase around 45 °C, and from the N-to-isotropic\nphase around 94 °C ( Figure S9 ), explaining\nthe two stages of the response seen both in Figure 3 a and in Figure S8 . Bare PEI itself does not show any shape changes within the evaluated\ntemperature range ( Figures S8 and S10 ).\nHence, as previously observed for bilayers, when one of the layers\nexperiences a contraction or expansion along the longitudinal axis,\nthe system bends parallel or perpendicular to it, respectively. 45 Thanks to the large contraction of LCEs, the\nfilm’s final configuration is a tight roll. In comparison,\nhighly cross-linked LC network (LCN) bilayer films (LCNs characteristically\nhaving 10% contraction due to their higher crosslink density) 3 , 16 typically only bend up to a full rotation, 18 and generally do not tightly roll up unless they have a wedge geometry. 46 Figure 3 Stimuli response of a patterned actuator (5 × 25\nmm 2 , 40% coverage, and 90 μm of an LCE on PEI, Figure 2 b). (a) Edge-on images\nof the\nfilm at different temperatures. The blue dashed lines represent the\nshape of the film at 25 °C. Scale bar represents 1 mm. (b) Edge-on\nimages displaying the light-driven bending motion of the patterned\nfilm to 365 nm (80 mW cm –2 ) and 455 nm light (145\nmW cm –2 ). (c) Tip displacement as a function of\ntime. Positive values mean that the tip bent away from its starting\nposition, that is, from the light source, and negative when bent toward\nthe light. Light Response Photoactuation of a patterned actuator\ncan be characterized either in air or underwater because the chosen\nLCE performs in either of these media. 44 Systematic studies of the photoactuation in air can be difficult,\nas both photomechanical and photothermal effects influence the actuation\n( Figures S11 and S14 ); 47 consequently, the light response was investigated underwater\nwhere photomechanical effects dominate. The 5 × 25 mm 2 , 40% covered, and 90 μm LCE-patterned film ( Figure 2 b) was suspended in 19 °C\nwater as shown in Figure S11 : at this temperature,\nthe LCE is in its SmC phase. When irradiated with a 365 nm light emitting\ndiode (LED) (80 mW cm –2 ), the film bent away from\nthe light source ( Figure 3 b and Movie S1 ), an expected performance\nfor this LCE. The bending away motion is the result of a photo-induced\nSmC-to-smectic A (SmC-SmA) phase transition and ca. 4% expansion parallel to the alignment direction. 44 A stationary state was attained after 10 min of illumination,\nafter which the light was switched off. Because of the photomechanical\nnature of the response, the deformation was only reversed when illuminated\nwith a 455 nm LED (145 mW cm –2 ), which induced the\nback isomerization from cis to trans ( Figure 3 c). This\nbending motion was reversible for at least 12 cycles with no apparent\nsigns of fatigue ( Figure S15 ). Additionally,\nno delamination of the LCE from the PEI was observed after the 12\ncycles, suggesting good adhesion of the stripe to the passive substrate.\nDespite being only 40% covered with the light-responsive LCE, the\nfilm showed a large, light-driven deformation, with a maximum tip\ndisplacement of 25 mm. A similar performance was observed for a film\nwith the stretching direction of the PEI foil perpendicular to the\nprinting direction rather than parallel ( Figure S15 ). The observed tip displacement is comparable to the displacement\nof a single-layer film made from the same LCE, 44 and to other single- and bilayer films having similar dimensions\nwith 100% surface area coverage, where maximum light-driven deformations\nare ranged from 18 to 26 mm. 5 , 21 , 23 , 43 , 47 This result suggests that large deformation can be obtained without\nfully covering the passive layer with an active material, inspiring\nnew designs of actuators to unlock their full potential. Influence of\nThickness, Shape, and Coverage on the Photoactuation There\nare three tunable parameters in the patterned actuator: the\nthickness of the active LCE ( t LC ), fractional\ncoverage of the passive PEI, and aspect ratio of the film. The effect\nof each variable on the maximum tip displacement was systematically\nstudied, see Tables S3–S5 . The influence of the thickness of the active LCE pattern, t LC , on photoactuation was investigated in a\nseries of films in which the dimensions (5 × 25 mm 2 ) and fractional coverage (40%) of the passive PEI layer were kept\nconstant ( Figure S16 and Table S3 ). Upon\nincreasing t LC from 50 μm to 150\nμm, the maximum tip displacement recorded after 10 min of illumination\nwith 365 nm light increased consistently until it reached a maximum\nvalue at a thickness of 95 μm, at which point it decreased.\nPreliminary simulations are consistent with this experimental data\n( Figure S17 ) and reveal the importance\nof the illumination conditions. When the active layer is illuminated,\ntwo regions are formed through the depth of the LCE: (1) an exposed\nregion at the surface that expands along the longitudinal axis and\n(2) the nonexposed, deeper region that remains inert. The extent of\nthe expanding region depends on several factors (including illumination\ntime, light intensity, and azobenzene concentration). In thin LCE\nlayers, the entire depth of the LCE stripe is exposed to the light,\ncausing for its entirety to expand. A further increase of t LC results in unexposed regions within the LCE\nthat increase the effective bending stiffness of the actuator, reducing\nthe maximum tip displacement, which is inversely correlated to the\npassive layer thickness cubed ( t 3 ). 43 The resulting optimal thickness value for the\nLCE for maximum tip displacement is, thus, determined by our specific\nillumination conditions. We investigated the effect of varying\nfractional coverage in the\nphotoactuation in a series of films with identical dimensions (5 ×\n25 mm 2 ) and deposited LCE thicknesses (90 μm) ( Figure 4 a and Table S4 ). By increasing coverage from 0 to 100%,\nthe maximum tip displacement plateaued at ca . 40%\ncoverage. The comparative rate of actuation during the initial 5 s\nof light exposure was roughly linear until 40% coverage where it peaks\nand roughly plateauing at greater coverages. Preliminary simulations\nwere generally consistent with the experimental data and help in understanding\nthe observed tendencies ( Figure S17 ). It\nappears that a minimum coverage is required to initiate rapid, extensive\nactuation: the degree of coverage required will be affected by the\nrelative physical characteristics of the passive and active layers\nand the illumination conditions. When increasing the fractional coverage\nof the LCE, the width of the stripe increases, resulting in increasingly\nsignificant perpendicular contraction and the generation of orthogonal\nstresses that oppose tip displacement, contributing to the plateau\nformation for displacement and decay in the absolute velocity. The\northogonal stresses that oppose tip displacement arise from the illuminated\nregion as it expands parallel to the alignment ( ca. 4%), 44 but it also contracts perpendicular\nto it ( ca. 2%, see Supporting Informtation for the calculations). Thus, the ca. 40% covered soft actuator outperforms 100% covered actuators with\nrespect to speed, while the maximum tip displacement is about the\nsame; this is an important finding to consider in future designs of\npatterned actuators, as it suggests less material is actually better. Figure 4 Maximum\ntip displacement and average velocity after 5 s of illumination\nas a function of (a) fractional LCE coverage and (b) PEI aspect ratio\nof the films with 40% LCE coverage, respectively. In both cases the\nillumination time with a 365 nm (80 mW cm –2 ) LED\nwas 10 min. Above the plots, schematics represent some of the films\nused for the studies. Yellow indicates the LCE and light gray represents\nthe PEI. In Tables S4 and S5 from the Supporting Informtation the reader can find the different dimensions,\nthickness, and prebends of the films used for this optimization study. Finally, the effect of the patterned actuator aspect\nratio on photoactuation\nwas evaluated. The widths ( w ) of the films were increased\nwhile maintaining 25 mm length, fractional coverage (40%), and thickness\n(90 μm) ( Figure 4 b and Table S5 ). By increasing the aspect\nratio from 1.3 to 5, the maximum tip displacement and the average\nspeed for the first 5 s of illumination increased linearly. When increasing\nthe aspect ratio, the width of the LCE layer ( w LCE ) is proportionally reduced. As a result, the ca . 2% contraction that occurs along w LCE that induces the bending opposing the tip displacement is minimized,\nbut the expansion ( ca. 4%) along the longitudinal\naxis responsible for the bending is maintained, as the length is not\naltered. Thus, increasing the aspect ratio results in larger and faster\nactuations as the forces opposing bending are reduced. Patterned Actuators\nwith Various Shape-Morphing Capabilities To fully explore\nthe potential of patterning, a 20 × 25 mm 2 film topped\nby two 4 × 25 mm 2 stripes of\nthe LCE separated by 4 mm was fabricated ( Figure 5 a and Movie S2 ). The film has 40% of the PEI foil covered by the LCE and an aspect\nratio of 1.3. Each of the two active layers can be individually triggered\nthanks to the space between them. Triggering one or the other of the\nLCE stripes by continual exposure up and down the stripe results in\na twisting motion either counter-clockwise or clockwise, depending\non which stripe is activated. Thus, by the simple printing of two\nstripes, additional motions can be induced in the film, also predicted\nby simulation ( Figure S18 ). Interestingly,\nwhen both stripes are irradiated simultaneously, the film bends, with\nthe bend significantly greater than for a film having an identical\nnet aspect ratio and total coverage fraction (40%), but with the active\nregion concentrated as a single stripe in the center ( Figure S19 and Movies S3 and S4 ). Such performance enhancements\nsuggest that the active layer’s distribution should also be\nconsidered when designing patterned actuators. We hypothesize that\nthis increased amplitude response is the result of reducing the aspect\nratio of the active layer, minimizing the counteracting forces from\nthe perpendicular expansion of the LCE stripe. Figure 5 Underwater performance\nof an actuator presenting multiple shape\nchanges. (a) On the left, a schematic drawing of the actuator: yellow\nindicates the LCE and light gray represents the PEI. The thickness\nof the LCE was found to be 90 μm. On the right, edge-on photographs\nof the actuator showing the different deformations accomplished when\nilluminated, partially or completely, with a 365 nm (80 mW cm –2 ) LED. The insets show which region of the actuator\nis illuminated in each case; the active part has been patterned with\npurple lines and colored in orange. (b) Series of images showing the\ndeformations that the actuator from (a) undergoes when no constraints\nare applied and light (from the top) and temperature are used to trigger\ndifferent stresses and induce phase changes in the active layer that\nresult in bending in opposite directions. Detailed snapshots of the\nshape changes from 19 to 50 °C can be found in Figure S20 . For both (a,b), snapshots of the simulated actuation\ncan be seen in Figure S18 . The light response of the LCE used here has a unique temperature\nsensitivity which controls the directionality of the light-driven\nbending motion, 44 as demonstrated when\nactuating the films in air at different light intensities ( Figures S13 and S14 ). Basically, when the network\nis illuminated in the SmC mesophase, it bends away from the light,\nwhile when illuminated in the N mesophase, it bends toward the light\nsource. We thus explored this property of the LCE to induce additional\nshape changes in patterned actuators (2 × 25 mm 2 ,\n40% coverage, and 90 μm of LCE, Figure 5 a) by combining light and temperature ( Figure 5 b). Rather than being\nsuspended, the film was left at 19 °C underwater, lying on the\nfloor of the container with the LCE side facing upward. Upon light\nillumination from the top at 19 °C, the film bent away from the\nlight source and eventually ended up standing. The light was switched\noff, and the temperature gradually increased from 19 to 28 °C\n( Figure S20 ). Initially, the film started\nto unbend toward its initial flat state, which was expected as temperature\ninduces a phase transition in the polymer network from Sm-to-N, resulting\nin a gradual contraction of the illuminated LCE 44 parallel to an opposing anisotropic shape change induced\nby temperature in the unexposed region of the network ( Figure S8 ). 44 However,\nfrom 30 to 50 °C, the film started to bend perpendicular to the\nlongitudinal axis, increasing its curvature with temperature. Illumination\nwith blue light had no effect on this deformation. Sudden removal\nfrom the water and placement on a 20 °C surface resulted in recovery\nof its initial bent state along the longitudinal axis. Placed again\nin water at 50 °C, the film immediately switched its bending\nto opposite of the longitudinal axis instead of along it. Such orthogonal\nbending motion induced by a combination of light and temperature was\nalso predicted by the simulations which supports the proposed mechanism\nbehind this atypical shape change ( Figure S18 ). This atypical perpendicular bending is the result of the temperature-induced\nLCE expansion occurring opposite to the alignment direction ( Figure S8 ), when the LCE layer is spread as two\nstripes, as opposed to having it concentrated in a single stripe as\nin Figure 3 , which\nresults in such a unique motion. Thus, two stripes plus a combination\nof light and temperature led to a film that reversibly switches from\nflat to bent along the longitudinal axis to bending opposite to it,\nmaking this the first actuator showing such shape-morphing capability. We also fabricated larger patterned actuators with alternative\ndeformations modes. A 25 × 36 mm 2 , 33% covered, and\n90 μm LCE film with four separate LCE stripes (one pair of 4\n× 25 mm and one pair of 2 × 25 mm) could be sequentially\ntwisted by gradually activating the stripes one by one, from bottom\nto top ( Figure S21 and Movie S5 ). Another 20 × 25 mm 2 , 16% covered,\nand 90 μm LCE film with two asymmetrical LCE lines (one stripe\nof 1.7 × 25 mm and one of 1.5 × 25 mm) resulted in different\ndegrees of twist ( Figure S22 and Movie S6 ). These examples embody the benefits\nof distributing the active material in discrete (symmetrical or asymmetrical)\nregions over the passive layer surface as it allows distinct, controlled,\nand reversible deformations modes. Light-Responsive Photonic\nActuator The dual-stripe\nactuator presented in the previous section (20 × 25 mm 2 , 40% coverage, and 90 μm of LCE, Figure 5 a) left a significant fraction of the PEI\nfoil uncovered. This construction allows for deposition of an additional\nfunctional ink in the blank regions. We opted to print a single 3\n× 24 mm 2 stripe cholesteric LC (CLC) ink, a photonic\npolymer network with a helical arrangement of LC molecules, 48 between the previously deposited light-responsive\nLCE stripes ( Figure 6 a). Optical and structural characterization of the 100 μm thick\nreflective CLC layer shows a broad reflection band centered at 564\nnm ( Figure 6 b,c). The\nedges of the layer are less defined than those observed in Figure 2 c, as the CLC ink\nhas a lower viscosity than the LC ink, so after extrusion it shows\nincreased spreading before polymerization. The difference in viscosity\nbetween the inks also explains the dewetting observed at the extremes\nof the reflective stripe. The film still displays large bending deformations\ntriggered by light despite the addition of the extra CLC layer ( Figure S22 ). The potential use as a dynamic reflector\nwas investigated ( Figure 6 d). The light-responsive-actuating reflector was suspended\nunderwater and the CLC stripe exposed to light from green (λ\n= 532 nm) or red (λ = 633 nm) lasers. At rest, red light was\ntransmitted while green light was partially reflected by the film,\nas expected ( Figure 6 b). 365 nm light was then used to trigger the responsive LCE in the\nstripes, causing film bending. As a result, the central reflective\nregion was removed from the path of both lasers, permitting a light\ncontrolled on-and-off transmission of the green laser light through\nthe water. This patterned actuator, having two stripes of a light-responsive\nmaterial and one of a photonic elastomer, shows the benefit of patterning\ndifferent materials to obtain multiple functionalities while having\nnumerous shape-morphing capabilities in the same film. Figure 6 Light-responsive photonic\nactuator. (a) Front photograph of the\nlight-responsive photonic actuator. The inset shows the molecular\nalignment of the CLCE line, which consist of a helix structure. Thanks\nto the periodicity of the full rotation of the helix, around 376 nm,\nthe middle line acts as the reflector for green light. (b) Fraction\nof reflected light as a function of the incident wavelength on the\nCLCE. (c) 3D profile of the edge of the CLCE and the PEI. (d) Series\nof edge-on photographs in which the selectivity to reflect a specific\nwavelength of the actuator and light-driven deformation are shown.\nThe images show the actuator underwater. The 365 nm (80 mW cm –2 ), 455 nm (145 mW cm –2 ), and laser\nlights were all incident from the left."
} | 7,870 |
39101034 | PMC11294248 | pmc | 8,398 | {
"abstract": "Sulfur oxidizing bacteria (SOB) play a key role in sulfur cycling in mine tailings impoundment (TI) waters, where sulfur concentrations are typically high. However, our understanding of SOB sulfur cycling via potential S oxidation pathways ( sox , r dsr , and S 4 I) in these globally ubiquitous contexts, remains limited. Here, we identified TI water column SOB community composition, metagenomics derived metabolic repertoires, physicochemistry, and aqueous sulfur concentration and speciation in four Canadian base metal mine, circumneutral-alkaline TIs over four years (2016 – 2019). Identification and examination of genomes from nine SOB genera occurring in these TI waters revealed two pH partitioned, metabolically distinct groups, which differentially influenced acid generation and sulfur speciation. Complete sox (c sox ) dominant SOB (e.g., Halothiobacillus spp., Thiomonas spp.) drove acidity generation and S 2 O 3 2- consumption via the c sox pathway at lower pH (pH ~5 to ~6.5). At circumneutral pH conditions (pH ~6.5 to ~8.5), the presence of non-c sox dominant SOB (hosting the incomplete sox , r dsr , and/or other S oxidation reactions; e.g. Thiobacillus spp., Sulfuriferula spp.) were associated with higher [S 2 O 3 2- ] and limited acidity generation. The S 4 I pathway part 1 ( tsdA ; S 2 O 3 2- to S 4 O 6 2- ), was not constrained by pH, while S4I pathway part 2 (S 4 O 6 2- disproportionation via tetH ) was limited to Thiobacillus spp. and thus circumneutral pH values. Comparative analysis of low, natural (e.g., hydrothermal vents and sulfur hot springs) and high (e.g., Zn, Cu, Pb/Zn, and Ni tailings) sulfur systems literature data with these TI results, reveals a distinct TI SOB mining microbiome, characterized by elevated abundances of c sox dominant SOB, likely sustained by continuous replenishment of sulfur species through tailings or mining impacted water additions. Our results indicate that under the primarily oxic conditions in these systems, S 2 O 3 2- availability plays a key role in determining the dominant sulfur oxidation pathways and associated geochemical and physicochemical outcomes, highlighting the potential for biological management of mining impacted waters via pH and [S 2 O 3 2- ] manipulation.",
"conclusion": "4 Conclusion Results here reveal new insights into the interplay between pH, SOB community composition and function (c sox dominant vs. non-c sox dominant), S cycling and acidity generation within primarily oxic mine TI waters over seasonal, annual, and mine operation scales. Our results highlight two key pH dependent niches characterized by the presence of either (i) c sox dominant SOB (e.g., Thiomonas spp., Halothiobacillus spp.) associated with lower pH values and lower [S 2 O 3 2- ] or (ii) non-c sox dominant SOB (i sox and/or r dsr pathways; e.g. Thiobacillus spp., Sulfuriferula spp.) associated with more circumneutral pH conditions and higher [S 2 O 3 2- ] ( Figure 5 ). The presence of the first part of the S 4 I pathway ( tsdA ; S 2 O 3 2- to S 4 O 6 2- ) was ubiquitous across pH and [S 2 O 3 2- ] niches; while possible subsequent processing of S 4 O 6 2- via tetH which can lead to S 0 , and S 2 O 3 2- regeneration, as well as SO 4 2- , was limited to Thiobacillus spp., observed more prevalently in circumneutral pH values ( Figure 5 ). Extrapolation of these results to broader environments via comparative analysis using M&A and environmental literature data revealed a specialized mining SOB microbiome characterized by elevated abundances of c sox dominant SOB. Elevated SOI concentrations, typical of mining environments, support the establishment and sustainment of c sox dominant communities unique to these environments. Our comprehensive study into SOB community dynamics, sulfur oxidation pathways, and the influence of geochemical and physicochemical factors in mining impacted waters, highlight the importance of thiosulfate availability, and pH constraints on sulfur oxidizing metabolism under oxic conditions, prevalent in TI contexts. This study highlights opportunities to manipulate TI SOB communities through pH adjustment and/or [S 2 O 3 2- ] management, offering potential avenues to reduce the risk of SOI discharge into receiving waters.",
"introduction": "1 Introduction Biological sulfur oxidation can present significant risks to the environment via acid generation, contaminant mobilization, and oxygen consumption (i.e., acid mine drainage) ( Edwards et al., 1999 ; Elberling and Damgaard, 2001 ; Baker and Banfield, 2003 ; Druschel et al., 2003 ; Johnson and Hallberg, 2003 ; Korehi et al., 2014 ). This biological acidity production occurs in both natural (termed acid rock drainage, found in alpine catchments; Lacelle et al., 2007 ; Zarroca et al., 2021 ) and anthropogenic (e.g., mine tailings impoundments and waste rock piles; Akcil and Koldas, 2006 ) environments, though the scale to which it occurs in anthropogenic environments is typically much larger. Base metal mine tailings impoundment (TI) wastewaters often have elevated sulfur concentrations due to the dominance of sulfide minerals in base metal ores [e.g., chalcopyrite (CuFeS 2 ), sphalerite ((Zn, Fe)S), etc.]. Sulfides can be partially oxidized during the grinding, flotation, and leaching steps of sulfide mineral ore extraction ( Liljeqvist et al., 2011 ), resulting in the production and subsequent release of sulfur oxidation intermediate compounds (SOI) from tailings streams to TIs. In addition to sulfide (ΣH 2 S), SOI commonly found in mining impacted waters include thiosulfate (S 2 O 3 2- ), tetrathionate (S 4 O 6 2- ), elemental sulfur (S 0 ), sulfite (SO 3 2- ), and a range of other polythionates (S x O y 2- ) ( Makhija and Hitchen, 1979 ; Miranda-Trevino et al., 2013 ; Whaley-Martin et al., 2020 ). The concentrations of individual SOI in TI wastewater can vary significantly spatially, seasonally, and amongst mining operations. SOI may be present in relatively high concentrations in tailings slurries, but are typically much lower in the TI waters due to dilution by other water inputs collected in these actively managed systems ( Thamdrup et al., 1994 ), though they are typically higher than concentrations observed in natural environments ( Foucher et al., 2001 ; Canfield and Farquhar, 2009 ; Silva et al., 2012 ; Camacho et al., 2020b ; Vincent et al., 2021 ). SOI can be reduced, oxidized, and disproportionated, resulting in differential SOI speciation and pH outcomes, via both abiotic and biotic reactions ( Philippot et al., 2007 ; Klatt and Polerecky, 2015 ; Houghton et al., 2016 ) further contributing to the complexity of the sulfur cycle in these environments. As mine TI systems continue to grow in number and size around the world, an understanding of the biogeochemical cycling of sulfur compounds occurring within these contexts, the SOB involved, and the influencing factors determining water quality outcomes is increasingly important. Few studies have addressed the coupling of sulfur oxidation metabolic pathways in TIs to SOB taxonomy, physicochemistry, and sulfur geochemistry ( Whaley-Martin et al., 2019 , 2023 ; Miettinen et al., 2021 ), contrasting more well-studied extremophilic acid mine drainage environments ( Bond et al., 2000 ; Johnson and Hallberg, 2003 ; Dold, 2014 ). Recent research has highlighted a divergence of microbial communities found in circumneutral mining impacted TI waters from those of acid mine drainage environments ( Whaley-Martin et al., 2019 , 2023 ; Camacho et al., 2020b ; Lopes et al., 2020 ; Miettinen et al., 2021 ). A number of sulfur oxidation pathways have been identified as being used by SOB potentially found in mining environments, though the physicochemical, geochemical, and/or ecological parameters governing which S oxidation pathway(s) occur are not well defined ( Han and Perner, 2015 ; Wang et al., 2016 ; Watanabe et al., 2019 ). These pathways include the sulfur oxidation ( sox ), reverse dissimilatory sulfite reductase (r dsr ), and tetrathionate intermediate (S 4 I) pathways ( Friedrich et al., 2001 ; Han and Perner, 2015 ; Klatt and Polerecky, 2015 ; Wang et al., 2016 ; Watanabe et al., 2019 ; Whaley-Martin et al., 2023 ). The sox pathway has seven structural genes which encode four proteins (soxXA, soxYZ, soxB, and soxCD) allowing this pathway to mediate S 2 O 3 2- , SO 3 2- , S 0 , and ΣH 2 S dependent cytochrome c reduction ( Friedrich et al., 2001 ). SoxAX catalyzes the attachment of S 2 O 3 2- to cysteine residue on the soxY of the soxYZ complex (forming soxZY-Cys-S - ) ( Friedrich et al., 2000 , 2001 ; Bamford et al., 2002 ; Watanabe et al., 2019 ). The soxCD complex then oxidizes the sulfate sulfur from SoxZY-Cys-S - to produce the sulfonate group as soxZY-Cys-SO 3 - ( Quentmeier et al., 2000 ; Zander et al., 2011 ; Watanabe et al., 2019 ). SoxB further hydrolyzes the sulfane sulfur from SoxZY-Cys-SO 3 - to a free sulfate ion ( Quentmeier et al., 2000 ; Zander et al., 2011 ; Watanabe et al., 2019 ). Free SOI are not produced by the sox pathway when found in its complete form (complete sox ; c sox ) as they are covalently attached to soxYZ until complete oxidation to sulfate (SO 4 2- ; Grabarczyk and Berks, 2017 ). An incomplete form of the sox pathway (lacking soxCD, i sox ) has also been identified which can form S 0 ( Frigaard and Dahl, 2008 ; Watanabe et al., 2019 ). This S 0 can be subsequently oxidized by r dsr, sulfur-oxidizing heterodisulfide reductase-like (s hdr ), or sulfur dioxygenase ( sdo ) to SO 3 2- ( Klatt and Polerecky, 2015 ). The r dsr pathway is composed of the same proteins as the dsr pathway ( sat , aprAB , dsrAB ) though the reductive and oxidative varieties are phylogenetically discernible ( van Vliet et al., 2021 ). SOI including S 2 O 3 2- , S 0 , and ΣH 2 S, can be oxidized through the r dsr pathway and generate free SO 3 2- ( Klatt and Polerecky, 2015 ). Sulfane sulfur generated from the i sox pathway can then be transported to the cytoplasm in the form of persulfides (R-S-S - ) to be further oxidized to HSO 3 - by dsrAB or s hdr ( Pott and Dahl, 1998 ; Frigaard and Dahl, 2008 ; Cao et al., 2018 ; Koch and Dahl, 2018 ). The HSO 3 - can be further oxidized to APS (adenosine 5’-phosphosulfate) by aprBA with electron transfer by aprM or hdrAACB and then to SO 4 2- by sat ( Meyer and Kuever, 2007 ; Loy et al., 2009 ). R dsr pathway presence is typically associated with high energy efficiency compared to the sox pathway ( Klatt and Polerecky, 2015 ). Several beta- and gammaproteobacteria have been found to utilize the S 4 I pathway (or Kelly-Trudinger pathway) which generates free S 4 O 6 2- ( Dam et al., 2007 ). The S 4 I pathway oxidizes S 2 O 3 2- to S 4 O 6 2- by thiosulfate dehydrogenase ( tsdA ) or thiosulfate:quinol oxioreductase ( TQO or doxD ; Brito et al., 2015 ; Wang et al., 2016 ; Hutt et al., 2017 ). Subsequent processing of free S 4 O 6 2- can be catalyzed by tetrathionate reductase ( ttrABC ) to produce S 2 O 3 2- or tetrathionate hydrolase ( tetH ) to produce S 2 O 3 2- , S 0 , and SO 4 2- , which are both common in SOB ( Wang et al., 2016 ; Camacho et al., 2020a ; Miettinen et al., 2021 ). A recent study ( Whaley-Martin et al., 2023 ) identified oxygen as a control on whether the c sox (high O 2 ) or r dsr (low O 2 ) pathway dominated in one TI with differing water quality outcomes, indicating that physicochemical and/or SOI substrate partitioning of SOB within TI may happen more broadly. The objectives of this cross-mine study were to identify mining TI associated SOB, examine functional differences in SOB communities, and align their associated S oxidizing repertoires to geochemical and physicochemical characteristics and outcomes in circumneutral TI waters of four base metal mines located across Canada (Manitoba, Newfoundland, Ontario). A better understanding of these genetic, geochemical, and/or physicochemical connections will inform biological management strategies as well as further understanding of S biogeochemical cycling more broadly. Four years (2016 – 2019) of S geochemistry, physicochemistry, and genus level community structure and function data from the four base metal mine TIs were examined. To determine if the findings of this study were site specific or reflective of broader environmental trends, comparisons were made to published studies on other mines and industrial environments, as well as natural environments.",
"discussion": "3 Results and discussion 3.1 SOB community 16S rRNA composition and metagenomic inferred function Microbial community composition was determined using 16S rRNA relative abundances for each TI sample ( n = 13 from Mine 1, n = 15 from Mine 2, n = 6 from Mine 3, and n = 8 from Mine 4). Nine major genera of SOB were identified in these 42 TI samples including: Halothiobacillus spp. (12.7 ± 20.5%), Sediminibacterium spp. (4.8 ± 9.1%), Thiobacillus spp. (3.9 ± 7.8%), Sulfuricurvum spp. (3.7 ± 11.5%), Thiovirga spp. (1.5 ± 3.2%), Sulfuritalea spp. (0.7 ± 2.9%), Sulfurimonas spp. (0.4 ± 1.3%), Sulfuriferula spp. (0.4 ± 0.7%), and Thiomonas spp. (0.1 ± 0.2%). Each genus occurred at >1% abundance in at least one sample ( Table 2 ). The presence and relative abundance of these nine SOB genera differed among mines and over time. Table 2 Average (± standard deviation), minimum and maximum 16S rRNA percent relative abundances across sulfur oxidizing bacteria genera data for Mine 1, Mine 2, Mine 3, and Mine 4 tailings impoundment waters (2016 – 2019). Mine 1 Mine 2 Mine 3 Mine 4 Thiomonas spp. Abundance (%) 0.0 – 0.7 (avg. = 0.1 ± 0.2) 0.0 – 1.2 (avg. = 0.1 ± 0.3) 0.0 – 0.01 (avg. = 0.002 ± 0.01) 0 Halothiobacillus spp. Abundance (%) 0.0 – 24 (avg. = 2.3 ± 6.3) 0.2 – 64 (avg. = 34 ± 21) 0.01 – 0.1 (avg. = 0.05 ± 0.04) 0.0 – 0.1 (avg. = 0.02 ± 0.02) Thiovirga spp. Abundance (%) 0.0 – 6.1 (avg. = 1.6 ± 2.2) 0.0 – 17 (avg. = 1.9 ± 4.1) 0.0 – 10 (avg. = 1.8 ± 3.8) 0.0 – 1.1 (avg. = 0.2 ± 0.4) Sulfuricurvum spp. abundance (%) 0.0 – 8.2 (avg. = 1.4 ± 2.9) 0.01 – 55 (avg. = 9.2 ± 18) 0.002 – 1.0 (avg. = 0.2 ± 0.4) 0.0 – 0.1 (avg. = 0.02 ± 0.04) Sulfurimonas spp. abundance (%) 0.0 – 8.0 (avg. = 1.2 ± 2.2) 0.0 – 0.2 (avg. = 0.1 ± 0.1) 0.0 – 0.1 (avg. = 0.01 ± 0.02) 0.0 – 0.02 (avg. = 0.003 ± 0.01) Thiobacillus spp. abundance (%) 0.0 – 39 (avg. = 7.0 ± 12) 0.0 – 14 (avg. = 1.1 ± 3.4) 0.1 – 12 (avg. = 5.1 ± 4.8) 0.2 – 12 (avg. = 3.0 ± 3.6) Sediminibacterium spp. abundance (%) 0.0 – 27 (avg. = 5.3 ± 9.0) 0.0 – 48 (avg. = 6.5 ± 12) 0.1 – 12 (avg. = 3.5 ± 3.9) 0.0 – 13 (avg. = 1.7 ± 4.3) Sulfuriferula spp. abundance (%) 0.0 – 0.9 (avg. = 0.1 ± 0.2) 0.04 – 3.1 (avg. = 0.6 ± 0.8) 0.0 – 0.5 (avg. – 0.1 ± 0.2) 0.0 – 2.4 (avg. = 0.7 ± 0.9) Sulfuritalea spp. abundance (%) 0.0 – 0.1 (avg. = 0.01 ± 0.02) 0 0.0 – 18 (avg. = 3.2 ± 6.6) 0.0 – 5.2 (avg. = 1.2 ± 2.0) Relative abundance of SOB (%) 0.0 – 65 (avg. = 19 ± 23) 12 – 77 (avg. = 53 ± 19) 1.6 – 23 (avg. = 14 ± 7.5) 1.0 – 19 (avg. = 6.7 ± 6.4) To identify the potential sulfur metabolizing pathways encoded by these SOB genera, 116 metagenomes were constructed from 38 samples collected at these four mines between 2016 and 2018 ( Supplementary Table 2 ). Genomes for Sulfurimonas and Sulfuritalea could not be reconstructed and thus interpretations for these two genera relied on literature reports. Three major pathways including sox , r dsr , and S 4 I were examined, as well as additional sulfur oxidation genes not specific to those three pathways ( Figure 1 ), based on the foundational understanding outlined by Watanabe et al. (2019) and expanded by Whaley-Martin et al. (2023) . Genes encoding the sox pathway, either complete (c sox ; soxXYZABCD ) or incomplete (i sox ; soxXYZAB and lacking soxCD ) were most common – occurring in eight of the nine genera. Thiomonas , Halothiobacillus, and Thiovirga genomes encoded the c sox pathway (resulting in generation of SO 4 2- ; Figure 1 ) which is consistent with published reports for Thiomonas and Halothiobacillus ( Veith et al., 2012 ; Lin et al., 2015 ). Outside of other published works from the mines included in this study, there are very limited data available in the literature for Thiovirga spp., Miettinen et al. (2021) identified possible Thiovirga spp. in zinc and copper ore processing facilities in Portugal that lacked only the soxC subunit, indicating there may be variability within the Thiovirga genus regarding the ability for complete oxidation to SO 4 2- via the sox pathway, unlike the findings presented here ( Figure 1 ). Sulfurimonas spp. have been reported to often possess the soxCD genes (i.e., c sox ) ( Lahme et al., 2020 ; Wang et al., 2021 , 2023 ) however, this genus may lack soxAB genes instead, which would not allow for complete oxidation via the sox pathway ( Lahme et al., 2020 ; Wang et al., 2021 ). At least one strain of Sulfurimonas (Strain NW10 T ) has been reported to possess genes for the c sox pathway but used the i sox pathway instead ( Wang et al., 2021 ). Figure 1 Metabolic potentials and pathways for sulfur oxidizing bacteria genera from the four TIs investigated here. Figure was adapted from Whaley-Martin et al. (2023) and based on Watanabe et al. (2019) . Genomes were reconstructed from TI samples collected from the four target mines between 2016 and 2018. Due to low abundance of some organisms, variable quantities of genomes were used including 3 Thiomonas , 31 Halothiobacillus , 3 Thiovirga , 31 Thiobacillus , 30 Sediminibacterium , 4 Sulfuricurvum , and 14 Sulfuriferula . Sediminibacterium was the only detected SOB in this study that did not encode any sox genes ( Figure 1 ). Two genera were identified as possessing the i sox pathway; Sulfuriferula and Thiobacillus ( Figure 1 ). Watanabe et al. (2019) reported Sulfuritalea hydrogenivorans sk43H, which at present is the only published pure culture strain, as possessing soxAX , soxYZ , and soxB but lacking soxCD, thus indicating the presence of the i sox pathway. Three of the reconstructed Sulfuricurvum spp. genomes encoded c sox, while one encoded no sox genes ( Figure 1 ), with the latter potentially due to the low quality of the genome. This finding of c sox encoding Sulfuricurvum from mining TI samples described here, and also reported in Whaley-Martin et al. (2023) , diverges from the current literature which typically reports Sulfuricurvum spp. found in the natural environment (terrestrial aquifer, geothermal springs) as possessing the i sox pathway (lacking soxCD ; Handley et al., 2014 ; Meziti et al., 2021 ). A S 4 I pathway gene for the first reaction generating S 4 O 6 2- , was the second most abundant identified across the seven genera with reconstructed genomes ( Figure 1 ). Two distinct catalysts, tsdA and doxDA, are responsible for the conversion of S 2 O 3 2- to S 4 O 6 2- ( Nguyen et al., 2022 ). However, doxDA was absent in the reconstructed genomes, while tsdA was present in the genomes of five of the seven genera, including >80% of Thiomonas , Halothiobacillus , and Sediminibacterium genomes and < 80% of Thiobacillus and Sulfuriferula genomes ( Figure 1 ). Generation of tetrathionate from tsdA activity can be subsequently disproportionated via tetH to form S 0 , S 2 O 3 2- , and SO 4 2- , which can then potentially feed both the incomplete/complete sox pathway and/or S 0 storage pathways (including sdo , SOR, and r dsr ; Friedrich et al., 2005 ; Frigaard and Dahl, 2008 ; Watanabe et al., 2019 ). Though tsdA was found in five genomes, tetH, which would be required for the completion of the S 4 I pathway, was only found in five of the thirty-one Thiobacillus genomes (<80%; Figure 1 ). This may indicate the potential for genera with only tsdA (such as Thiomonas , Halothiobacillus , Sediminibacterium and Sulfuriferula ) to couple with tetH encoding Thiobacillus to complete the S 4 I pathway. Limited tetH detection may also reflect current limitations in tetH reference material. TtrABC can catalyze the reduction of S 4 O 6 2- to produce S 2 O 3 2- though its components ( ttrA, ttrB , and ttrC ) were only variably present across Thiobacillus genomes ( Figure 1 ). Thiomonas spp., Sulfuricurvum spp., and Halothiobacillus spp. were found to possess ttrA, ttrB , and ttrC respectively, but there is no available research indicating these subunits are active in sulfur reduction individually. Watanabe et al. (2019) reported that doxDA , tsdA , and tetH genes were absent in Sulfuritalea hydrogenivorans sk43H indicating this Sulfuritalea species did not encode the S 4 I pathway. Currently, there is a lack of specific data regarding the presence or activity of the S 4 I pathway (and associated genes tsdA , doxDA , and tetH ) within Sulfurimonas . Consequently, it remains uncertain in results here and in available literature, whether Sulfurimonas possess the genomic capacity to support the S 4 I pathway. The genes for the complete rdsr pathway were detected in 29 of the 31 Thiobacillus genomes ( Figure 1 ). Sulfuriferula possessed the potential to express shdr (SO 3 2- production) and sat (SO 4 2- production through APS) but lacked the intermediate aprM/aprBA- I or aprBA-II/hdrAACB genes that mediate the reaction produce APS ( Figure 1 ), aligning with results presented in Watanabe et al. (2019) for Sulfuriferula sp. AH1 and Sulfuriferula thiophila mst6. These missing steps could be compensated for by the cytoplasmically oriented soeABC that was present in all Sulfuriferula genomes, which generates SO 4 2- directly from SO 3 2- ( Koch and Dahl, 2018 ; Figure 1 ). Alternatively, SO 3 2- can be transported to the periplasm via TauE -like transporter where it can spontaneously react with hydrogen sulfide if present ( Koch and Dahl, 2018 ). The lack of rdsr and shdr pathway genes in the SOB identified in this study, may reflect the relatively high dissolved oxygen concentrations across these water samples (averaging between 36.2 – 82.6 % saturation across the four mines; Table 1 ) which do not favour establishment of anaerobic and/or microaerophilic SOB which typically harbour the rdsr pathway ( Klatt and Polerecky, 2015 ). Isolated genomes of Thiobacillus contained aprM , aprBA-I , aprBA-II , and/or hdrAACB, which would allow them to further oxidize SO 3 2- to APS ( Figure 1 ). Watanabe et al. (2019) reported the presence of partial (lacking aprM and aprBA-I ) rdsr pathway genes (including dsrAB , dsrEFH , dsrC, dsrMKJOP, and sat ) in the model species Sulfuritalea hydrogenivorans sk43H . Furthermore, evidence reported by Purcell et al. (2014) indicates Sulfuritalea hydrogenivorans sk43H can carry out the rdsr pathway within Antarctic lake sediments. Sulfuritalea hydrogenivorans sk43H also hosts a partial s hdr pathway including aprBA-II and hdrAACB ( Watanabe et al., 2019 ). Second only to the Thiobacillus genomes presented here, Sulfuritalea hydrogenivorans sk43H hosts the most diverse set of r dsr pathway genes of the nine SOB genera identified. While there are several published Sulfurimonas genomes, detailed insights into r dsr pathway occurrence remains limited. This may be due to the absence of the rdsr pathway within the Sulfurimonas genus as Wang et al. (2023) reports Sulfurimonas sp. ST-27 lacks dsrAB , dsrE , dsrCD , dsrMKJOP, and aprBA, but does contain sat . The rdsr and shdr pathways may play an important role in bacteria harbouring the i sox pathway by facilitating the conversion of sulfane sulfur, produced as an intermediate, into SO 4 2- ( Xin et al., 2023 ). SOB that encode the i sox pathway without the r dsr or s hdr pathways, may produce and accumulate sulfane sulfur as a by-product of S 2 O 3 2- oxidation, though the mechanism for reducing oxidative stress from the accumulation of sulfane sulfur is not well delineated and may result in the accumulation of volatile H 2 S ( Xin et al., 2023 ). Additionally, sulfane sulfur has been found to exhibit toxicity in both bacteria and fungi, including when produced through the i sox pathway in Cupriavidus pinatubonensis ( Sato et al., 2011 ; Xin et al., 2023 ). Of the four genera in this study which contain the i sox pathway, only Sulfuritalea and Thiobacillus possess the ability to detoxify sulfane sulfur using either the r dsr or s hdr pathway. Additional cytoplasmic sulfur oxidative genes investigated include TST (S 2 O 3 2- to SO 3 2- ) and soeABC (SO 3 2- to SO 4 2- ; Figure 1 ). TST was broadly present across the SOB occurring in >80% of the Thiomonas spp., Halothiobacillus spp., Thiobacillus spp., Sulfuricurvum spp., and Sulfuriferula spp. metagenomes and in <80% of the Thiovirga spp. and Sediminibacterium spp. metagenomes ( Figure 1 ). To date, no information is available regarding the presence or absence of TST in Sulfurimonas or Sulfuritalea genomes. Wang et al. (2019) suggests TST may use S 2 O 3 2- produced by SOR to further oxidize it to sulfur and SO 3 2- . SoeABC transforms SO 3 2- to SO 4 2- in the cytoplasm and may even contribute to SO 3 2- oxidation as part of the r dsr pathway ( Dahl et al., 2013 ; Whaley-Martin et al., 2023 ). Greater than 80% of the Thiomonas, Thiobacillus and Sulfuriferula genomes contained soeABC ( Figure 1 ). SoeABC has also been identified in Sulfuritalea hydrogenivorans sk43H ( Watanabe et al., 2014 ) but no information on soeABC ’s occurrence in Sulfurimonas is currently available. Additional periplasmic sulfur oxidation genes identified in this study include sqr , fccAB, and sorAB ( Figure 1 ). Both sqr and fccAB are implicated in catalysis of hydrogen sulfide to S 0 which can form S 0 globules and be further oxidized to SO 3 2- via the r dsr pathway ( Nosalova et al., 2023 ). Sqr was widely present appearing in >80% of the genomes identified in all genera apart from Sediminibacterium and Sulfuriferula metagenomes where they were present in <80% of the metagenomes ( Figure 1 ). Sulfuritalea hydrogenivorans sk43H and various Sulfurimonas species were found to host the s qr gene ( Watanabe et al., 2019 ; Wang et al., 2021 , 2023 ). Among the reconstructed genomes, >80% of the Halothiobacillus genomes harbored the fccAB gene cluster while the fccAB cluster was identified in <80% of the Thiomonas and Thiobacillus genomes ( Figure 1 ). Sulfuritalea hydrogenivorans sk43H has been reported to contain fccAB ( Watanabe et al., 2019 ) and currently no Sulfurimonas species have been reported to contain fccAB . SorAB , which catalyzes the conversion of SO 3 2- to SO 4 2- in the periplasm, was found in >80% of the Sulfuricurvum genomes and < 80% of the Thiomonas, Thiobacillus , and Sulfuriferula genomes ( Figure 1 ). SorAB has also been reported in Sulfuritalea hydrogenivorans sk43H ( Watanabe et al., 2019 ) as well as eight of the eleven Sulfurimonas species examined by Wang et al. (2021) . The four most abundant genera in these four TI collectively possess the capacity for sulfur oxidation via all three universal pathways, c sox ( Halothiobacillus , Sulfuricurvum ), i sox + rdsr ( Thiobacillus ), and S 4 I ( Halothiobacillus, Sediminibacterium , Thiobacillus ) suggesting adaptation of these TI SOB communities to occupy all potential sulfur oxidizing niches that occur in these highly physicochemically and geochemically dynamic TI systems. 3.1.1 SOB functional classification Under acid mine drainage conditions (acidic, metal rich), Kuang et al. (2016) identified metabolic function as a better predictor of microbial community structure and function than taxonomy. Similarly, here, patterns in metagenomic data were used to classify these SOB genera into c sox dominant and non-c sox dominant SOB genera groupings. C sox dominant SOB genera identified here included Halothiobacillus spp., Thiovirga spp., Thiomonas spp., and Sulfuricurvum spp. based on the presence and high abundance of the c sox pathway ( Figure 1 ). Non-c sox dominant SOB genera were characterized by i sox gene pathway presence and increased abundance of genes associated with alternative pathways such as r dsr ( dsrABCEFH , aprAB , sat ; Whaley-Martin et al., 2023 ) and/or S 4 I ( tsdA, tetH ; Ghosh and Dam, 2009 ; Wang et al., 2019 ; Whaley-Martin et al., 2023 ) ( Figure 1 ). Thiobacillus spp., Sediminibacterium spp., Sulfuriferula spp., Sulfuritalea spp., and Sulfurimonas spp. were categorized as non-c sox dominant SOB genera ( Figure 1 ). Importantly, reactions catalyzed by the c sox pathway favour the complete oxidation of SOI to SO 4 2- , resulting in increased acidity and SO 4 2- production, while more energy efficient pathways (e.g., r dsr ; Klatt and Polerecky, 2015 ), favoured by non-c sox dominant SOB commonly generate free SOI and in some cases may consume H + ( Dam et al., 2007 ; Klatt and Polerecky, 2015 ; Hutt et al., 2017 ). Therefore, presence or absence of c sox vs. non-c sox dominant pathways may reflect, and in turn, result in physicochemically and geochemically distinct waters. Mine waters with non-c sox dominant SOB (i.e., Thiobacillus , which can generate free SOI) may be at greater risk of offsite acidification due to the ability of many SOI to pass through current treatment to receiving environments where their subsequent oxidation could release acidity. 3.1.2 Spatial and temporal trends in SOB community composition and function The abundance of the nine identified SOB genera differed spatially and temporally for individual mines as well as across the four mines ( Figure 2 ). The two oldest TIs, Mine 2 and Mine 1, had the highest SOB abundances ( Figure 2 ). Mine 2 TI had the largest total abundance of SOB genera across all samples (54 ± 19%) and was the only mine TI where c sox dominant SOB genera (predominantly Halothiobacillus spp.) had a higher average abundance than non-c sox dominant SOB genera (predominantly Sediminibacterium spp.; Figure 2 , Table 2 ). With more than 50% of the community on average consisting of SOB, Mine 2 TI had the lowest non-SOB genera abundance (38 ± 17%) and lowest unknown genera abundance (12 ± 9.4%; Figure 2 ). Based on evidence from this study and a previous study ( Whaley-Martin et al., 2023 ), the Mine 2 community had the potential to express the c sox pathway (via Halothiobacillus spp., Thiovirga spp., Thiomonas spp. or Sulfuricurvum spp.), i sox pathway (via Thiobacillus spp. or Sulfuriferula spp.), r dsr pathway (via Thiobacillus spp.), and/or the S 4 I pathway (via Thiobacillus spp.). The average c sox dominant SOB abundance was significantly higher in Mine 2 ( p < 0.001) compared to the other three mines, while the abundance of non-c sox dominant SOB were not significantly different across the four mines. Averaging a total SOB abundance 2.7 times lower than Mine 2, Mine 1 TI had the second highest total SOB abundance (20 ± 24%) ( Figure 2 , Table 2 ). SOB genera present at Mine 1 could express the c sox pathway (via Halothiobacillus spp., Thiovirga spp., or Sulfuricurvum spp.), i sox pathway (via Thiobacillus spp. or Sulfurimonas spp.), r dsr pathway (via Thiobacillus spp.), and/or the S 4 I pathway (via Thiobacillus spp.). Figure 2 16S rRNA relative abundance (%) of top nine sulfur oxidizing bacteria genera from 2016 – 2019, non-SOB genera and unknown genera for each of the four mines mapped to their sample locations across Canada. Individual samples are shown as bar graphs and mine average abundances are shown as pie graphs. “Non-SOB genera” are all other identified sequences and “unknown genera” are identified sequences not matched to genera in the Silva Database v138.1. Asterisks (*) denote samples with metagenome data included in this study. Mine 3’s TI waters averaged 14 ± 7.5% total SOB abundance ( Table 2 ), which was approximately 4 times lower than Mine 2 and 1.4 times lower than Mine 1. Mine 3 TI did, however, have the highest abundance of unknown genera (38 ± 12%) and the second highest abundance of identified non-SOB genera (48 ± 8.2%; Figure 2 ). The majority of TI SOB present in Mine 3 TI waters were identified as non-c sox dominant SOB and averaged 12 ± 6.2% primarily consisting of Thiobacillus spp., Sediminibacterium spp. and Sulfuritalea spp. ( Figure 2 , Table 2 ). The average abundance of c sox dominant SOB genera at Mine 3 was 2.0 ± 3.7% and was primarily Thiovirga spp. ( Figure 2 , Table 2 ). 16S rRNA abundance, metagenome data and present literature indicate that SOB genera from Mine 3 could express the c sox pathway (via Thiovirga spp.) and/or the i sox (via Thiobacillus spp. or Sulfuritalea spp.), r dsr (via Thiobacillus spp.), and S 4 I pathway (via Thiobacillus spp.). The lowest total SOB abundance was found in the Mine 4 TI samples, with an average abundance of 7.1 ± 6.3 and > 97% of the SOB present represented by non-c sox dominant SOB genera ( Table 2 ). Non-c sox dominant SOB genera averaged 6.5 ± 6.6% of the overall community abundance, while c sox dominant SOB abundance was the lowest of the four mines averaging 0.3 ± 0.4%, with only Thiovirga spp. ever reaching above 1% of the overall community abundance at Mine 4 ( Table 2 ). Mine 4 had the second highest contribution of unknown genera representing 27 ± 17% of the microbial community ( Figure 2 ). Mine 4 had the most limited S oxidation potential with no major SOB c sox abundance ( Table 2 ). Thiobacillus spp. present at Mine 4, contained tsdA enabling part 1 of the S 4 I pathway (S 4 O 6 2- formation), but did not contain tetH associated with the 2 nd part of the S 4 I pathway (S 4 O 6 2- disproportionation) and did contain genes for the i sox and r dsr pathway. 3.2 TI wastewater sulfur geochemistry Mean [total S 0.45μm ] and [SO 4 2- ] across the four mines display a pattern of increasing concentration with age, though this relationship is not proportional. Notably, Mine 1 and Mine 2 are most similar in age and Mine 2 and Mine 3 are most similar in concentrations ( Table 1 , Figure 3 ). However, [total S 0.45μm ] exhibited substantial variability across the four mines – ranging from 0.3 mM (Mine 4) to 16.8 mM (Mine 1) ( Table 1 ). ANOVA and post-hoc Tukey pairwise comparison tests revealed a significantly higher average total S 0.45μm concentration at Mine 1 (14 ± 1.6 mM) compared to Mine 2 (8.9 ± 0.7), Mine 3 (7.9 ± 1.0), and Mine 4 (1.9 ± 0.8 mM) ( Table 1 ; p < 0.001). [SO 4 2- ] displayed a similar trend where Mine 1 exhibited significantly higher concentrations than Mine 2 ( p < 0.05), Mine 3 ( p < 0.001), and Mine 4 ( p < 0.001) ( Table 1 , Figure 3 ). Elevated sulfur concentrations in Mine 1 and Mine 2 TIs can be attributed to their extensive use since the 1920s. Mine 3, while younger than Mine 1 and Mine 2, contains a comparatively higher [total S 0.45μm ], reflecting tailings additions from multiple mining operations and a shallow water cover depth. Mine 4, the smallest and youngest TI, has accumulated the lowest volume of tailings, resulting in the lowest [total S 0.45μm ]. Figure 3 Box and whisker plots and statistical analysis (ANOVA and post-hoc tukey pairwise statistical comparison) of cross mine pH, total S (mmol/L), S 2 O 3 2- (mmol/L), and S React (mmol/L). Box limits represent the first and third quartile of each dataset, with a black line indicating the median value and an “x” denoting the mean. Though not often reported, [total S] from various environments varies widely from ~10 μM to 800 μM observed in freshwater lakes ( Vincent et al., 2021 ) to ~28 mM in sea water ( Canfield and Farquhar, 2009 ; Vincent et al., 2021 ) with SO 4 2- typically being the largest contributor to the overall sulfur balance ( Canfield and Farquhar, 2009 ). Reported [SO 4 2- ] from other mining environments studies range from 3.6 mM to >250 mM ( Foucher et al., 2001 ; Silva et al., 2012 ; Kinnunen et al., 2018 ; Camacho et al., 2020b ). [Total S] (0.3 – 16.8 mM; Table 1 ) in this study are typically higher than the reported freshwater <1 μM value, while on the lower end of reported values for mining contexts ( Foucher et al., 2001 ; Kinnunen et al., 2018 ). Reactive sulfur concentrations ([S React ] calculated as [Total S] – [SO 4 2- ]; Whaley-Martin et al., 2020 ), representing all sulfur atoms capable of oxidation, were determined for each TI sample. ANOVA and post-hoc Tukey pairwise comparison tests revealed that Mine 1 had a significantly higher average S React compared to Mine 2 ( p < 0.001), Mine 3 ( p < 0.05) and Mine 4 ( p < 0.001) ( Figure 3 ). Considering the potential 10-fold variation in both [total S 0.45μm ] and [S React ] between samples, %S React ([S React ]/[Total S 0.45μm ] x 100) provides a useful metric to reflect the proportion of S React within each system’s individual total S pool. %S React ranged from 0 to 92.3% across the four mines, with the highest average %S React occurring at Mine 1 (28 ± 16.2%) followed by Mine 4 (27 ± 13%), Mine 3 (17 ± 4.4%), and Mine 2 (12 ± 9.5%). S React may include various sulfur species such as S 2 O 3 2- and SO 3 2- (quantified in this study), as well as additional unresolved sulfur species including but not limited to S 4 O 6 2- , S 2 O 4 2- , and S 0 which were not quantified here but could potentially contribute to or support microbial sulfur cycling ( Bak and Pfennig, 1987 ; Kelly et al., 1997 ). [SO 3 2- ] were generally low (often at or near the detection limit) with the highest concentrations found at Mine 1 (0.03 ± 0.05 mM; Table 1 ). S 2 O 3 2- however, was detectable across all sites with a significantly higher ( p < 0.001) average concentration at Mine 1 compared to the other three mines ( Figure 3 ). [S 2 O 3 2- ] was the largest measured S React contributor, constituting 0 – 100% (calculated %S 2 O 3 2- = [S 2 O 3 2- ]/[S React ] x 100) of the S React pool at each mine. While Mine 4 had the largest average %S React , it had the smallest average %S 2 O 3 2- (5.3 ± 8.2%) and therefore the largest proportion of unresolved sulfur species. On average, [S 2 O 3 2- ] comprised approximately one-fifth of both Mine 1 (22 ± 16%) and Mine 2’s (19 ± 27%) [S React ] pool. 3.3 TI wastewater physicochemistry 42 TI water cap samples were collected from the four mines between 2016 and 2019 during open water periods (early spring to late fall; Supplementary Figure S1 ) exhibited variable pH and DO values which are known to be important influencers of microbial ecological niches and associated sulfur oxidation pathways ( Chen et al., 2013 ; Whaley-Martin et al., 2023 ). High temporal variability in DO (% saturation) and pH occurred for each mine as well as across mines, reflecting morphometric differences between TI facilities (particularly depth; Table 1 ) and the dynamic nature of actively managed TIs ( Figure 3 ). Oxygen profiles from Mine 1, Mine 2, and Mine 3 demonstrated a steep oxygen gradient within TI water caps (Mine 4 profile data unavailable), ranging from <1 to >100% saturation ( Table 1 ) across the four mines. Cross mine comparison revealed Mine 4 had significantly higher %DO compared to Mine 1 and Mine 2 (ANOVA and a post-hoc Tukey pairwise comparison test, p < 0.05; Figure 3 ) and no statistically significant differences among the remaining mines %DO. Across the four mines, pH ranged from 5.1 to 11.8 ( Table 1 ; Figure 3 ). Lime (Ca(OH) 2 ) and alkaline tailings additions resulted in the highest average pH at Mine 1 (9.7 ± 2.0) and significantly higher pH values than Mine 2 ( p < 0.001), Mine 3 ( p < 0.05), and Mine 4 ( p < 0.05) ( Figure 3 ). Mine 2 TI waters had the lowest average pH (6.3 ± 0.9) with pH regularly falling below pH 6.5 (~64% of datapoints) and an observed decrease in pH from late spring to early fall. Mine 3’s TI water cap had the narrowest range of pH values (6.6 – 7.6, avg. = 7.1) with no observed seasonal fluctuations. Mine 4’s average pH was 8.0 ± 0.7 (7.2 – 9.0), with 75% of samples below pH 8.5, which was significantly higher than Mine 2 ( p < 0.05) and lower than Mine 1 ( p < 0.05; Figure 3 ). All four TIs exhibit dimictic lake mixing patterns, resulting in more homogenous dissolved oxygen and pH profiles during turnover events (spring and fall). The acidity to SO 4 2- ratio ([H + ]/[SO 4 2- ]) can be used as a proxy for discerning direct sulfur oxidation from disproportionation ( Bernier and Warren, 2007 ; Whaley-Martin et al., 2023 ). Higher [H + ]/[SO 4 2- ] values observed at lower pH values indicate greater dominance of the c sox pathway, indicating more direct and complete oxidation to SO 4 2- (higher acidity generation, minimal SOI generation; Supplementary Figure S3 ). Lower [H + ]/[SO 4 2- ] values observed at pH values >7, are consistent with more regeneration of SOI via disproportionation or partial oxidation attributed to i sox , r dsr , and S 4 I pathways activity, resulting in lower net acidity generation ( Supplementary Figure S3 ). While these ratio values may be underestimates at pH values >7, where buffering capacity is likely present, the observed pH dependent [H + ]/[SO 4 2- ] trend is consistent with both S 2 O 3 2- and SOB results. [S 2 O 3 2- ] was positively correlated to pH ( p < 0.0001, Pearson’s r = 0.80; Supplementary Figure S4A ), while total TI SOB abundance was negatively correlated with pH ( p < 0.01, Pearson’s r = −0.64; Supplementary Figure S4B ). Interestingly, average mining and anthropogenic (M&A) literature and environmental literature values were consistent with the pattern observed here ( Supplementary Figure S4B ). 3.4 pH effect on SOB community structure Further examination of the relationship between SOB and pH for these four mine TIs, highlighted a pH dependent pattern in the occurrence of c sox and non-c sox dominant SOB. While clear relationships between pH and total SOB abundance, [S 2 O 3 2- ], and [H + ]/[SO 4 2- ] ratios were observed, no discernible correlation was identified between dissolved oxygen and any of these parameters. Peak abundances of non-c sox dominant SOB were observed at circumneutral pH (pH ~6 to ~8.5; Figure 4Ai ) while the highest abundances of c sox dominant SOB were observed below pH ~6.5 ( Figure 4Aii ). Samples above pH ~8.5 had low abundances of both c sox and non-c sox dominant SOB ( Figures 4Ai,ii ). Current literature indicates at least two genera of SOB ( Thioalkalimicrobium and Thioalkalivibrio ) are capable of growth under alkaline conditions ( Sorokin et al., 2001 ) though neither were observed in this study’s TI samples. Lower pH value (pH < ~6.5) TI waters had the highest abundances of c sox dominant SOB, lower [S 2 O 3 2- ] and higher [H + ]/[SO 4 2- ] ratios; collectively consistent with direct oxidation via the c sox pathway ( Figure 4Aii ). At more circumneutral pH values (pH ~6 to ~8.5), where non-c sox dominant SOB dominated, higher [S 2 O 3 2- ] and low [H + ]/[SO 4 2- ] values were also observed, indicating distinct pH dependent SOB community structure and S pathway(s) ( Figures 4Ai,ii ). Figure 4 Genus level identification of (A–C) c sox dominant SOB relative abundances, (Ai, Bi, Ci) non-c sox dominant SOB relative abundances and (Aii, Bii, Cii) average 16S rRNA SOB communities identified in (A, Ai, Aii) tailings impoundment water samples from this study, (B, Bi, Bii) other mining and anthropogenic sample data from the literature and (C, Ci, Cii) environmental sample data obtained from the literature. “Non-SOB genera” are all other identified sequences in the samples and “unknown genera” are identified sequences not matched to genera in the Silva Database v138.1 (Data included in Bi, Bii, Ci , and Cii was collected from the following papers and can be identified using the superscript number: 1 Kadnikov et al., 2019 , 2-5 Miettinen et al., 2021 , 6-7 Auld et al., 2017 , 8-9 Lopes et al., 2020 , 10-12 Chen et al., 2013 , 13 Haosagul et al., 2020 , 14 Patwardhan et al., 2018 , 15 Arce-Rodríguez et al., 2019 , 16 Vavourakis et al., 2019 , 17-18 Meier et al., 2017 , 19-20 Reigstad et al., 2011 ). Black outlines indicate non-water samples (e.g., soil, rock, tailings, biofilm, etc.). 3.4.1 Comparison to broader environments The exploration of the pH – c sox dominant/non-c sox dominant SOB relationship extended across diverse contexts by integrating literature data (pH and SOB abundance) from broader mining environments, an industrial H 2 S bioscrubber, and a variety of natural environmental sites (e.g., hydrothermal vents, soda lakes, and thermal cave springs) ( Reigstad et al., 2011 ; Meier et al., 2017 ; Patwardhan et al., 2018 ; Arce-Rodríguez et al., 2019 ; Vavourakis et al., 2019 ) (see Supplementary Table S1 ). Genus level SOB abundances were categorized into c sox dominant SOB and non-c sox dominant SOB. Literature identified SOB (not found in this study’s TI samples) were classified as c sox dominant and non-c sox dominant SOB using available genetic data and consensus of the current literature based on the reported presence of a complete or incomplete sox pathway. Additional SOB genera included three c sox dominant clades Thiomicrospira spp., Thioalkalimicrobium spp., and Halothiobacilleaceae Family and seven non-c sox dominant clades ( Thiothrix spp., Thiomicrohabdus spp., Omnitrophica spp., Sulfurovum spp., Thermithiobacillus spp., Sulfobacillus spp., and Thioalkalispira spp.). Though class and family level abundances have been published for TI water cap samples, genus level data were not widely available in the literature. Data from various alternative sample locations (e.g., water from mill feed, solid tailings pore water), including both solid and water samples, were used. This approach aimed to broaden the range of geochemical and physicochemical characteristic investigated. A brief summary and comparison of the relevant data for the overall groups of samples are provided in Supplementary Table S1 as more complete descriptions for individual sites can be found in their respective publications. The collective M&A literature data ( n = 13) covered a pH range from 1.9 to 10.2 ( Supplementary Table S1 ), the environmental samples ( n = 7) ranged from pH 2.4 to 9.9 ( Supplementary Table S1 ), while this study’s TI samples ranged from pH 5.1 to 11.8 ( Table 1 ). Average total SOB abundance of this study’s TI samples (29 ± 26%) and M&A literature samples (31 ± 31%) were similar, while the average total SOB abundance for the environmental literature samples (42 ± 26%) was >10% higher ( Figures 4A – C ). Additional SOB genera (outside of the nine identified for TI in this study) represented ~15% (or ~ 50% of the total SOB) of the M&A literature and ~ 26% (or ~ 62% of the total SOB) of the environmental literature samples ( Figures 4B , C ). Both the M&A and environmental literature SOB communities diverged by at least 50% from those characterized here for base metal TI wastewaters, indicating notable disparities in microbial community composition suggesting SOB genera specific ecological niches. On average, this study’s samples had ~64% of the total SOB abundance consisting of c sox dominant SOB, while ~36% were classified as non-c sox dominant ( Figure 4A ). The M&A literature samples averaged slightly higher c sox dominant SOB contribution with ~72% of the total SOB abundance being attributed to c sox dominant SOB ( Figure 4B ). Conversely, the environmental literature samples averaged a much lower contribution of c sox dominant SOB, contributing only ~18% of the total SOB abundance ( Figure 4C ). Thiovirga spp. and Thiomicrospira spp., both classified as c sox dominant SOB, were the two most abundant SOB in the M&A literature samples, while the two most abundant SOB in the environmental samples were both classified as non-c sox dominant SOB ( Sulfurimonas spp. and Sulfurovum spp.) ( Figures 4B , C ; Supplementary Table S1 ). In contrast, the two most abundant SOB from this study’s TI waters were Halothiobacillus spp. (c sox dominant) and Sediminibacterium spp. (non-c sox dominant). Thiomicrohabdus spp., Omnitrophica spp., and Thermithiobacillus spp. were unique to M&A samples with Thiomicrohabdus spp. and Omnitrophica spp. only being found in a single sample ( Kadnikov et al., 2019 ) and Thermithiobacillus spp. identified in two samples from the same mine ( Lopes et al., 2020 ). Thioalkalispira spp. and Thioalkalimicrobium spp. were unique to one environmental sample ( Vavourakis et al., 2019 ). Four M&A literature samples were found to have elevated abundances (>10%) of non-c sox dominant SOB including three solid samples and one water sample ( Figure 4Bii ). Non-c sox dominant SOB abundances of the M&A literature samples were found to be highest in the only non-mining sample (13 SPM Swine Farm H 2 S Bioscrubber System; Figure 4Bii ). Three of the four M&A literature samples with elevated SOB abundances fell between pH ~6.5 and ~ 8, covering a similar range to the samples included in this study ( Figure 4Bi ). Among the seven M&A literature samples displaying elevated (>10%) c sox dominant SOB abundances, five were found to occur within a pH range of ~5 and ~ 7 ( Figure 4Bii ); a slightly higher upper range compared to this study’s TI results (< pH 6.5; Figure 4Aii ). The highest c sox dominant SOB abundances in M&A literature samples were identified in samples at pH ~5 and ~ 7 ( Figure 4Bii ) though no data between those two pH values are presently available in the literature. At pH >8, there were two M&A literature samples which both showed low abundances of both c sox dominant and non-c sox dominant SOB ( Figures 4Bi,ii ), aligned with the pattern of low SOB abundance at more alkaline pH observed in this study’s TI samples. Samples below pH 4 (closer to acid mine drainage conditions) from the M&A literature samples had very low SOB abundances ( Figures 4Bi,ii ) in both solid and water samples. Differences observed between the M&A literature data and this study’s TI data may reflect differences in S substrate availability due to the types or locations of samples included. The M&A mining-related solid sample’s primary source of sulfur would be sulfide-containing ores whereas SOI, aqueous dissolved species, prevalent in the TI waters investigated here, would be limited in solid samples. Solid tailings and feed samples may similarly exhibit different patterns of SOB community structure and abundance and associated sulfur oxidation genes due to these differences in available sulfur species, as only one of the three main pathways highlighted here ( sox ) utilizes sulfide ( Klatt and Polerecky, 2015 ). This may account for the elevated c sox dominant SOB abundances observed in some of the circumneutral solid samples included in Figure 4Bii . Extrapolating the pattern identified in this study’s samples to other anthropogenically impacted environments would require more data (both genus level abundance and pH data) for TI water caps which are not currently available. However, these comparative results, across a range of systems where data are available, show some consistencies in patterns with those observed for base metal TI wastewaters, namely low SOB abundances at elevated pH values (pH >8) and similar pH ranges for the peak functional pathway abundances of both non-c sox dominant (pH ~6.5 to ~8.5) and c sox dominant (~4.5 to ~7) SOB ( Figures 4Bi,ii ). The environmental literature samples that fell between pH ~6.5 and ~ 8.5 had elevated non-c sox dominant SOB abundances ( Figure 4Ci ) consistent with results for this study’s TI waters. No environmental literature samples were found between pH ~5 and ~ 6.5 precluding a direct comparison to the peak c sox dominant SOB abundances observed for TI here. Similar to the observed TI results here, however, three environmental literature samples (two samples <pH ~5 and one sample > pH ~8.5) had lower abundances of non-c sox dominant SOB than those within the more circum-neutral range (pH ~6.5 – ~8.5) ( Figure 4Ci ). The environmental literature group of samples was the only group where all samples had >1% non-c sox dominant SOB ( Figure 4Ci ). Unlike the previous two groups of samples, none of the environmental samples had high abundances of c sox dominant SOB (all <20% c sox dominant SOB abundance; Figure 4Cii ). This may reflect typically much lower environmental SOI concentrations ( O’Brien and Birkner, 1977 ) and therefore metabolisms that exhaust SOI by completely oxidizing them to SO 4 2- (i.e., c sox ) may not be favoured over those that disproportionate or recycle SOI such as the S 4 I or r dsr pathways (ie. non-c sox dominant SOB). The divergence of mining impacted environments (both in this study (oxic TI wastewaters) and literature data) from natural environments suggests a specialized mining specific microbiome with higher abundance(s) of c sox dominant SOB favoured by higher concentrations of S species in these contexts. 3.5 Factors influencing microbial community structure and function The i sox pathway or the i sox + r dsr pathway generates more ATP and is more efficient due to energy conservation via S 0 production and storage ( Klatt and Polerecky, 2015 ). While energy efficiency is important, it is not the sole determinant of success in any given environment ( Klatt and Polerecky, 2015 ). This is exemplified across the diverse set of environments presented here, where both the i sox or i sox + rdsr pathway, as well as the c sox pathway appear at different times. Microbial success can be determined by growth rate, which considers both the substrate uptake rate and the growth yield ( Klatt and Polerecky, 2015 ). Both growth yield and efficiency are directly related, but lower efficiency (such as that associated with c sox ) can be made up for by speed, resulting in equal or greater success ( Sorokin and Kuenen, 2005 ; Klatt and Polerecky, 2015 ). Whaley-Martin et al. (2023) provided laboratory enrichment data from the same four mines included in this study to showcase that c sox SOB ( Halothiobacillus spp. dominant) enrichments exhibit a considerably faster S 2 O 3 2- oxidation rate compared to the enrichments with r dsr containing SOB. Growth strategies which rely on speed are only likely to dominate in environments with unlimited substrates ( Klatt and Polerecky, 2015 ). Though mining and other industrial environments technically do not have unlimited SOI, there are consistent replenishments of SOI through tailings additions or continuous wastewater treatment/flow. The observed proliferation of SOB with the c sox pathway in mining and industrial contexts ( Figures 4Aii,Bii ), aligns with their demonstrated capability for rapid S 2 O 3 2- oxidation under primarily oxic conditions. This highlights their competitive advantage in these substrate-rich environments where consistent SOI inputs help sustain their growth. The transition from non-c sox dominant SOB (i sox or i sox + r dsr ) to c sox dominant SOB occurs within the range of 0.1 to 0.3 mmol/L [S 2 O 3 2- ] ( Supplementary Figure S4A ), suggesting a potential SOI threshold for the initiation of this metabolic shift. The associated pH decrease with the increased presence of c sox dominant SOB may occur as a by-product of the activity of the c sox pathway which produces SO 4 2- and acidity and may slowly decrease pH. Further, under natural environmental conditions, where SOI are not as readily replenished, we do not see the proliferation of the c sox pathway ( Figure 4Cii ). Across the four TIs investigated here, evidence of pH and [S 2 O 3 2- ] dependent occurrence of all four key metabolic pathways, c sox pathway, S 4 I pathway, i sox pathway, and r dsr pathway emerged ( Figure 5 ). pH partitioned three of these pathways: namely c sox pathway dominated at lower pH values (pH ~5 to ~6.5) while the i sox and r dsr pathways were more prevalent at circumneutral pH values (pH ~6.5 to ~8.5; Figure 5 ). The shift from circumneutral to more acidic pH may represent the onset of a positive feedback loop led by c sox dominant SOB such as Halothiobacillus spp. which has previously been implicated in catalyzing the shift to net acid generation in laboratory scale experiments ( Whaley-Martin et al., 2019 ). The c sox pathway, responsible for the direct oxidation of S 2 O 3 2- to SO 4 2- , which generates more acidity than the S 4 I, i sox , or r dsr pathway results in decreasing [S 2 O 3 2- ] and increasing [H + ]/[SO 4 2- ] values ( Figure 5 ). Uniquely, portions of the S 4 I pathway, namely the tsdA gene, were observed in SOB genera from across the entire pH range ( Figure 5 ). Figure 5 Conceptual model of pH bound metabolic activity with the associated responsible genera and geochemical pathway and outcomes in metal mining TI waters."
} | 14,354 |
22675593 | PMC3368403 | pmc | 8,399 | {
"abstract": "Paenibacillus sp. strain JDR-2, an aggressively xylanolytic bacterium isolated from sweetgum ( Liquidambar styraciflua ) wood, is able to efficiently depolymerize, assimilate and metabolize 4-O-methylglucuronoxylan, the predominant structural component of hardwood hemicelluloses. A basis for this capability was first supported by the identification of genes and characterization of encoded enzymes and has been further defined by the sequencing and annotation of the complete genome, which we describe. In addition to genes implicated in the utilization of β-1,4-xylan, genes have also been identified for the utilization of other hemicellulosic polysaccharides. The genome of Paenibacillus sp. JDR-2 contains 7,184,930 bp in a single replicon with 6,288 protein-coding and 122 RNA genes. Uniquely prominent are 874 genes encoding proteins involved in carbohydrate transport and metabolism. The prevalence and organization of these genes support a metabolic potential for bioprocessing of hemicellulose fractions derived from lignocellulosic resources.",
"introduction": "Introduction Paenibacillus sp. strain JDR-2 (Pjdr2) was isolated from wafers cut from live stems of sweet gum ( Liquidambar styraciflua ) placed in soil in an area populated predominantly by this tree species. The ability of this isolate to grow on 4-O-methylglucuronoxylose (MeGX) as the sole carbon source identified a metabolic potential not previously described. MeGX is released along with fermentable xylose during dilute acid pretreatment of lignocellulosic biomass. Since MeGX may represent 5 to 20% of the hemicellulose components from hardwoods and agricultural residues, this ability was of interest for increasing bioconversion yields of fermentable sugars from these resources [ 1 , 2 ]. Growth rates and yields of Pjdr2 with polymeric 4-O-methylglucuronoxylan (MeGXn) as substrate were much greater than with monosaccharides and oligosaccharides derived from MeGXn. These increases are presumably the result of a cell-associated multimodular GH10 endoxylanase that generates xylobiose, xylotriose, and the aldouronate, 4-O-methylglucuronoxylotriose (MeGX3), for direct assimilation and metabolism [ 2 ]. A cluster of genes was cloned and sequenced from Pjdr2 genomic DNA which contained two genes encoding transcriptional regulators, three genes encoding ABC transporters, and three sequential structural genes lacking secretion sequences encoding a GH67 α-glucuronidase, a GH10 endoxylanase catalytic domain and a putative GH43 β-xylosidase. The expression of these genes, as well as a distal gene encoding a secreted cell-associated multimodular GH10 endoxylanase, was coordinately responsive to inducers and repressors, leading to their collective designation as a xylan-utilization regulon [ 3 ]. Physiological studies defining the preferential utilization of MeGXn compared to MeGX and MeGX3 support a process in which extracellular depolymerization, assimilation and intracellular metabolism are coupled, allowing the rapid and complete utilization of MeGXn [ 4 ]. Pjdr2 was the first member of this genus to have its genome completely sequenced and made available for detailed analysis. The sequences of genomes of 2 strains of Paenibacillus polymyxa [ 5 , 6 ], “ Paenibacillus vortex” [ 7 ], and Paenibacillus sp. Y412MC10 (NCBI NC_013406.1, unpublished results) have since been completed. The incomplete genome sequence Paenibacillus larvae subsp. larvae , the causative agent of American Foulbrood disease of honey bees, has also been analyzed [ 8 ]."
} | 884 |
28216617 | PMC5316977 | pmc | 8,400 | {
"abstract": "Because the treatment of oily wastewater, generated from many industrial processes, has become an increasing environmental concern, the search continues for simple, inexpensive, eco-friendly, and readily scalable processes for fabricating novel materials capable of effective oil/water separation. In this study we prepared an eco-friendly superhydrophilic and underwater superoleophobic polyvinylpyrrolidone (PVP)-modified cotton that mediated extremely efficient separations of mixtures of oil/water and oil/corrosive solutions. This PVP-modified cotton exhibited excellent antifouling properties and could be used to separate oil/water mixtures continuously for up to 20 h. Moreover, the compressed PVP-modified cotton could separate both surfactant-free and -stabilized oil-in-water emulsions with fluxes of up to 23,500 L m −2 h −1 bar −1 —a level one to two orders of magnitude higher than that possible when using traditional ultrafiltration membranes having similar rejection properties. The high performance of our PVP-modified cotton and its green, low-energy, cost-effective preparation suggest its great potential for practical applications.",
"discussion": "Discussion We have fabricated an eco-friendly, low-toxicity superwetting material from a common bio-derived material (cotton) and a low-toxicity polymer (PVP). The PVP-modified cotton displayed superhydrophilicity and underwater-superoleophobicity; this material should have practical use in the effective separations of water-rich immiscible oil/water mixtures—even mixtures of oils and corrosive solutions—with extremely high separation efficiencies. The PVP-modified cotton exhibited excellent antifouling properties during long-term use. Moreover, the compressed PVP-modified cotton allowed effective separations of surfactant-free and -stabilized oil-in-water emulsions with ultrahigh fluxes (up to 23,900 L m −2 h −1 bar −1 ) and high separation efficiencies (oil contents in filtrated water: <10.0 ppm). Thus, this PVP-modified cotton has great potential for purifying immiscible oil/water mixtures and oil-in-water emulsions from industrial sources or found in our daily lives."
} | 534 |
31673108 | PMC6823497 | pmc | 8,402 | {
"abstract": "Despite recent advances in observational data coverage, quantitative constraints on how different physical and biogeochemical processes shape dissolved iron distributions remain elusive, lowering confidence in future projections for iron-limited regions. Here we show that dissolved iron is cycled rapidly in Pacific mode and intermediate water and accumulates at a rate controlled by the strongly opposing fluxes of regeneration and scavenging. Combining new data sets within a watermass framework shows that the multidecadal dissolved iron accumulation is much lower than expected from a meta-analysis of iron regeneration fluxes. This mismatch can only be reconciled by invoking significant rates of iron removal to balance iron regeneration, which imply generation of authigenic particulate iron pools. Consequently, rapid internal cycling of iron, rather than its physical transport, is the main control on observed iron stocks within intermediate waters globally and upper ocean iron limitation will be strongly sensitive to subtle changes to the internal cycling balance.",
"introduction": "Introduction Upper ocean primary production is limited by the availability of iron (Fe) over much of the ocean 1 . Even where nitrogen (N) and phosphorus (P) are the main limiting factors, Fe continues to play a key role by driving rates of N fixation 2 and acquisition of dissolved organic P 3 . Fe limitation ultimately arises owing to a deficiency in the supply of Fe, relative to N and P 4 . Away from regions of dust deposition, the dominant component of Fe delivery, relative to N or P, is its relative concentration in thermocline waters 5 . This is particularly apparent across the south Pacific Ocean, where transport by sub-Antarctic mode water (SAMW) and Antarctic Intermediate water (AAIW) plays a key role in setting thermocline nutrient levels 6 . Accordingly, any fluctuations in the relative balance between Fe and major nutrients N and P in mode and intermediate waters in response to changes in climate will influence upper ocean Fe limitation and consequently modify global carbon and nitrogen cycles. At present, there is low confidence in model projections of how modulations to climate will affect Fe supply to the upper ocean, as models generally show poor skill and substantial disagreement in their representation of the present-day ocean iron cycle. This lack of fundamental understanding of iron biogeochemistry is well illustrated by the order of magnitude inter-model variability in the residence time of iron in global models, despite aiming to reproduce the same ocean distributions from state of the art data sets 7 . Thus, despite a relatively long legacy of modelling the ocean iron cycle 8 , 9 , significant uncertainties in the magnitude of the major processes remain 1 , 10 . This means that although shifts in Fe inventories may indeed drive end-of-century trends in simulated productivity across much of the global ocean 11 – 15 , confidence in model projections is diminished by the lack of mechanistic constraints on their behaviour. The ocean iron cycle is affected by an array of processes that interact together to set the dissolved iron concentrations in different parts of the ocean 16 . In the past decade, continental margins and hydrothermal vents have been acknowledged to augment dust deposition as important external iron sources 17 , 18 . Perhaps, most striking has been the recognition that the internal cycling of iron is typified by a range of biotic and abiotic transformations linked to Fe uptake (by both phytoplankton and bacteria), recycling, regeneration, scavenging and colloidal dynamics 10 , 19 . These processes act to shuttle dissolved iron between soluble and colloidal phases 20 – 22 and drive transitions of particulate iron between biogenic, lithogenic and authigenic (i.e., the residual particulate Fe not accounted for by lithogenic and algal biogenic pools) components 23 , 24 . Despite these new insights, the relative magnitude of regeneration and scavenging, and crucially, the realised rate of net regeneration, is unknown at the spatial and temporal scales of mode and intermediate water transport. In part owing to these missing constraints, global ocean models used to assess the response of ocean ecology, biogeochemistry and the carbon cycle to environmental change are free to tune their internal iron cycle with residence times that vary from a few tens to a few hundreds of years 7 . Newly expanded data sets of dissolved Fe (DFe) distributions from international ocean survey efforts within the GEOTRACES programme 25 , 26 should facilitate model improvement, but only if quantitative insights into the governing processes can be determined. A particular challenge is to disentangle the balance between biogeochemical and physical processes in setting nutrient levels in the oceans’ interior. For example, total phosphate (PO 4 ) at depth is made of up of two components: one associated with physical transport to depth (preformed PO 4 ) and the other from the regeneration of P from organic matter degradation (regenerated PO 4 ), which is quantified using apparent oxygen utilisation (AOU) 27 , 28 . A similar framework can be outlined for Fe, but Fe may be decoupled from P as it is affected by additional processes, such as extra Fe inputs onto intermediate water surfaces, unique regeneration of Fe, or Fe removal by scavenging 1 , 10 , 29 . Although scavenging of Fe will add complexity to the two-component model used for P, its magnitude remains an unknown quantity in observations. This lack of understanding is encapsulated by the evolving view of the ocean iron residence time from models and observations 7 , 30 , 31 . Here we use observations to quantify the large scale modification of DFe, benchmarked to PO 4 , within the mode and intermediate waters of the south Pacific Ocean for the first time, using AOU to derive the role played by physics, regeneration and scavenging. We focus on mode and intermediate waters as they support the majority of global productivity through nutrient supply to surface waters 6 . This approach illuminates a highly dynamic interior ocean Fe cycle, within which the commonly measured DFe pool is only a small residual component. Consequently, additional measurements of the ocean iron cycle pools beyond DFe and in particular, Fe fluxes are necessary to better constrain internal cycling and reduce uncertainty in global climate model projections.",
"discussion": "Discussion Our results point to continual removal of regenerated iron, resulting in only a small accumulation of DFe within intermediate waters. The combination of the constant rain of new material and the disaggregation of sinking particles in the ocean interior may be able to maintain scavenging of released Fe as the increasing surface area:volume ratio provides new surfaces for scavenging. Indeed, the increase in the flux of small particles (11–64 μm, equivalent spherical diameter, ESD) off Bermuda, and the concomitant opposite trend for large (>64 μm ESD) particles at depth 51 , highlights the important role this may play in producing small particles. Similarly, number spectrum analyses (using underwater video cameras) across the upper 200 m of the water column in the S. Pacific Gyre reveal much higher abundances of small particles than larger ones 52 . As scavenging of trace metals like Fe is highly dependent on surface area 53 – 55 , these particle disaggregation/fragmentation processes can catalyse further scavenging of the dFe released by regeneration. Scavenging of regenerated Fe into authigenic phases may also enhance particle sinking rates by increasing the specific gravity of particles (as noted for lithogenic Fe 56 ). These abiotic processes may act in concert with the removal of solubilised Fe by heterotrophic bacteria operating within particle microenvironments 57 , 58 . If we take our median estimated regeneration rate of dFe and the estimated accumulation rate of dFe (Table 1 ), and then combining these with a typical intermediate water layer thickness of 300 m at 10 °S, requires net downward removal fluxes of ~0.39 μmol m −2 d −1 . Although these fluxes would be inconspicuous in the measurements spanning ~0.4–10 μmol m −2 d −1 from trace metal clean sediment traps 44 , 45 , they are crucial in shaping the basin scale internal cycling of dFe in intermediate water layers. We observe a small, but significant, accumulation of DFe with time (Fig. 2b ), suggesting that the net regeneration quantified by the slope of the DFe versus AOU relationship integrates the balance between regeneration and scavenging fluxes. Observed concentrations of weak Fe-binding ligands are typically well in excess of DFe levels, which would imply an ample capacity to stabilise regenerated Fe at much higher levels 59 – 62 and is not in agreement with our analysis. However, the muted increase in DFe we observe is very consistent with the apparent saturation of strong Fe-binding ligands by DFe pools in the south Pacific Ocean 60 . This would imply that strong Fe-binding ligands, rather than their weaker counterparts, may play a key role in shaping the dissolved Fe distribution in the oceans’ interior. An additional role may be played by the interplay between soluble and colloidal iron pools, which can also be part of the ligand pools 20 – 22 and in the future it may be useful to compare the net regeneration derived from the DFe-AOU slope to observations of colloidal iron. Finally, we emphasise that the putative production of authigenic Fe from the DFe solubilised during regeneration, that we term here as scavenging, might not occur in the water column, but instead within particles and their associated microenvironments 57 , 58 in a manner disconnected from the wider water column ligand pool. The DFe-AOU slope of 2.7 μmol DFe mol AOU −1 from our analysis (Fig. 2b ) permits us to examine what proportion of the DFe pool might be controlled by the net interplay between regeneration and scavenging (termed ‘internal cycling’ hereon). It is well understood that roughly two-thirds of the interior PO 4 signal is preformed (controlled by physical transport), with the remaining one-third due to regeneration 27 , 28 . In contrast to PO 4 , the proportion of the DFe pool controlled by internal cycling in intermediate waters (within the 26.8–27.2 isopycnal layer) across the entire available GEOTRACES data set 26 of DFe and AOU has a median value of 0.57 (Fig. 3 ). This implies that over half of the DFe concentration in intermediate water is in fact set by internal cycling (i.e., the interplay between regeneration and scavenging), with the remainder controlled by physical transport of preformed DFe (either from the ocean surface or laterally). The stronger role played by preformed PO 4 than preformed DFe arises owing to the higher unused PO 4 levels in the, typically Fe-limited, watermass outcrop regions. Thus, because DFe is drawn down to very low levels in regions of intermediate water formation, internal cycling has a larger imprint on the interior DFe concentrations across much of the globe than for PO 4 . This view agrees with the lack of clear watermass signals in large scale ocean DFe sections 63 and is at odds with simulations from iron models with a dominant physically transported component. Fig. 3 Quantifying the origins of dissolved iron in the GEOTRACES IDP2017 dataset. The fraction of the dissolved iron concentration from the IDP2017 explained by the regeneration—scavenging balance between the σ 0 = 26.8–27.0 isopycnal layers is quantified here. The magnitude of the regeneration—scavenging balance (Fe REG' , mol m −3 ) can be derived by using the slope of the apparent oxygen utilisation—dissolved iron relationship from the P16 transect (2.7 μmol dissolved iron per mol apparent oxygen utilisation) and the independent apparent oxygen utilisation and dissolved iron data sets from the GEOTRACES IDP2017. The estimated net regeneration of dissolved iron (Fe REG' ) is then divided by the observed total dissolved iron from the IDP2017 to quantify the fraction explained by the regeneration—scavenging balance. The median value of 0.57 is indicated with a vertical dashed line. This indicates that over half of the observed dissolved iron is explained by the regeneration—scavenging balance, with less than half explained by ocean physical transport Overall, the strong mismatch we find between the internal basin scale Fe cycle fluxes and the residual DFe pool that accumulates from their interplay explains why Fe models can produce such divergent residence times while trying to reproduce the same dFe data sets. Our analysis finds DFe to be rapidly cycled by Fe supply and removal processes, which supports those models parameterised with short residence times. The net regeneration that shapes the multi-decade accumulation of DFe in intermediate waters is likely controlled by some combination of strong iron-binding ligands, colloidal dynamics, bacterial and authigenic iron pools. Because of the dominance of internal cycling, the concentration of Fe (relative to major nutrients N and P) and hence upper ocean iron limitation, will be strongly sensitive to small changes in the gross fluxes that govern the net regeneration of Fe. For instance, the Fe content of upper ocean phytoplankton is highly variable and fluctuations due to changing iron supply or phytoplankton species composition will affect the gross regeneration fluxes. Alternatively, biological and chemical transformations of particles, strong iron-binding ligands, bacterial demand and/or iron speciation will modify gross scavenging rates. Both these examples would change the net regeneration rate and hence the relative supply of Fe to the upper ocean microbes. Our isopycnal framework also provides a mechanistic methodology to assess ocean biogeochemical models more rigorously in future model evaluation efforts. A new generation of in situ processes studies 1 , tracking the evolution of Fe biogeochemistry by measuring both fluxes and particulate and dissolved Fe pools within a coherent physical framework would offer the potential to further constrain the internal cycling mechanisms for inclusion into global biogeochemical models. This improved mechanistic understanding of the ocean Fe cycle is required to reduce uncertainties in how changes in climate will affect surface ocean Fe limitation of primary productivity over much of the global ocean."
} | 3,623 |
30181557 | PMC6123422 | pmc | 8,405 | {
"abstract": "The process of assembling astutely designed, well-defined metal-organic cube ( MOC ) into hydrogel by using a suitable molecular binder is a promising method for preparing processable functional soft materials. Here, we demonstrate charge-assisted H-bonding driven hydrogel formation from Ga 3+ -based anionic MOC ((Ga 8 (ImDC) 12 ) 12− ) and molecular binders, like, ammonium ion (NH 4 + ), N-(2-aminoethyl)-1,3-propanediamine, guanidine hydrochloride and β -alanine. The morphology of the resulting hydrogel depends upon the size, shape and geometry of the molecular binder. Hydrogel with NH 4 + shows nanotubular morphology with negative surface charge and is used for gel-chromatographic separation of cationic species from anionic counterparts. Furthermore, a photo-responsive luminescent hydrogel is prepared using a cationic tetraphenylethene-based molecular binder (DATPE), which is employed as a light harvesting antenna for tuning emission colour including pure white light. This photo-responsive hydrogel is utilized for writing and preparing flexible light-emitting display.",
"introduction": "Introduction Charge-assisted hydrogen bond (CAHB) is a type of non-covalent interaction (X-H (+) ···Y (−) ) that plays an important role in the structure-property correlation of bio-macromolecules and in various biological molecular recognition processes 1 – 4 . CAHB is also widely employed in the construction of discrete organic cages, extended crystalline metal-organic architectures 5 , 6 and soft supramolecular gels 7 . The reason behind such versatility of CAHB is essentially its intrinsic strength (stronger than neutral X-H···Y bond) and directionality, that results in a wide range of materials with an array of exciting and complementary properties 8 . In this regard, CAHB driven self-assembly of predesigned metal-organic polyhedra (MOPs) 9 – 29 that are discrete metal-organic cages with confined cavities and large number of connecting sites, into soft supramolecular hydrogel is yet to accounted. Among different classes of MOPs, metal organic cubes (MOCs) with a general formula [M 8 L 12 ] x ( x = 0, n-), comprising metal ions (M n+ = Ni 2+ , Zn 2+ , In 3+ , Cr 3+ ) as vertices and imidazoledicarboxylate (L) as edges of a cube have been well explored 30 . Aesthetic appeal, structural modularity and robustness pertaining to the MOCs showed great promise for diverse applications 31 . MOCs are neutral ( x = 0) or anionic ( x = n-, a -MOC) depending upon the charge balance between M n+ and L 32 . MOCs are exploited as molecular building blocks by connecting the peripheral free carboxylate oxygens with metal ions or with H-bond donor molecules and the resulting extended structures showed potential applications in gas storage/ separation and proton-conductivity 33 – 35 . Since exteriors of a -MOCs are decorated with free polar carboxylate groups, they could be soluble in polar solvents, like water 36 . We envisioned that interaction of soluble a -MOCs with the positively charged or neutral, H-bond donor molecular binders through CAHB interaction could result in extended, supramolecular network 37 , 38 . In aqueous solution, such supramolecular assembly between a -MOC and different molecular binders could result in hydrogels. Aida et al. have shown that CAHB interaction between anionic clay nanosheets and dendritic molecular binders containing multiple guanidinium ions facilitated the cross-linking of the clay nanosheets and formed hydrogels 39 . Recently, Johnson et al. and Nitschke et al. reported the self-assembly of soluble polymers having coordinating end groups with metal ions, leading to the formation polymeric gel that consisted of in-situ formed metal-organic cage at the junction of cross-linked polymers 40 – 42 . Although their approach is inspiring, use of water soluble, preformed a -MOC as a platform to study self-assembly in the presence of different molecular binders is yet to be accounted. We envision that introduction of different molecular binders would tune the nano-morphologies and functionalities of the a -MOC-hydrogels. For example, the surface charge of the hydrogel-nanostructure could be altered by choosing appropriate cationic/anionic binders, making the hydrogel useful for chromatographic separation of oppositely charged species. Moreover, suitably designed chromophoric molecular binder would result in a processable soft luminescent hybrid hydrogel. The a -MOCs could also act as an excellent template for immobilizing the multi-chromophoric donors and acceptor binders for light harvesting application. In such system the emission property can be tuned and even processable high quantum efficiency pure white-light-emitting materials can be realized. In addition, stimuli responsive molecular binders can also provide the stimuli-responsive a -MOC-hybrid gels. With the introduction of photoactive molecular binders, photo-responsive hydrogel could also be prepared, which can enable writing on the flexible displays by change in corresponding photochemical reactions. Herein, we report synthesis of a Ga 3+ based metal-organic cube, ((Me 2 NH 2 ) 12 (Ga 8 (ImDC) 12 )·DMF·29H 2 O) ( 1 ), extended into three dimension through CAHB interaction between anionic [Ga 8 (ImDC) 12 ] 12− ( MOC ) and Me 2 NH 2 + (DMA) cations. Compound 1 is highly soluble in water, where discrete MOC s remain intact in solution. This particular phenomenon provides an opportunity to crosslink the MOC s with a wide range of molecular binders that lead to the formation of charge-assisted hydrogels (Fig. 1a ). Different molecular binders are assembled with MOC and the resulting hydrogels show different morphologies and properties (Fig. 1b ). When ammonium cation (NH 4 + ) is used as molecular binder, the resulting hydrogel with negatively charged, tubular nanostructures is exploited for gel chromatographic separation of positively charged species. We also extend the concept to form stimuli responsive luminescent hydrogel by rationally designing an aggregation induced emission (AIE)-active molecular binder containing tetraphenylethene (TPE) core. The photoresponsive behaviour of this hydrogel is further exploited for writing on flexible displays based on photo-cyclization of TPE core. Such photoresponsive behaviour of the hydrogel is also explored for tuning the excitation energy transfer from TPE segment to encapsulated acceptor dye. Finally, a pure white-light-emitting hydrogel with Commission Internationale de L’Eclairage (CIE) co-ordinates of (0.33, 0.32) is achieved. Fig. 1 Molecular binder-driven self-assembly of MOC to hydrogels. a Schematic representation of CAHB driven self-assembly of MOC with small molecular binders towards the formation of hydrogel. b Self-assembly of MOC with different molecular binders (NH 4 + cation, AEPD, gua.HCl and β-ala) results in hydrogels with nanotube, nanobouquet, nanosheet and nanocube morphologies, respectively",
"discussion": "Discussion In summary, we have prepared water soluble, Ga 3+ bases anionic MOC which self-assemble to hydrogel in presence of different molecular binders through charge-assisted H-bonding interaction. Depending upon shape and geometry of the molecular binders, the hydrogels show different morphologies, such as nanotube, nano- bouquet, nanosheet and nanocube. Moreover, the properties of the hydrogels are tuned by selecting the suitable binders. Here we have exploited two different properties of the MOC -based hydrogels. In one hand, the surface negative charge of the nanotubes of MOC-G1 is exploited for gel-chromatographic separation of cationic dyes from anionic dye. On the other hand, we have documented a photo-responsive luminescent hydrogel based on AIE active chromophoric binder that is further exploited for light harvesting application. In addition, we also prepared a white-light-emitting hydrogel by tuning the donor-acceptor energy transfer efficiency using photo-cyclization of TPE as a tool. In short, MOC -hydrogels provide a platform to integrate different type of molecular binders with MOC to form self-assembled nanostructures, where the properties and corresponding functionalities can be deliberately tuned."
} | 2,052 |
27400232 | null | s2 | 8,406 | {
"abstract": "There is an urgent need to improve agricultural productivity to secure future food and biofuel supply. Here, we summarize current approaches that aim at improving photosynthetic CO"
} | 45 |
37449409 | PMC10484736 | pmc | 8,408 | {
"abstract": "Abstract \n Corynebacterium glutamicum is an important industrial workhorse for production of amino acids and chemicals. Although recently developed genome editing technologies have advanced the rational genetic engineering of C. glutamicum , continuous genome evolution based on genetic mutators is still unavailable. To address this issue, the DNA replication and repair machinery of C. glutamicum was targeted in this study. DnaQ, the homolog of ϵ subunit of DNA polymerase III responsible for proofreading in Escherichia coli , was proven irrelevant to DNA replication fidelity in C. glutamicum . However, the histidinol phosphatase (PHP) domain of DnaE1, the α subunit of DNA polymerase III, was characterized as the key proofreading element and certain variants with PHP mutations allowed elevated spontaneous mutagenesis. Repression of the NucS-mediated post-replicative mismatch repair pathway or overexpression of newly screened NucS variants also impaired the DNA replication fidelity. Simultaneous interference with the DNA replication and repair machinery generated a binary genetic mutator capable of increasing the mutation rate by up to 2352-fold. The mutators facilitated rapid evolutionary engineering of C. glutamicum to acquire stress tolerance and protein overproduction phenotypes. This study provides efficient tools for evolutionary engineering of C. glutamicum and could inspire the development of mutagenesis strategy for other microbial hosts.",
"introduction": "INTRODUCTION DNA replication and repair is one of the most important biological activities for all living organisms ( 1 ). On one hand, the fidelity of DNA replication and repair maintains the genome integrity by faithfully passing the genetic information to the daughter cells. On the other hand, the intrinsic error rate provides low-frequency random mutagenesis for evolution, which has extensively been proven essential for pathogenesis and development of drug resistance ( 2 , 3 ). To accelerate genome evolution for generation of desirable industrially and environmentally relevant phenotypes such as overproduction and stress tolerance, a higher mutation rate than spontaneous mutagenesis is required to provide a larger size of mutational population ( 4 ). The sophisticated DNA replication and repair machinery that restricts the DNA mutagenesis has been targeted for such purpose. In Escherichia coli , DnaQ (the ϵ subunit of the DNA polymerase III complex) possesses proofreading function during DNA replication by means of its 3′-5′ exonuclease activity ( 5 ). The MutS-MutL-based mismatch repair (MMR) pathway detects and removes incorrect mismatched nucleobases in a post-replicative manner ( 6 ). Installing certain mutations in DnaQ or deactivating both mutS and mutL can lead to increased mutation rates, which has been harnessed for continuous and efficient evolutionary engineering of microbes including E. coli , Clostridium acetobutylicum and Lactobacillus casei ( 5 , 7–10 ). In some bacteria such as Streptomyces lividans and Mycobacterium tuberculosis , DnaE1 (the α subunit of DNA polymerase III complex), not DnaQ, ensures replicative fidelity by using its histidinol phosphatase (PHP) domain with 3′-5′ exonuclease activity ( 2 , 11 ). Introduction of certain mutations to the PHP and catalytic polymerization domains of DnaE1 increased the mutation rate and allowed random activation of silent biosynthetic gene clusters in S. lividans ( 11 ). \n Corynebacterium glutamicum is a Gram-positive bacterium of significant importance in industrial biotechnology. This microbe is now used for producing over six million tons of amino acids per year for food, feed and pharmaceutical applications, accounting for approximately 60% of the global amino acid market. As an industrial workhorse, its product portfolio is continuously expanding to other bulk and fine chemicals, including organic acids, proteins, polymers, biofuels, and plant nature products ( 12 ). The development of advanced genome editing approaches, including the CRISPR-based ones, have greatly enhanced the capability of rational genetic engineering of C. glutamicum ( 13–15 ). When implemented in industrial bioproduction, engineered microbes require robust cell growth and metabolism to adapt to various harsh industrial conditions, such as the toxic compounds from complex substrates ( 16–18 ). However, the complexity of biological systems makes it challenging to obtain robust microbes by rational engineering of a specific gene or pathway ( 19–21 ). Conversely, untargeted mutagenesis enables genome evolution to obtain a desired phenotype by generating a combination of many mutations. However, despite the significant industrial importance of C. glutamicum , the development of advanced genome evolution approaches for C. glutamicum has lagged behind other industrial workhorses such as E. coli . Unlike E. coli with extensive knowledge of the proofreading process, the key components responsible for proofreading of DNA replication in C. glutamicum have not been identified so far. Fortunately, the machinery for MMR in C. glutamicum has been characterized. C. glutamicum has no identifiable mutS or mutL homologue. Instead, this microbe and most Actinobacteria employ an EndoMS/NucS-based non-canonical MMR system ( 22–25 ). Disruption of the nucS encoding the mismatch-specific endonuclease increased the spontaneous mutation rate by >200-fold in C. glutamicum . However, deletion of nucS is unsuitable for practical genome evolution application because the genotypes encoding the desired phenotypes cannot be stably inherited due to the complete loss of MMR. In this study, we first identified the crucial proofreading element of DNA replication and the amino acid substitutions that impaired the proofreading function of DNA polymerase in C. glutamicum . The post-replicative MMR machinery NucS was also mutated and variants allowing increased spontaneous mutagenesis were screened. Via inducible overexpression of DnaE1 and NucS variants, a binary mutator that simultaneously interfered with the DNA replication and repair machinery was developed to achieve up to 2352-fold higher mutation rate than the wild-type strain. The developed mutators were applied for evolutionary engineering of C. glutamicum for stress tolerance and protein overproduction. This strategy holds promise for accelerating strain improvement of C. glutamicum and guiding the development of mutagenesis strategy for other microbial hosts.",
"discussion": "DISCUSSION Genome hypermutation strategies for model microorganisms such as E. coli and S. lividans have been well-established, allowing the acquisition of desired phenotypes decided by a combination of many mutations ( 8 , 10 , 11 , 60 ). In this study, to accelerate the genome evolution of C. glutamicum , a genome hypermutation strategy was developed for the first time via controllably interfering with the DNA replication proofreading and the post-replicative MMR processes. The unary mutators based on engineered DnaE1 or NucS variants could increase the spontaneous mutation rate of C. glutamicum by up to 505-fold. To obtain a stronger mutagenetic effect, a binary mutator was developed by simultaneously manipulating the DNA replication and repair machinery. Engineered DnaE1 and NucS variants were installed in a curable plasmid under a controllable IPTG-inducible promoter, which ensured the controllability and mutagenesis efficiency. With IPTG induction, the binary mutator displayed 2352-fold higher spontaneous mutation rate than the wild-type strain. The increase in spontaneous mutation rate is comparable with the reported E. coli (up to 2839-fold) and S. lividans (up to 1000-fold) mutators, and higher than the C. acetobutylicum (up to 258-fold), and L. casei mutators (up to 3.3-fold) ( 9–11 , 61 ). The binary mutator facilitated the acquisition of stress tolerance and protein overproduction phenotypes in a short-term genome evolution, demonstrating the practicality of this strategy in evolutionary engineering of C. glutamicum . A previous study on mutator-assisted evolution of E. coli suggests that acquiring different physiological traits may require the assistance of mutators with different mutagenetic strengths ( 8 , 10 ). The binary mutator with very strong mutagenetic effect can be mixed with the various unary mutators based on DnaE1 and NucS variants to be used as a mutator pool for future evolution applications. In the process of developing the mutator, we found that the PHP domain of DnaE1, not the DnaQ homologs, executed the task of proofreading during DNA replication. This observation in C. glutamicum is consistent with those in M. tuberculosis and S. lividans , which all belong to the phylum of Actinobacteria ( 2 , 11 ). Considering that the Actinobacteria numbers also possess a non-canonical mismatch repair pathway administered by NucS/EndoMS ( 22 ), we speculate the conserved mechanism for maintaining the DNA fidelity by the PHP domain-mediated DNA replication proofreading of DnaE1 and the NucS/EndoMS-mediated post-replicative MMR among Actinobacteria . However, the key amino acid residues that maintain the biological function of DnaE1 and NucS may vary in different hosts. Regarding DnaE1, aspartate to asparagine substitutions at D23 and D226 of M. tuberculosis DnaE1 ( 2 ) and at the corresponding residues of C. glutamicum DnaE1 impaired the proofreading function and significantly increased spontaneous mutation rate. Although the corresponding D19 residual is conserved in S. lividans DnaE1, individually mutating this residual did not affect its function and combination with C154G and D634E mutations in the PHP and catalytic polymerization domains was required ( 11 ). However, adding these two mutations in the best C. glutamicum DnaE1 D20R and DnaE1 D20K variants did not further increase the mutation rate but almost disrupt the mutagenetic effects of DnaE1 D20R and DnaE1 D20K . The inconsistent effects of amino acid substitutions at conserved residues of different DnaE1 enzymes were also observed for engineering of NucS enzymes from different microorganisms. The R42A and W75S mutations of P. abyssi NucS and the D165A mutation of T. kodakarensis NucS could inhibit mismatch recognition and cleavage reaction in the original hosts, respectively ( 51 , 52 ). However, adoption of these mutations in the C. glutamicum NucS did not increase the mutation rate but affected the stability of plasmid harboring these variants. Based on the understanding of the mechanism for maintaining DNA fidelity, we suggest that interfering with DnaE1 and NucS may be a universal strategy for developing genetic mutators for Actinobacteria . However, engineering of tailored DnaE1 and NucS variants is considered necessary for the target microbial hosts. We present a high-throughput screening pipeline, which consists of construction of site-saturation or full-length random mutation libraries, enrichment of desired variants based on rifampicin-resistant selection (or other phenotype-based high-throughput selection), and NGS analysis. Such pipeline should be useful for engineering of tailored DnaE1 and NucS variants for the microbial hosts of interest."
} | 2,825 |
26971463 | null | s2 | 8,410 | {
"abstract": "Protein domains and peptide sequences are a powerful tool for conferring specific functions to engineered biomaterials. Protein sequences with a wide variety of functionalities, including structure, bioactivity, protein-protein interactions, and stimuli responsiveness, have been identified, and advances in molecular biology continue to pinpoint new sequences. Protein domains can be combined to make recombinant proteins with multiple functionalities. The high fidelity of the protein translation machinery results in exquisite control over the sequence of recombinant proteins and the resulting properties of protein-based materials. In this review, we discuss protein domains and peptide sequences in the context of functional protein-based materials, composite materials, and their biological applications."
} | 202 |
28515720 | PMC5413776 | pmc | 8,411 | {
"abstract": "LuxR solos are unexplored in Archaea, despite their vital role in the bacterial regulatory network. They assist bacteria in perceiving acyl homoserine lactones (AHLs) and/or non-AHLs signaling molecules for establishing intraspecies, interspecies, and interkingdom communication. In this study, we explored the potential LuxR solos of Archaea from InterPro v62.0 meta-database employing taxonomic, probable function, distribution, and evolutionary aspects to decipher their role in quorum sensing (QS). Our bioinformatics analyses showed that putative LuxR solos of Archaea shared few conserved domains with bacterial LuxR despite having less similarity within proteins. Functional characterization revealed their ability to bind various AHLs and/or non-AHLs signaling molecules that involve in QS cascades alike bacteria. Further, the phylogenetic study indicates that Archaeal LuxR solos (with less substitution per site) evolved divergently from bacteria and share distant homology along with instances of horizontal gene transfer. Moreover, Archaea possessing putative LuxR solos, exhibit the correlation between taxonomy and ecological niche despite being the inhabitant of diverse habitats like halophilic, thermophilic, barophilic, methanogenic, and chemolithotrophic. Therefore, this study would shed light in deciphering the role of the putative LuxR solos of Archaea to adapt varied habitats via multilevel communication with other organisms using QS.",
"introduction": "Introduction Quorum sensing (QS) is a specialized behavior of microorganisms to coordinate their activities via cell-to-cell communication ( Miller and Bassler, 2001 ; Rutherford and Bassler, 2012 ). It is driven by various species-specific QS signaling molecules (QSSMs) like acylated homoserine lactones (AHLs), QS peptides (QSPs), autoinducer-2 (AI-2), diketopiperazines (DKPs), autoinducer-3 (AI-3), etc. ( Rajput et al., 2015 , 2016 ). During the process, QSSMs are synthesized and secreted out from the cells, which further sensed by it or other cells to continue the cascade ( Parsek and Greenberg, 2000 ; Kim et al., 2005 ). These QSSMs help microbial world to establish diverse vital processes propelled by QS like biofilm formation, secretion of various virulence factors, sporulation, motility, bioluminescence, and many more ( Nealson et al., 1970 ; Henrichsen, 1972 ; Costerton et al., 1978 ; Rumbaugh et al., 1999 ; Yarwood and Schlievert, 2003 ; Parsek and Greenberg, 2005 ; Li et al., 2011 ; Perez-Velazquez et al., 2016 ). Acylated homoserine lactones are characterized as the major signaling language for interaction among Gram-negative bacteria. It is processed by various homologs LuxI/LuxR type QS system in bacteria ( Miller and Bassler, 2001 ). Two important proteins, LuxI and LuxR, control the expression of luciferase operon ( luxICDABE ), and thereof are the key regulators of QS circuit. LuxI homolog protein is AHL synthase that catalyzes the reaction between S -adenosyl methionine (SAM) and an acyl carrier protein (ACP) to produce AHL molecules ( Rutherford and Bassler, 2012 ). While, LuxR-like proteins activates the transcription of the target DNA by binding to its cognate AHL molecule ( Schauder and Bassler, 2001 ). Moreover, a LuxR homolog protein comprised of two domains, i.e., N-terminal region (response regulatory domain) that binds to its specific autoinducer and C-terminal region with Helix-Turn-Helix (HTH) motif responsible for binding the DNA and hence modulates the expression of genes ( Donaldson et al., 1990 ; Hanzelka and Greenberg, 1995 ). LuxR proteins are categorized as canonical LuxR (possessing cognate LuxI) and LuxR solos (lacks cognate LuxI) ( Fuqua, 2006 ). LuxR solos (or unpaired LuxR or bachelor LuxR or orphan LuxR) are proved to sense both AHL and non-AHL molecules and hence termed as AHL or non-AHL binders ( Eberhard et al., 1981 ; Subramoni and Venturi, 2009 ; Hudaiberdiev et al., 2015 ). For example, LuxR of Vibrio fischeri ( Eberhard et al., 1981 ), TraR of Agrobacterium tumefaciens ( Zhang et al., 2002 ), LasR, RhlR, and QscR of Pseudomonas aeruginosa ( Passador et al., 1993 ; Pearson et al., 1995 ; Chugani and Greenberg, 2014 ), etc. belonged to AHL binders. Whereas PauR from Photorhabdus asymbiotica senses dialkylresorcinols (DARs), PluR of Photorhabdus luminescens recognizes α-pyrones ( Brameyer and Heermann, 2015 ), PqsR regulator of P. aeruginosa binds to 4-hydroxy-2-alkylquinolines (HAQs) ( Bala et al., 2013 ), etc. LuxI/LuxR based mechanism for QS is extensively explored in Gram-negative bacteria both experimentally and evolutionarily. Various studies regarding the phylogenetic distribution of LuxI/LuxR in alpha, beta, and gamma classes of Proteobacteria (Gram-negative bacteria) was accomplished ( Gray and Garey, 2001 ; Lerat and Moran, 2004 ; Nasuno et al., 2012 ; Christensen et al., 2014 ). Moreover, small subgroups of Gram-negative bacteria like Vibrionaceae ( Rasmussen et al., 2014 ), Roseobacteriacea ( Cude and Buchan, 2013 ), Halomonadaceae ( Tahrioui et al., 2013 ), Aeromonas ( Jangid et al., 2007 ), etc. were also surveyed. Additionally, the autoinducer-binding domain of LuxR solos was analyzed in bacteria on the basis of their distribution and conservation ( Subramoni et al., 2015 ). However, in Gram-positive bacteria (Actinobacteria phylum), few phylogenomic studies were done to check LuxR regulators’ phylogenetic and functional diversity (C-terminal, HTH DNA binding) ( Santos et al., 2012 ; Polkade et al., 2016 ). Previously, we have developed a database named SigMol, which encompasses information of all QSSMs reported in prokaryotes ( Rajput et al., 2016 ). Interestingly, in the database few species of Archaea was reported to exploit QS phenomenon. For example, Paggi et al. (2003) studied the presence of intraspecies communication in Natronococcus occultus through AHLs and showed their correlation with the production of extracellular proteases. Later on, FilI/FilR regulators were known to process carboxy-AHLs in Methanosaeta harundinacea strain 6Ac for cell assembly and carbon metabolic flux ( Zhang et al., 2012 ). Moreover, some archaea like Methanosarcina mazei, Methanothermobacter thermautotrophicus ( Zhang et al., 2012 ), Natrialba magadii ( Montgomery et al., 2013 ), etc. were also proved to perform cross-talk through QS. However, there is a huge gap in the experimental exploration of QS potential among archaea due to difficulties in culturing them. Archaea are often considered as “ extremophiles ” found in diverse environmental niche like halophilic, acidophilic, thermophilic, psychrophilic, piezophilic, deep-sea, etc. ( Chaban et al., 2006 ; Sorensen and Teske, 2006 ; Teske, 2012 ). Although, biofilm formation is also reported in Archaeal species like Methanosarcina mazei, Methanothermobacter thermautotrophicus ( Orell et al., 2013 ), Ferroplasma acidarmanus ( Baker-Austin et al., 2010 ), Sulfolobus spp. ( Koerdt et al., 2011 ), Halobacterium salinarum DSM 3754 T ( Frols et al., 2012 ), Ignisphaera aggregans ( Niederberger et al., 2006 ), Thermococcus litoralis DSM 5473 T ( Rinker and Kelly, 1996 ), and many more. However, Archaea are exemplified to exhibit biofilm mode of growth mostly via syntrophic interaction with bacteria further proved their active involvement in QS cascade ( Frols, 2013 ; Orell et al., 2013 ; Perras et al., 2014 ; Pohlschroder and Esquivel, 2015 ). Therefore, there is a need to explore the fundamental and vital phenomenon of QS in archaeal species, to uncover various aspect of multilevel communication (intraspecies, interspecies, and interkingdom). Despite, various bioinformatics resources available for QS like Quorumpeps ( Wynendaele et al., 2013 ), QSPpred ( Rajput et al., 2015 ), SigMol ( Rajput et al., 2016 ), etc., the attempts to explore QS mechanism computationally and evolutionarily in Archaea is lacking. To best of our knowledge, this is the first study that focused on investigating QS in archaea kingdom through multidimensional perspectives. We performed stepwise analyses to unveil the QS potential of LuxR solos in Archaea via their distribution, similarity with bacteria, functional characterization followed by correlation between taxonomy and ecological niche.",
"discussion": "Discussion LuxR solos are diversely distributed transcriptional regulators in bacteria known to play an important role to sense and respond to environmental cues ( Venturi and Ahmer, 2015 ). They are able to sense internal as well as external signals and helps in the adaptation of microbes despite absence of cognate LuxI ( Hudaiberdiev et al., 2015 ). However, they are well established to involve in QS among bacteria ( Patankar and Gonzalez, 2009 ). Although, till date, they are extensively explored in the bacteria kingdom but their role in Archaea is unexplored. Therefore, in the present study, we tried to explore their distribution in Archaea, the similarity with bacterial LuxR, functional characterization, evolutionary trend and ecological relatedness. Subramoni et al. (2015) searched InterPro database to find the putative LuxR solos proteins. These LuxR solos have ABD and DNA binding domain in bacteria. Likewise, Santos et al. (2012) explored LuxI/LuxR in Pfam database to fetch putative proteins and identified the LuxR regulators with HTH transcriptional factors that involved in QS. We have used the similar strategy to searched LuxI/LuxR in InterPro database and recognized 110 LuxR solos in Archaea that lack ABD and possess only DNA binding, HTH domain. LuxR solos, well known to be involved in QS were fully characterized and established in Gram-negative bacteria followed by Gram-positive bacteria ( Subramoni et al., 2015 ). While searching their presence in archaea, we found that their frequency is uneven among species; varies from single to maximum seven. Multiple copies of LuxR regulators found in different species, e.g., Haloferax spp. followed by Haloquadratum walsbyi (04), Halonotius spp. (03), Haloterrigena turkmenica (03), Pyrobaculum spp. (08), Halolamina sediminis (08), etc. These archaea thrive in the different extreme environment like high salt, high and cold temperature, high pressure, ammonia and sulfur enriched, etc. and drives various biogeochemical cycles like sulfur, nitrogen, and carbon. Most of the sequences are from halophilic ( Enache et al., 2007 ) archaea followed by thermophilic ( Jaakkola et al., 2016 ), piezophilic ( Vannier et al., 2011 ), methanogenic ( Borrel et al., 2013 ), alkaliphilic ( Xu et al., 1999 ), ammonia oxidizing ( Mosier et al., 2012 ), etc. More than 95% archaea are the inhabitant of aquatic (marine and fresh-water) ecosystem and rest belongs to terrestrial one. Oldest archaea with LuxR domain containing protein is isolated from stromatolites (∼3 billion years) and early cretaceous (∼123 million years) halite was Halococcus hamelinensis 100A6 ( Goh et al., 2006 ) and Halobacterium hubeiense ( Jaakkola et al., 2016 ), respectively. Our analyses revealed that LuxR solos of Archaea shared similarity with bacteria and able to perceive small molecules. Although, some domains are not exclusive to archaea but also found in bacteria like Transcription regulator (LuxR, HTH), DNA binding domain, Signal receiver, etc. ( Santos et al., 2012 ; Subramoni et al., 2015 ). Although, LuxR containing archaeal proteins explored in our study contains various type of domains that indicates the relationship of archaea in signal transduction and its response to wide range of environmental modulators as reported in bacteria ( Skerker et al., 2005 ). Moreover, LuxR based QS signaling is different in Gram-negative (single transcription factors) and Gram-positive (two-component system) bacteria ( Sturme et al., 2002 ). Domains repertoire extracted by our study belonged to one-component and two-component system that are found in Archaea and/or bacteria. Interestingly, we extracted domain from putative LuxR solos of Archaea that are involved in two-component system, which is reported to be acquired via HGT from bacteria ( Koretke et al., 2000 ; Ulrich et al., 2005 ). From scanned domains, MerR and HTH_1 are the exclusive key component of one-component system extracted from putative LuxR solos of Archaea ( Ulrich et al., 2005 ). Whereas domains like GerE, PAS, HTH, etc. are involved in both one-component and two-component systems ( Taylor and Zhulin, 1999 ; Galperin et al., 2001 ). Moreover, we also found archetypal signal input ( small molecules binding ) domains like PAS, GAF, CheY in putative LuxR solos of Archaea ( Galperin et al., 2001 ; Ulrich and Zhulin, 2010 ). HTH motif ( RGL[TS]XEE[IV]A[ED]AL[GD][IV]SRSTV[LS]EH ) of GerE domain present at C-terminal of LuxR proteins in bacteria is reported to involve in signal sensing or QS. Moreover, this motif is also reported in HMM logo in Pfam (PF00196) and sequence logo from PROSITE (PS50043) database as LuxR_HTH motif with their implication in QS. This motif ( Motif 1 ) is also present in putative LuxR solos of Archaea. However, the majority of the motifs are conserved according to the ecological niche. Interestingly, alignment results showed 10–25% similarity of Archaeal LuxR solos with bacteria, which are almost same as found among bacterial LuxR solos (18–25%) ( Subramoni et al., 2015 ). Furthermore, we also found substitution among invariant amino acids of ABD that displayed the diversity of LuxR solos to sense wide range to autoinducers (AHLs or non-AHLS). Gene Ontotology annotation studies showed that Archaeal LuxR solos involved in regulation of transcription through autophosphorylation of a histidine kinase and transfer the phosphate moiety to aspartate that further acts as a phosphor donor to response regulator proteins. However, they also possess sigma factor activity that aids them in making sequence-specific contacts with the promoter elements. Further, they are also annotated to be functional intracellularly in the cell as that of bacterial LuxR regulators ( Santos et al., 2012 ). However, GO-based functional assignment showed that Archaeal potential LuxR solos involved in signal sensing mechanism. Furthermore, to examine their role in QS, the ligand-binding prediction was performed. Although, the analysis showed that Archaeal LuxR solos are functionally characterized by the ability to bind AHLs and non-AHLs ligands as bacterial LuxR solos ( Subramoni et al., 2015 ). It was further supported by MSA, which displayed that among 06 conserved key residues ABD, 05 are found conserved in few Archaea LuxR proteins (Supplementary Table S4 ). The substitution among invariant amino acids indicates their potential to sense a wide range of signaling molecules (AHLs and/or non-AHLs). However, the presence of AHLs as signaling molecules in Archaea was already reported in SigMol database and various other studies ( Paggi et al., 2003 ; Zhang et al., 2012 ; Rajput et al., 2016 ). Additionally, the presence of non-AHL ligands like DKPs was also established in the previous study ( Tommonaro et al., 2012 ). However, other non-AHLs ligands like, α-pyrones, dodecanoyl-CoA, pyruvic acid, amino acids, metal ions, etc. showed their similarity with bacterial LuxR solos ( Patel et al., 2013 ; Brameyer et al., 2014 ; Brameyer and Heermann, 2015 , 2016 ; Venturi and Ahmer, 2015 ). Our analysis revealed that bacteria are remote homologs of Archaeal LuxR protein. The phylogenetic analyses of the LuxR solos protein of Archaea and bacteria showed that they both evolved separately with less substitution per site in archaea as compared to bacteria. However, analyses further confirm the presence of few cases for the transfer of LuxR copies between bacteria and Archaea through HGT. Moreover, placement of multiple LuxR solos copies in same Archaea like Haloterrigena spp., Natronomonas spp., Halonotius spp., Haloferax spp. both distantly with different microbial strains and with each other indicates that they are acquired through HGT and gene duplication events, therefore, possessing diverse ligand binding properties like the bacterial LuxR solos copies ( Subramoni et al., 2015 ). Our study is based on exploring the archaea for an imperative and fundamental phenomenon known as QS. All the analyses showed that Archaea LuxR solos could bind to AHLs and non-AHLs ligands and participate in QS. However, experimental details need to confirm the ligand specificity but difficulties in culturing the Archaea led this kingdom under-explored. Therefore, we used computational approach to explore the extent and functionality of Archaea against QS cascade. Varied computational analyses like similarity, functional characterization and evolutionary history showed their involvement in QS through AHLs and/or non-AHLs ligands. Moreover, potential ecological niche of archaea was collated from literature and correlated with the outcome of our analyses for better understanding for the trend of QS being exploited via extremophiles. However, the extent of the diversification for QS in archaea is still a question that needs to be further explored. Simultaneously, the evidence reported in the literature for the occurrence of dominant microbial lifestyle, i.e., biofilms in archaea mostly via syntropic interaction with bacteria strengthen our findings that these extremophiles have capabilities perform intraspecies, interspecies, and even interkingdom cross-talks and thrive extreme environment through QS."
} | 4,387 |
38784195 | PMC11109602 | pmc | 8,412 | {
"abstract": "The utilization of industrial biomanufacturing has emerged as a viable and sustainable alternative to fossil-based resources for producing functional chemicals. Moreover, advancements in synthetic biology have created new opportunities for the development of innovative cell factories. Notably, Yarrowia lipolytica , an oleaginous yeast that is generally regarded as safe, possesses several advantageous characteristics, including the ability to utilize inexpensive renewable carbon sources, well-established genetic backgrounds, and mature genetic manipulation methods. Consequently, there is increasing interest in manipulating the metabolism of this yeast to enhance its potential as a biomanufacturing platform. Here, we reviewed the latest developments in genetic expression strategies and manipulation tools related to Y. lipolytica , particularly focusing on gene expression, chromosomal operation, CRISPR-based tool, and dynamic biosensors. The purpose of this review is to serve as a valuable reference for those interested in the development of a Y. lipolytica microbial factory.",
"introduction": "1 Introduction The study of oleaginous microorganisms has garnered significant interest due to its application to produce valuable fatty acids and derivatives [ 1 ]. Moreover, the resulting biodiesel derived from them are particularly important in terms of promoting clean energy while reducing the pollution associated with fossil fuels. Because of these benefits, oleaginous microorganisms are considered a highly promising option for sustainable renewable oil production. Among them, Yarrowia lipolytica is the most extensively studied, which possesses desirable qualities, such as high lipid content, robust cell growth, and compatibility with various substrates [ 2 ] (see Table 1 , Fig. 1 , Fig. 2 , Fig. 3 ). Table 1 Summary of synthetic biology tools in Yarrowia lipolytica . Table 1 Tools Characteristics Application References Promoter pTEF, pMnDH2, pPHO89 Endogenous promoters; constitutive; strong / [ 13 , 17 ] hp4d, nUAS1 XPR2 -LEU, nUAS1 XPR2 -TEF Hybrid promoters; derived from pXPR2; carries several tandem copies of UAS1 XPR2 / [ 21 , 22 ] hybrid RNA polymerase III promoters Hybrid promoters Improve sgRNA expression and CRISPR-Cas9 function [ 71 ] pXPR2 Inducible promoters; induced by peptone / [ 23 ] pEYK1 Inducible promoters; induced by erythritol and erythrulose / [ 26 ] pMT-1 to pMT-6 Inducible promoters; induced by Cu 2+ / [ 28 ] Terminator XPR2t, LIP2t, PHO5t Endogenous terminators Commonly utilized for the heterologous gene expression [ 22 ] Synth1t-synth30t Synthetic terminators; short; easily cloned Improve expression of heterologous genes [ 32 ] Multi-gene assembly One-step assembly Obtain multiple expression cassettes by overlap extension PCR (OE-PCR); simple; quick Integrate the β-carotene biosynthetic pathway into Y. lipolytica chromosome [ 37 ] Golden Gate assembly Rely on Type IIS restriction endonucleases; high efficiency; stable Assemble carotenoid pathway genes and improved the efficiency up to 90% [ 38 ] YaliBricks assembly Based on BioBrick assembly; rapid multi-component assembly Construct five-gene violacein pathway [ 15 ] Gene deletion Cre-loxP Sourced from the P1 phage; composed of cyclized recombinase (Cre) and loxP sites Integrated a flavonoid pathway into Y. lipolytica genome, and obtained different flavonoids [ 56 ] TALENs Recombinant restriction enzymes; fusion of the nuclease to the TAL effector DNA binding domains Generate mutants of the fatty acid synthase (FAS) gene [ 84 ] CRISPR tools CRISPR/Cas9 Composed of a Cas9 protein and the corresponding sgRNA Multi-gene targeting and marker-free integration [ 51 , 65 , 85 , 86 ] CRISPRi Used for gene repression via a catalytically deactivated Cas9 Repress NHEJ to enhance HR efficiency [ 70 ] CRISPRa Fusing dCas9 to transcriptional activators Activate the target genes [ 87 ] Genetic biosensors Fatty acyl-CoA biosensor Fatty acids as response factors; transcription factor FadR and manipulator fadO were derived from E. coli Regulate the cytochrome P450 enzymes that convert palmitate to ω-hydroxypalmitate [ 88 ] Naringenin biosensor Naringenin as response factors; flavonoid-sensing transcriptional activator FdeR; manipulator fdeO Improved cell fitness and pathway yield [ 79 ] Xylbiosensor Xylose as response factors; the activation factor XylR and the operator xylO were derived from E. coli Modulate naringenin synthesis with a yield of (715.3 ± 12.8) mg/L [ 80 ] Light-controlled biosensor Light as the response factor; fast response; non-destructive Application to the dynamic regulation of the biosynthesis and synthetic pathways of coumaric acid and naringenin [ 82 ] Fig. 1 Genetic technology applicable in Yarrowia lipolytica. A. Gene expression and multi-gene assembly strategies, including promoter and terminator engineering, plasmids expression system, and multi-gene assembly operation. B. Genomic chromosomal operations, including integrated expression, gene deletion, and CRSPR tools. C. The modification, performance, and application genetic biosensors. HR, homologous recombination. NHEJ, the nonhomologous end-joining. TF, transcription factor. Fig. 1 Fig. 2 The CRISPR/Cas genome editing platform for Yarrowia lipolytica. A. CRISPR/Cas9 method for gene knock-out/knock-in. When the sgRNA recognizes the targeted sequence, which is located before a protospacer adjacent motif (PAM) site, the Cas9 protein will catalyze the formation of a double-strand break (DSB) in the targeted DNA. B. CRISPR/dCas9 based gene editing. C. CRISPR/Cpf1 based gene editing. D. CRISPR/dCpf1 based gene editing. E. CRISPRi and CRISPRa methods for gene interference and activation. A catalytically deactivated Cas9 (dCas9), which has no cleavage activity, can be fused with different effector domains to control gene expression. When the targeted region is recognized, the dCas9 fusion protein with the transcriptional repressor domain binds the DNA to repress gene expression. Similarly, the fusion protein of dCas9 and the transcriptional activator domain binds to targeted regions to improve the gene expression level. CRISPRa, CRISPR/Cas based gene activation; CRISPRi, CRISPR/Cas based gene interference; crRNA, CRISPR RNA; DSB, DNA double strand breaks; PAM, protospacer adjacent motif; RNA P, RNA polymerase; sgRNA, single-guide RNA; TF, transcription factor. Fig. 2 Fig. 3 Dynamic regulation of gene circuits through biosensors in Yarrowia lipolytica. A. Transcription factor-based biosensor. Repressor-based biosensors. TF suppresses the expression of target gene expression. Activator-based biosensors. TF activates the expression of a target gene in the presence or absence of the target metabolite. The solid red line indicates inhibition and the green realized arrow indicates activation. B. Characterization of changes in the target metabolite concentration. C. Application of TF-based biosensor in Y. lipolytica. Fig. 3 Noticeably, Y. lipolytica is a non-conventional oleaginous yeast that holds the generally regarded as safe (GRAS) status [ 3 ]. It exhibits unique biochemical and metabolic characteristics, such as efficient acetyl-CoA metabolic pathway, the high flux of TCA cycle, and remarkable lipid accumulation, distinguishing it from Saccharomyces cerevisiae [ 4 , 5 ]. Y. lipolytica also possesses the ability to utilize a diverse array of low-cost, renewable substrates, including alkanes, fatty acids, organic acids, and proteins [ [6] , [7] , [8] ]. These distinguishing features make Y. lipolytica an ideal candidate for the biomanufacturing applications. In particular, strains W29 (CLIB89) and its derived strains Po1d, Po1f, Po1g, and Po1h, have been commonly employed as platforms for engineering research and industrial applications [ 9 , 10 ]. Moreover, noteworthy advantages of these strain series include: i) high levels of protein expression and secretion; ii) efficient utilization of inexpensive carbon sources; iii) the elimination of the endogenous alkaline extracellular protease to safeguard the degradation of expressed exogenous proteins. Currently, with the rapid development of synthetic biology, various innovative methods and strategies have been successfully implemented for gene regulation in Y. lipolytica . Moreover, genome editing techniques, such as Cre/loxP and CRISPR, have been effectively developed for use in Y. lipolytica . These genetic tools and strategies enable researchers to optimize cellular performance and confer the ability to synthesize novel chemicals. In this review, we emphasized the genetic manipulation tools and strategies developed in Y. lipolytica , including gene expression, chromosomal operation, CRISPR-based tools, and dynamic biosensors for metabolic engineering. Additionally, we discussed the limitations and challenges that need to be overcome, and explore emerging opportunities for Y. lipolytica in the context of synthetic biology and industrial applications."
} | 2,256 |
24894952 | PMC4189316 | pmc | 8,414 | {
"abstract": "The global demand for food, feed, energy, and water poses extraordinary challenges for future generations. It is evident that robust platforms for the exploration of renewable resources are necessary to overcome these challenges. Within the multinational framework MultiBioPro we are developing biorefinery pipelines to maximize the use of plant biomass. More specifically, we use poplar and tobacco tree ( Nicotiana glauca ) as target crop species for improving saccharification, isoprenoid, long chain hydrocarbon contents, fiber quality, and suberin and lignin contents. The methods used to obtain these outputs include GC-MS, LC-MS and RNA sequencing platforms. The metabolite pipelines are well established tools to generate these types of data, but also have the limitations in that only well characterized metabolites can be used. The deep sequencing will allow us to include all transcripts present during the developmental stages of the tobacco tree leaf, but has to be mapped back to the sequence of Nicotiana tabacum . With these set-ups, we aim at a basic understanding for underlying processes and at establishing an industrial framework to exploit the outcomes. In a more long term perspective, we believe that data generated here will provide means for a sustainable biorefinery process using poplar and tobacco tree as raw material. To date the basal level of metabolites in the samples have been analyzed and the protocols utilized are provided in this article.",
"introduction": "Introduction Population and economic growth have caused an increasing demand for food, water and fuels. Much of these supplies are produced, processed and transported using finite fossil-based means, such as petroleum. It is, however, clear that this practice is not sustainable, and the development of alternative resources will therefore be of great importance 1 . Many renewable resources are, to varying degrees, currently being exploited, including wind, water motion, solar, geothermal, and wave based energy sources. Another sustainable and largely untapped resource is the biomass from plants. This resource also offers a very cost efficient way to convert solar derived energy into fuels 2 . Apart from providing bio-based fuel, the plant biomass also offers unique opportunities for alternative products, including plastics, detergents, and valuable chemicals. The plant cell wall, which largely consists of sugar based polymers, makes up the main bulk of the plant's biomass and much effort is currently being invested in its efficient conversion into bioethanol. The remaining biomass may subsequently be processed into biogas and oil related products 3 . Much of the perennial plant species, including grasses and trees, that produce large amounts of cellulosic biomass typically grow best in the temperate zones. However, approximately 20% of the land area is semi-arid, and is therefore also prone to droughts 4 . Obviously, it would be of interest to also cultivate these arid lands with plants that could effectively contribute to the sustainable production of energy and material. These plants need to have an optimal water use efficiency and drought resistance and would include the tobacco tree ( Nicotiana glauca ) and species from the Agave genus. The MultiBioPro consortium aims to implement an integrated biorefinery pipeline, using the two important crop species, poplar and tobacco tree. Poplar has emerged as a promising biofuel crop as it is fast growing, easily clonally propagated and highly adaptable to a wide range of climatic and soil conditions. It also provides a wide range of wood, fiber, fuel wood, and other forest products 5 . The tobacco tree has also emerged as a suitable plant for biofuel and biorefinery purposes. It typically produces substantial quantities of biomass, contains high amounts of nonstructural carbohydrates 6 , and also has the rare ability to accumulate large quantities of readily extractable nonfood oils (including long chain C 29 -C 31 saturated hydrocarbons and triterpenoids) that are suitable for biodiesel production. The tobacco tree is, moreover, amenable to genetic improvement, has high sprouting capacity, and grows happily on semi-arid soils not used for food production. It therefore appears that both poplar and tobacco tree have intrinsic potential for multipurpose crops, i.e. as new high value feedstocks for an integrative bio-based industry. In this paper we focus on diverse set of approaches to discern how tobacco tree deposits long chain hydrocarbons. In an attempt to identify the underlying molecular machinery responsible for the production and secretion of the saturated long-chain hydrocarbons on tobacco leaves, we apply modern “omics” based technologies. This includes RNA seq of a developmental leaf series (ten stages), and multiplatform metabolite profiling approaches using LC- and GC-MS (for polar and nonpolar metabolites and lipidomics). These data will be used to mine for gene expression that correlates with, or precedes, the onset of biosynthesis of the molecules indicated above. Genes and pathways that appear promising from these endeavors will be used for functional testing in the model species Arabidopsis and could ultimately be amenable for biotechnological engineering in tobacco tree.",
"discussion": "Discussion The protocols presented here provide a comprehensive framework to analyze tobacco tree leaves for metabolites and transcripts. It is envisaged that these combined efforts should provide us with new insights into the processes underlying the synthesis and extrusion of the hydrocarbons and the high value compounds present in this tissue. These approaches should therefore give us a better understanding for how the compounds are being synthesized. In addition to the tobacco tree aspects of the work, it is also aimed to improve poplar biomass, especially targeting lignification of the secondary wall structure, but also to explore whether we can use the bark for extraction of valuable compounds. The methods presented in this paper are slight modifications of standardized methods for metabolite profiling. These methods are of course limited to known metabolic profiles, and it is possible that several new metabolic peaks may be obtained for which no compound is known. We hope to put these compounds in context to other metabolites by combining the behavior of metabolites and transcripts over the developmental time series. None of the methods presented here are significantly changed from methods typically used for plant materials. The interesting aspect lies in the combination of methods to understand the underlying framework for mainly long chain hydrocarbon production and modification in the tobacco tree leaves. One of the critical steps for obtaining this information is the subsequent combination of the different data types. We envision that the data as a first evaluation will be divided into different clusters based on the behavior of the metabolites/transcripts over the development and that these data will be used to infer transcript vs metabolite behaviors, and also to potentially assign certain metabolites to pathways. In addition, more elaborate network-based analyses are then envisioned to exploit causal relationships. The analytical protocols presented here will also provide a basis for field-trials and industrial exploitation of the biomass. To accomplish this, the MultiBioPro consortium contains several industrial partners that have the abilities to further explore the biomass, with the aim to deliver biodiesel, bioethanol and other high-value compounds. These types of biomass exploitation will be assessed based on; (1) testing the robustness and quality of the bio products produced (typical industry standard tests will be carried out to ensure the products generated have good market value), (2), an economic, social and environmental evaluation of the technologies will be performed using literature sources, interviews and material that is generated during field trials and pilot plant biorefinery assessments. These activities will include cost benefit and life cycle analysis, the generation of an environmental dossier and market and business strategies. We believe that this pipeline will become a useful blend of academia, applied science and industrial exploitation to further poplar and tobacco tree biomass for consumer end-products."
} | 2,095 |
35479176 | PMC9032048 | pmc | 8,419 | {
"abstract": "The demand for highly flexible and self-powered wearable textile devices has increased in recent years. Graphene coated textile-based wearable devices have been used for energy harvesting and storage due to their outstanding mechanical, electrical and electronic properties. However, the use of metal based nanocomposites is limited in textiles, due to their poor bending, fixation, and binding on textiles. We present here reduced graphene oxide (rGO) as an n-type and conductive polymer poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) as a p-type material for a wearable thermoelectric nanogenerator (TEG) using a (pad–dry–cure) technique. We developed a reduced graphene oxide (rGO) coated textile-based wearable TEG for energy harvesting from low-grade human body heat. The conductive polymer (PEDOT:PSS) and (rGO) nanocomposite were coated using a layer by layer approach. The resultant fabric showed higher weight pickup of 60–80%. The developed textile based TEG device showed an enhanced Seebeck coefficient of (25–150 μV K −1 ), and a power factor of (2.5–60 μW m −1 K −1 ). The developed TE device showed a higher potential to convert the low-grade body heat into electrical energy, between the human body temperature of (36.5 °C) and an external environment of (20.0 ± 5 °C) with a temperature difference of (2.5–16.5 °C). The wearable textile-based TEG is capable of producing an open circuit output voltage of 12.5–119.5 mV at an ambient fixed temperature of (20 °C). The rGO coated textile fabric also showed reduced electrical sheet resistance by increasing the number of dyeing cycles (10) and increased with the number of (20) washing cycles. The developed reduced graphene oxide (rGO) coated electrodes showed a sheet resistance of 185–45 kΩ and (15 kΩ) for PEDOT:PSS–rGO nanocomposites respectively. Furthermore, the mechanical performance of the as coated textile fabric was enhanced from (20–80 mPa) with increasing number of padding cycles. The thermoelectric performance was significantly improved, without influencing the breath-ability and comfort properties of the resultant fabric. This study presents a promising approach for the fabrication of PEDOT:PSS/rGO nano-hybrids for textile-based wearable thermoelectric generators (TEGs) for energy harvesting from low-grade body heat.",
"conclusion": "5. Conclusions The fabrication of rGO coated cotton fabric was successfully developed using the industrial-scale pad–dry–cure method. The developed highly flexible and washable textile-based thermoelectric generator coated with PEDOT:PSS–rGO nanocomposites is highly efficient and suitable for energy harvesting from human body heat. The results demonstrate that the graphene and conductive polymer nanocomposite coated textile-based TE nanogenerator showed an improved TE performance with improved Seebeck coefficient ( S ), power factor with increasing the content percent of rGO in PEDOT:PSS–rGO nanocomposites. The results show that the Seebeck coefficients higher than 60 μV K −1 was increased with increasing the number of dyeing cycles and the power factor was also significantly improved to 150 μW m −1 K −2 , with a temperature gradient of 16.5 °C (290–309 K) for PEDOT:PSS/rGO nanocomposite coated TE device. The output open circuit volts reached to 19.5–120 mV and 2.5–75 mV over a thermal difference between external and body temperature for series and parallel arrangements. This study demonstrates that fabrication of rGO and PEDOT:PSS/rGO nano-hybrids as thermoelectric materials on textile substrate by using pad–dry–cure method is as an effective approach for the development of thermoelectric textiles.",
"introduction": "1. Introduction The development of graphene coated textile based wearable energy harvesting and storage devices is highly anticipated with low grade body heat as a green and sustainable technique. The demand for such highly flexible and breathable wearable e-textiles is increasing due to their use in energy storage, conversion, and harvesting devices. 1 The human body is a highly stable form of energy source for energy harvesting, which can be used for wearable health monitoring devices. 2 We used a new approach in mass scale fabrication of graphene based wearable and washable e-textiles for future research in which the thermoelectric fabric is integrated as an energy harvesting device from human body heat. 3 The developed textile may replace metal based wearable textiles with advanced electronic and electrical properties and may provide a new window for the development of novel technologies which can be used for self-powered health monitoring devices, disease diagnostics, and prevention. 4 In brief, the TE material should possess an improved Seebeck coefficient S , as well as higher electrical conductivity σ . Whereas, the low thermal conductivity κ 5 is good for TE materials to operate a steady level of heat flow into electrical energy; because a high-performance thermoelectric material may hold a large thermal gradient as a temperature difference required to keep in the output electric potential difference. 6 The Seebeck voltage, is attained even though with a small temperature difference which is near to room temperature, even human body temperature. The TE parameters, σ and S are reciprocal to an electrical conductivity, 7 whereas the σ and κ e are directly related with each other. In connection of relationship to these invariable factors to enhance the dimension less figure of merit ZT , which is highly challenging to be improved. 8 On the other hand, the thermoelectric power factor is addressed as PF = S 2 σ , in which assess the concomitant effect of σ and S on TE performance. 9 Compared to inorganic counterparts, organic TE devices emerged as the potential candidates work at room-temperature and flexible (even wearable) TE power generation. During last few decades, extensive studies have been performed on the p- and n-type materials and devices to build up the inter-relationship among the TE parameters ( i.e. , electrical conductivity, Seebeck coefficient, thermal conductivity and power factors), demonstrating a great potential of organic TEs. 10 Hence, the use of organic materials such as carbon and its derivatives are highly anticipated for multifunctional finishing of textiles. The demand of carbon-based allotropes such as fullerene, carbon black (CB), carbon nanotubes (CNT), graphite oxide (GrO), graphene oxide (GO), reduced graphene oxide (rGO) and graphene is increasing in recent years for wearable flexible, bendable, breathable and washable electronic devices. 11,12 The use of conductive polymers with graphene and its derivatives is also highly attractive and considered as potential materials as compared to traditional metal-coated e-textiles. Therefore, the use of an organic conducting polymers such as: poly(3,4-ethylenedioxythiophene) (PEDOT) EDOT:FeCl 3 , 13 EDOT:FeToS, 14 polyaniline (PANI), 15 polypyrrole (PPy), 16 and poly(3-hexylthiophene) (P3HT) 17 are considered as promising materials for organic TE devices. The improved TE performance in these conducting polymers based of thiophene polymer poly(3,4-ethylenedioxythiophene) (PEDOT) has been reported. The intrinsic conductive polymer showed an enhanced TE performance with PF of 469 μW m −1 K −2 and dimension less figure of merit ZT value of 0.42 at room temperature when doped with poly(styrenesulfonic) (PSS) in conductive polymer. 18 Among all these conducting polymers, polyaniline (PANI) is also well known and most widely used polymer which have an electrical conductivity σ of 105 S m −1 , as comparable to the state-of-the-art for all inorganic TE materials. 19 The polymer is considered as highly environmentally stable and dipped with graphene, graphite, carbon black and CNTs have easy and high dissolution in PANI, which make them more easier to be dispersed in PANI. 20 As well as reported in study, that these polymer develops strong bonding and exists a strong π–π interaction between the polymer PANI intermolecular and intramolecular with carbon based derivatives. 21 The addition of these filler which would allows stable and enhanced growth of molecular chains for graphene in an ordered polymer matrix with PANI polymer chains with surface of carbon materials. 22 The another conductive polymer including polythiophene (PTh) showed a low Seebeck coefficient ( S ) and less electrical conductivity σ . The TE performance is significantly influenced by the size, shape, crystal growth, and orientation of the side chains with main chain structure in the nanocomposites. 23 The overall TE performance obtained with this polymer is nearly exhibiting a high S of ∼130 and 76 μV K −1 , σ ∼ 47 and 73 S cm −1 , under lower thermal conductivity κ of 0.17 and 0.15 W m −1 K −1 , respectively at room temperature. 24 The thermoelectric properties in fine-tuned by doping or de-doping of the polymeric materials, using various organic and inorganic filler, in which the Seebeck coefficient was significantly improved whereas, on the other hand the electrical conductivity of the developed nanocomposites was decreased, similar to attributes of inorganic materials. 24 However, as compared to inorganic thermoelectric materials, the intrinsic conductive polymers such as PEDOT:Tos films showed relatively less thermal conductivity of 0.2–0.25 W m −1 K −1 . On the other-hand, the power factor was also raised from 38 to 324 μW m −1 K −2 , with a figure of merit ( ZT ) value of 0.25 at room temperature. 25 The carrier concentration improved the filtering effect and showed a favorable TE results which contributed as higher Seebeck coefficient of (163 μV K −1 ) and power factor of (70.9 μW m −1 K −2 ) during doping in the conductive polymers i.e. poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate). 26 Familiarly, the development of inorganic doped in conductive polymer (PEDOT:PSS)/Te) introduced the Te nanorods in the nanocomposites with improved interfaces. 27 Similarly, the nanocomposites of PANi/graphene–PEDOT:PSS/PANi/DWCNT–PEDOT:PSS has been fabricated using layer-by-layer (LBL deposition) on different textile substrates. 28 The conductive polymer PEDOT:PSS was used developed nanocomposites using some surfactants i.e. (SDS) (SDBS) to disperse and stabilize the graphene and CNTs dispersion. The resultant nano-composite showed an electrical conductivity σ of 190 S cm −1 , Seebeck coefficient ( S ) of 120 μV K −1 and power factor PF of 2710 μW m −1 K −2 . Their results shows that, the best inorganic TE material Bi 2 Te 3 exhibits improved TE performance at room temperature. 29 The another work discuses, that the higher increase in carrier mobility efficiently increased the electrical conductivity σ , and reduced the TE performance in terms of Seebeck coefficient and power factor by the reduction of the carrier concentration with a decreased Seebeck coefficient, as well as favorable raising in the PF of 220 μW m −1 K −2 , which is twice higher magnitude than the PANI–CSA (∼1.8 μW m −1 K −2 these results are reported in previous studies performed on PANI nanocomposites. 30 The another study made on the optimization of TE performance in terms of Seebeck coefficient, PF for the PANI–CNT based nanocomposites. The results demonstrated that the TE performance was improved as compared to the pristine CNT and graphene film. Another study, researchers developed highly conductive polymer polyaniline (PANI)/with (SWCNTs) nanocomposites using in situ polymerization and template free approach for TE devices. 31 It was found that the solution processing and strong π–π interactions between the PANI and SWCNTs induced the PANI molecules to form a highly ordered structure. This improved degree of order of the PANI molecular arrangement increased the charge carrier mobility and thereby enhanced the electrical\ntransport properties of PANI. The PANI/SWCNT (65 wt%) composite films (10 μm) exhibited an electrical conductivity, Seebeck coefficient, and power factor of 1.44 × 10 3 S cm −1 , 39 μV K −1 and 217 μW m −1 K −2 respectively at room temperature. 32 This PF was more than 20 times the PF of pure PANI film. The improvement in TE properties is attributed to the highly ordered structure of PANI chains along the SWCNT via strong π–π interaction, 33 which turned into increased the carrier mobility. The quantum Hall effect measurement showed that the charge carrier mobility improves with increasing of concentration, while the charge carrier mobility increased three times with the increasing the content percent of CNT in the polymer nanocomposite. Furthermore, the thermal conductivity out of plane was very low of 0.2–0.5 W m −1 K −1 at room temperature. 34 Such polymer based composites are generally fabricated using in situ , chemical oxidation, vapor phase polymerization, and chemical exfoliation, 35–37 whereas the filler are widely mixed with different organic solvents, reducing agents, dispersing agents and stabilizers. 38 The features of the intrinsic conductive polymers are not only limited to their potential electrical and thermal properties but due to their light weight, flexibility and higher stability against water, and air makes them highly suitable for super capacitors, batteries, solar cells, and other energy storage and harvesting devices. 39 Furthermore, chemical functionalization and in situ solution processed polymer composite are commonly used including screen printing, ink jet printing dispenser printing; spray printing, and spin coating, 40,41 as an efficient, cost-effective method to fabricate TE devices on mass scale. The use of inorganic metals such as silver, copper, and gold based wearable devices become limited due to their higher rigidity, stiffness, and decomposition when exposed to water and air. 42 These metal-based electronic devices need intense care during fabrication and use as heavy batteries embedded in clothes are less flexible, highly stiff, non-biodegradable and toxic to human skin as well environment. 43 The use of graphene with these conductive polymer coated e-textiles has been used for various applications including super capacitors, batteries, pressure sensors, thermoelectric, triboelectric, piezoelectric, and nano-generators. 44–46 However, the production of graphene-based highly flexible, breathable and stretchable thermoelectric devices is challenging and need further improvements for mass production. The synthesis and fabrication of such a highly conducive, breathable, and washable textiles is complicated and limited due to their production on a commercial scale. The widely used fabrication processes are complicated, time-consuming and expensive. Some of these techniques include chemical, and thermal reduction of reduced graphene oxide (rGO) coated fabrics has been reported in the previous studies. 47 Thermal reduction is achieved at higher temperature range of 150–200 °C which is more costly and time-consuming. Whereas, synthetic fibers such as polyester, acrylic, Lycra, and spandex are susceptible to excessive heat, which results in deformation of the polymer chains at higher temperature range above glass transition temperature ( T g ) 160–200 °C. 48 Fabrication of graphene and other carbon-based materials require chemical and thermal treatment with strong reducing agents such as, hydrazine hydrate (HH), hydro iodic (HI), sodium borohydrate (NaBH 4 ). 49 The use of these reducing agents is insecure and limited, as wearable textile directly contacts with human skin can induce skin-irritation, and other allergic reactions. 50 The natural fibers such as cotton, jute, flex and hemp fibers lose strength due to the polymer degradation as their basic monomer is cellulose. The use of highly efficient and green reducing agents is needed for the production of rGO coated wearable e-textiles. 51 In this context, several studies have been made so far, by using green reducing agents such as, ascorbic acid (C 6 H 8 O 6 ), thio-urea (CH 4 N 2 S), and sodium hydrosulphite (Na 2 S 2 O 4 ). 52–54 Furthermore, different fabrication techniques has been reported so far; in previous studies, including; vacuum filtration, brush painting, spin coating, spray coating, screen printing, stencil printing, dip coating, heat transfer, vapor deposition, chemical vapor deposition (CVD) and wet transfer. 55 Among all, pad–dry–cure method is an efficient process with processing rate of 150 m min −1 , suitable for mass production of graphene-coated textiles. 56 Whereas, the graphene is hydrophobic and cannot be dispersed in water and other ionic solvents; therefore the fixation of carbon-based materials on textiles is also limited. As dispersion of graphene and carbon material is only possible with the addition of certain dispersing agents and anionic surfactants including, sodium dodecyl sulphonate (SDS), sodium dodecyl benzene sulphonate (SDBS), ammonium persulphate (APS) and cetyltrimethylammonium bromide; hexadecyltrimethylammonium bromide (CTAB). 57–60 Application and fixation of graphene is accomplished either by covalent or non-covalent chemical functionalization using ionic liquids including dimethyl sulphoxide (DMSO), and dimethyl formamide (DMF). 61–63 Several binders and thickeners such as polyurethane (PU), polyvinyl alcohol (PVA), polyvinylidene fluoride (PVDF), polystyrene (PS), polystyrene sulphone (PSS), carboxy methyl cellulose (CMC), sodium alginate, water-borne polyurethane (WPU), and polyvinylidenes (PVP) have also been used. 64–69 The use of such binders, thickeners and dispersing agents may influence the physical properties such as luster, feel, touch and end use properties for example; electrical, thermal, mechanical and comfort characteristics by altering transport of water vapors and air permeability of wearable e-textiles. 70–72 In this study, we used commercial-scale technique for the development of graphene coated cotton textile fabric for wearable thermoelectric nanogenerator for energy harvesting from low grade body heat. The as dyed fabric was converted into rGO using green reducing agent ( l -ascorbic acid) and lower scale thermal reduction at (90 °C) which is more convenient and efficient as compared other high temperature thermal reduction techniques. Herein, this study we applied GO as a dye in a water based dispersed solution of PEDOT:PSS without using any binder and dispersing agent. The application process uses already commercialized pad–dry–cure simple dyeing technique approach. This study demonstrates; that pad–dry–cure method is an alternative approach for the development of conductive textiles, which assists in better fixation and adhesion of graphene on textiles without using any binder or thickener."
} | 4,704 |
36333441 | PMC9636164 | pmc | 8,421 | {
"abstract": "Research focused on microbial populations of thermoalkaline springs has been driven in a large part by the lure of discovering functional enzymes with industrial applications in high-pH and high temperature environments. While several studies have focused on understanding the fundamental ecology of these springs, the small molecule profiles of thermoalkaline springs have largely been overlooked. To better understand how geochemistry, small molecule composition, and microbial communities are connected, we conducted a three-year study of the Five Sisters (FS) springs that included high-resolution geochemical measurements, 16S rRNA sequencing of the bacterial and archaeal community, and mass spectrometry-based metabolite and extracellular small molecule characterization. Integration of the four datasets facilitated a comprehensive analysis of the interwoven thermoalkaline spring system. Over the course of the study, the microbial population responded to changing environmental conditions, with archaeal populations decreasing in both relative abundance and diversity compared to bacterial populations. Decreases in the relative abundance of Archaea were associated with environmental changes that included decreased availability of specific nitrogen- and sulfur-containing extracellular small molecules and fluctuations in metabolic pathways associated with nitrogen cycling. This multi-factorial analysis demonstrates that the microbial community composition is more closely correlated with pools of extracellular small molecules than with the geochemistry of the thermal springs. This is a novel finding and suggests that a previously overlooked component of thermal springs may have a significant impact on microbial community composition.",
"introduction": "Introduction Thermoalkaline springs are unique environments of biological and industrial significance. Commercial applications are well documented in these systems and current thermoalkaline bioprospecting efforts are broad 1 . A successful example is the commercialization of a suite of thermostable enzymes including lipolytic and hydrolytic enzymes 2 , 3 . Of particular interest is the development of thermostable cellulolytic enzymes capable of converting lignocellulosic biomass into sugars and ultimately ethanol under industrial conditions 4 . The potential to develop thermo- and pH-stable enzymes for commercial applications and interest in the ecology of these systems has led to an accumulation of geochemical and microbial phylogenetic data 5 – 8 . Along with bioprospecting efforts, thermoalkaline ecology has also driven investigation. Through this work, temperature has been shown to be a large driver of microbial diversity with increasing spring temperatures translating into a decrease in microbial diversity 9 – 11 . Temperature increases have also been associated with increases in archaeal abundance and diversity 11 , 12 . Thermophiles have been shown to tolerate a wide range of pH 10 , 13 . pH has also been demonstrated to be an important factor in abundance and diversity in thermal environments, with circumneutral and alkaline springs supporting more diverse microbial populations 13 , 14 . However, these two factors alone do not completely explain the assembly of microbial populations in thermal systems 15 . Thermoalkaline springs have been shown to contain a wide range of bacteria and archaea, with several common clades residing in springs with variable geochemistry and geographical location 9 , 16 . Predominant microbes in described springs include Chloroflexi, Deinococcus, Nitrospiral, Cyanobacteria, Proteobacteria, Thermodesulfobacteria, Aquificae, Thermotague, Thermococcales and Crenarchaeota 9 , 14 , 16 , 17 . Despite increased interest and investigation in these systems, gaps in knowledge remain 2 . For example, the temporal dynamics of thermoalkaline microbial populations across multiple years have rarely been investigated 17 – 19 and to our knowledge, analyses combining the microbial with the intracellular metabolome and extracellular small molecule composition have not been conducted. This is especially true of winter sampling of hot springs in YNP where access is limited. Challenges associated with increasing knowledge in thermoalkaline springs include microbial culturing. To confirm the specific metabolic potential and ecological contributions of individual microbes, cultured isolates are often required. Extreme environments like thermoalkaline springs have presented challenges to isolation efforts when using traditional culturing practices, especially with respect to Archaea 16 , 20 . Gaining extracellular and intracellular-based understanding of thermoalkaline environments has broad implications including improving culturing efforts by providing insight into environmental small molecule and metabolic networks. Thermoalkaline springs can be found in several locations around the world, including Yellowstone National Park (YNP) where they are prevalent 1 , 21 , 22 . One such group of springs in YNP includes the Five Sisters (FS) hot springs, located in the White Creek Drainage (WCD). WCD is part of Lower Geyser Basin, the largest thermal basin in YNP. Thermal features in WCD have been previously identified as areas likely to contain distinct and dynamic environments 23 , 24 . The FS system is host to a variety of metabolic activity with photosynthesis occurring at the edges of the cooler pools in the spring and summer and heterotrophic activity driven by lignocellulose degradation 6 . In this study, the FS thermoalkaline springs were examined over the course of three years and an extensive analysis was conducted using high-resolution geochemical measurements, 16S rRNA microbial community sequencing, and liquid chromatography mass spectrometry (LCMS) based small molecule characterization that enabled the establishment of temporal microbial trends and a better understanding of driving factors underlying shifts in microbial population makeup and metabolism in this unique environment.",
"discussion": "Discussion This study was undertaken to explore thermoalkaline hot springs and the microbial life that inhabits them to better understand the ecology and small molecule biology of these unique environments. The described analysis allowed for the most comprehensive view to date of microbial life in the FS spring system. Our data indicate that a systemic change likely occurred between 2017 and 2018 which significantly impacted the environment and microbial ecology of the FS springs. Samples from 2017 to 2018 displayed coordinated extracellular small molecule, microbial and intracellular small molecule shifts. Yet, these general trends did not fully explain the pattern of archaeal decline relative to that of bacterial populations. This phenomenon of differential responses between bacteria and archaea to a shared environmental stimulus has been observed by Pala, et al., where contrasting archaeal and bacterial abundance shifts were the result of differing geochemical factors in the same aqueous environment 26 . However, the geochemical variables with significant differences over the course of the study did not exhibit a characteristic pattern reflective of the archaeal population data. To discover a correlative pattern for archaeal decline, we then examined the intracellular and extracellular small molecule profiles. An initial examination revealed a global shift between 2017 and 2018, suggesting environmental and metabolic change(s) in the microbial communities of the spring system during this time period. A closer investigation into the intracellular data revealed disruption in several specific nitrogen and sulfur cycle pathways. Pathways found to be impactful and significant included pyrimidine metabolism, glutamate and glutamine metabolism, arginine and proline metabolism, riboflavin metabolism and the citrate cycle. These pathways all have connections to both energy metabolism and nitrogen cycling as well as other metabolic impacts 27 , 28 . Disruption in the abundance of metabolites in these pathways could be due to or could cause shifts in archaeal abundance, especially with pathways impacting nitrogen cycling, where Archaea have been shown to play key roles 29 . Specifically, Archaea have unique enzymes and pathways in riboflavin synthesis as well as in nitrogen assimilation and dissimilatory pathways which could modulate the amino acid metabolism as well as pyrimidine metabolism and the citrate cycle 30 , 31 . Although not classified as impactful in our MetPA, sulfate metabolism was also found to be significantly (p < 0.05) dysregulated between years and is known to have different enzymes and pathways between bacterial and archaea metabolism 32 . While the initial analysis of intracellular and extracellular small molecules demonstrated a shift in metabolism and the spring environment between 2017 and 2018, it did not fully explain the consistent loss of Archaea from 2017 to 2019. Our analysis shows that the strongest correlations to archaeal decline are to specific nitrogen- and sulfur-containing small molecules. These small molecules decreased steadily from 2017 to 2019 and coincided with a general decrease in the relative abundance and diversity of Archaea. This observation led to the hypothesis that environmental shifts may induce disparate small molecule profiles with contrasting elemental composition and structure, which could result in thermophiles in the FS system adapting metabolic strategies to environmental changes. Availability of bioactive small molecules would then lead to differential success of specific organisms in the FS system, which could explain the shift in population makeup that occurred from 2017 to 2019 as defined by the observation of a decline in the relative abundance of Archaea. Although a general decline in Archaea was noted, two clades of Archaea exhibited a strong positive correlation to the number of nitrogen and sulfur compounds detected. These Archaea belonged to either Aigarchaeota or Crenarchaeota , both members of the TACK superphylum 33 . The top correlating Archaea did not include any members from the other superphyla DPANN, Euryarchaeota or Asgard 34 . Many members of the TACK clade have shown strong functional roles in nitrogen and sulfur cycles, such as ammonia oxidation ( Thaumarchaeota ), sulfur oxidation ( Crenarchaeota ), dissimilar sulfur metabolism ( Korarchaeota ) and facultative nitrate reduction ( Geoarchaeota ) 1 , 29 , 35 , 36 . The correlations to nitrogen- and sulfur-containing compounds coupled with previously discovered metabolic traits of the TACK superphylum demonstrate the need to further explore the relationship that may exist among the species of Aigarchaeota and Crenarchaeota in the FS system with respect to nitrogen and sulfur metabolism. In particular, the globally distributed but as of yet uncultivated Aigarchaeota 37 have consistently been found in high relative abundances within thermoalkaline sites within Yellowstone 17 , 38 . Based on metagenomics and single cell assemblies, this group has high metabolic versatility, including pathways autotrophy, dissimilarity sulfite reduction and carbon monoxide oxidation 39 , as well as diversity in carbon substrate utilization, including acetate, fatty acids and amino acids 40 , which could play a large role in nutrient cycling in these systems. While the decline in nitrogen- and sulfur-containing extracellular small molecules could be due to abiotic or biotic influences, Gonisor et al. concluded that the chemodiversity seen in nearby Octopus Spring was likely not microbially driven 41 . Their interpretation was based on the low biodiversity of thermoalkaline springs and the high level of unique extracellular small molecules. One potential explanation for the loss of nitrogen and sulfur compounds is the differential annual mixing of ground and surface water in the springs, where runoff, groundwater and snowpack can impact spring geochemistry and extracellular small molecule composition 41 , 42 . In our study, samples were collected in late February and early March. Our sampling coincides with very low surface runoff which occurs in late springs at this site. The average ambient temperature during our sampling times was still well below freezing so no appreciable addition melting should have occurred such that the spring system environment should be stable at the time of collection. However, the amount of groundwater recharge from the previous year may have a significant impact on resulting spring water composition the following winter. The area around WCD had unusually high snowpack in 2017 and 2018 (impacting 2018 and 2019 data) relative to the previous and following years according to the Water Resources Data System & State Climate Office of Wyoming ( www.wrds.uwyo.edu ). On March 1st, 2017 and 2018, the snow water equivalent amounted to 110% and 124% of the average for the time of year in, respectively. Snow water equivalent in 2016 was 88% of the median, potentially impacting 2017 data. This SWE trend indicating an increased snowpack from 2016 to 2018 correlates with the loss of unique nitrogen- and sulfur-containing compounds and archaeal decline. A Pearson correlation analysis found a strong negative correlation of − 0.99 between SWE and archaeal alpha diversity with a p-value of 0.0031. During the increased runoffs in the 2017 and 2018 season from high snowpack, the hot spring water composition likely went through a significant change leading to distinct differences in environmental small molecule compositions. Runoff from snowpack has been previously theorized to modulate specific populations of Archaea in hot springs. Campbell et al . observed temporal changes in Sulfolobus islandicus populations within different hot springs in YNP that did not correlate with measured geochemical changes observed in the springs 43 . These changes were hypothesized to arise from runoff or other hydrogeochemical perturbations. In conclusion, our analysis has characterized some of the complexity and dynamism of thermoalkaline spring ecology. It also revealed that thermophilic Archaea may be sensitive to small environmental perturbations. The combination of standard geochemical techniques with novel mass spectrometry small molecule analysis exposed the limitations of solely using elemental composition and standard geochemical measurements, such as dissolved oxygen, to assess changes in complex environmental systems. Mass spectrometry analysis identified specific changes in small molecule composition that correlated with archaeal relative abundance. Such associations would have been missed using a purely geochemical analysis. Our comprehensive study revealed that environmental events, possibly related to snowpack, occurred between 2017 and 2019 that shifted hot spring compound composition, manifesting in nitrogen- and sulfur-containing compound transitions, correlating with metabolic changes and a relative decrease in Archaea."
} | 3,761 |
25358460 | PMC4215306 | pmc | 8,422 | {
"abstract": "Magnetotactic bacteria biomineralize ordered chains of uniform, membrane-bound magnetite or greigite nanocrystals that exhibit nearly perfect crystal structures and species-specific morphologies. Transmission electron microscopy (TEM) is a critical technique for providing information regarding the organization of cellular and magnetite structures in these microorganisms. However, conventional TEM can only be used to image air-dried or vitrified bacteria removed from their natural environment. Here we present a correlative scanning TEM (STEM) and fluorescence microscopy technique for imaging viable cells of Magnetospirillum magneticum strain AMB-1 in liquid using an in situ fluid cell TEM holder. Fluorescently labeled cells were immobilized on microchip window surfaces and visualized in a fluid cell with STEM, followed by correlative fluorescence imaging to verify their membrane integrity. Notably, the post-STEM fluorescence imaging indicated that the bacterial cell wall membrane did not sustain radiation damage during STEM imaging at low electron dose conditions. We investigated the effects of radiation damage and sample preparation on the bacteria viability and found that approximately 50% of the bacterial membranes remained intact after an hour in the fluid cell, decreasing to ~30% after two hours. These results represent a first step toward in vivo studies of magnetite biomineralization in magnetotactic bacteria.",
"discussion": "Discussion Mitigating radiation damage is critical for electron microscopy of biological samples, especially for fluid cell STEM imaging, which has been shown to have a number of electron beam induced artifacts 41 49 . Cumulative electron doses on the order of magnitude used here (~0.1 electrons · Å −2 ) coincide with a recent fluid cell STEM study of gold nanoparticle uptake into eukaryotic cells; these doses were shown to induce only small changes in the cellular structure with a single STEM exposure 43 . The authors proposed that the first STEM image obtained at the outer cellular region contained information of the ultrastructure of the live cell and subsequent images in other areas likely contained information about the live cell as well. However, for the case of bacteria imaged at the low magnifications in this study, a single STEM exposure irradiates the entire cell at once. Therefore, in this paper, only the first HAADF-STEM images acquired of cells of M. magneticum have been shown as they are the most representative of the initial cellular structure. While we can use previous fluid cell STEM studies of radiation damage in eukaryotic cells as general guidelines for live imaging of magnetotactic bacteria, the question remains, are the bacterial cells alive following STEM imaging? The death of a bacterial cell is often defined as the inability of the cell to grow into a visible colony in media 50 . Another commonly used indicator of bacterial viability is cell mobility. In our experiments, it was necessary to attach the cells to the SiN windows to immobilize them for STEM imaging, so we could not evaluate bacterial viability on the basis of either of these criteria. This limits our indicators of radiation damage to cellular function to observations of bacterial shrinkage ( Fig. 4 ) and membrane viability via fluorescence labeling and imaging ( Fig. 6 ), neither of which can be used to directly assess enzymatic or reproductive damage of the cells. Intermediate fluorescence states identified in SYTO 9/propidium iodide labeled cells of Escherichia coli suggested that fluorescent labeling can indicate more than simply membrane integrity, but it was unclear exactly how these states related to specific cell functions and reproductive health of the bacteria 50 . The cumulative electron doses used in this study are at least four orders of magnitude higher than the lethal electron dose of ~10 −5 electrons · Å −2 shown to cause reproductive death and enzyme deactivation in E. coli 51 52 53 . The lethal electron dose was measured from pulsed electron beam experiments (E = 520 keV) of E. coli on wet Millipore filters 52 ; such a thick sample would absorb more ionizing radiation and facilitate more damage to the bacteria than for fluid cell STEM imaging, which allows for transmission of most of the electrons. Furthermore, calculation of the lethal electron dose assumed TEM irradiation 53 , which delivers a continuous flux of electrons over a micron sized area, whereas STEM imaging delivers electrons in a nanometer sized probe scanned serially across the sample. Due to these differences, we cannot conclude whether the same lethal electron dose exists for fluid cell STEM imaging of magnetotactic bacteria. Cellular processes in the cells of M. magneticum are likely altered or damaged, but further work on the viability of electron irradiated bacteria in the fluid cell will be needed to establish a lethal electron dose for STEM imaging. While the electron doses used here have been shown to cause various types of damage in cells of E. coli , they are two or more orders of magnitude below the established damage threshold for cryo-TEM and radiation damage of nucleic acids (~10 electrons · Å −2 ) 30 39 51 , and one order of magnitude below the damage threshold for amino acids (~1 electron · Å −2 ) 51 . Imaging at electron doses higher than the radiation damage threshold for amino acids induced damage in the form of magnetosome chain shrinkage, likely due to radiation damage of the bacterial cytoplasm ( Fig. 4 ). Importantly, this suggests that as long as the cumulative electron dose for a STEM image is <1 electron · Å −2 , the structure of the magnetosome exists in an environment representative of the initial cellular state, even if some of the cellular processes have been altered or arrested due to irradiation. The fact that we image at cumulative doses below the nucleic and amino acid damage thresholds further corroborates our observations of intact bacterial cell membranes following fluid cell STEM imaging. We expect these radiation damage thresholds to be generally applicable to fluid cell electron microscopy imaging of most prokaryotes. We must also consider changes to the bacterial cells due to fluid cell sample preparation. Potential causes of bacterial membrane damage in the fluid cell include confinement between the two SiN windows, the presence of dimethyl sulfoxide (DMSO) in the fluorescent dye, and non-ideal oxygen conditions in the growth media. Magnetotactic bacteria are cultured under microaerobic conditions (0.25 mbar is optimum for magnetosome biomineralization by M. Magneticum 46 ), and the small amount of media contained between the SiN windows in the fluid cell quickly saturates to atmospheric O 2 levels, creating a non-ideal, possibly toxic, environment for the bacteria 54 . However, this particular strain of M. \n magneticum is known to be relatively oxygen resistant, and the amount of DMSO in the media is likely below toxic levels, so confinement of the cells is likely the explanation for cell damage. Confinement of the bacterial cells in a hundreds of nanometers thick liquid layer creates compressive stresses on the bacterial cells. Compressive stresses have been shown to increase membrane permeability and damage efflux systems in E. coli , leading to increased absorption of propidium iodide 55 . High pressures have also been shown to lead to increased rates of cell lysis in Lactobacillus strains 56 . Our systematic fluorescence imaging showed increased propidium iodide absorption into the cells upon sample preparation and over the two hours in the fluid cell; however, cell lysis was not observed. While the exact cause of the rapid decrease in viability over two hours in the fluid cell is not entirely clear, it necessitates correlative fluorescence imaging after STEM imaging as well, to verify bacterial membrane viability in the electron microscopy images. Taken together, our bacterial viability experiments and the cumulative electron doses used strongly suggest that some bacterial cells remain alive in the fluid cell up to the point of electron beam irradiation. After STEM imaging, cellular function is altered or arrested due to radiation induced enzyme deactivation and possible reproductive death of the cell. Given these conclusions, we expect in vivo imaging of magnetosome biomineralization in initially live magnetotactic bacteria will be possible using fluid cell STEM imaging. The flow capabilities of the fluid cell will allow for controllable introduction of iron-replete growth media to an iron-deficient bacterial culture. While high resolution microscopy of the magnetosome structure in live bacteria currently may not be possible due to radiation damage, our results demonstrate that STEM imaging at magnifications up to M = 28,000 x allows for visualization of magnetosome structures while retaining an intact bacterial cell membrane. The electron doses used in this study are below that for radiation damage of the bacterial ultrastructure, so imaging of the magnetosome nanostructures in a native cellular environment is possible. High resolution TEM (HRTEM) of intact magnetosome structures in the fluid cell is also possible (cf. Supp. Fig. S1 ); however, the high electron dose will quickly kill the cells and damage the ultrastructure. One possibility is to perform a fluid cell HRTEM or HRSTEM study where the various steps of magnetosome growth are imaged subsequently in several cells, so that each image of a growing magnetosome is taken from a bacterial cell that is alive up to the moment of imaging. Previous studies have shown that exposure of magnetotactic bacteria to different types of radiation can affect magnetosome shape and size through direct DNA damage 57 . We didn't observe any direct damage to the magnetosomes by the electron beam in this work and there are no studies on the effect of electrons on magnetosome growth, but possible changes in the magnetosome structure due to high energy electron irradiation should be considered. Another possible route for this technique is in situ chemical imaging of soluble iron uptake by magnetotactic bacteria. We have recently visualized iron binding by micelles of bacterial membrane protein by observing the increase in Z-contrast of the micelles via fluid cell HAADF-STEM imaging 58 . This study indicates this technique shows promise for visualizing uptake of iron into the bacteria. In situ spectral imaging via EELS 59 or energy dispersive x-ray spectroscopy (EDS) 60 mapping may provide additional means for visualizing iron uptake, which has only been possible to date in dried bacteria samples using techniques like scanning transmission x-ray microscopy 61 . Finally, further correlative fluorescence and fluid cell electron microscopy studies utilizing fluorescent proteins could allow investigation of the connections between biomolecular processes and biomineralization of magnetosome magnetite nanocrystals. For instance, fluorescently tagged membrane proteins can be tracked during bacterial cell growth via fluorescence microscopy 20 , followed by fluid cell STEM imaging of the magnetosome structures to aid in elucidating the role of protein localization in the biomineralization process. In summary, cells of M. \n magneticum strain AMB-1 were fluorescently labeled with SYTO 9 and propidium iodide nucleic acid stains and imaged in media via fluid cell STEM and fluorescence microscopy. The magnetite magnetosomes provided high contrast labels to image bacterial cells using HAADF-STEM; in some cases the bacterial cell membrane was resolved when the liquid layer was sufficiently thin. We established the major cellular damage sources to be electron beam irradiation and compression of the bacteria between the SiN windows. The STEM electron dose could be kept sufficiently low to prevent detectable ultrastructure damage to bacteria. However, at high electron doses (>1.0 electron · Å −2 and after multiple STEM exposures), radiation damage in the bacterial cytoplasm was manifested as a contraction of the magnetosome chain length. Correlative fluorescence and STEM imaging indicated that cells of M. magneticum had intact membranes after STEM irradiation in the fluid cell, but could not establish other cell viability criteria such as reproductive ability or enzymatic function. We found that ~50% of the cells typically had intact membranes upon assembling the fluid cell, and approximately half of these bacteria sustained membrane damage due to compressive stresses over the next two hours in the fluid cell. This correlative fluid cell STEM and fluorescence microscopy technique is a first step in directly observing biomineralization of magnetite in viable magnetotactic bacteria. We expect this technique to be generally applicable for in vivo imaging of a wide range of biomineralizing organisms."
} | 3,235 |
34587163 | PMC8480890 | pmc | 8,423 | {
"abstract": "Due to increasing population growth and declining arable land on Earth, astroagriculture will be vital to terraform Martian regolith for settlement. Nodulating plants and their N-fixing symbionts may play a role in increasing Martian soil fertility. On Earth, clover ( Melilotus officinalis ) forms a symbiotic relationship with the N-fixing bacteria Sinorhizobium meliloti ; clover has been previously grown in simulated regolith yet without bacterial inoculation. In this study, we inoculated clover with S . meliloti grown in potting soil and regolith to test the hypothesis that plants grown in regolith can form the same symbiotic associations as in soils and to determine if greater plant biomass occurs in the presence of S . meliloti regardless of growth media. We also examined soil NH 4 concentrations to evaluate soil augmentation properties of nodulating plants and symbionts. Greater biomass occurred in inoculated compared to uninoculated groups; the inoculated average biomass in potting mix and regolith (2.23 and 0.29 g, respectively) was greater than the uninoculated group (0.11 and 0.01 g, respectively). However, no significant differences existed in NH 4 composition between potting mix and regolith simulant. Linear regression analysis results showed that: i) symbiotic plant-bacteria relationships differed between regolith and potting mix, with plant biomass positively correlated to regolith-bacteria interactions; and, ii) NH 4 production was limited to plant uptake yet the relationships in regolith and potting mix were similar. It is promising that plant-legume symbiosis is a possibility for Martian soil colonization.",
"conclusion": "5. Conclusion Martian colonization will be increasingly needed in the future, yet additional soil and atmospheric augmentation research will be required to develop astroagricultural techniques and allow for the greatest probability of success in Martian farming. This research demonstrates that based on regolith properties and its limited nutrients, Rhizobia can significantly increase plant growth in regolith. In addition, the relationship between Rhizobium spp. and plants differs when comparing regolith to soils; regolith interactions were positively correlated to plant biomass. Additional research focused on augmenting regolith would serve to reduce remaining ambiguities and to provide a broader understanding of how plants would function within Martian planetary dynamics.",
"introduction": "1. Introduction Given the circumstances of climate change, biological contagion, or other events that have potential to wipe out humanity, it is unlikely that humans will be able to remain a single planet species. With human populations growing and space for development and arable land becoming increasingly limited on Earth, off-world agriculture will likely be needed on celestial bodies such as Mars [ 1 ]. However, the harsh Martian environment challenges many of the basic tenets of biology found here on Earth. Plants will face some phytotoxicity in regolith (Martian soils), the atmosphere is significantly thinner with a different stoichiometry, temperatures can dip to below -100°C, the lack of atmosphere allows dangerous radiation to affect the planet surface, and plants will be reliant upon limited resources for survival [ 2 ]. To support a long-term colony and food production on Mars, it is imperative to establish an on-planet food source capable of feeding its inhabitants [ 2 ]. However, Martian regolith presents challenges for plant established and growth, which is especially true for nitrogen (N) availability. With a lack of critical nutrients in Martian regolith, particularly plant-available N, it will be necessary to find methods to supplement regolith in a cost-effective manner. Regolith is the only available on-site medium for growing plants on Mars. Given that it is not feasible to ship earth soils through space because of weight and cost, soil augmentation appears to be the most viable path forward. Regolith has been analyzed from several rover missions, and surveys have found no traces of plant-available N in regolith. In addition, no known significant organic material on the Martian surface has been identified that could supply plant-available N via microbial mineralization [ 3 , 4 ]. Another problem that requires further testing is the low diatomic N (N 2 ) content in the Martian atmosphere. With only around 1.9% of the Martian atmosphere being N 2 , the ability of the N-fixing microbial-plant association to utilize N 2 gas maybe hindered [ 5 ]. On Earth, approximately 78% of the atmosphere is N 2 , making N readily available for plant-microbial symbiotic associations. Nitrogen in Earth’s soils is partly made accessible by decomposers that mineralize organic N forms to release NH 4 , yet this is not the case on Mars as organic matter and microorganisms responsible for mineralization are lacking [ 4 , 6 ]. Furthermore, on Earth Rhizobium spp. form symbiotic relationships with leguminous plant roots to produce NH 4 from N 2 gas. The mechanism of rhizodeposition of N through root exudation has been shown in previous studies to provide 3–4.5% of fixed nitrogen from the plant-rhizobium symbiosis to the soil [ 7 ]. While Rhizobium sp. provide NH 4 readily available for plants to use in various functions (e.g., amino acid and protein production, DNA, RNA, ATP, chlorophyll) [ 8 ]; symbiotic microorganisms are likely lacking in regolith. Though plant-available N is lacking on the Martian surface, plants have been shown to grow in regolith simulant [ 9 ]. Prior experiments tested Lupinus sp., Vicia sp., and Melilotus sp. because these are common nodulating species that perform well in traditionally harsh soils. Although Wamelink et al. [ 9 ] did not inoculate these plants with their respective Rhizobium sp., it was posited as one method to increase plant biomass and regolith N availability over uninoculated regolith. In support of this contention, earlier studies using the JSC 1 regolith simulant have shown that at least one Rhizobium spp. can survive in a regolith simulant [ 10 ]. Regardless, the symbiotic relationship between leguminous plants and Rhizobium spp. is likely needed in materials, such as regolith, in order for both species to successfully thrive. It is well established that N-fixing bacteria (e.g., Rhizobium spp.) allow plants to indirectly acquire atmospheric N for their use and directly deposit excess N in the soil [ 11 ]. Host specificity has been observed in some species of N-fixing bacteria, and to ensure symbiosis, plants must be inoculated with their respective N-fixing symbiotes [ 12 ]. It is currently unknown if plants will benefit from Rhizobium inoculations in the harsh chemical and physical stress conditions of regolith, or if enough N will be synthesized to change regolith N content. In addition, it is not known how different Rhizobium spp. will respond in regolith. If N-fixing bacteria can be used to incorporate atmospheric N to Martian regolith, this could be used as a first step in creating a Martian astroagricultural system. Thus, the objectives of this study were to examine the: 1) relationship of nitrogen fixation and plant-microbe symbiosis of M . officinalis and S . meliloti in regolith versus potting soil; and 2) effects this relationship has on plant growth and soil N availability. We hypothesized that an increase in plant biomass would be observed in regolith inoculated with their respective N-fixing bacteria, and excess plant-available N would be deposited in the surrounding regolith via rhizodeposition of exudates, similar to the Rhizobium-legume relationship found on Earth.",
"discussion": "4. Discussion This research highlights the importance of using naturally forming partnerships between plants and their symbiotic bacteria to increase plant growth success in regolith, one of the first steps towards understanding the capability of establishing astroagricultural colonies on Mars. Though clover ( Melilotus officinalis ) had been previously demonstrated to grow in regolith [ 9 ], our study found that plant shoot and root growth was increased by over 75% when inoculated with S . meliloti compared to plants grown in uninoculated regolith. Our study highlights the importance of nitrogen as a major limiting factor for plant growth in regolith, suggesting that nitrogen-fixing bacteria can be used to reduce this limitation. Though we have demonstrated this using clover, this research may be the foundation for future research on other food producing crops. Rhizobia significantly enhanced plant growth in the regolith, suggesting that nitrogen is the major limiting resource for plants in this media. It has been shown that Rhizobium ’s survival and potential to fix nitrogen can be limited by soil stress [ 19 , 20 ]. In this study, effects of soil stress were demonstrated as the potting mix contained significantly greater NH 4 concentrations after plant growth, though plants in potting mix were not initially limited by nitrogen. At the end of the experiment, the mean difference in NH 4 between regolith and potting mix inoculated groups was 7.5 mg kg -1 ( Table 1 ). The limitation of the symbiosis to produce reactive nitrogen demonstrates previously studied challenges to the viability of Martian regolith as a in situ resource for agriculture. Chemical stress from Martian regolith has been shown to be extremely detrimental to plants [ 21 ]. The same study showed that without the addition of nutrients, Arapodobsis thaliana died within 10 days of germination [ 21 ]. However, when Hoaglands No. 2 nutrient solution was added, plants only experienced about a 10% die off. Additionally, while most simulants of Martian regolith are able to support plant life with the addition of nutrients and acidification, no current simulants account for the calcium perchlorate deposits on the Martian surface [ 21 ]. While this study is a step to correcting for nutrient deficiency and move in the direction of terraforming Martian regolith, more studies including phytoremediation or mycoremediation will be needed to correct for other toxicity issues in Martian regolith. While the above challenges remain, when in regolith, the addition of captured atmospheric nitrogen likely increased N for the plant, therefore decreasing this as limiting factor on plant growth [ 22 ]. However, despite less nitrogen being fixed in the regolith, NH 4 appeared more important for plant growth than in the potting mix because of its increased scarcity. The presence of nitrogen is required for nodulation and establishment of rhizobium, yet it is commonly used up by plants in low N environments, leading to N deficiency in the plant [ 23 ]. Micronutrient content could be another restrictive property as the regolith lacked Fe, Cu, Zn, and Mn. Although not analyzed in the regolith, Mo may have also been lacking as it is important for nitrogen fixation [ 24 ]. In essence, plants in the potting mix were likely not limited by soil nutrient deficiencies as compared to the regolith. We expected that in N-poor regolith, more root nodules would have been formed as compared to the potting soil, yet the opposite was observed. A relative reduction in regolith nodulation may have caused by other limitations such as pH or deficiency in almost every plant nutrient ( Table 1 ), and in particular, available Fe. Prior research established that plants with nodules require more Fe to sustain their relationship with rhizobium [ 25 ]. In addition, the rate of N fixation by rhizobia in some plants is positively correlated to available soil Fe concentrations [ 26 ]. Furthermore, the simulated regolith pH was fairly high (8.7), and thus may have impaired plant Fe intake, further reducing nodule formation and nitrogen fixation ( Table 1 ). Martian soils have between 5 and 14% iron oxide [ 27 ] but a soil pH of ~ 8.0 [ 28 ]. Given these conditions, plant-available Fe content would likely be less than 10 −24 M Fe 3+ and thus low, if not lower, than the initial regolith Fe concentrations [ 29 ]. Overcoming challenges in plant nutrient availability will need to be considered in order to effectively grow plants on Mars. With respect to plant biomass, the number of nodules had a negative relationship with potting mix plants and a positive effect with regolith plants ( Fig 2 ). Negative correlations between nodule formation and potting mix may have resulted from the presence of pre-existing nitrate. In nodulation, plants generally form associations with rhizobium at lower rates when nitrate (and/or NH 4 ) is abundant [ 30 ]; yet, plants grown in potting mix had significantly more nodules than regolith. Two likely explanations of this could be that nodules formation resulted in less biomass, or that available N at the initial condition in potting mix could have increased plant health and growth at the beginning of the experiment rendering nodules less effective. Interestingly, there was also less remaining NH 4 per nodule in the potting mix ( Fig 2 ). This may be explained by the larger size of the plants in the potting mix, as they were not limited by the other restrictive properties (compared to the regolith) and were able to use more available nitrogen. Potting mix nitrate could have leached over time due to watering, and it is also possible that plants may have assimilated nitrate prior to leaching and other normal soil N cycling processes [ 31 – 33 ]. Without measures of nitrate at the conclusion of the study, it is not possible to discern the cause, however it was not crucial to the main hypothesis this study tested. Soils commonly lose nitrate to leaching [ 34 ]. Nitrate is transformed from NH 4 when NH 4 is converted by nitrifying bacteria [ 35 ]. While the primary form of nitrogen that results from plant-bacterial symbiosis is NH 4 , in terrestrial soils nitrifying bacteria convert NH 4 to nitrite or nitrate. Because Mars has no bacteria observed in its regolith, there would likely be no Nitrosomonas or Nitrobacter to convert NH 4 to nitrite and nitrate, respectively. The loss pathway for nitrite and nitrate is most commonly leaching, while the leaching loss of NH 4 would likely be less of a concern when watering [ 36 ]. This could prove beneficial when raising and irrigating plants in regolith. However, given the high regolith pH (7.8), ammonia volatilization would likely be a more significant loss pathway concern on Mars. Ammonia volatilization occurs to greater extents as soil pH becomes more alkaline [ 37 ], increasing almost linearly above pH 8 [ 8 ]. However, over the long-term, ammonia volatilization drives pH down, and in the case of regolith, could make soils more suitable for plants and rhizobia [ 38 ]. Study results showed the symbiosis benefited plant growth and phenology, in both regolith and potting mix. Though we know that rhizodeposition occurs, NH 4 did not appear to accumulate in the soil. N-starved plants likely used all available soil NH 4 . Further, the lack of plant decay likely kept sequestered available N in plant roots. Given that the plants were not used as green fertilizer, the sequestered N was never released back into the soil. Thus, deposition from decay was not possible. An insufficient amount of experimental time could have been another factor as to why soil NH 4 concentrations did not increase, as suggested by others using regolith [ 39 ]. Companion cropping by using nitrogen fixing rhizobium and their plant symbiotes generally occurs at one year intervals [ 40 ]. In these cases, root exudation and root die off that result in subsequent release of nitrogen are thought to be integral to the transfer of nitrogen in these systems [ 42 ]. In order to overcome this issue in future studies, more plants per pot could be added to increase the amount of N fixation that occurs per volume of regolith and the experiment could be run for a longer duration. An additional option for increasing regolith N and its N storage capacity would be to till nitrogen containing plants, like clover, into regolith as green fertilizer [ 41 , 42 ], along with the addition of decomposer microorganisms to produce more bioavailable nitrogen via mineralization [ 6 , 43 ]. One study reported that the addition of organic matter in regolith, using grass clippings from Lolium perenne L., resulted in an improvement in plant growth displayed in plant phenology as plants grown in previous studies did not show seed or fruit production [ 44 ]. As fungi and bacteria are routinely placed in cold storage for archiving and research purposes, this process could be replicated for transport to Mars. Plant incorporation and decomposer cryogenesis/revival, followed by regolith application, should be quantified on Earth before use on Mars. Plants generally cannot grow without accessible nitrogen, and can only grow poorly in areas with scarce nitrogen [ 23 ]. However, the study by Wamelink et al. [ 9 ] showed that plants, other than nodulating plants, could grow in regolith, though the authors had a difficult time germinating seeds of nodulating plants. In comparison to nodulating plants used in Wamelink et al. [ 9 ], this current study showed nodulating plants inoculated with their respective Rhizobium sp. were able to survive for longer time periods. For comparison, the average M . officinalis survival rate after 50 days was roughly 50% as observed by Wamelink et al. [ 9 ]; whereas in our study, after 90 days, 100% of inoculated plants survived. Increasing study duration would aid in filling in gaps about persistence of plants in harsh conditions. While Rhizobium spp. fix atmospheric N 2 , regolith also contains bioavailable P and K, as well as some other micronutrients [ 3 , 4 ]. However, other methods of fixing or adding missing micronutrients will be needed for those which are not present. Specifically, Cu, B, and Mo are not present in regolith based on Mars rover analysis [ 3 , 4 ]. Another considerable issue that requires attention is how the atmospheric composition and density of Mars affects plant growth, plant gas exchange and ultimately N fixation. Mars has 31 times less atmospheric N at equal density than Earth. It seems prudent to test whether a condensed atmosphere of that composition would be able to support rhizobia N fixation [ 5 ]. Plants would likely have to be grown in a biosphere—an enclosed area with artificial heat and light. While a biosphere would be a necessity, it is unclear if only atmospheric composition would need to be altered or if also atmospheric density, as altering the stoichiometry of an enclosure could be energy taxing. The drastic difference between the stoichiometry of terrestrial and Martian atmosphere N content (78% versus 1.2%, respectively) may be pivotal for Rhizobia spp. and their ability to fix atmospheric N. A possible way to overcome this, should it be an issue, would be to breed plants and symbiotes for low N atmospheres. Interestingly, the lack of Nitrosomonas spp. and Nitrobacter in regolith would likely keep bioaccessible N in the form of NH 4 , with it not being converted to nitrate. Future experiments could focus on whether nitrification is a benefit to plants or if nitrifying bacteria addition is beneficial for N cycling in regolith. As observed at the end of our study, little NH 4 was remaining in regolith, and based on plant growth there was likely N within the plant, although this was not determined; future tissue quality analysis with inoculated and uninoculated plants could confirm this concept and could provide invaluable data for astroagricultural success on Mars."
} | 4,922 |
37267326 | PMC10287537 | pmc | 8,425 | {
"abstract": "Abstract Bacteria that form long-term intracellular associations with host cells lose many genes, a process that often results in tiny, gene-dense, and stable genomes. Paradoxically, the some of the same evolutionary processes that drive genome reduction and simplification may also cause genome expansion and complexification. A bacterial endosymbiont of cicadas, Hodgkinia cicadicola , exemplifies this paradox. In many cicada species, a single Hodgkinia lineage with a tiny, gene-dense genome has split into several interdependent cell and genome lineages. Each new Hodgkinia lineage encodes a unique subset of the ancestral unsplit genome in a complementary way, such that the collective gene contents of all lineages match the total found in the ancestral single genome. This splitting creates genetically distinct Hodgkinia cells that must function together to carry out basic cellular processes. It also creates a gene dosage problem where some genes are encoded by only a small fraction of cells while others are much more abundant. Here, by sequencing DNA and RNA of Hodgkinia from different cicada species with different amounts of splitting—along with its structurally stable, unsplit partner endosymbiont Sulcia muelleri— we show that Hodgkinia does not transcriptionally compensate to rescue the wildly unbalanced gene and genome ratios that result from lineage splitting. We also find that Hodgkinia has a reduced capacity for basic transcriptional control independent of the splitting process. Our findings reveal another layer of degeneration further pushing the limits of canonical molecular and cell biology in Hodgkinia and may partially explain its propensity to go extinct through symbiont replacement.",
"conclusion": "Conclusions The transcriptomes of cicadas’ bacterial endosymbionts, like their genomes, embody two opposite extremes. Sulcia exhibits highly conserved transcript abundance ratios and patterns of RNA-seq coverage that line up with biological expectations. Hodgkinia , meanwhile, shows diminished transcriptional control and transcribes genes in proportion to their sometimes wildly imbalanced DNA abundance. In either case, it is difficult to quantify the fitness consequences of these transcriptional outcomes. We expect that at least some of the outcomes we observe in Hodgkinia , such as widespread antisense transcription in D . near semicincta and failure to compensate for massive gene dilution in M. septendecim , are costly. The magnitudes of these costs are dependent on translational compensatory changes—if any occur—and, in the latter case, gene product transport. Both of these processes have yet to be characterized in Hodgkinia . As with the reproductive burden cicadas experience in order to transmit a complete Hodgkinia gene complement to their eggs following extensive lineage splitting, we speculate that these events are costly for the symbiosis and may tip the scales in favor of Hodgkinia extinction and replacement with a new endosymbiont ( Campbell et al. 2018 ; Matsuura et al. 2018 ; Wang et al. 2022b ).",
"introduction": "Introduction Vertically transmitted bacterial endosymbionts that form very stable and long-term association with host cells, including the ancestors of mitochondria and plastids, can lose most of the genes originally encoded by their free-living ancestors ( Andersson and Kurland 1998 ; Green 2011 ; Gray 2012 ). Endosymbiont genomes are often small in size, stable in structure, and densely packed with a core set of functional genes ( Boore 1999 ; Tamas et al. 2002 ; McCutcheon and Moran 2011 ; Graf et al. 2021 ). While such tiny, stable, and gene-dense endosymbiont genomes have evolved again and again in diverse host lineages, some endosymbiont and organelle genomes have secondarily become unstable, expanding in size through the accumulation or proliferation of non-coding and nonfunctional DNA. The cicada endosymbiont Candidatus Hodgkinia cicadicola (hereafter, Hodgkinia ) and the mitochondria of some sucking lice and flowering plants have all evolved multichromosomal genomes several times larger than those of closely related lineages despite virtually no change to their overall gene repertoire ( Shao et al. 2012 ; Sloan et al. 2012 ; Campbell et al. 2015 ; Campbell et al. 2017 ). In the case of Hodgkinia —and in contrast to mitochondria, where different chromosomes are mixed together throughout the mitochondrial compartments of a cell—genome fragmentation occurs in parallel with cellular diversification such that the total gene set is divided among distinct Hodgkinia cell populations which are present at different relative abundances in the host ( Van Leuven et al. 2014 ; Łukasik et al. 2018 ). As a result, genes critical both to Hodgkinia 's symbiotic role in nutrient biosynthesis along with genes central to basic bacterial cell function can differ in abundance by orders of magnitude within the same insect. This gene dosage problem raises the question of whether complex Hodgkinia can correct for large differences in gene abundance in some way, for example through transcriptional up-regulation of lowly abundant genes ( Campbell et al. 2015 ; Łukasik et al. 2018 ). The Hodgkinia genome has the expected single circular-mapping chromosome structure in many cicadas ( McCutcheon et al. 2009b ; Van Leuven et al. 2014 ; Łukasik et al. 2018 ). In some cicadas, however, Hodgkinia has independently undergone varying degrees of genome fragmentation via cell lineage splitting ( Łukasik et al. 2018 ; Campbell et al. 2017 ). Compared to the unsplit ancestral genome, individual split genomic lineages lack functional copies of many essential genes, but these losses occur in a complementary fashion such that the unsplit gene set is maintained at the level of the total Hodgkinia population in each cicada ( Van Leuven et al. 2014 ; Łukasik et al. 2018 ). The complementary genome erosion of each lineage enforces transmission of all Hodgkinia genomes to the subsequent host generation, resulting in an expansion of the total Hodgkinia genome from the perspective of the host ( Campbell et al. 2015 ). In extreme cases, this splitting process results in genome complexes consisting of at least a dozen lineages and totaling over 1.5 Mb in length, a more than tenfold increase in genome size relative to single-lineage Hodgkinia genomes ( Campbell et al. 2017 ). Importantly, comparisons between these largest Hodgkinia complexes show extreme variation in splitting outcomes with respect to the size and gene content of their constituent genomes, which suggests that splitting does not converge on a particular endpoint or optimum ( Campbell et al. 2017 ). While splitting results in an expansion of the total unique Hodgkinia genome found in each cicada, each individual genome lineage experiences only gene loss and genome reduction. Lineage splitting can therefore only decrease the overall abundance of functional Hodgkinia genes in the system, dependent on how many and which genome(s) a given gene resides. Importantly, gene products of even the mostly lowly abundant Hodgkinia genes must be shared among all lineages in a given host to preserve their collective function. While we have shown by in situ hybridization that Hodgkinia genomes and ribosomal RNAs are contained by their respective cell boundaries, the biochemistry of these cells must somehow behave as though these boundaries do not exist or are easily crossed ( Campbell et al. 2015 ; Łukasik et al. 2018 ). Genomics shows that many Hodgkinia genes within the same biochemical pathway have differential gene dosages that would result in 10- or 100-fold disruptions in pathway stoichiometry if left uncorrected. These differences are well in excess of those associated with dosage sensitivity and haploinsufficiency in eukaryotes and could introduce choke points in the enzyme kinetics of essential processes like nutrient biosynthesis ( Papp et al. 2003 ; Morril and Amon 2019 ). These dosage disruptions also far exceed the modest several-fold capacity for gene-specific transcriptional tuning exhibited by some insect endosymbionts in response to changes in their hosts’ nutrition or developmental stage ( Moran et al. 2005a , Stoll et al. 2009 ; Wilcox et al. 2003 ). Nevertheless, endosymbionts such as Hodgkinia remain able to somewhat regulate the expression of RNAs and proteins at a relative level within a genome, because transcripts such as those from rRNA, tRNA, RNase PRNA, and protein chaperones are more abundant than most other transcripts ( Wilcox et al. 2003 ; Van Leuven et al. 2019 ; Husnik et al. 2020 ). It is therefore possible that some baseline level of constitutive gene expression control remains in Hodgkinia . \n Hodgkinia 's predisposition to splitting may owe in part to its high rate of sequence evolution, a feature also observed in the huge, fragmented mitochondrial genomes of the angiosperms Silene conica and Silene noctiflora ( Sloan et al. 2012 ; Van Leuven et al. 2014 ). This is contrasted by the roughly 50–100 times lower nucleotide substitution rate exhibited by Hodgkinia 's partner endosymbiont, Candidatus Sulcia muelleri (hereafter, Sulcia ) ( Van Leuven et al. 2014 ). While Hodgkinia genomes are structurally unstable and vary widely in size, Sulcia tends to be much more stable. Following at least 250 million years of strict host-association, Sulcia genomes from distantly related hosts show almost perfect gene co-linearity and very similar gene sets ( Moran et al. 2005b ; Bennett and Moran 2015 ) [although a broader sampling of Auchenorrhynchan insects shows that several genomic inversions have occurred in different Sulcia lineages ( Deng et al. 2022 )]. Likewise, while several cicada groups have replaced Hodgkinia with fungal endosymbionts, Sulcia is retained in every cicada species examined to date ( Matsuura et al. 2018 ; Wang et al. 2022b ). Given the diversity of Hodgkinia genome size and organization and the relative structural stasis of Sulcia genomes in cicadas, this system constitutes an elegant natural experiment for evaluating the downstream transcriptional consequences of wild swings in gene dosage resulting from endosymbiont genome instability. To characterize the transcriptional activity of Hodgkinia and Sulcia genomes relative to their genomic abundance, we sequenced DNA and RNA from the symbiotic organs of 18 cicadas representing six species encompassing a spectrum of Hodgkinia complexity. We find that Hodgkinia exerts limited transcriptional control compared to Sulcia and is unable to transcriptionally compensate for the massive effect of gene dosage imbalance that is produced by lineage splitting.",
"discussion": "Discussion Hodgkinia Does Not Compensate for Transcriptional Consequences of Gene Dosage Imbalance The process of splitting into multiple interdependent cell lineages combined with complementary gene loss has resulted in varied and sometimes extreme gene dosage outcomes for the Hodgkinia populations contained in each cicada ( Campbell et al. 2017 ; Łukasik et al. 2018 ). We considered two possible outcomes that would reflect compensation for this change at the level of transcription: widespread overproduction of mRNA to guarantee sufficient transcript abundance (overcompensation, fig. 1 B ) and fine-tuned compensatory regulation to rescue the transcription levels of dosage-depleted genes (complementation, fig. 1 C ). Our analysis of genome relative abundance and transcription in Hodgkinia of multiple complexity levels shows that neither of these adaptive responses occurs. Rather, the transcriptional changes that occur in Hodgkinia composed of 2, 4, 5, and 12+ cell lineages consistently, strongly, and simply reflect the gene dosage of corresponding genes on their genomes. Some form of compensation could, in principle, occur at the level of translation. This would presumably rely on factors external to Hodgkinia , particularly since the least abundant Hodgkinia cell lineages in the species examined here tend to encode relatively limited complements of translation-related genes compared to more abundant lineages ( Campbell et al. 2017 ; Łukasik et al. 2018 ). Our observations could also be affected by the age of the cicadas sampled, which, as fully grown adults nearing the ends of their lives, may no longer be as reliant on the proper functioning of their nutritional endosymbionts in one or both sexes. However, given the lack of conservation in Hodgkinia transcript abundance across cicada species, and the relative conservation we see in Sulcia transcription, we favor the idea that Hodgkinia simply tolerates the transcriptional consequences of gene dosage changes, even quite extreme ones. This is not to say that such changes are always selectively neutral in Hodgkinia . Given that our results are most consistent with a lack of transcriptional response in Hodgkinia after splitting, our finding that genes that had been lost in one of T. undata 's two Hodgkinia lineages had lower per-cell transcription could suggest that lineage-specific gene losses are more likely to be fixed when they occur in lowly-expressed genes. Additionally, while the dosage outcomes in highly complex Hodgkinia are not deterministic, some mechanism seems to favor the retention of certain Hodgkinia genes at a greater total abundance, and this is reflected in their transcript abundance ( Campbell et al. 2017 ; Łukasik et al. 2018 ). Basic Transcriptional Control Shows Signs of Erosion in Hodgkinia The transcriptional machinery encoded by the bacterial endosymbionts with the tiniest genomes is extremely rudimentary ( McCutcheon and Moran 2011 ). Despite this, we found evidence for at least some degree of transcriptional control in two such endosymbionts. All Sulcia and Hodgkinia transcriptomes produced more alignments to genes in the sense orientation than in the antisense orientation, suggesting that open reading frames are preferably transcribed over random positions on the opposite strand of DNA. We also observed consistently high chaperone gene expression in both endosymbionts, a recurring feature of endosymbiont transcriptomes thought to be of functional importance to the endosymbiotic lifestyle ( Fares et al. 2002 ; Stoll et al. 2009 ; McCutcheon and Moran 2011 ; Luck et al. 2015 ; Medina Munoz et al. 2017 ). This occurs despite the loss of the rpoH -encoded σ 32 heat shock sigma factor, which modulates the expression of chaperone genes in free-living bacteria ( Neidhardt and VanBogelen 1981 ; Yamamori and Yura 1982 ; Grossman et al. 1987 ). Across all six host species examined, Hodgkinia and Sulcia differed in two potential indicators of transcriptional control. First, we found that RNA-seq coverage declined predictably at gene ends in Sulcia while the high coverage typical of transcriptional start sites frequently extended past annotated genes in Hodgkinia , possibly indicating transcriptional read-through. The gene contents of these two endosymbionts point to a potential mechanistic explanation: Sulcia , unlike Hodgkinia , retains rho and its cofactor nusA , which have well-characterized roles in transcription termination ( Schmidt and Chamberlin 1984 ; Richardson 2002 ; McCutcheon et al. 2009a , 2009b ; Łukasik et al. 2018 ). Second, Hodgkinia endosymbionts showed consistently higher levels of antisense transcription than their coresident Sulcia , although this may simply be a reflection of increased transcriptional read-through at genes located adjacent to a gene on the opposite strand. Surprisingly, the single-lineage Hodgkinia endosymbiont of D . near semicincta stood out in its apparent loss of transcriptional control, exhibiting a considerably higher proportion of antisense transcription than any other endosymbiont lineage we examined. The fact that antisense transcription in Sulcia from the D . near semicincta samples was not correspondingly high suggests that this effect is not a technical artifact. A previous comparative genomic analysis of endosymbiont RNA polymerases identified a deletion of seven amino acid residues long in the σ 3 subunit from Hodgkinia in a very closely related cicada species, D. semicincta , and predicted that this loss could impede recognition of an extended −10 box promoter element ( Rangel-Chávez et al. 2021 ). Promoter elements have not been characterized in Hodgkinia or in other endosymbionts with tiny genomes, and we have similarly found no recognizable sequence motifs upstream of Hodgkinia or Sulcia start codons regardless of transcription level ( supplementary figs. S7–S8, Supplementary Material online). However, we note that the rpoD gene in Hodgkinia from D . near semicincta , like D. semicincta , lacks this portion of the σ 3 subunit found in most other Hodgkinia genomes ( supplementary fig. S9, Supplementary Material online), although similar deletions in rpoD genes in Hodgkinia from M. septendecim are evidently not accompanied by a correspondingly high level of antisense transcription ( supplementary figs. S5 and S9, Supplementary Material online). We also found that Hodgkinia transcript abundances in D . near semicincta were weakly correlated with their homologs’ relative abundances in the single-lineage Hodgkinia of T. ulnaria in contrast to the strong correlation observed in the Sulcia transcriptomes of those cicadas. It is unclear to what extent host-specific losses in Hodgkinia transcriptional control may have contributed to this lack of conservation versus the 50+ million years of evolutionary divergence between these Hodgkinia lineages ( Marshall et al. 2018 ; Wang et al. 2022b ). Gene Products May Be Spread Extremely Thin in Complex Hodgkinia On one hand, the unresponsiveness of Hodgkinia transcription to extreme gene dosage outcomes is unsurprising given that Hodgkinia encodes no transcription factors or alternative sigma factors and has even accumulated functionally important losses to basic transcriptional machinery ( McCutcheon et al. 2009b ; Galán-Vásquez et al. 2016 ; Łukasik et al. 2018 ; Rangel-Chávez et al. 2021 ). Even in unsplit Hodgkinia lineages with uniform gene dosage, precise ratios of relative transcript abundance do not appear to be conserved. On the other hand, Hodgkinia 's unresponsiveness is surprising because of what it implies about its biology, specifically its apparent tolerance for extreme unbalancing of essential transcripts’ absolute abundance. In M. septendecim , where many genes may be present in fewer than ten percent of cells, we found no evidence for a generalized up-regulation of Hodgkinia transcription. In fact, Hodgkinia 's relative contributions to DNA and RNA coverage in this system imply an overall reduced transcriptional activity. While in situ hybridization has shown that rRNA and genomic DNA are not shared among cells in complex Hodgkinia , the endosymbiont's continued existence necessarily implies the movement of either mRNA, protein, metabolites, or some combination of these between cells by an unknown mechanism ( Campbell et al. 2015 ; Łukasik et al. 2018 ). The likelihood of a biologically important encounter between two Hodgkinia proteins could therefore be limited not just by the abundance of the genes by which they are encoded but also by those genes’ spatial distribution within the cicada bacteriome. In the absence of massive complementation or overcompensation at the level of protein synthesis, it is conceivable that the biochemistry of the most complex Hodgkinia occurs slowly or inefficiently relative to their single-lineage counterparts. Conclusions The transcriptomes of cicadas’ bacterial endosymbionts, like their genomes, embody two opposite extremes. Sulcia exhibits highly conserved transcript abundance ratios and patterns of RNA-seq coverage that line up with biological expectations. Hodgkinia , meanwhile, shows diminished transcriptional control and transcribes genes in proportion to their sometimes wildly imbalanced DNA abundance. In either case, it is difficult to quantify the fitness consequences of these transcriptional outcomes. We expect that at least some of the outcomes we observe in Hodgkinia , such as widespread antisense transcription in D . near semicincta and failure to compensate for massive gene dilution in M. septendecim , are costly. The magnitudes of these costs are dependent on translational compensatory changes—if any occur—and, in the latter case, gene product transport. Both of these processes have yet to be characterized in Hodgkinia . As with the reproductive burden cicadas experience in order to transmit a complete Hodgkinia gene complement to their eggs following extensive lineage splitting, we speculate that these events are costly for the symbiosis and may tip the scales in favor of Hodgkinia extinction and replacement with a new endosymbiont ( Campbell et al. 2018 ; Matsuura et al. 2018 ; Wang et al. 2022b )."
} | 5,281 |
40268922 | PMC12019592 | pmc | 8,426 | {
"abstract": "Organosulfur compounds (OrgS) are fundamental components of life’s biomass, yet the cycling of these compounds in the terrestrial deep subsurface, one of Earth’s largest ecosystems, has gone relatively unexplored. Here, we show that all subsurface microbial genomes reconstructed from Soudan Underground Mine State Park have the capacity to cycle organic sulfur species. Our findings suggest that OrgS degradation may be an integral link between the organic and inorganic sulfur cycle via the production of sulfite and sulfide. Furthermore, despite isolation from surface ecosystems, most Soudan microorganisms retained genes for dimethylsulfoniopropionate and taurine biosynthesis. Metagenomic analyses of an additional 54 deep subsurface sites spanning diverse lithologies revealed the capacity for OrgS cycling to be widespread, occurring in 89% of assembled metagenomes. Our results indicate that consideration of OrgS cycling may be necessary to accurately constrain sulfur fluxes, discern the energetic limits of deep life, and determine the impact of deep subsurface biogeochemical sulfur cycling on greater Earth system processes.",
"introduction": "Introduction The biogeochemical cycling of organic and inorganic sulfur (S) compounds is tightly intertwined and central to supporting microbial life through S uptake and incorporation into biomass or redox couplings that drive metabolic processes. These microbially-driven sulfur transformations regulate biogeochemical S cycling and couple with other globally important cycles such as Fe, C, and N 1 – 5 . Organosulfur compounds (OrgS) have been increasingly recognized as a central aspect of biogeochemical sulfur cycling in terrestrial and marine environments 6 – 8 . Investigations of OrgS cycling in low-sulfate systems such as Archean oceans 9 and modern Lake Superior 10 have demonstrated the importance of OrgS in the total sulfur budget and as an inorganic S source that drives chemoautotrophic or dissimilatory metabolisms. Given that microbial assimilation and mineralization of OrgS directly control the availability of inorganic S substrates (Fig. 1 ), considering these processes in tandem is paramount to obtaining a comprehensive understanding of biogeochemical sulfur cycling and, more broadly, the behavior of Earth’s Critical Zone across modern and deep time. Fig. 1 The biogeochemical sulfur cycle. A simplified diagram showing key transformations in the cycling of inorganic and organic sulfur. A selection of diverse, metabolically and environmentally important organosulfur compounds produced intracellularly is shown. DMS, dimethylsulfide; DMSO, dimethylsulfoxide; DMSP, dimethylsulfoniopropionate; Me, methyl group. The microbial production of OrgS compounds is ubiquitous in marine 11 – 13 , freshwater 14 , and terrestrial systems 15 , where they serve as building blocks for biomass 16 , 17 , osmoprotectants 18 – 20 , cryoprotectants 21 , protection against oxidative stress 22 , or as terminal electron acceptors in dissimilatory processes. Production of many OrgS compounds, such as dimethylsulfoniopropionate (DMSP) and taurine, was historically thought to be constrained to eukaryotes and is dominated by eukaryotes in extant marine ecosystems 16 , 20 , 23 . However, the identification of novel DMSP and taurine biosynthesis pathways in heterotrophic bacteria 20 , 24 , 25 , and evidence of these pathways occurring alongside mineralization pathways 26 , provide a mechanism for OrgS cycling to persist in prokaryote-dominated communities and expand the environments in which OrgS is expected to play an essential role. The terrestrial crust is one of Earth’s largest ecosystems, hosting 2–19% of total global biomass 27 – 29 . This biomass pool requires the synthesis of OrgS compounds, like cysteine, methionine, and coenzyme A, for growth. Likewise, as these organisms die and release OrgS extracellularly, their removal through reincorporation or mineralization is inevitable. However, the importance of OrgS cycling in crustal environments has received little attention despite the potential for OrgS to fuel metabolism in such settings. Several studies have suggested the occurrence of linked organic and inorganic S processes in these vast subsurface environments. Modeling-based approaches by Fakhraee and colleagues 6 demonstrate that OrgS mineralization could support deep biosphere communities of inorganic S reducers. Recently, sediments underlying low-sulfate waters (<30 μM) were observed to have increased sulfate concentrations fueled by in situ sulfate production via OrgS degradation 10 . This mechanism of sulfate production was supported by sediment incubation experiments that demonstrated sulfate production coupled to the degradation of OrgS compounds (i.e., taurine, sodium dodecyl sulfate, cystine, and methionine) 10 . Finally, the ability of sulfate-reducing bacteria to ferment OrgS compounds like cysteate to acetate, ammonia, sulfide, and sulfate 30 suggests these compounds are important substrates capable of fueling life. Although OrgS cycling genes, primarily involved in dimethylsulfoxide (DMSO) and taurine transport, have been observed in terrestrial deep biosphere studies 31 – 33 , our understanding of the extent and nature of microbial OrgS cycling in these systems is limited. Inorganic S cycling is well known to play a key role in deep terrestrial subsurface systems 34 – 39 and past studies have detected the presence of some OrgS cycling genes in these systems 31 – 33 . Therefore, we hypothesize that the cycling of OrgS compounds, such as DMSP and taurine, could serve as a key regulator of the inorganic S pool, and thus microbial metabolism, in the deep biosphere. A more extensive understanding of the functional capacity of the deep biosphere to cycle OrgS and the entwined nature of inorganic and organic biogeochemical S cycles encoded in these metagenomes will elucidate the role of OrgS in deep terrestrial subsurface systems. Here, we address this gap using shotgun metagenomic sequencing of anoxic groundwater and suboxic to oxic outflow channel fluids at Soudan Underground Mine State Park, as direct quantification of OrgS is difficult in these highly saline waters. Metagenomic insights into this microbial community reveal the genetic potential for widespread OrgS cycling, including pathways that link organic and inorganic sulfur cycling, and suggest that OrgS could play a significant role in supporting deep life. Further, we examine 54 additional terrestrial subsurface metagenomes to explore the ubiquity and diversity of OrgS cycling genes and demonstrate the widespread genetic potential for deep subsurface OrgS cycling. This work extends our understanding of biogeochemical S cycling in the deep terrestrial subsurface by underscoring the potential role of OrgS as a source and sink of inorganic sulfur and positing that microbial cycling of OrgS is likely globally significant.",
"discussion": "Results and discussion Microbial synthesis and utilization of OrgS likely occurred early in Earth’s history 40 and has diversified with time, resulting in extant organisms capable of intracellularly cycling OrgS compounds with varying carbon structures and sulfur redox potentials (Fig. 1 ). Furthermore, the sulfurization of labile organic molecules 41 creates a diverse pool of substrates for microorganisms to utilize. Synthesis of microbial biomass, including OrgS compounds such as cysteine, is typically a net-exergonic process under reducing conditions when oxidized S and N sources are available 42 . OrgS production through biomass creation is expected in deep subsurface systems such as Soudan Mine, where groundwaters are typified by their low reduction potential and the presence of sulfate 41 , 43 , 44 . Indeed, spectroscopic detection of reduced and oxidized OrgS in biofilm-associated mineral particulates from the Soudan Mine provides further evidence of OrgS synthesis (Fig. S2 ). OrgS catabolism is also expected, as the release of labile carbon backbones, such as the generation of pyruvate via cysteine degradation, provides substrates for central carbon metabolism 45 . Additionally, OrgS mineralization can generate either sulfite or sulfide, depending on the starting compound and pathway (Fig. 2a ). These inorganic S species sourced from organosulfur have been demonstrated to feed dissimilatory redox reactions and drive cellular energy production 46 , 47 . Thus, the cyclical metabolism of OrgS is predicted to play a central role in biogeochemical S cycling in diverse terrestrial subsurface systems. Fig. 2 Organosulfur cycling potential at Soudan Mine. Overview of proposed linked inorganic and organic sulfur cycles at Soudan Mine. a Schematic diagram depicting key intracellular pathways for dissimilation of inorganic sulfur (pink), transformation of taurine (blue), transformation of dimethylsulfide, dimethylsulfoxide, or dimethylsulfoniopropionate (DMS(O)(P), green), and cysteine and methionine production and utilization (yellow). Reactions show only key sulfur species, and not all intermediate steps are shown. Genes encoding the enzyme that catalyzes each reaction are indicated in italics, and the percentage of Soudan Mine metagenome assembled genomes (MAGs) able to carry out that reaction is shown in bold. b Taxonomic distribution of key genes found in the Soudan Mine. Phyla are listed based on evolutionary relationships, with the number of MAGs in that phylum in parentheses. Solid squares indicate the presence of the indicated gene or complete gene complex, while cross-hatched squares represent the presence of some but not all genes in the complex. The bottom row indicates the total number of MAGs in which that gene or group of genes was detected. See Supplementary Data S1 for gene abbreviations and Supplementary Data S2 for detailed reactions. Asterisks indicate genes that were identified using HMSS2 rather than METABOLIC. To see gene hits from METABOLIC, see Supplementary Data S7 (MAGs) and S8 (assemblies). While pathways generating both sulfide and sulfite appear in the Soudan Mine metagenome (Fig. 2a ), the cycling of sulfite-producing compounds is predicted to play a more important ecological role. Sulfide released via mineralization of cysteine or methionine may serve as an endpoint in Soudan groundwaters because it is quickly scavenged by and stably bound to reduced iron 48 . Conversely, compounds producing sulfite upon microbial decomposition, such as DMSO and taurine, are of interest because sulfite can be biotically oxidized to sulfate to generate ATP. Furthermore, sulfate (0.7 mM) is the most abundant dissolved terminal electron acceptor in Soudan groundwaters, and sulfate reduction coupled to CH 4 oxidation is predicted to be a highly energetically favorable metabolism at this site 43 , 49 . Here, we focus on aspects of the OrgS cycle that involve DMSO and taurine because they may fuel the sulfate reduction that is predicted to be the dominant process fueling this subsurface community. In our Soudan shotgun metagenomic datasets, we recovered a total of 65 medium to high-quality MAGs considered in further analyses. Reconstructed genomes are predominantly Bacteria, with two archaeal MAGs identified as Methanolobus . The Bacterial community spans 11 phyla, 16 classes, and 36 orders (Supplementary Data S3 and S6 ). Many MAGs align with previous diversity assessments at this site 44 . However, the detection of three previously unobserved classes (Kapabacteria, Spirochaetota, and UBP18) and several taxonomically novel MAGs (Supplementary Data S3 ) extends our understanding of site diversity and larger subsurface community membership 50 , 51 . The potential for OrgS cycling is similar in samples taken directly from anoxic boreholes and those taken from the suboxic to oxic fluid outflow channel. MAGs and bulk assemblies include protein-coding genes for OrgS production, utilization, and mineralization (Fig. 2 , Supplemental Data S7 and S8 ). Additionally, MAGs and metagenomes encode full or partial pathways involved in dissimilatory oxidation or reduction of inorganic S. Genes involved in OrgS catabolism commonly co-occur with dissimilatory S genes, and 25 MAGs (38.5%) have the capacity to degrade OrgS to sulfite and then reduce or oxidize that sulfite (Fig. S3 ). This includes genes involved in the catabolism of sulfoacetaldehyde, isethionate, or cysteate ( xsc , islA , cuyA ) appearing alongside genes enabling sulfite reduction to sulfide ( cysIJ , dsrAB red , asrAB ) or sulfite oxidation to sulfate ( soeABC , aprAB ox + sat ox , sorAB ). The co-occurrence of organic and inorganic S cycling genes within MAGs suggests that a single organism has the metabolic potential to couple OrgS catabolism to the dissimilatory oxidation or reduction of sulfite or sulfide. This direct, intracellular link between the inorganic and organic S cycles would provide energy for cells in this nutrient-depleted subsurface system. Furthermore, the established importance of metabolic handoffs in deep subsurface settings 35 suggests that even cells in which pathways for sulfite generation and metabolism do not co-occur directly could be integral to the cycling of organic and inorganic S intermediates. This connection highlights the relevance of OrgS degradation as a source and sink of inorganic substrates within the deep terrestrial biosphere. DMS(O)(P) cycling is ubiquitous across deep subsurface metagenomes The cycling of DMSP and the related molecules dimethylsulfide (DMS) and DMSO, referred to collectively as DMS(O)(P), has been demonstrated to play an integral role in biogeochemical sulfur cycles in marine and terrestrial systems 15 , 52 – 54 . Genes involved in DMSO reduction have also been detected in two deep biosphere settings, suggesting that DMS(O)(P) cycling could be an important process in the deep subsurface 31 , 32 . Protein-coding genes involved in DMS(O)(P) cycling were found in all bulk assemblies and 95.4% of Soudan MAGs (Figs. S3 , S4 ). Four MAGs, appearing in 17 of the 21 metagenomes investigated, encode dysB , a canonical gene indicative of DMSP biosynthesis 24 . A BLAST search of Soudan protein sequences against the related DMSP biosynthesis gene dysGD , which does not currently have an HMM, revealed several hits; however, the low amino acid identity to known DsyGD sequences (<50%) and lack of full coverage suggest these proteins in Soudan may not be functional 20 . Additionally, 6 MAGs can transport DMSP into the cell via the transporter gene dddT . We detected two pathways for DMSP catabolism: demethylation to methylmercaptopropionate (MMPA) via DMSP demethylase ( dmdA ) and cleavage to DMS via DMSP lyase ( dddL, dddP, or dddD ) (Fig. 2 ). The former was detected in one MAG from the order Rhodobacterales and appears predominantly in outflow channel metagenomes, while DMSP lyases were detected in six MAGs spanning Rhodobacterales, Acetobacterales, Pseudomonadales, and Nevskiales. An additional 3 MAGs encode dmdBCD to degrade MMPA and produce methanethiol (MeSH). Every bulk assembly and 25 MAGs encode methanethiol S-methyltransferase ( mddA ), which methylates MeSH and generates DMS. The most common pathway for DMS oxidation to DMSO identified in this dataset involves dimethylsulfide dehydrogenase ( ddhAB ), which was detected in every bulk assembly and 35 MAGs (Figs. 2 , S4 ). Together, MAGs encoding DdhAB comprise 20.7%–65.4% of the total microbial community in each sample (Supplementary Data S4 ). Pathways using dimethylsulfide S-monooxygenase ( dsoBED ) and trimethylamine monooxygenase ( tmm ) for DMS oxidation were also detected in three and one MAG, respectively (Fig. 2 ). Seven MAGs, spanning all 21 metagenomes, can transform DMSO to methanesulfonate via msuE , and three encode msuD , which produces sulfite via desulfonation of the intermediate methanesulfonate. One MAG encodes the catalytic subunit of DMSO reductase ( dmsA ), allowing DMSO to serve as a terminal electron acceptor in anaerobic respiration. Four additional MAGs contain the DMSO reductase subunit involved in electron transfer ( dmsB ) but were missing dmsA . Genes encoding Cym-type DMSO reductase ( cymA ) and TMAO reductase ( torAC ), which serve a functionally similar role to dmsAB , were not detected in any samples. The absence of DMSO reduction suggests sulfite could eventually be produced via methanesulfonate, a process which requires molecular oxygen (Fig. 2a ). Though the reaction rate of this pathway may be constrained in Soudan’s anoxic groundwater, experimental and modeling-based studies that demonstrate H 2 production via the radiolysis of water in similar Precambrian Shield brines suggest that limited oxygen production may be possible in this environment 55 , 56 . Additionally, chlorite and nitric oxide dismutation have been suggested as a source of molecular oxygen in ancient groundwaters 57 , and the detection of the chlorite dismutase gene ( cdo ) in four Soudan metagenomes supports this as a potential oxygen production mechanism in Soudan Mine. These findings are well-aligned with previous studies of the terrestrial deep biosphere, which have also detected dmsAB 31 , 32 , but widely expand the array of DMS(O)(P)-related genes documented in these settings. The detection of dysB provides evidence of DMSP biosynthesis potential in a deep subsurface environment. The genetic potential to produce DMSP and mineralize it to DMS provides a mechanism to replenish and support a continuous, biotically driven DMS(O)(P) cycle. The ubiquity of genes involved in DMSP synthesis or uptake and high relative abundance of MAGs encoding these genes (9.1%–62.2% of borehole microbial communities) underlines the relevance of these processes in the deep biosphere (Supplementary Data S4 ). While DMS oxidation to DMSO occurs primarily through tmm in marine environments 12 , 13 , our findings indicate that the dominant pathway for DMS oxidation in Soudan relies on ddhAB . The differences between the two systems likely reflect differing redox conditions. The tmm is a dioxygenase enzyme requiring oxygen to drive DMS conversion to DMSO. While the production of molecular oxygen is likely possible in Soudan groundwaters 55 – 57 , the use of the ddhAB facilitates energy conservation because this complex translocates hydrogen ions and creates a reduced cytochrome. The presence of DMSP production alongside DMS oxidation, and ultimately sulfite production in the presence of oxygen, underscores the importance of continued investigation into organosulfur cycling in the deep subsurface. Furthermore, our results suggest that incorporation of OrgS into thermodynamic and reactive transport models is necessary because OrgS cycling could fuel inorganic S delivery and subsequent S redox. Pathways to cycle taurine are diverse and prevalent in the deep biosphere Taurine is ubiquitous in marine settings where it serves as an important source of C, N, S, and energy for marine heterotrophic bacteria 16 , 17 . The anaerobic dissimilation of taurine has been proposed as a mechanism to drive sulfite respiration 17 , and microbial communities in freshwater sediments are known to degrade taurine and accumulate sulfate 10 . Thus, taurine is of particular interest as a potential energy source in the oligotrophic terrestrial deep biosphere. We identified protein-coding genes involved in the import and degradation of taurine and related intermediates in 64.6% of reconstructed MAGs and all bulk assemblies (Fig. 2 , S4 ). The capacity to import taurine-related compounds was detected in twelve MAGs, including two that encode the taurine importer TauABC and eleven capable of general alkanesulfonate uptake via SsuABC. Additionally, we detected two pathways for taurine degradation to sulfite, either directly or via the intermediate sulfoacetaldehyde (Fig. 2 ). Three MAGs, all within the phylum Pseudomonadota, encode taurine dioxygenase ( tauD ) to degrade taurine directly to sulfite. These MAGs are more abundant in outflow channel samples, comprising 3.4–5.4% of the total microbial community, compared to borehole fluids (<0.1% of the community). TauD allows taurine to be used as a sulfur source during aerobic growth and appears in 90.9% of metagenomes from the more oxic outflow channel and only 20.0% of metagenomes from anoxic borehole fluids. Seven MAGs contain genes to transform taurine to sulfoacetaldehyde, including three MAGs encoding taurine:oxoglutarate aminotransferase ( toa ), two encoding taurine:pyruvate aminotransferase ( tpa ), and four encoding taurine dehydrogenase ( tauXY ). The ability to desulfonate sulfoacetaldehyde and produce sulfite via sulfoacetaldehyde acetyltransferase ( xsc ) appears in eight MAGs (Fig. 2 ). Five deep terrestrial subsurface MAGs fully encode this pathway for taurine catabolism to sulfite via sulfoacetaldehyde, while six additional MAGs can complete only one of the two necessary steps. In total, 71% of metagenomes, found in every sample except those from DDH951, encode a full pathway for taurine degradation to sulfite. Metagenomes from DDH951 encode only toa , with one also encoding islAB to mineralize the related OrgS compound isethionate to sulfite (Fig. S4 ). The ability of microorganisms to degrade taurine and related OrgS compounds in oxic and anoxic prokaryote-dominated deep, terrestrial subsurface systems is intriguing, as taurine production is primarily attributed to eukaryotes 58 . The previous identification of a cysteine sulfinic acid decarboxylase in bacteria 25 ., however, provides direct evidence for prokaryotic taurine biosynthesis and should be further investigated to understand the prevalence of this process under in situ conditions. The prevalence and diversity of taurine degradation pathways in Soudan Mine metagenomes would suggest that pathways for taurine production are also present in this system; however, we did not detect key genes (hypotaurine monooxygenase, cysteamine dioxygenase, cysteine sulfonic acid decarboxylase) involved in known microbial taurine biosynthesis pathways (Supplementary Data S1 ). These pathways require oxygen to degrade cysteine to taurine via the intermediates cysteine sulfinic acid or cysteamine, but the highly reducing fluids of Soudan Mine and other subsurface systems may necessitate a pathway that can operate under anoxic conditions. One potential anaerobic taurine synthesis route that has been documented in higher eukaryotes involves the formation and subsequent decarboxylation of cysteate to form taurine 59 . Cysteate synthesis is one of the first steps in the production of coenzyme-m, and is known to occur in bacteria and archaea 60 . The cysteate synthase gene ( cysS ) that facilitates this reaction appears in two archeal and one bacterial MAG, and all bulk assemblies from anoxic borehole fluids. No cysS genes were detected in the more oxic outflow channels, supporting the notion of anaerobic taurine synthesis in the deep subsurface. Additionally, a promiscuous glutamate decarboxylase enzyme (EC. 4.1.1.15) has been documented to use cysteate as a substrate 61 and the gene encoding this enzyme ( gadAB ) was detected in five Soudan MAGs (Fig. 2 ). While neither of these genes ( gadAB or cysS ) was found to co-occur in a single MAG, this could be due to MAG incompleteness. Even so, 30.0% of the metagenomes from anoxic borehole fluids have the capacity to complete this pathway at the community scale via the exchange of intermediate substrates (Fig. S4 ), a common metabolic strategy in the deep terrestrial subsurface 35 . The confirmation of multiple taurine degradation pathways strongly suggests a source of taurine in this system, and the protein-coding genes cysS and gadAB provide one possible mechanism for biosynthesis in an anoxic setting. As such, microbes may use taurine as an osmolyte 58 to offset the high salinity of Soudan groundwaters. The hypersaline conditions of the Soudan groundwaters may promote the synthesis and uptake of taurine, which would make it abundant and available to the microbial community. The common and diverse nature of genes involved in taurine uptake, transformation, and degradation underscores the potential importance of taurine in supporting microbial growth and diversity in deep subsurface environments. Dissimilation of inorganic sulfur is a key metabolism in the deep biosphere While the coupling of OrgS degradation to energy-generating inorganic sulfur redox reactions has received little attention, inorganic redox processes are a well-established cornerstone of metabolism in many subsurface environments, including Soudan Mine 44 , 48 , 49 . Previous work investigating the inorganic sulfur cycle at Soudan has noted genes related to redox and disproportion of intermediate oxidation S species, as well as sulfate and sulfide 44 , 48 . Here, we focus on expanding existing knowledge by incorporating the potential for OrgS to impact the biogeochemistry of the system. We focus on sulfate, sulfite, and sulfide because they are common links between the organic and inorganic sulfur cycles. The presence of genes such as phsA suggests that other intermediate forms of sulfur (thiosulfate, S(0)) are cycled here as well, but these pathways are predicted to predominantly add to the sulfite and sulfide pools. Thus, they are not expected to inhibit the processes of mineralization or dissimilatory sulfite redox focused on here. We detect protein-coding genes involved in reductive or oxidative sulfur metabolisms in every metagenome and 72.3% of MAGs, with four MAGs encoding a complete dissimilatory sulfate reduction pathway and an additional six capable of sulfide oxidation to sulfate (Fig. 2a ). These metabolisms are likely coupled to the oxidation of ferric iron and nitrate or the reduction of methane, H 2 , and more complex organics in this system 44 , 48 , 49 . A total of 26 MAGs encode sulfate adenylyltransferase ( sat) , with 17 of these identified as the reductive-type sat red responsible for the activation of sulfate to adenosine 5’-phosphosulfate (APS). Four MAGs encode the reductive-type APS reductase ( aprAB red ) to reduce APS to sulfite, three of which also contain a complete quinone-modifying oxidoreductase ( qmoABC ) complex. The QmoABC complex shuttles electrons between AprAB and the quinone pool, coupling APS reduction to energy production 62 . Dissimilatory sulfite reductase ( dsrAB red ) is present in six MAGs, and anaerobic sulfite reductase ( asrAB ) and coenzyme F 420 sulfite reductase (fsr) are each present in one, indicating the ability of those MAGs to metabolize via sulfite reduction. Every bulk assembly sampled from anoxic borehole fluids and 81.8% of those from outflow channels have the genetic potential ( sat , aprAB , qmoABC , and dsrAB all encoded ) to fully reduce sulfate to sulfide (Fig. S4 ). While only four MAGs can fully reduce sulfate to sulfide independently, seventeen others encode a subset of the genes involved. This could be an artifact of missing sequences in incomplete MAGs or could reflect a reliance on metabolic handoffs to complete this pathway as a community 35 . Six Soudan MAGs have the capacity to fully oxidize sulfide to sulfate, encoding an oxidative-type dsrAB ox for sulfide oxidation to sulfite. These and six additional MAGs are capable of sulfite oxidation to sulfate via the oxidative-type aprAB ox and sat ox (5 MAGs) or the sulfite dehydrogenases soeABC (11 MAGs) or sorAB (2 MAGs). Genes involved in the SOX pathway for thiosulfate oxidation were identified in 23 MAGs, of which four encode the complete Sox pathway ( soxAX , soxZY , soxB , and soxCD ) to oxidize sulfide to sulfate 63 . Five MAGs encode an “incomplete” sox pathway lacking the soxCD gene (Fig. 2 ). These microorganisms can oxidize thiosulfate to sulfate or oxidize sulfide to elemental sulfur (S(0)), then further oxidize the intermediate-valence S(0) to sulfate using other pathways such as reverse dissimilatory sulfite reductase (dsrAB ox ) 64 , 65 . One Gammaproteobacteria MAG in the order Burkholderiales appears capable of using this approach, as it encodes both an incomplete SOX pathway and the oxidative-type DsrAB. MAGs encoding a full sulfur oxidation pathway were identified in all outflow channel bulk assemblies and 70.0% of those from anoxic borehole fluids. Except for DDH951, MAGs capable of sulfur oxidation make up a significant portion of the microbial community in anoxic borehole fluids (8.2–41.0%) and likely support an active inorganic sulfur cycle in this deep subsurface system. Overall, the presence of both oxidative and reductive S metabolisms agrees with findings at other terrestrial subsurface sites and underlines the dynamic nature of sulfur cycling in these settings 35 , 37 , 44 . The coupling of organosulfur mineralization to dissimilatory redox reactions can fuel deep life Organosulfur degradation may produce reduced forms of inorganic sulfur (e.g., HS - ) that can be oxidized or intermediate forms of inorganic S (e.g., SO 3 2- ) that can be either further oxidized or reduced. Thus, OrgS mineralization can generate redox-reactive S species under oxidizing and reducing conditions 66 . This process extends the metabolic framework of the sulfur cycle and calls attention to a previously overlooked pool of S in organic compounds. It also bolsters our understanding of the extensive hidden, or cryptic, S cycle uncovered by metagenomic approaches when geochemical approaches alone may not adequately reveal the full extent of cycling 2 . The deep terrestrial biosphere at Soudan Mine has the genetic potential to acquire, create, transform, and mineralize a wide range of OrgS compounds (Fig. 2 ). Only a fraction of this OrgS is expected to be used in biosynthesis based on the rarity of sulfur-containing amino acids, with cysteine and methionine previously reported to comprise just ~1.1% and ~2.9% of total cellular proteins, respectively 17 , 67 . Thus, the majority of S in these compounds remains available for energy production via respiratory processes. DMSO can serve directly as a terminal electron acceptor 68 , but most other OrgS compounds must first be mineralized. The fate of sulfur in DMSP, taurine, and other OrgS is often sulfite, which can then fuel energy production 46 . As an example, taurine degradation by TauD frees sulfite to act as an electron acceptor through the interaction with the dissimilatory sulfite reductase (DsrAB), a process that could occur cryptically (Fig. 3 ). Given the highly reactive and toxic nature of sulfite, OrgS may provide a safer mechanism to carry sulfur at intermediate oxidation states until it is mineralized and used for energy production. Protein-coding genes involved in DMS(O)(P) and taurine cycling co-occur with inorganic sulfur cycling genes in 57 MAGs (87.7%), with bulk assemblies from every outflow channel sample and 70% of borehole fluid samples encoding complete pathways for OrgS degradation to sulfite as well as sulfite dissimilation (Fig. S4 ). Fig. 3 Proposed cryptic S metabolism. Example of coupled organic and inorganic sulfur cycles, with catabolism of taurine providing the substrate used in dissimilatory sulfite reduction. The Gibbs energy yield of several potential sulfite metabolisms under in situ conditions was evaluated to determine the feasibility of these strategies in Soudan Mine (Fig. S6 , Supplementary Data S10 ). These metabolisms included the oxidation and reduction of sulfite coupled to known electron acceptors (NO 3 - ) and donors (CH 4 , H 2 ). Thermodynamic calculations suggest that all investigated reactions are exergonic, with the coupling of nitrate reduction to ammonia with sulfite oxidation to sulfate providing the highest energy yield per electron. Despite being less exergonic per mole of electron, energy density calculations suggest that the reactions coupling sulfite reduction with hydrogen or methane oxidation have the potential to provide a similar amount of energy per kg of water due to the limited availability of nitrate in Soudan borehole fluids. Thus, linking OrgS mineralization to dissimilatory oxidation or reduction of the inorganic byproduct would allow a diverse array of OrgS compounds, including those lacking reductases, to serve indirectly as electron donors or acceptors and to play a cryptic but central role in deep terrestrial subsurface energy production. Organosulfur cycling is important across the terrestrial subsurface While the microbial community at Soudan is clearly capable of OrgS cycling, the iron-rich formation is not representative of all deep terrestrial environments. To investigate whether OrgS serves a central role in other terrestrial subsurface systems, we searched 286 MAGs generated from 54 metagenomes obtained through the Census of Deep Life, a program within the Deep Carbon Observatory that sampled the deep biosphere around the planet 69 . We observe patterns similar to those identified in Soudan MAGs across diverse lithologies (Fig. 4 and Supplementary Data S5 ). Genes involved in S assimilation were detected in 283 MAGs (99.0%). Genes related to the synthesis, utilization, and degradation of DMS and taurine are present across all sites, identified in 241 (84.3%) and 118 (41.3%) MAGs, respectively. Of note, 42 MAGs (14.7%) have the capacity to produce sulfite via mineralization of organosulfur, with 16 (5.6%) of these MAGs also encoding pathways for either dissimilatory reduction or oxidation of that sulfite. The gene ddhAB is the most commonly occurring OrgS gene outside of those involved in sulfur assimilation, suggesting that 120 MAGs (42.0%) are capable of oxidizing DMS to DMSO. These results support the ubiquity of OrgS cycling genes in deep subsurface metagenomes and the possibility that OrgS compounds could act as an energy source for the deep biosphere, especially considering the availability of new and ancient sulfurized organic matter in these settings 10 , 70 . Fig. 4 Global relevance of organosulfur cycling. Number of Census of Deep Life (CoDL) metagenome assembled genomes (MAGs) of the total 286 analyzed that contain each gene or complex of interest. Pink bars indicate genes involved in the dissimilation of organic sulfur. Blue bars indicate genes involved in taurine transformations. Green bars indicate genes involved in the transformation of dimethylsulfide, dimethylsulfoxide, or dimethylsulfoniopropionate (DMS(O)(P)). Yellow bars indicate genes involved in the production and utilization of cysteine and methionine. The inset map depicts sample locations (black dots) of the 54 CoDL samples assembled into MAGs. See Supplementary Data S1 for gene abbreviations and Supplementary Data S9 for all gene hits with METABOLIC. This study demonstrates the diverse, cosmopolitan genetic potential for OrgS cycling throughout the terrestrial deep subsurface, highlighting pathways encoded in deep biosphere metagenomes that link the organic and inorganic sulfur cycle and have the potential to fuel the energy-limited deep biosphere. The biosynthesis and subsequent cycling of OrgS can directly impact the availability of redox-reactive inorganic S species that act as electron donors or acceptors, and so OrgS cycling could provide an underappreciated control on terrestrial deep subsurface S fluxes. The growing recognition of OrgS as an integral aspect of biogeochemical sulfur cycling across diverse environments 6 – 8 , 10 underscores the need for a fundamental revision to our conceptualization of S cycling to acknowledge the potential impact of both the inorganic and organic aspects of this cycle. These efforts will expand our understanding of microbial community functioning and inform other avenues of research, including the search for novel biomarkers and attempts to culture the uncultured microbial majority. Continued investigation into the activity of microbial OrgS transformations is necessary to understand the full extent of deep subsurface OrgS cycling and its larger impact on Earth’s Critical Zone."
} | 9,045 |
39656810 | PMC11836721 | pmc | 8,427 | {
"abstract": "Abstract The thermoacidophilic red alga Cyanidioschyzon merolae survives its challenging environment likely in part by operating a carbon-concentrating mechanism (CCM). Here, we demonstrated that C. merolae 's cellular affinity for CO 2 is stronger than the affinity of its rubisco for CO 2 . This finding provided additional evidence that C. merolae operates a CCM while lacking the structures and functions characteristic of CCMs in other organisms. To test how such a CCM could function, we created a mathematical compartmental model of a simple CCM, distinct from those we have seen previously described in detail. The results of our modeling supported the feasibility of this proposed minimal and non-canonical CCM in C. merolae. To facilitate the robust modeling of this process, we measured and incorporated physiological and enzymatic parameters into the model. Additionally, we trained a surrogate machine-learning model to emulate the mechanistic model and characterized the effects of model parameters on key outputs. This parameter exploration enabled us to identify model features that influenced whether the model met the experimentally derived criteria for functional carbon concentration and efficient energy usage. Such parameters included cytosolic pH, bicarbonate pumping cost and kinetics, cell radius, carboxylation velocity, number of thylakoid membranes, and CO 2 membrane permeability. Our exploration thus suggested that a non-canonical CCM could exist in C. merolae and illuminated the essential features generally necessary for CCMs to function.",
"conclusion": "Conclusions The extremophilic red microalga C. merolae operates a CCM, as evidenced by this alga having gas-exchange behavior which was not explained by its rubisco properties. Mathematical modeling suggested that this CCM could consist of a minimal mechanism. Robust parameter exploration and statistical analysis, aided by the use of a surrogate model, allowed us to quantify the sensitivity of our model to parameter uncertainties, identify important parameter interactions, and identify key determinants of CCM efficiency. Therefore, in addition to supporting the presence of a non-canonical CCM in C. merolae , our results shed light on what conditions must be met for this CCM to function and the essential elements of biophysical CCMs in general.",
"introduction": "Introduction \n Cyanidioschyzon merolae is a red microalga found in moist environments surrounding geothermal sulfur springs. This species is extremophilic, with optimal laboratory growth conditions including low pH (∼2) and high temperatures (∼42°C) ( Miyagishima and Wei 2017 ; Miyagishima et al. 2017 ). C. merolae and other thermoacidophilic red algae draw interest for their unique biology and simple characteristics, which position them as useful model organisms and as candidates for biotechnology applications ( Rahman et al. 2017 ; Miyagishima and Tanaka 2021 ; Seger et al. 2023 ; Villegas-Valencia et al. 2023 ). For example, C. merolae is of interest because it is one of few organisms which relies on photosynthesis in geothermal spring environments, where hot and acidic conditions restrict the availability of inorganic carbon and challenge biological carbon fixation ( Gross 2000 ; Miyagishima et al. 2017 ). Notably, organisms of acid waters can only access approximately 10 μ M inorganic carbon, as the inorganic carbon pool at acid pH is primarily the volatile species CO 2. In comparison, organisms of near-neutral and alkaline waters may have access to several millimolar of inorganic carbon, due to accumulation of the involatile bicarbonate ( Oesterhelt et al. 2007 ). \n C. merolae is thought to survive in its challenging environment in part by operating a carbon-concentrating mechanism (CCM) ( Zenvirth et al. 1985 ; Rademacher et al. 2017 ; Steensma et al. 2023 ). CCMs boost carbon-fixation efficiency by concentrating CO 2 around rubisco, providing ample substrate for carbon-fixation and inhibiting a competing oxygen-fixation reaction of rubisco. Evidence supporting a CCM in C. merolae includes measured accumulation of radiolabeled carbon in the cell, d 13 C consistent with a CCM, transcriptional response of potential CCM genes to CO 2 fluctuations, and substantial CO 2 assimilation at low environmental CO 2 concentrations ( Zenvirth et al. 1985 ; Rademacher et al. 2017 ; Steensma et al. 2023 ). However, many of these indications of the CCM are not definitive: in particular, it is not known how much of C. merolae 's ability to assimilate CO 2 efficiently could be explained by the affinity of C. merolae rubisco for CO 2 . Thus, we here provide further evidence for the CCM in C. merolae by demonstrating that the affinity of C. merolae cells for CO 2 is better than could be explained by the affinity of C. merolae rubisco for CO 2 . \n C. merolae 's CCM may be described as a “non-canonical” CCM, since the C. merolae CCM must operate differently from the few CCM types which are well-characterized. For example, unlike algae and cyanobacteria with well-characterized CCMs, C. merolae is not able to take up external bicarbonate, and C. merolae lacks anatomy associated with the pyrenoid CCM organelle ( Zenvirth et al. 1985 ; Badger et al. 1998 ; Misumi et al. 2005 ; Steensma et al. 2023 ). The absence of these CCM features in C. merolae challenges our understanding of what components are required for a functional CCM, and presents the opportunity to define essential CCM components. While previous work has discussed CO 2 as a source of carbon for the CCM ( Fridlyand et al. 1996 ; Price 2011 ), there has been little quantitative exploration of whether a CCM could function while lacking both facilitated carbon uptake and specialized compartments such as the pyrenoid or carboxysome. We thus used mathematical modeling, informed by our experimental measurements, to explore how the C. merolae CCM may function. Research on CCMs has long employed mathematical models to understand the components of functional CCMs in model cyanobacteria and algae, with a particular area of interest in CCM modeling being the possibility of boosting crop productivity by engineering CCMs into crops which lack CCMs ( Price et al. 2013 ; McGrath and Long 2014 ; Fei et al. 2022 ; Kaste et al. 2024 ). By developing modeling approaches to robustly describe CCMs in organisms where biochemical data is limited, such as extremophile algae, we can better understand how organisms survive environmental challenges. Here we add to these engineering efforts by modeling a heat-tolerant CCM with minimal components which offers unique possibilities for plant synthetic biology ( Misumi et al. 2017 ). To draw robust conclusions about cellular characteristics which can support a CCM, we used state-of-the-art statistical methods to define the effects of model parameters on the predicted photosynthetic phenotype while limiting unwarranted a priori assumptions. We demonstrate an interdisciplinary modeling approach which efficiently sampled from large parameter spaces and identified features (e.g. compartment permeability, pH, enzyme characteristics) that determine the function and energy cost of a simple CCM. This adds a useful tool for compartmental photosynthetic modeling, and could facilitate effective use of models to inform experiments and rational engineering. Some sets of model input parameters produced model outputs which met empirically based criteria for functional carbon concentration and efficient energy usage, and we identified input parameters which have substantial impacts on the model outputs. Overall, our model of a hypothetical biophysical CCM which requires minimal enzymes and anatomical features ( Fig. 1 ) appears to represent a feasible CCM structure in C. merolae , which invites further research into the sources of environmental resilience in extremophile algae. Figure 1. Cross-section of model structure. This model describes fluxes (indicated by arrows and “V#” notation) and pools (indicated by molecular formulas) of a simplified dissolved inorganic carbon system (CO 2 , HCO 3 − ) and of oxygen (O 2 ). Fluxes are as defined in Supplementary Methods section “Model fluxes.” Molecule pools can be present in several well-mixed compartments: the bulk external medium surrounding the cell, an unstirred boundary layer of medium around the cell, the cytosol, or a central stromal space of the chloroplast. Circles mark enzymatically-catalyzed fluxes. Compartments are not drawn to scale. PR = photorespiratory CO 2 release, R L = respiration in the light. All fluxes are reversible and are assigned an arbitrary direction, except those fluxes which represent producing or consuming material.",
"discussion": "Results and discussion Rubisco kinetics demonstrated that C. merolae operates a CCM In previous work, we determine that if C. merolae has rubisco kinetics similar to other red algae, then this alga must operate a CCM to maintain its measured photosynthetic efficiency. Alternatively, its measured photosynthetic efficiency could be explained by unprecedented rubisco kinetics, meaning enzyme properties favoring carbon-fixation over oxygen-fixation to an unprecedented degree ( Steensma et al. 2023 ). Here we confirmed that C. merolae rubisco kinetics are similar to those of other red-type (Form 1D) rubiscos ( Read and Tabita 1994 ; Uemura et al. 1997 ; Whitney et al. 2001 ). C. merolae rubisco had a strong affinity for CO 2 (low K C ), a poor affinity for O 2 (high K O ), and a slow carboxylation rate (low kcat C ) ( Fig. 2 ). Consistent with other studies, kcat C and K C were higher when measured at increased temperature, while K O was lower. Although K O is in the denominator of rubisco specificity ( S c/o ) and S c/o decreases with increased temperature, in vitro K O is observed to decrease with increased assay temperature in some species ( Jordan and Ogren 1984 ; Uemura et al. 1997 ; Prins et al. 2016 ). Figure 2. Experimental data incorporated into the model. A, B) . Response of net assimilation in C. merolae to A) CO 2 availability and B) light availability. Points are mean ± SE ( n = 3), and parameters calculated from the data are indicated in the upper left corner of each plot as mean ± SE. Dashed lines indicate trend fits used to determine Michaelis–Menten constant of CO 2 fixation (K C ) and respiration in the light (R L ). The linear fit used to determine the CO 2 compensation point (Γ CO2 ) is not pictured but is described in Methods. C) Kinetic properties of C. merolae rubisco. Rubisco turnover rate for CO 2 fixation (kcat C ), Michaelis–Menten constant of CO 2 fixation (K C ), and Michaelis–Menten constant of O 2 fixation (K O ) were measured at 25 and 45°C. Data are mean ± SE, n = 4. These kinetics findings indicated C. merolae does operate a CCM, as C. merolae cells had higher affinity for CO 2 than C. merolae rubisco (8.71 ± 1.7 µ M cell K C vs. 24.9 ± 3.2 µ M rubisco K C at 45°C, P = 0.008 by two-sample t -test) ( Fig. 2 ). This result adds to the evidence of a CCM in C. merolae ( Zenvirth et al. 1985 ; Rademacher et al. 2017 ; Steensma et al. 2023 ). Quantitative modeling showed that a hypothesized CCM can explain C. merolae 's carbon-concentrating behavior To explore how the C. merolae CCM may operate, we constructed a functional model of a CCM ( Fig. 1 ). This model demonstrated that there were parameter sets consistent with the empirical literature that result in a functional CCM, despite the minimal model structure lacking structures like a pyrenoid or carboxysome ( Fig. 3 ). Cyanobacterial CCM models have also supported reduction to a simple model with only two compartments from the cell membrane inwards ( Mangan and Brenner 2014 ). Figure 3. Values of key model outputs. A) Parameter sets are organized into a 2-dimensional histogram according to their output values of Γ CO2 and ATP per CO 2 , with dashed lines indicating bounds for acceptable values of these outputs. The dataset for the figure was the 240,000 total parameter sets. However, 80 parameter sets (0.03% of the total) are not pictured in this panel, as they produced negative ATP per CO 2 values and could not be log-transformed. B) Percentages of parameter sets meeting various combinations of output criteria. Model output criteria and associated units are as defined in Methods: Definition of reasonable model output values. Our results provided quantitative support for a CCM taking inorganic carbon from the environment solely through CO 2 diffusion into the cell without specialized compartments, which we term a “non-canonical” CCM due to its differences in structure and function from CCMs that have been characterized in detail. C. merolae has a different structure and environment than the “canonical” CCMs of Chlamydomonas reinhardtii and of model cyanobacteria, which allowed us to explore a biology and a parameter space which are different from those in previous CCM models. Though there is speculation that extremophilic red algae may use a C 4 -like CCM, it has been previously proposed that acidophile algae may accumulate carbon by a “bicarbonate-trap” or “acid-loading” mechanism similar to our modeled CCM ( Gehl and Colman 1985 ; Fridlyand 1997 ; Gross 2000 ; Rademacher et al. 2016 ; Curien et al. 2021 ; Fei et al. 2022 ). Briefly, this mechanism would involve bicarbonate being concentrated for enzymatic action by bringing inorganic carbon speciation near equilibrium in near-neutral cellular compartments, since the predominant inorganic carbon species from pH ∼6 to ∼10 is the poorly-membrane-permeable bicarbonate. Various facilitated CO 2 uptake mechanisms exist in CCM-containing organisms, such as the NAD(P)H dehydrogenase-I complexes in cyanobacteria and the periplasmic carbonic anhydrase (CA) system in algae ( Fridlyand et al. 1996 ; Moroney et al. 2011 ; Price 2011 ). We here test a different model where inorganic carbon enters the cell solely by passive CO 2 diffusion into the cytosol, followed by the action of non-vectorial cytosolic carbonic anhydrase. In contrast to the well-studied cyanobacterial and algal systems, where growth under limiting CO 2 is supported by active bicarbonate uptake and the accumulation of cytosolic bicarbonate above equilibrium levels ( Price and Badger 1989 ; Price et al. 2004 ; Duanmu et al. 2009 ), our model functions as a CCM without taking any bicarbonate from the environment. Another unique feature of our model is the nature of the diffusion barrier surrounding rubisco. Cyanobacteria encapsulate rubisco in a proteinaceous shell called the carboxysome, which is thought to provide a diffusion barrier to CO 2 ( Price et al. 2008 ). The model alga C. reinhardtii aggregates rubisco into an organelle called the pyrenoid, which in wild-type cells is surrounded by a starch sheath that may serve as a diffusion barrier. In contrast to the well-studied system of C. reinhardtii , there has been comparatively less investigation into algae which lack starch sheaths or lack pyrenoids entirely ( Morita et al. 1999 ; Barrett et al. 2021 ). Thus, to broaden our knowledge of CCM anatomy, we modeled an arrangement where rubisco is diffuse within a series of concentric thylakoid membranes. This allowed us to further investigate whether membranes, which are thought to be highly permeable to CO 2 ( Gutknecht et al. 1977 ; Missner et al. 2008 ), could impact carbon concentration, and how carbon concentration could function without a carboxysome or pyrenoid. To investigate these and other features of interest, we used two strategies to deeply explore the model parameter space and ensure that our conclusions were robust. First, the model included our experimental data on gas-exchange and rubisco parameters central to photosynthetic efficiency ( Fig. 2 ). Second, we developed a method for thoroughly assessing the model's sensitivity to the value of model parameters of interest. Specifically, we were interested in 19 of the 43 model parameters which were biologically interesting in relation to the function of a hypothetical C. merolae CCM and which were not well-characterized physical constants ( Supplementary Table S1 ). We thus sampled input parameter sets with varying numbers for these parameters of interest. We sampled parameter sets through a Latin hypercube design ( McKay et al. 1979 ) which enhanced analysis accuracy by mitigating sampling bias, as it produced parameter sets distributed throughout the 19-dimensional parameter space of interest. Then, each input parameter set was used to parameterize the model and to generate a set of outputs for analysis. Some of the input parameter sets produced outputs consistent with a functional CCM with reasonable energy cost. Of particular interest were the parameter sets which met all the empirically based criteria for a realistic and functional CCM (criteria selection described in Methods Supplementary S1 Supporting Information). Of 240,000 parameter sets, 13,998 (6%) parameter sets fulfilled the two competing objectives of functional carbon concentration (corresponding to outputs of low Γ CO2 , high stromal CO 2 , and low v o /v c ) and efficient energy usage (corresponding to output of low ATP per CO 2 ) ( Fig. 2 ). The generated parameter sets allowed us to explore the tradeoffs associated with various features related to the CCM. For example, adding additional concentric thylakoids slightly improved carbon concentration by presenting barriers to CO 2 leakage out of the chloroplast, but incurred additional energy costs of carbon transport ( Fig. 4 , Supplementary Figs. S1 and S2 ). This is consistent with other modeling studies indicating that thylakoid membranes could affect inorganic carbon diffusion and with observations of pyrenoids surrounded by layers of thylakoids in hornworts ( Thoms et al. 2001 ; Fei et al. 2022 ; Robison et al. 2024 ). Figure 4. Effect of select input parameters on key model outputs. A, B) Effect of model input parameter membranes ( x -axis) on key model outputs. Distribution of parameter set outputs for each value of membranes is represented by a box plot overlaid on a violin plot. Shaded areas represent unacceptable values of outputs. A) Effect of membranes on model output Γ CO2 . B) Effect of membranes on model output ATP per CO 2 80 parameter sets (0.03% of total) are not pictured in this panel, as they produced negative ATP per CO 2 values and could not be log-transformed. C, D) Effect on key model outputs where carbonic anhydrases or bicarbonate transport activity at the chloroplast boundary are removed from the model. Distribution of parameter set outputs for each scenario is represented by a box plot overlaid on a violin plot. Shaded areas represent out-of-bounds values of outputs. The same sampling of input parameter sets was run through models representing each scenario. C) Γ CO2 in model scenarios where various model features removed, with an indication of how many parameter sets met output criteria in each scenario. D) ATP per CO 2 in model scenarios where carbonic anhydrases or bicarbonate transport activity at the chloroplast boundary are removed from the model. 6,991 parameter sets producing negative ATP per CO 2 values (0.6% of total) are not pictured in this panel. All panels: Model output criteria and associated units are as defined in Methods: Definition of reasonable model output values. 240,000 total parameter sets were graphed for each panel of the figure, except where otherwise noted. The plots pictured were created with geom_violin and geom_boxplot in the R ggplot2 library. geom_boxplot visualizes the median, two hinges marking the first and third quartiles of the data, and two whiskers extending to the largest and smallest values no farther than 1.5 interquartile ranges from the hinge. Machine-learning-based surrogate models identified the parameters that most influence CCM efficiency Like most mathematical models of photosynthetic systems, this model faced the challenge of drawing robust conclusions while using parameters which, although bounded by their relationship to physical processes, have substantial uncertainty ( Supplementary Table S1 ). To model a system with limited biochemical data while not constraining input parameters to a greater degree than was supported by the literature, it was important to assess uncertainties which seemed likely to have substantial and interdependent effects on the model. For example, the input parameter describing the permeability of a lipid bilayer to CO 2 ( Plip CO2 ) has reported values ranging over several orders of magnitude ( Supplementary Table S1 ). Furthermore, the effect of Plip CO2 in the model depended on the value of other parameters, such as the number of lipid bilayers which pose a barrier to carbon moving between the stroma and cytosol ( Membranes ). Various sensitivity analyses are available for ordinary differential equation (ODE) models, but Plip CO2 and similar parameters were unlikely to be satisfactorily explored by classical local sensitivity analyses, which involve tracking model outputs when individual parameters are varied by a set fraction of the parameter's original value. Therefore, to reveal which model conditions were necessary for the modeled CCM to function biologically, and to identify interesting directions for future investigation, we used statistical methods to identify impactful parameters and to identify which input spaces corresponded to target output ranges. These statistical methods involved training a surrogate machine-learning model on our CCM model inputs and outputs. Interpretations of this surrogate model identified which zones in the input parameter space contained the most combinations fulfilling output criteria ( Fig. 5 lower left ), quantified how much each input parameter affected the prediction of outputs by the surrogate model ( Fig. 5 upper right ), and visualized the response of model outputs to inputs ( Supplementary Figs. S4 to S7 ). Figure 5. Statistical investigation of parameters affecting model output. Upper right bar plots: Mean absolute SHapley Additive exPlanations (SHAP) plots showing which inputs most affect each output criterion based on the 240,000 samples of the surrogate model. Lower left density plots: Density plots of parameter sets meeting all output criteria, based on the 240,000 samples of the surrogate model and organized by selected pairwise input parameters (input parameters pictured are those input parameters with high SHAP values for all output criteria). Scale bars by each density plot indicate relative densities, with darker areas indicating areas where more parameter sets meeting criteria occur. (Scales of color vary for each plot.) Parameters and associated units are as defined in Supplementary Table S1 . Some input parameters had little impact on model outputs with the tested input ranges. For these parameters, values from across the input range were evenly represented in the parameter sets meeting all output criteria. The parameters with relatively little impact on outputs included values related to carbonic anhydrase concentration and kinetics ( [CA] , CAkcat , Km CO2 , and Km HCO3− for carbonic anhydrases), chloroplast pH, and values related to bicarbonate membrane permeability ( Plip HCO3− , Q10 PlipHCO3− , Fig. 5 , Supplementary Figs. S4 to S8 ). While it is possible that these aspects of the CCM may become impactful if varied beyond the tested range (e.g. if engineering efforts produce carbonic anhydrase concentrations falling outside the range of literature values we used), these parameters did not emerge as particularly impactful in our exploration. Due to how fast the interconversion of inorganic carbon species by carbonic anhydrase is, the enzyme is likely capable of keeping inorganic carbon species close to their equilibrium concentrations across the range of values we explored for its kinetics. Given this, it is unsurprising that model outputs varied little with respect to carbonic-anhydrase-related parameters, even though the complete absence of these enzymes was deleterious ( Fig. 4 ). Other parameters were more constraining in the model, indicating their importance in producing a functional CCM. For example, six parameters appeared to impact all four of the target model outputs in the mean absolute SHapley Additive exPlanations (SHAP) plots: V c , Vmax pump , Km pump , pH in the cytosol, PlipCO 2 , and Membranes ( Fig. 5 ). Sobolʹ analysis ( Sobol′ 2001 ) of the surrogate model produced similar results ( Supplementary Fig. S9 ). As might be expected in a model relying on a cytosolic bicarbonate trap followed by bicarbonate pumping, parameter sets that successfully and efficiently concentrated carbon tended to have cytosolic pH at or above the pH where bicarbonate predominates (cytosol pH above 6) and tended to have a lower ATP cost of pumping bicarbonate (low Pump cost ), as well as faster and higher-affinity bicarbonate pumps (high Vmax pump , low Km pump ) ( Fig. 5 ). Other features enriched in parameter sets meeting output criteria were a cell radius in the middle of the input range (moderate Radius cell ) and a lower CO 2 membrane permeability (low Plip CO2 , Fig. 5 , Supplementary Figs. S4 to S9 ). This suggested an important relationship between the volumes where metabolism occurs and the surface areas which present diffusion barriers between compartments. As the radius of the cell increases, CO 2 loss from R L may overcome the ability of the cell to acquire carbon through passive diffusion into the cell. Conversely, as the radius of the cell decreases, less absolute bicarbonate pumping would be necessary to achieve high rubisco saturation, especially when rubisco is slow (low V c ). In low-radius scenarios, “over-pumping” bicarbonate could reduce energy efficiency. \n In silico knockouts identified experimental targets for further characterization of the C. merolae CCM The modeling also suggested interesting directions for investigating enzymatic components of the CCM. Alternative models with CCM enzymes removed (carbonic anhydrases or bicarbonate pumping not functional) were less likely to meet the criterion of a Γ CO2 indicative of functional carbon concentration, but tended to have lower ATP per CO 2 cost than the model with all enzymes present ( Fig. 4 , Supplementary Figs. S1 and S2 ). The modeled CCM functioned without fine details of cellular structure that support photosynthesis in other organisms, such as rubisco aggregation into an area smaller than the stroma, carbonic anhydrases with restricted distributions and directions (i.e. lumenal and vectorial carbonic anhydrases), recapture of mitochondrially respired CO 2 , and perforations or interconnections in concentric thylakoids ( Nevo et al. 2007 ; Rademacher et al. 2017 ; Barrett et al. 2021 ). Our work thus expands on previous models with detailed chloroplast geometry ( Fei et al. 2022 ) by demonstrating that efficient carbon capture may occur in a simple case when rubisco and carbonic anhydrase are diffuse within a series of concentric thylakoid spheres. It may still be of interest to explore what chloroplast structures support photosynthesis in C. merolae and to investigate the biochemical and molecular basis for this non-canonical CCM. Further applications of surrogate modeling and uncertainty quantification More broadly, the statistical approach adopted in this paper represents an advance in metabolic and biochemical modeling. By training a surrogate model on the parameter space of mechanistic biological models, we can understand and account for high-dimensional uncertainty in model parameters. Metabolic modeling in general, especially complex metabolic modeling, has been highlighted as a particularly promising application of surrogate modeling, as metabolic modeling has biotechnological potential but is challenged by the complexity of metabolism and by the “trial and error” process which is often required to produce a working metabolic model ( Gherman et al. 2023 ). Surrogate modeling has found uses in dynamic flux balance analysis and process modeling for bioprocesses ( Mountraki et al. 2020 ; de Oliveira et al. 2021 ). Our work expands on these investigations by demonstrating what is to our knowledge the first application of surrogate modeling to ODE-based compartmental modeling of biological systems. Our methods may be particularly valuable for models that have poorly defined parameters or are extremely computationally expensive. For example, the implementation of surrogate modeling described here could alleviate current limitations in interpreting reaction-diffusion models and genome-scale metabolic models ( Gherman et al. 2023 ). Even for our relatively-simple model, the run time for 240,000 simulations was several hours and required the use of a computing cluster. In contrast, surrogate modeling could be run locally on a laptop computer and was able to generate 240,000 predictions for all four outputs of interest in less than 10 s, easily creating a large dataset for analysis and allowing for precise sensitivity estimation. We compared this with Sobolʹ sensitivity analysis ( Sobol′ 2001 ) performed with the original model with a sample size of n = 163,840, comparable to the number of parameter sets and outputs used to train the surrogate model. Despite the generation of these samples taking several hours of computation time, this approach yielded extremely imprecise and uninterpretable results, suggesting that substantially more computational investment would be necessary to achieve acceptably precise sensitivity estimates ( Supplementary Fig. S10 ). With normalized root-mean-square error (NRMSE) below 1.5% in our validation ( Supplementary Table S2 ), the computational gains associated with the surrogate modeling approach outweighed the near-negligible potential error introduced by an inexact surrogate. Important considerations in any surrogate modeling application include the sample size required to train the model and limitations of surrogate models for out-of-sample predictions. Surrogates should be used cautiously for out-of-sample predictions, particularly in high-dimensional settings where training data are limited ( Forrester et al. 2008 ). Regarding the sample size, early studies ( Chapman et al. 1994 ; Jones et al. 1998 ; Loeppky et al. 2009 ) suggested using around 10 d samples, where d is the input dimension, for building an accurate Gaussian Process (GP) surrogate model. GP surrogates are particularly effective for small datasets and provide uncertainty quantification, which is valuable for assessing the confidence of out-of-sample predictions ( Gramacy 2020 ). If the desired accuracy is not achieved, one can improve the model by increasing the sample size through adaptive strategies such as active learning ( MacKay 1992 ), which allows for more efficient use of additional data to further enhance accuracy. Recent studies have also provided guidance on determining the run size required for a GP surrogate to achieve a pre-specified level of out-of-sample prediction accuracy ( Harari et al. 2018 ). In scenarios where high extrapolation performance is critical, one may consider using physics-informed surrogates, which tend to be more reliable in out-of-sample contexts. These surrogate models incorporate physical laws into their training process and offer improved performance for out-of-sample predictions, especially when physical dynamics are a key feature of a model. Examples of physics-informed surrogates include a manifold-constrained GP surrogate that adheres to an underlying ODE system ( Yang et al. 2021 ) or Physics-Informed Neural Networks (PINNs) ( Raissi et al. 2019 ). Effective parameter exploration and analysis may generally be useful in confronting global challenges. Here, we used statistical sampling, surrogate modeling, and uncertainty quantification methods to investigate how a particular aquatic organism achieves the high photosynthetic efficiency that enables them collectively to be responsible for approximately half of global photosynthetic CO 2 consumption ( Field et al. 1998 ). Similar modeling techniques may be applied effectively to any system: for example, as part of engineering efforts for bioproduction, crop resilience, and other goals, it may be useful to in silico determine which features of a system are essential or inflexible throughout ranges of interest before devoting resources to in vivo experimentation."
} | 8,186 |
37023039 | PMC10079068 | pmc | 8,428 | {
"abstract": "Edge effects resulting from adjacent land uses are poorly understood in agroecosystems yet understanding above and belowground edge effects is crucial for maintaining ecosystem function. The aim of our study was to examine impacts of land management on aboveground and belowground edge effects, measured by changes in plant community, soil properties, and soil microbial communities across agroecosystem edges. We measured plant composition and biomass, soil properties (total carbon, total nitrogen, pH, nitrate, and ammonium), and soil fungal and bacterial community composition across perennial grassland-annual cropland edges. Edge effects due to land management were detected both aboveground and belowground. The plant community at the edge was distinct from the adjacent land uses, where annual, non-native, plant species were abundant. Soil total nitrogen and carbon significantly decreased across the edge ( P < 0.001), with the highest values in the perennial grasslands. Both bacterial and fungal communities were different across the edge with clear changes in fungal communities driven directly and indirectly by land management. A higher abundance of pathogens in the more heavily managed land uses (i.e. crop and edge) was detected. Changes in plant community composition, along with soil carbon and nitrogen also influenced the soil fungal community across these agroecosystems edges. Characterizing edge effects in agroecosystem, especially those associated with soil microbial communities, is an important first step in ensuring soil health and resilience in these managed landscapes.",
"conclusion": "5. Conclusions In our study, we saw differences across the edge aboveground and belowground; changes included plant community composition, soil total N and C, and soil microbial community composition. Aboveground, weedy species were most abundant at the edge and appeared to have a positive response to the edge, where conditions from the cropland and grassland made it ideal for those species [ 107 ]. Belowground, soil C and N were lowest in the cropland, but NO 3 was highest in the cropland and edges. Soil microbial community composition across the edge was different, and fungi had more apparent differences in community composition than bacteria. A more in-depth analysis on fungi, showed some genera were more abundant in the cropland, edge, or grassland. For a holistic understanding of agroecosystem impacts, future studies need to consider the interrelated effects of management on soil properties and plant communities as these factors are often driving changes in soil microbial communities [ 110 , 129 ]. Further knowledge of the interactions between the soil microbial community, soil properties, plants, and edges in the agroecosystem will help to develop more sustainable agricultural practices and build healthier more resilient agroecosystem. *Raw sequence fasta files and the associated metadata can be found at the National Center for Biotechnology Information (NCBI) under Bioproject PRJNA588061",
"introduction": "1. Introduction Habitat fragmentation is a leading cause of biodiversity loss [ 1 , 2 ] and agriculture has caused extensive habitat fragmentation [ 3 , 4 ]. Highly fragmented landscapes have a high proportion of edges, which affect various ecological aspects [ 5 ]. Edges can be high contrast such as a forest abutting a pasture, or a more gradual low contrast edge like a shrub patch adjacent a meadow. Edges have edge effects which are abiotic and biotic changes occurring at the bounds of an ecosystem or habitat patch [ 5 , 6 ] influencing properties including microclimate, moisture, soils, plant or animal community composition and distribution [ 7 – 9 ]. Some factors that influence edges are orientation [ 10 , 11 ], time [ 12 ], patch size [ 13 , 14 ], edge contrast [ 15 ] and matrix composition [ 16 , 17 ]. Ecological dynamics and patterns around edges can be understood through four essential mechanisms, ecological flows across edges, resource distribution, resource mapping, and unique species interactions [ 6 ]. Expansion and intensification of agriculture has induced change in nearby habitats, and have been observed in both plant communities and soil properties [ 18 – 20 ]. Agricultural intensification is thought to magnify edge effects [ 19 ] further altering vegetation and soil biodiversity in these systems [ 21 ]. Commonly, edges in the agroecosystem are inhabited by non-native undesirable plants, here called weeds, or other invasive species [ 22 ]. Plant communities at the edge may be of concern to farmers, where weeds can compete with crops [ 23 ]. While aboveground vegetation changes at the edge are evident, belowground changes are also occurring. Underlying gradients of soil properties have been found at edges, including soil pH, nitrogen (N), and carbon (C) [ 24 , 25 ], though these studies are limited to forest edges. Aboveground and belowground interactions are important to consider because those interactions determine ecosystem function, and in particular agroecosystems, where land management has effects beyond the field boundary. However, the extent and characteristics of edges and their effects in agroecosystems remain poorly understood belowground. Two major land uses in the agroecosystem are cultivated croplands and grasslands; they each have characteristics that affect the soil microbial community. Nutrient dynamics between the two are quite different; for instance, croplands often have lower soil C than grasslands [ 26 , 27 ] while grasslands have more soil C and are frequently correlated with higher microbial biomass [ 28 ]. Various environmental factors affect soil microbial community composition and function [ 29 ], but agricultural practices directly alter environmental conditions affecting soil microbes [ 30 ]. These agricultural practices include but are not limited to, soil amendments [ 31 , 32 ], tillage [ 33 , 34 ], herbicides [ 35 , 36 ], and crop type [ 37 ]. However, the magnitude to which these factors influence the soil microbial community are complex [ 38 – 40 ]; considering edge effects and the interactions with agricultural practices is essential to understand soil microbial community dynamics in these landscapes. Aboveground edge effects provide insight into belowground conditions and ultimately the soil microbial community. Plant species can have specific microbial associations affecting microbial community composition, such as mycorrhizal associations with plant roots [ 41 ]. Additionally, invasive plant species can alter the soil microbial community through changing inputs of litter quality and quantity [ 42 ]. Knowing how and what alters the soil microbial community is important, as soil microorganisms are critical in maintaining ecosystem function, especially through nutrient cycling, disease suppression, and plant growth promotion [ 43 , 44 ]. Understanding how the soil microbial community responds to edge effects is crucial, as the soil microbial community is essential for maintaining ecosystem function, especially with intensification of agricultural lands [ 44 ]. To investigate edge effects in agroecosystems above and belowground, we measured vegetation composition and biomass, and soil physicochemical and microbial properties across perennial grassland and annual cropland edges in central Saskatchewan, Canada. Our goal was to determine if changes in land use altered the plant community and soil properties at agricultural edges, and if so, how these changes influenced the microbial community across the edge. Considering the interrelated effects of management on soil properties and plant communities, and their impacts on soil microbial communities, will better our understanding of agroecosystem edges and their ecosystem function.",
"discussion": "4. Discussion We investigated soil properties, vegetation community, and the soil microbial community across edges of perennial grasslands and annual croplands. Land management had direct and indirect influences on the soil microbial community through changes in vegetation and soil properties. Edges acted as an intermediate and unique environment between the two land uses, composed of predominately non-native weedy plants and the edge was more similar to cropland than grassland in both plant and soil. 4.1. Aboveground changes across the edge Differences in plant community composition and biomass across the edge was largely determined by land use type. Three different vegetation communities were observed: the perennial grassland, the edge (~1 m in width), and the cropland. Unsurprisingly, cropland vegetation was strongly influenced by the crop seeded; B . napus at CLC and L . usitatissimum at SDNWA. Living biomass was greatest in grasslands, which were dominated by brome species ( B . inermis and B . biebersteinii ) that were seeded in previous years. Both brome species were primary contributors to biomass, as grass constituted 88% of total living biomass. Plant community composition at the edge was a mixture of grassland plants, crops, and weedy species. Weed population densities are highest near, or at, an edge [ 78 ] because these plants are disturbance tolerant [ 79 ]. Non-native plant presence in agriculture frequently increases plant species richness in these settings and is driven by agronomic activities [ 80 , 81 ]. Agronomic activities including general mechanical disturbance such as mowing, crop sowing, and harvesting disturb the edge [ 82 ]. While our study sites were no-till systems, croplands still experienced a higher level of disturbance than grasslands throughout the growing season. In-field herbicide and fertilizer application can have unintended effects on adjacent areas [ 83 ]. Herbicide and fertilizer drift can reach beyond cropland edges and affect the plant community [ 84 , 85 ]; for example, fertilizer drift can promote faster growing competitive plant species that outcompete others [ 84 , 86 , 87 ]. In addition to higher nutrient availability, cropland edges have open space allowing undesirable weedy species to establish [ 82 , 88 ]. These edge effects lend advantages to these plant species that may compete with crops, reducing yields [ 89 ] and facilitate invasion of undesirable plants into adjacent, more natural, land use types [ 90 ]. Management practices, such as using herbicides or doubling sown crop density are effective in reducing weed populations at edges [ 91 ]. However, conventional eradication attempts may bring more detriments to larger agroecosystem, herbicide can drift into non-target areas and weedy species can become herbicide resistant [ 92 ]. Field edges can act reservoir for invasive weeds and other undesirable microbial pathogens [ 93 ]. However, the reverse is also true, a diverse weed community can provide ecosystem services and habitat to beneficial species [ 82 , 94 , 95 ]. Multiple management strategies are needed to successfully manage edge habitats valuable to many aspects of the agroecosystem. 4.2. Belowground changes across the edge Land management practices indirectly influenced soil physiochemical properties across perennial grassland-cropland edges through modification of aboveground plant community, and directly through fertilizer application. We found total C and N were highest in the perennial grasslands and lowest in the cropland; this is common in agroecosystems as soil quality is often poorer in cultivated land compared to non-cultivated land [ 96 – 98 ]. At our sites, perennial grasslands had plant species with relatively high-quality litter that likely influenced soil properties through the deposition of rich C sources. For example, at our sites in the perennial grasslands, B . inermis and M . stavia produce large amounts of litter that quickly degrades and is high in N content with a low C:N, which can increase soil organic C and rates of soil N cycling [ 99 – 101 ]. In addition, while the cropland is relatively productive, the majority of aboveground biomass is removed, not allowing the plant based C to return to the soil, which is a major source of soil C [ 102 ]. Edges are subjected to fertilizer applied to the cropland, evidenced by high spikes of NO 3 in both in cropland and edges. Inorganic N amendments, applied over both long and short time periods, can increase soil total N and NO 3 [ 103 – 105 ]. Nitrate concentrations in edge soils were more similar to croplands, likely due to the close proximity of the edge to the cropland and inputs from surface runoff [ 106 ]. However, our observation was only at one time point and may not provide a complete picture of N dynamics and seasonal fluctuations of NO 3 in this system. Regardless, edges in agroecosystems appear to act as a buffer for nutrient movement from managed croplands into adjacent land use types. 4.3. Soil microbial community across the edge In our study, land management appeared to have a strong influence on soil microbial community composition, as the direct pathways from land management to microbial communities were mostly significant in the SEMs. We chose to focus on community composition rather than a metric like richness, because in cases where richness is not affected, composition can detect more discreet changes [ 2 , 107 ]. Management practices can directly and indirectly affect soil microbial communities [ 108 – 110 ] and long term practices have selective forces on the soil microbial community, thus changing the microbial community composition as it adapts to these disturbances [ 111 ]. Fungal community composition was different in the grassland than cropland, as denoted by a ‘negative’ impact by the perennial grassland and a ‘positive’ by the cropland and edge. Fungal community composition was also different between the edge and cropland, though not as pronounced. Bacterial community composition was also different in the perennial grasslands compared to edge or the cropland, however patterns of response across the land uses were not as clear for bacteria as fungi ( Fig 3 ). Bacterial communities may respond less than fungal communities to changes in land use and vegetation, similar patterns were found in no-till cropland and native prairie in Kansas [ 26 ] and in comparing native and exotic grasslands [ 112 ]. Direct relationships between land management and the microbial community is likely driven by underlying changes of soil and plants associated with land use types. Plants are an important factor affecting microbial communities, especially at our study sites, land management created three distinct plant communities across the edge. Plant species can influence soil microbes through symbiotic relationships, root exudates, and plant litter inputs [ 112 ]. A key difference in plant community across the edge was the dominance of annual plants in the cropland and edge, while the grassland was composed of nearly all perennial plants. Brassica species, like the B . napus planted at CLC are non-mycorrhizal plants, which would greatly affect both the quantity and quality of AMF hyphae and spores observed [ 113 , 114 ], thus could be an aspect shaping fungal community composition. The distinction between annual and perennial plants is important as McKenna et al., (2020) [ 113 ] found that soil fungal community composition was similar under two different perennial vegetation types a seeded monoculture of intermediate wheatgrass ( Thinopyrum intermedium (Host) Barkworth & D.R. Dewey) grassland and a native prairie. However, both perennial fungal communities were different than the fungal community under annual crop rotation. Root architecture and activity may be largely responsible for differences between annual and perennial plants, as perennial grasslands have greater root biomass and more evenly distributed and deeper roots than annual croplands [ 26 ]. Annual plants dominated the cropland and edges, which had similar direct effects on fungal community ( Fig 4A ), suggesting that the life history strategies of dominant plants influence the fungal community. Although we did not observe significant pathways from soil nutrients to fungi or bacteria, we did observe a strong influence of land use on soil nutrients. The perennial grassland had more total N likely due to more biomass, but high NO 3 and NH 4 were observed in the cropland. Fertilizers containing N can reduce fungal diversity and fungal richness, possibly related to NO 3 [ 32 ] . However, others have found no effect of N fertilizers on fungal diversity or richness [ 114 , 115 ], but differences in fungi community composition [ 31 ]. Increased N availability, specifically NO 3 , may be disrupting natural plant-soil feedback relationships [ 31 , 116 ]. By increasing the N available to soil fungi or interrupting available C exudates via N available to plants, NO 3 can alter community composition by promoting or suppressing fungi with different life history strategies based on altered soil conditions [ 104 , 117 ]. Higher NO 3 levels in the cropland and edge may have been an important driver of microbial community composition, specifically fungi at our study sites. One aspect not considered directly in this analysis, was the soil C to N ratio. The C:N is crucial for microbial functioning [ 118 , 119 ] and linked to soil microbial community composition [ 120 – 122 ]. Considering the soil C:N explicitly in the future would aid in understanding soil microbial community composition across the edge. Examining abundant fungal genera revealed further insight into the effect of land management on the fungal community. Plants and soil fungi often develop a stable environment together as their interactions can provide mutual benefits, such as aid in nutrient acquisition for plants and carbon sources for fungi through plant exudates [ 118 ]. Different plant species can affect soil fungi differently, likely due to unique soil microbiomes associated with each plant species [ 119 ]. For example, plant species with litter high in C:N can promote Basidomycota fungi to aid in decomposition, changing fungal community composition [ 120 , 121 ]. Fungal genera Gibberella and Paraphoma were significantly more abundant at the edge and likely reflect the presence of both crop species and grasses. Many Gibberella species are plant pathogens that can cause significant crop diseases, such as head blights in grain crops and ear rot in corn ( Zea mays L.) [ 122 ]. Paraphoma are common soil fungi and frequently associate with monocots [ 123 ]. Furthermore, at the edge we found P . chrysanthemicola , a plant pathogen [ 124 , 125 ] known to affect plants in the Asteraceae and Rosaceae families [ 126 ] which were found at the edge. Significant fungal genera abundant in the cropland were mostly pathogenic, including Sarocladium [ 127 ] and Parastagonospora ; P . nodorum , a major wheat pathogen, which was identified to the species level [ 128 ]. Others have hypothesized that edges can act as a reservoir for undesirable microbial pathogens [ 93 ]. In our study the difference between fungal communities in cropland and edges, compared to perennial grasslands, was driven by the abundance of pathogens in these more heavily managed land uses supporting this hypothesis."
} | 4,821 |
39906146 | PMC11793158 | pmc | 8,429 | {
"abstract": "Marine heatwaves (MHW) are intensifying, posing a grave threat to coral reefs. We exposed\n Lutjanus carponotatus to MHW conditions (+1°C and + 2°C) for 4 weeks\nand found increased oxygen consumption and recovery time, among other physiological\nchanges. Interestingly, several effects persisted for at least 2 weeks post-MHW.",
"introduction": "Introduction The continued anthropogenic release of greenhouse gases is raising the Earths’ average\ntemperature, along with increasing the frequency and intensity of extreme climatic events\n( Hoegh-Guldberg et al ., 2018 ;\n Meehl et al ., 2000 ; IPCC 2022 ). The oceans are estimated to have taken up\nmore than 90% of the excess heat in the climate system, leading to unabated ocean warming\nsince the 1970s ( Hoegh-Guldberg et al .,\n2018 ). Both the increasing average temperature of the ocean, and anomalous warming\nevents called marine heatwaves (MHWs), are of concern to marine ecosystems and the\nbiodiversity they support. MHWs are caused by a combination of atmospheric and oceanographic\nprocesses with common drivers including persistent high-pressure systems, ocean currents\nthat create a build-up of warm water and air-sea heat flux that transfers atmospheric heat\ninto the sea surface ( Hobday et al .,\n2016 ; Holbrook et al .,\n2019 ; Gupta et al. ,\n2020 ). Specifically, MHWs are defined as an anonymously warm event, exceeding the\n90 th percentile of a 30-year average that lasts for at least 5 days ( Hobday et al ., 2016 ). Over the last\ncentury MHWs have increased in both frequency (34%) and duration (17%) resulting in a 54%\nincrease in the number of MHW days ( Oliver\n et al ., 2018 ). The frequency, intensity and duration of MHWs are\nexpected to further increase as anthropogenic climate change continues ( Frölicher et al ., 2018 ). Consequently,\nMHWs are considered a more imminent threat to marine organisms than the gradual average\nincrease of sea surface temperature ( Oliver, 2019 ;\n Smale et al ., 2019 ; Guo et al ., 2022 ). Most marine organisms are ectotherms; therefore, an increase in water temperature can\nresult in thermal stress when it exceeds their thermal optima ( Somero, 1995 ; Mora & Maya,\n2006 ; Pinsky et al .,\n2019 ). Recent MHWs have caused significant mortalities in invertebrates ( Garrabou et al ., 2009 ), loss of\nseagrass meadows ( Marbà & Duarte, 2010 ), mass\nbleaching and mortality in corals ( Hughes\n et al ., 2017 ) and are predicted to reduce the biomass of some\nfish populations ( Cheung & Frölicher, 2020 ).\nWhile many marine ecosystems are affected by MHWs ( Wernberg et al ., 2013 ; Cavole\n et al ., 2016 ; Smale\n et al ., 2019 ), the impacts have been most acutely observed on\ncoral reefs where mass coral bleaching and mortality due to anomalous temperatures have\noccurred with increasing frequency, magnitude and geographical extent over the past 30 years\n( Heron et al ., 2016 ; Donner et al ., 2017 ; Le Nohaïc et al ., 2017 ; Hughes et al ., 2018 ; Dietzel et al ., 2020 ). For example,\nmass coral bleaching events occurred on the Great Barrier Reef in the summers of 2016 and\n2017 where coral mortality was estimated to exceed 50% when averaged over the entire reef\n( Hughes et al ., 2018 ; Stuart-Smith et al ., 2018 ). However,\ndocumented effects of MHWs on other coral reef organisms are more restricted ( Bernal et al ., 2020 ). Tropical marine\nspecies are predicted to be more sensitive to extreme temperatures than higher latitude\nspecies because they have evolved in more thermally stable environments ( Tewksbury et al ., 2008 ; Sunday et al ., 2012 ; Comte & Olden, 2017 ) and are often living close to\ntheir thermal optimum in summer ( Rummer\n et al ., 2014 , Rodgers\n et al ., 2018 ). Yet, the direct effects of MHWs on most coral\nreef organisms remain unknown (e.g. exceptions; Brown\n et al ., 2021 , Haguenauer\n et al ., 2021 ; Tran &\nJohansen, 2023 ). Increased water temperature can have broad physiological effects on fish from direct\nthermodynamic effects on biochemical reaction rates through to changes in whole organism\ntraits such as swimming performance ( Clarke &\nJohnston, 1999 ; Farrell et al .,\n2009 ; Little et al .,\n2020 ). Increasing rates of cellular processes in warmer water results in rising basal\nmetabolic rates ( Clarke & Johnston, 1999 ; Gillooly et al ., 2001 ) and elevated\nenergetic costs for physical activities ( Johansen &\nJones, 2011 ; Hein & Keirsted, 2012 ),\nresulting in increased recovery time ( Lee\n et al ., 2003 ; Yanase &\nArimoto, 2009 ) and energetic cost post-exercise ( Lee et al ., 2003 ; Zeng\n et al ., 2010 ). Furthermore, the additional energy requirement\nfor physical activities may not always be possible through aerobic processes. For instance,\nthe mitochondria, which have a key role in all ATP production, can decrease in efficiency\ndue to enzyme thermal sensitivity, compromising their ability to meet ATP demands when\noptimal temperatures are surpassed ( Brand & Nicholls,\n2011 ). When oxidative metabolism is insufficient, individuals can increase\nanaerobic metabolism to meet energetic demands ( Jacobs,\n1986 ; Omlin & Weber, 2010 ; Iftikar et al ., 2014 ) resulting in\ncostly accumulation of byproducts like blood lactate ( Jain\n& Farrell, 2003 ; Zakhartsev\n et al ., 2004 ). Water temperature above the thermal optimum can\nalso reduce aerobic capacity ( Nilsson\n et al ., 2009 ; Rummer\n et al ., 2014 ; McMahon\n et al ., 2020 ) through an inability of the cardio-vascular system\nto keep pace with maximum oxygen demands ( Farrell\n et al ., 2009 ; Pörtner\n et al ., 2017 ). While research to date provides some understanding of the likely impacts of elevated\ntemperature on tropical marine fish ( Donelson\n et al ., 2010 ; Johansen &\nJones, 2011 ; Rummer et al .,\n2014 ), much of the work has been designed to explore the effects of longer-term\nincrease in average water temperature (i.e. exposure of mid-century to end of century\nprojections for months to generations) rather than shorter duration warming events and\npotential recovery trajectories afterwards ( Hollowed\n et al ., 2013 ; Lefevre,\n2016 ). However, the rate of change and intensity of MHWs present an acute stress\ncompared to longer warming experiments. Research to date has shown that MHWs in tropical and\nsub-tropical coral reefs have an average duration of 5–10 and 10–15 days, respectively;\nhowever, these durations are expected increase over the current century ( Oliver et al ., 2018 ). This presents a\npotential issue as the plasticity of a species to cope with longer, slow changes in\ntemperature may indicate capacity to cope with acute MHW stress. Additionally, current\nresearch has shown that coral reef species may not seek thermal refuge during MHWs as\npreviously hypothesized ( Haguenauer\n et al ., 2021 ). Recent work has also shown that species may be more\nsensitive to MHWs over a larger period of the year than previously expected ( Tran & Johansen, 2023 ) which could present\nunexpected physiological challenges as MHWs increase in duration. The physiological recovery\nof individuals after MHWs is also an historically overlooked area ( Grimmelpont et al ., 2023 ); therefore, research\nencompassing MHWs and post-MHW recovery may provide us with further insight into how marine\nspecies will cope with future challenges. To date, most research on the effects of elevated water temperature on coral reef fishes\nhas focused on smaller bodied, site-attached species, yet the impacts to larger reef fishes\nmay not be easily extrapolated from this. Research into the effects of MHWs on larger\npredatory fishes, which play an important role in ecosystem function ( Ritchie and Johnson, 2009 ; Hixon,\n2015 ; Hempson et al .,\n2017 , 2018 ), is particularly lacking. Our\ncurrent knowledge of thermal sensitivity of larger bodied predatory coral reef fish is\nrestricted to a focus on coral trout ( Plectropomus leopardus ), which\nexhibit high sensitivity to a temperature increases between +1.5 and 4.5°C that results in\ndecreased activity, aerobic scope and survivorship ( Johansen et al ., 2014 ; Johansen\n et al ., 2015 ; Messmer\n et al ., 2017 ; Pratchett\n et al ., 2017 ). However, this research used temperatures of 30°C\nand 33°C for a period of 4–6 weeks and consequently may overestimate the impacts of current\nand imminent MHWs which are on average shorter (5–15 days) in in the species distribution\n( Oliver et al ., 2018 ). We do not\nyet understand how common MHW conditions may impact reef fish nor how they will recover from\nthese events. Maximal physiological performance is beneficial for fisheries and bycatch species as the\naerobic stress of capture can lead to post-capture mortality, which in recreational catch\nand release fishing is estimated to be 3–30% depending on the species ( Diggles & Ernst, 1997 ; Frisch\n& Anderson, 2000 ; McLeay\n et al ., 2002 ; Sumpton\n et al ., 2010 ). Survival post-capture is important for a number\nof coral reef fisheries species (e.g. grouper, snapper, trout.) as they are more valuable\nwhen sold live (Sadovy et al. 2013). Additionally, while the majority of\nline bycatch is thrown back alive, the sudden and intense stress of capture can have\nsignificant physiological impacts and consequently impact survival post-release ( Cooke et al ., 2014 ; Wilson et al ., 2014 ; Raby et al ., 2018 ). For example, in a\nsingle species fishery, such as coral trout, there are a range of important mesopredator\nspecies (e.g. lethrinids, haemulids, lutjanids and serranids) that are caught as bycatch and\nthrown back ( Walton et al .,\n2021 ). Interestingly, wild populations have also been found to be more susceptible to\nfishing efforts during MHWs ( Brown\n et al ., 2021 ) which could have unforeseen consequences on the\nhealth and management of these populations. Further research is essential to understand the\nimpacts of MHWs to larger bodied reef fish including the potential for capture stress\nassociated with fishing during MHWs exacerbating the effects. Tropical snappers from the family Lutjanidae are among the most abundant mesopredators on\ncoral reefs ( Newman et al ., 1996 ).\nThey play an important role in ecosystem function ( Ritchie and Johnson, 2009 ; Hixon, 2015 ;\n Hempson et al ., 2018 ) and are\ncomponents of both commercial and recreation fishing catches ( GBRMPA, 2014 ). To test the physiological effects of MHWs on a coral\nreef mesopredatory fish, we subjected adult Lutjanus carponotatus (Spanish\nflag snapper) to two different magnitudes of simulated heatwave conditions, +1°C (29.5°C)\nand + 2°C (30.5°C) above summer average, for a total of 4 weeks. At 2 weeks of exposure, 4\nweeks of exposure and 2 weeks of post-exposure, we measured resting oxygen consumption,\ncapture oxygen consumption, recovery time and associated blood chemistry responses (lactate,\nglucose, haematocrit and haemoglobin). In a subset of fish not exposed to the MHW\ntreatments, we explored the thermal preference temperature and avoidance temperature to\ndetermine how their behavioural thermal optimum range relates to the simulated MHWs\ntemperatures. This experimental design allowed us to measure the physiological effects of\nMHW conditions on adult Spanish flag snapper, both during and following the period of\nelevated temperature. If MHW conditions are pushing the species beyond their thermal\noptimum, we would expect to see signs of increased reliance on anaerobic pathways (e.g.\nchanges in oxygen consumption, recovery times, post exercise oxygen consumption etc.) and\nbiochemical markers (e.g. increased lactate). The repeated measured design allowed\ndetermination of whether the magnitude and duration of the warming event had substantive\neffects on the physiological impacts, as well as whether recovery was possible within 2weeks\nfollowing a MHW or if there was evidence for lag effects.",
"discussion": "Discussion The increased frequency and intensity of MHWs pose a significant threat to marine\norganisms, especially those adapted to stable thermal environments like coral reefs. Our\nstudy revealed distinct physiological responses in adult L. carponotatus \nunder simulated MHW conditions, at modest temperature increases of +1 to +2°C above the\nsummer average. During MHWs, fish exhibited higher metabolic activity and signs of elevated\nstress, yet mostly recovered within 2 weeks post-exposure. The heightened basal cellular\ncosts (resting MO 2 ), prolonged recovery time after capture, and elevated blood\nlactate indicate negative physiological impacts of MHW conditions. Depending on MHW duration\nin nature, this could potentially reduce body condition and hinder the ability to escape\npredators or find prey ( Killen et al .,\n2015 ; von Biela et al .,\n2019 ). Interestingly, haemoglobin and haematocrit remained elevated at 2 weeks\npost-exposure, and capture MO 2 significantly increased, suggesting a shift in\nrelative aerobic and anaerobic energy production, corroborated by post-capture lactate\nlevels. Thermal stress manifested in multiple physiological measures during MHW exposure. Resting\nMO 2 , recovery time, EPOC and baseline blood lactate levels all significantly\nincreased in both MHW treatments over the 4-week exposure period but showed recovery within\n2-weeks after returning to control temperature (i.e. post-exposure). Resting MO 2 \nand basic metabolic costs generally rise by ~2–14% (Q10: 1.48–3.71) with every degree of\nwarming during summer for tropical coral reef fish ( Nilsson et al ., 2009 ; Johansen\nand Jones, 2011 ; Rummer et al .,\n2014 ; Messmer et al .,\n2017 ; Pratchett et al .,\n2017 ). In L. carponotatus , resting MO 2 increased by\n10–20% per degree Celsius (Q10: 2.86–7.93), indicating a high degree of thermal sensitivity\ncompared to other coral reef species, which is similar to the larger-bodied mesopredator,\ncoral trout (10–14% increase per degree celsius) ( Messmer\n et al ., 2017 ; Pratchett\n et al ., 2017 ). Additionally, we observed prolonged recovery time\nfor fish to reach resting MO 2 after simulated capture stress, with individuals\ntaking ~2 hours longer in both +1°C and + 2°C MHW conditions. This extended period not only\nincurs elevated metabolic rate costs but also exposes them to increased predation risk due\nto reduced aerobic escape capacity ( Killen\n et al ., 2015 ). Fish from both MHW treatments also exhibited\nhigher baseline lactate levels (27–48% higher than controls), indicating a greater reliance\non anaerobic glycolysis for energy production. This is perhaps due to thermal effects on\nmitochondrial efficiency or simply an inability to meet required energy aerobically as\nresting MO 2 also increased ( Jacobs,\n1986 ; Omlin and Weber, 2010 ; Iftikar et al ., 2014 ). Furthermore,\nthere was a trend of accumulating blood lactate, and potentially stress, over time in\nMHW-exposed fish, as indicated by a general increase (15–20%) in lactate between the 2- and\n4-week exposure periods. Nevertheless, these fish were capable of processing lactate once\ntemperatures returned to normal. During exercise, blood lactate levels can rapidly increase to supplement energy demand for\nswimming ( Jones, 1982 ; Weber et al ., 2016 ). Under MHW conditions, blood\nlactate levels increased significantly (67–109%) due to simulated capture stress compared to\nthe baseline increase (27%) at the same MHW exposure time point. Since capture\nMO 2 remained consistent across all treatments during MHW exposure, it is likely\nthat individuals in the elevated MHW treatments lacked sufficient aerobic capacity, leading\nto increased anaerobic energy production as compensation ( Drucker and Jensen, 1996 ; Svendsen\n et al ., 2010 ). The additional lactate may have contributed to\nthe observed longer recovery times as aerobic metabolism remained elevated to oxidize\nlactate from the blood ( Wells et al .,\n2009 ; Ohlendieck, 2010 ). Interestingly,\nwhen comparing baseline and post-capture lactate, distinct patterns emerged over time.\nBaseline lactate tended to increase at the 2-week exposure mark and remained high at\n4-weeks, whereas post-capture lactate decreased between the 2- and 4-week exposure periods.\nThis suggests that no physiological mechanisms were induced to offset the fundamental\ncellular costs of functioning under elevated conditions. However, it is plausible that other\nunmeasured physiological mechanisms were initiated, resulting in a reduction of lactate\nfollowing the capture event (e.g. upregulation of lactate dehydrogenase; Larios-Soriano et al ., 2020 ). Haemoglobin and haematocrit levels were elevated during MHW treatments, with residual\neffects observed post-exposure. Increased production of red blood cells (RBCs), indicated by\nhigher haematocrit, may have been prompted to meet the elevated respiration and energy\ndemands caused by higher temperatures ( Gillooly and\nZenil-Ferguson, 2014 ). Additional RBCs would enhance oxygen transport capacity and\ngill diffusion efficiency in lower dissolved oxygen concentrations at higher temperatures\n( Wells and Baldwin, 1990 ; Gallaugher and Farrell, 1998 ). We might expect a corresponding shift in\naerobic performance, including increased resting and capture respiration. Resting\nMO 2 followed expectations, showing increased demand during MHW, however,\nMO 2 during simulated capture stress did not increase, despite the presence of\nadditional RBCs. While capture swimming costs may not have risen in MHW treatments (i.e. if\nfish were within their thermal optimal range), higher lactate levels and extended recovery\ntime indicate increased energy demand for swimming in MHWs. Therefore, swimming physiology\nfactors, rather than oxygen delivery, likely limited capture MO 2 ( Pörtner, 2010 ; Pörtner et al ., 2017 ). For example, capture MO 2 may\nrepresent the maximum capacity of aerobic swimming ( Schulte, 2015 ), which remained consistent across the three treatments during MHW.\nAlternatively, our measurements may represent the limit of aerobically generated energy\n(i.e. Krebs cycle: Krebs, 1950 ). While the specific\nphysiological mechanism remains unclear in this study, it suggests the existence of an\naerobic capacity limit is unaffected by MHW treatment. While the additional RBCs did not alter capture MO 2 during MHW conditions, they\nmay have had an effect once temperature returned to normal in the 2 weeks following. Capture\nMO 2 was 25% higher than control, and higher than these fish during the MHW, in\nboth +1°C and + 2°C MHW treatments at 2-week post-exposure while haemoglobin and haematocrit\nremained elevated. The lifespan of a RBC is thought to be ~ 60–120 days ( Franco, 2012 ; Shrestha et al ., 2016 ) and if the initial temperature increase\ninduced the production of additional RBCs, they would not be destroyed or discarded if\nhealthy. Thus, the elevated proportion of RBCs (haematocrit) post-exposure are likely to be\na legacy of MHW exposure rather than an active response. Interestingly, alongside this\nelevated aerobic response, higher lactate levels and recovery time returned to control\nlevels in MHW fish post-exposure. This suggests that the relative production of aerobic to\nanaerobic energy during swimming capture has shifted, and perhaps the increase in RBC is\nassisting the increase of aerobic energy production ( Mairbäurl, 2013 ; Saunders\n et al ., 2013 ). This point is strengthened by the reduced EPOC\nfound at 2 weeks post-exposure as while there is higher aerobic demand seen during the\ncapture event there is not a significantly higher debt post exercise, which may indicate\nreduced reliance on anaerobic energy. Although we observed some persistence of physiological effects post-MHW exposure, the fish\ngenerally showed limited stress and costs 2 weeks post-MHWs. The only physiological\nattribute that suggested a continuing cost was resting MO 2 (~9% higher), however,\nthis was not statistically significant. Baseline lactate and recovery time, while\nsignificantly higher during the MHW exposure phase, both returned to control levels within 2\nweeks of fish returning to control conditions. There may have been other energetic or\nphysiological costs associated with MHWs that were not measured here, such as the release\nand replacement of hormones ( Iwama\n et al ., 1998 ; Iwama\n et al ., 1999 ; Alfonso\n et al ., 2020 ), production of proteins ( Foster et al ., 1992 ; Jonassen et al ., 1999 ; Larios-Soriano et al ., 2020 ; Johansen et al ., 2021 ), and cell repair from oxidative stress\n( Lepock, 2005 ; Lushchak and Bagnyukova, 2006 ; Madeira\n et al ., 2013 ; Birnie-Gauvin\n et al ., 2017 ) that may still impose an energetic cost on fish\nfollowing a MHW. It is also possible that we would have observed costs in the metrics we\ninvestigated had we measured closer to the end of exposure (e.g. 1-week post-exposure).\nHowever, it is encouraging that while L. carponotatus is sensitive to MHWs,\nthere is a relatively rapid recovery of the physiological system within 2 weeks. This\nsuggests that MHWs similar to the length and duration used in this study may not have\nsignificant, longer-lasting effects on these fish. Physiological changes measured in this study are expected to be a result of thermal effect\non molecular processes, cellular stress response and cellular homeostasis response ( Kültz, 2003 ; Kültz,\n2004 ). When stressful conditions persist, genes may be up or down regulated to\nadjust cellular and whole organism phenotype in response to the altered environmental\nconditions. Only one study so far has investigated the molecular response of coral reef fish\n(damselfish and cardinalfish) to a natural MHW ( Bernal\n et al ., 2020 ). Interestingly, Bernal et al . (2020) found that during a MHW fish exhibited gene\nexpression changes that related to metabolic processes, cell damage and cell repair. At the\npeak of the MHW gene expression differences were associated with processes including\nmitochondrial activity, adenosine triphosphate activity and cholesterol and fatty acid\nmetabolism, which may indicate genomic level effects to the higher level metabolic processes\nin the present study. The average preference temperature of L. carponotatus (29.8°C) resembled\nthe +1°C MHW treatment (29.5°C) used in this study, where stress indicators (baseline\nlactate) and increased energetic costs (resting MO 2 and recovery time) were\nobserved. Similar temperatures have been recorded for other reef fish species from the GBR,\nsuch as the five-lined cardinalfish (29.5°C, Nay\n et al ., 2015 ) and the blue green damselfish (28.9°C, Habary et al ., 2017 ). It is possible\nthat other unmeasured physiological responses, such as reproduction and enzyme activity, may\nbe enhanced at warmer conditions, optimizing the overall fitness of this species at 29.8°C.\nHowever, preference temperature may not always align with natural selection due to\necological factors (territory, shelter, food) influencing body temperature ( Hugie and Dill, 1994 ; Lindberg et al ., 2006 ; Eurich\n et al ., 2018 ). This preferred temperature suggests that shorter\nwarming events (<2 weeks) may not pose a physiological challenge, but further\ninvestigations are needed. The avoidance temperature for L. carponotatus \nwas 3.4°C above the summer average and 1.1°C above preference, indicating a narrow thermal\nwindow. Concerningly, GBR reefs have already experienced temperatures exceeding this\navoidance threshold during MHWs in 2016, 2017 and 2020 ( AIMS, 2017 ; Hughes et al .,\n2018 ; Huang et al .,\n2024 ), suggesting some reefs have already been close to the species’ thermal limits.\nWhile our experimental evidence that fish will actively avoid warm temperature is perhaps\nencouraging for survival in nature, thermal refuges may not be available or used in nature.\n L. carponotatus has evolved in shallow tropical reefs and individuals may\nsimply remain in their established home ranges during MHWs and endure the physiological\nchallenges. The recovery speed from MHWs we observed may support this hypothesis, similar to\nthe strategy observed in the coral reef mesopredator, Plectropomus\nleopardus , which reduced activity but increased feeding rate as temperatures rose\nabove the summer average ( Scott et al .,\n2019 ). Regardless of the strategy employed, managing the energetic requirements\nduring MHWs in nature could be challenging. Interestingly, the physiological effects of the two different magnitudes of simulated MHW\n(+1°C and + 2°C) were often similar. The temperature increase of +1°C elicited a significant\nresponse in baseline lactate, post-capture lactate (exception 4-weeks of exposure), resting\nmetabolic rate, and recovery time, which was not proportionally increased further at +2°C\n(i.e. there was not an additive effect of each 1°C temperature increase in the MHW\ntreatments). This suggests that not all physiological processes will be affected linearly by\nMHWs which supports the hypothesis that tropical species are sensitive to relatively small\ntemperature increases and are living close to their thermal optimum during summer ( Tewksbury et al ., 2008 ; Sunday et al ., 2012 ; Rummer et al ., 2014 ; Comte and Olden, 2017 ; Rodgers et al ., 2018 ). The complexity of these\nvarious thermal physiological responses indicates the importance of understanding a range of\nphysiological traits when investigating the effects of future MHWs on wild populations, as\nno single metric is sufficient to comprehend the whole animal physiological response to\nelevated temperature. Due to the increasing frequency and intensity of MHWs, there is an immediate need to\nunderstand the sensitivity of organisms both during and following MHWs. This study shows\nthat while short-term (2–4 weeks) exposure to MHWs has significant effects on the\nphysiological response of a coral reef snapper there is a relatively rapid restoration to\nbaseline levels post-exposure within 2 weeks. One point to consider about these findings is\nthat the individuals used in this study were all caught by hook and line. This could\nintroduce bias towards specific phenotypes (e.g. bolder, more active, etc.) potentially\nleading to over- or under-representation of effects within the wild population. While there\nis limited literature on this, if phenotypic selection has occurred, we might expect that\nthese bolder individuals will have higher metabolic rates ( Fu et al ., 2021 ) and are the most thermally sensitive ( Norin et al . 2024 ). Nevertheless,\nwhile fish appear able to alter their physiological processes to cope with MHWs, the\nelevated resting metabolic rate suggests that fish still need to obtain about 20% more\nenergetic resources to sustain basic maintenance during MHWs. For mesopredators, this would\nultimately mean increased predation of smaller reef organisms, which could flow on to affect\nabundance and assemblage composition of lower trophic levels ( Boaden and Kingsford, 2015 ), or conversely a decline in condition if\nenergy requirements are not met ( Hempson\n et al ., 2018 ). If not able to be met with intake, these\nincreased energy demands could mean a trade-off by decreasing other activities like growth\nor reproduction, which might influence population dynamics of L.\ncarponotatus . Additionally, the stress on individuals during MHWs may be\ncompounded by fishing pressure. As MHWs induce a range of physiological responses in\nindividuals, additional pressure from fishing activities may exacerbate the physical strain\non this species. Catch-and-release fishing, whether recreational or commercial bycatch,\nduring MHWs may significantly impact individuals’ health and survival rates, and therefore\naffect a population’s overall health and persistence. In planning conservation measures, it\nwould be prudent to consider implementing fishing restrictions during heatwaves to alleviate\nthese stress effects. Further research into how the duration and intensity of MHWs affect\nthe physiology of a range of coral reef fishes would help us identify the physiological\nlimits, processes and costs that future MHWs may impose on these important ecosystems and\nfood resources globally."
} | 7,026 |
35643082 | null | s2 | 8,430 | {
"abstract": "Symbiosis is one of the most important evolutionary processes shaping the biodiversity on Earth. Symbiotic associations often bring together organisms from different domains of life, which can provide an unparalleled route to evolutionary innovation."
} | 62 |
39865091 | PMC11770067 | pmc | 8,431 | {
"abstract": "Terrestrial geothermal springs, reminiscent of early Earth conditions, host diverse and abundant populations of Archaea. In this study, we reconstructed 2,949 metagenome-assembled genomes (MAGs) from 152 metagenomes collected over six years from 48 geothermal springs in Tengchong, China. Among these MAGs, 1,431 (49%) were classified as high-quality, while 1,518 (51%) were considered as medium-quality. Phylogenomic analysis revealed that these MAGs spanned 12 phyla, 27 classes, 67 orders, 147 families, 265 genera, and 475 species. Notably, 575 (19%) MAGs represented new taxa at various taxonomic levels, and 2,075 (70%) lacked nomenclature and effective descriptions. The most abundant phyla of archaeal genomes were Thermoproteota , Thermoplasmatota , and Micrarchaeota . The DRTY, ZMQ, and ZZQ geothermal springs were predominated by Archaea, particularly by Thermoproteia and Thermoplasmata . These draft genomes provide new data for studying species diversity and function within terrestrial geothermal spring archaeal communities, thus contributing to the conservation and utilization of thermophilic and hyperthermophilic microbial resources."
} | 290 |
31292503 | PMC6620340 | pmc | 8,433 | {
"abstract": "Robust, buoyant, superhydrophobic PVB/SiO 2 coatings were successfully formed on wood surface through a one-step solvothermal method and a nanoimprint lithography method. The as-prepared PVB/SiO 2 /wood specimens were characterized by scanning electron microscopy (SEM), X-ray diffraction (XRD), Fourier transform infrared (FT-IR), thermogravimetric/differential thermogravimetric (TG–DTG) analyses. The superhydrophobic property and abrasion resistance of rose-petal-like wood were measured and assessed by water contact angle (WCA) and sand abrasion tests. The results show that PVB/SiO 2 /wood not only exhibited a robust superhydrophobic performance with a WCA of 160° but also had excellent durability and thermostability during the sand abrasion tests and against corrosive liquids. Additionally, the as-prepared PVB/SiO 2 /wood specimens show high buoyancy.",
"conclusion": "Conclusion In conclusion, natural wood with a robust, water resistant, mechanically stable, highly thermostable and highly buoyant performance was successfully fabricated by solvothermal deposition of hydrophobic monodispersed nano-SiO 2 microspheres, followed by a nanoimprint lithography treatment. This method introduced SiO 2 microspheres into a PVB solution and plated them onto a wood surface to be replicated. Thus, the resultant not only exhibited superhydrophobic performance with a water contact angle of 160° but also had high buoyancy. In addition, PVB/SiO 2 /wood exhibited excellent superhydrophobic properties in the liquids, sandpaper and ultrasonic tests. PVB/SiO 2 /Wood also presented excellent mechanical stability and thermostable properties.",
"introduction": "Introduction Natural wood, a low-cost and earth-abundant material, is ubiquitously used as a structural material across the globe 1 , 2 . Historically, wood timber has been extensively applied in vzrious daily applications, such as construction, indoor decoration, translation mining, transportation, furniture 3 , etc. however, it is susceptible to water absorption or water vapor, resulting in cracking, deformation, decay and insect-damage 4 , 5 , and thus, the application area is limited. One possible efficient solution would be to prepare a superhydrophobic coating for the surfaces of wood timber to prevent water adsorption 3 , 6 . Over the past few years, some interesting coatings with various properties, such as self-cleaning, anti-corrosion, superhydrophobic or fire resistance properties, have been studied 5 – 7 . However, most researchers reported only the method of superhydrophobic wood analysis, and only a few robust superhydrophobic wood surfaces have been developed 6 . However, few researchers have focused on superhydrophobic wood surface-outgrowth-induced high buoyancy performance. Nevertheless, developing robust, buoyant, superhydrophobic surfaces on wood substrates with a superhydrophobic coating will have great potential advantage for theoretical research and practical applications. Humans can learn from nature. After billions of years of evolution, nature creates countless mysterious living organosms that demonstrate almost perfect structures and functions. Learning from nature, many researchers have discovered many functional interfacial materials for applications, including materials with zwitter-wettability, superhydrophobicity, adhesion, self-cleaning, anti-snow, anti-fogging/icing, anti-oxidation, and corrosion-resistancepropeerties 8 – 18 . The example in nature include the self-cleaning superhydrophobic surface of the lotus leaf, toro leaf, and water lily 3 , 18 – 20 ; the directional catchment effect in a spider web 21 ; directional adhesive and self-cleaning superhydrophobic cicada’ wings, butterfly wings, and peacock feathers 22 , 23 anisotropic superhydrophobic rice leaves 24 , 25 ; high adhesion superhydrophobic peanut leaves 26 , 27 ; the superhydrophobic, high adhesive, and reversibly adhesive gecko foot 26 ; anti-freezing penguin’ wings 28 ; the robust, superhydrophobic water strider leg 29 , 30 ; the superhydrophobic, highly adhesive, and structural colour red rose petals 31 . Inspired by the approaches in nature, we will develop strategies to design and fabricate micro-nanoscale wood surfaces with superwettability to prevent the loss of performance in outdoor environments. Recently, more attention has been paid to the fabrication of superhydrophobic wood surface inspired by biological materials 7 . A variety of methods, such as hydrothermal, solvothermal, soft-lithography, sol−gel, photolithography, spraying, template, layer-by-layer, and self-assembly, have been used to replicate the biomimetic micro/nanostructures of surfaces 3 , 7 , 8 , 32 – 34 . As a technique for preparing micro/nanostructures, a solvothermal treatment combined with nanoimprint lithography methods can overcome many of the shortcomings of the above methods. To the best of our knowledge, there are no reports about the fabrication of a superhydrophobic surface with outgrowth-induced high buoyancy on wood by a two-step method of solvothermal deposition and nanoimprinted lithography. In this study, a petal-like PVB/SiO 2 superhydrophobic wood was successfully fabricated by using red rose petals, and cross-linked PDMS was used as the master template and stamp for nanoimprinted lithography after solvothermal deposition. The as-prepared PVB/SiO 2 /wood surface not only had robust superhydrophobic performance during the ultrasonic rinse and sand abrasion tests, but also stable super-repellency towards commonly used liquids, including brine, tea, milk and vinegar. Meanwhile, the buoyancy of a superhydrophobic surface is a very important influence on the properties of the material. Therefore, petal-like PVB/SiO 2 superhydrophobic wood could effectively prevent moisture from penetrating wood and improve the dimensional stability, which can satisfy our daily life application needs",
"discussion": "Results and Discussion The mechanism of the condensation reaction of hydrophobic monodispersed nano-SiO 2 microspheres can be expressed by the reaction in Equations ( 1 )–( 3 ): 1 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\rm{Si}}\\mbox{--}{({{\\rm{OC}}}_{2}{{\\rm{H}}}_{5})}_{4}+4{{\\rm{H}}}_{2}{\\rm{O}}+{{\\rm{NH}}}_{3}\\cdot {{\\rm{H}}}_{2}{\\rm{O}}\\to {\\rm{Si}}\\mbox{--}{({\\rm{OH}})}_{4}+4\\,{{\\rm{C}}}_{2}{{\\rm{H}}}_{5}{\\rm{OH}}$$\\end{document} Si – ( OC 2 H 5 ) 4 + 4 H 2 O + NH 3 · H 2 O → Si – ( OH ) 4 + 4 C 2 H 5 OH 2 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\rm{Si}}\\mbox{--}{({\\rm{OH}})}_{4}+{\\rm{Si}}\\mbox{--}{({{\\rm{OC}}}_{2}{{\\rm{H}}}_{5})}_{4}\\to \\equiv {\\rm{Si}}\\mbox{--}{\\rm{O}}\\mbox{--}{\\rm{Si}}\\equiv +4\\,{{\\rm{C}}}_{2}{{\\rm{H}}}_{5}{\\rm{OH}}$$\\end{document} Si – ( OH ) 4 + Si – ( OC 2 H 5 ) 4 → ≡ Si – O – Si ≡ + 4 C 2 H 5 OH 3 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\rm{Si}}\\mbox{--}{({\\rm{OH}})}_{4}+{\\rm{Si}}\\mbox{--}{({\\rm{OH}})}_{4}\\to {\\rm{Si}}\\mbox{--}{\\rm{O}}\\mbox{--}{\\rm{Si}}\\equiv +4{{\\rm{H}}}_{2}{\\rm{O}}$$\\end{document} Si – ( OH ) 4 + Si – ( OH ) 4 → Si – O – Si ≡ + 4 H 2 O According to previous results, a possible mechanism could be as follows. First, ethyl orthosilicate is hydrolysed to SiO 2 in aqueous ammonia. Hydrophobic monodisperse nano-SiO 2 microspheres are formed by the alcohol condensation and water condensation reaction. On the basis of the results above, a schematic illustration of the process of a replicating a biomimetic red rose petal through nanoimprint lithography is depicted in Fig. 1a . First, a mixture of curing agent and PDMS prepolymer (Sylgard 184 Silicone Elastomer Kit, Dow Corning) with a weight ratio of 1:10 was cast onto a fresh red rose petal to prepare the PDMS stamp, and the thickness was 3 mm. Then, the stamp was degassed in a vacuum container to remove air bubbles under the petal. After curing at 70 °C for 4 h, the solidified PDMS stamp was peeled off from the master template; thus, inverse petal structures were obtained. Tthen, (Fig. 1b ) natural wood was placed into the precursor PVB/SiO 2 solutions at 60 °C for 24 h in a Teflon-lined autoclave to increase bonding strength via furthersolvothermal treatment. This coating was plated on a natural wood surface, with the same replicate process, but the inverse petal structure cross-linked PDMS stamp was used as the master template to produce a petal-like structure. Finally, the PDMS template was peeled off the wood surface, and biomimetic PVB/SiO 2 /wood was fabricated. Figure 1 Schematic illustration of ( a ) the replication process of the PDMS stamp with a negative nanostructures of biomimetic red rose petals, ( b ) the fabrication process for PVB/SiO 2 /wood. Figure 2a shows SEM images of the red rose petal, inverse petal structure PDMS stamp, untreated wood, and biomimetic PVB/SiO 2 /wood. In Fig. 2a , the pristine red rose petal surface is covered by micro/nano-papillae with grooves and folds at the top of each papilla. This micro/nanostructure on the red rose petal surface allows for high adhesive superhydrophobicity because of the large adhesive force between the liquid droplet and the red roses petal surface. Figure 2b shows the SEM images of the as-prepared PDMS stamp surface, and the inverse petal micro/nanobiomimetic structures were observed on the surface of the red rose petal. In Fig. 2c , many traditional open vessel-to-vessel mesoporous structures can be observed on natural wood surface in a across section. After solvothermal treatment, the surface of the untreated wood was uniformly covered by a compact PVB/SiO 2 coating, as shown in Fig. 2d . After the second nanoimprint lithography step, Fig. 2e shows that micro/nano-papillae and nanofolds were found on the PVB/SiO 2 /wood surface. The top of each micro/nano-papilla includes nanofolds oriented towards the centre. The surface topography of the rose-petal-like PVB/SiO 2 /wood was similar to that of the fresh red rose petal surface. In addition, the small, superhydrophobic, monodispersed nano-SiO 2 microspheres can be found in the higher-magnification images (Fig. 2f ). Therefore, these results indicate that the nanoimprint lithography method successfully reproduced and retained the macrostructure and micro-nanostructure of red roses petal onto the wood surface. Figure 2 SEM images of the surface of red rose petals surface ( a ), inverse petal structures PDMS stamp ( b , c ) natural wood, PVB/SiO 2 coating ( d ), and biomimetic PVB/SiO 2 /wood ( e , f ) with different magnifcation. Figure 3 displays the XRD patterns of the untreated wood and PVB/SiO 2 /wood. In Fig. 3a , two strong characteristic crystalline diffraction peaks located at approximately 15° and 22° appear in the natural wood spectrum, and originate from the crystalline region of cellulose in natural wood. There was no other obvious characteristic peaks 2 , 35 . As shown in Fig. 3b , new strong diffraction peaks were observed after treatment, and these diffraction peaks at 34° could be well-indexed to the standard diffraction pattern of nanostructure SiO 2 (JCPDS 46-1045), except for the characteristic peaks of the untreated wood. This result suggests that the as-prepared PVB/SiO 2 solution had a high purity and no impurities. Figure 3 XRD patterns of ( a ) natural wood and ( b ) PVB/SiO 2 /wood. Figure 4 shows the FTIR spectra of the untreated wood and PVB/SiO 2 /wood. As shown in Fig. 4a , the main absorption bands in the FTIR spectra are located at ~3450 cm −1 , ~2908 cm −1 , ~1700 cm −1 and ~1429 cm −1 , corresponding to the O–H, C–O, C=O and C–H 3 stretching vibrations, respectively, and are attributed to the untreated wood. In Fig. 4b , the main absorption peaks of PVB/SiO 2 /wood at 3500 cm −1 are attributed to the stretching vibration of the hydroxyl groups (-OH). The absorption peaks at 3400 cm −1 become increasingly stronger and are mainly attributed to the stretching vibration of silicon hydroxyl groups. This result indicates that more -OH reacted with the superhydrophobic PVB/SiO 2 coating. The C–H 3 stretching vibration absorption at 2977 cm −1 and stronger absorption can be due to methyl groups. The absorption peaks at 1350 cm −1 indicate the stretching vibrations of C-F groups, and the peak at 1730 cm −1 corresponds to the C=O bond stretching vibrations. In addition, the strong absorption peaks at 810 cm −1 can be assigned to Si-O-Si and Si–CH 3 , which indicated that the monodispersed SiO 2 microspheres were deposited onto the wood surface and enhanced the untreated wood hydrophobic performance. The results demonstrate that the PVB/SiO 2 coating was successfully placed on the wood surface and existence of a long-chain-alkyl group on the surface, both of which were obtained through a one-step solvothermal method and a nanoimprint lithography method. Figure 4 FTIR spectra of ( a ) natural wood and ( b ) the PVB/SiO 2 /wood. The thermogravimetric and differential thermogravimetric analysis (TG–DTG) curves of the untreated wood and PVB/SiO 2 /wood are shown in Fig. 5 . As shown in Fig. 5a , a small weight loss of 2–3% was observed at 37–170 °C in both samples, which was attributed to the evaporation of adsorbed water. After the combined PVB/SiO 2 coating was added onto the wood surface, the initial decomposition temperature of PVB/SiO 2 /wood was approximately 303 °C, which was 30.5 °C higher than that of the untreated wood 35 , 36 . In Fig. 5b , the maximum pyrolysis rate of PVB/SiO 2 /wood occurred at 345 °C, and the pyrolysis rate was 51% lower than that the untreated wood, which might be attributed to depositing the PVB/SiO 2 coating on the wood surface. During the whole process of weight loss, the mass percentage of pyrolysis residue for untreated wood and PVB/SiO 2 /wood was approximately 8.9% and 29.8%, respectively. These results show that the thermal stability of PVB/SiO 2 /wood improved because of the strong interaction between the natural wood and the PVB/SiO 2 coating. Figure 5 TG-DTA curves of ( a ) natural wood and ( b ) the PVB/SiO 2 /wood. The surface wettability of the as-prepared PVB/SiO 2 /wood specimens were evaluated by WCA measurements with a volume of approximately 5 μL for the water droplets. Figure 6a shows the water droplets behavior on the untreated wood surface. The surface had a hydrophilic performances with a WCA of 10°. Such a small WCA was attributed to the hydroxyl groups on the natural wood surface and the low surface energy of open lumina of wood. Without nanoimprint lithography(Fig. 6b ), the solvothermal deposition of PVB/SiO 2 coating exhibited hydrophobic properties with a WCA of 107°. When a droplet was dropped onto the as-prepared PVB/SiO 2 /wood specimens surface was formed an approximate sphere, and couldn’t roll-off even when the wood was turned upside down (the inset). And the WCA value measured on the surface of PVB/SiO 2 /Wood was 160° (Fig. 6c ). The results show that the superhydrophilic surface of the natural wood was directly transformed into a superhydrophobic surface after solvothermal treatment with nanoimprint lithography. The Cassie and Baxter equation can theoretically be used to explain the superhydrophobicity of the as-prepared PVB/SiO 2 /wood. Figure 6 Superwettability of ( a ) natural wood, ( b ) wood with PVB/SiO 2 coating, and ( c ) PVB/SiO 2 /wood. To confirm the stability and robust superhydrophobic properties of PVB/SiO 2 /wood, 4 types of liquids, including brine, tea, milk, and vinegar, were used to examine the surface repellency toward commonly used liquids after an ultrasonically rinsed test for 24 h (Fig. 7a ). In addition, sandpaper abrasion tests were performed using awood surface weighting 150 g was, which was placed on sandpaper (standard sandpaper, grit no. 320 cW) and moved 20 cm along a ruler (Fig. 7c ). As shown in Fig. 7a,c , it is obvious that the surface of PVB/SiO 2 /wood repels the tested liquids, that is, the WCAs were closed to 150° under harsh conditions after various artificially accelerated ageing tests (Fig. 7b,d ). The high WCAs of the PVB/SiO 2 /wood surface can be attributed to the PVB/SiO 2 coating, which significantly demonstrated a robust superhydrophobic performance. Figure 7 Robust superamphiphobic performances of PVB/SiO 2 /wood measured by ( a ) an ultrasonic rinse for 24 h and ( b ) a sand paper abrasion test. ( c , d ) The corresponding WCAs of the PVB/SiO 2 /wood with liquid droplets of brine, tea, milk and vinegar. Water absorption, a crucial characteristic of natural wood, determines the ultimate applications. In this study, the water resistance of natural wood and PVB/SiO 2 /wood was investigated. The experiments were carried out by immersing natural wood and PVB/SiO2/wood samples in water for 45 days. As shown in Fig. 8a , the natural wood sample sank within 45 days after the specimen was fully immersed in water. The water absorption processes appeared during the typical five condition over the range of floating on the water surface to sinking into water. However, the PVB/SiO 2 /wood samples floated on the water surface for 45 days (Fig. 8b ). These performances can be attributed to the superhydrophobic paint coating, including the wood density and moisture content of the wood. The density of the cell wall material is approximately 1.5 g/cm 3 , whereas wood has a density of less than 1.0 g/cm 3 , allowing floating in water. Moisture can exist in natural wood as free water in cell lumens and as bound water within cell walls 37 . When the moisture content in both is saturated with water, natural wood samples will sink in water. Figure 8 Water absorption properties of ( a ) natural wood and ( b ) PVB/SiO 2 /wood, which are consistent with the schematic illustration of water absorption over 45 days. To determine the buoyancy of PVB/SiO 2 /wood induced by the superhydrophobic coating, two samples were used. As shown in Fig. 9 , the natural wood and PVB/SiO 2 /wood samples that retain wettability had buoyancy. Figure 9a shows that the surface energy of natural wood is higher than that of water, indicating that natural wood tends to attract water. However, PVB/SiO 2 /wood has a lower surface energy than the water. This result indicates that the water cohesion force of PVB/SiO 2 /wood was reduced by the superhydrophobic coating. The water surface close to the PVB/SiO 2 /wood sample has a convex meniscus, showing that the cohesion forces between PVB/SiO 2 /wood and water molecules are greater than the adhesion forces between the water molecules. The PVB/SiO 2 /wood loading ability determined by weight loading tests is showed in Fig. 9c,d . When the load was removed, wood floating was clearly observed on the water surface(Fig. 9d ), which was attributed to the water surface tension and number of air bubbles entrapped at the PVB/SiO 2 /wood surface. According to the Cassie-Baxter equation, PVB/SiO 2 /wood can be in an impregnating wetting state which water penetrate the micro/nano-papillae, and the air remaining in the nanofolds. However, the high buoyancy of PVB/SiO 2 /wood could be attributed to the surface of the superhydrophobic PVB/SiO 2 coating and the embedded air bubbles on the wood samples. Detailed analysis of the surface tension of the air-water interfaces and superhydrophobic coating of the PVB/SiO 2 /wood surface shows excellent buoyancy. Figure 9 Schematic illustration of the adhesion and cohesion forces of water molecules on ( a ) natural wood and ( b ) PVB/SiO 2 /wood surface. Supporting buoyant forces of PVB/SiO 2 /wood (water surface tension and air bubbles) applied to whole surface-coated samples after loading with weight. ( c , d ) Schematic illustration of the PVB/SiO 2 /wood sample in water after loading and removal loading weight."
} | 5,092 |
25945739 | PMC4444528 | pmc | 8,435 | {
"abstract": "The origin of the eukaryotic cell remains one of the most contentious puzzles in modern biology. Recent studies have provided support for the emergence of the eukaryotic host cell from within the archaeal domain of life, but the identity and nature of the putative archaeal ancestor remain a subject of debate. Here we describe the discovery of ‘Lokiarchaeota’, a novel candidate archaeal phylum, which forms a monophyletic group with eukaryotes in phylogenomic analyses, and whose genomes encode an expanded repertoire of eukaryotic signature proteins that are suggestive of sophisticated membrane remodelling capabilities. Our results provide strong support for hypotheses in which the eukaryotic host evolved from a bona fide archaeon, and demonstrate that many components that underpin eukaryote-specific features were already present in that ancestor. This provided the host with a rich genomic ‘starter-kit’ to support the increase in the cellular and genomic complexity that is characteristic of eukaryotes."
} | 253 |
37908647 | PMC10613949 | pmc | 8,437 | {
"abstract": "Reversed-electrowetting based droplet electricity generator (REWOD-DEG) shows merits in high power densities, tunable output formats, and wide adaptability to diverse mechanical energies. However, the surface charge trapping and dielectric failure, which are also common challenges for electrowetting system, hinders the development of reliable REWOD-DEGs for long-term running. We innovatively introduce a slippery lubricant-infused porous surface (SLIPS) into REWOD-DEG. Benefits from the significant inhibitory effect for surface charge trapping and ambient contamination, self-healing characteristic given by SLIPS, and robust reversed-electrowetting based energy harvesting were achieved. The SLIPS enhanced REWOD-DEG experienced 100 days of intermittent energy harvesting without deterioration. In addition, the device shows robust performances when exposed to a variety of extreme working conditions, like low temperature, pH, humidity, fouling, and even scratching. This work may address the core application challenges of REWOD based devices, and inspire the development of other robust droplet-based electricity generators.",
"conclusion": "3 Conclusion In this paper, a lubricant-infused porous surface (SLIPS) is introduced in the reversed-electrowetting based droplet electricity generator (REWOD-DEG), which has low contact angle hysteresis and low charge trapping characteristics and is capable of stable power generation for a long time at high bias voltage. Based on vibrating plate type REWOD-DEG, adjusting the conductive droplet concentration, vibration frequency and vibration amplitude to regulate the output power, the best matching load of the system was obtained, and the maximum output power of the generator was 143 nW. The SLIPS enhanced REWOD-DEG experienced 100 days of intermittent energy harvesting with no degradation in performance. In addition, when exposed to a variety of extreme operating conditions, such as low temperature, pH, humidity, fouling, and even scratching, the device shows robust performance.",
"introduction": "1 Introduction Water in nature contains tremendous kinetic energy, which is clean, renewable, and collectible; the advancement of green energy and micro/nano manufacturing technology has led to a growing interest in researching the harvesting of mechanical energy in liquid environments. 1–4 The conversions include triboelectric nanogenerators (TENG), 5–8 electrokinetic effect generator (EKEG), 9,10 piezoelectric nanogenerator (PENG), 11,12 electrical double layer capacitor (EDLC), 13,14 and reversed electrowetting (REWOD) based electricity generator, 15,16 etc. Among them, reversed-electrowetting based droplet electricity generator (REWOD-DEG) shows great advantages in high power densities (up to 10 3 W m −2 ), tunable output formats (several volts to tens of volts), and wide adaptability to diverse mechanical energies, such as energy harvesting from human movement and high-power generation from mechanical vibration, which supports promising application in portable, wearable devices. 17–19 As a solid/liquid capacitive power generation device with bias voltage, the performance of REWOD-DEG is limited by wetting behavior and the inherent charge trapping and dielectric failures of the electrowetting system. 18–20 Up to date, oil infused slippery surface has been introduced in variety EWOD or droplets energy harvesting applications, 21–24 its unique advantages given by the fluidic nature, such as low contact angle hysteresis, antifouling, self-healing and liquid repellency have been extensively studied. However, the dielectric properties of SLIPS film as a liquid dielectric layer was rarely reported. In our previous studies, we found the ion barrier effect of SLIPS in an electrowetting system, 25 which lights up the potential to address the crucial issue of charge trapping in reverse electrowetting (REWOD) based devices. This work innovatively introduced SLIPS into REWOD based droplet electricity generator for robust mechanical energy harvesting. The stability in long-time running and the tolerance to extreme operating conditions for the SLIPS enhanced REWOD-DEG were systematically studied.",
"discussion": "2 Results and discussion 2.1 Characterizations of SLIPS The slippery surface was prepared by spin-coating a lubricant onto a polytetrafluoroethylene (PTFE) film covered on an indium tin (ITO) glass (see Fig. 1a ). The lubricant with low surface tension easily infuses into the porous PTFE ( Fig. 1b ) and forms SLIPS, which is immiscible with most liquids. 26 The SLIPS film shows much higher transmittance compared to the original PTFE film ( Fig. 1c ). As shown in the insert, the text below can be clearly seen through the SLIPS covered glass. The lubricant liquid fills the air pockets in the porous film, reducing light scattering at the surface/air interface, 27 which can explain the enhancement in transmittance. Fig. 1 Characterization of the slippery surface (SLIPS). (a) Schematic drawing of SLIPS fabrication process, (b) SEM image of the PTFE membrane, (c) transmittance measurements of SLIPS film and original PTFE film, insert shows a SLIPS covered glass (inside the white dashed rectangular box) siting on a paper with printed characters, (d) the electrowetting behavior on pre-charged PTFE and SLIPS. A harsh charge trapping test of SLIPS and PTFE was conducted using the electrowetting-assistant surface charging method. 28,29 The surface of samples was charged under a 400 V bias applied between sample electrode and the aqueous solution (ESI: Fig. S1 † ). As widely agreed, the trapped charges on the surface will directly cause the electrowetting curve deviation from the theoretical prediction by the Young–Lipmann equation. 30–34 The data in Fig. 1d shows a linear and reversible electrowetting behavior on SLIPS, while a significant decline in droplet manipulation performance occurs on PTFE. The obvious symmetry differences of electrowetting curves (Fig. S2 † ) also indicates the introduction of SLIPS blocks surface charge trapping, which can be explained by the extra insulation effect given by the perfect oil-infusion to the pores inside PTFE. 2.2 SLIPS based REWOD-DEG performance The REWOD-DEG comprises a conducting liquid and a parallel conducting electrode substrate covered with a dielectric layer. A droplet is placed between two electrodes, and the distance between them is altered through mechanical modulation in the form of sinusoidal vibrations, achieved by vertically vibrating the bottom plate ( Fig. 2a ). The electrodes are linked to an external power source and the output of the system is tested by measuring the voltage drop across the load resistor R L . It should be noted that the wetting characteristics of the two conductive plates differ entirely. Specifically, the upper plate is covered with SLIPS, while the lower plate consists of a purely conductive ITO. Therefore, in this experiment, the upper plate is hydrophobic and the lower plate is hydrophilic. The circuit of the system is represented by an equivalent variable capacitor and a resistor connected in series, while an applied bias power supply is present ( Fig. 2b ). In terms of charge transfer, the process works as follows: initially, the lower electrode plate is positively charged due to the positive power supply. At the interface, the droplet and the lower electrode plate are electrostatically induced and negatively charged, forming a double-layer capacitance. The lower electrode plate and the droplet also form another double-layer capacitance to maintain system balance. As the lower plate vibrates, it increases the contact area with the droplet, attracting more positive charge from the power source. This movement of charge creates a current. Conversely, as the contact area decreases, the charge flows back and a reverse current is formed ( Fig. 2c and d ). Fig. 2 Schematic diagram of the experiment and circuit model. (a) Experimental setup, (b) circuit model of REWOD-DEG (inset: video image of water bridge), (c) when the droplet start to contact the plate and (d) at the very moment when the two plates are approaching each other. In REWOD-DEG, the electrical energy generated per unit area at the interface of the liquid-thin film dielectric-solid is directly proportional to the interface capacitance. 16 Therefore, the AC current generation in REWOD-DEG dependents on various parameters, including droplet concentration, vibration frequency and amplitude, applied voltage, and load resistance. The capacitance in the dielectric material is directly or indirectly affected by these parameters, which in turn affects the output of the energy collector. The REWOD-DEG process can be optimized by choosing the right conductive liquid and polymer coating to reduce the effects of contact angle hysteresis and charge trapping. 35–37 Common ionic liquids such as ultrapure water and NaCl solution (0.5 mol L −1 , 1 mol L −1 ) were used in this experiment. As expected, higher concentrations resulted in slight increase in V rms (the root mean square value of the voltage distributed on the load resistance V L ) compared to pure water ( Fig. 3a ). This can be attributed to the increase in mobile ion density, resulting in higher capacitance. Based on the essence of a variable capacitor for REWOD-DEG, high concentration solutions with high conductivity are preferred for pursuing high power output. 16,17 Therefore, in all subsequent experiments, unless otherwise specified, the high concentration NaCl solution of 1 mol L −1 was applied as the liquid choice. We studied the influence of driving waveform applied to vibrator on the energy output of REWOD-DEG. By comparing square wave, sine wave, and triangular wave, we found that sine waves give more stable and higher energy outputs, especially at a frequency of 5 Hz perform best (Fig. S3 † ). It is obvious that regardless of the type of droplets, the output voltage increases with the frequency increases ( Fig. 3a ). Limited by the hydrodynamic response, 38 the vibration frequencies used in our experiments did not exceed 20 Hz which are common movement frequency in daily life. Fig. 3 The energy output performance of REWOD-DEG under various working conditions. (a) The effects of vibration frequency and solution differentiation on the output V rms , R L = 10 MΩ, L = 1.2 mm, V = 20 V. (b) Energy generated per vibration cycle as a function of bias voltage at different vibration amplitude, R L = 10 MΩ, f = 5 Hz. (c) Load resistance and vibration frequency effect on V rms , V = 20 V, L = 1.2 mm. (d) Load resistance effect on power and current outputs. f = 20 Hz, V = 20 V, L = 1.2 mm. For all graphs, the droplet volume is 10 μL. The energy generated per vibration cycle as a function of bias voltage at different vibration amplitude is presented in Fig. 3b . As we discussed, to a REWOD-DEG, the appliable output energy is highly correlated to the energy changes in the equivalent capacitor, which can be written as . Therefore, a higher bias voltage or a larger change in solid/liquid contacts lead to higher amount of energy generated per vibration cycle. The trend of V rms outputs changing with load resistance was presented in Fig. 3c , in which the V rms outputs shows monotonic increase with vibration frequency and load resistance. By carefully choosing the working conditions, particularly the load condition, we got the maximum peak power of 143 nW ( Fig. 3d ). The REWOD-DEG shows potential application by lighting up LED lights based on droplet array approach (Fig. S4 † ). 2.3 Robust energy harvesting The accumulative trapped charges in the dielectric layer will decline the stability of the current generation of the REWOD-DEG system. 39 As shown in Fig. 4a , the SLIPS based REWOD-DEG shows very stable electricity output V L withstanding 100 000 vibration cycles. In the meanwhile, the electricity generation on the PTFE based REWOD-DEG presents obvious deterioration with the time, in particular at high bias voltages. The contrasting results improve the effective ion-blocking effect of the SLIPS, which supports the long-term running stability of the system. Fig. 4 The stability of power generation for SLIPS based REWO-DEG. (a) Long-term operating stability with one hundred thousand vibrations under different bias voltages. (b) The scratching damage of SLIPS can be restored within 20 minutes. (c) The output V L of the SLIPS based REWOD-DEG shows no decline after scratching test. (d) Long-term running performance for 100 days. R L = 10 MΩ, L = 1.2 mm, f = 5 Hz. To prove the self-healing potential coming from SLIPS, the DEG performances before and after physical scratching were studied. In its original state, the lubricant with ultra-low surface energy wets and fills the porosity, and further forms a defect-free complete film ( Fig. 4b ). Thanks to the fluidic nature of the SLIPS, once encountering physical damage, the wound can be self-healed by nearby lubricant through capillary wicking effect and high mobility. In our experiments, the light scratches on the surface repaired themselves within 1 second. After 20 minutes, the SLIPS was completely self-healed (Movie S1 † ). The power generation performance of the DEG before and after scratching ( Fig. 4c ) is comparable, which demonstrates the immunity of SLIPS covered DEG to physical damage. 23 To further investigate the stability of power generation of SLIPS, intermittent REWOD-DEG energy harvesting tests were performed over a period of 100 days, and no weakening of the output voltage was observed ( Fig. 4d ). SLIPS in extreme environments such as low temperature, high humidity, and acid–base conditions, lubricant injection still allows for excellent stability. 40–42 At room temperature (23 °C), SLIPS produces a slightly higher V L than PTFE ( Fig. 5a ), which is consistent with our conclusion above that SLIPS has a lower charge trapping rate. The two samples were then placed on a cooling platform at −4 °C and tilted 45°. Droplets were released directly over the samples with an injection pump, and ice coverage was recorded every 20 minutes with a camera. Remarkably, during the 80 minute test, the surface of SLIPS became less transparent only due to the condensation of water vapor in the air, and due to the ultra-smooth nature of the surface, no ice formed that would have resulted from the release of droplets. PTFE, on the other hand, exhibited many small condensation droplets and a layer of ice after 20 minutes, and the ice formation became more evident as the test duration progressed ( Fig. 5b and Movie S2 † ). We immediately REWOD-DEG the two samples tested at low temperature and found that PTFE performance decreased significantly because the ice layer prevented effective charge transfer on the surface. After that, samples are thawed at room temperature for 1 hour before REWOD-DEG. SLIPS still maintain stable output and have good low temperature resistance. However, the adhesion between PTFE and ITO substrate is obviously weakened after this process, the energy harvesting effect is affected by the failure to maintain steady contact with the droplet during vibration. Fig. 5 Extreme condition test. (a) SLIPS based REWOD-DEG performed (upper set of data) more stable under low temperature compares to PTFE based device. (b) SLIPS shows antiicing merit at low temperatures, and the impinging droplet slides off in a timely manner. In contrast, a droplet impinging on PTFE was easily pinned because of the formation of ice layer. The release height was 30 cm and the flow rate was 5 mL min −1 . (c) Stable power generation under high humidity conditions. (d) pH tolerance test. For all graphs the droplet volume is 10 μL, bias voltage = 20 V, R L = 10 MΩ, L = 1.2 mm, f = 5 Hz. Similarly, when considering practical applications such as energy harvesting during the rainy season, the lifetime of the superhydrophobic surface is susceptible to excessive humidity. 41–43 Therefore, we mimicked the conditions for energy harvesting at normal humidity (55%) and humidity during the rainy season (80%) at room temperature (23 °C) ( Fig. 5c ). The results demonstrate that SLIPS maintain stable output even in high humidity environments, without experiencing attenuation phenomena. Benefit from the excellent chemical inertness, liquid repellency and “slippery” properties, the SLIPS based REWOD-DEG can work with droplets over a wide range of pH ( Fig. 5d ). In addition, it also shows anti-fouling properties, which can realize self-cleaning (Fig. S5 † ) and repelling complex liquids (Fig. S6 † ). The above results clearly demonstrated that the introduction of SLIPS effectively improves the stability of REWOD-DEG."
} | 4,191 |
22363321 | PMC3282476 | pmc | 8,438 | {
"abstract": "Experimental approaches to identify horizontal gene transfer (HGT) events of non-mobile DNA in bacteria have typically relied on detection of the initial transformants or their immediate offspring. However, rare HGT events occurring in large and structured populations are unlikely to be detected in a short time frame. Population genetic modeling of the growth dynamics of bacterial genotypes is therefore necessary to account for natural selection and genetic drift during the time lag and to predict realistic time frames for detection with a given sampling design. Here we draw on statistical approaches to population genetic theory to construct a cohesive probabilistic framework for investigation of HGT of exogenous DNA into bacteria. In particular, the stochastic timing of rare HGT events is accounted for. Integrating over all possible event timings, we provide an equation for the probability of detection, given that HGT actually occurred. Furthermore, we identify the key variables determining the probability of detecting HGT events in four different case scenarios that are representative of bacterial populations in various environments. Our theoretical analysis provides insight into the temporal aspects of dissemination of genetic material, such as antibiotic resistance genes or transgenes present in genetically modified organisms. Due to the long time scales involved and the exponential growth of bacteria with differing fitness, quantitative analyses incorporating bacterial generation time, and levels of selection, such as the one presented here, will be a necessary component of any future experimental design and analysis of HGT as it occurs in natural settings.",
"introduction": "Introduction Bacteria in natural populations are known to import and integrate exogenous genetic material of diverse, often unidentified, origins (Eisen, 2000 ; Ochman et al., 2000 ; Lawrence, 2002 ; Nakamura et al., 2004 ; Didelot and Maiden, 2010 ). Bacterial genomes can be exposed not only to the multitude of sources of exogenous DNA present in their natural environments (Levy-Booth et al., 2007 ; Nielsen et al., 2007 ; Pontiroli et al., 2007 ; Pietramellara et al., 2009 ; Rizzi et al., 2012 ), but also to introduced sources of novel DNA such as the fraction of recombinant DNA present in genetically modified organisms (GMOs). Such exposure can potentially lead to horizontal gene transfer (HGT) events of GMO recombinant DNA, dependent on the multitude of parameters that govern HGT processes in various environments (Bertolla and Simonet, 1999 ; Bensasson et al., 2004 ). However, for long-term persistence of infrequently acquired genetic material in new bacterial hosts, a conferred selective advantage is considered necessary (Feil and Spratt, 2001 ; Berg and Kurland, 2002 ; Johnsen et al., 2009 ; Kuo and Ochman, 2010 ). Experimental investigations have shown that most HGT events that integrate into the bacterial chromosome are deleterious (Elena et al., 1998 ; Remold and Lenski, 2004 ). Thus, in terms of the persistence of its signature and its effects on fitness, HGT processes resemble routine mutational processes that take place at similarly low frequencies in bacteria and that are eventually lost from the population (Kimura and Ohta, 1969 ; Jorgensen and Kurland, 1987 ; Lawrence et al., 2001 ; Mira et al., 2001 ; Johnsen et al., 2011 ). However, rare HGT events and mutations can be positively selected under particular conditions and are the sources of bacterial adaptation and evolution (Imhof and Schlötterer, 2001 ; Townsend et al., 2003 ; Orr, 2005 ; Barret et al., 2006 ). HGT is particularly well known for playing a central role in the evolution of resistance to antibacterial agents (Bergstrom et al., 2000 ; Heinemann and Traavik, 2004 ; Aminov and Mackie, 2007 ; Aminov, 2010 , 2011 ). The detection of HGT events in a given bacterial genome can be performed retrospectively through bioinformatics-based comparative analyses (Ochman et al., 2000 ; Spratt et al., 2001 ; Nakamura et al., 2004 ; Didelot and Maiden, 2010 ). Alternatively, events may be detected via focused experimental efforts on defined bacterial populations under controlled conditions in the laboratory or monitoring efforts on subsamples taken from bacterial populations present in various environments, e.g., from soil, water, wounds, or gastrointestinal tracts (GITs; Nielsen and Townsend, 2004 ; Thomas and Nielsen, 2005 ; Pontiroli et al., 2009 ; Aminov, 2011 ). The latter approach can enable the identification of HGT events as they occur in the context of complex interactions of diverse bacterial communities. Its main limitation is sensitivity due to restricted sampling capacity of large bacterial populations, other methodological limitations, and cost of analysis. Representative analysis of HGT events in bacterial communities also depends on knowledge of the structure and population dynamics of the population and the sequence of the DNA transferred. Detection strategies frequently rely on hidden or implicit assumptions regarding the distribution and proportion of the individual cells in the sampled larger bacterial population that would carry the transferred DNA sequences (Keese, 2008 ; Heinemann et al., 2011 ). Large-scale cultivation of genetically modified plants (GM-plants) result in multitudinous opportunities for bacterial exposure to recombinant DNA and therefore opportunities for unintended horizontal dissemination of transgenes (EFSA, 2004 , 2009 ; Nielsen et al., 2005 ; Levy-Booth et al., 2007 ; Wögerbauer, 2007 ; Pietramellara et al., 2009 ; Brigulla and Wackernagel, 2010 ). In laboratory-settings, experimental studies of single bacterial species have demonstrated that bacteria can take up DNA fragments from plants and integrate them into bacterial genomes under highly optimized conditions (e.g., Gebhard and Smalla, 1998 ; De Vries et al., 2001 ; Kay et al., 2002 ; Ceccherini et al., 2003 ). In contrast, in natural settings, sampling-based studies of agricultural soils, run-off water, and GIT contents have found spread of transgenes from GM-plants, but negative or inconclusive evidence for HGT (Gebhard and Smalla, 1999 ; Netherwood et al., 2004 ; Mohr and Tebbe, 2007 ; Demanèche et al., 2008 ; Douville et al., 2009 ). Most research on HGT from GM-plants to bacteria has been performed via an assay after a limited time period following transgene exposure, perhaps in part because only limited explicit considerations of the population dynamics of HGT events have been presented to guide sampling design and data analysis (Heinemann and Traavik, 2004 ; Nielsen and Townsend, 2004 ; Nielsen et al., 2005 ). Given the low mechanistic probability of occurrence, horizontally transferred non-mobile DNA will initially be present at an exceedingly low frequency in the overall population. It may therefore take months, years, or even longer for the few initial transformants to divide and numerically out-compete non-transformed cells of the population to reach frequencies that can be efficiently detected by sampling efforts. The generation time of bacterial populations is therefore of high importance for detection efforts. Cell division time varies with species and environments and can be as short as <1 h in nutrient rich environments such as the GIT and up to several weeks in nutrient limited environments such as soil. The time lag between initial occurrence and potential detection will be present even though the relevant HGT events lead to positive selection of transformants (Nielsen and Townsend, 2001 , 2004 ; Heinemann and Traavik, 2004 ; Pettersen et al., 2005 ). Quantifying this time lag and determining the relationship between HGT frequencies and probability of detection requires mathematical models with dependency on several key parameters: HGT frequencies, changes in relative fitness of the transformants, bacterial population sizes, and generation times in nature. A few studies have accordingly begun to characterize the effects of natural selection and the probability of fixation of HGT events in bacterial populations (Nielsen and Townsend, 2001 , 2004 ; Pettersen et al., 2005 ). Here we integrate previous theory into a cohesive probabilistic framework that addresses current methodological shortcomings in the detection of HGT events and guides experimental design of future sampling of bacterial populations. Our analysis yields a simple formulation for the probability of detection given that a HGT actually occurred, and facilitates computation of the statistical power of an experimental sampling design. We apply the model to four different scenarios that are relevant for experimental monitoring of complex bacterial communities, accounting for both the adaptive dynamics of natural selection and the unknown timing of HGT events. In scenarios 1 and 2, the effects of variable DNA exposure are considered (i.e., exposed sub-population versus the total population of bacteria). Sampling occurs at the end of the DNA exposure period. In scenarios 3 and 4, the sampling is delayed until sometime after the DNA exposure of the bacterial recipients has ended. The total population size ( N ) and the strength of selection ( m ) varies in the scenarios (between N = 10 6 –10 12 , and m = 10 −10 –1). The m parameter represents the relative cost or advantage conferred by the HGT event to the transformant bacterium compared to untransformed members of the same population. In nature, m values would range from the reciprocal of the population size (weak positive selection) to near infinity (strong positive selection). The latter would for instance be caused by antibiotic treatment leading to death of all susceptible non-transformed cells. However, for most traits much lower values of m are expected. The m value of a given trait is not a constant and will depend on the environmental conditions. For instance, an antibiotic resistance trait can be highly advantageous in the presence of antibiotics (high positive m ) but confer a fitness cost in the absence of antibiotics (negative m value; c.f. Johnsen et al., 2011 ). Modeling Immediately following an HGT event into a large bacterial population, the lineage of bacterial cells carrying the novel transferred gene is highly vulnerable to extinction due to natural stochasticity in cell survival over the first generations (Fisher, 1922 , 1930 ; Haldane, 1927 ; Johnson and Gerrish, 2002 ; Pettersen et al., 2005 ). Subsequently, after the transformant population has established at higher numbers, it can be assumed to follow a fairly deterministic path, given continued directional selection. For a selected variant in transit to fixation with Malthusian relative fitness m per generation over t g generations, the current frequency p of a mutant starting at frequency p 0 can be modeled deterministically as (1) p 1 - p = p 0 1 - p 0 e m t g (Hartl and Clark, 1997 ; Nielsen and Townsend, 2004 ). With a single HGT event, the frequency of the transformant in a haploid population becomes 1/ N where N is the overall number of bacterial cells in the population of interest. Thus, for the frequency of a transformant ( p ) subsequent to a HGT event, (2) p 1 - p = 1 N - 1 e m t g (Nielsen and Townsend, 2004 ). Solved for p , Eq. 2 yields (3) p = e m t g N - 1 + e m t g . Because HGT events are relatively rare and presumably independent, we assume that the time delay until a HGT occurs is exponentially distributed, parameterized by a rate that incorporates the number of bacteria exposed, x , the rate of HGT per exposed bacterium, r , and the time, t x , during which exposure may occur (See Nielsen and Townsend, 2001 for a detailed description of these factors). Accordingly, the time to the next HGT, T x , would be distributed as (4) f T x t x = r x e - r x t x . The probability of fixation of a new variant gene in a haploid population has been characterized as (5) 1 - e - 2 m 1 - e - 2 N m , (Kimura, 1957 , 1962 ; Moran, 1961 ; Gillespie, 1974 ; see Patwa and Wahl, 2008 , for a review of alternate cases). Following the exponentially distributed occurrence rate (Eq. 4), filtered by the fixation process (Eq. 5), the timing until the occurrence, T , of the first HGT that is to be eventually fixed in the population, would be distributed as (6) f T ( t ) = 1 - e - 2 m 1 - e - 2 N m r x e - 1 - e - 2 m 1 - e - 2 N m r x t . Given that an HGT occurs that is on its way to fixation, what is the probability that such a transfer will be detected? This probability depends in part on the sample size of the monitoring effort, n . Here, n is treated as the number of bacteria in the environment sampled in a perfect assay for possession of the HGT event. If the frequency of the primary transformant and its offspring in the population at a given time is p , then the probability of detection is (7) 1 - ( 1 - p ) n . Because the frequency p of a HGT event/transformant that is under strong positive selection deterministically increases with time until it is fixed in the population (Figure 1 ), the probability of detection depends on the amount of time t since the first HGT event occurred, which depends on the time of first exposure to DNA of concern, T . The later the samples are taken, the greater the probability that a selected HGT event on its way to fixation will be detected. Figure 1 Scenario 1: HGT in large populations, no sampling delay; weak positive selection . Probabilities of detection of transformants in a large bacterial population N = 10 12 with HGT rates ranging from r = 10 −14 to 10 −5 , and weak positive selection of transformants ranging from m = 10 −10 to 10 −3 . The proportion of DNA exposed bacteria is low, medium and high ( x = 10 6 , 10 8 , and 10 10 , respectively, out of the 10 12 total bacterial population, from top to bottom), and the time period of DNA exposure is the same as the time to sampling: t x = t s = 20, 10 3 , and 10 5 , from left to right. Sample size n = 10,000 bacteria. The probability of detection for a HGT event on its way to fixation with selection coefficient m at time t g after the original transfer event is derived by substituting Eq. 3 for p into Eq. 7 (c.f. Nielsen and Townsend, 2004 ). Assuming the value of Eq. 3 is very small (i.e., population size is large and selection coefficient is sufficiently small), a useful approximation for the probability of detection of a HGT event on its way to fixation is 1 - 1 - e m t g N - 1 + e m t g n ≈ 1 - 1 - n e m t g N - 1 + e m t g = n e m t g N - 1 + e m t g . (8) However, for practical implementation, the probability term from Eq. 3 may not be known to be small. Furthermore, the unknown timing of the successful HGT is a key factor in the probability of detection. Therefore it would be best to integrate over all possible timings in order to calculate a representative probability of detection of HGT events. Noting that in this case t g = t s − t , this integration, from Eqs 4, 5, and 8, is ∫ 0 t x 1 - e - 2 m 1 - e - 2 N m r x e - 1 - e - 2 m 1 - e - 2 N m r x t × 1 - 1 - e m t s - t N - 1 + e m t s - t n d t , (9) or, moving factors that do not depend upon time t out of the integral, 1 - e - 2 m 1 - e - 2 N m r x ∫ 0 t x 1 - 1 - e m t s - t N - 1 + e m t s - t n × e - 1 - e - 2 m 1 - e - 2 N m r x t d t . (10) Equation 10 yields a prediction of the probability of occurrence and detection of a HGT event, and may be parameterized across a range of rates of HGT. For experimental design purposes (or for prediction for policy purposes), it may be important to calculate not just the full probability of detection, but also the restricted, higher probability of detection given that a successful HGT has occurred. This calculation can be achieved by dividing the result of Eq. 10 by the probability of any successful HGT event over the time t x , (11) 1 - e - 2 m 1 - e - 2 N m r x ∫ 0 t x e - 1 - e - 2 m 1 - e - 2 N m r x t d t ≈ 1 - e - 1 - e - 2 m r x t x . The approximation is valid provided N is large compared to m . Setting this approximation aside for generality, the larger probability of detection given that a successful HGT has occurred is then (12) ∫ 0 t x 1 - 1 - e m t s - t N - 1 + e m t s - t n e - 1 - e - 2 m 1 - e - 2 N m r x t d t ∫ 0 t x e - 1 - e - 2 m 1 - e - 2 N m r x t d t .",
"discussion": "Discussion We have presented a probabilistic framework for detection of initially rare HGT events/transformants by sampling of larger bacterial populations. This result expands on earlier studies (Nielsen and Townsend, 2001 , 2004 ; Pettersen et al., 2005 ) to derive a quantitative approach for analysis of the time scale over which HGT events take place and can be detected. Our population genetic framework facilitates practical implementation as well as a more detailed examination of the relative role of the key factors determining the fate of horizontally acquired genes in bacterial populations. The utility of the quantitative approach presented here is, although dependent on some knowledge of the rates of the relevant processes, independent of the specific mechanism of HGT (e.g., transduction, conjugation, transformation). The model is therefore equally applicable to understanding HGT processes between bacterial species/strains/cells as it is applicable to HGT events occurring between unrelated species. Quantitative adjustments to the DNA exposure and HGT rate can accommodate diverse mechanisms. Furthermore, the model identifies parameter values that should guide further hypothesis formation and experimental design. Selection The design of sampling approaches aimed at detecting rare HGT processes is deeply challenging, because the fundamental task is to detect a very low probability event with a very small sample size in a very large population. Field studies over limited time periods are correspondingly not likely to identify rare HGT events in large and complex bacterial communities (Figure 7 ). The detection of HGT events is therefore most often feasible only if the few initial transformants have a growth advantage so they increase their relative proportion in the overall population. However, there are methodological challenges to the implementation of defined selective conditions at the DNA exposure stage when rare transformants arise. In laboratory systems, the use of antibiotics at concentrations below the minimal inhibitory concentrations (MIC) can possibly apply such directional selection, and hence enrichment of rare transformants present in large, complex microbial communities. However, it is a non-trivial problem to experimentally achieve sub-lethal concentrations of antibiotics that confer directional selection of rare transformants without simultaneously limiting the viability of the overall bacterial population. In field systems, directional selection and enrichment of initially rare transformants will depend on the prevailing environmental conditions. There are usually few opportunities to introduce directional selection with controlled selection coefficients. As exemplified in this study, directional selection typically dominates determination of the probability of detection. Strong sampling designs would therefore avoid focus on the detection of the initial HGT events (and associated HGT frequencies), but rather attempt to detect positively selected descendants of the primary transformants. A shift in focus to the detection of descendants precludes precise determination of HGT frequencies. However, frequencies are poor predictors of the short and long-term (evolutionary) impact of HGT events. As long as such events occur repeatedly, other factors will determine the biological impact of these events (Pettersen et al., 2005 ). Our calculations are based on a fixed selection coefficient m . However, the strength of selection will frequently fluctuate over space and time due to environmental variables, as well as variability among bacterial genotypes attributable to gene-by-environment interactions (Kimura, 1954 ; Barker and Butcher, 1966 ). Thus, selection coefficients will be inexact and will rarely be amenable to robust quantification over variable environments. Furthermore, the genome of a given bacterial transformant will be exposed to other HGT events and mutational processes that may change the initial beneficial fitness effects of a given HGT event (Lenski et al., 1991 ; Gerrish, 2001 ; Heffernan and Wahl, 2002 ; Rozen et al., 2002 ; Barret et al., 2006 ; Johnsen et al., 2011 ). In practice, host and environmental variation prevents precise and meaningful quantification of m values. Theoretical modeling approaches, however, offer the opportunity to examine the effects of broad ranges of m , therefore providing opportunities to identify threshold values and to predict the dynamics of rare HGT events in larger bacterial populations. Fixation Our approach quantifies detection of HGTs that are on their way to fixation (Kimura, 1962 ), whereas the biological importance of HGT events arises at population proportions much less than one. For instance, the prevalence of a pathogenic strain carrying an HGT event encoding antibiotic resistance is of highest interest when its relative proportion among sensitive strains is <0.1–0.3; as higher proportions will lead to changes in clinical prescription guidance for first line antibiotic therapy (Daneman et al., 2008 ). Random or seasonal variations in local population sizes may also cause particular genotypes (e.g., transformants) to fluctuate at low frequencies above or below detection for long periods of time (Gerrish and Lenski, 1998 ). Genetic drift and uneven survival rates in structured bacterial populations are important in determining the fate of transformants (Heffernan and Wahl, 2002 ; Pettersen et al., 2005 ). The event of key importance is therefore when the transformant proportion rises to the point where subsequent evolution is largely deterministic based on the current level of directional selection (Rouzine et al., 2001 ). The probability of fixation of a transformant by genetic drift alone is governed by the inverse of population size. Given geographically dispersed and large population sizes, the fixation of a horizontally acquired gene/transgene in a bacterial population has been viewed as unlikely (Berg and Kurland, 2002 ). However, see also views by Majewski and Cohan ( 1999 ), Cohan ( 2002 , 2005 ), and Novozhilov et al. ( 2005 ). The likelihood of fixation of a neutral HGT event may differ from the likelihood of fixation of a neutral mutation; this is because mutations occur routinely and repeatedly in large bacterial populations, whereas HGT events may be much more tempo-spatially variable. Spatial considerations Although our model accounts for the effects of natural selection over time, it contains no inherent spatial component. Samples should be collected with consideration that rare horizontal transfers are not expected to occur and be distributed evenly in large, structured bacterial populations. Similarly, antibiotic resistance genes or transgenes are likely to be initially present only in a limited number of patches (e.g., patients/hospitals, or soil sites/fields); representing metapopulations of the larger global population (Maynard Smith et al., 2000 ). Initial frequencies will match their occurrence, but subsequent frequencies will correspond to the outcome of spatially variable directional selection and genetic drift. Migration between patches may also be of variable intensity and directionality. Uneven distribution patterns need to be considered in the sampling design. Field monitoring The analyses of published GMP field monitoring studies (Figure 7 ) indicated that detection of HGT events could only be achieved under circumstances of strong positive selection of the hypothesized transformants. Selection coefficients as high as m = 0.05, by evolutionary genetic standards, represent an extraordinary adaptive event. In the laboratory, selection coefficients as small as m = 0.01 can be measured, and over evolutionary time, selection coefficients as small as the inverse of the effective population size (here, this would be as small as 10 −10 bacteria per gram sample) are of importance in determining the genome composition of organisms. The retrospective analyses of these studies also suggest that increasing the sample size massively does little to increase the probability of detecting a HGT event that has occurred. However, increasing the delay between exposure and testing permits detection of HGT events characterized by much lower selection coefficients. Most of the field sampling-based HGT studies published so far have been based on a number of implicit assumptions on the characteristics of the biological system investigated. A more formal theoretical analysis of the population genetic aspects of the system investigated will contribute to make these assumptions explicit; and therefore provide improved clarity and robustness to future experimental design. The model presented here aims to provide guidance on future field-based sampling incorporating key population genetic factors. The multiple levels of, and importance of population genetic considerations in understanding horizontal gene flow have recently been reviewed by Baquero and Coque, 2011 , and references within) and Zur Wiesch et al. ( 2011 ). Clinical settings The Research topic for this particular issue of the journal is on resistance genes in the open environment, not in clinical settings. The practical scenarios examined in this study are therefore taken from non-clinical environments. However, the general insight of the presented study is also conceptually relevant to the general aspects of the population genetics of horizontal gene flow in clinical environments. This generality arises because, as we indicate in the Section “Discussion,” the model design does not rely on a given DNA transfer mechanism or particular environmental conditions. Our model examines the relationship between the four essential components determining the fate of initially rare HGT events in larger populations: (exposed) population size, HGT rates, bacterial generation time and selective advantage. These four population parameters are essential to the fate of HGT events occurring both in clinical settings among pathogens, as well as in non-clinical settings among non-pathogens. Despite the general insight to clinical scenarios that our model might provide, certain characteristics of the lifestyle of clinical pathogenic populations render specific calculations based on our model to be inappropriate. These characteristics include: Exceptionally strong selective environments are caused by the use of high doses of antibiotics for treatment of bacterial infections. An acquisition of a resistance gene under antibiotic treatment is exceptionally advantageous, as, 100% of the susceptible population is likely to die. Thus, the relative growth advantage is immense, leading to very rapid population expansion of the transformant population and the absence of competitors. Such strong positive selection is not comparable to the much weaker levels of positive selection for most other traits in non-clinical environments. Moreover, clinical antibiotic usage is also highly time-limited, producing strong fluctuations in the selection for a given resistance trait over time, that would require dynamic epidemiological modeling. We assume constant selection over time in our model. The infectious lifestyle of some pathogens leads to exceptionally rapid changes in their population sizes (during infections) followed by strong bottlenecks (during transmission). Thus, depending on the pathogen in question, the transformed cells may or may not be competing with non-transformed members of their populations. Thus, the fitness effects may have a different context in clinical environments depending on the characteristics of the infectious pathogen in question. The infection pattern of the pathogen in question will also determine its initial population size N , a value that would be very low for a strict pathogen (e.g., tuberculosis) but perhaps initially somewhat larger for opportunistic pathogens (e.g., Clostridium difficile ). However, our model assumes a more stable environment with a constant large population size, and is based on a competitive growth advantage of the transformant (relative to non-transformed members of the same populations; present in the same environment). This growth advantage will materialize as higher cell division rates for the transformant; the rate difference expressed through the m (Malthusian fitness parameter) value. Thus, a materialized growth advantage requires the presence of a much larger non-transformed population. In summary, our model is not designed to capture the intense short-term positive selection, population expansions (infections), coupled with bottlenecks (insufficient antibiotic treatment, and or transmission of a few bacteria to the next patient) that lead to different ranges of population genetic parameters and other model assumptions. From our point of view, such characteristics cannot be included in our model without addressing the etiology of infections and resistance patterns of individual pathogenic strains. Such developments are of high interest for further work. From the application of our model in the examination of various environmental scenarios, cases and literature examples, it can be concluded that some interspecies HGT is likely to occur over time and spatial scales not amenable to direct experimental observation. The model suggests sampling-based detection of the descendants (offspring) of the initial transformants is achievable; emphasizing that the probability of detection can only correspond to a calculable level of selection, and that a powerful experimental design requires a delayed sampling strategy. The recent publications by Gallet et al. ( 2012 ) and Toprak et al. ( 2012 ) present innovative laboratory approaches for quantification of weak positive selection, or for selection of initially rare but positively selected bacterial phenotypes."
} | 7,589 |
35496021 | PMC9049636 | pmc | 8,439 | {
"abstract": "Bacteriochlorins are crucial to photosynthesis in bacteria. Studies of air-stable, meso -substituted bacteriochlorins are rare. We herein report the synthesis, properties, and photovoltaic performance of three new air-stable, meso -substituted bacteriochlorins bearing a dioctylfluorenylethyne (denoted as LS-17), a dioctylaminophenylethynylanthrylethyne (LS-43), and a diarylaminoanthrylethyne (LS-45) as the electron-donating groups. Among these LS-bacteriochlorins, LS-17 displays sharp UV-visible absorption bands whereas LS-43 and LS-45 give rise to broadened and red-shifted absorptions. Electrochemical and DFT results suggest that the first oxidation and reduction reactions of these bacteriochlorins are consistent with the formation of the cation and anion radicals, respectively. For dye-sensitized solar cell applications, photovoltaic performance of the LS-45 cell achieves an overall efficiency of 6.04% under one-sun irradiation.",
"conclusion": "Conclusions In this work, we report the synthesis, fundamental properties, and photovoltaic performance in DSC of three new meso -diphenylbacteriochlorins. The LS-17, LS-43, and LS-45 bacteriochlorins bear a dioctylfluorenylethyne, a dioctylaminophenylethynylanthrylethyne, and a diarylaminoanthrylethyne as the electron-donating groups, respectively. By using the LS-11 dye from our previous work as a reference, LS-17 shows sharp UV-visible absorption bands whereas LS-43 and LS-45 exhibit broadened and red-shifted absorptions. Electrochemical results and DFT-calculated MO patterns suggest that the first oxidation and reduction reactions are consistent with the formation of bacteriochlorin cation and anion radicals, respectively. The energy level diagram suggests that the new bacteriochlorins are suitable for DSC application. The J – V and IPCE studies show that LS-45 outperforms others in the series with a PCE of 6.04%. Superior PCE of LS-45 may be attributed to the greater J sc and V oc values. A greater J sc is consistent with a broadened and red-shifted IPCE spectrum with no apparent dips. EIS, IMPS, and IMVS measurements suggest that the higher V oc of the LS-45 cell may be attributed to a greater charge recombination resistance. As to remedy the possible issue of molecular aggregation, synthesis of more soluble version of the LS-bacteriochlorins are underway.",
"introduction": "Introduction Dye-sensitized solar cells (DSCs) have received substantial attention as an alternative and renewable power source because of the benefits of low cost, convenient fabrication procedures, and versatile choices of dyes. 1–3 Power conversion efficiencies (PCE, η ) greater than 10% have been well demonstrated with the use of ruthenium complexes, 4–6 organic dyes, 7–10 and porphyrins. 11–17 It has been suggested that the photovoltaic performance of DSCs may benefit from red-shifted absorptions of the dyes. 18 Therefore, efforts have been devoted into developing dyes with near-IR absorptions. Examples have been reported to reach 900 nm while maintaining the PCE above 10%. 16,17 However, very complicated chemical structures of the dyes are required in order to achieve near-IR absorption bands. Bacteriochlorins are planar and conjugated macrocycles with multiple absorption bands. 19 In addition to the absorption bands in near-UV (B or Soret bands) and visible region (Q x ), bacteriochlorins show very intense and characteristic absorptions in near-IR region (Q y ). Owing to this unique and important property, bacteriochlorins play key roles in the bacterial photosynthetic reaction centers, including light harvesting, energy transduction, charge separation, and electron transfer, etc. 20–24 Inspired by the pioneer work of Lindsey and co-workers, 25–28 we recently reported the synthesis, fundamental properties, and photovoltaic performance of two air-stable, meso -substituted bacteriochlorins (denoted as LS-01 and LS-11, Fig. 1 ). 29 To the best of our knowledge, there have been very few reports of air-stable, meso -aryl-substituted bacteriochlorins . 28,29 In our previous work, PCE of the DSC reached 4.67% with the use of the LS-01 bacteriochlorin. This value outperforms that of a reference porphyrin dye with a similar structure (H2PE1, 2.06%, structure not shown). In addition, a greater PCE of 5.36% was achieved with the use of the LS-11 dye. Importantly, we demonstrated that light harvest of DSC reached 850 nm with dyes having simple structures. Fig. 1 Chemical structures of the LS-bacteriochlorins. In order to improve photovoltaic performance of bacteriochlorin dyes in DSC, we prepared three new meso -substituted, push–pull bacteriochlorins (denoted as LS-17, LS-43, and LS-45, Fig. 1 ) and studied their fundamental properties as well as photovoltaic performance in DSC. As shown in Fig. 1 , LS-11 is included in this work as a reference dye to compare with the new dyes. LS-11 employs a dioctylaminophenylethyne as the electron-donating group. This substituent has been used in an efficient porphyrin dye. 30 For LS-17, LS-43, and LS-45, they adopt electron-donating substituents from several efficient porphyrins in literature. 31–33 For LS-17, dioctylfluorenylethyne 31 is highly fluorescent, and the two octyl chains may improve solubility of the dye in organic solvents. For LS-43, it employs a dioctylaminophenyl-ethynylanthrylethyne as the electron-donating group. This donor group has been shown to largely red-shift absorption bands of a porphyrin dye. 32 For LS-45, diarylaminoanthrylethyne was chosen to be the electron-donating group because it also broadens absorption bands of a chromophore. This dual-functioned donor has been reported to be effective for both porphyrin and organic dyes. 33,34 As will be shown below, DSC sensitized with the LS-45 dye outperforms other dyes in this work with a PCE of 6.04%. This value is a noticeable improvement over that of reference LS-11 dye. 29",
"discussion": "Results and discussion Synthesis Preparation of the LS-bacteriochlorin precursor is as that described in our previous report, 29 except for the dihydrodipyrrin. Synthesis of the dihydrodipyrrin is sensitive to the composition of the TiCl 3 solution. Therefore, knowing the composition of the TiCl 3 solution is crucial to the reaction. 35 The procedure of dihydrodipyrrin synthesis is slightly adjusted in this work due to the change of TiCl 3 solution source (ESI † ). Syntheses of the LS-17, LS-43, and LS-45 bacteriochlorins are similar to those of related porphyrins in the literature. 29–33 Details of the syntheses and characterization data are put in the ESI. † Note that we encountered difficulties with LS-43 during chromatographic separation and 13 C-NMR measurement because of the lower solubility in organic solvents. This may be related to its more extended pi-conjugation and planar chemical structure (see below). UV-visible absorption and fluorescence emission spectra \n Fig. 2 compares the (a) UV-visible spectra and (b) normalized fluorescence emission spectra of the LS-bacteriochlorins in THF. Related data are collected in Table 1 . First of all, B (or Soret), Q x , and Q y bands of the LS bacteriochlorins are found around 390 nm, 600 nm, and 770 nm, respectively. This is consistent with the literature reports. More specifically, the B bands are observed at 388 nm for LS-11, 385 and 403 nm for LS-17, 386 nm for LS-43, and 383 nm for LS-45. In addition, LS-17 exhibits split B bands whereas LS-11, LS-43, and LS-45 show one broadened B bands. For the Q bands, LS-17 gives rise to a sharper and unified Q x band at 594 nm whereas LS-11 (576 and 615 nm, centered at 596 nm), LS-43 (609 nm), and LS-45 (598 nm) exhibit broadened and split Q x bands. In near-IR region, the Q y bands are sharp for all LS bacteriochlorins. Although the differences are small, we observe a trend of the Q y wavelengths as LS-43 (779 nm) > LS-45 (775 nm) > LS-17 (770 nm) ∼ LS-11 (769 nm). Among the LS-bacteriochlorins, LS-43 exhibits the most red-shifted Q y band in the series. This phenomenon may be related to its more extended pi-conjugation and is consistent with the literature report. 32 As for the weak absorptions between the B and Q x bands of LS-43 and LS-45, these bands may be attributed to the anthracene moieties in the chemical structures. This phenomenon is consistent with that of an anthracene-modified porphyrin in literature. 36 Fig. 2 (a) Absorption and (b) normalized fluorescence spectra of the LS-bacteriochlorins in THF. Absorption wavelengths, fluorescence maxima, and the first macrocyclic ring redox potentials in THF/TBAP Entry Absorption/nm (log ε , M −1 cm −1 ) Emission b /nm ( ϕ f c , %) \n E \n 1/2 / V vs. SCE Ox(1) Red(1) LS-11 a 388 (5.08), 576 (4.51), 615 (4.54), 769 (5.05) 774 (8.32) +0.75 d −0.95 LS-17 385 (5.01), 403 (5.04), 594 (4.78), 770 (4.96) 774 (7.83) +0.81 −0.92 LS-43 386 (5.02), 480 (4.51), 609 (4.76), 779 (5.04) 784 (6.07) +0.81 −0.85 LS-45 383 (4.99), 598 (4.68), 775 (4.99) 781 (6.74) +0.82 −0.84 a Taken from ref. 29 . b Excitation wavelength/nm: LS-11 (388), LS-17 (403), LS-43 (386), LS-45 (383). c The quantum yields were estimated by comparing with that of 1,3,3,1′,3′,3′-hexamethyl-2,2′-indotricarbocyanine iodide (HITCI) at 688 nm. d Potential determined by differential pulse voltammetry due to overlapped waves. For fluorescence emission, the emission bands appear to be the mirror images of their corresponding Q y bands. Also, the trend of the emission maxima is LS-43 (784 nm) > LS-45 (781 nm) > LS-17 (774 nm) = LS-11 (774 nm), consistent with the trend of the Q y wavelengths mentioned above. As shown in Table 1 , LS-43 was estimated to have a lower fluorescent quantum yield than other LS-bacteriochlorins. This may be attributed to its ease of molecular aggregation. As mentioned above, we encountered difficulties during chromatographic separation and 13 C-NMR measurement of LS-43 due to its lower solubility in organic solvents. Electrochemistry, energy level and molecular orbital patterns \n Fig. 3 shows cyclic voltammograms (CV) of the LS-bacteriochlorins in THF/TBAP. The first redox potentials are collected in Table 1 . One reduction reaction was observed for each LS-bacteriochlorin. These reduction reactions appear as reversible redox couples at −0.95, −09.2, −0.85 and −0.84 V vs. SCE for LS-11, LS-17, LS-43, and LS-45, respectively. On the other hand, two oxidation waves were observed. The first oxidation potentials of LS-11, LS-17, LS-43, and LS-45 were found at +0.75, +0.81, +0.81, and +0.82 V vs. SCE, respectively. These first reduction and oxidation potentials are consistent with those of the formation of a bacteriochlorin anion and cation radical, respectively. 29,37 Because the LS-bacteriochlorins share a common electron-withdrawing anchor ( Fig. 1 ), differences in the redox potentials may be attributed to the various electron-donors. Fig. 3 Cyclic voltammograms (CVs) of the LS dyes in THF/0.1 M TBAP, showing (a) reduction and (b) oxidation reactions. CVs of the blank solution (THF/0.1 M TBAP) are shown as the dotted lines. \n Fig. 4 depicts a diagram comparing energy levels of the highest occupied molecular orbital (HOMO), the lowest un-occupied molecular orbital (LUMO) of each dye, the conduction bands (CB) of TiO 2 , and the redox energy of the electrolyte. The first oxidation and reduction potentials of the LS-bacteriochlorins were used to estimate energy levels of the HOMOs and LUMOs, respectively. Importantly, LUMO levels of the dyes are considerably higher than the CB of TiO 2 whereas the HOMO levels are noticeably lower than redox energy of the electrolyte. Therefore, all LS-dyes should be capable of injecting electrons to the CB of TiO 2 upon excitation and the resulting cations can be regenerated by the electrolyte. Fig. 4 Energy level diagram of TiO 2 , LS dyes, and the electrolyte ( I − / I 3 − ). \n Fig. 5 depicts the frontier molecular orbital patterns of the LS-bacteriochlorins calculated by density-functional theory (DFT) at the B3LYP/6-31G(d,p) level. 38 Five orbitals are illustrated for each compound, from one orbital below the HOMOs to two levels above the LUMOs (or from HOMO−1 to LUMO+2). These patterns represent the pi-electron densities/probabilities of each orbital. As expected, the frontier MO patterns are consistent with the Gouterman's model with deviation. 19 For example, the HOMO−1 and HOMO resemble the so-called a 2u and a 1u orbitals, and the LUMO resembles one of the e g orbitals. For deviation, pi-electron densities/probabilities of these MOs delocalized towards the electron-donating/withdrawing substituents, owing to the pi-conjugation through the ethynyl bridges. Also, large probabilities at the LUMO+1 levels appear at the anthryl moieties of LS-43 and LS-45. They represent the anti-bonding orbitals of the anthracene, and are consistent with the literature report. 36 Importantly, the HOMO patterns concentrate at the macrocyclic cores and the electron-donating substituents, whereas the LUMO, LUMO+1, or LUMO+2 patterns largely reside at the macrocycle and the electron-withdrawing substituents. This translates to a push–pull tendency of the dye, pushing the electron density from the donor groups toward the anchoring groups upon excitation. This is a welcome merit of a dye for n-type DSC application. For the optimized geometry of LS-17 and LS-45, the alkyl chains are calculated to be pointing out of the bacteriochlorin planes upwards and downwards (ESI, Fig. S7 † ). This would increase the steric hindrance and help the dyes dissolve into organic solvents. In contrast, the alkyl chains of LS-43 are calculated to be extending outwards into the bulk, providing little steric hinderance. This would leave the core structure susceptible to pi–pi interaction/molecular aggregation, possibly resulting in a lower solubility. This suggestion is consistent with the above-mentioned difficulties that we encountered during chromatographic separation and 13 C-NMR measurement of LS-43. Fig. 5 Frontier molecular orbitals of the LS dyes, calculated by DFT at B3LYP/6-31(d,p). For clarity, the octyl chains were represented by ethyl groups in calculations. Photovoltaic properties \n Fig. 5 shows (a) the current density–voltage ( J – V ) curves under the irradiance of 100 mW cm −2 simulated AM1.5 sunlight (or one-sun irradiation, solid lines) and in the dark (dotted lines) and (b) the plots of incident photon-to-electron conversion efficiency (IPCE) as a function of wavelengths of DSCs sensitized with the LS-bacteriochlorins. Related data are put in Table 2 . Parameters of the LS-sensitized solar cells a Dye \n J \n IPCE \n sc \n b (mA cm −2 ) \n J \n sc (mA cm −2 ) \n V \n oc (V) FF \n η (%) LS-11 14.90 16.13 ± 0.20 0.52 ± 0.00 0.64 ± 0.01 5.35 ± 0.03 LS-17 13.83 14.90 ± 0.51 0.54 ± 0.00 0.64 ± 0.00 5.16 ± 0.15 LS-43 13.86 14.67 ± 0.23 0.52 ± 0.00 0.61 ± 0.00 4.63 ± 0.09 LS-45 15.26 17.43 ± 0.42 0.54 ± 0.01 0.64 ± 0.01 6.04 ± 0.05 a The photovoltaic parameters were obtained under simulated AM-1.5 G illumination (power density 100 mW cm −2 ). The active area was 0.25 cm 2 with a black mask of area 0.16 cm 2 for each cell. These parameters are averaged values of 5, 4, 4, and 5 cells for LS-11, LS-17, LS-43, and LS-45, respectively. b To compare with the J sc obtained from the J – V measurements, J IPCE sc is derived via wavelength integration of the IPCE spectra. For PCE, we observed a trend as LS-45 (6.04%) > LS-11 (5.35%) > LS-17 (5.16%) > LS-43 (4.63%). These values are comparable with other bacteriochlorin systems reported in literature. 29,39–41 Significantly, PCE of the LS-45 cell (6.04%) outperforms those of other devices in the series, especially that of the LS-11 cell (5.35%). Superior performance of the LS-45 cell may be attributed to its greater short-circuit photocurrent density ( J sc ) and a higher open-circuit voltage ( V oc ). For J sc , we observed a trend as LS-45 (17.43 mA cm −2 ) > LS-11 (16.13 mA cm −2 ) > LS-17 (14.90 mA cm −2 ) > LS-43 (14.67 mA cm −2 ). The greater J sc value of the LS-45 cell may be related to its broadened and intense IPCE spectrum with no apparent gaps/dips. Between LS-45 and LS-11, Fig. 6b shows that IPCE of the LS-45 cell outperforms that of the LS-11 device around 480 nm and beyond 800 nm. As a result, the LS-45 cell exhibiting a greater integrated J sc value ( J IPCE sc , Table 2 ). For V oc , V oc of the LS-45 cell is higher than that of the LS-11 device. This is consistent with its later occurrence of the dark current of the LS-45 cell ( Fig. 6a , red dotted line). Although the overall efficiency of the LS-43 cell is the poorest in the series, it does give rise to the most red-shifted IPCE spectrum, arriving at 920 nm. Fig. 6 (a) J – V curves and (b) IPCE spectra of LS dye-sensitized solar cells. Overall efficiencies of these specific cells are 5.36% for LS-11, 5.15% for LS-17, 4.63% for LS-43, and 6.02% for LS-45. Dye-loadings on the photo-anodes may also affect photovoltaic performance of DSCs. Therefore, dye-loadings of the LS-dyes on the photo-anodes were estimated by soaking the fully loaded photo-anodes in 0.01 M of tetrabutylammonium hydroxide (TBAOH) in THF. The results show that 110, 230, 280, and 180 nmol cm −2 of LS-11, LS-17, LS-43, and LS-45 molecules were de-sorbed from the TiO 2 anodes, respectively. Note that the LS-43 photo-anode carries the most dye molecules in the series. Yet, the LS-43 cell gives rise to the poorest photovoltaic performance. This may be caused by dye aggregation of LS-43 on the photo-anode. This suggestion is consistent with the above-mentioned difficulties during chromatographic separation and 13 C-NMR measurement of LS-43. Because LS-45 outperforms other dyes in this work, we therefore focus on comparing only LS-45 with LS-11 in the following experiments. Electrochemical impedance spectroscopy (EIS), intensity-modulated photovoltage spectroscopy (IMVS) and intensity-modulated photocurrent spectroscopy (IMPS) analyses were carried out in order to better understand the superior performance of the LS-45 cell over that of the LS-11 cell. Fig. 7 compares Nyquist plots of the LS-11 and LS-45 cells (a) under one-sun irradiation and (b) in the dark. Plots of (c) chemical capacitance ( C μ ) and (d) recombination resistance ( R rec ) vs. applied voltages were also obtained based on the equivalent circuit model (inset in Fig. 7d ). As shown in Fig. 7a , EIS responses from both cells are similar under one-sun irradiation. For Fig. 7b , the difference is more apparent in the dark. Three semicircles are expected in this figure. One small semicircle in the lower frequency region (from 0.1 Hz to 2 Hz) is associated with ion diffusion in the electrolyte. One larger semicircle in the middle frequency region (from 2 Hz to 1 kHz) is related to the charge-recombination at the TiO 2 /dye/electrolyte interface. And another small semicircle at the higher frequency region (>1 kHz) is corresponding with the charge-transfer processes at the Pt/electrolyte interface. 42 Occasionally, as in the case of Fig. 7b and a large middle-frequency semicircle may obscure the smaller semicircles. A larger resistance in the middle frequency region suggests a lower tendency of charge recombination, contributing to a higher V oc and a lower dark current. This is consistent with what we observed for the LS-45 and LS-11 cells in Fig. 6a and Table 2 . In Fig. 7c , the fitted C μ values are similar for both cells, suggesting that chemical capacitance may not be a major factor affecting the photovoltaic performance. In Fig. 7d , one observes a greater R rec value of the LS-45 cell than that of the LS-11 cell at a fixed applied voltage. This is consistent with the LS-45 cell having a greater V oc . For IMVS and IMPS, Fig. 8 compares (a) lifetime ( τ r ), and (b) collection time ( τ c ) against applied voltages. The plot in Fig. 8a shows a trend of τ r as LS-45 > LS-11 at a fixed applied voltage. A greater τ r value of the LS-45 cell is consistent with its greater recombination resistance ( R rec , Fig. 7d ), contributing to a higher V oc . For collection time, Fig. 8b shows little difference between the LS-45 and LS-11 cells. Again, this suggests that collection time may not be a major factor affecting the overall efficiency. Fig. 7 Nyquist plots of the LS-11 and LS-45 cells (a) under one-sun irradiation and (b) in the dark. Based on the equivalent circuit model (inset in (d)), (c) chemical capacitance ( C μ ) and (d) recombination resistance ( R rec ) vs. applied voltages were also obtained. Fig. 8 Plots of (a) life time ( τ r ), and (b) collection time vs. applied voltages for LS-11 and LS-45 cells."
} | 5,189 |
29755734 | PMC5946826 | pmc | 8,442 | {
"abstract": "We report on a comprehensive study of the unique adhesive properties of mats of polymethylmethacrylate (PMMA) nanofibers produced by electrospinning.",
"conclusion": "4. Conclusions We have studied the wetting and the unique adhesive properties of electrospun PMMA fibers. The fiber diameter has been intentionally varied in order to disentangle eventual fiber size-effects affecting the overall surface wettability behavior. On randomly oriented fibers, the apparent contact angle of deposited water drops is ∼130°, regardless of the fibers size. More interestingly, very large adhesion forces capable to hold water drops as large as 60 μL, more than twice the characteristic values obtained with hairy surfaces, are achieved. Aligned fibers present anisotropic wetting behavior, and a maximum volume of water drops retained in the direction perpendicular to the fibers up to 90 μL. Measurements carried out on free-standing fiber mats indicate that the presence of the glass substrate plays a marginal role in determining the above mentioned features. This work suggests electrospun polymers as very promising tool to tailor surface wetting behavior, through up-scalable production of nanofibers which might exhibit modulated interactions with liquids.",
"introduction": "1. Introduction Wettability is one of the most important properties of solids, affecting their surface mechanics, tribology, resistance, and biocompatibility, and being governed by both the chemical composition and the morphology of the involved interface. 1 , 2 In this respect, surfaces that have attracted a lot of attention in recent years are those exhibiting superhydrophobicity, namely an apparent static contact angle, θ , formed by water drops greater than 150° inspired by many plants and insects, which provides new and versatile ideas for designing materials with self-cleaning and antifouling properties, and drag reduction. 3 However, the dynamic behavior of these bioinspired surfaces can vary significantly. 4 – 6 On a lotus leaf, water drops roll off very easily even at inclination well below 10°, removing dust particles present on the surface (self-cleaning or lotus effect). 7 In contrast, large water drops stick to rose petals even though they are tilted upside down (petal effect). 8 Key features to achieve a specific superhydrophobic behavior involve both a proper chemical composition of the surface and an appropriate roughness at the micro/nanometer scale, 2 , 9 , 10 since in untextured surfaces θ is generally below ∼120°, which is the value characteristic of fluorinated materials. 1 Various physical and chemical methods have been employed to realize either self-cleaning 11 – 15 or sticky 8 , 16 – 19 superhydrophobic surfaces. In this framework, electrospinning provides a simple and practical way to tailor surface roughness over large areas through coatings made of fibers with diameters ranging from tens of μm to tens of nm, which are produced from polymer solutions with sufficient molecular entanglements. 20 – 22 To this aim, a high voltage is applied to the solution, which is extruded from a spinneret as an electrified jet. 23 The resulting materials, in a variety of forms ranging from individual nanofibers to non-woven mats with large area, have found application in many fields, including the realization of self-cleaning, superhydrophobic and superoleophobic coatings. 24 – 35 Their wettability has been mainly assessed by measuring the apparent contact angle (≅150°) and the roll off angle (≤10°), and explained in terms of the standard Cassie model with the drop contacting a composite landscape of trapped air and solid substrate. 1 Here, we focus on a different property, studying highly sticky hydrophobic fibers, and investigate a feature scarcely addressed 36 – 38 hitherto. The large adhesion is attributed to the water drop partially penetrating the surface texture according to a Cassie impregnating model. 1 We electrospin polymethylmethacrylate (PMMA) fibers onto different substrates, varying the diameter and the relative orientation of the polymer filaments (random vs. aligned configuration). Very large adhesion forces, capable to hold water drops as large as 90 μL are found. To better understand the role played by the substrate, the wetting data are compared with those obtained on free-standing mats. This study suggests new routes to produce coatings with tailored wetting properties which can easily cover extended surface areas. Potential applications of these findings include the design and the fabrication of new adsorbing media, catalytic surfaces, delivery of fluids with reduced or no volumetric loss, lab-on-chip architectures, and materials supporting water remediation."
} | 1,175 |
40287646 | PMC12034179 | pmc | 8,445 | {
"abstract": "Background The accessibility of sequencing technologies has enabled meta-transcriptomic studies to provide a deeper understanding of microbial ecology at the transcriptional level. Analyzing omics data involves multiple steps that require the use of various bioinformatics tools. With the increasing availability of public microbiome datasets, conducting meta-analyses can reveal new insights into microbiome activity. However, the reproducibility of data is often compromised due to variations in processing methods for sample omics data. Therefore, it is essential to develop efficient analytical workflows that ensure repeatability, reproducibility, and the traceability of results in microbiome research. Results We developed metaTP, a pipeline that integrates bioinformatics tools for analyzing meta-transcriptomic data comprehensively. The pipeline includes quality control, non-coding RNA removal, transcript expression quantification, differential gene expression analysis, functional annotation, and co-expression network analysis. To quantify mRNA expression, we rely on reference indexes built using protein-coding sequences, which help overcome the limitations of database analysis. Additionally, metaTP provides a function for calculating the topological properties of gene co-expression networks, offering an intuitive explanation for correlated gene sets in high-dimensional datasets. The use of metaTP is anticipated to support researchers in addressing microbiota-related biological inquiries and improving the accessibility and interpretation of microbiota RNA-Seq data. Conclusions We have created a conda package to integrate the tools into our pipeline, making it a flexible and versatile tool for handling meta-transcriptomic sequencing data. The metaTP pipeline is freely available at: https://github.com/nanbei45/metaTP .",
"conclusion": "Conclusion The study of microbiomes using meta-transcriptome sequencing enables the analysis of the activity of microbial communities that may have important roles in their environments. Over the past decade, omics technology has provided a theoretical foundation for understanding the distribution patterns and functional mechanisms of microorganisms under various conditions. Integrating different bioinformatic tools for this high-throughput data has become the burden for biologists. The data processing pipeline has strong potential to improve the reproducibility in meta-analysis studies. In our study, metaTP provides an analytical environment with reproducible workflows that efficiently process raw data into a gene expression matrix with reference-independent quantification methods. MetaTP also provides downstream analysis and visualization methods including functional enrichment and gene co-expression network analysis, which contain network topology calculations. Our hope is that this tool will serve in the future as a valuable resource for researchers."
} | 728 |
40263549 | PMC12015454 | pmc | 8,446 | {
"abstract": "Chickpea, a widely cultivated legume, actively fix atmospheric nitrogen in root nodules through a symbiotic relationship with rhizobia bacteria. A recombinant inbred line (RIL) population, progressing from F 2 to F 7 generations, was developed in a short-period of 18 months using the Rapid Generation Advancement (RGA) protocol. The F 7 RILs were evaluated during the 2020-21 and 2021-22 crop seasons under typical field conditions to quantify the effects of nodulation on seed yield (SY) and its associated traits. The analysis of variance revealed a highly significant difference ( P < 0.01) among genotypes for seed yield and other agronomic traits, with no significant seasonal effect. In the pooled analysis, nodulating genotypes (NG) exhibited a substantial increase ( P < 0.01) in SY (62.55%), 100-seed weight (SW100; 12.21%), harvest index (HI; 6.40%), number of pods per plant (NPPP; 39.55%), and number of seeds per plant (NSPP; 44.37%) compared to non-nodulating genotypes (NNG). Both NG and NNG exhibited a significant ( P < 0.01) positive correlation between SY and NPPP ( r = 0.64 and 0.63), NSPP ( r = 0.66 and 0.61), HI ( r = 0.27), and number of primary branches per plant (PBr) ( r = 0.31), respectively. The top-performing genotypes for yield and related traits were predominantly nodulating. Genotype-trait bi-plot analysis identified nine nodulating genotypes as the most adaptable across the two seasons—six for SY, plant height, SW100, and three for days to first flowering and maturity. These findings underscore the critical role of nodulation in maximizing chickpea yields and the significant yield penalties associated with non-nodulation. To boost chickpea production, future breeding efforts should focus on developing genotypes with high compatibility with rhizobium strains. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-025-98965-2.",
"conclusion": "Conclusion The RIL mapping population was developed in less than two years, the first report in chickpea to show the practical application of the RGA protocol in developing breeding material. The influence of nodulation on yield was distinctly favorable in nodulating genotypes, characterized by increased SY, biomass, and PH compared to NNG. The stable performing genotypes with high yield and early flowering nature were mostly nodulating in nature depicting the beneficial effect of nodulation on the crop growth under moisture stress conditions. The promising genotypes identified can serve as donors for use in chickpea breeding programs. Furthermore, the results emphasize the critical role of compatible Rhizobium strain/s in achieving optimal yields. The study highlights that the absence of nodulation can lead to substantially lower yields in chickpea. Therefore, the distribution and presence of Rhizobium strains in cultivated fields emerge as influential factors affecting final yield levels. To enhance the understanding of nodulation in chickpea, further research on identifying the genomic regions/QTLs/markers associated with nodulation is in progress to enhance our knowledge and provide valuable insights into the molecular mechanisms controlling nodulation in chickpea, ultimately contributing to improved crop productivity and promoting sustainable agricultural practices.",
"introduction": "Introduction Legumes have a specific mutualistic relationship with soil bacteria that fix the atmospheric nitrogen into plant-usable ammonical form in root nodules 1 , in contrast to other plants. Development of nodules and nitrogen fixation depends majorly on the legume cultivar and its specific rhizobial strain 2 – 4 which will in turn improve the overall crop growth, productivity 5 , 6 and soil health. Around 20 to 22 million tonnes of nitrogen per annum is fixed in the agricultural systems by the Leguminosae family members 7 which contributes significantly towards reducing the global carbon footprint. Chickpea ( Cicer arietinum L.) is a highly nutritious, diploid (2n = 2x = 16) legume crop grown in an area of 14.2 million hectares across 56 countries globally 8 , of which, India is the largest producer and consumer. It is considered a storehouse of proteins, complex carbohydrates, vitamins, and micronutrients that are required for human nutrition 9 – 13 . Being a leguminous crop, chickpea fixes the atmospheric nitrogen by association with Rhizobium species 14 which differentiates it from other cereal crops 15 . This ability to establish a symbiotic relationship is critical for legumes as it provides them with a readily available source of nitrogen, an essential macronutrient for plant growth and development. It stores the fixed nitrogen in the nodules present in its root system and converts it into ammonia that can be used by the plant 16 , 17 . It was estimated that about 70 kg of nitrogen per hectare is fixed annually by chickpea 18 , which helps in providing nitrogen not only to the host but also to the subsequent crops grown 19 thereby, helping the farmers in reducing the cost of production. Root nodulation is a complex process requiring the recognition of symbiotic bacteria, Nod-factor induced infection, and root growth 20 – 22 . Polyphenols named flavonoids released from the roots, stimulate Nod-factor production in rhizobia thereby initiating curling and colonization of root hair and the formation of root nodules 23 . Phytohormone signaling plays a major role during this process between the bacteria and the host 24 . The symbiotic nitrogen fixation (SNF) efficiency is dependent on the host, rhizobial strains, soil conditions 25 , availability of phosphorous 26 , and environmental conditions 27 . Under water stress, the biochemical activity in the nodules will get disrupted resulting in the senescence of nodules 28 – 30 along with downgraded leg-hemoglobin content and nitrogenase activity 31 . The ability of the nodules to supply energy, transport, regulate oxygen molecules, and assimilate ammonia to the plants will help in increasing plant growth and yield 32 . However, the consequences of the chickpea roots being unable to make a symbiotic relationship with rhizobium are not well understood and documented. Nitrogen is a key component of proteins, nucleic acids, and chlorophyll, all vital for plant function. Hence, effective nodulation directly impacts legume productivity and nutritional value. Previous studies in legumes have identified numerous genes and signaling pathways involved in nodulation, shedding light on the molecular basis of signaling between the plant and the bacteria 33 , 34 . Conversely, the genetic basis of non-nodulation in certain legume species has been linked to mutations in key nodulation genes or disruptions within the signaling cascades vital for establishing this symbiosis 35 . For example, studies in Medicago truncatula have revealed that mutations in the NFP (Nod Factor Perception) gene result in nodulation deficiency 36 . Unraveling the genetic mechanisms underlying both successful nodulation and its absence is crucial not only for enhancing nitrogen fixation in crops but also for broadening our understanding of the intricate plant-microbe interactions that underpin sustainable agriculture. Several studies reported in chickpea 14 , 37 – 39 as well as in other leguminous crops viz., soybean 40 , cowpea 30 , lupin 41 , groundnut 42 , etc., were focused majorly on the external application of biochar, artificial fertilizers, growth hormones and the findings of strain-specific effect on nodulation and yield-related traits. However, the information on the impact of nodulation over non-nodulation on yield and yield-contributing traits was minimal. In this context, the current study aimed to (1) broaden the knowledge on the association between nodulation, yield, and its associated traits; and (2) quantify the value gain of the traits in a RIL population segregating for nodulation trait. In addition, we successfully showcased the utility of rapid generation advancement methods for developing mapping populations in a short period.",
"discussion": "Discussion The formation of rhizobium nodulation is a key symbiotic mechanism in legume crops for their adaptation to marginal environments. The quantitative assessment of their impact on plant growth and economic yields is crucial for cultivar improvement and optimizing agricultural productivity. In this study, the selected parents are landraces, distinguishing one as a non-nodulating mutant (ICC 4918NN) derived from the other germplasm (ICC 4918). By employing the RGA approach, we successfully generated a RIL population within a mere 18 months, a testament to the efficiency and robustness of the methodology utilized 43 . While the potential of RGA has been acknowledged in other crops such as pea 44 , pigeon pea 45 , barley and wheat 46 , and canola 46 , its application in practical breeding programs remains largely unexplored. The variance analysis revealed a significant difference ( P < 0.01) among the RILs for all the traits under the study. This indicates the presence of ample variability for the traits in the population (Table 1 ). Significant genetic variability among genotypes was observed in earlier studies evaluated for nodulation-related traits in chickpea 47 , 48 . A notable finding in our study is the substantial increase in yield (62.55%) in NG compared to NNG. The improvement was majorly contributed by the increase in the NPPP (39.5%) and NSPP (44.4%). On a moderate level, nodulation has reduced the flowering time and maturity and enhanced the PH, SBr, SW100 and HI (Table 1 ). These results indicate that NGs were more efficient in synthesizing photosynthates, which led to produce more number of flowers for generating a large number of pods and seeds 49 , 50 than NNG. In addition, the yield advantage reflects not only the inherent genetic potential of the legume plant but also the synergistic effects of the established symbiosis between roots and rhizobial bacteria. The result agrees with an earlier study in chickpea, which reported a 31% higher yield in NG compared to NNG counterparts 51 , 52 . The current study, with a more pronounced yield advantage, emphasizes the significance of rhizobial nodulation in optimizing chickpea productivity. The poorer performance of NNG may be attributed to the deficiency of nitrogen fixation through nodulation, emphasizing its crucial role in chickpea productivity. The reliance on alternative nitrogen sources becomes crucial for NNG genotypes, and supplementing nitrogen in the form of fertilizers may be necessary to enhance yield 53 . This aligns with previous findings in chickpea and groundnut, where NNGs in nitrogen-rich soils could attain yields on par with NG’s 54 , 55 . Though the degree of reliance on nodulation for nitrogen fixation varies throughout legumes, it is perpetually present. For instance, in soybean, nodulation can contribute to a significant proportion of the plant’s nitrogen needs, with non-nodulating variants often displaying stunted growth and reduced yields 25 . Similarly, nodulation is critical for optimal growth in common beans, especially under nitrogen-deficient conditions 56 . Further investigations into the performance of NNG and NG under varying nitrogen availability can offer insights into their adaptability and yield potential across different soil conditions. The present investigation aimed to bridge a crucial gap in the existing knowledge base by assessing the impact of nodulation on agro-morphological and yield traits in chickpea. Our results align with existing literature on the positive effects of rhizobial nodulation on various growth characteristics, emphasizing the importance of this symbiotic association. The observed higher values in PH, PBr, biomass, and yield traits in NG compared to NNG underscore the pivotal role of rhizobial nodulation in enhancing plant growth and productivity in chickpea (Table 2 ) and mobilization of insoluble nutrients in the soil, leading to improved nutrient uptake in other legumes 57 , 58 . In addition, the absence of rhizobium nodulation resulted in a significant reduction in various growth parameters which might be due to a deficit in (a) host-dependent strain fitness 59 , (b) up-regulated expression of nif genes related to flavonoid synthesis 60 , and (c) maintenance of plant Pi (Inorganic phosphate) levels 61 . The estimation of genetic variability and inheritance through GCV, PCV and heritability allows the breeders to identify the traits with substantial genetic control and potential for selection in crop improvement programs. For phenological traits, the small difference between GCV and PCV values suggests a predominant influence of genetic factors on their variance. HI exhibited low estimates of both PCV and GCV, indicating a more substantial influence of environment (Supplementary Tables 2 and 3). This aligns with previous studies highlighting the major role of genetic components in the inheritance of flowering, maturity, and the environmental factors in determining HI 62 – 65 . For the traits NPPP, NSPP, HI, PBr, and SBr, a magnitude of low to high heritability was observed, suggesting several genetic factors controlling the inheritance of these traits. In particular, SY demonstrated a high magnitude of GCV, PCV, and heritability emphasizing its potential for genetic improvement through simple selection even in early generations efficiently 63 , 65 , 66 , 70 , 71 . This diverse heritability pattern underscores the importance of trait-specific breeding strategies to achieve improvements in chickpea agronomics 64 , 65 , 72 , 73 . High heritability coupled with high GAM was recorded for SY and SW100 across seasons (Supplementary Tables 2 and 3), indicating the predominance of additive gene action for these traits. Similar findings were reported under diverse genetic backgrounds in chickpea 71 , 74 , 75 , blackgram 76 as well as in cowpea 77 , 78 . Interestingly, the association among SY, NPPP, and NSPP was significantly positive in both NG and NNG (Fig. 2 ). This indicates the possibility of simultaneous improvement of multiple traits in chickpea genotypes and cowpea advanced breeding lines 39 , 62 , 78 , 77 . The ability to enhance multiple traits concurrently is crucial for developing improved chickpea varieties with enhanced agronomic performance. The scatter plot analysis (Fig. 3 ) demonstrated distinct performance patterns for yield and yield component traits across various seasons (Pooled, Rabi 2020-21 & 2021-22) in the NG and NNG indicating their superior performance in their respective seasons (Fig. 3 ). The use of scatter plots to differentiate genotypes based on their mean performance and its strategic application enhances the accuracy of genotype selection based on mean performance which is consistent with previous research conducted in chickpea 66 , 67 and mungbean 68 , 69 . Further analysis of the genotypes for their yield and trait interactions using GT bi-plot identified the checks, Phule Vikram and RVG 204 on its vertex with high values for all the traits under the study which can be considered the best adaptable genotypes (Fig. 4 ). For SY, PH, and SW100, the NG #68, 159, 13, 17, 187, and 123 while 106, 109, and 46 for DFF and DM were best adaptable over the seasons. Comparable research of this kind in various crops 79 – 81 , provide a valuable insight for the selection of superior genotypes with enhanced adaptability and nodulating nature in the chickpea breeding programs. The impact of rhizobium nodulation on grain yield and its association with yield-related traits was prominent in different legume crops. Earlier studies in chickpea emphasized the role of nodulation in enhancing seed yield and pod development 62 , 71 . Similarly, the seed yield was significantly improved by 40% in cowpea 82 , 33% in common bean 56 , and 45.6–50% in faba bean 83 , 84 compared to control or non-inoculated treatments, indicating the critical role of symbiotic nitrogen fixation in achieving higher seed yields in legumes. Moreover, the crop-specific strains, soils, and weather factors are crucial in achieving yields. These shared trends highlight the conserved nature of the nodulation mechanism inherited historically for thriving legume crops under diverse agro-ecologies. The comprehensive assessment of genetic variability, heritability, and genetic advance in this study provides a robust foundation for future chickpea breeding programs. The identification of traits with high heritability and substantial genetic advance, coupled with positive correlations among yield-related parameters, highlights avenues for effective genetic improvement. However, it’s crucial to recognize the crop-specific nature of these findings and to tailor breeding strategies accordingly. Future research should delve deeper into the molecular mechanisms underpinning nodulation in chickpea and explore comparative studies across leguminous crops to enhance our understanding of these complex interactions."
} | 4,269 |
37805632 | PMC10560283 | pmc | 8,447 | {
"abstract": "The architecture design and multi-scale learning principles of the human brain that evolved over hundreds of millions of years are crucial to realizing human-like intelligence. Spiking neural network based Liquid State Machine (LSM) serves as a suitable architecture to study brain-inspired intelligence because of its brain-inspired structure and the potential for integrating multiple biological principles. Existing researches on LSM focus on different certain perspectives, including high-dimensional encoding or optimization of the liquid layer, network architecture search, and application to hardware devices. There is still a lack of in-depth inspiration from the learning and structural evolution mechanism of the brain. Considering these limitations, this paper presents a novel LSM learning model that integrates adaptive structural evolution and multi-scale biological learning rules. For structural evolution, an adaptive evolvable LSM model is developed to optimize the neural architecture design of liquid layer with separation property. For brain-inspired learning of LSM, we propose a dopamine-modulated Bienenstock-Cooper-Munros (DA-BCM) method that incorporates global long-term dopamine regulation and local trace-based BCM synaptic plasticity. Comparative experimental results on different decision-making tasks show that introducing structural evolution of the liquid layer, and the DA-BCM regulation of the liquid layer and the readout layer could improve the decision-making ability of LSM and flexibly adapt to rule reversal. This work is committed to exploring how evolution can help to design more appropriate network architectures and how multi-scale neuroplasticity principles coordinated to enable the optimization and learning of LSMs for relatively complex decision-making tasks.",
"introduction": "Introduction The brain is a highly heterogeneous and powerful network of tens of billions of neurons possessing unparalleled feats of cognitive functions. Traditional artificial intelligence models are predominantly built on networks with hierarchical feed-forward architectures, different from the highly recurrent connected biological network in the brain 1 , making it difficult to match the results of natural evolution in terms of function and efficiency. Macro-scale reconstruction studies of human brain structure 2 confirmed the existence of a large number of non-hierarchical structures in the brain, such as modular structure 3 – 6 , hub structures 7 , 8 , small-world structures 6 , 9 . These topological properties enable the brain to better coordinate multiple cognitive functions to adapt to complex and dynamic environments and are also unconventional structures missing in existing brain-inspired AI models. Motivated by this, this work focuses on a network structure called Liquid State Machine (LSM) which can generate complex dynamics like the brain and facilitate the processing of real-time tasks. LSM 10 is a spiking neural network (SNN) structure that belongs to the reservoir, with randomly connected liquid layers and readout layers whose weights can be modified, as shown in Fig. 1 . Reservoir computing has achieved some progress in different fields, such as speech recognition 11 – 13 , image recognition 14 – 16 , robot control 17 , 18 , etc. Existing hardware designs for spiking neural networks can not only carry LSM 19 – 21 , but also adaptively switch synaptic plasticity 20 , and have been proven to be energy-efficient 22 , 23 . Figure 1 In the traditional definition of the LSM, randomly connected liquid layer neurons receive time-varying signals from external inputs and other nodes. Recursive connection patterns enable input signals to be converted into liquid layer dynamics and then abstracted by the readout layer. Some LSM models use fixed weights for the liquid layer, probably because its complex recurrent structure is difficult to be trained and optimized, which limits the learning capability and wide application of LSM 17 , 24 , 25 . Most of these existing models used gradient-based approach 26 – 30 to train the readout layer without training the liquid layer, resulting in a gap with the real learning mechanism in the brain. Some approaches 31 – 34 tried to train the liquid layer through local synaptic plasticity such as Spike-Timing-Dependent Plasticity (STDP) 35 or Hebb 36 , which is limited to simple tasks. In summary, there is still a need to explore biologically plausible learning rules applicable to LSM to optimize its liquid and readout layers. In addition, from the structure perspective, the liquid layer is usually fixed after initialization, simply serving as a way of high-dimensional encoding. Some methods 30 , 33 , 37 inspired by deep neural networks superposed multiple LSM layers as a deep LSM to solve machine learning tasks. These approaches have not explored the studies about dynamic LSM’ structure search in order to adapt to the changing tasks. And in fact, the human brain evolved rather than followed a specific design, which is different from current common AI algorithms. Evolution allows the brain’s nervous system to be continuously optimized and eventually evolve into non-hierarchical, highly-efficient structures. Inspired by this, some studies 25 , 26 , 28 , 32 , 38 proposed evolutionary methods for optimizing the parameters and structures of LSM. Previous study 17 assessed LSM according to three LSM’s properties 10 , and this work encoded the three LSM properties into the chromosome, and optimized the separation property (SP) as the objective function. Using SP as fitness is reasonable because it could reflect the role of evolution in the network dynamic adjustment. However, this work is limited to simple control tasks. A study 26 developed a three-step search method based on the genetic algorithm (GA) to search the network architecture and parameters of LSMs. Some researchers 26 , 28 , 32 directly used the experimental data set as a criterion for evaluating the fitness of LSM. These approaches lack effective exploitation of the internal dynamics of LSM. Considering the various limitations of existing LSM’s studies mentioned above, in this paper, we present a brain-inspired LSM with evolutionary architecture and dopamine-modulated Bienenstock-Cooper-Munros (DA-BCM) Rule. We consider the optimization of LSM from structure and function respectively. Structurally , we optimize the architecture of liquid layer according to an evolutionary perspective to obtain a more brain-inspired effective structure with a higher separation property. Functionally , instead of the gradient-based method, we propose a biologically plausible learning method with the combination of local trace-based BCM 39 synaptic plasticity and global dopamine regulation. The experimental results show that the proposed evolved DA-modulated LSM is able to learn the correct strategy faster and flexibly adapt to rules reversal on multiple reinforcement learning tasks. As reservoir computation exhibits complex dynamics consistent with activity in brain neural circuits, the evolvable LSM based on DA-BCM provides us with a powerful and biologically realistic tool to delve deeper into the learning process of complex networks. This work provides new opportunities for developing more brain-inspired complex and efficient network models, building adaptive learning frameworks, and revealing the evolutionary mechanisms of brain structures and functions.",
"discussion": "Discussion Evolution has not only designed the brain’s general connectivity patterns but has also optimized a multi-scale plasticity coordinated learning rule, endowing the brain with the ability to flexibly adapt to reversal learning and enabling efficient online learning. Inspired by this, this paper proposed a structurally and functionally optimized LSM model that incorporates adaptive structural evolution and biologically plausible DA-BCM learning rule. Experimental results demonstrated that the structural evolution of the liquid layer and the DA-BCM regulation of the liquid layer and the readout layer significantly improved multiple decision-making tasks. Most existing works 26 – 34 used backpropagation-based methods (which are suitable for hierarchical networks) to optimize the readout layer without considering the optimization of the liquid layer, or only adopted unsupervised STDP to optimize the liquid layer. Our model proposed a DA-BCM learning rule for both the liquid layer and the readout layer, which shows more biologically plausible. In addition, unlike existing structural search methods that directly search for the highest-performing structure, we took inspiration from the evolutionary mechanism and optimized the structure of the LSM according to its internal properties. Here, we would like to compare our approach with other reinforcement learning models, including the classical Q-learning 51 ,DQN 52 , and LSTM 53 (learning via policy gradient algorithm) with recurrent structure. In LSTM configuration, the network consists of one layer of LSTM with 128 hidden neurons and one fully connected layer. The Bellman equation Q-learning uses as Eq. 15 , where \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\gamma =0.9$$\\end{document} γ = 0.9 , \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\alpha =0.1$$\\end{document} α = 0.1 . Agent’s action is selected according to the \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\epsilon$$\\end{document} ϵ -greedy algorithm ( \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\epsilon =0.8$$\\end{document} ϵ = 0.8 ), which means that there is a probability of 0.2 for each selection to explore the action space randomly. The reward discount value \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\gamma$$\\end{document} γ and learning rate \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\alpha$$\\end{document} α are set to 0.99 and 0.1, respectively, in DQN. The loss function of the Q network is constructed in the form of mean square error, as shown in Eq. 16 . The DQN network, which is fully connected, consists of three layers, the input layer, the hidden layer (with a size of 50), and the output layer. In Eq. 16 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\gamma$$\\end{document} γ is set to 0.86. Learning rate in the used adam optimizer is set to 0.1. For fairness, multiple experiments are performed for each comparison algorithm, and the performance is averaged. The results for LSTM, Q-learning, and DQN are averaged over multiple runs ( \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=20$$\\end{document} n = 20 ), where LSTM and DQN run 1000 episodes each. 15 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\begin{aligned} Q(s, a)&=Q(s, a)+\\alpha \\left[ R(s, a)+\\gamma \\max _{a^{\\prime }} Q^{\\prime }\\left( s^{\\prime }, a^{\\prime }\\right) -Q(s, a)\\right] \\end{aligned}$$\\end{document} Q ( s , a ) = Q ( s , a ) + α R ( s , a ) + γ max a ′ Q ′ s ′ , a ′ - Q ( s , a ) 16 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\begin{aligned} \\omega ^{*}&=\\arg \\min _\\omega \\frac{1}{2 N} \\sum _{i=1}^N\\left[ Q_\\omega \\left( s_i, a_i\\right) -\\left( r_i+\\gamma \\max _{a^{\\prime }} Q_\\omega \\left( s_i^{\\prime }, a^{\\prime }\\right) \\right) \\right] ^2 \\end{aligned}$$\\end{document} ω ∗ = arg min ω 1 2 N ∑ i = 1 N Q ω s i , a i - r i + γ max a ′ Q ω s i ′ , a ′ 2 Figure 6 shows policy maps of the agent under the four models in the T-maze experiment. It can be seen that the well-performing model (Fig. 6 a, our model) learns to avoid poison and obtain food faster. And poorly performing models not only get more poison, but hang around along the way. Table 4 and Fig. 7 compare the average reward of the evolved \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$N_{opt}$$\\end{document} N opt individuals under different learning rule applications in detail. From the results, it can be seen that the efficiency of our proposed model is better than the comparison algorithms in terms of both mean and stability (variance). In fact, by combining Table 3 and Table 4 , it can be found that three evolutionary LSMs (DA-BCM+DA-BCM, STDP+DA-BCM, none+ DA-BCM) outperform LSTM and Q-learning in two tasks. We can also see that on the T-maze task, the performance of LSTM and DQN are significantly weaker than other models, and the variance of LSTM is very large, which may be caused by too many parameters that bring overfitting in a small sample learning task. In Flappy Bird, although DQN performance is better than LSTM and Q-Learning, the variance is very large. The overall efficiency is not as good as our model. Table 4 The final performance (average reward R and its variance) for the T-maze and Flappy Bird tasks and computational cost of different methods. Learning methods T-maze Flappy Bird Computational cost Q-learning 51 348.4±12.81 5.19±0.04 84 LSTM 53 169.95±72.61 4.54±0.40 68483 DQN 52 80.1±11.68 5.36±0.62 403 Ours 464.4±5.15 5.43±0.03 81.92 Computational cost analysis We also consider the impact of the computational cost of the model on fairness. Take T-maze for example. In our experiments, the three-layer LSTM needs to take into account the weights of 4 gates, so the total number of trainable weights is 68483. For Q-learning, only a state-action table of size \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$28*3=84$$\\end{document} 28 ∗ 3 = 84 needs to be stored. The number of trainable weights of the fully connected network inside the three-layer DQN is \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$4*50+50+50*3+3=403$$\\end{document} 4 ∗ 50 + 50 + 50 ∗ 3 + 3 = 403 . As for the LSM model we proposed, considering that the connection density of the evolved model is less than 2%, the number of connections of the liquid layer is up to about \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$64*64*0.02=81.92$$\\end{document} 64 ∗ 64 ∗ 0.02 = 81.92 . Including the number of connections between the liquid layer and the input and output, the total number of parameters is about 109.92 on average, of the same magnitude as Q-learning. Therefore, the computational cost of our proposed model belongs to a low level compared to DQN and LSTM. To sum up, this work breaks through the fixed deep hierarchical network structure that relies on BP optimization used in AI, and develops a multi-scale biological plasticity coordinated learning rule (instead of BP) and an efficient structure evolution for LSM. Because the proposed model borrows the information processing mechanism of the brain from the structure and function, it is more biologically plausible, more flexible and efficient, and naturally more suitable for developing human-like cognitive intelligence. Although this paper demonstrates the superiority of our proposed evolutionary LSM model in terms of model efficiency and computational cost, there are still some limitations. For instance, while reservoirs trained with gradient methods 26 – 34 are typically limited to the readout layer, applying them to the liquid layer remains challenging. Further exploration of gradient-based approaches suitable for the reservoir architecture may be a direction to improve efficiency. There is still more room for exploration for developing brain-inspired models in learning algorithms, and many neural mechanisms are waiting for us to investigate and fully apply in the AI field. This paper focuses on the small sample learning environment. Other application scenarios can also be used to further explore more energy-efficient self-organizing brain-inspired evolutionary spiking neural networks."
} | 4,670 |
20886243 | PMC3077745 | pmc | 8,452 | {
"abstract": "Summary Mycorrhizal species richness and host ranges were investigated in mixed deciduous stands composed of Fagus sylvatica , Tilia spp., Carpinus betulus , Acer spp., and Fraxinus excelsior . Acer and Fraxinus were colonized by arbuscular mycorrhizas and contributed 5% to total stand mycorrhizal fungal species richness. Tilia hosted similar and Carpinus half the number of ectomycorrhizal (EM) fungal taxa compared with Fagus (75 putative taxa). The relative abundance of the host tree the EM fungal richness decreased in the order Fagus > Tilia >> Carpinus . After correction for similar sampling intensities, EM fungal species richness of Carpinus was still about 30–40% lower than that of Fagus and Tilia . About 10% of the mycorrhizal species were shared among the EM forming trees; 29% were associated with two host tree species and 61% with only one of the hosts. The latter group consisted mainly of rare EM fungal species colonizing about 20% of the root tips and included known specialists but also putative non-host associations such as conifer or shrub mycorrhizas. Our data indicate that EM fungal species richness was associated with tree identity and suggest that Fagus secures EM fungal diversity in an ecosystem since it shared more common EM fungi with Tilia and Carpinus than the latter two among each other. Electronic supplementary material The online version of this article (doi:10.1007/s00572-010-0338-y) contains supplementary material, which is available to authorized users.",
"conclusion": "Concluding remarks It is well known that soil properties, climatic conditions, and physiological factors such as tree age affect EM fungal species richness in a stand (Conn and Dighton 2000 ; Wardle 2006 ). In accordance with other studies (Kernaghan et al. 2003 ; Ishida et al. 2007 ; Tedersoo et al. 2008 ), we found that the number of mycorrhizal species increased with increasing number of host tree species. Since the contributions of different tree taxa to mycorrhizal species numbers varied considerably, our study highlights that the increment in fungal species numbers depended on tree species identity or tree social status but was not simply a function of tree species numbers (Dickie 2007 ). Therefore, stand composition is important for below-ground mycorrhizal community species richness. We found clear host preferences. However, overall generalist fungi colonized the major fraction of root tips.",
"introduction": "Introduction In boreal and temperate forests of the palearctic realm, most tree species form ectomycorrhizal (EM) associations with a high number of fungal taxa. In these ecosystems, EM fungal species richness has been mainly studied with host trees of economic importance such as pine ( Pinus sylvestris ), spruce ( Picea abies ), oak ( Quercus spp.), and beech ( Fagus sylvatica ). High-throughput sequencing revealed extremely high fungal species diversity in soils of these forests (Reich et al. 2009 ). Meta-analysis across ecosystems indicated that each of these tree species can be colonized by 160 to 226 different EM fungal taxa (De Roman et al. 2005 ). Within a given ecosystem, there is also considerable EM fungal diversity. For example, in beech forests, roots of old-growth trees are colonized by about 80 to 90 EM fungal taxa (Buée et al. 2005 ; Rumberger et al. 2005 ; Pena et al. 2010 ). By far less, i.e., only 10 to 15 EM fungal species have been identified for other deciduous European tree species that occur in mixed forests together with Fagus such as Tilia spp. ( Tilia cordata Mill., Tilia platyphyllos Scop.) or Carpinus betulus (De Roman et al. 2005 ; Timonen and Kauppinen 2008 ). It is unknown whether Tilia and Carpinus are indeed associated with lower numbers of different mycorrhizal species than Fagus or whether these figures simply reflect differences in research intensity. There is also only limited information on the contribution of arbuscular mycorrhiza (AM)-forming tree species such as Fraxinus excelsior L. and Acer spp. ( Acer pseudoplatanus L., Acer platanoides L.) to mycorrhizal species richness in mixed deciduous forests. To date, tree species with a wide ecological amplitude such as Tilia , Carpinus , Fraxinus , and Acer (Ellenberg 1996 ; Marigo et al. 2000 ) are gaining importance for silvicultural management since mixed forests with these tree species may be more suitable to withstand climate change with lower summer precipitation anticipated for Central Europe (Gessler et al. 2007 ). It is expected that increasing tree species richness of forests will increase mycorrhizal fungal species richness due to host preferences of the fungi. This has been reported for boreal and temperate mixed coniferous–deciduous forests (Kernaghan et al. 2003 ; Ishida et al. 2007 ) as well as for wet, sclerophyllous forests in Australia (Tedersoo et al. 2008 ). Surprisingly, information on the importance of host species (i.e., root attachment) for mycorrhizal fungal taxa in mixed deciduous Central European forests is missing. To uncover host–fungus interactions that shape mycorrhiza diversity in mixed deciduous forests, we characterized mycorrhizal species richness in a forest composed of members of five tree families (Fagaceae, Tiliaceae, Betulaceae, Oleaceae, and Aceraceae). We hypothesized that (1) multi-host fungal species are dominant with respect to root colonization, (2) increasing richness of EM-forming tree species increases EM fungal species richness because of fungal host specificity, and (3) AM-forming tree species contribute little to mycorrhizal fungal species richness in mixed forests. To test these hypotheses, we have chosen mixed deciduous forests containing Fagus , Tilia , Carpinus , Acer , and Fraxinus in the National Park Hainich (Thuringia, Germany). The National Park is covered with old-growth forests, which have not been managed for several decades (Meinen et al. 2009 ). Study plots identified in same climatic conditions with similar forest and edaphic structures (Leuschner et al. 2009 ; Meinen et al. 2009 ) were used for multiple samplings in different seasons to investigate mycorrhizal fungal species richness and their host preferences in this ecosystem.",
"discussion": "Discussion A comparison of mycorrhizal species richness of different host taxa in a mixed deciduous temperate forest This study provides a comprehensive analysis of mycorrhizal community species richness in beech forests mixed with Tilia , Carpinus , Fraxinus , and Acer . Tilia has occasionally been reported to form AM (Harley and Harley 1987 and references therein), but this was not observed here. Both Tilia and Carpinus shared more common EM fungal taxa with Fagus than among each other. Therefore, we suggest that Fagus , which is the potentially dominant tree species in most Central European forests (Ellenberg 1996 ), secures EM fungal species richness and is therefore ecologically important as a warrantor of EM fungal diversity. Tilia was colonized by the same set of abundant EM fungi and moreover hosted a high number of unique EM fungal species with low abundance as did Fagus . Tilia was, thus, ecologically equivalent in fostering high EM fungal community species richness. This result is important because, in the light of the current debate on prevention of biodiversity erosion, our data suggest that Tilia is recommendable as a host taxon able to maintain high mycorrhizal diversity. Based on the data of our study, Carpinus appears to be less useful in this respect. If we corrected for the higher sampling intensities of Tilia and Fagus , their EM fungal species richness would decrease only marginally (−5 and −3 EM fungal taxa) and, thus, would still be 32% and 40% higher than that of Carpinus . However, on our study plots Carpinus was a subordinate tree species (Jacob et al. 2010 ). Overstory plants can influence EM fungal diversity of understory plants (Kennedy et al. 2003 ). Since EM fungal species richness, especially that of rare species, is strongly dependent on plant carbon productivity and supply with recent photoassimilates (Druebert et al. 2009 ; Pena et al. 2010 ), it is possible that carbohydrate allocation to the below-ground compartment was too limited for Carpinus to maintain high EM fungal species richness. This would suggest that, in addition to the host tree species composition, the stand structure might also have influenced mycorrhizal community richness. To unravel the factors controlling mycorrhizal fungal species richness, this aspect will deserve further attention in future studies. In contrast to EM fungi, the contribution of AM fungal taxa to mycorrhizal species richness was low (5%). Their host tree species Fraxinus and Acer were as abundant as Tilia and Carpinus , respectively, in this ecosystem (Meinen et al. 2009 ) and were found here in 66% and 23% of the soil cores. We can, therefore, assume that AM fungi have ample opportunities for root colonization. Employing pyrosequencing for the detection of fungi in different forest soils, Bueé et al. ( 2009 ) also found only one operational taxonomic unit (OTU) belonging to the glomeromycota compared to 33 OTUs classified as potential EM fungal species. These observations support earlier notions that the species richness of AM fungi is generally lower than that of EM fungi (Smith and Read 2008 ). Host ranges and preferences of EM fungi Host range and host specificity are important determinants of EM fungal community composition in mixed forests. In our study, the number of multi-host EM fungal species, i.e., fungi associated with Fagus , Carpinus , and Tilia , constituted only a small fraction (8%) of the total mycorrhizal species richness. Since these multi-host EM fungi colonized the largest fraction of the root tips of EM-forming host trees, our data support that mycorrhizal species with a large host range are strong competitors (Horton and Bruns 1998 ; Cullings et al. 2001 ; Kennedy et al. 2003 ; Richard et al. 2005 ; Walker et al. 2005 ; Ishida et al. 2007 ; Twieg et al. 2007 ; Hubert and Gehring 2008 ; Tedersoo et al. 2008 ). The advantage may be that plants in different environments can always find suitable fungal associates (Bruns et al. 2002 ). But the differences in colonization found here for different tree taxa suggest that multi-host EM fungal species still exhibit host preferences in a given ecological context or that their competitiveness differs on different host trees. The majority of EM fungi in this ecosystem showed clear host preferences. The category of fungi with narrow host range contained specialists, for example fungi typically associated with Fagaceae such as Lactarius blennius , Lactarius fluens , Russula fellea , Russula raoultii , and Russula solaris (Brand 1991 ; Beenken 2004 , Agerer in http://www.deemy.de/ ). An advantage of specialized associations may be improved adaptation to host physiology for nutrient exchange (Baxter and Dighton 2001 ; Hobbie et al. 2005 ) or other host or fungal benefits. Since the “host range” categories used in this study were based on colonization patterns, they reflect realization of ecological niches on the background of genetic affinities or barriers to certain plant–fungus interactions (Molina and Trappe 1982 ; Molina et al. 1992 ; Dickie 2007 ). Therefore, they are flexible rather than fixed entities. For example, various EM fungal species which have previously been documented only on Fagus or Quercus ( Inocybe umbrina ; UNITE, http://www.unite.ut.ee ), Peziza michelii (Tedersoo et al. 2006 ), Russula pectinatoides (Agerer, http://www.deemy.de ; Dickie and Reich 2005 ), Tomentella terrestris (Kjøller 2006 ), and C. cristata (Buée et al. 2005 ; 2007 ; Kjøller 2006 ) were found here for the first time on Tilia or Carpinus . Therefore, our study shows that the host range of these fungi is greater than previously known. Otherwise, fungi species known from the literature to associate with Fagus , e.g., Russula cyanoxantha (Agerer in http://www.deemy.de ; DeRoman et al. 2005 ; Grebenc and Kraigher 2007 ) and Cortinarius infractus (Garnica et al. 2003 ) were not colonizing Fagus roots but Tilia and Carpinus . This suggests that their host preferences are also strongly affected by ecological factors. The category of fungi with narrow host range furthermore included taxa previously not known as colonizers of Fagus : Cortinarius anomalus is a documented associate of shrubs of maqui in semi-arid climate ( Cistus sp.) and of early succession tree species such as Salix sp. (Watling 2005 ; Barden 2007 ); Xerocomus badius is a typical EM fungus of spruce (Gronbach 1988 ) and Inocybe asterospora of the orchid Cephalathera longifolia (Leake 2004 ). Colonization of Fagus roots with putative non-host EM fungal taxa has been reported previously (Pena et al. 2010 ). These non-host EM fungal associations occurred only at low frequency and were very labile when the photoassimilate supply was interrupted (Pena et al. 2010 ). It has been suggested that Fagus provides ecosystem services by maintaining non-host fungi which may constitute the insurance for future forest development and adaptation to changing environmental conditions (Pena et al. 2010 ). Concluding remarks It is well known that soil properties, climatic conditions, and physiological factors such as tree age affect EM fungal species richness in a stand (Conn and Dighton 2000 ; Wardle 2006 ). In accordance with other studies (Kernaghan et al. 2003 ; Ishida et al. 2007 ; Tedersoo et al. 2008 ), we found that the number of mycorrhizal species increased with increasing number of host tree species. Since the contributions of different tree taxa to mycorrhizal species numbers varied considerably, our study highlights that the increment in fungal species numbers depended on tree species identity or tree social status but was not simply a function of tree species numbers (Dickie 2007 ). Therefore, stand composition is important for below-ground mycorrhizal community species richness. We found clear host preferences. However, overall generalist fungi colonized the major fraction of root tips."
} | 3,573 |
38032216 | PMC10734540 | pmc | 8,457 | {
"abstract": "ABSTRACT While methane is typically produced under anoxic conditions, methane supersaturation in the presence of oxygen has been observed in both marine and fresh waters. The biological cleavage of methylphosphonate (MPn), which releases both phosphate and methane, is one pathway that may contribute to this paradox. Here, we explore the genomic and functional potential for oxic methane production (OMP) via MPn in Flathead Lake, a large oligotrophic freshwater lake in northwest Montana. Time series and depth profile measurements show that epilimnetic methane was persistently supersaturated despite high oxygen levels, suggesting a possible in situ oxic source. Metagenomic sequencing indicated that 10% of microorganisms in the lake, many of which are related to the Burkholderiales (Betaproteobacteria) and Actinomycetota, have the genomic capacity to cleave MPn. We experimentally demonstrated that these organisms produce methane stoichiometrically with MPn consumption across multiple years. However, methane was only produced at appreciable rates in the presence of MPn when a labile organic carbon source was added, suggesting that this process may be limited by both MPn and labile carbon supply. Members of the genera Acidovorax , Rhodoferax , and Allorhizobium , organisms which make up less than 1% of Flathead Lake communities, consistently responded to MPn addition. We demonstrate that the genomic and physiological potential for MPn use exists among diverse, resident members of Flathead Lake and could contribute to OMP in freshwater lakes when substrates are available. IMPORTANCE Methane is an important greenhouse gas that is typically produced under anoxic conditions. We show that methane is supersaturated in a large oligotrophic lake despite the presence of oxygen. Metagenomic sequencing indicates that diverse, widespread microorganisms may contribute to the oxic production of methane through the cleavage of methylphosphonate. We experimentally demonstrate that these organisms, especially members of the genus Acidovorax , can produce methane through this process. However, appreciable rates of methane production only occurred when both methylphosphonate and labile sources of carbon were added, indicating that this process may be limited to specific niches and may not be completely responsible for methane concentrations in Flathead Lake. This work adds to our understanding of methane dynamics by describing the organisms and the rates at which they can produce methane through an oxic pathway in a representative oligotrophic lake.",
"introduction": "INTRODUCTION Globally, freshwaters account for approximately 20% of methane (CH 4 ) emissions to the atmosphere ( 1 ). The conventional understanding of biological methanogenesis is that it occurs exclusively under anoxic conditions ( 2 , 3 ). However, in the upper ocean and the epilimnia of freshwater lakes where oxygen is present, concentrations of methane can be supersaturated with respect to atmospheric equilibrium ( 4 – 8 ). A growing number of studies using measurements of methane stable isotopic composition and physical transport modeling indicate that supply from anoxic habitats may not be responsible for these elevated concentrations ( 7 , 9 ). While many mechanisms for methane production in oxic waters have been suggested, including photosynthesis, the metabolism of methylated amines, and within cryptic anoxic niches ( 10 – 13 ), the degree to which these pathways contribute to methane supersaturation is unknown. Our understanding of the production of CH 4 and its contribution to atmospheric concentrations is important as methane flux from lakes is projected to increase in future climate change scenarios ( 14 , 15 ). Many lakes appear phosphorus limited ( 16 , 17 ), potentially inducing the use of alternative phosphorus sources other than phosphate by microorganisms. A portion of the dissolved organic matter pool in marine and freshwater environments resides as dissolved organic phosphorus (DOP), a poorly characterized reservoir of carbon and phosphorus ( 18 ). DOP can include phosphonates, compounds characterized by a stable C-P bond that is synthesized as part of lipid headgroups, exopolysaccharides, and glycoproteins ( 19 ). These compounds are produced by some of the most abundant lineages in the global ocean, including SAR11, Prochlorococcus , and Nitrosopumilus ( 20 – 23 ). While phosphonates and the transcription of genes involved in their synthesis have been detected in freshwater ecosystems ( 10 , 24 – 29 ), quantitative determinations of their types, abundances, and distributions are sparse. One type of phosphonate is methylphosphonate (MPn), the demethylation of which releases not only phosphate but also methane. Oxic methane production (OMP) via the degradation of MPn has helped explain the methane paradox in the upper ocean ( 30 – 34 ). Organisms in freshwater and meromictic lakes can also cleave MPn, including members of the Alphaproteobacteria, Gammaproteobacteria, and Betaproteobacteria ( 35 – 37 ). However, the contribution of this process to methane production in phosphorus-limited freshwater systems and the diversity of organisms catalyzing it have not been well documented. We explored the dynamics of methane and the microbial capacity to use MPn in Flathead Lake, Montana, one of the largest (surface area ~500 km 2 , maximum depth 116 m) natural freshwater lakes in the western United States. Its short hydrologic residence time (~2.2 years) ( 38 ) and environmentally protected, largely undeveloped montane watershed make Flathead Lake oligotrophic; soluble reactive phosphorus levels are typically below detection, and N:P stoichiometric ratios are elevated ( 39 ). Experimentally, plankton growth in the lake can be limited by nitrogen, phosphorus, or light ( 39 – 42 ). A large fraction of the available nitrogen and phosphorus may be in the form of dissolved organic matter ( 39 , 43 ). In this study, we sought to address four questions: (i) How do concentrations of methane vary over space and time in Flathead Lake? (ii) What is the taxonomic and temporal distribution of organisms that have the genomic capacity for methylphosphonate-mediated methane production? (iii) Can we show experimentally that these organisms can perform OMP? (iv) What factors may limit the magnitude of OMP in situ ?",
"discussion": "DISCUSSION We explored methane dynamics and the potential for MPn-mediated methane production in Flathead Lake. Despite high oxygen concentrations, methane was consistently supersaturated relative to the atmosphere. These findings indicate in situ methane production or a source which supplies methane to the lake. Methane concentrations ranged from 20 to 500 nM, relatively low compared to those reported in the oxygenated water column of other lakes in which OMP has been a focus [e.g., references ( 6 , 7 , 9 )]. Concentrations were highest in April when the lake was fully mixed, possibly due to fluvial CH 4 input. However, the oft-reported subsurface methane maximum also appeared in Flathead Lake as the summer progressed, suggesting an in situ source. Consistent with the oligotrophic nature of Flathead Lake, methane production rates in the epilimnion, based on unamended control samples, were low and averaged 0.69 ± 0.18 nmol CH 4 L −1 d −1 . These rates are significantly slower than in other lakes where OMP has been studied, including Lake Stechlin (26–236 nmol L −1 d −1 ) ( 9 ) and Lake Hallwil (110 nmol L −1 d −1 ) ( 67 ). Given high oxygen concentrations and the apparent absence of in situ anaerobic methanogenesis, we sought to determine if the cleavage of MPn could be responsible for methane production. Phosphonates can comprise up to 25% of the oceanic DOP pool ( 18 ) and have been detected in lakes, although quantitative estimates are lacking and indicate they may be low ( 25 , 26 , 28 ). We estimate that ~10% of Flathead Lake microorganisms have the potential for MPn cleavage via C-P lyase. These abundances are similar to those in phosphorus-limited oceanic sites ( 23 , 68 ), suggesting that MPn could be an important source of P in oligotrophic systems. Members of the Betaproteobacteria and Actinomycetota were the most well-represented lineages, consistent with findings in other lakes ( 10 , 35 , 37 ). Although the biosynthetic potential for phosphonate production appears widespread ( 29 ), genes for this process ( pepM , ppd , and mpnS ) in Flathead Lake were relatively rare. Similar findings have been reported in the ocean, where genes for the production of phosphonates are less abundant than those for its consumption ( 22 , 23 , 69 ). It is striking that while some of the most abundant lineages in marine systems appear capable of producing or consuming MPn, including SAR11 and Prochlorococcus ( 21 , 22 , 70 , 71 ), we did not identify MPn cycling genes within Flathead Lake from members of the freshwater SAR11 Fonsibacter or cyanobacterial Synechococcus -related Cyanobium . Given the presence of putative phosphonate transporters in genomes of freshwater cyanobacteria ( 72 ), including those in Flathead Lake, it is possible these organisms are capable of cycling MPn using currently undescribed pathways. Regardless, it is evident that diverse microorganisms can use MPn in this system. Following the observation of methane supersaturation and the identification of widespread genomic potential for MPn use, we experimentally demonstrated that the functional capacity for OMP exists in Flathead Lake. We found that substantial methane was only produced when both MPn and glucose were added (102.6 ± 19.4 nmol CH 4 L −1 d −1 across all experiments). Methane was not produced following amendment with only carbon and nitrogen or MPn and nitrogen, indicating that these communities may be limited in situ not only by MPn substrate availability but also by organic carbon. OMP was also not observed at high rates in amendments with phosphate, even in the presence of MPn and carbon. Phosphate repression of C-P lyase has been observed in both isolates and environmental samples ( 35 , 37 , 73 , 74 ), consistent with regulation by the Pho regulon ( 69 ). Our finding that methane production in Flathead Lake may be carbon limited is consistent with other studies which have shown that methane production is stimulated by other nutrients, including in the presence of particles and labile dissolved organic matter and through the addition of carbon, nitrogen, and even iron ( 31 , 37 , 75 ). Given that dissolved organic carbon in Flathead Lake is ~100 µM, similar to concentrations in the open ocean ( 76 ) where MPn is thought to contribute significantly to methane production, the type and lability of the carbon available may be important in how organisms respond to MPn. MPn cleavage may, therefore, occur at labile carbon hotspots, for example, associated with zooplankton detritus ( 77 ) or in the phycosphere, and may be a more prominent source of OMP in eutrophic lakes with excess C and N. Future work should consider the importance of the types of carbon in how and which organisms respond to MPn. Using metagenomic sequencing, we identified the dominant microorganisms responding to MPn as members of the genera Acidovorax and Rhodoferax . Closely related strains (>98% ANI) of Acidovorax responded across both replicates and years and showed similarity to genomes from other lakes. These findings highlight the consistent response of Acidovorax to MPn, nitrate, and glucose temporally and the potential widespread distribution of this organism. Notably, members of this genus can produce methane through other aerobic pathways ( 78 ), further solidifying their role in OMP. We found that the genes for MPn use are widely distributed in this genus and did not co-occur with those for MPn synthesis, consistent with metagenomic analyses of a wide diversity of organisms from the ocean ( 23 ). In our experiments, unique strains of Acidovorax and Rhodoferax appeared to respond differently to the amendments, indicating that Flathead Lake is home to a large diversity of closely related members of the Burkholderiales that appear to differ in their nutrient and organic matter preferences. The organisms that responded to MPn amendment are rare in situ , likely representing <1% of the Flathead Lake community. Future transcriptomic sequencing of Flathead Lake communities will determine if these microorganisms, which show robust phosphonate utilization, are active in situ . While we have shown that heterotrophic microorganisms are able to perform OMP in Flathead Lake, we conclude that net OMP is altogether very low in this system (mean 0.34 ± 0.07 nmol CH 4 L −1 d −1 in all treatments with no added MPn; range, −2.81 to 1.59 nmol CH 4 L −1 d −1 ). We are unable to rule out alternative sources of methane that could contribute to the observed supersaturation in Flathead Lake. One possibility is that methane may be produced by or associated with certain photosynthetic organisms ( 10 , 11 , 79 – 81 ). However, we did not observe strong differences in methane production in treatments in complete darkness or those with light (0.58 ± 0.20 and 0.83 ± 0.23 nmol CH 4 L −1 d −1 , respectively; t -test, P > 0.5) or in treatments where photosynthetic organisms responded. This would also not be consistent with high methane concentrations observed during April when the lake is still fully mixed. Alternatively, lateral transport from anoxic habitats (littoral, fluvial) could be a source of methane to Flathead Lake [e.g., references ( 82 – 85 )]. While we provide evidence that the capacity for MPn use exists under the right conditions, namely, when both MPn and labile organic carbon are available, future work will be needed to fully understand the sources and sinks of methane in Flathead Lake."
} | 3,473 |
27899631 | PMC5314764 | pmc | 8,458 | {
"abstract": "We present a conceptually new reversible nanosensor regulated by a DNA nanoswitch. This system is not only responsive to external stimuli (e.g. ATP) but also can be reversibly switched between ‘OFF’ and ‘ON’ states via toehold mediated strand displacement reactions. It functions like a molecular net woven by DNA to capture or release the target molecules. As a proof-of-principle experiment, ATP is here chosen as the model to demonstrate our new strategy, which holds great promise for applications such as switchable DNA nanomachines and nanocarriers for drug delivery.",
"introduction": "INTRODUCTION In the past decades, DNA nanosensors have attracted much attention for their wide applications in environmental monitoring ( 1 ) and biological diagnosis ( 2 ), including the specific and sensitive detections of metal ions ( 3 , 4 ), proteins ( 5 , 6 ), nucleic acids ( 7 ), small molecules ( 8 , 9 ), etc. Recently, several DNA nanodevices have been built upon DNA nanosensors such as DNA logic sensors ( 10 ), DNA sensing dendrimers ( 11 ) and DNA sensing nanomachines ( 12 ). It reveals an emerging interest in these fields of nucleic acids. Nevertheless, only a few of reported DNA nanosensors can be reset ( 13 , 14 ). There remains a challenge in the development of a reusable nanosensor. Hence, it is of great significance to design a reversible nanosensor which can be reset like a switch. In general, a DNA nanoswitch undergoes a reversible structural change upon addition/removal of external stimuli such as metal ions ( 1 , 15 ), light ( 16 , 17 ), pH value ( 18 , 19 ), electrons ( 20 , 21 ), etc. Incorporating the properties of DNA switches into sensors, herein we devise a conceptually new reversible DNA nanosensor where capturing and releasing adenosine triphosphate (ATP) molecules are switched by toehold mediated strand displacement reactions. This is in contrast to previous studies ( 8 , 22 , 23 ) that typically utilize ATP as a target to trigger DNA strand displacement reactions. In our design, a fluorescent ligand for an ATP-binding aptamer ( 24 ), Thioflavin T (ThT), is utilized as an indicator for the binding and release of ATP, as it serves as a competitive reporter and reflects ATP binding by a fluorescence decrease. Scheme 1 briefly depicts this strategy. A functionalized contractile three-way structure was designed by incorporating three 27-mer ATP aptamer sequences ( 25 ) (in blue, Scheme 1 , I) into single-stranded motifs in the center of the three-way junction. The resulting functional DNA nanostructure behaves as a specific ‘ATP-responsive three-way nanosensor’ (ARTWN) whose structural integrity was confirmed by native polyacrylamide gel electrophoresis (PAGE). Upon addition of ThT molecules, the embedded ATP aptamer sequences undergo an inter-loop folding instead of an intra-loop one, namely, each pair of ATP aptamers in two adjacent loops forms an active bimolecular structure with minor groove binding pockets where ATP binds (Scheme 1 , II). As a result, the folded ARTWN strongly resembles a Watson–Crick base-paired three-arm junction rather than the six-arm, as verified by PAGE and fluorescence quenching (Figure 1 ). Meanwhile, the fluorescent ligand ThT is replaced by ATP and removed from ARTWN, accompanied by a sharp decrease in the fluorescent intensity of ThT (Scheme 1 , III). By this means, ARTWN is able to sensitively probe ATP with high specificity. Scheme 1. Illustration of the working principle of the DNA nanoswitch-controlled reversible nanosensor. ( I ) The nanostructure of ARTWN is self-assembled from equivalent of three component strands A 1 , A 2 and A 3 . ( II ) The initial state of nanosensor. ThT molecules intercalate into the two adjacent embedded ATP aptamers intermolecularly, giving rise to a fluorescence increase of ThT. ( III ) The ‘OFF’ state of the switch cycle. The bound ThT molecules are displaced by ATP, accompanied by a sharp fluorescence decrease. ( IV ) The ‘ON’ state of the switch cycle. Upon introduction of the release strand, the central ATP aptamers are fully hybridized thereby releasing the ATP molecules. The free ThT molecules bound to G-quaduplexes serves as a fluorescence indicator for the switch cycle. Figure 1. Native PAGE electrophoretogram for analyzing the formation and structural transformations of ARTWN in the absence ( A ) and presence ( B ) of 300 μM ATP. Lanes 1–8: A 1 control, A 1 , (A 2 +A 3 ) control, A 2 +A 3 , (A 1 +A 2 +A 3 ) control, A 1 +A 2 +A 3 , three-way junction (3WJ) control, six-way junction (6WJ) control. ( C ) Fluorescence quenching experiment using the Cy3-BHQ2-labeled ARTWN. To drive ARTWN to operate repeatedly, two single-stranded DNAs were further used to release and capture ATP. The release strand consists of the complementary sequence of the ATP aptamer (Scheme 1 , marked in green) and a DNA G-quaduplex PW17 (in purple). The G-quadruplex not only serves as a toehold to facilitate the strand displacement reactions, but also complexes with ThT ( 26 ) to indicate the repeated operations. The capture strand is fully complementary to the release strand. When all bound ThT molecules are replaced by ATP, ARTWN is in the ‘OFF’ state (Scheme 1 , III) with low fluorescence intensity. Upon addition of the release strand, the central ATP aptamers are fully hybridized to form rigid duplexes and consequently release the ATP molecules (Scheme 1 , IV). Simultaneously, the free ThT molecules bind to G-quaduplexes, giving rise to a fluorescence increase to indicate the ‘ON’ state of the switch cycle. By introducing the capture strand to remove the release strand, the system can return to the ‘OFF’ state. As the ATP molecule is released from the DNA structure in the ‘ON’ state (Scheme 1 , IV), namely, it's free and unbound in the system, it can be easily separated from the DNA structure using centrifugal filter devices thereby removed from the working system. This will allow ARTWN to be reset to the initial state (Scheme 1 , II) of the ATP biosensor upon addition of the capture strand. Although a similar strategy using ATP and toehold strand displacement reactions has been reported previously ( 22 ), there are main differences between ours and the reported approach. The DNA structural integrity is here maintained throughout the whole conformational changing processes (illustrated in scheme 1 ), which guarantees a good reversibility of our designed DNA nanosensor. Furthermore, the reported approach ( 22 ) focused on the ‘hidden toehold’ on a metastable DNA bulge-loop structure, whereas our strategy highlights the ‘reversible nanosensor’ itself, which can be reset after the sensing processes and thus it holds great promise for applications such as switchable nanomachines, nanocarriers for drug delivery and other related DNA nanodevices. In a sense, the designed nanoswitch behaves like a molecular net woven by DNA to capture or release ATP molecules through toehold strand-displacements.",
"discussion": "DISCUSSION In summary, we proposed a new principle to construct DNA nanoswitch-controlled reversible nanosensors. The designed nanosensor can be switched between the ‘OFF’ and ‘ON’ states via toehold strand displacement reactions. In a sense, it functions like a molecular net woven by DNA to capture or release ATP molecules. Removing free ATP molecules from DNA nanostructure, the final state of the switchable system will be returned to the initial state of the ATP biosensor. Such a novel reversible nanosensor holds great promise for the applications to switchable nanomachines ( 18 , 30 ), nanocarriers for drug delivery ( 11 ) and other related DNA nanodevices ( 31 )."
} | 1,910 |
30275496 | PMC6167322 | pmc | 8,459 | {
"abstract": "While the phrase ‘foraging bumblebee’ brings to mind a bumbling bee flying flower to flower in a sunny meadow, foraging is a complicated series of behaviors such as: locating a floral patch; selecting a flower-type; learning handling skills for pollen and nectar extraction; determining when to move-on from a patch; learning within-patch paths (traplining); and learning efficient hive-to-patch routes (spatial navigation). Thus the term ‘forager’ encompasses multiple distinct behaviors that rely on different sensory modalities. Despite a robust literature on bumblebee foraging behavior, few studies are directly relevant to sensory-guided search; i.e. how workers locate novel patches. The first step in answering this question is to determine what sensory information is available to searching bumblebees. This manuscript presents a computational model that elucidates the relative frequency of visual and olfactory cues that are available to workers searching for floral resources under a range of ecologically relevant scenarios. Model results indicate that odor is the most common sensory cue encountered during search flights. When the likelihood of odor-plume contact is higher, odor-encounter is ubiquitous. While integrative (visual + olfactory) cues are common when foragers are searching for larger flowers (e.g. Echinacea ), they become rare when foragers are searching for small flowers (e.g. Penstemon ). Visual cues are only encountered in isolation when foragers are seeking large flowers with a low odor-plume contact probability. These results indicate that despite the multisensory nature of floral signals, different modalities may be encountered in isolation during search-behavior, as opposed to the reliably multimodal signals encountered during patch-exploitation or nectar/ pollen acquisition.",
"introduction": "Introduction Bumblebee populations are sensitive to decreases in foraging efficiency Bumblebees are critical pollinators in both agricultural and native ecosystems 1 – 3 . Unfortunately these keystone species have experienced alarming declines alongside the highly publicized drops in honeybee numbers 4 – 7 . Critical work exposing the negative effects of neonicitinoid pesticides on bumblebee fitness indicates that pesticide exposure lowers rates of reproduction due, at least in part, to a drop in foraging efficacy of both workers and the colony as a whole 8 , 9 . This provides a critical link showing that the modification of worker behavior scales up to impact colony level fitness – a result that is consistent with seminal work showing that a colony’s ability to produce reproductive individuals is directly correlated with their size 10 . Better foragers provide more resources to rear young at the hive, which can increase the size of a colony during a foraging season. Given the current environmental pressures on bumblebees, developing a deeper understanding of their foraging behavior is relevant to conservation efforts. How do foragers search for flowers? While the term “forager” can be defined as an animal locating and consuming food resources, it is a complicated series of behaviors. In bumblebees this includes: locating a floral patch 11 , 12 ; selecting a flower-type 13 , 14 ; learning handling skills for pollen and nectar extraction 15 ; determining when to move-on from a patch 16 , 17 ; learning within-patch paths (traplining) 12 , 18 ; and learning efficient hive-to-patch routes (spatial navigation) 11 , 12 , 19 . Thus the term ‘forager’ encompasses multiple distinct behaviors that rely on different sensory modalities 20 – 22 . A critical component of foraging theory is the search phase 23 ; which would be floral patch location in the case of pollinators. This phase is comprised of: (1) movement through the environment; and (2) recognition of resources, which should terminate the search. There is a wealth of literature analyzing forager search paths, from bumblebees to albatrosses 24 – 27 . While there is some controversy over the precise algorithms that accurately describe these search paths 28 – 33 , there is consensus that search paths can be reasonably represented with stochastic models of forward-biased motion (i.e. while turning events happen, complete direction reversal will be rare). Once a searching forager recognizes a resource, their behavior should transition from random-search to approach and feeding. In bumble bees the ability to recognize floral resources will be dependent upon perception of floral signals. Flowers provide complex sensory displays, including color, shape, nectar guides, odor and morphology 34 – 37 . In the case of pollinators searching for novel patches, only those sensory cues capable of operating at a distance will factor into recognition and subsequent sensory-guided navigation. Morphological cues are only relevant upon physical contact with flower and are thus not useful for search. Complex patterns on flowers, such as nectar guides or visible stamens, are only resolvable at close distances (4–45 cm) 38 . Thus shape, color and odor are the sensory signals most likely to be available for patch recognition. Odor pollution impacts forager behavior, but the effects on foraging efficiency are unclear Several studies over the past decade have indicated that anthropogenic odor pollution is both modifying floral odor plumes 39 and subsequent behavioral responses of bees 40 , 41 . While this work is interesting from a neuroethological standpoint, it is currently unclear how drastically natural foraging populations are impacted by odor pollution. Understanding the potential impact of odor-pollution first requires an understanding of odor’s role in foraging. There is a substantial body of work indicating that olfaction is important in patch exploitation; however, the precise role that odor plays is not completely understood. PER studies indicate that bumblebees are capable of associative odor learning 42 – 44 , generating the logical hypothesis that floral odor could be used to identify rewarding flowers. Multimodal studies investigating both vision and olfaction indicate that stimulation of odor pathways improves foraging accuracy, regardless of whether or not floral signals have differentiating odor stimuli 45 . Field experimentation on floral morphs showed that bumblebees prioritized visitation of a learned visual (color) signal over the learned odor 46 . These findings might imply that any odor is effective, and that precise odor identity might be irrelevant. However, work by Leonard et al . showed that when flowers differ in both visual and olfactory modalities, foraging accuracy was higher 35 – pushing back against the idea that odor identity is unimportant. Social odor cues – tarsal scent deposits on flowers, reduce bumblebee visitation rates. This is an example of a ‘contaminating odor’ that increases energy gain by reducing visitation to recently emptied flowers. It is likely that scent marks are perceptually distinct from the floral odor, rather than modifying the floral blend-structure such that it becomes unrecognizable to the bumblebee, as behavioral data have been relatively consistent across multiple flower species 47 and with unscented artificial flowers 48 . Therefore, it appears that the precise odor identification of tarsal scent-marks is quite important to foraging behavior. Given the contradictory nature of current data on odor usage, it is difficult to predict the effects of pollution on foraging efficiency during patch exploitation. There is a paucity of work looking at the impact of odor on navigation to food resources in bumblebees. Several lab-based studies indicate that odor alone is sufficient to facilitate navigation 40 , 49 . However, the relative roles of odor and vision (which could have implications for how drastic the effects of odor pollution might be) have never been investigated at a spatial scale that would shed light on the role of odor in patch location. For example, lab studies are typically in arenas that are less than 3.6 m in their largest dimension 13 , 35 , 40 , 48 , 50 – 59 . However, the foraging range of a bumblebee can reach up to 1.75 km from their nest 60 – a distance that is orders of magnitude larger than typical sensory-behavior studies, even those that are based in the field 46 . An understanding of odor-pollution’s impacts requires a better understanding of the relative roles that vision and olfaction play in navigation to floral resources. If a searching-forager is consistently encountering an odor signal before a visual signal, it stands to reason that odor-guided navigation will bring that animal within visual range of a flower. Given that odor plumes are theoretically available at a much greater distance from a flower 39 , 61 than visual cues 38 , 62 this is a logical assumption. However, odor plume contact is stochastic, and some empirical measurements of odor-plumes indicate much shorter distances travelled 63 . This manuscript presents a computational model that moves beyond assumptions and asks – given the probabilistic nature of odor plume contact- what is the likelihood of a bumblebee encountering resolvable visual versus olfactory cues?",
"discussion": "Results and Discussion This model explored the sensory signals available to bumblebee foragers searching for novel resources by calculating the relative probability of workers encountering the visual and/or olfactory signal from a floral resource while searching in a relatively low-resource environment. The parameters varied in this model were: plant density, number of blooms (and thus the strength of sensory signals from an individual plant), the probability of odor plume encounter, and the visual acuity of the searching “bumblebee”. Olfaction is the dominant sensory modality available to searching bees Looking holistically at all tested scenarios we see that odor dominated as the available sensory modality; with odor alone representing floral sensory encounter in 179/350 scenarios, an integrated odor-visual signal available in 136/350 scenarios, and vision alone as the dominate modality in only 35/350 (Figs 4 and 5 ). Odor information is therefore available for decision making in 90% of successful floral encounters, while visual information is only present in 49%. While there is substantial work indicating that vision is vitally important for patch exploitation behaviors 35 , 46 , 84 , it is likely that odor is crucial in patch location behavior. Figure 4 Heat maps indicating the relative likelihood of encountering a resolvable olfactory (blue), visual (yellow), or integrated olfactory and visual (green) sensory signal. These likelihoods were calculated for:1. multiple plant sizes, indicated by a variable number of flowers on the x axes; 2. multiple plant densities, indicated on the y axes; 3. two different plant species, Echinacea (top diagram) and Penstemon (bottom diagram); 4. two different odor probabilities, with high encounter probability represented in the left row and low on the right; and 5. three different visual acuities, labelled with their angular resolution on the right hand side of the figure. Each model scenario was run 1000 times. The number of failures – runs where a bee searched for 1.5 km without encountering a sensory signal- are indicated on the plots themselves. The absence of a number means that all 1000 runs resulted in a successful sensory encounter. Figure 5 Results of model runs using a Levy-walk distribution of step lengths for bumblebees with a visual resolution of 3.5 ° searching for Echinacea . Despite the difference in search-path calculation methods, the results are nearly identical to those depicted in Figs 4 and 6 . Levy-walk searches do lead to a slight reduction in failure rates for low-odor probability scenarios. Odor landscapes are changing, which could have a considerable impact on bumblebee foraging behavior Model runs with a higher probability of odor contact demonstrated a larger discovery distance, with bumblebees contacting a resolvable sensory signal in the range of 25–40 meters, as opposed to 2–20 meters (Fig. 6 ). Additionally, decreased probability of odor-contact drastically increased the likelihood that forager searches would end in failure (Figs 4 and 5 ). Failure rates overall were higher for the smaller bloom size ( Penstemon ), as the larger plant and flower size of Echinacea afforded a better ability to transition to visual navigation when odor was unavailable. These computational results are commensurate with laboratory investigations on visual search time in bumblebees, where bees who have been restricted to solely visual information have higher search times to locate smaller flowers 85 . The low-odor probability tested in this model decays rapidly, transitioning to zero before 10 meters from the point source (Fig. 3 ) 63 . This empirical measurement may be underestimating plume strength due to environmental conditions: previous work has shown that odor plumes can rise in altitude 86 and the Riffell et al . measurements were taken at a consistent elevation from the ground. However; the results from this odor fit are relevant to consider in light of work examining the impacts of anthropogenic pollution on floral odor-plumes. Seminal work by McFrederick et al . indicates that environmental pollutants can interact with floral odorants, reducing their distance travelled by an order of magnitude drop: odorants that historically could travel 1000 m before dropping to 80% of their original concentration would only make it 100 m in worst case scenarios. While McFrederick’s study was computational, subsequent experimental studies have been equally concerning. Girling et al . found that diesel exhaust degrades select floral odorants, modifying odor blend structure 87 . Likewise Farre-Armengol et al . found that ozone decreases floral odorant concentrations 41 . Based on our model results it is reasonable to hypothesize that bumblebees will experience higher failure rates in locating flowers when searching in polluted environments, particularly if available floral resources are comprised of plants with smaller bloom size and lower bloom number. Indeed, failure to locate a floral signal only occurred in the low-probability odor scenario - when odor information is readily available searching events are universally successful. Figure 6 Heat maps indicating average distance at which a resolvable sensory signal was encountered in successful model runs. These likelihoods were calculated for: 1. multiple plant sizes, indicated by a variable number of flowers on the x axes; 2. multiple plant densities, indicated on the y axes; 3. two different plant species, Echinacea (two left columns) and Penstemon (two right columns); 4. two different odor probabilities (labelled by column); and 5. three different visual acuities, labelled with their angular resolution on the right hand side of the figure. Bloom size, number of blooms per plant and plant density impact both available sensory modality and distance at which plants are found Unsurprisingly, plant size and density impact the likelihood that bumblebees will encounter a resolvable sensory signal (Figs 4 and 5 ). Increased plant density reduced failure rate in low odor probability situations for both large ( Echinacea ) and small ( Penstemon ) flowers. However, when plants with small bloom sizes are in low density patches they were only reliably ‘found’ in model runs with a higher probability of odor plume contact. Echinacea simulations were moderately less susceptible to density effects as they can be seen from a greater distance, but higher flower number was still associated with an increased discovery distance. Interestingly, field data on Penstemon indicated that they were typically found at the higher densities this model tested – the lower densities tested here were included purely for comparative purposes. In light of anthropogenic modulation of odor environments, bumblebees may passively select for larger bloom size and higher plant density in polluted environments by virtue of not being able to locate smaller flowers, or those with larger nearest-neighbor distances. Effects of search-path type The outcome of model runs using a power-law distribution for step lengths (Levy-walk) (Fig. 5 ) is nearly identical to the results from constant step lengths (Fig. 4 ). Odor information is ubiquitous in the high-odor probability scenarios, with visual information not being encountered in isolation until the low-odor probability scenarios. As in Figs 4 and 6 , a shift to low-odor probability both decreases the distance at which flower-signals are encountered and increases failure rates in search flights. The predominant difference between the two search-path methods is a slight decrease in failure rates when using variable step lengths, a finding that is consistent with the fact that the latter method ran for approximately double the distance, creating a longer search path. Limitations and Future Directions It is worth emphasizing that this experiment was done in silica. As such it is limited by the assumptions used to generate model results. These model results are strongly driven by visual and olfactory resolution: on the plant side from the strength of floral signal, and on the pollinator side from sensory sensitivity. While all of these variables were parameterized based on the ecology and physiology of the relevant plant-pollinator relationships, the absolute values returned by the model are less relevant than the trends. These trends raise interesting questions for future experimental work. For example, the indication that bumblebees with lower visual acuity first encounter smaller floral displays via odor plumes begets the question, will bumblebees searching for novel resources navigate with odor information alone? This phenomenon has previously been demonstrated on a small spatial scale 40 , 49 , but remains to be shown at field-realistic scales. The substantial number of model runs finishing with simultaneous odor and visual signal encounter raises the question, does odor information make a minimally resolvable visual cue more salient? Again, work on small spatial scales shows improved learning and recognition of food resources with multimodal sensory information 51 , 88 , but how this operates on large spatial scales is less clear. Finally, this computational model provides an alarming context for recent work on odor pollution in bee behavior. While that work has largely focused on laboratory investigations, decreasing plume distance is likely to have profound impacts on foraging efficiency in bumblebees and other odor-guided pollinators. These results, in combination with recent computational findings on air pollution decreasing distance travelled by floral scent 67 , strongly indicate that relevant field work to ground-truth theoretical concerns is necessary."
} | 4,746 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.